Judgment Approved by the court for handing down Deutsche Bank v Sebastian Holdings
(subject to editorial corrections)
Neutral Citation Number: [2013] EWHC 3463 (Comm)
Case No: 2009 Folio 83
IN THE HIGH COURT OF JUSTICE
QUEEN'S BENCH DIVISION
COMMERCIAL COURT
Royal Courts of Justice
Strand, London, WC2A 2LL Date: 8th November 2013
Before:
MR JUSTICE COOKE
- - - - - - - - - - - - - - - - - - - - -
Between:
Deutsche Bank AG Claimant
- and - Sebastian Holdings Inc. Defendant
- - - - - - - - - - - - - - - - - - - -
- - - - - - - - - - - - - - - - - - - -
David Foxton QC, Sonia Tolaney QC, Henry King, James MacDonald (instructed by
Freshfields Bruckhaus Deringer) for the Claimant
David Railton QC, Simon Birt, Thomas Plewman SC, Oliver Jones and Max Schaefer
(instructed by Travers Smith) for the Defendant
Hearing dates: 22nd, 23rd, 25th, 26th, 29th, 30th April,
1st, 2nd, 3rd, 7th, 8th, 9th, 10th, 14th, 15th, 16th, 20th, 21st, 22nd, 23rd May,
4th, 5th, 6th, 10th, 11th, 12th, 13th, 14th, 17th, 18th, 19th, 20th, 24th, 25th, 26th, 27th June,
1st, 2nd, 3rd, 4th, 8th, 9th, 15th, 16th July,
2nd August 2013
- - - - - - - - - - - - - - - - - - - -
Judgment Approved by the court for handing down
(subject to editorial corrections)
.............................
CONTENTS Section Heading
Introduction
The Key Issues
The Bank Audit Report
The New York Action
The Law of the Case
The Key Witnesses
The Contractual Documents
7(a) The Principles of New York Law applicable to the
construction of contracts and the implication of terms therein
7(b) The Said Letter of Authority
7(c) The Nature of FX Prime Brokerage and the Expert
Evidence thereon
7(d) The Foreign Exchange Prime Brokerage Agreement
(the “FXPBA”)
7(e) The FX ISDA, the Schedule and the CSA
7(f) The Pledge Agreement
7(g The Limited Power of Attorney
Implied Terms
8(a) Paragraph 38(1)
8(b) Paragraph 38(2)
8(c) Paragraph 38(3)
8(d) Paragraph 38(3A)
8(e) Paragraphs 38(4) and (4A)
8(f) Paragraph 38(4B)
8(g) Paragraph 38(4C) 8(h) Paragraph 38(4D)
8(i) Paragraph 38(5)
8(j) A further Implied Term of the FX ISDA
The Principles of the New York Law of Tort
9(a) Concurrent duties of care and the Economic Loss Rule
9(b) Negligent Misrepresentation
9(c) Damages
The Alleged Oral Agreements
10(a) The Capital Limitation Agreement
10(b) The Pledged Account Limit (PAL)
10(c) The Oral Agreements as to the types of trade
10(d) Convention, Acquiescence and Rectification
10(e) The Collateral Warning Agreement
The Meaning of Currency Options and Structured Options,
the Said Letter of Authority and the FXPBA
11(a) The Expert Evidence
11(b) Clause 2(iii) of the FXPBA
The VaR Parameters
12(a) The Changed Parameters
12(b) The Computer Models
The problems created by the OCTs and the EDTs for DBAG’s systems
Mr Said’s Evidence on Affidavits, Depositions and in his Timeline
Mr Said’s Agreement to the Non Reporting of the EDTs,
MTMs and Margin calculations which included them
15(a) The 5th May 2008 telephone call between Messrs
Quezada, Walsh and Said
15(b) The 22nd July telephone conversation between Messrs
Walsh and Said
15(c) The 8th September meeting between Messrs Quezada,
Spokoyny and Said
The History of Mr Said’s Trading and Mr Vik’s knowledge
thereof
The 2008 Agreements
17(a) The Equities PBA
17(a)(i) The First Issue of Construction
17(a)(ii) The Second Issue of Construction
17(a)(iii) The Third Issue of Construction
17(a)(iv) The Fourth Issue of Construction
17(b) The Listed F&O Agreement
17(c) The Master Netting Agreement
Ratification
Mr Vik’s FX Trading with DBAG and its collateralisation
19(a) The Course of Events in 2007-2008 relating to Mr
Vik’s FX trading
19(b) The Pattern of Mr Vik’s FX trading
19(c) The 3rd September email
19(d) Agreement, estoppel by convention, acquiescence and waiver
Mr Vik’s F&O transactions and their collateralisation
The DBS Counterparty Issue
The Alleged Misrepresentations
22(a) The first implied representation at the meeting of 7th May 2008
22(b) The second alleged misrepresentation arising from emails relating to the withdrawal of cash from the FX account
22(c) The third alleged misrepresentation on 6th October 2008
22(d) The fourth alleged misrepresentation at the 7th October
2008 meeting
The GEM Terms and Conditions of Use
Inducement of Breach of Contract
The FX Margin Calls
25(a) The Ninth Argument
25(b) The First Argument
25(c) The Second Argument
25(d) The Third Argument
25(e) The Fourth Argument
25(f) The Alleged Events of Default or Potential Events of
Default
25(g) The Fifth Argument
25(h) The Sixth Argument
25(i) The Seventh Argument
25(j) The Eighth Argument
The Equities Margin Call
Termination of the Contracts
Wrongful Transfers from SHI’s accounts
FX Close Out
Equities Close Out
30(a) The American Shipping Shares
30(b) The Floatel Shares
30(c) The Scorpion Shares
30(d) The Seajacks Shares
30(e) The Standard Drilling Shares
30(f) The Thule Shares
30(g) The Yantai Shares
The Covenant of Good Faith and Fair Dealing
DBAG Claims
SHI’s Damages Counterclaim
33(a) SHI’s Available Funds
33(b) Mr Vik’s trading in September and October 2008 and
the losses claimed in respect of the forced close out
33(c) The hiatus and the starting fund for the Hypothetical
Portfolio
33(d) The Hypothetical Portfolio
33(e) Bars to Recovery
DBAG’s alleged duty to account
Disclosure
The nature of DBAG and SHI’s trading
Conclusions
Annexes
Extracts from ISDA Master Agreement and Schedule
FXPB Organisational Chart
The margin figures for Mr Said’s FX Trading as calculated by the Forensic Accountants
DBAG’s ARCS Monte Carlo VaR Methodology
Mr Justice Cooke:
1. Introduction
The claimant bank (“DBAG”) is incorporated in Germany with branches around the world including London and New York. Its wholly owned subsidiary Deutsche Bank Suisse SA (“DBS”) is based in Geneva. The defendant (“SHI”) is a special purpose vehicle, incorporated in the Turks and Caicos Islands, which has at all material times been owned and controlled by Mr Alexander Vik, who is its sole director and a man
of considerable means, (a multi-billionaire) with recognised business acumen and money-making skills.
SHI was from 2003 onwards a private wealth client of DBS, dealing with its Private Wealth Management section (PWM), on an “execution only” basis. Mr Vik carried out different types of investment and trades through DBS, including investments in shipping, drilling and related equities (where he utilised the services of Mr Harald Hanssen), a special situation hedge fund and currency transactions (FX). He also engaged Mr Bokias as his full-time investment manager who provided him with financial analysis and market views by reference to it, with regular updates, graphs and spreadsheets. His major contact at DBS in 2005 was Mr Meidal, who left DBS in about September 2007. Mr Brügelmann, who had worked with Mr Meidal then became Mr Vik’s point of contact and his “go to” man for the effecting of trades through DBS and later in 2008 through DBAG in respect of his (Mr Vik’s) own trading.
In May 2006 SHI and DBAG entered into an ISDA Master Agreement (the Equities ISDA) essentially for the purpose of allowing SHI to carry out a CDS transaction with DBAG, which DBS was not equipped to facilitate.
In November 2006 SHI and DBAG concluded a Prime Brokerage Agreement enabling SHI to act as DBAG’s agent in executing FX transactions and Precious Metals transactions, including in particular currency and Precious Metals options. The main purpose of this agreement (the “FXPBA”), was to allow Mr Klaus Said, an individual with experience in FX, to trade in FX on SHI’s behalf, using the medium of the Prime Brokerage arrangements furnished by DBAG. Another ISDA Master Agreement was concluded between SHI and DBAG and signed by Mr Vik on 28th November 2006 (the “FX ISDA”), with a Schedule of specific terms and a Credit Support Annex (“CSA”). Both that Schedule and an Amendment Agreement to the 2006 Equities ISDA, also signed on 28th November 2006 made it plain that “unless otherwise agreed between the parties, any transactions other than Foreign Exchange Transactions and Currency Options Transactions shall be governed by the ISDA Master Agreement dated 8th May 2006” and that the November 2006 ISDA Master Agreement should govern only Foreign Exchange Transactions and Currency Option Transactions. As part and parcel of the arrangements for trading in FX and currency options, a Pledge Agreement was concluded dated 28th November 2006 by which SHI, DBAG and DBS agreed that all assets deposited or relating to an account of SHI with DBS would be held as collateral for all claims that DBAG might have against SHI.
On the same date (or, at least, as it appears from the document itself) Mr Vik for SHI and Mr Said signed a letter of authority (the Said Letter of Authority), addressed to DBAG authorising Mr Said to trade FX transactions and currency options on behalf of SHI. On 7th December DBS undertook, without SHI being privy to it, to hold assets in the pledged account in accordance with the terms of the Pledge Agreement and to monitor the lending value according to its own margining requirements and to advise DBAG if that value should fall below US$35 million (the TPMCA). The NOK equivalent of the sum of US$70m (approximately) was paid into the account.
From this point on, Mr Said did carry out FX trading through the Prime Brokerage arrangements set up by DBAG in New York, although the account was technically a London account. It was generally referred to as Mr Said’s account or the New York account. Mr Said was not technically employed by SHI, but by another creature company of Mr Vik’s which provided him with health benefits and a salary which were to be set off against financial compensation paid to him by SHI in the shape of 10% of the profits he made for SHI on his trading. Mr Said worked (when not working from home or from his holiday house during the month of August) in the office annexed to Mr Vik’s wife’s house in Greenwich Connecticut, where others engaged by SHI also worked and where Mr Vik himself also had an office, on a different floor, from which he worked when not in his main Monaco office or travelling around the world. Mr Vik was in the USA for 60 days a year or less, because of his residency in Monaco. He had three personal assistants.
Mr Vik continued to trade, on behalf of SHI, through DBS in equities and other investments, including FX transactions of his own which were concluded as over the counter (OTC) trades with DBS. To effect these different investments, the normal pattern was for him to give instructions to Mr Meidal or Mr Brügelmann who would then execute the deals for SHI. Wishing to indulge in more sophisticated forms of transaction, and in particular the shorting of equities, Mr Vik caused SHI to conclude a further Prime Brokerage Agreement with DBAG on 30th January 2008, but in this case operating through its London branch (the Equities PBA). This prime brokerage (often referred to as the “London Account”) related to “transactions in respect of Securities” as defined in the agreement itself, including any instrument as agreed between the parties from time to time. Under the Equities PBA, DBAG was to provide financing and settlement services to SHI against cash or securities to enable SHI to enter into transactions, primarily in stocks or bonds, but also futures and options in respect of either. Under this arrangement, SHI could enter into purchase and sale transactions of Securities with third parties, nominating DBAG as its agent for settlement purposes, whilst DBAG could, at its discretion, provide cash or securities financing, against collateral in the Securities Account held by it, with a security interest over those assets.
On the same date SHI and DBAG also concluded a Listed Futures and Options Agreement, “the F&O Agreement” under which DBAG would enter into transactions as principal (with a back-to-back transaction with SHI) when entering into exchange traded listed Futures and Options on the orders or instructions of SHI.
Additionally, on the same date a Master Netting Agreement was made between DBAG and SHI in respect of the Equities ISDA Master Agreement of 8th May 2006, the Equities PBA and the F&O Agreements. These agreements are together referred to as “the 2008 Agreements”.
DBAG treated Mr Said’s FX transactions alone as governed by the FXPBA, which was administered out of New York and New Jersey, although the account was technically a London account. All transactions concluded by Mr Vik were, prior to 30th January 2008, effected through DBS and, after January 2008, his FX transactions, his transactions in equities and in options and listed futures and options were all treated as governed by the suite of agreements reached on 30th January 2008 and handled in London on what was referred to as the GPF Platform, which provided for margining of these different types of transaction together under what was known as the DBX system. As appears hereafter there is an issue between the parties as to whether Mr Vik’s FX transactions were governed by the FX ISDA to which the FXPBA (which indubitably governed Mr Said’s FX transactions) referred and any collateral agreements relating thereto, or by the 2008 Agreements. Mr Brügelmann was authorised to execute trades on the GPF account on Mr Vik’s instructions and had access to DBX.
In the global financial storm in the autumn of 2008, the FX trades conducted by both Mr Said and Mr Vik became severely loss making. DBAG issued margin calls totalling over US$500 million in respect of Mr Said’s trading. These were mostly paid by SHI by transfers of cash from the GPF account from available assets and proceeds of sales, in order to meet the premium needed to close out Mr Said’s FX trades and the collateral required whilst doing so, whilst SHI also closed out substantial loss making positions on Mr Vik’s own FX trades and profitable positions on his Futures trades, until the money available to SHI in its DBAG and DBS accounts ran out. DBAG claims sums as due to it, amounting to US$118,474,958, under the FX ISDA and the FX Close Out Agreement allegedly reached with SHI and US$125,523,086 under the 2008 Agreements, as sums due on the close out of transactions.
DBAG now admits a number of failings on its part in the course of handling SHI’s trades. First, in circumstances which I shall describe more fully later, it accepted, when it had a discretion whether to accept or reject them, particular types of trades under the FXPBA (EDTs and OCTs), which, by reason of their terms, could not be booked, valued or margined on its GEM and ARCS VaR computer systems at the time. When the market moved against SHI in the course of 2008, and in particular in the autumn of that year, on its case DBAG became entitled to call for margin in substantial sums. It did not do so until it became apparent to it, as a result of another bank, Morgan Stanley (MS), seeking collateral in respect of Mr Said’s trades with it (using DBAG’s name and credit under the FXPBA) and from 2 direct trades that Mr Said had concluded with DBAG, that there was a deficiency in margin put up by SHI, which then led to the margin calls under the FXPBA to which I have already referred. These were made between October 13th and October 17th 2008 and were paid between 14 and 22 October 2008, without protest.
Furthermore, in the course of reporting the situation to SHI and making these margin calls, DBAG discovered two other major computational errors on the GPF account under the DBX system (the Russell Multiplier Error and the Ignored Payments Error) which reflected other deficiencies in that computer set-up. Although the latter would have impacted on earlier trading in amounts not calculated by the experts, on discovery on 16 October and 22 October 2008 respectively the errors were corrected. The effect of the errors was that assets in the Cash and Securities Accounts on the GPF platform had been overstated by US$115m and US$315m respectively. It is in these circumstances that a deficit emerged on the GPF account which led to a disputed margin call on that account on 22 October 2008. DBAG was less than straightforward about its margining and computational difficulties.
It is agreed between the expert forensic accountants that none of the FXPBA margin calls were overstated on the true state of the FX account with margining effected in accordance with the contractual position on new computer models built by the parties’ experts. On DBAG’s case, on 22nd October 2008 it also issued a margin call in respect of the GPF account but SHI denies ever receiving this. In circumstances which are the subject of dispute, the balance of Mr Said’s FX transactions were closed
out and assets subject to the Pledge Agreement and other SHI funds held by DBAG were applied to recoup the losses incurred, so far as possible, leaving the amounts now claimed by DBAG as the alleged deficit under all the agreements.
DBAG also informed DBS that it was enforcing the Pledge Agreement on 23rd October and served notice of termination of the FXPBA the same day. By close of business on 24th October all the FX positions had been closed out.
On 30th October there was a meeting in London between Mr Vik and high ranking DBAG personnel where DBAG sought payment of sums claimed to be due and Mr Vik raised complaints about DBAG’s actions in allowing SHI to incur such large losses when he had allocated only US$35m to the Pledged Account as capital for Mr Said’s trading. He sought but did not obtain a report on the daily MTM and margin requirements for the “Structured Options” traded by Mr Said under the FXPBA from August 1st 2008 – October 13th 2008 (as to which, see below). DBAG did not admit the deficiencies referred to in paragraph 12 above.
On 4th December DBAG sent a letter to SHI terminating the Equities PBA on the ground of failure to pay the GPF margin call of 22 October, stating that 4th December 2008 was the termination date applicable. It also demanded immediate payment of the deficit on the FXPBA, said to be US$120,650,166. On 20th January 2009 DBAG sought payment from SHI in respect of the Equities PBA, the Equities ISDA Agreement and the Master Netting Agreement, in the sum of US$125,523,086.
SHI alleges that a series of trades, referred to in this action as “Exotic Derivative Transactions” (EDTs) and referred to by other banks and Mr Said by a variety of names, “TPFs” (“Target Profit Forwards”), “Target Forward Structures”, “TARNs”, “Pivot Range accruals” and “Range Bets”, were all concluded by Mr Said outside the scope of the authority given to him under the Said Letter of Authority or FXPBA and are not binding on SHI. These 41 EDTs were a major contributor to the huge margin figures to which DBAG claimed to be entitled under the terms of the FXPBA, the FX ISDA and its CSA, as calculated by the expert forensic accountants, and to the losses incurred on close out. Additionally, although not responsible for any losses (in fact giving rise to a small profit), it is said by SHI that there were 53 “Other Complex Transactions” (“OCTs”) which also fell outside Mr Said’s authority. SHI maintains that none of such transactions were “currency options” within the meaning of the Said Letter of Authority or the FXPBA. This gives rise to issues of construction, as affected by relevant market usage, of the terms “currency options” and “Structured Options” as those terms appear in those contractual documents.
Furthermore SHI alleged two additional oral agreements, concluded at or about the time of the FXPBA which limited Mr Said’s authority to trade on behalf of SHI:
The first is said to have limited Mr Said to trading “vanilla options” only, which are said to be straightforward options involving nothing more than a put or call or a series of put or call options.
The second is said to have limited Mr Said to concluding transactions which did not give rise to losses in excess of US$35m. This is also expressed as an agreed trading limit of US$35m or an agreement restricting DBAG’s recourse to that sum (the Capital Limitation Agreement).
Such agreements are said to have been concluded between Mr Vik and Mr Said and between SHI (in the person of Mr Vik) and DBAG, in the persons of Mr Meidal and/or Mr Brügelmann (as pleaded) or Mr Meidal alone (according to Mr Vik’s evidence).
A yet further oral agreement is alleged as between Mr Vik and Mr Meidal/Mr Brügelmann to the effect that all SHI’s FX trading would be subject to a trading limit constituted by the amount standing to the credit of the Pledged Account plus SHI’s FX profits, which were or should have been paid into that account, insofar as not removed from the account on Mr Vik’s instructions. This became known as the Pledged Account Limit or PAL.
SHI denies liability for the sums claimed by DBAG and pursues a counterclaim for damages in excess of US$8 billion based on these agreements. Although these agreements are put forward as oral agreements, it was alleged that Mr Vik understood them to be encapsulated in the written agreements which he signed, because he was assured that this was the effect of their terms. Additionally, by an amendment made in 2012 SHI also maintains that it was agreed that DBAG would give a warning to SHI whenever the margin approached US$35 million (the Collateral Warning Agreement).
SHI further alleges that, in accordance with the FX ISDA Schedule and the Amendment Agreement, Mr Vik’s own FX transactions for SHI (although concluded directly with DBAG, rather than with third parties through DBAG, using its name) after January 2008, were governed by the FX ISDA and “the Pledged Account Limit” (“the PAL”) and were not governed by the Equities PBA, under which DBAG had in fact been operating his FX trading in 2008 and, according to its own understanding of margining procedures, margining those transactions. It is said that DBAG wrongly margined Mr Vik’s own FX transactions with his and SHI’s other transactions on the GPF platform (supposedly under the Equities PBA), which led him to close out various Equities Futures and FX transactions which he would not otherwise have done, on receiving notification on September 3 2008 from Mr Brügelmann that he was near to SHI’s GPF margin limits, as calculated by DBAG. If DBAG had margined Mr Vik’s FX transactions separately from the rest of his transactions on the GPF platform, there would have been no such warning as it was those FX transactions which were the subject of large market movements, with the consequent effect on the calculation of variation margin on the Equities PB Account. SHI maintains that, in the absence of any warning on 3rd September, its Equity Futures and FX trades would otherwise have been held until profitable once again, and it thus incurred losses from the wrong margining and the wrong warning. These failures were exacerbated by one of the two huge computational errors made by DBAG in respect of the GPF account.
Moreover, if Mr Vik’s transactions had been margined under the PAL, that limit would have been reached much earlier than September 3, with consequent effect on Mr Vik’s and potentially Mr Said’s trading.
SHI contends that the GPF margin warning was wrongly made and that DBAG failed to give a warning on the FXPBA margin situation. It maintains that the FXPBA margin calls were wrongly made, because of the breaches already outlined, so that SHI’s other positions were closed out when they should not have been. There would have been no Equities Margin call either. Had none of this happened, SHI would
have continued with its trading strategies and, over the course of the years from 2008 to date, made profits of over US$7 billion in accordance with a Notional Portfolio which SHI has maintained during that period, showing the trading decisions it would have made.
In response, DBAG contends that the parties agreed by word or conduct or conducted themselves on the basis of a common assumption (regardless of the words of the FX ISDA Schedule or Amending Agreement) that Mr Vik’s FX trading for SHI was to be effected under the terms of the Equities PBA and the 2008 Agreements and margined accordingly. Alternatively DBAG relies on some form of waiver or estoppel in this respect.
SHI also contends for implied terms in the FXPBA to the effect that DBAG was obliged properly and accurately to record the details of the transactions concluded by Mr Said, properly to record the cash flows, the profit and loss, the mark to market value (MTM) and the collateral required (margin) for Mr Said’s trading and to make accurate reports to Mr Said and/or Mr Vik of these matters.
It has been accepted by DBAG since February 2012, but not before, that it did not accurately record the details of the EDTs and OCTs and that it did not accurately value them on a mark to market basis nor require collateral in respect of them at any stage prior to the margin calls in October 2008. The computer system operated by DBAG and its Prime Brokerage (PB) desk was not capable of recording transactions of such complexity nor of producing valuations on a MTM basis nor of calculating margin requirements by reference to exposure (MTM) or what is known as VaR (value at risk), which is intended to reflect the loss involved in the liquidation of the assets in question. SHI says that, in consequence of this inability to record and report on the exposure and margin generated by Mr Said’s trading, whether or not there was an agreed limit on Mr Said’s authority to US$35 million worth of exposure or an agreement to limit SHI’s liability to that figure, the effect of DBAG’s failures was that SHI was ignorant of the true figures (in the persons of both Mr Said and Mr Vik) and was unable to appreciate the risks being run on the EDTs. Had there been any appreciation that the exposure and margin requirement exceeded US$35 million, and in particular to the extent that it did, steps would have been taken to close out those transactions and reduce the exposure to manageable levels. Although the forensic accountancy experts instructed by the parties differ on matters of detail, both agree that by April 2008 the collateral requirement of Mr Said’s trading, if properly calculated in accordance with DBAG’s maximum entitlement would have been of the order of US$90 million and, after dropping below the US$35 million limit in May and June 2008, from July onwards fluctuated between US$50-100 million until September and October when it leapt, with some fits and starts, to US$400 million and, at its peak at the time of the margin calls, to US$800-$900 million.
Questions of fact arise as to what would have happened if DBAG had valued the TPFs and OCTs and provided reasonably accurate MTM figures to Mr Said and sought an increase in collateral to match the exposure. If additional collateral had been required, Mr Vik would have had to be asked to procure its provision.
2. The Key Issues
First, as should be plain from the above summary, there are key issues relating to oral agreements reached between Mr Vik on the one hand and, as pleaded, with Mr Meidal and/or Mr Brügelmann on the other. Those agreements are said to be linked with agreements between Mr Vik and Mr Said to the same effect.
Mr Meidal acted as DBS’ relationship manager for the SHI accounts until July 2007, whereupon Mr Brügelmann who had been working almost exclusively on the SHI accounts since September 2006 as an investment manager, took on Mr Meidal’s role. On Mr Vik’s evidence, Mr Meidal, since leaving DBS in November 2007, has been angling for a business relationship with Mr Vik, both during his time at Goldman Sachs and now whilst at Lombard Odier. Mr Vik said that Mr Meidal, who had moved to London, agreed with his complaint against DBAG in the course of a number of meetings he had had with him since 2007. He had asked him to give evidence but he had declined to do so without the approval of DBS, which had not been forthcoming. Mr Vik considered that his evidence would be helpful but had not subpoenaed him.
In his witness statements, Mr Vik said that the oral agreements he reached in relation to limits on Mr Said’s trading were made with Mr Meidal in the course of telephone conversations and, he thought, two meetings at his wife’s home in Greenwich Connecticut, with Mr Said present for part of those meetings. No longer was it said that Mr Brügelmann was a party to these agreements.
Mr Meidal was, of course, an employee of DBS, not DBAG but Mr Vik said that he did not draw any such distinction, particularly given the “one bank” philosophy advanced by DBAG and its subsidiaries in advertising. Mr Meidal was not called by either party to give evidence nor was he deposed in the New York proceedings which SHI began on 24 November 2008 in which it served its complaint on 20th January 2009.
Mr Said was also not called as a witness by either party. He had formerly been engaged by SHI to trade on its behalf. He was however extensively deposed in the United States in the context of the New York litigation, the subject matter of which overlaps with this action and to which I shall refer later in this judgment. Both parties relied on various parts of his deposition under the Civil Evidence Act whilst denying the truthfulness of other parts. In circumstances where Mr Vik alleges that he made agreements with Mr Meidal and Mr Said in the context of the latter’s FX trading, and where Mr Brügelmann is now no longer said to be a party to any such arrangement, the key issue of the oral agreements turns essentially upon the credibility of Mr Vik’s evidence in the light of the written agreements and their proper construction, the surrounding documents and the commercial probabilities, as well as Mr Said’s deposition evidence.
Secondly, the proper construction of the various agreements to which I have referred is a key issue, particularly in the context of the disputes about the oral agreements to which I have just referred. The question of implied terms in those agreements also looms large because of the failure of DBAG to book the EDTs and OCTs correctly and/or at all (until they settled, whereupon cash settlement entries could be made). In consequence DBAG failed to value or margin these transactions or to report their MTM or to include them in the margin requirements reported to Mr Said and SHI.
The terms about booking, valuing, margining and reporting are, as alleged, to be
implied into the FXPBA which is governed by the law of New York. Experts in the New York law of contract were not called to give oral evidence for cross-examination but their reports on New York law were in evidence before me.
The third key issue which arises relates to Mr Said’s authority to bind SHI in a number of different contexts. The question of his authority to conclude the EDTs and OCTs turns on construction of the Said Letter of Authority (arguably in conjunction with the FXPBA) and evidence of market understanding as to what is included in the term “currency options”, a subject upon which I heard expert market evidence from Mr Malik and Professor Wystup. Questions also arise as to the authority of Mr Said to bind SHI to new margin terms and to waive compliance with the implied terms alleged of booking, valuing, margining and reporting on such matters to SHI.
The fourth key issue relates to the agreement by words or conduct or common assumption between DBAG and SHI, as alleged by DBAG, about the collateralisation of Mr Vik’s FX trading with DBAG on the GPF account. The Amendment Agreement to the Equities ISDA and the Schedule to the FX ISDA (both concluded in November 2006) provided that, “unless otherwise agreed between the parties”, transactions other than FX transactions and currency options transactions should be governed by the Equities ISDA and that FX transactions and currency options transactions (as defined in the 1998 ISDA FX and Currency Option Definitions) should be governed by the FX ISDA. DBAG maintains that, notwithstanding this, the parties proceeded on the basis of Mr Vik’s FX trading being governed by the Equities PBA concluded in 2008.
The existence of the agreement or understanding of the parties to Mr Vik’s FX trading being governed by or margined under the 2008 Agreements turns essentially on the evidence of Mr Brügelmann and Mr Vik as to the conduct of the parties and their understanding and on the contemporaneous documents, which reveal the exchanges between them. Although there was a dispute as to how the margining was done and whether it was truly cross margining or merely an aggregation of FX margin with margin under the Equities PBA and the Listed F&O Agreement, it is clear that DBAG did margin Mr Vik’s FX trading under the suite of 2008 Agreements and not under the FXPBA, whether by reference to the Capital Limitation Agreement or PAL. In cross-examination, Mr Vik said that he did not apply his mind to this issue at the time and that it was an argument for the lawyers about construction and contemporaneous exchanges between the parties. There is here therefore a question of objective construction of documents to show what was or was not “otherwise agreed”, and questions of fact as to the conduct of the parties and/or the understanding of the parties, in the context of common assumption, waiver or estoppel.
A considerable number of other issues relating to liability were raised by the parties in a trial involving 45 days of hearing, opening submissions in writing of 930 and 845 pages and closing submissions of 1530 and 1336 pages. The parties’ stance on some of these issues was not to my mind realistic and I invited the parties to reconsider that stance on more than one occasion early on in the trial. There were no concessions of any substance at any stage, as revealed by the extensive closing submissions which took just about every possible point, however good or bad, including many fall-back arguments. The parties and their representatives are to be commended for adhering to the scheduled timetable for witnesses but the extent of their closing submissions meant that neither was in a position to analyse the other’s volumes of argument in the
time available before scheduled final oral submissions. I therefore allowed one day, on the second day of the long vacation, for making key submissions about the credibility of the witnesses and similar major points and had to postpone further argument until the new term, when it was agreed between the parties that short additional submissions in writing on new points would be made, as they then were in mid-October 2013.
A range of further issues arise in the context of the quantum of DBAG’s claim and SHI’s counterclaim in the event of success on liability. Much labour and expense were incurred on these points, on which the parties adduced expert evidence from experts in valuation, forensic accounting and computers. It was recognised that much of this work would inevitably prove to be redundant, depending on my findings. I have sought to deal not only with all the points necessary to determine this action but also some other major issues which would have arisen had my findings been different.
3. The Bank Audit Report
Following the events of 2008, DBAG conducted an internal “Group Audit” in which investigations were carried out to ascertain the lessons to be learned from the events of October 2008 in relation to SHI. Many of the witnesses who gave evidence before me were interviewed for that process. A draft of that report was disclosed by DBAG but the final version was subject to redaction by lawyers and is less informative. The draft available, dated 19th January 2009, referred to “inadequate FXPB systems infrastructure, business and support functions management oversight and a lack of clear roles and responsibilities [resulting] in four Critical Issues.”
The first of those issues was the acceptance by FXPB of TPF products without access to the right systems to manage those trades. No FXPB New Product Approval Process had been completed for the product. FXPB had since conducted a review and confirmed that no other TPFs were held by it and had stopped accepting complex trades including TPFs. The second issue referred to the lack of defined and implemented clear roles and responsibilities over the set up of new accounts for margining purposes, as between GPF, GES (responsible for F&O), and IBO (the IT team). A review was being conducted of existing accounts to ensure that they were being margined correctly as well as implementing control over the set up of new accounts. The third issue was that inadequate credit risk management technology was utilised to manage the risks as a result of products traded by PWM (Private Wealth Management) clients operating primarily on the GM (Global Markets) infrastructure. In particular the specific processes used to stress the exposures in a credit environment were not employed. The fourth issue was the need for clear definition of PWM and GM responsibilities for clients managed under the PIC (Private International Client) desk framework. It was said that PWM had no defined clear guidance on acceptable email correspondence with clients and in particular portfolio overviews were provided without clearly defining the scope of the reports or including disclaimers.
In the discussion of detailed issues and action plans relating to TPFs, it was said that the systems infrastructure and trade approval and monitoring processes for TPF trades were inadequate. Specifically it was said that FXPB had accepted TPF transactions without completing the NPA process, that FXPB systems’ functionality was insufficient to capture the margin or report to the client and that TPF give-up trades
were accepted from the client/executing brokers without an adequate approval and oversight review by Front Office (by which was meant the business/sales side of FXPB). It was further said that IBO did not have a process to track and escalate trades booked (as proxies) for review and approval (TPFs booked as Resurrecting Fader Options) and a report to monitor the timely capture of proxy trades was absent. System feed exceptions reports were incomplete and the GEM margin system contained trades with valuation errors marked as “N/A” which were not included on the IBO exception report.
In relation to Credit Risk Oversight, it was said that PWM Credit Risk Management’s (CRM’s) focus was on stressing the collateral rather than stressing the transaction as well with stress tests. TPFs had not been stressed since the first transaction was concluded and “no independent function was assigned specific responsibility for stressing transaction exposures and assessing the results against the margin levels and collateral held”.
Mr Roesch, the regional head of Group Audit, set out, in an email, his “final” reading of the investigation. He referred to a “multi organ failure” which had contributed elements to the risk situation. The key contributors the conclusion of into TPFs without NPA and appropriate infrastructure report and the resulting failure to margin TPFs at all which were booked as proxies as Resurrecting Faders in circumstances where they were not margined until October 10th/13th. Other contributors were late bookings (although that was almost irrelevant in the light of the failure to margin), incorrect mapping of GES into the DBX system (the Ignored Payments error), the Russell Multiplier error and “the back-and-forth” with regards to PIC, GM and PWM.
Mr Eggenschwiler, the Global Head of CRM PWM commented that the suggestion that margin levels were insufficient to collateralise the exposures told less than half the story because the reality was that the FXPB process and systems did not and could not reflect the true position or exposure. As their systems could not handle the TPFs, it did not matter how diligent CRM was or how much collateral they might have had because the margin requirement was unknown as a result of the absence of proper booking. He made much the same point in relation to stress testing, commenting, as indeed was the fact, that what was booked on GEM had very probably been stress tested but the TPFs were not correctly reflected in GEM or were completely absent which meant that they would not be subject to such tests. If trades were not booked properly onto the systems, everything was affected downstream so far as CRM was concerned. These appear very fair points to make.
4. The New York Action
On 24 November 2008 SHI commenced an action in New York against DBAG by the filing of a Summons with Notice. SHI’s Complaint was filed and served on January 20th 2009. The key allegations in that complaint relate to what is there referred to as the “Collateral Limitation Agreement” which corresponds in large measure with the “Capital Limitation Agreement” alleged in the English action. It was alleged that between September 2006 and November 2006 there were discussions between DBAG and Mr Said, sometimes with a representative of SHI present, about a prime brokerage arrangement for the bank to provide services to SHI in connection with Mr
Said’s FX trading. This would include the provision of information about the performance of Mr Said’s trading account upon which SHI would rely in order to allow it to monitor the risks involved in the account. In those discussions it was agreed that US$35 million would be pledged as collateral in Geneva and that “Sebastian Holdings’ maximum exposure in connection with the FX trading … was limited to US$35 million”. Reliance was placed upon Clause 2 of the FXPBA and it was alleged that DBAG was obliged to report the net exposure on the account to SHI on at least a daily basis to ensure that its collateral limitation or exposure would not exceed US$35 million.
The Complaint alleged that DBAG had acted in breach of contract in failing to report the MTM figures and net exposure of SHI on the FX trading account on a net basis and that if there had been such reporting, Mr Said would not have been permitted by the bank to make the trades at all or, if SHI had sufficient collateral and chose to make the trades, they would have been liquidated earlier than they were with less or no loss. It was further alleged that the margin calls were wrongful, that DBAG had converted SHI’s assets in consequence of such calls and in closing out SHI’s transactions had acted in breach of fiduciary duty. Both fraud and negligence were alleged against DBAG and claims in restitution were also made.
The only reference to “Structured Options” in the Complaint was the allegation that DBAG had informed Mr Said that it was willing to accept the transactions and had approved them under the FXPBA but had never advised SHI that such trades would create the need for collateral in excess of the US$35 million limit. No allegation was made in the Complaint that such trades were outside the scope of Mr Said’s authority nor was any allegation made in relation to Mr Vik’s FX transactions and the arrangements for them, save that it was said that the margin call of 23rd October 2008 on the GPF account had never been received. Mr Said’s FX trading was expressly said to be entirely separate from any other trading of SHI, which was referred to as taking place in London. There was no trace of the multiple allegations which SHI has pursued in the English action.
DBAG filed a motion to dismiss the Complaint or, in the alternative, to stay the action and on December 10th 2009 Justice Kapnick issued a decision dismissing SHI’s causes of action for breach of fiduciary duty, fraudulent concealment, fraud and negligent misrepresentation, whilst denying DBAG’s motion to dismiss the Complaint based on forum non conveniens and its request for a stay on that basis. In denying the motion to dismiss on the grounds of forum non conveniens, she found that the contracts alleged to have been breached in SHI’s Complaint were the FXPBA and an oral Collateral Limitation Agreement which was alleged to have been made in New York and related to the FXPBA. She dismissed the cause of action for breach of fiduciary duty because, in her view, SHI was a sophisticated investor engaged in arm’s length transactions with DBAG and such transactions did not give rise to fiduciary duties under New York law. Without a fiduciary or other special relationship, the claims for fraudulent concealment and negligent misrepresentation also failed.
The First Department of the Supreme Court, Appellate Division, affirmed Justice Kapnick’s decision on 9th November 2010. This decision was not appealed.
On 10th January 2011 SHI filed an Amended Complaint alleging fourteen causes of action, eight of which were for breach of contract, including breach of the Said Letter of Authority, breach of the implied covenant of good faith and fair dealing, negligence, conversion, money had and received and unjust enrichment. Many of these allegations mirror the allegations made in the current action in this country.
DBAG filed a motion to dismiss some of the claims in the Amended Complaint and, on 8th November 2012, Justice Kapnick granted DBAG’s motion to dismiss one of the breach of contract claims, a second breach of contract claim to the extent that it relied on a purported breach of the Said Letter of Authority and claims for the breach of the implied covenant of good faith and fair dealing, negligence, conversion, money had and received and unjust enrichment. The other causes of action for breach of contract survived on the basis that there were factual issues to be resolved. The reasons for dismissing the negligence claim are discussed in this judgment in the context of the New York law of tort, but in essence, the judge held that there was no independent duty upon DBAG outside the parties’ contracts. She also accepted DBAG’s submissions that the Said Letter of Authority acted as a “complete defence to [SHI’s] allegations that [DBAG] could be held liable for Said’s trading activities”. DBAG had submitted that SHI had expressly granted Mr Said the authority to trade on behalf of SHI and that DBAG assumed no responsibility for that trading under the terms of the Said Letter of Authority. It was incumbent upon SHI to rein in Mr Said’s trading and it could not disclaim knowledge of its own agent’s actions.
On 2nd July 2013 the First Department of the Supreme Court, Appellate Division, upheld Justice Kapnick’s decision, holding that SHI’s sixth and ninth claims for breach of contract arising from unauthorised trades were properly dismissed: “The agreements expressly absolved defendant (DBAG) from any liability for unauthorised trades by the plaintiff’s agent. Indeed, as a general matter, the agent’s knowledge and conduct would have been imputed to plaintiff at any rate, under basic agency principles.” The court held that the negligence claim was properly dismissed, not only as being duplicative of the contract claims but because DBAG was not shown to be subject to duties outside the written contracts.
On 15th October 2013 the First Department of the Supreme Court, Appellate Division denied SHI’s motion for leave to appeal against its rulings to the Court of Appeals.
5. The Law of the Case
It is agreed between the New York lawyers instructed by the parties that, as a matter of New York law, there is no res judicata or collateral estoppel applicable to the decisions of Justice Kapnick or the Appellate Division, First Department in the New York action since those decisions were not final. No res judicata or issue estoppel can arise in this jurisdiction unless the effect of the decisions in New York is effective in New York to preclude further argument on the subject matter of those decisions.
There is a doctrine under New York law, recognised by both sets of New York lawyers, known as the law of the case. The New York Court of Appeals explained this doctrine in People v Evans 94. NY.2d 499 (2000): “The law of the case addresses the potentially preclusive effect of judicial determinations made in the course of single litigation before final judgment”. It is “a judicially crafted policy that expresses the practice of courts generally to refuse to reopen what has been decided.”
It is not however a limit on the court’s power but directs a court’s discretion. Thus the law of the case doctrine precludes the relitigation of questions already decided by the same court or a court of co-ordinate jurisdiction unless there are extraordinary circumstances warranting a departure from the principle.
Despite the submissions of SHI’s lawyers, it appears to me that the decisions taken in the New York action to dismiss various grounds of complaint are decisions on the merits of the issues put before the courts for decision, though not decisions on the merits of the action as a whole. Nonetheless, the application of the doctrine is a discretionary one and, if an amended pleading is subsequently filed, it supersedes the original pleading and the allegations made in it, unless identical in practice to those previously dismissed, are not affected by the prior decisions of the court. The issues which then arise are different. SHI states that it is in the process of reformulating its complaints in the New York action to tally with many of the complaints made in this jurisdiction. Furthermore, the decision of the Appellate Division on 2nd July 2013 apparently can be the subject of appeal after final judgment in the New York action.
There are essentially therefore two reasons why I cannot find that the “law of the case” has the effect of conclusively determining any of the issues which I have to decide.
The first is the nature of the law of the case doctrine which is recognised as directing a court’s discretion, not its authority. It is not a mandatory doctrine but acts as guidance that will generally but not invariably be followed. A change in the circumstances or a change in the law could give rise to a court later departing from its earlier decisions. During the pendency of an action every New York court retains jurisdiction to reconsider its prior, intermediate or interlocutory orders, however rare it may be for any departure from them subsequently to take place.
The second reason is the approach set out in Carl Zeiss Stiftung v Rayner (No. 2) [1967] 1 AC 1853. In order to found issue estoppel in this country there must be a judgment which is final and conclusive and a decision on the merits of an issue which is identical in the two sets of proceedings and which was necessary for decision in the foreign court. The House of Lords emphasised the need for a cautious approach in regard to issue estoppels based on foreign judgments in the light of different procedures and rules in other jurisdictions. There is a need for the English court to be satisfied that the issues in question cannot be relitigated in the foreign country.
I cannot be satisfied about that in relation to the decisions taken. Not only are pleadings to be amended in New York and the appeal process not yet exhausted in relation to the most recent decision of the Appellate Division but the applicability of the principle is in any event, as I have just set out, a discretionary one, which does not give rise to the necessary finality for this Court to accept that issue estoppel arises in relation to any finding made.
As matters stand however, there is a considerable overlap between the allegations made in the current New York pleadings and the allegations made in the English action. I have had the advantage of hearing many days of evidence, both factual and expert, on the allegations made in the English action whereas the New York court had to assume the truth of the facts alleged when deciding whether or not specific claims should be dismissed. Nonetheless the best evidence of New York law in relation to allegations which are common to both actions must be the decisions of the New York court, to which I am entitled to have regard, without treating them as being binding upon me.
6. The Key Witnesses
In addition to the absence of Mr Meidal and Mr Said as witnesses who, on SHI’s case, were key participants in the oral agreements (and for whose absence each party criticised the other) SHI noted the absence of Mr Quezada and Ms Liau, both of whom were employed by DBAG with responsibilities in relation to FXPB. Ms Liau was the Chief Operating Officer of Fixed Income Prime Brokerage and Head of Product for FX Prime Brokerage whilst Mr Quezada performed the same role as Mr Giery (who did give evidence) as a member of the Sales and Product Team in FX and
Electronic Trading under Ms Liau (the FXPB front office, sometimes referred to as “Business”). Those three worked at DBAG’s offices at 60 Wall Street, New York in close proximity to one another. The FXPB Operations Team was based in New Jersey where Steven Kim was the Global Head of FXPB Operations and was ultimately, through intermediate line managers, responsible for Mr Walsh. The latter’s job title was “analyst” but his task was to oversee the trade flow of FXPB accounts by checking that trades had been matched in the TRM system (the automated matching engine used by DBAG) and, when required, by booking trades into DBAG’s FX booking system, RMS, on behalf of clients for whom he was the designated Client Services Representative. He was the Client Services Representative for SHI and was in contact with Mr Said in relation to his FX trading. Mr Said also had dealings with Mr Quezada to whom Mr Walsh looked for assistance and guidance in relation to the EDTs and OCTs. Whilst the other persons mentioned in this paragraph did give evidence, neither Mr Quezada nor Ms Liau did, although once again passages in Mr Quezada’s deposition were relied on by both parties. An organisational chart of DBAG’s Fixed Income Prime Brokerage (FIPB) and FXPB personnel appears in Annex 2.
Additionally, SHI criticised DBAG for not producing Mr Gunewardena to give evidence (although he was deposed). He was at the time the Global Head of Fixed Income Prime Brokerage to whom Ms Liau and Mr Quezada were ultimately responsible. The comment can be made that Mr Walsh, who took up permanent employment at DBAG from about the middle of December 2006, having graduated earlier that year, was left exposed to the full force of criticism for the acceptance of EDTs and OCTs which were known by Mr Quezada, Mr Kim, Mr Manrique and possibly others to be incapable of being properly booked, valued or margined by DBAG on its systems at the time, leaving this junior employee with an insoluble problem which he resolved by booking them under a “place holder” as “Resurrecting Faders” or, from mid 2008 onwards, largely by not booking them at all until the transactions had concluded with cash settlement, whereupon the cash entries could be and were made. When all this came to light in October 2008 because MS were asking for collateral from DBAG in respect of exotic trades booked with SHI, Mr Walsh hurriedly booked the outstanding unbooked EDTs and OCTs, once again as
Resurrecting Faders, as place holders.
I did not have any difficulty in accepting the vast majority of the evidence of Mr Walsh who was about 24 or 25 at the material time and out of his depth when dealing with someone like Mr Said who was a dominant, forceful personality experienced in
FX trading and a man who insisted on getting his own way, insisted on conducting the trades that he wished to do and threatened to complain to Mr Vik and take SHI’s business elsewhere if Mr Walsh, Mr Quezada and DBAG did not play ball. Mr Walsh’s own understanding of the transactions effected by Mr Said which gave rise to the major issues was, it appears, somewhat limited, since he had to have those trades explained to him by Mr Said in October 2008, with the possibility of huge losses far from his mind until the collateral call came in from MS for sums in excess of US$100 million. He was, it seemed to me, straightforward in the evidence he gave, although, not unsurprisingly, he could not recall detailed day to day events in 2007-8. It was plain from the contemporaneous recorded telephone calls that he went through agonies at the time in not knowing what to do and he squarely faced up in the witness box to his failings, without casting blame on to Mr Quezada or Mr Kim, as well he might have done, with justification. As will appear from a more comprehensive review of the evidence later in this judgment, the issue as to whether or not to accept these types of transaction was something well above Mr Walsh’s pay grade and was effectively taken by Mr Quezada, with acquiescence from Mr Kim, in the full knowledge of the problems in booking the trades and the hope of finding a solution to them in time. Mr Quezada too was anxious not to upset Mr Said and to go along with his demands that he should be able to effect these trades through the prime brokerage arrangements. The extent to which Ms Liau was aware of what was going on remains uncertain but since Mr Quezada sat next to Mr Giery and they both sat opposite Ms Liau, about a metre or so away, it is hard to believe that they were all unaware of at least some of the problems which arose in relation to the disputed transactions, although Mr Giery’s focus was different in 2008 and he had left by September of that year. Whilst attempts were made to resolve the booking issues, Mr Walsh was essentially left to manage the situation as best he could, whilst those issues remained unresolved and had little prospect, it would appear, of resolution, whether in a period of 6 months, as envisaged by Mr Kim, or at all. Mr Quezada and to a lesser extent Mr Spokoyny were aware of this.
Mr Brügelmann, as the client relationship manager at DBS and the man with whom Mr Vik dealt most, was a man who sought to manage relationships and smooth over issues. He sought to find solutions rather than simply face customers with a problem or hard facts. He was polite, calm and personable and would seek to avoid confrontation. He was undoubtedly in awe of Mr Vik and conscientiously sought to fulfil his every instruction. I found him essentially an honest witness although I did not think he had much real recollection of the detailed history. His failure to recall that he had books of notes/jottings until shortly before the trial was extraordinary but not, I am satisfied, sinister. Much of his evidence was a learned recital of the documents which he had read with care. He sought to reconstruct what had happened, at times following his reading of the documents when answering questions, which led him into errors both favourable and unfavourable to DBAG.
Further, it seemed to me that there were times during the history of events when he told people what they wanted to hear rather than the unvarnished truth. He was loyal to Mr Vik and throughout the relevant history often took his part when dealing with others in DBAG but, when everything went wrong in October 2008, he, along with his superiors who had taken control of the situation by then, toed the party line. Although he came to know of the issues after the margin calls, he was not straightforward in dealing with Mr Vik and telling him of the Ignored Payments Error
or of the booking, valuation and margining problems that DBAG had and their failure to margin the disputed trades properly and/or at all during the preceding months. At that stage more senior people at DBAG were controlling what was said and he was not to know how much Mr Vik’s knowledge was superior to his own with regard to Mr Said’s FX trading.
Where however there was any issue between him and Mr Vik in relation to instructions given by Mr Vik and his fulfilment of them, I have no hesitation in accepting his evidence that he would only ever do what Mr Vik instructed him to do. However when transferring NOK 290m from DBS in August 2008 to the GPF account, I find that he did so without express instructions at the time, moving assets in order to avoid a problem for Mr Vik, based on what he thought Mr Vik would want and what would solve the issue, relying on an outdated instruction for transfer at the opening of the GPF account. When seeking to justify transfers in October 2008 he seized on a “standing instruction” which was nothing of the sort, but, as it transpired, there was in fact a telephone transcript that showed that he had Mr Vik’s agreement to the transfer.
There was an attempt by DBAG not to reveal to Mr Vik in the period from October 13th onwards the fact that DBAG’s systems had not been capable of booking, valuing or margining the trades or reporting on them. A number of witnesses employed by DBAG and DBS participated in this attempted “cover up”, which was in any event futile because Mr Said was well aware of the position, as should have been obvious to the individuals concerned, and he had in fact told Mr Vik. Given the size of the margin calls, it should also have been obvious that he must have told Mr Vik the position. It is the more senior employees who must take responsibility for this, as Ms Serafini’s evidence confirmed, because decisions were being taken at a high level about the margin calls, about the basis on which they were to be made and the explanations to be given in an effort to keep the payments coming. By the time of the second margin call, the problems were known to all the FXPB team who were dealing in any way with SHI. They knew that recent market movements alone could not account for the sums now required by way of margin. Attempts by Mr Gunewardena and Ms Liau to explain the calls on this basis were ridiculous. The approach adopted by DBAG does it no credit at all. Misleading and inaccurate statements were made and made to the knowledge of those concerned which was both dishonest and senseless in the circumstances. If they thought that Mr Vik was being deceived, they were themselves deluded. They were, it seems, surprised that Mr Vik paid up on the first five margin calls but the reason was not that they had deceived him but that he considered that he had entrusted Mr Said with considerable trading authority for SHI and felt that SHI was bound by his commitments, as indeed it was.
In October 2008 the financial world was in turmoil. There was doubt as to the solvency of banks of considerable size. Losses of the amounts involved in this action could cost individuals their jobs if responsibility fell on them for such losses. FXPB was keen not to let on to PWM in the shape of Mr Brügelmann or its Credit department, CRM, Messrs Halfmann and Lay that it had been at fault. Mr Gunewardena must take the blame for his own attempts to deceive SHI in telephone calls to Mr Vik, but Ms Serafini, Ms Liau, Mr Quezada, Mr Kim and Mr Spokoyny were not forthcoming in internal dealings with DBS and CRM, nor in conversations with Mr Vik, insofar as they were involved. What they were seeking to do was to
hide the deficiencies of FXPB for which each must have considered he or she or the FXPB team had some responsibility. Those who gave evidence were protective of their own positions. Ms Serafini sought to suggest that Mr Cloete or some other similar figure lay behind the decision not to tell Mr Vik the truth about failures to margin but Mr Cloete denied that he played any part in that and the evidence did not go far enough for me to conclude who might have been involved, apart from those I have named.
DBAG’s other witnesses were all straightforward in the evidence they gave, which was, in so far as they had real recollection, reliable.
Mr Quezada did not give evidence but SHI relied on his answers in deposition, as did DBAG. I was unable to place any reliance on anything much he said in those answers because of the inconsistency with the record revealed in the contemporaneous documents. He appears to be doing business currently with Mr Said and, as the prime individual responsible for deciding to take in the EDTs, knowing that they could not be booked, he had good reason to pretend ignorance of matters which were, on the documents, plainly known to him, to be defensive of his reputation and to seek to blame others at DBAG rather than to blame Mr Said or to tell the whole truth. When the extent of Mr Said’s MTM position came to light in October 2008, in revealing conversations with Mr Said, Mr Quezada wanted matters kept under the radar.
At first sight Mr Vik had a genuine grievance in respect of Mr Said’s trading which exposed SHI to large losses. He provided collateral of US$35 million for Mr Said’s FX trading.
He knew the basis upon which banks operated and how margin was ordinarily required to support such trading. He could readily have expected that, as DBAG calculated the margin requirements for Mr Said’s FX trading, if and when the requirement rose to US$35 million, Mr Said and SHI would be notified and asked to put up further collateral or reduce positions so as to bring margin down within the US$35 million provided. According to the bank’s promotional documents relating to its systems, MTM valuation was meant to take place about every 15 minutes of the day so that the scope for collateral requirements to exceed the US$35 million figure by much before notification was very limited if all trades were being booked, valued and margined correctly.
On October 13th Mr Vik was faced with a margin call for approximately US$98.8 million. The following day there was a further call for about US$202 million. This was followed by a call for US$125 million (approximately) the next day which was rolled into a margin call the following day of US$175 million and on 17th October an additional call was made of nearly US$35 million. Since Mr Vik had already decided the previous weekend to close down transactions as opposed to just putting up additional margin, these figures included large premium figures for the closing out of transactions, but, on the face of it, without some knowledge of what was going on, he might have expected margin calls to be limited to a reasonably small excess over and above US$35 million, even given the extraordinary state of the markets in October 2008, following the collapse of Lehman Brothers in September and the events which followed.
Mr Vik was given a clear idea of the likely total of calls on 16th October in a telephone conversation with Mr Gunewardena and others. Mr Vik paid all the FXPBA margin calls and only refused to come up with more cash when told of the bank’s accounting errors (amounting to some US$430 million) and/or when faced with the GPF margin call which he maintains he did not receive.
Mr Vik had a genuine grievance but at issue here is whether that grievance was properly directed at Mr Said, SHI’s agent, at DBAG or at both. Furthermore the question arises as to whether he himself was also to blame because of his knowledge and/or approval of Mr Said’s trading. The timing of his knowledge of different aspects of Mr Said’s activities may assume some importance in this connection. As Mr Said himself said at the time, however, there is no doubt that the prime cause of the losses was Mr Said’s own market misjudgments in the torrid circumstances of the autumn of 2008.
As appears below, by the time Mr Vik caused SHI to pay margin calls in excess of US$500 million, he must have known from Mr Said why those calls were being made and why there had not been any earlier substantial margin requirement – namely that DBAG’s systems were incapable of accurately valuing and margining the Structured Options, the EDTs and OCTs, which gave rise to the huge margin requirements, as suddenly brought home to DBAG by MS’s demand for inter-bank collateral in respect of some of them. Whatever his state of knowledge at an earlier stage, as a result of exchanges with Mr Said, he knew by 9th October that DBAG had failed to margin the EDTs and by October 10 that closing out the trades was likely to cost hundreds of millions of US$. That was the course which, over the weekend of 10th-13th October, he decided to take. By the time that DBAG issued its first margin call, Mr Said must have alerted him to the true position, in so far as it was known to Mr Said. So it was that thereafter Mr Vik kept pressing DBAG to inform him of the history of margining, knowing that it could not do so without revealing its inability to book, value and margin the disputed trades correctly. He must have known, because Mr Said must have told him, that Mr Said had agreed to DBAG not providing MTM figures in respect of the disputed trades with consequent impact on margin reports. He knew that SHI had been getting a “free ride” on margin because Mr Said told him that. Because of the terms of the authority which had been given to Mr Said by SHI, Mr Vik must have known that there were problems for SHI in relation to waiver of compliance with any implied terms as to accurate booking, valuation, margining and reporting of the trades.
Consequently SHI has looked for other ways of putting the case which would not run foul of any arguments based upon Mr Said’s authority to bind SHI. I regret to say that in these circumstances I have concluded that Mr Vik has invented oral agreements that were not and could never have been made by Mr Meidal, or Mr Brügelmann, and has instructed his lawyers to pursue contrived arguments which bear no relation to the agreements made in writing, nor to reality.
It is possible that he has deceived himself into thinking that some such agreements must have been made because of the US$35 million collateral put up in the first place, which he would, at the outset, have expected to be close to the maximum he would actually lose on Mr Said’s trading. He undoubtedly considered that he had allocated a US$35m fund as capital to support Mr Said’s trading, which would act as a something of a brake, if DBAG’s calculations of margin reflected the market situation with a
degree of accuracy and if it demanded collateral according to its maximum entitlement, but, along with most traders, he regarded collateral as something to be minimised in order to obtain maximum leverage against it and he entrusted Mr Said with authority in dealing with DBAG on issues of margin as well as trade.
At all times he knew the difference between liability for trading losses on the one hand and the provision of margin as collateral in respect of any liability on the other. He knew that the two were not coincident and that assessment of market value (MTM) was not a science. If he had understood the concept of 95% VaR he would have realised that, ex hypothesi, it was not expected to cover 5% of the situations which might arise (although the application of a multiple to the VaR figure might be anticipated to cover the deficit in most situations, to which October 2008 might well be an exception). If Mr Said incurred trading losses on trades otherwise within his authority, the amount of collateral supplied could not be or act as a trading limit in itself, though it might in practice act as a brake on losses, because of DBAG’s expected desire to insist on protecting itself by obtaining security in respect of potential losses caused by SHI’s trading. The fact that some DBS and DBAG individuals, including Mr Brügelmann, referred to the US$35m as a “risk budget”, a “VaR budget”, an “equity chip”, a “maximum position limit” or even as a “trading limit” has provided SHI with something on which to seize to support an argument but it merely reflects a loose use of language and the view that this was the amount of capital Mr Vik wished to provide by way of collateral for Mr Said’s trading. They understood that Mr Vik had set up Mr Said with US$35m as trading capital – as the money available to him as security for his trading – but neither he nor they could reasonably regard that as a limit on SHI’s liability for his trading, if it was otherwise authorised.
No mitigation is available in relation to the allegation of a specific agreement that Mr Said would only trade vanilla options. If that had been the case, the email exchanges between Mr Vik and Mr Said in 2008 about the very disputed transactions, where reference is made to “range bets”, “range trades” and the discussion of their terms, could not have taken place. These were, as described, obviously not vanilla trades in anyone’s parlance and Mr Vik could not have thought otherwise. Furthermore, if he had thought that these were outside the scope of Mr Said’s authority, Mr Vik would have told Mr Said and DBAG that this was the case when he says he first heard of them. He did not. On the contrary, he appears (in email exchanges) to have supported and encouraged Mr Said to continue in these trades in the light of the profits then being made which it was hoped would continue from mid 2008 onwards and he raised no issue about them throughout the margin calls, when it is clear that he knew that they were the main cause of the losses.
It is noteworthy that, in the original Complaint filed by SHI in New York, the only claim put forward is made on the basis of the US$35 million collateral limit, constituted by the CLA. This Complaint was filed in January 2009, by which time on any view Mr Vik must have had a full picture of what Mr Said had done and Mr Said’s state of knowledge of what had gone on at DBAG. There is no suggestion of any of the other alleged agreements restricting Mr Said to vanilla trades, to the PAL, to margining his own trades under the FXPBA or FX ISDA or anything else. The Complaint is based on an agreement to a US$35m limit of exposure and DBAG’s failure to monitor the risk and keep SHI informed of margin calculations of its maximum entitlement with consequent failure to prevent Mr Said trading beyond such limit by demanding more margin.
Furthermore, the fact that Mr Vik caused SHI to pay over US$500 million in margin calls in respect of the transactions and to close out those transactions that he now says were unauthorised, in circumstances where he now maintains that he had an agreed “trading limit” with DBAG for Mr Said’s FX trading of US$35 million, speaks for itself. If Mr Vik had made the agreements to which he testified in the witness box, his conduct is inexplicable.
SHI disclosed no notes of any conversations or meetings of Mr Vik with others and no internal documents of the kind that would be expected to exist as set out elsewhere in this judgment. Mr Vik’s evidence, however, in lengthy statements and in crossexamination was argumentative in the sense that he put forward a case by reference to documents which he had not seen at the time. His actual recollection of events was limited and much of his statements consisted of a running commentary on DBAG’s documents rather than evidence of what he saw, heard or knew at the time.
Later in this judgment I will explore in more detail the reasons why I cannot accept Mr Vik’s evidence on these and other areas, but I am driven to the conclusion that he is not a reliable witness and that where there are conflicts of evidence I would need to be very careful in accepting anything Mr Vik said in preference to the evidence of other witnesses. Of course, a witness may be inaccurate or lie about one matter and not be inaccurate or lie about another. The motivation for lying in one area is highly relevant in determining whether a witness is lying on another aspect. Apart from the desire to protect SHI’s assets by refusing disclosure and making misleading statements the only motivation here is the advancement of SHI’s case against DBAG in respect of the trading losses incurred by SHI’s own agent, Mr Said, and I am clear that this was Mr Vik’s motivation. Although Mr Vik sought, when he came to give evidence, to avoid some of the conflicts of evidence presented by the way that SHI’s case had been framed in the pleadings, whether by limiting the party to the oral agreements to Mr Meidal in his witness statements or by diluting some of the evidence which was in his six witness statements, the documents, the written formal contracts and the commercial realities meant that his evidence, even when standing uncontradicted by any DBAG or DBS witness, is incapable of acceptance on the crucial issues. I have concluded that in some respects he was simply dishonest.
Mr Said, whilst still employed by one of Mr Vik’s companies, swore an affidavit (one of three) for use in the New York proceedings, in support of SHI’s case about the types of trade he effected. When he came to give evidence on deposition, he abjured that part of that affidavit, recognising that the “under orders” defence would not mitigate his responsibility for making false statements and saying what he had. Whilst SHI maintained that the stance he took in his depositions was adopted to defend himself against potential claims by SHI and his affidavit was truthful in all matters and respects, the documents and all the commercial probabilities support his evidence in those depositions about the absence of any agreement with Mr Vik limiting his trading in any relevant way and about the extent of his discussions with Mr Vik concerning the different types of exotic trades that he pursued. Once again, when it comes to a conflict between Mr Vik’s evidence and Mr Said’s evidence, albeit that the latter’s evidence is on deposition alone, there is good reason to accept his evidence, backed up as it is by the documents, in preference to that of Mr Vik who is
looking to defeat DBAG’s claim and to pursue his own counterclaim against DBAG for sums vastly in excess of anything that he could possibly think of recovering from Mr Said, who could not be good for a fraction of the sums in issue in this action, were he to be sued for breach of authority, and who would not therefore be likely to fear such an action against him.
I am conscious that many of the answers given by Mr Said on deposition however were self-justificatory. He was happy to blame others for the results of the trading judgments he made, which, at the time, he had acknowledged were the sole cause of the losses, combined with the “perfect storm” of the market conditions of October 2008. With the passing of time, his “bounce” had returned and, as with most issues in this action, it is the documents which provide the surest guide to the truth, coupled with the commercial probabilities.
7. The Contractual Documents
87. The starting point must be the written contracts concluded between the parties. I need not dwell on the Equities ISDA at this stage, nor the Amending Agreement. It is the contracts concluded with the date of signature of 28th November 2006 with which I am most concerned – the Said Letter of Authority, the FXPBA, the FX ISDA, Schedule and CSA and the Pledge Agreement. As they were entered into as part of a suite of contracts which were intended to govern the situation between them, each forms part of the matrix for the others and each can be read in the light of the others. Whilst each must be examined for its own language and terminology and its provisions and obligations construed for what they are, they could be expected to fit together, so that a commercial reading would give rise to a consistency of result. The TPMCA was an additional agreement between DBAG and DBS, to which SHI was not a party, the terms of which were unknown to SHI and therefore falls into a different category with no effect on construction of the other agreements.
7(a) The Principles of New York Law applicable to the construction of contracts and the implication of terms therein
It is agreed that both the FXPBA (expressly) and the Said Letter of Authority (impliedly) are governed by the law of New York. The principles of construction applicable are agreed by the parties’ respective experts in large part and, equally, for the most part, are familiar to English lawyers, subject to different emphases and nuances.
The first basic principle is known as the “four corners rule”: “When parties set down their agreement in a clear, complete document, their writing should as a rule be enforced according to its terms. Evidence outside the four corners of the document as to what was really intended but unstated or misstated is generally inadmissible to add to or vary the writing.” W.W.W. Associates Inc v Giancontieri 77 N.Y.2d 157 (1990). Both the professors of law engaged by the parties were aware of no case in which the court had admitted evidence “as to what was really intended” outside the four corners of a clear and complete document in order to add to or vary an agreement set down in that document. They were agreed that contracts fall to be construed in accordance with the manifested intent of the parties in the light of the surrounding circumstances.
It is further agreed that contemporaneous agreements made between the same parties relating to the same subject matter are to be read together and interpreted as forming part of one and the same transaction.
If a term is ambiguous however, and its meaning is not revealed by examination of the written contract, extrinsic evidence of the parties’ intentions may be considered, including evidence of custom and usage in the relevant trade. There was a measure of difference between the experts as to the need for ambiguity for evidence of trade usage to be admissible, but, in the context of the use of terminology such as “currency option” and “Structured Options” where it is alleged that there is a trade meaning and the terms themselves are open to interpretation, there is ambiguity which would undoubtedly allow for the admission of evidence of that trade usage. It is agreed that evidence of custom and usage in the relevant trade may be admitted to explain the meaning of words used in a contract if the custom and usage is reasonable, meets appropriate standards of consistency and of being widely known in the particular trade, and does not conflict with the express terms of the contract or a rule of law. Courts use a variety of adjectives to express these requirements, such as “uniform”, “well-settled” and “established” for the requirement of consistency and “well-known” and “notorious” for the requirement of being sufficiently widely known.
In addition to the admissibility of evidence of trade, custom and usage, where a contract is ambiguous, extrinsic evidence may be adduced, not just of the parties’ intentions as outwardly expressed at the time of contracting, but of the dealings of the parties leading up to execution of the agreement, including the drafting history, evidence of the parties’ relevant conversations and negotiations, the parties’ relationship and the purpose of the contract. Furthermore, the parties’ course of performance (or course of practical construction) and course of dealing are admissible as aids to construction of ambiguous provisions.
Ambiguity is a question of law for the court to decide but a contractual term is ambiguous if it is reasonably susceptible of more than one interpretation. One of the professors considered that it was possible that a court might go beyond the four corners of the contract to determine whether it contained a “latent ambiguity” but, for the reasons I have already given, in the context of the current action, there was no disagreement that expert evidence would be admissible as to what the terms “currency option” and “Structured Options” were generally used to mean according to the established custom and usage in the relevant trade.
If a contract that contains a “no oral modifications” clause has been modified orally (or in a manner otherwise inconsistent with the clause) it was agreed that the clause will be given effect in order to prevent the modification only if the modification is completely executory – i.e. it has not been performed by either party. The principle of estoppel may also apply if there is reliance by a party seeking modification sufficient to prevent the party opposing the modification from doing so. Partial performance that avoids the requirement of a written and signed modification must be “unequivocally referable” to the alleged modification. Determination of whether a modification is completely executory and whether partial performance is unequivocally referable to an alleged modification is a matter for the court.
So far as concerns implication of terms, the experts disagreed on the importance of the distinction between terms implied in fact and terms implied by law and on the
scope of implication and the willingness of the courts to imply terms. They agreed that, as a matter of law, there is under New York law an implied covenant of Good Faith and Fair Dealing. A breach of this covenant may occur where a defendant has acted with intent to deprive a plaintiff of its contractual rights or if the defendant has acted in reckless or neglectful disregard of such rights. Both agreed that acting without appropriate care in carrying out a contractual obligation could violate this implied duty of good faith and fair dealing, if the lack of care reflected an intention to cause harm to the other party. It was also agreed that acting without appropriate care, which was merely accidental and lacking intention, did not violate the implied duty. Whether or not the acts of a defendant were in such bad faith or in such wilful or neglectful disregard of the rights of a plaintiff as to constitute a breach of this implied term would depend on the facts presented to the Court.
Professor Cohen said that the duty of good faith and fair dealing was not selfdefining. There have been various elucidations of the concept by New York courts but the most common formulation is that the implied covenant “embraces a pledge that neither party shall do anything which will have the effect of destroying or injuring the right of the other party to receive the fruits of the contract”. Professor Fishman stated that the duty of good faith not only proscribes undesirable conduct but may require affirmative duties, including a duty not only to refrain from hindering or preventing the occurrence of the party’s own duty or performance of the other party, but also to do whatever is necessary to enable him to perform. He went on to say that the claim of implied duty of good faith and fair dealing brings to light implicit duties to act in good faith which are already present but not necessarily specified in the contract. He said that “the claim cannot create new duties under a contract or substitute for an insufficient contract claim”, footnoting this by saying that in order for the implied duty of good faith and fair dealing to stand as a cause of action, there has to be an underlying contractual obligation between the parties.
In my judgment, as Professor Cohen points out, the formulation to which he referred is made clear by subsequent New York cases that constrain the courts from going beyond the contract in applying the implied covenants. The duty of good faith cannot add to, detract from or alter the terms of the contract itself and cannot be used to create independent obligations beyond those agreed upon and stated in the express language of the contract.
The effect of both professors’ views is that a MacKay v Dick type of obligation in English law is encompassed by the implied covenant of good faith and fair dealing. A party must not put it out of his power to perform his obligations or prevent performance by the other party and may be required to co-operate to enable the other party to perform. The implied covenant may attach to the exercise of a discretion or the exercise of a right as the authorities cited by Professor Fishman in his footnote 93 exemplify. Although in that sense the implied covenant is free-standing, it does directly relate to the terms of the contract and the manner in which existing rights and obligations of the parties are to be performed. It thus does not create independent substantive obligations.
As Professor Cohen points out, compliance with the duty of good faith in a contract is often described by the absence of its opposite, namely bad faith: “A breach of the covenant depends upon a finding that the defendant acted with intent to deprive the plaintiff of his rights under the agreement to which the defendant was a party, or, if
the same was brought about by conduct of the defendant in such reckless or neglectful disregard of plaintiff’s contract rights as to justify an inference of bad faith”.
Similarly, with fair dealing: “The law contemplates fair dealing and not its opposite. Persons invoking the aid of contracts are under an implied obligation to exercise good faith not to frustrate the contracts into which they have entered. The rule is grounded in many cases that in every contract there is an implied undertaking on the part of each party that he will not intentionally and purposely do anything to prevent the other party from carrying out the agreement on his part.”
There is thus a requirement of intentionality and purpose in breaching the implied covenant. Acting without appropriate care in carrying out a contractual obligation can only violate a party’s implied duty of good faith and fair dealing if the lack of care reflects an intention to harm the other party or something akin to reckless disregard of another party’s rights.
A breach of this covenant may occur where a defendant has acted with intention to deprive a plaintiff of its contractual rights or if the defendant has acted in reckless or neglectful disregard of such rights. It is a question of fact whether or not an inference of bad faith can be justified. The line between intention and pure accident, which can be difficult to draw, is for the determination of the court. In the context of the present action, this distinction would not appear to matter.
There was greater disagreement about the circumstances in which a court would imply other terms. Thus SHI, on the basis of its own expert’s view, contended that a term will be implied into a contract where a reasonable person in the position of a party to a contract would be justified in understanding that such a term was included (although not set out explicitly) in the contract. A court may imply an essential omitted term that the parties did not consider at the time of drafting the agreement or when circumstances have changed since the agreement was formed or which was implicit in the agreement at the outset. DBAG, on the basis of its expert’s view, contended that, when a contract is made between sophisticated parties at arm’s’ length, courts are extremely reluctant to imply a term that the parties have neglected specifically to agree and (subject to custom and usage) will only do so when it is clear that the parties must have intended to include such a term but failed to do so. When parties set down their agreement in a clear, complete document, their writing should be enforced according to its terms and it is the role of the courts to enforce the agreement made by the parties and not to add, excise or distort the meaning of the terms they chose to include, thereby creating a new contract under the guise of constructional implication. Thus, a term will only be implied into a contract where a reasonable person in the position of the parties to a contract would be justified in understanding that such a term was included and where such a term, at the time of contracting, was in fact implicit in the agreement viewed as a whole. Notwithstanding that, a court would not imply a term that conflicted with the express terms of a written agreement. SHI’s only real qualification of this was to say that a term could be implicit in the agreement viewed as a whole where the parties could not necessarily be taken to have intended to include such a term but failed to do so. It was sufficient if the parties had not turned their minds to a particular topic but the Court determined that, had they done so, they would have agreed to the implied term alleged.
There was disagreement also about the implication of a duty of reasonable care and skill. Such a term is applied to building and construction contracts and has been applied to the provision of professional services where causes of action for malpractice and breach of contract are concurrently alleged. The authorities where such an implication has been upheld typically refer to “services” or “work” to be performed under a contract where the specified duty is described more in terms of a process than in terms of a specific result. Where a contract sets out obligations which are absolute as to the result required, questions of due care and skill are irrelevant. There is no case known to either of the experts where a promise to use reasonable care and skill has been implied in connection with a contract such as the FXPBA. In practice, in the context of the current action, it would not appear that there is any real issue as to the implication of a duty of care and skill in relation to any express obligation in the FXPBA. The question is whether or not a term should be implied requiring DBAG accurately to book, value and margin the trades in SHI’s portfolio and to report those matters to SHI or whether a term should be implied that DBAG should exercise reasonable care in carrying out those functions or whether no such term falls to be implied at all.
Under New York law, a waiver occurs where a contracting party dispenses with the performance of something that it has a right to exact or could have demanded or insisted upon if it chose to do so. A valid waiver requires no more than the voluntary knowing and intentional abandonment of a known right which, but for the waiver, would have been enforceable. A waiver may arise by either an express agreement or by such conduct or failure to act which evidences an intent not to claim the purported advantage. The experts agreed that “mere silence, oversight or thoughtlessness in failing to object” to a breach could not amount to a waiver although where acquiescence featured in this spectrum was in issue. The burden of proof of a waiver is upon the person asserting it and there must be a clear manifestation of intent to relinquish the contractual protection. Waiver does not need consideration to be effective.
None of the differences between the experts on the New York law of contract are significant for any decision I have to make in the light of my findings of fact in this action.
7(b) The Said Letter of Authority
On 28th November 2006, as part and parcel of the FXPB arrangements, SHI and Mr Said each signed a letter addressed to DBAG. It was headed “Re: Klaus Said”. It is common ground that this Letter is governed by New York law, as is the FXPBA. The key paragraphs read as follows:
“We, the undersigned, the directors of Sebastian Holdings Inc., (the "Company") hereby authorize Mr. Klaus Said, Vik Brothers, 10 Ashton Drive, Greenwich, CT 06831 (the "Agent"), to trade on behalf of the Company for the purpose of executing spot, tom next and forward foreign exchange transactions and currency options (herein, 'FX and Options Transactions") with Deutsche Bank AG ("DBAG"). We hereby authorize the Agent to sign and deliver on our behalf and in our name any documentation related to the execution of any such FX and Options Transaction, including, without limitation, ISDA master agreements, schedules, confirmations, credit support annexes, security interests or other credit support documentation (herein, as amended from time to time, "Documentation").
We hereby recognize and agree that Deutsche Bank AG ("DBAG") will rely upon this letter in connection with FX and Options Transactions. We further hereby agree that we will be subject to the terms and obligations of, and liabilities contained in, any FX or Options Transaction or related Documentation executed by the Agent on our behalf to the same extent as if we were directly executing such FX or Options Transaction or were directly the signatory of any such Documentation. We further hereby recognize and agree that DBAG shall have no duty to inquire as to the nature of the relationship between us and the Agent nor as to any restrictions upon the activities of the Agent in connection with the Agent's execution of FX and Options Transactions on our behalf.”
The best evidence of New York law on the issue of construction of this Letter might be thought to be the decision of Justice Kapnick in the Supreme Court of the State of New York dated 9 November 2012, as upheld by the Appellate Division on 2 July 2013. The Letter sets out the terms of the authority given by SHI to Mr Said to bind SHI in its dealings with DBAG to trade, and to sign and deliver documentation.
Justice Kapnick, in determining the sixth and ninth causes of action in the New York proceedings, accepted DBAG’s argument based upon the Said Letter of Authority and the specific recognition of the absence of any duty on the part of DBAG to enquire as to the nature of the relationship between SHI and Mr Said or as to any restrictions upon Mr Said’s activities in connection with the execution of FX and Options Transactions on SHI’s behalf. She referred to DBAG’s submission that, if SHI had intended to rein in Mr Said’s trading, it was incumbent upon SHI to do so itself in accordance with the revocation procedures in the Letter whilst DBAG assumed no responsibilities thereunder. The third paragraph of the Letter, which I have not set out in full, simply provided that the letter should remain in full force and effect and the authorisation contained in it would be irrevocable “until receipt of notice in writing” from SHI to Mr Said and to DBAG, stating that the authorisation had been revoked. Despite SHI’s reliance upon the FXPBA and the alleged limitation to “vanilla” FX transactions, Justice Kapnick found that the Letter of Authority acted as a complete defence to SHI’s allegations that DBAG could be held liable for any of Mr Said’s trading activities. She held that DBAG had no duty to regulate Mr Said’s trading activities and could not be found liable for permitting him to engage in the EDTs.
On appeal, the Appellate Division held that “the plaintiff’s sixth and ninth claims for breach of contract arising from unauthorised trades were properly dismissed. The agreements expressly absolved defendant from any liability from unauthorised trades by plaintiff’s agent.”
Whilst I do not consider, as set out elsewhere in this judgment, that these decisions create an “issue estoppel” which binds me from deciding differently in this action, the
best evidence of the application of New York principles of construction to the Said Letter of Authority must be the decisions of Justice Kapnick and the Appellate Division. These courts read the first two paragraphs of the Letter as giving authority to Mr Said to trade in the types of transactions to which they referred whilst absolving DBAG from any obligation to enquire whether or not the transactions concluded by Mr Said did or did not fall within the description of the FX and Options Transactions referred to. The Appellate Division went on to say that “As a general matter, the agent’s knowledge and conduct would have been imputed to plaintiff (SHI) at any rate under basic agency principles.”
Before me, DBAG did not restrict itself to this approach. DBAG submitted that the EDTs were “currency options” within the meaning of the first paragraph of the Letter. If reliance on the Letter meant that DBAG was on notice that Mr Said’s authority was restricted to the FX and Options Transactions referred to, as opposed to permitting him to trade in such transactions, without any duty on DBAG to enquire further about the agency relationship between SHI and Mr Said, it was clear that the only question which could arise was whether or not the trades in question did fall within the definition. On the wording of the second paragraph there could be no possible obligation to enquire about any other internal restriction, whether relating to a US$35 million trading limit or a limitation to vanilla transactions, unless that was the true meaning of the Letter itself.
It is common ground between the parties that the EDTs and OCTs do not fall within the description of “spot, tom next and forward foreign exchange transactions”. The issue is whether or not they constitute “currency options” within the meaning of the first paragraph of the Letter, either as a matter of language or market understanding. That issue may turn on questions of legal analysis of the individual trades, whatever the terminology used to describe them, and in particular whether they truly have an element of optionality within them. Alternatively, if the market understands such trades to be options, does the authority conferred upon Mr Said by SHI encompass all that the market understands to be options or understood from time to time to be options, or only what a lawyer would properly regard as options? SHI contends that the trades have no optionality whilst DBAG, with the exception of two correlation swaps, contends that they do. Moreover, whether or not there is express optionality, DBAG maintains that they have embedded optionality and as a matter of market practice in the world of FX, they are seen as “currency options” so that the wording used in the Said Letter of Authority must be construed to encompass them. The parties adduced expert evidence on this issue.
It can be seen that there is no express reference to “Structured Options” as there is in the FXPBA (see below) where much the same form of words is used when setting out the authority given by DBAG to SHI to act as its agent in executing FX transactions. In the context of the FXPBA, Structured Options are plainly seen as being “currency options” because the definition of a Structured Option is “any Option other than one which is a (i) put or call that does not have special features, or (ii) single barrier option”. Such Structured Options, under the terms of Clause 2 of the FXPBA, required the approval of DBAG at the time when SHI concluded the deal. Whether or not an item is a Structured Option, whatever the terminology commonly used, must again depend upon the same criteria to which I have already referred. To be a Structured Option, it must first of all be an option, but DBAG prays in aid the use of
the terminology of Structured Option for a large variety of transactions which are said to include those at issue in this action. DBAG submits that the Said Letter of Authority should be read in conjunction with the FXPBA and that reference to the latter can be made as an aid to the construction of the former.
I set out later in this judgment the different types of trade which are in dispute and the expert evidence relating to them and the manner in which Mr Said referred to these trades when discussing them with Mr Vik in email exchanges between them, but it is the EDTs or TPFs upon which attention is focussed because they were such a major contributor to SHI’s losses.
Mr Said’s authority, as set out in the Said Letter of Authority, was to trade on behalf of SHI and extended to the signature and delivery of documentation relating to the execution of “FX and Options Transactions” “including, without limitation, ISDA master agreements, schedules, confirmations, credit support annexes, security interests or other credit support documentation”. Mr Said thus had authority to bind SHI by signing and delivering documents which set out the terms upon which DBAG would extend credit or require margin or collateral. Subject to the terms of such instruments, therefore, he had authority to vary, waive or modify their terms.
Furthermore, the second paragraph of the Letter expressly stated that DBAG was under no duty to enquire as to the nature of the relationship between Mr Said (the Agent) and SHI, nor as to any restrictions upon his activities in connection with the execution of FX and Options transactions which were to bind SHI exactly as if SHI had executed the documents itself, whether those documents were confirmations of individual trades or documentation of a more general nature falling within the ambit of the first paragraph of the Letter.
DBAG was therefore entitled to look to Mr Said and to treat him as SHI for the purpose of the “FX and Options Transactions” subject only to the potential limitation of his authority to trades which fell within that definition. If there was some internal SHI limit upon Mr Said’s authority which differed from the terms of this letter, such a limit was expressly to have no impact upon the liability of SHI in respect of FX and Options Transactions executed by him.
7(c) The nature of FX Prime Brokerage and the Expert Evidence thereon
The issues which arise in this action centre upon Mr Said’s trading for SHI in FX transactions and in particular the EDTs and OCTs which were concluded through the Prime Brokerage of DBAG. The way in which Foreign Exchange Prime Brokerage (FXPB) works was set out by the expert instructed by DBAG in this area, Mr Quinn, in his report. This was largely uncontroversial but, to the extent that there was any dispute about it, I accept his evidence, founded as it was upon years of experience of FXPB and its operation. I found him a careful and thoughtful expert witness and reliable on other areas where there was more significant dispute.
By comparison, Ms Rahl, the expert instructed by SHI, had no direct experience of working on the buy or sell side of FXPB, although she had been consulted on matters where FXPB played a part in the scope of the consultation. As appears elsewhere, I found her change of stance in relation to customs of the market and market practice revealing, inasmuch as she initially failed to draw any distinction between direct
trades to which DBAG and SHI were parties and indirect trades effected by SHI through FXPB arrangements with DBAG, when considering the need for an explanation to be given to someone other than the trader of the risk level of products that he was contemplating. She later “clarified” that she intended always to refer to direct trades only in this context and this led to an amendment of the pleadings by SHI. I came to the conclusion that I could not rely upon her evidence, which was not candid about that change of stance, that her original stance showed a lack of understanding of how FXPB operated and that she argued a case, rather than giving independent expert opinion based on market knowledge.
Mr Quinn described the operation of FXPB in the following way:
“23. FXPB is essentially a clearing function (although others may refer to it as a credit intermediation function) offered by a bank to its clients to facilitate a client's trading activities on an agency basis. The bank permits the client to use the bank's interbank credit lines, enabling the client to trade directly with several executing brokers but consolidating all positions and risk with the bank (which acts as the FX Prime Broker). 24. In summary, the process of FXPB operates as follows:
a client enters into a trade with an executing broker in the name of the FX Prime Broker;
contemporaneously the client enters into an equal and offsetting transaction with the FX Prime Broker (these transactions are known as "give up" trades) and by so doing the bank takes a credit risk on the client in respect of which the client usually posts collateral with the bank;
the service enables the client to post a single pool of collateral with the FX Prime Broker rather than posting collateral with each executing broker with whom the client wishes to trade. This provides a valuable benefit to the client, not just because a single pool is more convenient but, because utilizing a single pool posted with the FX Prime Broker usually allows the client to trade against a lower amount of collateral, while still being in a position to negotiate terms and conditions with a number of different executing brokers. The amount of collateral posted is a matter for negotiation by the client with the FX Prime Broker (and all sorts of factors are relevant to this, including commercial considerations); and
the FXPB service is therefore a highly client demand-driven service. In exchange for the authority to trade in its name, the FX Prime Broker typically charges the client a fee per transaction.
25. The basic steps of the "give-up" process are as follows:
the client selects the FX trade it wishes to enter into. The sales/trading team at the executing broker negotiates the terms and conditions of the FX trade with the client (and may provide advice in this respect) and executes the FX trade on behalf of the client;
after the trade has been executed, the client notifies the FX Prime Broker of the trade details. The client provides this notification using automated or manual systems put in place with the bank;
the executing broker communicates the trade details to the FX Prime Broker;
the FX Prime Broker confirms matching details between the executing broker and the client. If there are any mismatches, the FX Prime Broker advises both the client and the executing broker of those mismatches;
the FX Prime Broker also enters into an equal and offsetting trade with the client so that its net exposure is zero and inputs the back-to-back trades; and
the FX Prime Broker clears and settles the trade on the settlement date.
The diagram below highlights the key steps in an FX Prime Brokerage give-up transaction:
It is apparent from the description I set out above that the FX Prime Broker is not providing any advisory service to the client: the FX Prime Broker does not even know what trades the client has entered into until the client informs it.
Rather, FXPB is a low-profit margin operationally-focused business where revenues are generated by the bank charging transaction fees for trades done with other executing brokers. It is a service business that takes no market risk and differs from FX trading, which is a business that revolves around taking market risk. The service is often offered to enhance the overall FX franchise and increase the amount of trading done with the bank directly. The FXPB department within a bank is operationally-focused group which may have its own sales function and has minimal interaction in its daily operational functions with FX Sales or Trading. FXPB tends to physically sit in a different part of the bank due to confidentiality issues.
It is useful to compare and contrast the functions of the other various separate, but sometimes overlapping, departments involved in the FX businesses for a bank, all of which reported to the head of the FX division:
FX Trading: This is an internal proprietary-focused department, which allows banks to express a view on the FX markets which is translated into market risk in the hope of being profitable on a trade-by-trade basis. This internallyfocused group can generate and execute ideas through its own research team or can participate in the trade flow of clients brought to them by the FX Sales Department, and provide liquidity to a client by enabling it to transact in the product, quantity, and direction it needs. It will make a market on a proposed client trade and if they win the trade, will take on the proprietary risk of that position.
FX Sales: This is a bank-offered service that manages, sources, and communicates with clients about trading opportunities. FX Sales communicates with FX Trading, which sets a price level when a client wants to execute an instrument. This group is responsible for getting the client to transact with the bank, based on its service orientation and ability to work with its FX Trading group to offer competitive pricing and liquidity on FX instruments. This FX Sales function should not be confused with any sales function which exclusively promotes the FXPB product itself.”
In DBAG, the Trade Desk, Execution Desk or Franchise Desk for Complex Options was headed up by Mr Hutchings, and Mr Chin was the relevant trader who entered into EDTs with Mr Said as direct trades. Mr Geisker was the salesman at the Sales Desk, with whom Mr Said would make contact. FXPB was part of a combined group which handled FXPB and FIPB (Fixed Income Prime Brokerage). This was headed
up by Mr Gunewardena as Global Head of FIPB. On the credit side Ms Serafini was answerable to him and below her Mr Spokoyny was answerable to her. On the Product or Business (Sales) side of FXPB, Messrs Quezada and Giery, as Product Specialists in FXPB were responsible to Ms Liau as Chief Operating Officer of FIPB and Head of Product FXPB. She in turn was accountable to Mr Gunewardena. All of these individuals worked in the 60 Wall Street offices of DBAG in New York.
FXPB Operations were based in New Jersey where they were run by Mr Steven Kim, as Global Head of FXPB Operations. Under him in 2008 were two line managers who were responsible for supervising five or six client services representatives (CSRs), including Mr Walsh, the CSR for SHI and, amongst other CSRs, Ms Ng, with whom he shared confidences and the problems with which he was faced. Also under Mr Kim were Ms Ogilvie to whom Mr Manrique was accountable and Ms Greenberg, who unlike the rest of the Operations team was based in New York and was responsible for “onboarding” and transition of clients on to the FXPB systems. Thus, apart from Ms Greenberg, there was physical separation between FXPB Operations in New Jersey and the FXPB Sales (Product) team in New York, as well as the “Chinese wall” between FXPB generally and DBAG’s Trading/ Sales/ Franchise desk.
Mr Quinn’s report continued thus:
“30. Further, and as I explain below, there was no one-sizefits-all FXPB service. On the contrary, in the relevant period from November 2006 to November 2008, the FXPB service offered by banks acting as prime brokers varied from bank to bank and from client to client. This is because the FXPB service offered by a bank to a client would differ from client to client and take into account the client's characteristics and set up.
FXPB was first conceived in 1994. A convenient summary of the evolution of the FXPB market is set out in a document that was presented to institutions in 2005 by the Federal Reserve (the Product Overview and Best Practice Recommendations) as follows:
“Prime brokerage emerged in the early 1990s with the use of semi-formalized “give-up” arrangements initiated by a few financial institutions. The product gained momentum in the late 1990s when several banks entered the prime brokerage business with dedicated market and sales efforts as well as tighter and more formal operational controls, procedures and processes. This focus laid the foundation for a rapid expansion of the client base.”
FXPB, as it is now known, was commonly called "FX Clearing" at this point. It was only in the 2000's that it became commonly known as FXPB.
By 1998, there were three institutions that offered the
FXPB/FX Clearing product as a designated offering: ABNAMRO, AIG and Deutsche Bank/Bankers Trust. These institutions offered a dedicated FXPB team (separate from FX Trading and Sales), a senior FXPB relationship point, and carved out "give-up" lines that were exclusive for the client's use. Every bank has a line of credit with other banks. The FX Prime Broker would use a portion of these credit lines between the banks for the client's exclusive use to trade FX transactions.
By 2004, the number of FXPB service providers had grown to more than 20 institutions, including banks, investment banks, insurance companies, and niche broker-dealers, which provided some form of FXPB offering to more than 500 clients. The product had evolved from simple FX spot, FX forwards, and short dated swaps "give-ups" to include more complex transactions.
By 2008, more than 25 entities offered some sort of FXPB. This list included US and European commercial banks, investment banks, insurance companies, retail FX firms, small broker dealers and some corporate entities. Every firm that offered FXPB seemed to have a different product offering. The nature of their offering correlated with the firm's core client base, technology and operational robustness, research focus and overall credit rating. The main users of FXPB services were hedge funds that actively traded FX as an asset class for their portfolio. Other users of FXPB emerged over the period 19942008 and typically included small commercial banks, small broker-dealers, high frequency traders, European and Asian private banks, large and small corporations and retail-focused web-based FX trading firms.
Although I am not a compliance expert, I had a broad understanding of the regulatory framework in New York. My experience was that as more banks offered FXPB, the Federal Reserve and the Office of the Comptroller of the Currency started taking a more proactive interest by asking more questions about the business in FXPB starting in early 2001. These regulators were interested as FXPB was situated within the banking (rather than securities or broker dealer) entities of banks.
In or about 2005, the Product Overview and Best Practice Recommendations were presented by the Federal Reserve addressing the role and structure of FXPB. This overview did not constitute binding or enforceable guidelines. Rather, the document emphasized that the recommended best practices "were intended as goals rather than binding rules".
Other regulatory bodies like FINRA, the SEC, the NYSE and the CFTC had no direct authority over establishing or enforcing regulations with respect to the FXPB product specifically, or licensing professionals involved in offering the product, and I never came across such regulators in my FXPB practice.
Given the absence of formal regulation or binding guidelines, there was substantial variance in arrangements which clients negotiated with the various banks individually, which led to big differences in the operational processes of the participants involved.
However by 2006 there were some key facets to executing "give-up" transactions with FX Prime Brokers, which had similar characteristics or practices across industry participants. Some of these included:
Use of Client Documentation: Prior to trading through an FX Prime Broker, the client would typically sign a prime brokerage agreement outlining the terms and legal structure of the service, which may include a list of approved "give-up brokers" and associated financial limits with each broker, and other items like permitted currencies. Other documents that were customarily used included an ISDA Master Agreement (ISDA) and Credit Support Annex (CSA). While documentation was put into place to set out the terms of the relationship with an FX Prime Broker, the actual content and terms of these documents varied greatly from bank to bank and client to client, reflecting substantial variation in the business model used by each bank and applied to each client.
Use of Bank to Bank Documentation: The FX Prime Broker negotiated a "give-up" agreement with the executing brokers [referred to in this action usually as a Counterparty Agreement].
Formalized access to specific banks' credit lines by each FXPB client: The FX Prime Broker allowed the client to use its interbank lines of credit so that it could transact contracts with numerous executing brokers as specified in the "give-up" agreement. Very often, an FXPB arrangement would require the client to post collateral with the FX Prime Broker as a safeguard against the total risk of the client's trading positions, for the bank's protection.
Trade confirmation timing: It was market practice for there to be a period of time in which an executing bank would confirm a trade with the FX Prime Broker for a transaction. The FX Prime Broker would not be regarded as having a trade until both parties had reported the trade. After the suggested period had elapsed, it was common practice for an FX Prime Broker to reach out to the party that had not reported for confirmation. If neither party had reported the trade, it was impossible for an FX Prime Broker to know a trade had been executed. Reporting methods included: Reuters 3000 Machine (FX chat and communication system), fax, phone, File Transfer Protocol file (FTP) or email. Some banks who were on the Triana platform offered this system to their clients. By 2006, it was commonplace to receive an FTP file that was straight through processed from the client into the bank's internal FX booking systems for all FX spots, vanilla options and forwards. These FX spots, vanilla options and forwards generally represented more than 90 per cent of all transactions done by clients and, due to the degree of automation, the FXPB team had minimal manual involvement.
Clearing and settlement functions: The FX Prime Broker would be solely responsible for matching, reconciling, and inputting the trades into the bank's internal systems. Matching a trade is meant to ensure that the client and the executing broker have exactly the same trade details. If a trade is matched properly, the FX Prime Broker then inputs an equal and opposite transaction between itself and the client to ensure the FX Prime Broker ends up with a flat position. While the FX Prime Broker takes credit risk by the provision of a credit line which gives the clients access to the FX markets, the ending net position for the FX Prime Broker is flat, reflecting the complete offset between the client trade and give-up received. If a trade was matched it would have then been settled in accordance with its terms.
Reporting: Typically, some basic reporting of the trades executed would be provided by the FX Prime Broker to the client on a daily basis. The type, method and detail of reporting would vary between institutions and clients. As I have already explained, the FXPB service provided would differ from client to client and that was particularly true of any reporting arrangements made. Any reporting in the context of an FXPB service was primarily to provide a record of the trades the client had selected and executed in the name of the FX Prime Broker.
Access to client information: Customarily, FXPB staff sat on a different floor and used different systems (with restricted access for FXPB staff) to the FX sales/trading department. This ensured a separation of duties for operational control, and confidentiality to preserve the integrity of client information (so that only FXPB had access to clients' position information). The effect of this was that sales and traders had no access to clients' position information, aside from what was traded directly with the bank.
Client interaction: FXPB professionals were solely responsible for the client-facing operational aspects of the FX client relationship. FXPB would not customarily be involved when clients sought advice or research from FX sales coverage.
Even though there was some consistency with regard to the broad roles and functions of FXPB providers, each FXPB facility was typically very different from the next. The lack of regulation and any specific regulatory guidance resulted in substantial variances in the FXPB service provided to different clients by different banks.
Most FXPB facilities would take into account a client's existing operational and trading infrastructure. One client might do their "forward rolls" (which is taking a currency position from a near date to a date further in the future) at the end of the day, another client might do them at 12pm EST and another might do them trade by trade.
Operationally, the process conducted by FXPB for a client was not standardized due to the diversity of each client's trading strategies. One client might have preferred to receive trade confirmations via fax, another by telephone, another by website, and another by FTP file.
In addition, the size of a client's operation and the client's level of sophistication and customization needs typically informed the FXPB process. For example, larger clients who had a technology infrastructure and strong operational support were more streamlined tended to use automated processes to interact with their FX Prime Broker. Smaller clients with fewer resources tended to use more manual processes, both in operations processes internally and how they interacted with the FX Prime Broker. However the provision of the Triana system by some FXPBs enabled clients to automate the process for certain types of vanilla trades.
A broad range of documents for creating an FXPB facility were or could be used. These documents were the sole source of governance of the relationship between the FX Prime Broker and the client, in the absence of any specific regulations that were applicable to FXPB. Furthermore, there was no Exchange/Central Clearing Counterparty or regulatory margin policies that existed for FXPB. The use of specific documents was a function of a bank's policies in addition to both the type of client and the nature of their proposed trading activities. Those documents could include (though these were not necessarily always used) an FXPB agreement, ISDA, CSA, give-up documents, Collateral Agreement, and Trader authorization list.
Since the actual provisions in these documents were not required by the terms of an exchange (as was the case in futures markets), there was great flexibility in the language and customization of these bilateral documents. The ISDA and give-up documents were fairly industry standard, but these and the other documents were always subject to negotiations based on the bank's unique credit or risk characteristics and the client's priorities, needs, and negotiating power.”
It should be noted, therefore, that Mr Said concluded FX transactions for SHI by initiating contact with the Sales/Trading team at other banks such as Credit Suisse (CS), Goldman Sachs (GS) and MS or with DBAG’s own Sales/Trade Desk. Having agreed the terms of a transaction, the terms of that deal would be notified to DBAG’s FXPB department by Mr Said and by the other bank with whom DBAG had a Counterparty Agreement. That trade took effect as a transaction between the other counterparty bank and DBAG, through the agency of Mr Said and, subject to matching of the details provided, an offsetting trade would then come into effect between DBAG and SHI (the Give-up or Agency Transaction). This structure would commonly be referred to as an Indirect Trade. Where Mr Said, for SHI, concluded a trade with DBAG’s Trade Desk, however, the structure was necessarily different. Self-evidently DBAG’s Trade Desk could not enter into a contract with DBAG’s FXPB desk as they were each part of the same legal entity but the mechanics of the transaction operated in the same way. This was referred to as a Direct Trade, but was treated by FXPB as if there were two distinct legs to that trade.
The difference in function between the Trade Desk and the FXPB desk has significance in the context of some of the allegations made by SHI. FXPB was a high volume, low cost clearing arrangement whereby SHI gained access to DBAG’s credit status in order to effect its own transactions with the benefit of putting up such collateral as DBAG demanded to DBAG alone, as opposed to having put up separate collateral to each of the banks with whom it had negotiated the trades in question. According to Mr Quinn, the FX prime broker was not understood to play and did not play any role in assessing or advising on the market risk or possible gains or losses from the transactions that the customer concluded. Ordinarily, the FX prime broker would only know about the transaction after it had been agreed by the customer and the Counterparty, on receipt of notification from the customer and Counterparty, whether by automated means, electronically or otherwise. The customer, acting as agent for the Prime Broker in concluding the agreement with the Counterparty, instigated the trade and notified the Prime Broker of it with the offsetting trade between the Prime Broker and the Agent springing up without anything in the nature of a trading decision on the part of the Prime Broker itself and without any opportunity for any discussion of the trade or the risk involved in it between the Prime Broker and the Customer.
In the case of Structured Options concluded under the FXPBA, prior approval had to be obtained by SHI from DBAG before they could be Accepted Transactions falling within the ambit of the FXPBA but the individual treatment of these items cannot change the basic nature of the relationship between the Prime Broker and the customer. The FXPBA envisaged most of the trades being effected in the ordinary way, by automated entry by Mr Said or by manual inputting of information supplied
by Mr Said, with matching of this by DBAG against the details provided by the Counterparty.
Mr Quinn expressed his understanding of the role of the FX Prime Broker in this way:
“51. The role of an FX Prime Broker was understood in the market to be to facilitate the clearing of client-traded products which were approved in the documentation. The FX Prime Broker was not understood to play, and in my experience did not play, any role in assessing or advising on the market risk or possible gains or losses from those transactions. Indeed, customarily the FX Prime Broker would only know about the transaction after it had been executed by the client. …
52. As providers of a clearing service whose primary function is to match and settle trades, it was not customary for FX Prime Brokers to offer risk management services to clients, even though I became aware some FX Prime Brokers offered limited portfolio metrics. By "portfolio metrics" I mean data such as volatility estimates about the underlying currencies in a client's portfolio. If an FX Prime Broker did offer any limited portfolio metrics, those would only be to aid the client's own risk management and not intended as a risk management service. Some FXPB clients outsourced their risk management
(although many utilized their own proprietary risk systems) and there were a number of outside vendors who offered third party risk management software, such as Risk Metrics, SunGard, Algorithmics and others. The provision of risk management was not, however, customarily a service provided by FX Prime Brokers even as an add-on feature available for an additional fee.”
The expert appointed by SHI, Ms Rahl, was in broad agreement with this description but in cross-examination said that risk reporting obligations could be undertaken by a Prime Broker and, where they were, they should be fulfilled. She considered that the position of a bank which entered into direct trades or incorporated a PWM department should be seen differently from the position of a stand alone FX Prime Broker. Ms Rahl also considered that a bank would be subject to obligations in respect of margining. She was prepared to accept that the function of margin was primarily to provide credit protection for the FX Prime Broker but did not accept that it could not also represent a calculation to be used by the client as an assessment of the market risk of the positions it had cleared through the FXPB. Mr Quinn did not see any direct relationship between risk reporting by the Prime Broker, even if that was undertaken as an obligation, and calculation of margin since they were distinct functions effected by the FX Prime Broker for different purposes. The calculation of margin was for the Prime Broker’s protection whilst risk reporting was for the client’s purposes. Ms Rahl disagreed on the basis that if margin was “risk based”, which she took to be the case with VaR, it could be used as a risk assessment tool. It was agreed between them that there was no regulatory standard in respect of determining the amount of margin for FX OCT derivatives.
In her first report, Ms Rahl gave her opinion that it was at all material times the custom and practice of a reasonably prudent and competent investment bank to have a policy or policies regarding high risk derivative transactions to require that the bank explain the risks of the instrument to some person at the client who was not the immediate trading contact – i.e. a supervisor or a Chief Risk Officer. That second person could be more senior than the trading contact or could be independent of the trading contact. This evidence was adduced in support of a plea by SHI that, under New York law, on the true construction of the FXPB Agreement, in the light of factual background and/or as an implied term, it was a term of the FXPBA that the bank would ensure that, in circumstances when Mr Said proposed to enter into a high risk product, particularly a leveraged derivative (such as an EDT), an accurate explanation of the risk level of the product would be given to Mr Vik; alternatively the bank would take reasonable care to ensure that in such circumstances an accurate explanation of the risk level of the product was given to Mr Vik. The plea related specifically to the FXPBA and Ms Rahl’s first report drew no distinction at all between direct trades with DBAG and indirect trades with other Counterparties. Following the sequential service of Mr Quinn’s report, Ms Rahl responded by saying that the custom and practice to which she had referred in her first report regarding high risk products applied only to direct trades and SHI amended its case accordingly in Further Information served on 1st February 2013. Ms Rahl agreed with Mr Quinn in the Joint Expert Memorandum of 31st January 2013 that the FX Prime Broker was not providing any advisory service to the client and that it was not customary for an FX Prime Broker to seek assurances that clients had adequately considered the market risks of the transactions which they sought to clear with the FX Prime Broker on “stand alone” trades as opposed to direct trades.
Her evidence was, however, that she considered that there was a subset of transactions that were so unusual and risky that they would, even if permitted under the terms of a Prime Brokerage Agreement, require FXPB approval. She maintained this in crossexamination but there seemed to be no logical basis for it at all save for the contractual provision in the FXPBA in relation to Structured Options. It was a feature of Ms Rahl’s report that, instead of giving expert evidence on the market, she sought to refer to the specific contractual provisions and the facts surrounding the DBAG/SHI relationship, as she saw them.
It was common ground between the experts that it was standard market practice for transactions to be booked accurately by the Prime Broker on the day of trading, although it was not uncommon in Mr Quinn’s view to have trades which were not booked on the same day that they were traded by reason of the complexity that some trades involved. Some required manual inputting with a careful review of the terms of the contract. Ms Rahl considered that although there was no uniform identifiable set of standards which would be customarily applied and although the FX market was self-regulated, clients could reasonably expect accurate and timely booking and confirmation of trades, reporting and margining, irrespective of the systems actually used by the Prime Broker. Although she agreed that a trading professional would generally maintain his own records, that did not change her opinion that the Prime Broker had an obligation to keep accurate records and there was no reason why a trading professional should not be able to rely on an FX Prime Broker to assist him with his records. She considered that the accurate booking, valuing and recording of transactions was the minimum standard of service and the very basis of the prime brokerage business, whilst Mr Quinn did not consider these were core services offered by an FX Prime Broker to its client. The services that were offered would of course depend upon the contract between the parties and no analogy with bank “checking account statements” was appropriate.
Regulatory material relied upon by Ms Rahl was all for the purposes of audit of the bank by the examiners of the Office of Comptroller of Currency (OCC), seeking to ensure that banks were run properly with a view to avoiding counterparty risk for the bank’s own protection. In my judgment, as Mr Quinn said, the only relevant guidance appeared in the Federal Reserve Foreign Exchange Prime Brokerage Product Overview and Best Practice Recommendations of 2005. These were not binding or enforceable. The core services offered by a prime broker to a client were there said to be “leverage, access to market, liquidity, and consolidated settlements, clearing and reporting”. This, on Mr Quinn’s evidence, corresponded with his view as to the Prime Broker’s functions, namely that of allowing the client to use the Prime Broker’s credit lines and the clearing and settlement of trades. The reporting to the client might take many formats but was to enable the client to reconcile its own records with those of the prime broker. There was no custom or practice which applied in respect of the reporting, save basic reporting of this kind. Booking a trade in the prime broker’s books was done for its own internal management purposes, namely management of risk and exposure. Any entity would seek to do this accurately for its own benefit but how it chose to do so was entirely a matter for itself. Valuation was not part of any core service provided by an FX Prime Broker and, typically, traders did their own valuation internally with the aid of a third party system, or by employing an administrator who would be responsible for managing valuations and striking monthly Net Asset Values if that was required. It was the custom and practice for FXPB clients to keep their own valuations.
Mr Quinn’s opinion was that there was no formalised industry standard regarding booking and recording of FX transactions and that in the 2006/2008 timeframe there were many new products which were inevitably going to be difficult to fit into existing systems of booking. If the trade fell within the ambit of the Prime Brokerage Agreement, the Prime Broker had no option but to take the trades in and to book and record them as best it could. Similarly, valuation and margining were arguably the most challenging aspects of FXPB throughout 2008, especially for complex structured FX options. There was no standard market practice in valuation of the products, nor in the imposition of margin based thereon. It was in his view customarily understood in the market that the function of margin was to provide credit protection for the FX Prime Broker and was not to be a calculation for the client of the market risk of the positions it had cleared through that Broker. There was no regulation governing specific collateral levels and clients therefore appreciated that the FX Prime Broker might be flexible about margin. There was intense pressure between 2005 and 2008 between banks, in seeking to match or better competing banks’ credit/margin terms in order to win new business. There were also situations in which an FX Prime Broker might choose not to call for all the margin to which it was entitled. It was customary for clients to have their own risk system, or to utilise the services of a third party technology provider to evaluate their overall exposure to the market. Where the Prime Broker did provide trading information, there was dialogue between the Prime Broker and the customer to reconcile mismatch errors but not to reconcile risk evaluation. Mr Quinn was not aware of any FX Prime Brokers who offered any form
of comprehensive risk exposure or metrics reporting during November 2006 to November 2008.
In cross-examination Ms Rahl stated that, for direct trades, there was clearly an obligation on an investment bank to have proper records. She said that for “give-up trades” she had not formed an opinion. She went on to say however that FXPB reporting was not simply to provide a record of trades the client had selected and executed in the name of Prime Broker and that she considered that whatever the client received by way of report should be accurate and whatever reports were provided were the services upon which the parties had agreed.
So far as MTM and exposure were concerned, she considered that all that mattered to the client was what the prime broker thought about it for collateral purposes. The client was reliant upon that reporting. She agreed that valuation was not a core service but stated that it was part of leverage margining and that leverage for a collateralised product assumed an ability on the part of the Prime Broker to calculate margin. Valuation was therefore not just for the Prime Broker’s own purposes. It was a regulatory requirement for the calculation of margin. She agreed that margin was sought by FX Prime Brokers to protect themselves against the risk of client default and came forward with no other reason for it. She agreed it was not put in place to protect the client from incurring loss but to give the Prime Broker credit protection. She agreed with Mr Quinn that the provision of collateral was conventional and that margin was for the protection of the bank and not a service to the client. They both agreed that there were no regulatory standards for the manner in which margin was to be charged in FXPB.
In my judgment it is clear that the core services of a Prime Broker are as Mr Quinn described them. It is also clear that a Prime Broker, when conducting valuations of trades, does so for its own benefit and protection, so that it can charge margin in relation to those trades. In the ordinary way, a trader manages its own risk by reference to its own valuations and there is no obligation on the Prime Broker to calculate or call for margin, since any entitlement to do so is provided for the bank’s own benefit and not for that of the customer. Indeed the customer’s own interests are usually directly contrary to those of the Prime Broker, since the customer will wish to minimise the amount of margin to be put up in order to obtain the greatest amount of leverage and the greatest scope for trading. The standard position when margin is being negotiated is for the client to seek to restrict the figure to as low a level as the Prime Broker will wear.
Of course there is scope for agreement to the provision of additional services, over and above those which obtain in any Prime Brokerage relationship. The parties can expressly agree to specific forms of reporting, whether of the trades done, the valuation of those trades or the calculation of margin in relation to them. That is a matter of freedom of contract. In the absence of specific agreement to a broader scope however, I accept Mr Quinn’s evidence that, as a matter of market practice, FX Prime Brokers’ duties were to provide basic information for reconciliation of trades and positions and that traders who trade institutionally in the market know their own positions, track them and risk manage them. If a service is contractually agreed, it should however be performed.
Ms Rahl accepted that the FX Prime Broker was not obliged to call for margin but suggested that it had to communicate the fact if it was not calling for it. She said she did not know if the bank could call for less collateral than that to which it was entitled but said that it was unacceptable to make partial calls without saying what it was that was being done. She maintained that it was standard practice in margining that the client would be informed of the total collateral entitlement, the extent of any demand by reference thereto and the reasons for not calling for collateral in full and the circumstances in which further demand for it might be made. She considered it misleading not to call for collateral when collateral was due and, although a bank could decide to waive its entitlement, it ought to inform the client that it was so doing.
All of this was said to be based upon what Ms Rahl had seen done in the past but she could not remember if that was in an FXPB context or not. She considered it standard practice in margining and that there should be no distinction for FXPB. I was unable to accept any of this evidence as giving rise to a market practice which could affect contractual rights. Ms Rahl had no relevant experience of the FXPB market to bring to bear on the subject and once it is accepted that any provisions for margining are for the benefit and protection of the Prime Broker, and not the customer, it is hard to see how there can be any applicable custom which could give rise to an obligation affecting the exercise of that entitlement.
All in all I am unable to find that there is any sufficiently well known custom or market usage which could impact upon the contractual obligations of the parties as set out in the FXPBA which is governed by New York law, nor the other 2006 Agreements which are governed by English and Swiss law. The question whether or not DBAG agreed to provide additional services, beyond the core services inherent in FX Prime Brokerage depends upon the proper construction of the FXPBA, in the light of the nature of FX Prime Brokerage and the surrounding circumstances known to the parties and/or upon the implication of terms into the FXPBA in accordance with the relevant principles of New York applicable to the construction of contracts and the implication of terms therein.
7(d) The Foreign Exchange Prime Brokerage Agreement (the “FXPBA”)
The introductory paragraph to this Agreement sets out its purpose in clear terms:
“This Agreement describes the arrangement pursuant to which Deutsche Bank AG London ("DBAG") authorizes Sebastian Holdings, Inc. ("Agent"), to act as its agent in executing spot, tom next, deliverable and non-deliverable, forward foreign exchange transactions with a maximum tenor of 24 months ("FX Transactions"), gold, silver, platinum and palladium ("Precious Metals") spot and forward transactions with a maximum tenor of 24 months which provide for settlement without physical delivery of metal ("Precious Metals Transactions"), and currency and Precious Metals options with a maximum tenor of 24 months and which provide, in the case of Precious Metals options, for settlement without physical delivery of metal ("Options") (collectively, the "Counterparty Transactions") with the Counterparties listed in Annex A hereto (each, a "Counterparty") and on the terms set forth in Annex B hereto. Capitalized terms not defined herein shall have the meanings assigned to them in the 1998 FX and Currency Option Definitions (as published by the International Swaps and Derivatives Association, Inc., the Emerging Markets Traders Association and The Foreign Exchange Committee).”
The FXPBA operates to confer authority on SHI to commit DBAG to transactions with identified Counterparties, subject to its terms. The original named Counterparties in Annex A were Citibank, N.A., Credit Suisse First Boston, Goldman
Sachs International and Den Norske Bank although additional Counterparties, such as MS, Société Generale and Lehman Bros, were later added. SHI was thus enabled to act as DBAG’s agent in concluding contracts (“Counterparty Agreements”) between DBAG and the Counterparties in accordance with the terms of the FXPBA. This SHI did, concluding transactions with the named Counterparties, of which it would then inform DBAG (“Counterparty Transactions”). Under the terms of the FXPBA, DBAG was contemporaneously to enter into an equal and offsetting transaction with SHI (an “Agent Transaction”). Although there was a provision which would have allowed DBAG to “give up” such transactions to approved third parties, there were no approved third parties for this purpose, with the result that each offsetting transaction was always between SHI and DBAG as an Agent Transaction.
The FXPBA contained express limits on the authority thus given to SHI to commit DBAG to these Counterparty Transactions and thus to the offsetting transactions between SHI and itself. The introductory paragraph, set out above, referred to the types of authorised transactions, which, so far as is material, included “FX Transactions” as defined and “Options” as defined (including specifically currency options with a maximum tenor of 24 months).
Additionally, paragraph 1 of the FXPBA provided as follows:
“1. This authority is expressly limited for each Counterparty in that (a) for any Settlement Date the Net Daily Settlement Amount for such Counterparty may not exceed the Settlement Limit as specified in Annex A hereto and (b) the Counterparty Net Open Position may not exceed at any time the Maximum Counterparty Net Open Position as specified in Annex A hereto. The Settlement Limit and the Maximum Counterparty Net Open Position shall apply to all Counterparty Transactions entered into between DBAG and the Counterparty branch specified in Annex B.”
There was no Counterparty Net Open Position specified for any Counterparty in Annex A but at a later stage limits were set in respect of GS and Société Generale. In Annex A, the Net Daily Settlement Amount for each Counterparty was specified as US$200 million and the Maximum Counterparty Net Open Position limit was set later and then increased in May 2008 to US$600m and US$800m respectively for those two Counterparties. There was thus an express agreed restriction to the trading of SHI to these amounts on a daily basis for the net amount owing to DBAG by a Counterparty for any settlement date. The Net Daily Settlement Amount was defined as the net amount owing to DBAG by a Counterparty (excluding options prior to their exercise) on any Settlement Date. The Counterparty Net Open Position was defined
as the “aggregate amount owed by the Counterparty to DBAG” calculated by reference to the MTM of each FX or Precious Metals Transaction and the Dollar Countervalue of the delta equivalent of each sold or bought option.
Clause 1 went on to refer to “Netted Options” and “Structured Options”. Netted Options were (excluding Structured Options), options sold by DBAG and owned by the Counterparty which could be discharged and terminated together with an Option (other than a Structured Option) sold by the Counterparty and owned by DBAG, subject to certain criteria (e.g. being of the same currencies, duration and style and having the same strike prices). Structured Options were excluded from the calculation of the Net Open Position. They were defined thus:
“"Structured Option" means any option other than one which is a (i) put or call that does not have special features, or (ii) single barrier option.”
Clause 2 provided that SHI should monitor the Net Daily Settlement Amount and the Counterparty Net Open Position for each Counterparty (although there was no Counterparty Net Open Position originally specified), doubtless with a view to ensuring that it (SHI) incurred no breach of the limits set out by the FXPBA. DBAG submits that its agreement to provide “a summary of the outstanding trades and the net exposure with respect to each Counterparty up to two times on each Business Day during which there are Counterparty Transactions outstanding” must be read in this context. The full provisions of Clause 2 appear later in this judgment, when considering what obligations DBAG owed SHI under the FXPBA.
Clause 3 provided as follows:
“Prior to entering into any Counterparty Transactions, DBAG shall have executed a Counterparty Agreement with such Counterparty. Agent [i.e. SHI] shall promptly communicate trade details of each Counterparty Transaction by notifying via facsimile or other electronic means an area of DBAG separate from trading and marketing personnel. Each Counterparty Transaction between DBAG and a Counterparty shall be confirmed and settled in accordance with the terms of the applicable master agreement between DBAG and the
Counterparty (a "Counterparty Master Agreement").”
DBAG had concluded Counterparty Agreements with the Counterparties set out in Annex A and with later additional Counterparties included by agreement with SHI. It was for SHI to notify DBAG of the transactions it concluded with such Counterparties in the name of DBAG which would then be confirmed and cleared in accordance with the terms agreed between DBAG and the relevant Counterparty. It was Clause 4 that then provided for the corresponding offsetting transaction between DBAG and SHI, in the following terms:
“4. In connection with entering into each Counterparty Transaction, DBAG shall contemporaneously therewith enter into an equal and offsetting transaction or transactions with, at the discretion of Agent (i) Agent (each, an "Agent Transaction") or (ii) one or more of the give up parties listed in Annex C hereto …”
As previously mentioned there were no such parties specified in Annex C so that every Counterparty transaction concluded by SHI gave rise to an Agent transaction between DBAG and SHI.
Clause 4 went on to provide:
“Each Agent Transaction shall be an FX transaction precious metals transaction or option under, and subject to and governed by, the applicable ISDA Master Agreement or other master agreement between Deutsche Bank AG and the Agent, including the Credit Support Annex which is a part thereof (the "Agent Master Agreement''). Agent shall be required to post collateral with respect to its obligations under the Agent Master Agreement (including the Agent Transactions) in accordance with terms and provisions of the Credit Support Annex. DBAG and Agent agree that any breach of this Agreement by Agent shall constitute an Event of Default under the Agent Master Agreement.
….
Each Agent Transaction and each Give Up Transaction shall be subject to and settled in accordance with any market practice … applicable to, or adopted by, DBAG and the Counterparty in connection with the Counterparty Transaction for which it is offsetting notwithstanding any provision in a confirmation for an Agent Transaction or Give Up Transaction that may be to the contrary.”
Thus, the effect of the conclusion by SHI of a Counterparty Transaction was to bring into being a contract between DBAG and the Counterparty which, in accordance with the Counterparty Agreements, meant on ISDA terms (as modified by agreement) with margin and collateral requirements as agreed. The offsetting Agent Transaction between SHI and DBAG to which DBAG and SHI were thereby committed was also to be governed by ISDA terms, in accordance with the ISDA Master Agreement between them, including the Credit Support Annex (the “CSA”) which provided for the posting of collateral with respect to SHI’s obligations. Any breach of the FXPBA by SHI, but not by DBAG, was to constitute an Event of Default under the Agent Master Agreement.
The issues which arise in relation to the FXPBA concern the meaning of the words “FX Transactions”, “currency options” and “Structured Options” in the context of the authority initially given to Mr Said to conclude transactions on behalf of both DBAG with the Counterparty and SHI with DBAG. The primary instrument by which SHI gave authority to Mr Said is the Said Letter of Authority which uses the same expression “currency options”, but makes no reference to Structured Options. It is on that document and those words that SHI’s arguments centre, since DBAG has settled all the Counterparty Transactions with the Counterparty banks on the footing that all
the transactions which were concluded by Mr Said with such Counterparty banks were authorised by DBAG and binding upon it. DBAG has thus accepted that Mr Said had authority to bind it to the Counterparties under the FXPBA. It is SHI which contends that he had no authority to bind SHI to DBAG in respect of the EDTs and OCTs.
The other central issue which arises is the existence or non-existence of duties imposed upon DBAG by the FXPBA with regard to the booking, recording, clearing and settlement of the transactions and in particular the reporting to SHI of the MTM of the trades and its own margin and collateral requirements in respect of them, in circumstances where the FX ISDA agreement and the CSA entitled DBAG to calculate such margin and demand Credit Support in respect of the total portfolio on a net basis.
It is in this context that Clause 2 of the FXPBA falls to be considered in the context of the FXPBA as a whole and the market evidence as to the role and function of a Prime Broker. The FXPBA does not expressly oblige DBAG to book transactions, to record them, to calculate MTM valuations, exposure or margin or to report any of these to SHI, save as set out in Clause 2, which, in its entirety reads as follows:
“2. Agent acknowledges and agrees that it shall monitor the Net Daily Settlement Amount and the Counterparty Net Open Position for each Counterparty and that DBAG shall not be responsible for any Counterparty Transaction executed by Agent on behalf of DBAG unless (i) giving effect to such Counterparty Transaction does not cause the Settlement Limit or the Maximum Counterparty Net Open Position to be exceeded (without DBAG's prior written consent or recorded verbal consent (confirmed by fax immediately thereafter)); (ii) such Counterparty Transaction meets the criteria set forth in Annex B and (iii) if such Counterparty Transaction is a Structured Option, DBAG shall have approved the particular Structured Option transaction proposed by Agent, including the Counterparty and principal amount of such Structured Option, and such approval shall have been effective when the Structured Option was executed by Agent (an "Accepted Transaction"). DBAG agrees to provide Agent with a summary of the outstanding trades and the net exposure with respect to each Counterparty, up to two times on each Business Day during which there are Counterparty Transactions outstanding. Each Accepted Transaction shall be valid and binding upon DBAG, enforceable against DBAG in accordance with its terms. The dealing arrangement with respect to each Counterparty shall be set forth in a Foreign Exchange Prime Brokerage Counterparty Agreement (a "Counterparty
Agreement"). Each such Counterparty Agreement is substantially in the form of a template which DBAG will send you on your request.”
It can be seen that “DBAG agrees to provide” SHI with a summary of the outstanding trades and the net exposure with respect to each Counterparty up to two times on each Business Day during which there are Counterparty Transactions outstanding. DBAG submits that the purpose of this provision is plain, because it appears in the context of a clause which requires SHI to monitor the Net Daily Settlement Amount and the Counterparty Net Open Position for each Counterparty in order to ensure that the contractual limits are not exceeded. The contents of the “summary” are not specified, but bearing in mind that purpose, must at least include the date of each trade, some identifying description of it, the identity of the relevant counterparty and the sums involved in the transaction so that settlement figures and Counterparty Net Open Position figures can be monitored against the limits specified - the Settlement Limit and the Maximum Counterparty Net Open Position.
Whereas the Settlement Limit refers to sums owing by a Counterparty on any Settlement Date, the Counterparty Net Open Position is to be calculated in accordance with a series of provisions set out in Clause 1 (A) to (E), in order to assess the “aggregate amount owed by [each] Counterparty to DBAG”. This is to be done by summing the respective short and long positions on each currency in all the open individual FX Transactions between DBAG and the Counterparty (concluded by SHI), expressed in US$ and aggregating the result with the US$ Countervalue of the delta equivalent of Options purchased or sold by the Counterparty, taking into account Netted Options.
The “net exposure with respect to each Counterparty” to which Clause 2 refers must mean the amount owing by such Counterparty in relation to these two different limits, with their different foci and methods of calculation. DBAG’s agreement to provide the figures for net exposure is clearly for the purpose of enabling SHI to perform its monitoring obligation in circumstances where the effect of not doing so could be to vitiate the authority given by DBAG to SHI to conclude Counterparty Transactions. Clause 2 itself provides that DBAG is not to be responsible for any Counterparty Transactions executed by SHI if giving effect to such transactions causes the Settlement Limit or the Maximum Counterparty Net Open Position to be exceeded without its prior written consent or recorded verbal consent confirmed by fax immediately afterwards.
DBAG’s agreement was to provide the information “up to two times on each Business Day during which there are Counterparty Transactions outstanding”. The phrase “up to two times” appears to provide a degree of latitude in the obligation. DBAG contends that it is to be construed as a provision giving a choice to SHI to ask for such information, but with DBAG only required to provide it no more than twice. If no request is made, then there is no obligation to provide the information. SHI contends that it provides an obligation on DBAG to provide the information once a day, and, at SHI’s option, a second time, if so asked. It does not appear that SHI in fact ever did ask for such information, whether because Mr Said never felt the need to do so, being confident of being well within the limits in question, or because he did not want to raise the issue. Where there was no specified Maximum Counterparty Net Open Position, there would be no reason to ask and, as will be noted, Structured Options fell outside this regime in any event.
In my judgment, since the purpose of the provision of such information was, as DBAG submits, to inform SHI so it could monitor its trading against the limits, so that it was not at risk of exceeding them and running the risk of concluding trades in the name of DBAG which DBAG would not acknowledge, the Clause must be read as requiring DBAG to produce such information once or twice a day if and as requested by SHI. If there was no request, there was no obligation to provide the information.
It will be noted that the information which DBAG agrees to provide is “a summary of the outstanding trades and the net exposure with respect to each Counterparty”. The information to be given relates to outstanding Counterparty Transactions and to realised profit and loss (unpaid cash settlement figures between DBAG and the Counterparty banks) on such trades and to unrealised profits and losses (in the shape of MTM on those trades between DBAG and the Counterparty banks). The latter would appear to correspond to the margin requirement between DBAG and the Counterparty banks – essentially MTM on the trades between them (variation margin), without any initial margin, but it does not relate to margin calculations or margin requirements as between DBAG and SHI. Those are to be found in the FX ISDA and the Schedule and CSA thereto.
The figures thus to be produced by DBAG, if asked by SHI, represent global figures in respect of the totality of outstanding Counterparty Transactions, per Counterparty, rather than individual amounts for each trade. The figures do not represent SHI’s total portfolio upon which margin could be charged under the FX ISDA, the Schedule and the CSA, with both initial margin and variation margin. The whole of Clause 2 relates to the position between DBAG and its Counterparties and the effect of SHI’s actions when concluding transactions, as between DBAG and the Counterparties.
Nor is there any reference here to transactions between SHI and DBAG directly without the involvement of any Counterparty, which would give rise to their own margin requirements which would then fall to be taken into account in any calculation of margin and collateral as between SHI and DBAG under the ISDA Agreements, along with the indirect transactions (the Agent transactions) as part of the total SHI portfolio.
Furthermore, it is common ground that, if the EDTs and OCTs were currency options at all, and if they fell within the ambit of the FXPBA, they would all be “Structured Options” within the meaning of the FXPBA. This means that they did not fall to be taken into account in assessing the Maximum Counterparty Net Open Position nor require SHI to monitor them. Each “Structured Option” was to be the subject of specific agreement by DBAG, as provided by Clause 2, at the time SHI concluded any such trades with a Counterparty. Thus “Structured Options” fell into a special class of options, outside the Counterparty Net Open Position provisions, each of which had to be approved before being concluded by SHI and taken in by DBAG, in order to be an “Accepted Transaction”.
Since the monitoring obligation was to monitor the Counterparty Net Open Position and Structured Options fell outside that, the requirement to provide a summary of the outstanding trades and the net exposure with respect to each counterparty, if asked, could not include the EDTs and OCTs and, for the reasons set out above, did not require any report of MTM valuations on individual trades or margin requirements for the portfolio as a whole.
Both the Settlement Limit (initially set at US$200 million, but increased later for some if not all Counterparties) and the Counterparty Net Open Position (set in 2008 for two Counterparties at US$600m and US$800m respectively) are provided in the FXPBA for the protection of DBAG. Each involves the aggregation of figures on a “per Counterparty” basis and Clause 8 gives DBAG the right unilaterally to amend those limits on 20 days notice or to do so immediately in the event of reasonable grounds to believe that the Counterparty is unable to perform any of its obligations under the Counterparty Master Agreement. Moreover, under Clause 11, if SHI were to exceed one of those two limits without DBAG’s prior consent, DBAG would be entitled to terminate the FXPBA immediately. The limits are provided for DBAG’s benefit and are capable of unilateral modification by it. The agreement to provide a summary of outstanding trades and net exposure in relation to each Counterparty was for the purpose of enabling SHI to monitor its obligations and to ensure that it stayed within the limits provided by Clause 2. No request was ever made for a report of the net exposure with respect to a counterparty for the purpose of monitoring that limit let alone for a report of net exposure on Structured Options which fell outside that limit, so that DBAG cannot be in breach of Clause 2. In short, the express terms of Clause 2 are of no direct assistance to SHI in its allegations of breach of a duty on the part of DBAG to book, value and record trades and to margin them and report such figures. SHI has never alleged any breach of the Clause, as so construed.
Clause 2 does not therefore include an obligation to provide information on MTM values or exposure for each individual transaction nor for the SHI/DBAG position as a whole. A fortiori, there is no obligation to provide any information with regards to any collateral that DBAG might require. Any obligation on the part of DBAG to provide information of this kind is not to be found in Clause 2 and could only be founded on an implied term of the FXPBA as a whole, in circumstances where Clause 2 provides for a much more limited type of information to be given.
Clause 4 of the FXPBA provided that each Agent Transaction should be subject to and governed by the terms of an applicable master agreement between DBAG and SHI, namely the FX ISDA, with its Schedule and CSA.
7(e) The FX ISDA, the Schedule and the CSA
169. The most important provisions of the FX ISDA, the attached Schedule and the CSA relating to Credit Support, collateral and margin (though the last named term is not used in these documents) are as follows:
“The FX ISDA
…
5. Events of Default and Termination Events
(a) Events of Default. The occurrence at any time with respect to a party or, if applicable, any Credit Support Provider of such party or any Specified Entity of such party of any of the following events constitutes an event of default (an "Event of Default”) with respect to such party:-
(i) Failure to Pay or Deliver. Failure by the party to make, when due, any payment under this Agreement or delivery under Section 2(a)(i) or 2(e) required to be made by it if such failure is not remedied on or before the first Local Business Day after notice of such failure is given to the party;
…
(iii) Credit Support Default.
(1) Failure by the party or any Credit Support Provider of such party to comply with or perform any agreement or obligation to be complied with or performed by it in accordance with any Credit Support Document if such failure is continuing after any applicable grace period has elapsed;
…
(3) the party or such Credit Support Provider disaffirms; disclaims, repudiates or rejects, in whole or in part, or challenges the validity of, such Credit Support Document;
6. Early Termination
(a) Right to Terminate Following Event of Default. If at any time an Event of Default with respect to a party (the "Defaulting Party") has occurred and is then continuing, the other party (the "Non-defaulting Party'") may, by not more than 20 days notice to the Defaulting Party specifying the relevant Event of Default, designate a day not earlier than the day such notice is effective as an Early Termination Date in respect of all outstanding Transactions.
The Schedule
Part 1
…
(l) "Additional Termination Event'' will apply. The following shall constitute Additional Termination Events …
…
(iii) Failure to Provide Additional Collateral. If Party A [DBAG] shall for any reason deem that there is/are insufficient Eligible Assets held pursuant to the terms of the Credit Support Document [defined in the Schedule as the Pledge Agreement] and available to satisfy Party B's [SHI’s] present or future obligations under this agreement or Party B's present or future obligations under any other agreement or arrangement between Party B and Party A or its affiliates, Party B shall within two Local Business Day's notice thereof deliver additional collateral assets of such type specified by Party A (which collateral assets shall be delivered and secured pursuant to any existing Credit Support Document or other arrangement in a form satisfactory to Party A in its sole discretion) in an amount as may be required by Party A. If Party B fails to deliver such additional collateral assets, such failure shall constitute an Additional Termination Event with respect to Party B and Party B shall be the sole Affected Party and all Transactions shall be Affected Transactions.
…
Part 5 Other Provisions
...
10. Credit Support Balance
All payments required to be made by Party A in respect of any FX Transaction or Currency Option Transaction under this Agreement (except for payments required to be made under the Credit Support Annex or Section 6(e)) shall be made by way of credit to the Credit Support Balance of the relevant amount and Party A's obligation to make such payment shall be satisfied and discharged in full.”
“The Credit Support Annex (The CSA)
Paragraph 2. Credit Support Obligations
Delivery Amount. Subject to Paragraphs 3 and 4, upon a demand made by the Transferee on or promptly following a Valuation Date, if the Delivery Amount for that Valuation Date equals or exceeds the Transferor's Minimum Transfer Amount, then the Transferor will transfer to the Transferee Eligible Credit Support having a Value as of the date of transfer at least equal to the applicable Delivery Amount (rounded pursuant to Paragraph 11(b)(iii)(D)). Unless otherwise specified in
Paragraph ll(b), the "Delivery Amount" applicable to the Transferor for any Valuation Date will equal the amount by which:
the Credit Support Amount exceeds
the Value as of that Valuation Date of the Transferor's Credit Support Balance (adjusted to include any prior Delivery Amount and to exclude any prior Return Amount, the transfer of which, in either case, has not yet been completed and for which the relevant Settlement Day falls on or after such Valuation Date).
…
Paragraph 3. Transfers, Calculations and Exchanges
(a) Transfers. All transfers under this Annex of any Eligible Credit Support, Equivalent Credit Support, Interest Amount or Equivalent Distributions shall be made in accordance with. the instructions of the Transferee or Transferor, as applicable, and shall be made:
(i) in the case of cash, by transfer into one or more bank accounts specified by the recipient; …
Subject to Paragraph 4 and unless otherwise specified, if a demand for the transfer of Eligible Credit Support or Equivalent Credit Support is received by the Notification Time, then the relevant transfer will be made not later than the close of business on the Settlement Day relating to the date such demand is received; if a demand is received after the Notification Time, then the relevant transfer will be made not later than the close of business on the Settlement Day relating to the day after the date such demand is received.
Calculations. All calculations of Value and Exposure for purposes of Paragraphs 2 and 4(a) will be made by the relevant Valuation Agent as of the relevant Valuation Time.
…
Paragraph 11. Elections and Variables
(b) Credit Support Obligations
…
(C) "Credit Support Amount means, for any Valuation Time, (i) the Transferee's Exposure for that Valuation Time plus (ii) the aggregate of all Independent Amount applicable to the Transferor, if any, minus (iii) the Transferor's Threshold; provided, however, that the Credit Support Amount will be deemed to be zero whenever the calculation of the Credit Support Amount yields an amount less than zero. (h) Other Provisions
…
(ii) Provisions relating to the Allocated Portion, the Eligible
Assets and the Pledged Accounts
Party A may from time to time notify Party B of the Allocated Portion orally or in writing or electronically by email or other similar electronic messaging system or internet based reporting system.
On a Valuation Date, the Credit Support Balance shall be the Allocated Portion as of such Valuation Date (and the definition of "Credit Support Balance” shall be construed accordingly).
Where a transfer obligation arises under Paragraph 2(a) any Eligible Credit Support transferable by the Transferor pursuant thereto shall be transferred into the Pledged Account. (For the avoidance of doubt, the Transferee acknowledges that, notwithstanding Paragraph 5(a) and (b), any Eligible Credit Support so transferred by the Transferor into the Pledged Account will thereafter be held subject to the security created by, and in accordance with the terms of the Credit Support Document.)
The following definitions shall be inserted into the appropriate place in Paragraph 10:
“Allocated Portion” means such amount of the assets standing to the credit of the Pledged Account as is calculated by Party A in its sole discretion and notified by Party A to Party B to constitute the "Allocated Portion" as of such Valuation Date.
"Eligible Assets" means any assets which (i) would be eligible for transfer into the Pledged Account and (ii) would constitute satisfactory collateral pursuant to the terms of the Credit Support Document.
“Pledged Account” means all accounts containing Eligible Assets and which are subject to the terms of the Credit Support Document.
…
(v) Additional Termination Events
The following shall constitute an additional Termination Event with respect to Party B (for which purpose Party B shall be the sole Affected Party):
The Net Collateral Value is equal to or less than the Value at Risk multiplied by the Close-Out Ratio.
The above Additional Termination Event shall apply irrespective of whether or not Eligible Credit Support has been requested by Party A, or is being delivered to Party A pursuant to the terms of this Annex. For purposes of determining whether such an Additional Termination Event has occurred at the discretion of Party A, Exposure, Net Collateral Value and Value at Risk may be calculated at any time on any calendar day, and if such day is not a Valuation Date, the Value of Posted Credit Support may at the discretion of Party A be calculated based on the Value on the preceding Valuation Date.
Furthermore, Party A may use foreign exchange rates as of the close of business on the Local Business Day preceding any calendar day on which Exposure and Value at Risk are calculated when the calculation of such amounts is otherwise determined as of a time on such day.
Notwithstanding any provision of this Agreement that may be to the contrary, if the Additional Termination Event specified in this Credit Support Annex shall occur with respect to party B, Party A shall be entitled to designate an Early Termination Date with respect to all Transactions with immediate effect. Without limiting such right, Party A agrees to use reasonable efforts to deliver to Party B notice of such designation in accordance with Section 12 of this Agreement.”
There is a hierarchy set out in these contractual documents for the resolution of any inconsistencies between their component parts. Under Clause 1(b) of the FX ISDA, the Schedule is to prevail over any other provisions of the FX ISDA, if there is inconsistency between them. The introductory words of the CSA provide that it is part of the Schedule and that, in the event of any inconsistency between it and other parts of the Schedule, it is to prevail. In the event of any inconsistency in the CSA itself, paragraph 11, which is the paragraph dealing with “Elections and Variables”, is to prevail.
Under the FXPBA, DBAG allowed SHI to use its name to conclude FX trades. The provision of collateral under the ISDA was a one way feature in support of SHI’s obligations to DBAG arising out of such trades. Paragraph 2 of the CSA set out SHI’s Credit Support Obligations by reference to the concepts of the Delivery Amount, the Credit Support Amount and the Credit Support Balance, the last of which was the Allocated Portion of the Pledged Account as at any Valuation Date. Under paragraph 11 of the CSA, the Credit Support Amount required to support SHI’s trading was the aggregate of the Independent Amount and DBAG’s Exposure. That Exposure was represented by the amount payable by SHI to DBAG on the Agent Transactions as if they were terminated at the date of the valuation (namely the marked to market valuation or MTM/premium to close out the transactions). The Independent Amount was initially 200% of VaR, representing the potential loss of value which might occur in the course of unwinding those transactions in a given period. In the event that the value of the Credit Support Balance (the Allocated Portion of the assets standing to the credit of the Pledged Account), fell below the Credit Support Amount (constituted by the sum of the Exposure and Independent Amount), the differential, namely the Delivery Amount, was to be made good by SHI with Eligible Assets which would then constitute Eligible Credit Support. DBAG was entitled to notify SHI of the Allocated Portion “from time to time” orally or in
writing, the Allocated Portion being such amount “as calculated by DBAG in its sole discretion and notified” by it to SHI.
SHI contended at one point in its arguments that only sums in the Pledged Account could constitute Eligible Assets, for transfer into the Allocated Portion which is constituted by assets in the Pledged Account, as calculated by DBAG and notified by it to SHI. Paragraph 11(h)(ii) of the CSA provided that where a transfer obligation arose under paragraph 2(a) of the CSA, any Eligible Credit Support should be transferred into the Pledged Account and that, once transferred, it would be held subject to the terms of the Pledge Agreement. Only Eligible Assets can constitute Eligible Credit Support and Eligible Assets are defined as “any assets which (i) would be eligible for transfer into the Pledged Account and (ii) would constitute satisfactory collateral, pursuant to the terms of the Credit Support Document” (i.e. the Pledge Agreement). The Pledged Account is defined in the Pledge Agreement as Account 2011084. What is plain, contrary to that submission of SHI, is that “Eligible Assets” are not confined to assets which are already in the Pledged Account because they are defined as assets which “would be eligible for transfer into the Pledged Account”.
The reason for the convoluted construction given by SHI to the definition of “Eligible Assets” which are capable of constituting “Eligible Credit Support for SHI’s trading under the FXPBA and FX ISDA” is the desire to support, by an alternative route, SHI’s case of a maximum limit of loss of US$35 million and/or the Pledged Account Limit (PAL), which are primarily said to arise as a result of a separate oral agreement between Mr Vik and Mr Meidal. Contrary to its submission that the only available assets which could constitute collateral for the envisaged FX trading were those already in the Pledged Account, the very definition of Eligible Assets envisages that the assets may lie outside the Pledged Account, because it refers to assets which “would be eligible for transfer into the Pledged Account” and “would constitute satisfactory collateral, pursuant to the terms of the [Pledge Agreement]”.
Thus a margin demand made of SHI for a “Delivery Amount” requires a transfer by SHI of “Eligible Assets” into the Pledged Account, if rendered necessary because there are inadequate assets in it, to top up the Allocated Portion which constitutes the Credit Support Balance so that it matches the Credit Support Amount required at the relevant time. Of course if there are other Eligible Assets in the Pledged Account which are not part of the Allocated Portion and are available to become the Allocated Portion (and therefore the Credit Support Balance), then there would be no need for a transfer from outside but in the absence of that, there is no doubt that the CSA provides for margin calls to be met by SHI by payment into the Pledged Account to form part of the Allocated Portion.
Paragraphs 5(a) and (b) of the CSA provide that all right, title and interest in and to any Eligible Credit Support transferred by SHI to DBAG under the CSA is to vest in DBAG free of any encumbrances and that nothing in the CSA creates any security interest in any assets transferred by SHI to DBAG. However by paragraph 11 which is to prevail over any inconsistent provision in the CSA, sub-paragraph (h)(ii) expressly provides that, notwithstanding the terms of paragraph 5(a) and (b), any Eligible Credit Support transferred as margin under paragraph 2(a) of the CSA into the Pledged Account will thereafter be held subject to the security created by and in accordance with the terms of the Pledge Agreement.
SHI submitted that funds could only be paid from the unallocated part of the Pledged Account to the Allocated Portion and that, when this occurred, title to those funds would transfer from SHI to DBAG. In this connection it relies upon the use by parties of the ISDA form of the English law CSA which, in a footnote, states that the document is not intended to create a charge or other security interest over the assets transferred under its terms. Reliance is then placed on paragraphs 5(a) and (b) but paragraph 11(h)(ii) prevails over this and SHI’s argument falls foul of the express terms of that sub-paragraph under the definition of Eligible Assets, to which I have already referred. SHI refers to the use of “clumsy language” and “an imprecise use of language”, when the language is plain and it is SHI’s construction which is clumsy in its attempts to avoid its obvious meaning.
The terms of the Pledge Agreement itself make it plain that any and all assets whatsoever deposited or relating to account 2011804 of SHI with DBS, at present and in future, are pledged by SHI to DBAG to “serve as collateral to [DBAG] for all claims that [DBAG] has and/or will have against [SHI]”. The assets in the account are held by DBS but are, by Clause 4 of the Pledge Agreement, at DBAG’s disposal and “may be sold, exchanged and repaid only with [DBAG’s] express or implied consent”. By Clause 7 the assets are to be administered in accordance with the safe custody regulations of DBS and SHI remains responsible for taking any precautions necessary to protect the value of the assets pledged. In the event that SHI is in arrears of any existing or future credit facility or agreement with DBAG, or is in default of any obligation towards DBAG, DBAG is entitled automatically to realise the assets under the terms of Clause 9. By Clause 11 SHI undertakes if necessary to participate in the transfer of assets to a new purchaser as required by DBAG and by Clause 14 irrevocably and unconditionally instructs DBS to follow any and all instructions from DBAG regarding the assets, including sale, realisation and transfer instructions.
The Pledge Agreement is governed by Swiss law but there is no suggestion of any difference between Swiss law and English law. It is a pledge agreement where title remains with SHI and DBAG has a security interest in accordance with the terms of the pledge. It is, moreover, by Clause 16 effected “in addition to and independently of any existing or future securities/collateral” and is to remain in force until DBAG’s claims are met in full. Furthermore, by Clause 17 DBAG is not responsible for not exercising its rights under the Pledge Agreement nor for any consequential damages for actions taken to realise the security.
SHI’s purpose in advancing these arguments on construction was to seek to negate DBAG’s argument that it had the right but not the obligation to demand collateral from SHI and to argue that SHI’s ability to trade was limited to the capital in the Pledged Account which DBAG was obliged to manage. SHI submitted that DBAG had an obligation under Clause 11(h)(ii) of the CSA to calculate the Allocated Portion on each Local Business Day. I am unable to see how this obligation can arise when paragraph 11(h)(ii) expressly provides that DBAG “may from time to time notify [SHI] of the Allocated Portion” and contains no requirement that it should do so.
That is the governing wording so that although the CSA provides that on a Valuation Date (each Local Business Day) the Credit Support Balance is to be the Allocated Portion at such a date, that adds nothing and is neither here nor there.
Paragraph 3(b) of the CSA provides that all calculations of Value and Exposure “for the purposes of Paragraph 2” (the calculation of the Delivery Amount and the Credit Support Balance) are to be made by DBAG as of the close of business on the business day preceding the Valuation Date. There is, however, no obligation on DBAG to notify SHI of those calculations of value and exposure at all, because the standard provision to that effect in paragraph 3(b) has been deleted. A calculation “for the purpose of paragraph 2” is, on its own terms, a calculation made in order to make a demand – a margin call. If there is to be no demand, there is no independent requirement or need for a calculation to be made “as of the relevant Valuation Date or Time”.
The express terms of paragraph 2, paragraph 3(b) and paragraph 11(h)(ii) of the CSA do not therefore create an obligation on DBAG to notify SHI of the Allocated Portion on a daily basis, nor to make a demand of a Delivery Amount, nor to effect margin calculations, unless a demand is to be made. If DBAG requires further collateral, which is a matter for it to decide, since the collateral is for its protection, it must, if it so chooses, then calculate the Delivery Amount in accordance with the contractual methodology and notify SHI of an increase in the Allocated Portion of the Pledged Account, so that a greater part of that account is then to be treated as available collateral under the FX ISDA. If there are insufficient Eligible Assets in the Pledged Account to constitute the Credit Support Amount (the MTM deficiency plus 200% VaR) it can make a margin call by giving notice of its calculation of a new figure for the Allocated Portion and by demanding a transfer of the Delivery Amount for payment into the Pledged Account so as to increase the size of the Allocated Portion.
Paragraph 11(h)(v) of the CSA provides for an Additional Termination Event where the Net Collateral Value (representing the Allocated Portion less the MTM deficiency) becomes equal to or less than the VaR, representing the loss that would be incurred on liquidating the assets. Where therefore DBAG is at risk of a deficit if the assets had to be liquidated and the collateral realised, DBAG has the right to terminate the FX ISDA and to do so “with immediate effect”. The sub-paragraph goes on to state that this applies “irrespective of whether or not Eligible Credit Support has been requested by [DBAG] or is being delivered to [DBAG] pursuant to the terms of this [CSA]”.
The very terms of this sub-paragraph predicate a situation where DBAG is entitled to demand margin but may not have asked for it under paragraph 2 (in circumstances where the Credit Support Amount is constituted by the aggregate of the MTM and 200% of VaR). For this Additional Termination Event to occur, a situation under paragraph 2(a) of the CSA must necessarily have arisen and the terms of paragraph 11(h)(v) therefore envisage that occurring without a call being made.
There is nothing uncommercial about this facility to make, or the effect of making, a declaration of an Early Termination Date in such circumstances. Everything in the Pledged Account (including the Allocated Portion) remains in the ownership of SHI but, as it is collateral for SHI’s obligations to DBAG and is held under the Pledge Agreement, it is susceptible to realisation by DBAG under the terms of that Pledge Agreement in order to satisfy any indebtedness of SHI. If there is any balance remaining, which in such circumstances would appear unlikely, SHI’s entitlement to that balance remains undisturbed.
A further aspect of the arrangements between the parties in relation to the Credit Support Balance is the requirement in paragraph 10 of part 5 of the Schedule to the
FX ISDA. This required all payments by DBAG in respect of any of SHI’s FX trades under the FX ISDA to be made by way of credit to the Credit Support Balance. As the Credit Support Balance, in accordance with paragraph 11(h)(ii) of the CSA, is the Allocated Portion on any Valuation Date, the trading profits, if paragraph 10 of part 5 of the Schedule is taken at face value, would go to increase the collateral available as part of the Allocated Portion, without any notification of an increase in the Allocated Portion. As paragraph 11(h)(ii) provides that the Allocated Portion is constituted by the amount of assets standing to the credit of the Pledged Account, as calculated by DBAG in its sole discretion and notified by it to SHI, a notification of payment of profits into the Pledged Account can be taken as a notification of an increase in the Allocated Portion. Since accumulated profits can only be paid into the Allocated Portion, the effect of notification of such payment must be to increase the collateral even if there is no express notification that the Allocated Portion has been increased. The collateral available to support the FX trading is thus increased by the profits earned, as long as they remain in the Pledged Account, because they automatically form part of the Allocated Portion. The accumulation of profits in the Allocated Portion would serve to increase the scope for collateralised trading by Mr Said above and beyond what was possible with US$35 million.
In fact, no change in the Allocated Portion was ever notified by DBAG and most of the profits earned on Mr Said’s trading were, with DBAG’s consent, withdrawn from SHI’s trading account (in one instance via the Pledged Account) without, it appears, any consideration of these provisions of the CSA. Some US$30m was withdrawn in late 2007 and some US$75m (including a sum in Euros) in mid 2008 but, under the terms of the ISDA, once the money was in the Pledged Account, whether as part of the Allocated Portion or not, it could only be released from it with the consent of DBAG, whilst DBS maintained its monitoring role to ensure that the lending value of the assets pledged was always US$35 million or more. On 6 October 2008, Mr Said agreed to new margin terms, which, for the reasons I have already set out in the context of the Said Letter of Authority, he had authority to do. No notification of any change in the Allocated Portion was ever made to allow for this, even though the suggestion at the time was that the effect of the change would be to increase the margin required from about US$20m to US$40m, some US$5m more than the notified Allocated Portion. This aspect was soon overtaken by events and the margin calls were met by payments from other sources than the Pledged Account, with Mr Vik signing transfers for payment direct to DBAG’s own accounts.
There is thus no basis in the FX ISDA, the Schedule or the CSA for the contention that DBAG was restricted to collateral of US$35m or to what was in the Pledged Account at any time. There was a procedure to be followed in the provision of more collateral, in the shape of a margin call and a requirement to increase the Allocated Portion of the Pledged Account by “allocation” of assets within it, or transfer of external Eligible Assets to it, but, on the face of these Agreements, there was nothing to prevent DBAG from seeking additional margin, calculated in accordance with the CSA, whenever the situation arose. As is recognised by SHI, the effect of its argument is that DBAG would be assuming responsibility for any drop in the market value of the assets in the Allocated Portion or the Pledged Account, if it, when aggregated with the Initial Margin, gave rise to a figure which fell below the required margin level calculated in accordance with the CSA. This is an unlikely and unbusinesslike construction of a provision setting out a requirement for security.
It is self evident that the reason for the provision of “the Allocated Portion” was because the Pledged Account was also intended to serve as collateral for other transactions apart from Mr Said’s FX transactions. Only a part of that Pledged Account was intended to serve as collateral for Mr Said’s FX trading, namely the NOK equivalent of US$35m, whilst other types of transaction could be secured on other assets in the Pledged Account, as permitted by DBS in accordance with its separate requirements. Initially the documents show that the NOK equivalent of another US$35m (approximately) was put into the Pledged Account as security for other types of trading that Mr Said had in mind, such as Fixed Income trading and trading of the kind he effected during the lead up period to the FXPBA, in Credit Default Swaps and Argentinian Bonds.
The reason for the maintenance of the Pledged Account at DBS, as opposed to security placed with DBAG, appears to have been both Mr Vik’s desire to maintain his cash balances in NOK to the greatest extent practicable in the context of trading in US$ as the base currency and the PWM connection with DBS. DBS had undertaken by the TPMCA to monitor the sums in the Pledged Account so as to ensure that the assets, denominated in NOK, did not fall below the US$35 million figure, but this was a private arrangement between itself and DBAG, to which SHI was not a party. The terms of the Pledge Agreement governed the basis upon which assets within it could be dealt with.
Furthermore, under Part 1(l)(iii) of the Schedule to the FX ISDA, if, for any reason, DBAG deemed that there were insufficient Eligible Assets held in the Pledged Account to satisfy SHI’s obligations under the FX ISDA or any other agreement between it and DBAG or one of its affiliates (including for this purpose DBS), it was entitled to demand additional collateral assets to be “delivered and secured pursuant to” the Pledge Agreement or pursuant to any “other arrangement in a form satisfactory [to DBAG] in its sole discretion” in such amount as DBAG required. A failure on SHI’s part to provide such additional collateral within 2 days would constitute an Additional Termination Event. This provision was expressly additional to any rights given to DBAG to designate an Early Termination Date under section 5(a)(iii) of the FX ISDA which made a failure to provide margin an Event of Default.
In these circumstances, SHI’s contentions on the construction of the FX ISDA, the Schedule and the CSA are untenable and its allegations of breach based thereon fall to the ground.
7(f) The Pledge Agreement
I have made reference to the Pledge Agreement in construing the FX ISDA which refers to it as the Credit Support Document. It was a tripartite agreement between SHI, DBAG and DBS. As Pledgor, SHI pledged its assets, claims, tangible property and rights now or later held by DBS, including its claims on DBS itself, as collateral to DBAG for all claims that DBAG had or would have against SHI. The pledged assets were listed in article 22 as being “[any] and all assets whatsoever deposited or relating to Account 2011084 of SHI with DBS, at present and in future”.
The pledge included, by article 3, all rights that had already fallen due and all future rights attaching to the pledged assets and rights (such as interest, dividends, coupons, warrants etc). Article 4 provided that, as the pledged assets were at the disposal of
DBAG (London branch), the pledged items were only to be sold, exchanged and repaid with DBAG’s consent. They were to be held and administered in accordance with the safe custody regulations of DBS. Under article 9, if SHI was in arrears under any existing or future credit facilities or agreements, as referred to in article 6, or in default with regard to any obligations towards DBAG, DBAG was entitled, but not obliged, to realise the assets forthwith at its discretion, either freely by private sale (including to itself) or by force of law and to apply the proceeds in full or partial reimbursement of the debts owing.
SHI authorised DBS to provide DBAG with unrestricted information about the assets and its banking relationship with DBS and irrevocably and unconditionally instructed DBS to follow all instructions from DBAG regarding the assets, including the sale and realisation thereof, without liability on the part of DBS.
Under articles 16 and 17, the pledge was stated to be in addition to and independent of any other existing or future security/collateral and was to remain in force until all sums owing to DBAG were met. DBAG was not to be held responsible for not fully exercising rights accruing to it under the Pledge Agreement nor for any consequential damage for actions taken in realising the assets under article 9.
The Pledge Agreement was governed by Swiss law but no evidence was adduced by either party of its contents.
It is plain from the language of the Pledge Agreement that property in the assets in the Pledged Account remains with SHI and that they stand pledged as security for indebtedness to DBAG, whilst deposited with DBS. By the separate Third Party Monitored Collateral Agreement (the TPMCA), to which SHI was not a party, DBS agreed with DBAG to hold the assets pledged in DBAG’s favour under the Pledge Agreement with SHI, the holder of account number 2011084 and pledgor for margin lines extended to it at the London branch of DBAG. The pledged assets amounted to a figure just short of US$70 million. DBS stated that it had blocked and would safeguard the pledged assets held at its Zurich office and would ensure that they were not released prior to receipt of authorisation from DBAG. DBS agreed to monitor the loanable value of the pledged assets in accordance with its own margin requirements and procedures and undertook to advise DBAG should the loanable value of the assets fall below US$35 million so that DBAG could take the necessary action to maintain the required loanable value (under Paragraph (i)(iii) of Part 1 of the Schedule). The TPMCA referred to the value of the assets in both US dollars and CHF but in fact the instructions from Mr Vik were that all sums should be held in NOK, save in so far as they were required for the fulfilment of obligations in other currencies.
7(g) The Limited Power of Attorney
198. On 28th November 2006 Mr Vik on behalf of SHI also signed a limited power of attorney, addressed to DBS, appointing Klaus Said attorney “in respect of all the assets and valuables deposited” in SHI’s account with DBS. The customer reference number was given as 2011084, which was the account number for the Pledged Account. The limited power of attorney comprised “the most extensive possible powers of management and administration, to the exclusion of all powers of disposal”. It permitted Mr Said to buy and sell securities, to invest and reinvest, to encash and convert, to exercise and realise subscription rights and to use coupons constituting assets or valuables in the account. What Mr Said could not do was to withdraw anything from the account or create any additional pledges over assets in it or contract loans. All Mr Said’s actions, performed within the limits of the mandate, were to bind SHI and DBS was to be relieved of all liability in respect of them. In the absence of wilful misconduct or gross negligence, within the limits stipulated by the law, DBS was to incur no liability to SHI, with an indemnity from SHI to DBS in respect of complaints lodged against it by anyone else in connection with the power of attorney.
8. Implied Terms
As appears from the discussion of New York law pertaining to contract, as set out above, whatever reluctance a court may have to imply a term that the parties have neglected specifically to agree if the parties are sophisticated business entities dealing at arm’s length, a term will be implied into a contract where a reasonable person in the position of a party to it would be justified in understanding such a term to be included and such term at the time of contracting was implicit in the agreement when viewed as a whole. It will not however imply a term which conflicts with the express terms of the written contract. The reservations expressed by SHI’s expert about any narrowness in this test, with particular reference to terms implied by custom and usage, make no difference to the determination of the issues in this action. The exact formulation of the test for implication of terms is not significant in the context of the implied terms alleged when set out against the express terms of the contracts and the factual background.
The implied terms of the FXPBA for which SHI contends are set out in paragraph 38 of its re-re-re-amended defence and Part 20 counterclaim (the “RRRADC”). These implied terms are extensive and read as follows:
“38. Under New York law, on the true construction of the FX PB Agreement in the light of the aforesaid factual background and/or as an implied term, it was a term of the FX PB Agreement that:
The Bank would obtain from each Counterparty, retain and/or provide to SHI on demand and/or within a reasonable period of time after each trade a confirmation in writing of each FX Transaction that accurately recorded each FX Transaction.
The Bank would ensure that each FX Transaction was booked, valued and recorded accurately in SHI’s account (i.e., trading book) with the Bank relating to such FX trades (“the FX Account”). Alternatively, the Bank would exercise reasonable care to ensure that each FX Transaction was booked, valued and recorded accurately in the FX account.
The Bank would ensure that each FX Transaction was performed by the Bank and the Counterparty strictly in accordance with the terms of such confirmation.
(3A) No equal and offsetting transaction is capable of arising under clause 4 unless the transaction entered into with the Counterparty was within the scope of the relevant Counterparty Master Agreement as defined in clause 3.
The Bank would allocate capital in the Pledged Account in respect of each FX Transaction in accordance with the FX ISDA Agreement and would carry out complete and accurate calculations of the capital required in order to allocate capital appropriately and would notify SHI of the capital requirements of its FX trading when calculated, alternatively would take reasonable care to ensure that calculations of the capital required were carried out completely and accurately, and to notify SHI of the capital requirements of its FX trading when calculated.
(4A) The Bank would inform SHI of any inability or failure to:
book and/or to record and/or value accurately or at all SHI’s transactions in the FX Account of SHI or in its reporting systems, and/or
carry out any, or any complete or accurate, calculations of the capital required in order to allocate capital appropriately, and/or failure to allocate capital in the Pledged Account properly or at all.
(4B) The Bank would ensure that, in circumstances when it was proposed to enter into a high risk product directly between SHI (through Mr. Said) and the Bank, particularly a leveraged derivative (such as the Exotic Derivatives Transactions, as referred to below), Mr Vik understood the risk level of the product; alternatively the Bank would take reasonable care to ensure that, in such circumstances, Mr Vik understood the risk level of the product. Such a term is to be implied from the fact that it was at all material times the usual custom and practice of an investment bank (such as the Bank) to have a policy or policies providing that, in respect of high risk products, in particular leveraged derivatives (which would include transactions such as the Exotic Derivatives Transactions), that it was proposed to trade directly between the client and the bank, the bank would ensure before entering into the transaction that an appropriate individual in the client organisation other than the usual trading contact understood the risk level of the product. This would typically involve an explanation of the risk level of the product being given to an appropriate individual in the client organisation other than the usual trading contact, unless the bank knew, by reference to previous trading, the client's risk tolerance or other relevant facts, that the said appropriate individual already understood the risk.
(4C) The Bank would not allow any trading in a product not approved through the appropriate New Product Approval Processes. Such a term is to be implied from the fact that it was at all material times the usual custom and practice of an investment bank (and SHI understands that it was also a policy and practice of the Bank) not to allow any trading in a product not approved through appropriate New Product Approval Processes.
(4D) The Bank would not enter into any transaction with
SHI, or accept any transaction under the FX PB Agreement (alternatively, the Bank would not accept under the FX PB Agreement any Structured Option (by refusing to give its approval pursuant to clause 2(iii) of the FX PB Agreement)) which the FX Prime Brokerage division was unable to book, value and record accurately in the FX Account or in its reporting systems and in respect of which it was unable to carry out any, or any complete or accurate, calculations of the capital required to support such trading and to report accurately to SHI such capital requirements (or, alternatively, in circumstances where it was unable to do any one or more of these tasks). Alternatively, the Bank would take reasonable care to ensure that transactions were not entered into in the said circumstances.
(5) The Bank was subject to a duty to act in good faith and a duty of fair dealing in the course of its performance such that the Bank was required not to do anything which would have the effect of depriving or injuring the right of SHI to receive any of the intended benefits for which it bargained under the FX PB Agreement; such term being implied as a matter of New York law as the governing law of the FX PB Agreement.”
None of these 10 alleged implied terms are readily associated with any of the express terms of the FXPBA, save for those which bear on the provisions of Clause 2. The FXPBA set out the arrangement by which DBAG authorised SHI to act as its agent in concluding transactions with Counterparties and contained qualifications and restrictions on SHI’s authority to do so by reference to the types of trade, tenor of trade and the Settlement Limit and the Maximum Counterparty Net Open Position Limit. It provided for DBAG to enter into Counterparty Agreements with Counterparty banks before any Counterparty transactions could be concluded by SHI and for DBAG to enter into offsetting trades with SHI when SHI concluded authorised Counterparty Transactions. As already set out, the FXPBA did not expressly provide for any obligations on DBAG beyond that, save for those set out in Clause 2. DBAG did not expressly agree in the FXPBA to provide any other services than those to which I have referred.
The provision under Clause 2 of “a summary of the outstanding trades and the net exposure with respect to each Counterparty” requires a reporting capability on the part of DBAG, namely an ability to furnish the basic details of each trade and the MTM valuations of those trades, in order to provide the summary and the Counterparty net exposure position. In practice this would necessitate the maintenance of records of the trades which contained the relevant information although the obligation, is a reporting obligation, rather than a record keeping obligation and applies when a request is made. It is undoubtedly the case however that the expectation of the parties would be that such records would be maintained and it was not suggested otherwise by DBAG.
Mr Quinn’s evidence, as set out above, was that the core functions of a Prime Broker were to lend its name and credit to the trader and to clear and settle the trades by matching the information supplied by the trader and the Counterparty in relation to the trades effected between them and by effecting the monetary settlements which resulted from the contracts so concluded. The manner in which the Prime Broker booked or recorded these trades was an internal issue for itself alone and not a matter for the client. The reporting function involved only reporting of basic trade details in order to enable the client to reconcile his own records with those of the Prime Broker in circumstances where every trader could be expected to retain his own records of trades and positions, to watch the market and to manage his own risk. No trader would rely upon the Prime Broker’s collateral requirements to assess the market risk being run in relation to liability on trading positions. (On his own evidence Mr Said was no exception to this.) Any trader would be able to tell you, at any point in time, how his trades stood with respect to the current state of the market, from his own records and from following the market himself. Risk management was a matter for the client and not for the Prime Broker whilst the Prime Broker’s entitlement to call for collateral in the shape of margin was a provision for the Prime Broker’s benefit alone and one which there was no obligation on the Prime Broker to utilise or enforce.
Whilst however these were the core functions inherent in any Prime Brokerage, Mr Quinn accepted that the provision of additional services could be agreed between the Prime Broker and the client and although market practice did not require any additional services to be provided, an agreement could be concluded to that effect.
It is the fact that DBAG did provide additional services in the shape of web reporting and, in circumstances to which I will refer later in this judgment, the provision of spreadsheets created manually by Mr Walsh which set out vanilla transactions with MTMs. On what basis were these services supplied? SHI contends that they were supplied as part of the services afforded by DBAG as a Prime Broker, that the FXPBA was the governing contract in respect of Prime Brokerage and that terms must be implied into the FXPBA in relation to the provision of those services. SHI submits that the function carried out by DBAG in practice, however badly, in booking, recording and valuing SHI’s trades and in reporting those trades, their MTM value and the margin requirements of the portfolio were so fundamental to the Prime Broker relationship between SHI and DBAG that the court ought to hold that they were implied terms of the FXPBA. It is said by SHI that a reasonable person in the position of SHI would be justified in understanding that there were implied terms in the FXPBA to that effect, since without them the framework in which SHI operated as DBAG’s agent and entered into offsetting transactions would not exist.
It is in this connection that SHI rely upon the promotional literature and brochures produced by DBAG and sent to SHI prior to conclusion of the FXPBA and to the evidence of DBAG’s own witnesses as to the role and function of a Prime Broker.
There was a wealth of evidence from DBAG about the role and function of a FXPB desk. None of DBAG’s witnesses drew any distinction between the FXPB desk’s functions of booking, recording, valuing and margining trades or margining the portfolio or the reporting of the details of the trades, the MTM values and the margin requirements. None of them appeared to have any knowledge of the contractual provisions which operated between DBAG, DBS and SHI, although some of them appreciated and advanced the view that margining was an entitlement which rested in DBAG for its own protection alone.
DBAG charged a monthly fee to SHI in relation to the Counterparty Transactions concluded during that month on the basis of US$8 per US$1 million of the Notional Amount of each transaction (although the Notional Amount of the TPFs could, in reality, only be assessed after knock out or maturity). This level of fee was suitable for the routine transactions which were dealt with in an automated fashion. In the ordinary way, with such routine vanilla transactions, the trades would be booked with apposite trade details in DBAG’s automated system, by both Mr Said for SHI and the Counterparty. The system would match the transactions and confirmations could be sent to both parties showing the trade between DBAG and the Counterparty and DBAG and SHI. Under DBAG’s automated systems, this would lead to the creation of initial margin at the outset by the use of DBAG’s ARCS VaR FX II (ARCS VaR) system in the case of SHI’s portfolio of trades (and by its NOP system in GEM for many other customers). Initial margin was a figure intended to represent the losses which the portfolio would incur during the time it would take to liquidate it. Thereafter the trades and assets would be valued (marked to market or MTM) and variation margin would be calculated by the same VaR engine. In practice therefore, if margin was calculated and required on a daily basis, the variation margin payable on any given day would represent the shortfall between the aggregate MTM valuation of the portfolio on that day as compared with the valuation the previous day. The aggregate amount of the Initial Margin and the Variation Margin was thus to be the subject of collateral available to DBAG, as provided by SHI.
With simple trades the client would input the trade details in the computer system by use of a programme called TRM which would feed into DBAG’s Risk Management System (RMS) and into its Web Reporting System (GEM) by which the customer could access, 24 hours in the day, the details of the trades done, the MTM valuation (supposed to be updated every 15 minutes of the day) of those trades, the net asset position and the current collateral situation. The Counterparty would input details of the same trade by use of another programme called Harmony which would automatically match those trades. In the event of a failure to match, because of a discrepancy (a break), human intervention was needed to resolve the issue.
Equally, with more complex trades, for which the automated system was not specifically designed, human intervention was needed for inputting the trade details and matching, using the menus available within the system. DBAG had an office in India which was used to effect such inputting. If however the transaction was yet more complex, intervention would be needed on the part of DBAG’s FXPB operations department in New Jersey, if necessary with the assistance of personnel in the front office (business/sales) in 60 Wall Street, New York.
The valuation and reporting of MTM values and margin levels depended upon the correct inputting of the trade details. Once the trade details fed in by the customer, the Indian office or the FXPB operations department had been matched by the Counterparty, whether through Harmony or by email with a short form trade confirmation, the transaction would be booked in RMS and its details would then appear in GEM. Valuation for margining on the VaR basis was effected by the ARCS VaR engine which fed through details of its calculations to GEM, with the MTM values as well as the Initial Margin. Margining on the NOP basis, with MTM valuations for that purpose, came from a different source than ARCS VaR to GEM. As appears later in this judgment, there were problems in the feed of information from ARCS VaR to GEM throughout the period from June 2007 to August 2008 so that, on ordinary vanilla transactions, the MTM seen on the GEM Web Reporting site was inaccurate and did not reflect the current state of the trades done. In consequence, a work around was effected whereby a new reporting account was set up in GEM. This was intended to replicate the trade details which appeared in the Outstanding Trade Details part of the website so that the trade information in this account auto-generated the same information in what became known as the P&L Reporting Account. MTM valuations from the NOP system were then fed into the P&L Reporting Account which was not accessible to SHI on the GEM website, being only available to DBAG personnel. At the close of business each day in the periods when these problems arose, Mr Walsh would send Mr Said, by email, a spreadsheet which represented a screenshot of the vanilla trades (with some adjustment) and the MTM position on each as they then appeared. Mr Said’s main concern here was with the vanilla cash transactions, not the options, as set out later in this judgment.
During the course of the relationship between SHI and DBAG, Mr Said increasingly was to effect trades which would not fit into the automated system and which were recognised by him and DBAG to be “Structured Options”. He sought DBAG’s consent for these and, when obtained, provided details of the trades by email. Manual intervention was then needed in order to book the trades into DBAG’s systems although, as appears elsewhere, there were insufficient fields available in the computer systems for the EDTs to be properly booked. Nonetheless, in the light of the increasing manual work required in particular to match trade confirmations, a special fee was negotiated and agreed with Mr Said, namely US$1,200-$1,500 in respect of each EDT. Nonetheless, in total, DBAG received less than US$400,000 in contractual remuneration for allowing SHI to trade in its name, for giving SHI the benefit of its net margin arrangements with Counterparty banks whilst requiring SHI only to put up collateral to DBAG alone and, in addition, providing whatever services it did provide to SHI, which SHI calls “back office services”. These figures can be seen in the context of realised profits of about US$40 million in Mr Said’s first year of trading and some US$81 million in eight months in 2008 until the market moved against him with resultant losses in excess of US$600 million in October 2008.
Notwithstanding this low level of remuneration, DBAG did provide services beyond the “core services” to which Mr Quinn referred.
Prior to the conclusion of the FXPBA, DBAG supplied SHI with a “Pitch Book”, which Mr Said described, in his deposition, as a “standard” for a bank offering prime brokerage services. This Pitch Book set out the benefits of prime brokerage as including the efficient use of capital and increased operational efficiency with a centralised FX clearing facility, centralised credit arrangements, centralised settlement, position and trade reporting, error rate and market risk reduction through timely trade matching and confirmation, auto-matching technology (for Straight Through Processing by TRM), online real-time reporting and a simplified documentation process. “Operational risk management” was to be achieved by the TRM auto-matching engine in the computer systems and website reporting with trade data feeds. “Product depth” specifically included “FX options” and “Exotic FX options”. In the Basic Trade Flow diagram, reference was made to the single credit line, single collateral relationship, consolidated reporting and single reconciliation achieved by the prime brokerage system operated by DBAG. Trade matching, credit monitoring, allocations and valuations were specifically mentioned. As part of the client service there was “on-line reporting” which was described as a real-time suite of trade reports which included trade detail, cash flow summaries, collateral summaries, historical reports, bank credit lines with all reports marked to market every 15 minutes. Electronic summary trade files were said to be available for automatic reconciliation purposes. Reference was made to efficient back office trade capture and reconciliation, and, on the page dealing with “Credit Terms”, under the heading “Margin Requirements”, the explanation was given that credit terms were specific to each client’s needs with Net Open Position based margining (NOP) on a tiered currency basis with percentages determined by credit. VaR methodology was also said to be available. As part of the seamless integration into DBAG’s FXPB system, the first step was set out as agreement on the product scope for the service.
DBAG’s Foreign Exchange Web Reporting brochure stated that the web reports allowed access to information relating to “trades, positions, cash flows, collateral account balances and P&L”. Available reports included Margin Status, Open P&L, Collateral Reports including Assets on Deposit, Collateral Summary Reports and Net Equity Breakdowns with Close of Business reports including the history position of trades and cash flow, Collateral Summary and Open P&L. Key features included revaluation every 15 minutes. Whilst the margin status reports referred to the NOP method of valuation, the Open P&L screenshot shown in the brochure referred to a display of all outstanding option positions, latest revaluation rate and current market value of outstanding positions. The Collateral Summary screenshot referred to drilling down to a Net Equity Breakdown Report and showed current market values of open trades for available collateral purposes, the available collateral sum of assets on deposit and their current market value and the collateral required based on total current NOP and margin ratios. The amount by which the required collateral exceeded or fell short of the net equity was then set out with the total of cash and options positions netted against each other.
In its promotional literature, DBAG thus held itself out as being capable of running a “back office” system for the recording of transactions, including exotic FX options, for their valuation and margin calculation with reporting to the client. All the reports were said to be marked to market every 15 minutes in a real-time suite of reports which included trade details, cash flow summaries and collateral summaries.
Although it is argued by DBAG that it undertook no obligations beyond those which appear in the FXPBA itself, such an argument would suggest that all the services referred to in the Pitch Book and the Website Brochure, when provided, were not contractual services and fell outside the terms of the FXPBA. DBAG did not go so far as to suggest that they were provided on an ex gratia basis, without responsibility, whilst denying the existence of the implied terms alleged by SHI.
Whilst therefore the FXPBA included very limited express obligations on DBAG, and the obligation under Clause 2 to provide, if asked, a summary of outstanding trades and the net exposure per Counterparty to enable SHI to monitor the Net Daily Settlement Amount and the Counterparty Net Open Position was likewise very limited, with the latter figure specifically excluding Structured Options, the promotional literature referred to a wide range of services including reporting services to be provided by DBAG, largely by means of the GEM Web Reporting System.
It cannot have been intended that these services were freestanding, offered and accepted as an act of generosity on the part of DBAG. These services were proffered by DBAG in its capacity as a Prime Broker and as part and parcel of Prime Brokerage services, as advertised in its promotional literature, in order to attract customers to use it as a Prime Broker. Whilst the remuneration for Prime Brokerage services was limited in itself, there was a potential spin off, of which personnel in the FXPB department of DBAG were aware, namely that SHI might conduct direct trades with DBAG’s Trading/Franchise desk, thus generating profits for DBAG in a different department. Moreover, DBAG recognises that these services were not freestanding, in as much as it contends that the web reporting services were governed by the GEM Terms and Conditions. It thus accepts that there was a contract between SHI and DBAG in relation to web reporting, albeit that it relies upon those terms and conditions as excluding it from liability in relation to inaccurate reporting. One way or another, it is in my judgment clear that there was an agreement of some kind to provide the services which were actually carried out. These services involved the provision of information, not the internal booking and recording of the trades by DBAG, although, of necessity, in order to report, DBAG had to maintain some systems which would enable it to do so. The GEM web reporting agreement cannot be seen in isolation from the FXPBA – the two are plainly linked, as both parties recognise.
In this context it is to be borne in mind that the FXPBA cannot be seen in isolation from the FX ISDA, the Schedule and the CSA, to which it refers. Under Clause 4 of the FXPBA, every Agent Transaction was to be “subject to and governed by the applicable ISDA Master Agreement, including the Credit Support Annex”. Specifically the Agent was to be required to post collateral in accordance with the terms and provisions of the CSA, it being agreed that any breach of the FXPBA on the part of the Agent (SHI) should constitute an Event of Default under the FX ISDA. The FX ISDA and the CSA are governed by English law.
With these matters in mind, I turn to the specific alleged terms set out in paragraph 38 of the RRRADC, there being no basis for contending that there are any express terms to that effect.
8(a) Paragraph 38(1)
The first implied term for which SHI contends is that the DBAG would obtain from each Counterparty and retain and provide on demand or within a reasonable period, a
trade confirmation in writing for each FX Transaction concluded under the terms of the FXPBA and that such confirmation should accurately record the terms of each transaction.
DBAG submits that such a term would be inconsistent with the structure of the FXPBA which authorises SHI to execute transactions with the Counterparties and, by Clause 3, to communicate trade details of each Counterparty Transaction to DBAG (its FXPB desk in practice, as opposed to its Trading or Franchise Desk) by fax or electronic means. Only SHI therefore knew the terms agreed between it and the Counterparty which gave rise to the offsetting transaction between SHI and DBAG. Clause 7 of the FXPBA provides that SHI is to indemnify DBAG against loss resulting from any error in any information provided by it to DBAG. In practice, in the case of simple transactions SHI could feed the information directly into DBAG’s computer system (RMS) which would carry out a matching process with information fed by the Counterparty. On more complex transactions human agency was required to effect the matching in the systems. Structured Options would be the subject of special prior agreement before being concluded however so that DBAG would be aware of at least the main terms before giving its approval, but the same provisions of Clause 3 and Clause 7 applied nonetheless. The Prime Broker was not responsible for the terms of the trades.
It is accepted by DBAG that a core function of Prime Brokers is to clear trades, by matching them, on the basis of information supplied by the Agent and the Counterparty. How this is done is a matter of mechanics and, following email exchanges and/or the supply of short form term sheets, it appears that the normal process involved matching trade confirmations from each party. Where that could not be done, because there was an unresolved issue as to the terms of the trade, no deal would be concluded. As this is a matter of mechanics however I do not see how the test for implication of terms is met, although it seems to me that an implied term to fulfil the core function of a Prime Broker of clearing by matching the Counterparty Transaction and Agent Transaction (on the basis of information supplied) is to be implied because any reasonable person in the position of the parties to the FXPBA would be justified in understanding that there was a term to that effect, without which the Prime Brokerage function could not be fulfilled. As a matter of fact, nothing turns on this implied duty in any event, but the terms of Clause 4 (last sentence) and of Clauses 5(a), (c) and (d) of the FXPBA provide circumstances in which action is said to be contractual notwithstanding any provisions to the contrary in a trade confirmation. The implied term is inconsistent with these provisions and cannot stand.
8(b) Paragraph 38(2)
225. The second implied term alleged by SHI is that DBAG was obliged to ensure (or to take reasonable skill and care to ensure) that Transactions entered into by SHI were booked, valued and recorded accurately in the accounts it maintained in relation to SHI. As a third implied term (in paragraph 38(4A)) it is further alleged that DBAG was obliged to inform SHI of any inability or failure to book and/or record and/or value accurately any of SHI’s transactions in its accounts or recording systems. I am unable to see how these terms can properly be implied either, since the additional functions which DBAG actually performed, and represented that it would perform, in accordance with its promotional literature were reporting functions, rather than internal functions of booking and recording which led to that reporting of trades and valuation of trades. Moreover, if there was such a duty it would be an oddity to be under an obligation to give notice of any breach and, if there was no such duty, a failure to book or record would again be irrelevant. Booking, as such, was an internal matter for DBAG, as Mr Quinn said.
8(c) Paragraph 38(3)
226. The fourth implied term, set out in paragraph 38(3) of SHI’s pleading, is that DBAG would ensure that each FX Transaction was performed by it and the Counterparty strictly in accordance with the terms of such confirmation. This appears to me to be an impossible implication to make. DBAG cannot guarantee the performance of an FX transaction by the Counterparty with whom SHI has brought it into contractual relations. So far as its own performance is concerned, DBAG would inevitably be in breach if it failed to perform the Agent Transaction. No implication of any term is required for this and the alleged implied terms would not cross the mind of a person in the position of a party to the FXPBA.
8(d) Paragraph 38(3A)
227. As to the implied term in paragraph 38(3A), the test is once again not met. Mr Said was, by the Said Letter of Authority, authorised by SHI to trade on its behalf under the FXPBA. The FXPBA in turn authorised SHI to act as DBAG’s agent in binding it to trades with Counterparties. It was always open to DBAG to agree with a Counterparty that a trade conducted by SHI in DBAG’s name, though falling outside the scope of the relevant Counterparty Master Agreement as defined in Clause 3 of the FXPBA, was to be binding on it. The principal can ratify unauthorised acts effected by its agent. Whether such a trade would be binding upon SHI under the FXPBA and/or the Said Letter of Authority is another matter but, even if it was not within his actual or ostensible authority, it could be ratified by SHI. The only question which would matter in such a context was whether or not Mr Said had such authority, which depended upon the FXPBA and the Said Letter of Authority or on ratification by SHI of something done outside that.
38(e) Paragraphs 38(4) and (4A)
Paragraphs 38(4) and (4A) set out the alleged implied term which lies at the heart of these proceedings. The implied term requires DBAG to calculate margin in accordance with the FX ISDA and CSA and to notify SHI of its maximum entitlement to margin in accordance with those calculations and to require allocation of capital in the Pledged Account (presumably by notifying an increase in the Allocated Portion). Alternatively a term is to be implied that DBAG would take reasonable care to ensure that such calculations were carried out and notified to SHI. In the yet further alternative, DBAG was obliged to inform SHI of any inability to carry out complete and accurate calculations of margin in order to call for its maximum entitlement and/or notify an increase in the Allocated Portion.
In my judgment, a reasonable person in the position of either party to the FXPBA would be justified in understanding that, in accordance with DBAG’s promotional literature and the universal understanding of DBAG’s witnesses, DBAG had undertaken a service additional to the core services referred to by Mr Quinn. The
service in question which DBAG must be taken to have agreed to provide was a webbased reporting service which set out details of the trades concluded, the MTM valuation placed upon those trades by DBAG and the margin calculations relating to the SHI portfolio as a whole.
DBAG argued and I accept that the margin provisions of the FXPBA, and the FX ISDA, existed for its own benefit and that it was under no obligation either to calculate or demand margin at any stage. Nor could it be obliged to calculate or demand its full margin entitlement at any particular time. There could however be a significant difference between the provision of information on valuation on a MTM basis on the one hand, which a trader might wish to know for any number of reasons, including the hedging of the item or closing out a transaction, and on the other hand a demand for margin, which constitutes security to DBAG in respect of the exposure revealed by such valuations. Any trader in the ordinary way would want to know how his trades were faring on an MTM basis in order to make decisions about future trading strategy and, in the case of more complex transactions, such as the EDTs or OCTs (as opposed to vanilla trades and forwards, where valuation was straightforward from readily available sources and with readily available tools) a sophisticated computer model might be required to provide MTM, with the only alternative being to seek a price in the market to unwind the trade in question. Although the evidence of Mr Quinn was that all traders kept their own valuations as part of their risk management, there is no reason why a trader might not value a crosscheck on ordinary trades and might look for some guidance in relation to complex trades where a computer model was needed. As will be seen later, Mr Said did not expect this for the EDTs because he knew that DBAG’s systems were not able to value such complex trades and he had no regard for their MTM status in any event, regarding them as buy to hold options. Mr Said did, of course, as any trader would, retain his own trading records and had his own pricing tool which would operate for vanilla trades.
The margin provisions were undoubtedly provisions included in the FX ISDA and CSA, and incorporated by reference in the FXPBA, for the protection and benefit of DBAG alone, so it was entitled but not obliged to call for margin (representing security in relation to potential debts owing by SHI). If however it wished to avail itself of the margin provisions of the FX ISDA and CSA, it was in my judgment obliged to notify SHI of its margin calculations for the portfolio as a whole and, as provided by the promotional literature, the MTM valuation of each trade (which when added would give rise to the Counterparty Net Open Position in respect of each Counterparty, as well as providing an element in the calculation of margin as between DBAG and SHI).
DBAG was clearly expected to provide these web-based services as part of its FXPB arrangements and it was to those services that the promotional literature referred. Mr Said, on his evidence, expected to receive such services for his ordinary trades, however little he relied on them and however well he could manage without them. All the FXPB personnel saw it the same way. Although as a matter of English law the CSA did not require any calculation of margin or notification, save in the event of a demand for a Delivery Amount, the whole system of FXPB as operated by DBAG, and as advertised in order to attract customers, included the web reporting service with these elements. What use the client might make of them was another matter, but
DBAG made it plain that this is what the website would provide. The essence of the additional service was this “reporting”, in the sense of it being made available for access by the customer on the web-based system. The duty was not to record the information but to make it available for use by the customer. Although the FXPBA did not spell out any of the “back office” services which DBAG supplied to SHI, it is the fact that such services were supplied in terms of booking, clearing and settling the vanilla trades, reporting on market values and collateral requirements and the net equity or collateral position as DBAG’s brochures spelt out.
As to the basis upon which that website service would be provided, there is no clue in the FXPBA, as the FXPBA makes no reference to it. Is the implied duty an absolute one, or one of reasonable care in reporting? Whereas, as a matter of English law, to my mind it would go without saying that a Prime Broker who provides services for reward impliedly undertakes to act with reasonable care and skill in the provision of those services, that does not appear to be the stance adopted by New York law, which does not so readily imply duties of reasonable care into contracts, save those of a specified kind. As a matter of commercial good sense, regardless of any notions of “fair dealing”, both DBAG and SHI must have entered into the FXPBA with the expectation that trades would not just be accurately cleared and settled, but reported on the website, together with MTM valuations according to DBAG’s systems and margining in accordance with the agreed terms set out in the CSA. No customer however would expect perfection in the sense of 100% accuracy at all times and it appeared to be common ground between Ms Rahl and Mr Quinn that errors were to be expected. Computer systems being prone to fail, that might go without saying.
The difficulty is resolved because, as Ms Rahl effectively recognised, access to computer systems of this kind invariably involves acceptance of conditions of use. DBAG’s own terms and conditions of use of the GEM appeared on it, and although there is no direct evidence of Mr Said reading them or acknowledging their existence, reference to them was prominent enough to bring them to his attention when accessing the website.
SHI accepted the facts and matters relating to the website set out in Deloitte’s letter of 3rd May 2013. Mr Said used the GEM website where the page loaded immediately after a user had logged onto the site (the “Home page”) contained a summary disclaimer referring to the GEM terms and conditions, with a hyperlink to those terms and conditions. That summary disclaimer also appeared on the “Margin Report – Help” page and the “Global FX Reports” page as well as appearing on downloads in Excel format.
In circumstances where there is an implied term in the FXPBA for the provision of these web based services, and the web based services themselves purport to be provided on the basis of terms and conditions, those terms and conditions must be, like the FXPBA, governed by the law of New York. The very terms of the exclusion set out below reinforced that because, as a matter of public policy, New York law prevents the contractual exclusion of liability for damages which result from a party’s own “wilful misconduct or gross negligence”. Moreover, reference is made to the concept of punitive damages which is not known to English law. It is common ground between the parties that “gross negligence” is different in kind and not just degree from ordinary negligence. It amounts to conduct which evinces a reckless disregard for the rights of others or smacks of intentional wrongdoing.
The following limitation of liability Clause appears in the terms and conditions:
“Except to the extent that liability under any applicable law or regulation cannot be excluded and to the extent of its own wilful misconduct or gross negligence, Deutsche Bank is not liable for loss or damage of any kind whatsoever arising as a result of (1) information published on the Website or (2) any errors or omissions from the Website, including any made in computing or disseminating valuations, and under no circumstances shall Deutsche Bank be liable for any damages whatsoever, whether direct, indirect, punitive special or consequential, that are directly or indirectly attributable to the use of, or the inability to use, the Website, even if advised of the possibility of such damages or if such damages were foreseeable.”
There could to my mind be no implied term which went wider than the terms of use of the GEM system, because that was the service which was advertised as part of the PB services. Whether this is seen simply as an implied term of the FXPBA or as a separate agreement to the use of the GEM web system is neither here nor there, but I have no doubt that an implied term in the FXPBA must be subject to this limitation.
If Mr Said, for SHI, concluded a transaction which fell within the ambit of his authority and that of SHI to bind DBAG, such a transaction should be the subject of reporting on the website, with valuation and margining together with other trades. A failure to carry out such functions would be a breach of contract, but I can see no basis for holding that there was an implied term requiring DBAG to admit to a breach. Despite the willingness of DBAG’s witnesses to accept that the customer should be told if the systems could not handle the trade, that is not a term which a person in the position of either party to the contract would consider was included in it – nor would it be implicit in the agreement, viewed as a whole. In any event, as appears later, Mr Said was told and knew of the limits of DBAG’s systems where the EDTs were concerned.
DBAG’s Pitch Book included Exotic FX options in its array of products which could be administered by the FXPB desk, but the FXPBA made special provision for Structured Options, which as appears elsewhere in this judgment (see Mr Said’s Timeline) were thought of as presenting particular problems for DBAG’s systems and had therefore to be the subject of DBAG’s prior consent in order to be an Accepted Transaction. If, for whatever reason, DBAG could not accurately report such a trade and set out its value on the website and include it in the margin calculations, it had the choice before taking it in of refusing that trade or agreeing special terms in relation to its reporting. An inability to report as required would constitute a very good reason for agreeing special terms. If it failed to do that, accepted the trade and then failed in its reporting duties, there would be a breach of contract but once again, for the same reasons as set out above, I see no basis for an implied term that DBAG should confess to its own wrongdoing.
DBAG accepts that there were failures on its part in a number of different respects. DBAG has, since February 2012, accepted that at no time before 13th October 2008 did FXPB accurately book or otherwise record the indirect EDTs in its RMS system at
the time the trades were executed, that its GEM web reporting did not accurately state MTM valuations in respect of EDTs and that no information was provided to SHI which contained accurate MTM valuations for them. Neither did DBAG perform accurate calculations of VaR in respect of the EDTs and could not do so at the time. In order to do so accurately, DBAG has had to construct a computer model to do that which its own computer systems at the time could not do. Questions of categorisation of the failures could arise here – was there wilful misconduct or gross negligence on DBAG’s part?
8(f) Paragraph 38(4B)
The next alleged implied term is that found in paragraph 38(4B), namely that DBAG would ensure that, before it entered into a direct trade of a “high risk product” with Mr Said for SHI (and in particular any EDT), Mr Vik understood the risk level of the product, alternatively that the bank would take reasonable care to ensure that he understood the risk level of the product. It is said that such a term is to be implied because it was the usual custom and practice of an investment bank to have a policy providing that it would ensure (before concluding a “high risk” transaction of this kind) that an appropriate individual in the client organisation other than the usual trading contact understood the risk level of the product. Some explanation would therefore have to be given to the appropriate individual unless the bank was aware, by reference to previous trading or other circumstances, that this individual already understood the risk.
There are a number of oddities about this implied term and I have no hesitation, when applying the test in New York law for the implication of terms by reference to any alleged custom or usage or otherwise, in rejecting it. In the first place, it is hard to see how the internal policies of the bank, designed for its own protection, in order to avoid counterparty risk or loss of reputation, could give rise to an implied term of the kind alleged.
The “usual custom and practice” put forward by SHI, to which it referred in its pleadings and which is required to meet the appropriate standards of consistency and wide knowledge in the trade in order to qualify as a relevant custom and usage, has changed during the course of the action. The difficulty which SHI has had in defining this custom suggests that it is not “uniform”, “well settled”, “established”, “wellknown” or “notorious”. In the Amended Defence and Party 20 Counterclaim, when the point was first pleaded, SHI stated that the relevant custom of an investment bank, such as DBAG, was to have a policy providing that in respect of high risk products (in particular EDTs) “an explanation of the risk level of the product would be given to, and authorisation for the trading for such products required from, at least one level of seniority, if not two levels of seniority, of decision-maker in the client, above the usual trading contact”. At a Case Management Conference, I ordered a sequential exchange of reports in relation to this alleged custom. Ms Rahl, on behalf of SHI, expressed her opinion that “the custom and practice of a reasonably prudent and competent investment bank” was to have a policy “to require that the bank explained the risks of the instrument to some person at the client who [was] not the immediate trading contact, i.e. a supervisor or a Chief Risk Officer”. That person had to be independent of the trading contract but not necessarily more senior. The reason for such a policy was, she said, to seek to prevent clients from unknowingly incurring large risks “and thus exposing the bank itself to financial, legal and reputational risk” inasmuch as informing additional persons within the customer’s organisation would provide a check or balance that was intended to reduce or control the risk of a rogue trader.
In the Re-Amended Defence and Part 20 Counterclaim, SHI modified the alleged custom and practice, stating that an explanation of the risk level of the product would be given to “an appropriate individual in the client organisation other than the usual trading contact”.
It was following Mr Quinn’s report served in response to that of Ms Rahl, to which I have made extensive reference earlier in this judgment, and after the Joint Experts’ Memorandum prepared by the two experts, that the position changed once again. Ms Rahl agreed in the Joint Experts’ Memorandum that the FX Prime Broker did not provide any advisory service to the client, which ruled out any question of advice on market risk, as a matter of custom of the FX PB market. As Mr Quinn pointed out, customarily, the FX Prime Broker would only know about any transaction after it had been executed by the client and would be bound to accept that transaction if it fell within the ambit of the agreement between them. Moreover, the alleged custom related to authorisation of “the trading” of “high risk” products and FX Prime Brokers would not be considered to be involved in trading since the whole point of the arrangement was that the Prime Broker’s position was “flat”, with the trading risk being taken by the Counterparty and the client respectively, between whom the deal had been concluded. It was agreed between the experts that the FXPB desk had minimal interaction with the FX Sales or Trade Desk and charged low level fees for its operationally focused business. In the Joint Memorandum Ms Rahl, when agreeing that the Prime Broker did not even know what trades the client had entered into before the client informed it, noted that this was not true in the case of direct trades not involving a Counterparty bank or executing broker. Where, for example, SHI, in the person of Mr Said, concluded an EDT with DBAG’s Trade Desk, through Mr Geisker at the Sales Desk, the position was different to trades with Counterparty banks effected as agent for DBAG through the Prime Brokerage mechanism. (As appears elsewhere in this judgment, of the forty one EDTs concluded by Mr Said, six were direct trades concluded with DBAG.)
In Further Information served on 1st February 2013, the day after agreement in the Joint Experts’ Memorandum, SHI set out a proposed further amendment to plead the custom which now appears at paragraph 38(4B) of its current pleading.
This was a volte-face since paragraph 38 of the pleading has at all times referred to the “true construction” of the FXPBA “in the light of the aforesaid factual background” and/or terms to be implied into the FXPBA. That was the issue which Ms Rahl had been addressing in her first report, without reference to any distinction between direct trades with DBAG and indirect trades with Counterparties. In her Report in response to that of Mr Quinn, she was less than candid when saying that the custom and practice she had referred to in her original report applied to direct trades only. In cross-examination she stated that this was obvious and that she had only felt the need to clarify the position in her second report following the experts’ meeting. I regret to say that I have concluded that Ms Rahl was advancing the case for a custom in her first report, in the context of an implied term in the FXPBA (albeit by reference to the surrounding circumstances which included a PWM relationship with DBS), which she then abandoned when faced with Mr Quinn’s evidence about the nature of
FXPB, an area of which she had very little knowledge and no working experience. For the first time, in her Reply Report the custom (which had already changed shape from that originally pleaded, doubtless on her instructions) was said to refer to direct trades which had not previously been mentioned.
In the light of this changed case, DBAG sought the permission of the court to adduce evidence on custom relating to direct trades, as opposed to FXPB and, on being given permission, adduced the report of Ms Mandell who drew attention to the fact that the FX market was self-regulating, despite various bodies opining on best practice for sales and trading as well as market and counterparty risk management. She referred to the OCC and the Federal Reserve’s high level guidelines for FX sales trading in this context. In consequence, banks adopted standardised customer classification, although this sometimes merely split them into those considered sophisticated or unsophisticated, whilst other banks would have a wider spectrum. Classifications might indicate the types of product which a client could trade and the need for independent approval for products outside that range but this was once again variable as between banks.
Most importantly however, she drew attention to the standard Master Agreements most commonly used in FX trading, which developed over the years so that, in 2005, terms such as ISDA were in common usage with the express provisions therein of the kind to be found in part 5, paragraph 1 of the FX ISDA in the present case. In such provisions, the client acknowledges that it has made its own independent decisions as to whether the transaction was appropriate or proper for it based upon its own judgment and is not relying upon any communication given by the other party to the transaction as investment advice or as a recommendation to enter into it. It acknowledges its capability of assessing the merits of and understanding the terms, conditions and risks of the transaction and acknowledges that it does assume risks. The other party is acknowledged as not acting as a fiduciary for or adviser to it in respect of the transaction. (In the present case there are various other provisions in part 5 paragraph 1(b) in which SHI acknowledged that the transactions to which the ISDA refers may at times be volatile and subject to complex and substantial risks that can arise without warning with losses occurring quickly and in unanticipated magnitude. The terms of part 5 paragraph 1 are set out in full in Annex 1 to this judgment). Ms Mandell continued by saying that sales people for sophisticated clients with IFXCO/ISDA type agreements would not customarily question the motivation for trading a particular instrument, especially if the request came from the client (reverse-enquiry) or if it was known to the bank that the client had traded the instrument before (thereby implying that the client was capable of booking and managing the risk).
For exotic derivatives, approval from the bank’s credit department might be required and on occasion that approval might be issued contingent upon certain disclosures or signatures from the client. She stated that ensuring that an appropriate individual in the client’s organisation, other than the usual trading contact, understood the risk level of a product prior to trading, was not in her experience a standard requirement of the sales process. It could conceivably be a specification attached to the authorisation to trade but it would not be customary to require the sales force to do this for every high risk transaction. In entering into a transaction, an FX Sales/Trade Desk would pay regard to internal guidelines and to the agreed documentation with the client. The
desk would be guided both by any internal information regarding the nature of the client and its sophistication and what other transactions the client was known to be transacting with other market counterparties. Consequently, in her view, there was no customary policy that the bank would ensure that a person other than the trading contact of the client understood the risk level of a high risk product (however that was defined) and would explain the risk level to such a person. Nor was there any general or regulatory requirement to do so. Whether it actually took such steps in any particular situation would depend upon all the circumstances, including the nature of the client, its trading history, whether there was any evidence that the trader was engaged in improper practice and the terms of the client documentation.
In her report in reply to that of Mr Quinn, Ms Rahl endorsed the term now pleaded in paragraph 38(4B) of SHI’s current pleading but two paragraphs further on stated that if there was any ambiguity about whether the specified proposed transactions were aligned with the intentions of the client, a reasonably prudent and competent investment bank “would clarify with a senior person at the client which transactions are acceptable before engaging in them.” In cross-examination she agreed that the trading contact might himself be the senior person in question.
The alleged custom thus acquired something of the character of a moveable feast. Whatever form it took from time to time, there was essentially nothing to support it save for Ms Rahl’s assertion. There was no binding regulatory guidance to support it and the industry writings that she relied upon were all aimed at the steps which a bank might take for its own protection and the steps that bank examiners might take when auditing the processes that the bank had put in place for that purpose.
By contrast, the evidence of Ms Mandell, set out in her report and maintained under cross-examination, was entirely convincing. From the perspective of a Sales Desk of a bank, it is irrelevant whether or not the other party has an FXPBA with the same bank. There are sales standards which apply to FX sales persons as to other sales persons in a bank and banks are subject to Federal regulation in the US. The FX market however is self-regulated. She accepted that regulations were principles based but those that existed were all about the risk that the bank assumed, namely the risk of counterparty default and it was in that context that a bank might wish to ascertain the effectiveness of a counterparty’s risk management systems and capabilities in order to preserve the institution’s reputation in the market place by avoiding situations that created unjustified expectations. Steps taken to ensure that counterparties understood the nature and risks inherent in transactions were taken for the bank’s own protection. Some higher risk transactions would be referred to the senior management of the client or the senior sales person at the bank with this in mind. This was particularly the case where the client was unsophisticated. A bank should assess the client’s ability to understand the risk in the relevant product and would want to have a procedure whereby it was confident that the client management understood the nature of the transaction, but often it would be the manager or a senior person who would be executing that transaction and a bank would not go to another individual to inform them of it. There was moreover a higher degree of confidence that the client was prepared to undertake complex higher risk transactions if it had done such trades before, passed them through its systems and confirmed and settled them. It was not therefore standard practice on the sales desk or a standard practice of the sales process to present an outline of the risks associated with complex derivatives to someone
other than the person who was executing the transaction, whether someone away from the Trade Desk or away from the executing client manager, in order to ensure that someone other than the trading contact understood the product in a high risk transaction, however that was defined.
I accept Ms Mandell’s evidence in preference to that of Ms Rahl, because it is inherently more credible, is coherent in the light of ISDA standard terms and the approach to bank policies and because I found her, unlike Ms Rahl, measured and objective in her approach. I find therefore that the alleged custom is not made out in relation to the existence of bank policies in respect of direct trades of “high risk” products.
Moreover, as pointed out earlier, the allegation is not that, as a matter of custom and practice, investment banks owed their clients an obligation to ensure that someone other than the trading contact understood the risk level of the product (or that Prime Brokers owed such duties). The allegation is of a custom and practice about bank policies. It does not follow that, even if banks did have such internal policies, the terms of those policies would be customary terms of contracts with those dealing with the institution. To my mind the notion is inimical to the internal nature of such policies designed for the bank’s own protection and I can see no basis for importing a term into the FXPB or the FX ISDA to the effect alleged.
Moreover, any such customary term in the arrangements between SHI and DBAG would be inconsistent with the express terms of the contracts between them. The FXPBA makes no distinction between “high risk” trades and other trades. The opening paragraph of the FXPBA lists the type of trades which DBAG authorises SHI to transact as its agent. Structured Options, as defined, are more complex transactions and require DBAG’s approval prior to execution by DBAG but, as with the vanilla trades, the initiative for concluding the transactions rests with SHI which deals with Counterparties before approaching DBAG. As pointed out by Mr Quinn, the alleged custom could have no place in the context of an FXPB arrangement.
By the Said Letter of Authority, SHI specifically authorised Mr Said to trade on behalf of the company for the purpose of executing the types of transaction referred to in the first paragraph and, by the second paragraph, expressly acknowledged that DBAG should have no duty to enquire as to the nature of the relationship between SHI and Mr Said, nor as to any restrictions upon his activities in connection with his execution of such transactions. SHI was, by the terms of this Letter, expressly clothing Mr Said with authority to act on its behalf as its trader and agreeing to be bound by transactions executed by him “to the same extent as if we were directly executing such FX or options transactions”. SHI thus equated Mr Said with itself, held him out to be the manager of the business referred to and absolved DBAG from any duty to enquire further about SHI’s interrelationship with him. The requirement to ensure that someone other than Mr Said within SHI, whether Mr Vik, a contact other than the trader or another senior figure, understood the risk level of a product would necessarily involve an inquiry as to the understanding of persons other than Mr Said, an enquiry into the nature of his relationship with SHI as a “manager” or “senior person” and an enquiry which questioned his authority to execute transactions on SHI’s behalf.
Justice Kapnick in her decision of 8th November 2012 held, as a matter of New York law, that the Letter was “a complete defense” in relation to SHI’s claim that DBAG could be liable for Mr Said’s trading activities and she was upheld by the Appellate Division on this point. The best available evidence of the application of New York law to this Letter effectively rules out a possibility of any custom of the kind alleged.
These points hold good in relation to the modified custom which is said to relate solely to direct trades between DBAG’s Sales/Trade Desk and Mr Said on behalf of SHI. Those trades, in practice, were dealt with by the FXPB desk in exactly the same way as a trade by the Counterparty and the implied term is one to be implied, according to SHI, into the FXPBA. However the claim is put, it is nonetheless completely inconsistent with the terms of the FX ISDA and its Schedule, as set out in Annex 1 and referred to above, which applied to every transaction concluded under the FXPBA and to the direct transactions with DBAG as much as the indirect trades. The acknowledgement that SHI had made its own independent decisions as to the appropriateness and propriety of the transaction and was not relying upon any advice or views of DBAG in relation thereto and had determined that the transaction in question was suitable for SHI in the light of its investment objectives, financial situation and level of investment sophistication, when combined with the acknowledgement that transactions may be volatile, subject to complex and substantial risks with losses which occur quickly and in unanticipated magnitude, render the implication of a term of this kind impossible. SHI specifically represented and acknowledged that it needed no explanation or advice for the conclusion of the transactions in question.
Not only is the alleged obligation inconsistent with the terms of the contracts between SHI and DBAG but it is also inconsistent with the parties’ practical construction of those agreements and their performance of the contracts in question. There was no such expectation or understanding on the part of any of the individuals involved, whether at the FXPB Desk, the Sales/Trade Desk or on the part of Mr Said or Mr Vik. As appears elsewhere in this judgment, Mr Vik knew that Mr Said was entering into complex risks, the nature of which was explained to him by Mr Said, but never looked to FXPB or the Trade/Sales Desk for an explanation of the risks involved. Nor was it ever suggested that any such approach on their part was required. Mr Geisker’s evidence was that, in his seventeen years as a salesman, he had never called anyone else other than the trading contact he had in relation to any trade being discussed. He knew of no market practice whereby investment banks, before selling leveraged products such as the EDTs to a client, ensured that an appropriate person at the client, other than the trader, understood the risks involved. He said that it never occurred to him to contact Mr Vik and there was never any discussion of the need for that at the bank. Not only did his evidence support the absence of any custom of the kind alleged but his evidence showed that the course of performance and the parties’ practical construction of the agreements, in so far as anybody had reference to them, was inconsistent with any such customary terms. The salesmen would assume that a Counterparty employing a trader was receiving reports from the trader and it would be difficult and embarrassing to approach anyone else over the trader’s head and would undermine the relationship between them.
8(g) Paragraph 38(4C)
A further implied term alleged by SHI (paragraph 38(4C) of the RRRADC) is that DBAG would not allow any trading in a product not approved through the appropriate NPA processes. This term is alleged to be implied from the fact that it was at all material times the usual custom and practice of an investment bank (and also the policy and practice of DBAG) not to allow any trading in a product not approved through appropriate NPA processes.
It will be recalled that the draft internal Audit Report referred specifically to the absence of any NPA process by FXPB in respect of the EDTs (there referred to as TPFs). There is no dispute that carrying out an NPA process was standard practice in the banking world. It was also common ground between the experts that NPA processes were internal matters, conducted by banks as part of their internal risk management and that the processes instituted and adopted by such banks were confidential to them. The processes varied from bank to bank and there was no identifiable set of standards, according to Mr Quinn, which would have been customarily applied, nor an identifiable set of steps which would be taken by a reasonably prudent and competent investment bank engaged in operating an FXPB arrangement. Whilst therefore a customer might be aware that banks habitually carried out NPA processes, the customer could not know what those processes involved nor how effective they were in ensuring that the bank in question had minimised the risks to it from trading in such products.
The same underlying difficulty arises with this implied term as with the previous implied term in relation to explaining the risk level of a transaction to a person other than the trader. It simply does not follow from the fact that a bank has an internal practice of reviewing a product, before being prepared to trade in it, that a term is to be implied into a contract with a third party that it will not permit itself to enter into a trade with that third party unless and until that process, however effective or ineffective it is, has been carried out. Once again, the test enunciated by the New York law experts is not met, namely that, for a term to be annexed to a contract as a matter of trade usage, it must be shown that the parties to the contract are aware of the usage or that the existence of the usage in the business to which the transaction relates is so notorious that they should have been aware of it. The custom and usage must be such that the parties are taken to have contracted by reference to it.
The effect of SHI’s submission is that DBAG is taken to have promised SHI that it would not allow SHI to enter into trades, in its name, with Counterparties, which had not been reviewed and accepted under the NPA process. This, once again, is directly inconsistent with the terms of the FXPBA and the Said Letter of Authority. Both of those instruments contain their own definitions of the types of trades which Mr Said was authorised to conclude for DBAG on the Counterparty Transaction and/or SHI on the Agent Transaction. If a trade fell within the ambit of the FXPB and the Said Letter of Authority, DBAG was not entitled to refuse to accept such a trade. The fact that DBAG might not have gone through its own internal NPA process is an irrelevance in that regard. It had agreed to accept the transactions set out. The fact that there was a special category of Structured Options which had to be approved by DBAG before becoming an “Accepted Transaction” does not affect the overall position in relation to the alleged implied term. The implied term, as formulated, would require DBAG to refuse to trade a product which otherwise fell within the scope of what it had specifically agreed to authorise Mr Said to conclude on its behalf
and what it had agreed to trade with SHI itself in the offsetting transactions. Whilst the absence of an NPA process in relation to a Structured Option would provide a good reason for DBAG to refuse to accept it, that was a matter for DBAG’s choice, without any fetter being placed upon the exercise of that choice by the terms of the FXPBA.
From whence then can any obligation arise in the context of Prime Brokerage which Ms Rahl accepted did not involve trading? The Prime Broker’s role is much more passive than that but the implied term, as framed, requires DBAG to refuse “any trading in a product not approved through the appropriate New Product Approval processes”.
The uncertainty which surrounds the nature of this implied term requiring rejection of a product which has not been subjected to “an appropriate New Product Approval procedure” is problematic for the implication of a customary term, which, in order to qualify as such, must be uniform, well-settled and well-known or notorious. This is illustrated by Ms Rahl’s own evidence. In her first report she was insistent that the Prime Brokerage desk should conduct an NPA process before handling the new product, regardless of the position of the Trade Desk. In her second report she said that the application of the NPA process to a Prime Brokerage desk could be different from that applied to the Trade Desk but that the Prime Brokerage desk might be able to rely on an NPA process conducted elsewhere in the bank. In the Joint Memorandum with Mr Quinn she agreed that NPA was primarily done for the bank’s protection but stated that it was a standard procedure and clients could reasonably expect that all products entered into by a bank would be subject to an NPA “or some other approval process” that ensured that their operations and systems could handle the products. This point was repeated in her third report where she again referred to “some other similar process” to ensure that an FXPB group could properly handle the trade “without making sustained errors in valuation, margining and reporting” and in cross-examination, having stated that approval at the FX level could not possibly cover FXPB, she accepted that a shorter process, such as a “sign-off by systems and operations and credit” could suffice. Whilst she was insistent that it was standard practice to seek to achieve what was set out in Best Practice Number 51 issued by the New York Federal Reserve in its publication, Management of Operational Risk in Foreign Exchange, she recognised that there were no regulatory requirements for NPA and only guidelines, none of which referred to NPA processes being conducted for the client’s benefit and none of which set out the details of the process to be followed. External regulatory examiners would audit banks for compliance with their own procedures, but no more than that.
What an “appropriate” NPA process involves is not specified and is uncertain. DBAG’s own evidence seeking to establish that there had been an NPA process in respect of the EDTs, by reference to an Approval AB103563 which was completed in 2004 and covered a range of product types with particular economic features, illustrates the uncertainty in any application of the term alleged. SHI maintains that this Approval was insufficient for the EDTs and the reference to the technical security type “KOMultiDualCCY” did not incorporate the EDTs, although Mr Chin’s evidence was that EDTs were treated by DBAG as falling within this category. If this was a New Product Approval, it suggests that the process was inadequate, certainly as far as the FXPB Desk was concerned, even if it was sufficient for the Trade desk,
because of the difficulties involved in booking, valuing and margining the EDTs in DBAG’s FXPB systems. Whether or not this process would have been applicable to FXPB, as opposed to the Trading Desk, only creates another issue of what is “appropriate” as an NPA process.
In circumstances where, even if the trader is aware of an NPA policy applicable to the Prime Broker or the Trading Desk, it cannot know what the process involves or whether the process followed is adequate or inadequate in ensuring the capability of the institution to handle the product in question, the uncertainty surrounding the content of the alleged implied term is reinforced. In the context of internal procedures, designed for the bank’s own protection, and no regulatory requirement for NPA or the manner in which the NPA process is to be conducted, the “appropriateness” of such a process is too uncertain and too unknown to give rise to any customary obligation.
Mr Vik, at paragraph 44 of his first witness statement, expressed his belief, from his own experience, that DBAG must have had NPA policies and operated in accordance with them. The balance of the paragraph which follows on merely states that, whenever he wanted to trade something he had not traded before, there was a ritual of process and procedure which had to be gone through which involved form filling, legal documentation and significant delay. He did not attribute this however to any NPA process. Given my other findings about the unreliability of Mr Vik’s evidence, I am not inclined to accept this evidence from him either, being contained in one sentence alone without anything to support it. There was no evidence from Mr Said that he had any such expectation at all. The evidence of DBAG’s witnesses in relation to NPA was all given in the context of procedures designed for DBAG’s own protection. Mr Quinn’s evidence was that NPA processes were for the bank’s benefit without any client involvement. NPA process was a standard practice but a matter of internal risk management to ensure that banks had proper controls, operation, technology and finance. Moreover, FXPB, which was not involved in the trading of new products, maintained a flat position on business it was bound to take in under the Prime Brokerage documents agreed and would not necessarily know that a product had or had not been approved on a firm-wide basis when taking in a trade. He said that, particularly in the years 2006-2008, Prime Brokers were often presented with new products that a client was trading in an evolving market. If the trade fell within the contractual documents, the broker was bound to take it in and fulfil its role as Prime Broker.
Ms Mandell’s evidence was to the same effect in the context of direct trades, stating that NPA processes were conducted as part of internal risk management.
Ms Rahl stated in her second report that the purpose of the NPA process was to ensure that a bank did not authorise transacting in products for which it was unprepared. The NPA process was intended to assess whether the relevant departments within the bank had the necessary knowledge and expertise to book and account for the product, whether it had the knowledge and expertise properly to risk manage the product and whether transacting in the product would expose it to unnecessary reputational risk. She agreed in cross-examination that the NPA process was an internal process and that different banks would have different processes though the intent and the major components, she said, would be the same. She agreed that banks conducted NPA processes for their own risk management purposes but maintained that it was also “a regulatory imperative”. She appeared to accept that the regulatory guidance referred to in her report focused on what banks did for their own protection without any suggestion that it was being done for the benefit of the client. She said that it was “not done for the clients per se, but it is done to ensure that the product can be properly processed and handled and by definition therefore it tells a client that a bank is not going to enter into an activity that it is ill-prepared to handle properly.” Given however the absence of any knowledge on the part of the customer as to how any NPA process would be conducted and how effective it would be, it could not tell the client anything of the kind.
The position is therefore that, notwithstanding the fact that it was customary for banks to conduct NPA processes in relation to new products and that DBAG had a policy itself to do so, that does not translate into a customary term of an FXPBA not to permit the Agent to conclude trades which have not been through such a process, let alone an “appropriate” process. The evidence does not establish that an obligation to refuse trades which had not been through a Prime Broker’s appropriate NPA process was a customary obligation in the trade to be incorporated into the contract because the parties must be supposed to have made their contract by reference to it. It is inconsistent with the internal, confidential and proprietary nature of an NPA process, inconsistent with the terms of the FXPBA (and the Said Letter of Authority) and with the terms of the FX ISDA and Schedule (as set out in Annex 1) inasmuch as SHI thereby acknowledges that it assumes the risks of the transactions, relies on its own judgment in entering into such transactions and reaches its own decisions as to the suitability of them whilst acknowledging that DBAG is neither a fiduciary nor an adviser on whose advice or communications it relies.
8(h) Paragraph 38(4D)
Paragraph 38(4D) of SHI’s latest pleading alleges a further implied term which was introduced in draft amendments on 1st March 2013. The implied term alleged is that DBAG would not enter into any transactions with SHI or accept any transactions under the FXPBA (alternatively any Structured Options) which the FXPB desk was unable to book, value, record, margin and report accordingly. The alternative formulation is that the bank would take reasonable care to prevent such transactions being concluded. Once again I have to apply the test for implication of terms as set out by the New York professors of law engaged by the parties. The oddity about this implied term is that, once again, it imposes an obligation upon DBAG to prevent SHI from conducting transactions which otherwise fall within the authority given to SHI under the FXPBA and to Mr Said under the Said Letter of Authority. This is not, by any means, the opposite side of the coin from the implied term alleged under paragraph 38(2) of the RRRADC which relates to DBAG’s obligations in relation to trades concluded by Mr Said that fall within the ambit of the authority given to him. The great advantage of the implied term alleged in paragraph 38(4D), from SHI’s perspective, is that it would cast responsibility on DBAG for the transactions which have led to the major losses (the EDTs), all of which resulted from Mr Said’s trading decisions.
As framed, in relation to “any transaction with SHI” and “any transactions under the FXPBA” the implied term is plainly inconsistent with the terms of the FXPBA, for
much the same reasons as I have already set out in relation to other alleged implied terms. If a trade fell within the ambit of the FXPBA (and the Said Letter of Authority) it was a transaction which DBAG had authorised SHI to conclude as its agent (and which SHI had authorised Mr Said to conclude as its agent). DBAG would therefore be obliged to accept such trades as binding upon it in relation to the Counterparty and to conclude an offsetting transaction with SHI on the same terms under Clause 4. For ordinary transactions, other than Structured Options, DBAG had no power to accept or reject and was bound to process them under the FXPBA, if they were within its terms. DBAG was simply not in a position to refuse a transaction which was authorised under the FXPBA and would be in breach if it did so. It was bound to process the authorised trade which Mr Said had concluded, in its name, with the Counterparty and it was bound then to enter into the corresponding transaction with SHI. There simply is no room for the implication of a term to the contrary effect not only allowing but obliging DBAG to refuse transactions on the basis of its inability to book, value, record and margin properly or to report accordingly. It matters not whether this is expressed as an absolute duty or as a duty to take reasonable care to prevent such transactions taking place. Either form of the obligation is inconsistent with the express terms of the FXPBA.
The same issue arises in relation to direct trades with the DBAG Trade Desk. The implied term is framed by reference to the inability of the FXPB desk properly to book, value, record, margin and report to SHI. Direct trades were, in fact, treated by the FXPB desk in exactly the same way as indirect trades, from an operational point of view, although the legal position was, self-evidently, different. DBAG, as one and the same entity, could only have one contract with SHI and there could be no separate Counterparty Transaction and Agent Transaction, since the DBAG FXPB desk could not contract with the DBAG Trade Desk, as a matter of law. Nonetheless this was how the direct trades were treated by the FXPB desk.
The implied term proceeds on the basis that the transactions entered into by the Trade Desk would be booked by the FXPB desk and not by the Trade Desk itself. The Trade Desk was in fact capable of booking and valuing the EDTs by using DB Analytics and margining could be done but not on the VaR basis by the ARCS VaR model, as the FXPBA and the CSA required. DBAG is thus said to be in breach of the implied term in so far as the Trade Desk concluded direct trades with SHI which the FXPB desk could not properly process. It is not suggested that the direct trades are governed by anything other than the FXPBA and thus it is an implied term which is alleged into the FXPBA in relation to what the FXPB desk could or could not do. If a trade had been treated as outside the FXPBA and CSA, then the Trade Desk would have instituted separate processes for trade confirmation, valuation and margining on a different basis, such as trade level margining.
Once it is accepted that the direct trades are as much governed by the FXPBA and CSA as the indirect trades, although the Counterparty Transaction and the Agent Transaction collapse into one transaction only, Mr Said and SHI’s authority to transact direct business must be taken as coincident with their authority to transact indirect trades. Thus the self same inconsistency between the implied term and the contractual authority applies and DBAG could have no choice whether or not to comply with the authority it had conferred because of the inadequacy of its own systems.
As DBAG submits, the matter can be looked at from the opposite perspective. If DBAG did owe a duty to book, value, margin and report on the trades concluded by SHI under the FXPBA, it would be in breach of contract if it failed to do that. How then would the New York law test for implication of terms apply? There could be no basis for the implication of a term that DBAG would prevent SHI from concluding Counterparty Transactions and Agent Transactions or direct trades where DBAG’s failures would mean that it was in breach of contract. If no such obligations were assumed, there would be no basis for implying a term that DBAG would prevent SHI from using the authority given by the FXPBA to conclude trades in circumstances in which DBAG could not carry out various functions that it was not obliged to carry out anyway. In circumstances where I have held that there was an implied term requiring DBAG to report trades and their valuation and the margin requirements of the portfolio as a whole on the website, DBAG would be in breach if it failed to do so though its liability might be restricted by reference to the GEM website terms and conditions. It would still be no basis for any implied term of the kind alleged.
With regard to the more limited implied term requiring DBAG to refuse to accept any Structured Option by refusing approval under Clause 2(iii) of the FXPBA in circumstances where it could not carry out the specified functions relating to those Structured Options, the New York law test can, once again, not be met. The same logic applies as for trades which are not Structured Options. If DBAG owed an implied obligation to book, value, record and margin Structured Options or the more limited obligation that I have held to exist, namely to report on such Structured Options, DBAG would be liable for failing to do so if it has approved such transactions so that they became Accepted Transactions. If it owed no such duties, there would be no basis for saying that it was obliged to tell SHI that it could not do what it was not obliged to do. Either way, there is no basis for thinking that a reasonable person in the position of a party to the FXPBA would be justified in understanding that such a term was included or that such a term was implicit in the Agreement viewed as a whole.
The way in which Clause 2(iii) of the FXPBA worked was to give an entitlement to DBAG, at its option, to approve or not approve Structured Options proposed by SHI. DBAG would not be responsible for any Counterparty Transaction executed by SHI on its behalf unless such approval was obtained and was effective at the time of execution of the trade. In the absence of DBAG’s consent, it would not be bound by Mr Said’s conclusion of the trade with a Counterparty and would not be obliged to enter into any offsetting transaction with SHI. The requirement of consent was therefore a pre-condition to DBAG’s liability for such transactions and obligations in respect of them (and DBAG could impose additional conditions for its consent). The Clause does not involve an obligation to refuse consent in circumstances where DBAG was unable to comply with any enforceable obligations which would arise if it gave its approval. Once again there can be no basis for implying an anterior duty which has the effect of restricting DBAG’s ability to give consent, whether or not the giving of consent for transactions it cannot properly handle gives rise to future breaches of contract.
Again, the same logic applies to a lesser duty to exercise reasonable care to prevent the conclusion of such trades with SHI.
Any reliance upon statements made by DBAG witnesses in their evidence to the effect that DBAG should not accept trades which it could not book, value, margin or report is misplaced, in much the same way as their reliance upon statements that a client should be told of DBAG’s inability to perform these functions in relation to the trades. As a matter of prudent and internal risk management, any employee would rightly think that a bank should not take on trades that it could not properly process in its systems. That however is very different from a contractual obligation, about which they were not being directly questioned and on which they were not qualified to give an answer. A bank which takes in trades that it cannot properly process may be liable for any loss caused by such failure but there is no basis for the implication of an independent obligation not to accept such trades and it is in fact a powerful factor against the suggestion that such a term should be included and was implicit in the agreement taken as a whole.
8(i) Paragraph 38(5)
The last implied term alleged by SHI appears at paragraph 38(5) of the RRRADC. This alleges that DBAG had a duty to act in good faith and a duty of fair dealing in the course of its performance of the FXPBA with the effect that it was “required not to do anything which would have the effect of depriving or injuring the right of SHI to receive any of the intended benefits for which it bargained under the [FXPBA]”.
As set out in the section of this judgment relating to the principles of the New York law of contract, the two professors of New York contract law agreed upon the existence of an implied covenant of good faith and fair dealing and on the appropriate test to ascertain whether or not it has been breached. They disagreed however as to whether the implied covenant created additional freestanding duties in the sense of substantive obligations which were independent of the other obligations imposed on the parties by the contract, whether express or implied. Both agreed that the implied covenant could not give rise to duties which conflicted with the express terms of the contract and I have found that the difference between the professors was not as wide as might at first sight be supposed. I have also set out my conclusions on the point.
Professor Cohen stated that the implied covenant of good faith and fair dealing, as part of the contract, does not create a cause of action that is separate from an action for breach of contract. He opined that, as noted by many New York courts, “The duty of good faith and fair dealing is implicit in the performance of contractual obligations to the extent that a separately stated cause of action asserting breach of that duty is routinely dismissed as redundant.” In the present case, that is exactly what has happened in the New York action. The seventh cause of action alleging breach of the implied covenant of good faith and fair dealing was dismissed by Justice Kapnick on 9th November 2012 as duplicative of SHI’s more specific breach of contract claims on the basis that they arose from the same facts. The other wrongdoings alleged deprived SHI of the intended benefits for which it bargained, including the rights provided by the terms of the FXPBA and FX ISDA. The alleged breach of the implied covenant therefore had no independent substance and added nothing.
That decision was upheld by the Appellate Division, First Department on 2nd July 2013. In my judgment, on the facts here it can add nothing in the context of the breaches alleged in this action either, as appears hereafter.
8(j) A Further Implied Term of the FX ISDA
As can be seen from the earlier section of this judgment dealing with the construction of the FX ISDA there are three provisions relating to collateral which could result in the termination of the FX ISDA. The first was section 5(a)(iii) of the FX ISDA itself which provided that a failure by SHI to comply with its obligations under the CSA, continuing after any applicable grace period had elapsed, amounted to an event of default in respect of which, if continuing, DBAG was entitled to serve a twenty day notice specifying the default and designating an Early Termination Date. The second provision was found in Part 1(i)(iii) of the Part 1 of the Schedule where DBAG could, if for any reason it deemed there were insufficient Eligible Assets held under the Pledge Agreement, give notice obliging SHI to deliver additional collateral, either into the Pledged Account or in some other form satisfactory to DBAG. A failure on SHI’s part to comply would amount to an Additional Termination Event. Thirdly, under paragraph 11(h)(v) of the CSA, DBAG had a right of immediate termination if the value of the collateral in place fell below 100% of VaR.
These provisions, when seen in the light of the other terms to which I have made reference, are inconsistent with a further implied term alleged by SHI. SHI submits that DBAG had an obligation to calculate the Allocated Portion every day and an obligation to ensure that it contained sufficient collateral to support SHI’s FX trading on the basis of 200% of five day VaR. SHI submits that once there was insufficient collateral in the Pledged Account for the Allocated Portion to support the FX trading on a fully collateralised basis, there was no scope for the parties to enter into any further FX trades without a request for additional collateral being made and fulfilled. If a request was made and SHI did not provide the additional collateral, DBAG could close down SHI’s FX trading. In such circumstances of termination, on SHI’s construction, DBAG would retain the Allocated Portion in its entirety because ownership in that had been transferred to it. The effect of this, if DBAG had been performing its calculations correctly, would be that most or all of the losses would be covered even if market movements were sudden, since the margin calculated was intended to cover this very situation and, as soon as the limits were exceeded, the request would be made and met within a day during which time prices would not be expected to fall substantially. SHI further submitted that it had no liability for any outstanding transactions above and beyond the collateral provided. It submitted that “[g]iven the structure of the FX CSA that was set up, where the Bank has no entitlement to demand additional collateral (only the ability to close down SHI's FX trading where a request for additional collateral was refused or where the collateral fell below 100% of the VaR level), and where the Bank retained the Allocated Portion, notwithstanding that it might (and indeed was very likely to) overcompensate the Bank on any loss-making transactions, it must follow that SHI was not to have any outstanding liability for transactions". DBAG’s only recourse was therefore to the Allocated Portion and there would be little sense in the agreed limitation on DBAG’s ability to seek further collateral if SHI had remained liable for the underlying transactions. Accordingly it is said that there was an implied term of the FX ISDA that SHI’s liability on its FX transactions was limited to the Allocated Portion, working on the assumption that DBAG had complied with its obligations with regard to the calculation of margin.
It was accepted by SHI that this would have been an unusual arrangement whereby DBAG assumed responsibility for the mechanism under which payment of SHI’s collateral requirement was satisfied, namely by reference to the Allocated Portion, and was unable to recover any losses that exceeded the balance in the Pledged Account, but this is said to be the result of the provision for the Allocated Portion and the transfer of ownership of it to DBAG.
As already set out in section 7(c) of this judgment, however, ownership in the Allocated Portion remains vested in SHI. Furthermore the express terms of the termination provisions gainsay any implied term just as much as they gainsay any construction of the FX ISDA which has the effect of limiting SHI’s liability to US$35 million. DBAG is given a right of termination in the event that SHI fails to put up additional collateral but this in no way touches upon SHI’s obligation to do so under Clause 2 of the CSA. There would be no purpose served in allowing for time to comply with an obligation before the termination provisions kicked in, if there was no obligation to produce margin over and above the US$35 million. The alleged implied term is inconsistent with the express terms of the contract.
Some of the above alleged implied terms which I have rejected run counter to the express terms of the FXPBA in which they are said to be implied. Others run counter to the essential nature of the FXPB relationship or may be seen as contrary to Clause 12 of it or to Clause 9 of the FX ISDA. Each in the form pleaded fails for the reasons set out.
9. The Principles of the New York law of Tort
293. It is agreed between the parties that the existence of the tortious duties alleged by SHI in paragraphs 38A-38D of the re-re-re-amended Defence falls to be assessed by reference to the law of New York. Those duties are as follows:
“38A. Further or alternatively, the Bank owed a duty of care in tort to SHI in the following respects:
A duty to take reasonable care to ensure that each FX Transaction was booked, valued and recorded accurately in the FX Account.
A duty to take reasonable care to ensure that calculations of the capital required to support SHI’s FX trading were carried out completely and accurately, and to notify SHI of the capital requirements of its FX trading when calculated.
A duty to take reasonable care to ensure that information communicated to SHI in relation to its FX trades and its accounts was in all material respects accurate and complete.
A duty to take reasonable care to inform SHI of any inability or failure to:
book and/or to record and/or value accurately or at all SHI’s transactions in the FX Account of SHI or in its reporting systems, and/or
carry out any, or any complete or accurate, calculations of the capital required in order to allocate capital appropriately, and/or failure to allocate capital in the Pledged Account properly or at all.
A duty to take reasonable care to ensure that, in circumstances when it was proposed to enter into a high risk product directly between SHI (through Mr. Said) and the Bank, particularly a leveraged derivative (such as the Exotic Derivatives Transactions, as referred as referred to below), Mr Vik understood the risk level of the product.
A duty to take reasonable care not to enter into any transaction with SHI, or accept any transaction under the FX PB Agreement, (alternatively, not to accept under the FX PB Agreement any Structured Option (by refusing to give its approval pursuant to clause 2(iii) of the FX PB Agreement)) which the FX Prime Brokerage division was unable to book, value and record accurately in the FX Account or in its reporting systems and in respect of which it was unable to carry out any, or any complete or accurate, calculations of the capital required to support such trading and to report accurately to SHI such capital requirements (or, alternatively, in circumstances where it was unable to do any one or more of these tasks).
38B. The said duty of care arose out of the facts and circumstances pleaded at paragraphs 3 to 14, 16 to 21 and 38(4B) above, in particular SHI relies upon the following matters:
The Bank’s presentation of itself with key values and priorities in relation to risk management and monitoring of risk and in the provision of customised solutions to clients, as set out at paragraphs 5 and 6 above.
The fact (as pleaded at paragraphs 7 and 14 above) that, as the Bank knew, SHI did not have employees or front, middle or back office operations dealing with its investments and so would be reliant upon the Bank for such services, including in particular the services set out in paragraph 7 above.
The Bank at all times held itself out to SHI as being able to provide the prime brokerage service required by SHI, as set out at paragraph 19 above, and was aware that SHI did not have access to the data, models and systems referred to at paragraph 19 above; and the discussion and agreement (pleaded at paragraph 20(2) above) that FX trading would be facilitated by the provision of prime brokerage services.
The way in which SHI managed its risk, as set out at paragraphs 9, 10, 12, 14 and 17 above, including the parties’ reasonable expectations as pleaded at paragraph 10 above, and SHI’s requirements as pleaded at paragraph 17 above (which were discussed with representatives of the Bank, as pleaded at paragraph 18 above).
That SHI never requested nor agreed to any “trading on credit” with the Bank in relation to FX trading, as set out at paragraphs 11 and 12 above, the standard practice (pleaded at paragraph 12 above) to ensure that the amount of capital due in respect of any trades was always in place on a timely basis, and the discussion and agreement (pleaded at paragraph 20(1) above) that SHI’s FX trading would be supported by capital and not by credit.
SHI was always treated by the Bank as a Private Wealth Management client, as set out at paragraphs 4 and 14 above.
It was discussed and agreed, as pleaded at paragraphs 20(3) and (5) above, that Mr Said’s FX trading would be limited as there set out, the amount of capital to be provided by SHI to support SHI’s FX trading was discussed and agreed as pleaded at paragraph 20(4) above, and the Bank and SHI discussed the matters relating to Mr Said’s FX trading pleaded at paragraphs 20(6) and (7) above.
The essential purpose of the FX prime brokerage agreement, as pleaded at paragraph 21 above.
The custom and practice pleaded at paragraph 38(4B) above.
38C. In the premises, the said duty of care arose on the basis:
(1) that the Bank assumed responsibility to SHI for:
booking, valuing and recording accurately each FX
Transaction in the FX Account;
carrying out calculations of the capital required to support SHI’s FX trading completely and accurately;
communicating to SHI information in relation to its FX trade and its accounts that was in all material respects accurate and complete;
informing SHI of any inability or failure to (i) book and/or record and/or value accurately or at all SHI’s transactions in the FX Account of SHI or in its reporting systems, and/or (ii) carry out any, or any complete or accurate, calculations of the capital required in order to allocate capital appropriately, and/or failure to allocate capital in the Pledged
Account properly or at all; and
in circumstances when it was proposed to enter into a high risk product directly between SHI (through Mr. Said) and the Bank, particularly a leveraged derivative (such as the Exotic Derivatives Transactions, as referred to below), ensuring that Mr Vik understood the risk level of the product;
not entering into any transaction with SHI, or accepting any transaction under the FX PB Agreement, (alternatively, not accepting under the FX PB
Agreement any Structured Option) which the FX Prime Brokerage division was unable to book, value and record accurately in the FX Account or in its reporting systems and in respect of which it was unable to carry out any, or any complete or accurate, calculations of the capital required to support such trading and to report accurately to SHI such capital requirements (or, alternatively, in circumstances where it was unable to do any one or more of these tasks). And/or
(2) that (i) it was reasonably foreseeable that if it did not take reasonable care in the respects set out in paragraph 38A above, SHI may suffer loss, (ii) there was a relationship of proximity between the Bank and SHI, and (iii) it was in all the circumstances fair, just and reasonable that the Bank owed the duties of care set out in those paragraphs.
38D. Further or alternatively, the Bank owed a duty of care in tort to SHI (which arose out of the facts and circumstances pleaded at paragraphs 2 to 14 and 16 to 21 and in particular those pleaded at paragraph 20(3A) above, and paragraphs 44A, 45, 62 and 66A below) to inform Mr Vik on behalf of SHI whenever the collateral requirements of SHI’s trading with the Bank were approaching the upper limit of the collateral then available for that trading. The said duty of care arose on the basis:
that the Bank assumed responsibility to SHI for informing Mr Vik on behalf of SHI whenever the collateral requirements of SHI’s trading with the Bank were approaching the upper limit of the collateral then available for that trading; and/or
that (i) it was reasonably foreseeable that if it did not take reasonable care in the respects set out in this paragraph, SHI may suffer loss, (ii) there was a relationship of proximity between the Bank and SHI, and (iii) it was in all the circumstances fair, just and reasonable that the Bank owed the duties of care set out in those paragraphs.”
294. It can be seen that the duties of care alleged in paragraph 38A(1) and (2) are identical to the alternative implied terms set out in paragraph 38(2) and 38(4) of the same pleading. There is no contractual counterpart to the duty to take reasonable care to ensure that information communicated to SHI in relation to its FX trades and its accounts was in all material respects accurate and complete (as alleged in paragraph 38A(3)) but it overlaps with the implied term that I have found to exist in relation to the reporting of trades, valuations and margin on the GEM website. The implied term that I have found is a straightforward obligation to report on the web, subject to the GEM terms and conditions, whereas the duty alleged here is one of reasonable care. The duties alleged in paragraph 38A(4) and (5) are the same as the implied terms alleged in paragraphs 38(4A) and 38(4B) so that, once again, these allegations allege a duty of reasonable care rather than the absolute duty of the implied terms. The duty of care alleged under paragraph 38A(5) draws on the same alleged custom and practice as the implied term alleged under paragraph 38(4)(b) which I have found not to exist.
9(a) Concurrent duties of care and the Economic Loss Rule
I had the benefit of two reports from Professor Catherine Sharkey, two reports from Professor Benjamin Zipursky, their Joint Memorandum of Agreement and
Disagreement and a day of oral evidence in which they were both cross-examined. As recorded in the Joint Memorandum of Agreement and Disagreement, it was agreed that, under New York law, there is a duty of care in law to avoid causing physical bodily injury and property damage, but that such a general duty of care does not apply to cases involving pure economic harm. There are exceptions to this principle but the experts were not at one in relation to tort claims “at the borderland” between tort and contract, although both agreed that there were two New York Court of Appeals decisions which provided a framework for deciding when New York law would recognise a tort claim for economic harm in circumstances in which the parties are in a contractual relationship. Those two decisions are Clark-Fitzpatrick, Inc. v Long Island R.R. Co. 70 NY2d 382 (1987) and Sommer v Federal Signal Corp. 79 NY2d 540 (1992). Within this framework, it was agreed that a pre-requisite for recognition of a negligence claim is the existence of a “legal duty independent of contractual obligations”. There was disagreement as to the effect of three further decisions of the New York Court of Appeals to which I shall refer later, but there was recognition between them that “professionals” have been found to hold such an independent legal duty and that lawyers and accountants qualified as professionals for this purpose.
There was agreement that New York law recognised an “economic loss rule” in products liability cases but disagreement as to whether it extended in the same way to contract cases.
Professor Zipursky’s views suffered from two main defects. First, they ran counter to the decisions of Justice Kapnick and the Appellate Division in the New York Action.
Secondly, they essentially ignored the fact that the parties had chosen to govern their relationship by a series of contracts with detailed terms and conditions which set out the rights and obligations of the parties. If terms could not be implied into those contracts imposing the duties in question there was no scope for tortious duties, absent strong policy considerations requiring the existence of an independent duty. As Professor Sharkey put it: “the New York courts … are extremely wary about imposing extra-contractual duties in contexts where sophisticated parties have set forth their contractual arrangements”. Cases where there was no contract between the parties are of limited assistance when considering the imposition of a duty and Professor Zipursky’s views were largely founded on “near privity” cases where there was a special relationship akin to contract without any applicable terms and conditions governing the parties’ rights and duties.
Here, the alleged tortious duties do not fit with the non reliance clause and exclusion of any fiduciary or advisory role on the part of DBAG in the standard ISDA conditions agreed between the parties. Moreover, the parties agreed detailed terms relating to their dealings to govern their mutual rights and obligations.
In Clark-Fitzpatrick (ibid) it was undisputed that the relationship between the parties was defined by a written contract which provided for the improvement of the Long Island Railroad by the addition of a second track and that this contract provided for project design changes with appropriate adjustments in compensation. The court concluded that the two causes of action sounding in negligence were properly dismissed stating that:
“It is a well established principle that a simple breach of contract is not to be considered a tort unless a legal duty independent of the contract itself has been violated … This legal duty must spring from circumstances extraneous to, and not constituting elements of, the contract, although it may be connected with and dependent upon the contract.”
The court went on to say that the plaintiff had not alleged the violation of a legal duty independent of the contract but had alleged that the defendant failed to exercise due care in designing the project, locating utility lines, acquiring necessary property rights and informing the plaintiff of problems with the project before construction began. This was alleged to be gross negligence but the court held that each of these allegations was merely a restatement, albeit in different language, of the implied contractual obligations asserted in the cause of action for breach of contract. Furthermore, the damages allegedly sustained as a result of the breach of the duty of due care were clearly within the contemplation of the written agreement as indicated by the design change and adjusted compensation provisions in it. It was said that “merely charging a breach of a duty of due care, employing language familiar to tort law, does not, without more, transform a simple breach of contract into a tort claim.”
In Sommer (ibid) the plaintiff claimed against a defendant fire alarm company for extensive fire damage to his property as a result of the defendant’s failure to transmit a fire alarm to the fire department. This had led to extensive property damage but the defendant relied upon exemption and limitation of liability clauses in the contract, arguing that its liability should be limited to the sum of US$55.50. As a matter of New York law, such a clause would be effective in respect of negligence, but not in
respect of gross negligence, as had been pleaded. As Professor Sharkey testified, the decision in Sommer is critical in dealing with what was described as “the borderland between tort and contract” and as setting several guide posts for separating tort from contract claims. It was maintained by the defendant that the plaintiff was restricted to claims under the contract. The court referred to the basis of the claims in tort and to the duties imposed by law as a matter of social policy. By contrast, contract duties arise from the parties’ consensual undertaking. The borderland situations most commonly arise when the parties’ relationship initially is formed by contract but a claim is then made that the contract was performed negligently. That was the position in Sommer. There was no duty owed prior to the contract being made but, once made, the defendant had obligations which included a duty to make timely reports to the fire department. The effect of such a claim being put in tort or in contract could be significant in the context of statutes of limitation, proof and measure of damages and the availability of exemption clauses or contribution rights. The court set out the following relevant guideposts:
“[1] A tort may arise from the breach of a legal duty independent of the contract, but merely alleging that the breach of a contract duty arose from a lack of due care will not transform a simple breach of contract into a tort. …
A legal duty independent of contractual obligations may be imposed by law as an incident to the parties' relationship. Professionals, common carriers and bailees, for example, may be subject to tort liability for failure to exercise reasonable care, irrespective of their contractual duties. …. In these instances, it is policy, not the parties' contract, that gives rise to a duty of care.
In disentangling tort and contract claims, we have also considered the nature of the injury, the manner in which the injury occurred and the resulting harm … In Bellevue, we rejected plaintiff's attempt to ground in tort a claim that defendants supplied defective floor tiles, noting that the injury (delamination of tiles) was not personal injury or property damage; there was no abrupt, cataclysmic occurrence; and the harm was simply replacement cost of the product. Thus, where plaintiff is essentially seeking enforcement of the bargain, the action should proceed under a contract theory.”
With those guideposts in mind the court held that the plaintiff was not limited to a claim for breach of contract but might also claim for breach of tortious duty. In the light of the guidelines the court highlighted the nature of the services to be performed in the light of the scheme of fire safety regulations which applied to buildings in New York. Fire alarm companies were said to perform a service affected with a significant public interest where failure to perform that service carefully and competently could have catastrophic consequences. The conclusion that a tortious claim was available therefore rested on the special relationship of the parties in the context of the services to be rendered and the public interest policy which required a duty of reasonable care to be imposed that was independent of the defendant’s contractual obligations. The conclusion was also held to rest upon the manner in which the injury arose and the resulting harm, both of which were said to be typical of tort claims, in contrast to the delamination of tiles in the Bellevue case and the damages claim in Clark-Fitzpatrick, which amounted to no more than the benefit of the contractual bargain. In Sommer what was sought was recovery of damages for a fire which had spread out of control – the sort of “abrupt, cataclysmic occurrence” that had been referred to in Bellevue.
Two further decisions of the New York Court of Appeals were relied on by Professor Sharkey, namely New York University v Continental Insurance Co. 87 NY2d 308 (1995) and Abacus Federal Savings Bank v ADT Security Services Inc. 18 NY3d 675 (2012). She considered that the decision in AG Capital Funding Partners LP v State Street Bank and Trust Company 896 NY2d 61 (2008) was a particular decision relating to the position of Indenture Trustees under the Federal Indenture Trustee Act and did not reflect any different principles from those referred to in Clark-Fitzpatrick, Sommer and the other two decisions upon which she relied. There was a difference of view between her and Professor Zipursky on these authorities.
I found Professor Sharkey’s explanation of the principles which underlay the authorities entirely convincing. Whereas Professor Zipursky effectively maintained that the criteria which governed the nature of a relationship which would support a third party negligent misrepresentation case applied to a negligence claim between two parties to a contract, the authorities did not support him in this. It is not enough to establish the requirements which obtain for a negligent misrepresentation case on the part of a non-contractual claimant in order to found a liability in negligence for carelessly performing contractual duties, whether they be express or implied.
In the New York University action it was alleged by a plaintiff insured that its insurer had failed adequately to investigate the claim and failed to renew the policy, in violation of the insurance law of New York. It was held that this amounted to no more than a claim based on the alleged breach of the implied covenant of good faith and fair dealing and the use of tort language in the pleading could not change the cause of action to a tort claim in the absence of an underlying tort duty. Allegations of a sham investigation and bad faith practices did not change the position so as to claim punitive damages in tort.
The court held that:
“A tort obligation is a duty imposed by law to avoid causing injury to others. It is “apart from and independent of promises made and therefore apart from the manifested intention of the parties” to a contract … Thus defendant may be liable in tort when it has breached a duty of reasonable care distinct from its contractual obligations or when it has engaged in tortious conduct separate and apart from its failure to fulfil its contractual obligations.”
A tort claim can only arise for breach of a contractual obligation where the very nature of that contractual obligation and a public interest in seeing it performed with reasonable care require it, in accordance with the decision in Sommer.
In referring to Sommer, the court explained that in that case it had been held that the alarm company’s duty, separate and apart from its contractual obligations, arose from
the very nature of its services – to protect people and property from physical harm, whilst noting the catastrophic consequences that could flow from the defendant’s failure to perform the contractual obligations in that case. The fire safety regulations reflected the public interest in the careful performance of the fire alarm services contract. Protecting the fiscal interests of insureds was “simply not in the same league as the protection of the personal safety of the citizens. As compared to the fire-safety regulations cited in Sommer, the provisions of the Insurance Law are properly viewed as measures regulating the insurer’s performance of its contractual obligations, as an adjunct to the contract, not as a legislative imposition of a separate duty of reasonable care …”. The claim therefore could sound in contract alone.
In Abacus a bank sued defendant security services contractors to recover damages for losses incurred during a burglary of a branch of the bank. Abacus sought to recover some US$590,000 cash which had been stolen and the value of property valued at some US$927,000 taken from the safe deposit boxes, together with damages for lost business, lost reputation in the community and punitive damages. Additionally the costs of repair of the vault of some US$85,000 plus additional security costs were claimed. There were therefore claims both for lost property, damage to property and pure economic loss. Distinguishing Sommer, the New York Court of Appeals dismissed the tortious claim saying:
“Finally, we conclude that the complaint did not allege conduct that would give rise to separate liability in tort. Here, the allegations that a breach of contract occurred as a result of gross negligence does not give rise to a duty independent of the contractual relationship (see Clark-Fitzpatrick …; c.f. Sommer [the plaintiff’s breach of contract claim against the defendant fire alarm company may also sound in tort where the defendant’s alleged failure to act with due care affected a significant public interest independent of its contractual obligations]).”
The focus in that succinct paragraph is on the nature of the burglar alarm service to be supplied under the contract, as opposed to the duty to report fires to the fire brigade in Sommer and the lack of any public interest involved which required an independent tortious duty to be owed as a matter of policy. The underlying decision of the Appellate Division was upheld (77 AD3d 431) where the court referred to Sommer, saying that the separate duty arose from the very nature of the services provided there to protect people and property from physical harm, the catastrophic consequences that could flow from the fire alarm company’s failure to perform its obligations with due care and the public interest in the careful performance of the fire alarm services contract. The contrast was drawn specifically between Sommer and the matter under decision because “no public interest is implicated here” and therefore no basis for tort liability arose.
I accept Professor Sharkey’s evidence in relation to the AG Capital decision of the New York Court of Appeals. There is a degree of uncertainty about the exact identity and nature of the contract which existed in that case between the claimant note holders and the defendant which was the Indenture Trustee, whose position was regulated by the Federal Indenture Trustee Act. Claims were made in contract for breach of the Indentures and the Additional Secured Indebtedness Registration Statements [“ASIRS”], for breach of fiduciary duty and for negligence in failing to deliver the Registration Statement to the Collateral Trustee of earlier loan instruments. There is considerable discussion about the position of an Indenture Trustee and reference to a number of authorities where it was held that “prior to default, Indenture Trustees owe note holders an extra-contractual duty to perform basic, nondiscretionary ministerial functions redressable in tort if such duty is breached”. The claims for fiduciary duty were dismissed but it was held that the duty to perform basic non-discretionary ministerial functions did give rise to a tortious duty to act with due care. The contract claims could not run because of the terms of a release that had been agreed but the court decided that there were issues of fact which had to be resolved in relation to the tortious duty established by prior authorities. It was said that there were “issues of fact as to whether State Street, separate and apart from its contractual duty under the ASIRS, undertook and breached a duty of care, ‘connected with and dependent upon the [ASIRS]’, to act in accordance with the ASIRS and the CTA registration requirements to protect plaintiffs’ security rights in the CTA collateral and whether plaintiffs sustained significant losses as a result of this alleged breach.”
Professor Sharkey’s evidence was that, under the Federal Indenture Trustee Act, there was no room for any exclusion of negligence on the part of the Trustee and the policy of the Act was therefore to create a liability to perform basic non-discretionary ministerial functions for the protection of the note holders. Although therefore there was no reference to policy as such, in the decision, once the peculiar position of Indenture Trustees was recognised, no different principle was being enunciated from those set out in the four authorities to which I have already referred.
Professor Zipursky’s evidence was that it was always a question of fact for the court, whether or not there was a “special relationship” sufficient to give rise to a tortious duty of care, regardless of the existence of a contract. It was in that context that he relied upon the authorities relating to negligent misstatement, to which I refer later.
He relied on a Federal decision of the US Court of Appeals, Second Circuit, in
Bayerische Landesbank v Aladdin Capital Management LLC 692 F.3d 42 (2012). There, the claimant was not a party to a contract but was able, by virtue of the decision of the court, to bring a third party beneficiary claim. The question then arose as to whether or not a tortious duty could arise in the context of an allegation of gross negligence in the management of a portfolio on behalf of the investors. The court referred to the need for a legal duty independent of the contract to be violated for a tort claim to arise, citing Clark-Fitzpatrick and New York University. It was held that the claimant could be taken to have alleged material representations which had induced Bayerische to purchase the notes, including a representation of future contractual performance with care and skill. It was held that the claimant had sufficiently established, in order to withstand a motion to dismiss, a legal duty independent of contractual obligations imposed by law as an incident to the parties’ relationship, in accordance with Sommer. Although the duty was assessed largely on the standard of care and other obligations set forth in the contract (to which the claimant was not a party), that duty arose out of the independent characteristics of the relationship between Bayerische and Aladdin in circumstances where the claimant purchased notes linked to the Portfolio that Aladdin was to manage under the contract itself. The duty was connected with and dependent upon the contract but sufficiently sprang from circumstances extraneous to and not constituting elements of it, within the meaning of the expressions used in Clark-Fitzpatrick.
The court however went on to say this was not the end of the inquiry because of the absence of privity. It was therefore necessary to examine the ambit of the duty to third parties which the court then did by reference to the requirements for recognising liability of “professionals” to third parties in the analogous context of negligent misrepresentation claims.
Having therefore applied the Clark-Fitzpatrick and Sommer principles, additional principles were then considered, taken from the negligent misstatement cases to ensure that that there was nothing in those third party misstatement cases which would militate against the application of the principles applicable to contractual parties, arising from Clark-Fitzpatrick and Sommer, to a third party beneficiary claim. There is here no additional principle to those set out in Sommer.
It will be recalled that in Sommer in the context of the second guidepost referring to the potential imposition by law of a legal duty independent of contractual obligations as an incident to the parties’ relationship, the position of “professionals, common carriers and bailees” was given as an example. In such circumstances tort liability for failure to exercise reasonable care would be imposed irrespective of contractual duties. That was said to be a result of policy.
There are a number of authorities which deal with the characterisation of “professionals”. Both the experts agreed that lawyers and accountants were so characterised but there was disagreement as to the significance of this and the ambit of any further “special relationships” which could give rise to the imposition by law of a legal duty independent of contractual obligations.
There are a number of first instance decisions where the imposition of the duty has turned upon whether or not the defendant was to be regarded as “a professional” or undertaking “professional services”. In Robin Bay Associates LLC v Merrill Lynch & Co 7 Civ. 376 (2008) the Southern District Court of New York characterised the defendant as a “placement agent to secure funding” as opposed to “a financial adviser” and concluded that there was no public interest element as was required for liability for professional malpractice in tort, in additional to contractual liability. The loss was purely economic which was at odds with the “long standing New York rule that economic loss is not recoverable under a theory of negligence”. The court decided it need not decide whether financial advisers were subject to professional malpractice suits. The earlier decision in TD Waterhouse Investor Services Inc. v Integrated Fund Services Inc, No. 01 Civ. 8986 (2003) was cited. Professionals “subject to malpractice liability have extensive formal learning and training, licensure and regulation indicating a qualification to practice, a code of conduct imposing standards beyond those accepted in the market place and a system of discipline for violation of those standards.”
In TD Waterhouse (ibid) a firm retained to perform accounting duties was found not to have the status of a professional under New York law because of the absence of most of these features. After citing Clark-Fitzpatrick and Sommer and the need for a significant public interest for the imposition of an independent duty in tort where the injury alleged was economic in nature (as opposed to physical injury or damage to property and/or a cataclysmic occurrence) the court found that there was no room for the negligence claim in the provision of accounting advice and that the services provided were not “professional services”.
A further example is provided by another decision of the United States District Court in Deutsche Bank Securities Inc. v James M. Rhodes No. 06 Civ. 413 (DC) (2008) where the issue was whether an investment bank’s performance of a contract for financial services could give rise to a cause of action for malpractice. Once again reference was made to Clark-Fitzpatrick and Sommer and the principles there set out. It was said that professionals could be subject to tort liability on the basis of an independent legal duty imposed by law as an incident to the parties’ relationship. It was then said that “in such cases it is public policy, not the contract, that gives rise to the duty of due care”, clearly by reference to Sommer once again. It was then pointed out that the New York Court of Appeals had never found financial institutions such as investment banks to be professionals for these purposes but that, on the contrary, courts had found in actions involving the contractual duties of corporations and financial institutions, that an action in negligence would not lie and that only contractual remedies applied. The court found that the parties in the action in question were “sophisticated entities that entered into a time limited financial contract” and that no significant public interest was linked to the provision of investment banking services that would weigh in favour of imposing a tort liability for public reasons. There are therefore three first instance decisions which militate against the imposition of any tortious duties in the absence of a true “professional” relationship.
New York courts have held that in certain specified circumstances financial institutions may assume non-contractual duties if they have the status of “fiduciaries”. This can arise where a financial institution has discretion and authority with respect to a current account, as opposed to merely accepting instructions on a non-discretionary account with occasional advice. The point is made clear in the Federal decision in De Kwiatkowski v Bear Sterns & Co Inc. in the United States Court of Appeals 2nd Circuit 306 F. 3d 1293 (2002). There it was said that a duty of reasonable care would apply to the broker’s performance of obligations to customers with non-discretionary accounts only in relation to the individual transactions undertaken. The claim advanced however presupposed an ongoing duty of reasonable care, giving rise to obligations between transactions. It was held that in establishing a non-discretionary account, the parties ordinarily agreed that the broker had narrowly defined duties that began and ended with each transaction. The court said it was unaware of any authority for the view that in the ordinary case a broker could be held to have an open-ended duty of reasonable care to a non-discretionary client that would encompass anything more than limited transaction by transaction duties. In the ordinary non-discretionary account the broker’s failure to offer information and advice between transactions could not constitute negligence.
The giving of advice was an unexceptional feature of the broker/client relationship but that did not alter the character of the relationship by triggering an on-going duty to advise in the future or between transactions, or to monitor all data potentially relevant to a customer’s investment. There would have to be evidence of a broker undertaking a substantial and comprehensive advisory role of the kind which arises in a discretionary account for such liability to arise. Where there is “complete discretion
and authority over a claimant’s investment account”, a legal duty may be owed independent of contractual obligations as appears in Assured Guaranty (UK) Ltd v JP Morgan Investment Management Inc. 80 A.D. 3d 923 and Ambac Assurance UK Ltd v JP Morgan Investment Management Inc. 88 A.D. 3d 1.
What is plain from the authorities, notwithstanding attempts to argue the contrary, is that the “economic loss” rule which applies to strict product liability under New York law to restrict the end purchaser of a product to his contract remedies and prevent him claiming from the manufacturer for economic loss in tort, operates effectively in the ordinary contractual situation also, to prevent claims in tort for such losses. Such a claim can be put in contract but not in tort.
Although the New York Court of Appeals in Madison Avenue Gourmet Foods Inc. v Finlandia Center Inc. 750 N.E.2d 1097 stated that the rule had no application to a negligence claim for failure to keep premises in reasonably safe condition and to protect from economic loss in the absence of property damage, in King County Washington v IKB Deutsche Industriebank AG, IKB No. 09 Civ. 8387 (2012) the United States District Court for the Southern District of New York pointed out that in practice the principle had been applied broadly in negligence actions.
Professor Zipursky said, in evidence, that he taught his students that there was no general tortious duty not to cause economic loss, to which the exception was “a special relationship”. In King County the court referred to “the economic loss rule” and the “economic loss doctrine” as preventing a plaintiff from recovering in tort for purely economic losses caused by a defendant’s negligence, whether that was founded on a limited scope of duty or otherwise. Recovery of economic loss lay in the nature of breach of contract, as opposed to tort, but the unavailability of a contract remedy does not trigger an exception to the rule and the doctrine may apply where there is no contract at all between the parties. The presence of a contract or a financial transaction that is in the nature of the contract can however be a strong indicator that a plaintiff was not owed a legal duty that was separate and apart from the obligations for which he bargained and obtained in the transaction. Furthermore, the court referred to the limited exception of malpractice which was to be read narrowly to apply to professionals such as attorneys, engineers, accountants or architects with a contract to provide professional services.
In Bocre Leasing Corporation v General Motors Corporation 645 N.E. 2d 1195 (1995) the New York Court of Appeals, in deciding a product liability case by reference to the foundational decision in East River S.S. Corp v Transamerica de la Val (476 US 858) stated that the particular seller and purchaser were in the best position to allocate risk at the time of entering into their contract and that allocation was normally manifest in the selling price. To allow a purchaser to recover in tort for what was in sum and substance a commercial contract claim would be to grant the purchaser more than the benefit of the bargain to which the purchaser had agreed. If the purchaser has not protected itself with warranties in the contract, it should not be permitted to “fall back on tort when it had failed to preserve its … remedies.” This principle is, if anything, more applicable to a direct claim between parties to a contract than to a claim by an end purchaser under a chain of contracts who pursues a claim against the original manufacturer. The judgment went on to say that, because the allocation of risk was fixed by the parties at the time of purchase, the plaintiff should be deemed to have assumed the risk of loss above and beyond his contract
rights. Courts should not therefore later modify the plaintiff’s commercial contractual risks by imposing a belated tort benefit.
In my judgment, on the basis of the evidence that I have heard, the position is clear. Where there is a contract between the parties, a claim for economic loss will be governed by the contract and an independent duty in tort not to cause such loss will not arise unless there are policy reasons to impose an independent duty, such as those which apply in the case of fiduciaries, professionals, common carriers or bailees or where the nature of the services provided, the manner in which the injury arose and the nature of injury and resulting harm require, in the public interest, that a tortious duty be imposed.
None of those criteria are met in the relationship between DBAG and SHI, in the context of FX Prime Brokerage. This was an arm’s-length Prime Broker relationship, where DBAG was not a “fiduciary”, nor a “professional”, offered no advisory services and fulfilled operational functions, whilst allowing SHI to trade in its name. The nature of Prime Broker relationships is set out earlier in this judgment. The nature of the services provided and the way in which loss was caused have nothing to do with physical harm nor with the kind of issues to which tort law is directed and no issues of public policy arise in that context, requiring an extra-contractual duty to be imposed. The complaint is made about the performance of alleged duties under the FXPBA and resultant economic loss – the loss incurred as a result of an alleged breach of the bargain made between the parties. There is no cataclysmic loss of the type to which tort law is addressed. (New York courts have rejected the events of 2008 as constituting a cataclysmic occurrence). There were a series of contracts between DBAG and SHI concluded in November 2006 which governed their inter- relationship with the detailed terms of Part 5 of the ISDA schedule, as set out in Annex 1, showing the absence of any reliance by SHI on DBAG as an adviser or fiduciary in respect of any transaction concluded by Mr Said. Economic losses suffered by SHI are either recoverable under such contracts or are irrecoverable, as a matter of the Law of New York.
On 8th November 2012 Justice Kapnick dismissed SHI’s eighth cause of action for negligence, referring to Clark-Fitzpatrick and Sommer and the guidelines set out in the latter and referring also to the Bayerische decision. The judge referred to the facts pleaded by SHI and its contention that it had detailed the wrongdoings of DBAG as Prime Brokers and private bankers which were independent from any explicit contractual obligation. It contended that the position of a Prime Broker was akin to that of professionals, common carriers and bailees and that the result of DBAG’s alleged wrongdoing was catastrophic losses in an “abrupt, cataclysmic occurrence” represented by the global financial meltdown in September/October 2008. Thus DBAG’s position was said to be like that of the fire alarm monitoring company in Sommer in failing to perform the risk management duties it had undertaken. The judge rejected these arguments and accepted DBAG’s submissions based on Sommer, where the court’s conclusion rested in part on the nature of the injury, the manner in which the injury arose and the resulting harm which were typical of tort claims such as personal injury and property damage. She said that SHI was essentially seeking to recover economic losses sustained under the contract where a plaintiff sought enforcement of the bargain or merely alleged that the breach of contract arose from a lack of due care and no claim lay in tort, notwithstanding the use of familiar tort language in the pleadings. Moreover, she accepted DBAG’s contentions that SHI was attempting to circumvent the court’s earlier ruling dismissing its breach of fiduciary duty claims, which had already been affirmed by the Appellate Division by that time. An allegation of reliance upon DBAG’s superior knowledge and expertise in connection with its FX trading account ignored the reality that the parties had engaged in arm’s length transactions pursuant to contracts between sophisticated business entities that did not give rise to fiduciary duties. Repeating those claims with a negligence label did not alter the fact that the only duties between them were based in contract.
On 2nd July 2013 the Appellate Division, First Department upheld Justice Kapnick’s decision, stating that the negligence claim set out in the eighth cause of action was properly dismissed as duplicative of the contract claims, by reference to ClarkFitzpatrick. The court also held that there was no “showing that the defendant was subject to duties beyond the roughly thirteen written agreements between the parties”.
Whilst of course the pleas put forward in New York by SHI are not identical to those made in the English action, the essential principles which underlie the decision of Justice Kapnick and the Appellate Division are those which I have found to apply here. The best evidence of New York law must be the decisions of the courts there. The pleadings in each action may not be identical but the relationship between SHI and DBAG, the contracts between them and the alleged type of duties owed are the same so that, even though there is no issue estoppel, the decisions of the New York court carry huge weight. They proceeded on the basis of SHI’s allegations, which doubtless put forward its best case. I have proceeded on the basis of the facts established by evidence and have come to the conclusion that the criteria for establishing any exception to the economic loss rule and/or an independent tortious duty outside the contracts have not been met. None of the tortious duties alleged can therefore stand.
9(b) Negligent Misrepresentation
The New York law of tort Professors agreed that the existence of a “special relationship” between parties is a threshold requirement to make out a claim for negligent misrepresentation under New York law. Furthermore, they agreed that the fact that the parties are in a contractual relationship or near contractual privity is insufficient to establish the “special relationship” required for that purpose. It was further agreed that Kimmell v Schaefer 89 N.Y.2d 257 (1996) “is an important negligent misrepresentation case in which the New York Court of Appeals analyses what will satisfy the special relationship requirement”.
There the court stated that, in a commercial context, a duty to speak with care exists when the relationship of the parties, arising out of contract or otherwise, is such that in morals and good conscience the one has the right to rely upon the other for information. The reliance must be justifiable as a casual response given informally does not stand on the same legal footing as a deliberate representation for the purpose of determining whether an action in negligence has been established. The court said
that since the vast majority of commercial transactions are comprised of such casual statements and contacts, it recognised that not all representations made by a provider of services would give rise to a duty to speak with care. Liability for negligent misrepresentation was therefore imposed only on those persons who possessed unique or special expertise or who were in a special position of confidence and trust with the injured party, such that reliance upon the negligent misrepresentation was justified. Professionals, such as lawyers and engineers, by virtue of their training and expertise might have special relationships of confidence and trust with their clients, and in certain situations liability had been imposed for negligent misrepresentation when they had failed to speak with care. In this context reliance was placed upon older authorities such as Ultramares Corporation v Touche 225 NY 170 and Glanzer v Shepherd 233 NY 236.
The court went on to say that the analysis in a commercial case was necessarily different from those cases because of the absence of obligations arising from the speaker’s professional status and that in order to impose tort liability there had to be some identifiable source of a special duty of care. The existence of a special relationship might give rise to an exceptional duty regarding commercial speech and justifiable reliance upon it.
There is therefore an issue of fact as to the nature of the relationship between the parties and whether it is such to justify reliance upon a negligent representation. The court said that a fact finder should consider whether the person making the representation held or appeared to hold unique or special expertise; whether a special relationship of trust or confidence existed between the parties; and whether the speaker was aware of the use to which the information would be put and supplied it for that purpose.
That principle is not at issue between the parties. It is clear, as Professor Sharkey says, that this is a principle which applies to negligent misstatement. It is not one which carries over into other claims of negligence for economic loss although Professor Zipursky sought to draw on those principles in the context of negligence actions, as I have already mentioned.
In the proceedings between SHI and DBAG in New York, on 9th November 2010, the Appellate Division affirmed the first instance judge’s dismissal of SHI’s negligent misrepresentation claim because of the “absence of a fiduciary relationship”. In a terse paragraph, the court upheld Justice Kapnick’s dismissal of SHI’s claims for breach of fiduciary duty, fraudulent concealment and negligent misrepresentation, stating that the reality was that the parties engaged in arm’s-length transactions pursuant to a contract between sophisticated business entities that did not give rise to fiduciary duties. Professor Zipursky pointed out that the lack of fiduciary duty was not enough for dismissal of the misrepresentation claim since the test was “a special relationship”.
Justice Kapnick’s judgment of 10th December 2009 had dismissed the same claims on the basis of the absence of a fiduciary or other special relationship between the parties and it was that decision which was upheld by the Appellate Division, albeit using different words. She referred to an earlier authority which stated that the plaintiff in that case had failed to plead a fiduciary or other special relationship sufficient to sustain the cause of action for breach of fiduciary duty and negligent representations.
She came to the same conclusion for SHI. Whether or not there is such a special relationship is obviously a question of fact but it would be a highly surprising result if this court were to come to a different conclusion from that reached by the Appellate Division of New York, albeit that the conclusion was reached on an interlocutory basis and on a pleaded case that was not identical to that pleaded or established here. (The allegations of misrepresentation in the US action related to SHI’s exposure, DBAG’s collateral requirements, the equity in the FXPB and Pledged Accounts and the amount of SHI’s losses.)
New York law also recognises that the existence of express contractual disclaimers of reliance upon representations defeats that recognition of a “special relationship” upon which a negligent misrepresentation claim might be based. So, if contracting parties expressly agree that they are dealing with one another “at arm’s-length”, New York courts have refused to recognise negligent misrepresentation claims. In HSH Nordbank AG v UBS AG 95 A.D. 3d 185 (2012) the Appellate Division held as a matter of law that the plaintiff could not succeed in asserting a special relationship where “The parties expressly agreed that they were dealing with each other at arm’slength, that UBS was not acting as HSH’s financial or investment adviser and that HSH was not relying (for the purposes of making any investment decisions or otherwise) upon on any advice, counsel or representations … of UBS.”
On the basis of applicable New York law, the special relationship required for liability between commercial parties to a contract does not exist as between DBAG and SHI. Once again reference should be made to the nature of the Prime Brokerage relationship as set out above, which is not affected by the matters pleaded by SHI in support of any basis of liability in tort.
Furthermore, the existence of the FX ISDA and the Schedule, as set out in Annex 1, conclude this issue against SHI in respect of any transactions that SHI says it would not have concluded but for some representation made by DBAG, whether on the GEM web-site or by other means. (In the case of the GEM website, there are further limitations on liability set by the GEM terms and conditions, in any event).
9(c) Damages
As far as damages are concerned, the experts agree that, under New York law, consequential damages in negligence actions must be “actual” and “ascertained with reasonable certainty” as opposed to being remote or speculative. Professor Sharkey’s view is that a plaintiff can recover lost profits only if it can be demonstrated “with certainty” that the lost profits were caused by the defendant’s negligence and if the amount of lost profits can be proved “with reasonable certainty”. Professor Zipursky’s view is that the phrase “with reasonable certainty” applies to both limbs of the issue. Nothing is likely to turn on this.
Furthermore, both parties agree that under New York law, damages for negligent misrepresentation are governed by the “out of pocket” rule which limits damages to “actual pecuniary loss sustained as a direct result of wrong”. Lost profits are therefore not recoverable in negligent misrepresentation actions to recover economic loss.
10. The Alleged Oral Agreements
There are a number of different oral agreements alleged by SHI, both agreements between SHI and Mr Said on the one hand and agreements between SHI and DBAG on the other. These different agreements relate to the types of trade which Mr Said was authorised to conclude on SHI’s behalf and to financial limits imposed upon his FX trading and Mr Vik’s FX trading and to a warning to be given by DBAG as margin limits were approached.
Mr Vik in his statement said that between September and December 2006, discussions took place between him and Mr Meidal on behalf of the bank which included several telephone conversations and two meetings at his wife’s home which Mr Said also attended in part. The agreements which he reached with Mr Meidal were made mainly on the telephone. As set out in his statement or as stated in cross-examination, the following specific points were agreed:
Mr Said’s trading account would be segregated from SHI’s non-FX trading accounts.
SHI’s exposure would be limited to US$35 million and DBAG would have no recourse against SHI in respect of Mr Said’s trading beyond that.
The collateral for Mr Said’s trading would take the form of a pledge to DBAG in a separate and segregated account with DBS.
DBS would provide DBAG with a guarantee of US$35 million to support Mr Said’s trading.
DBAG’s recourse against SHI in relation to SHI’s FX trading (whether through Mr Said or anyone else) would be limited to the amount of capital secured in favour of DBAG in the separate and segregated account, including built-up profits made on FX trading (the PAL).
Mr Said would be permitted to trade only “plain vanilla” FX transactions in the nature, essentially, of FX forwards and options.
Mr Said’s trading would be monitored and reports made through the prime brokerage arrangements, effectively creating a “back office”.
The collateral for SHI’s FX trading would be managed and Mr Vik would be told whenever the collateral requirements were approaching the limit.
If the collateral in the separate and segregated account became insufficient trading could not continue.
There would be a new FX ISDA Master Agreement which would reflect the above points and would govern Mr Said’s FX trading and any other FX trading in which SHI might engage.
Mr Vik’s evidence in his statements, depositions and under cross-examination was that, after the relevant written agreements had been produced at around the end of November 2006, Mr Meidal told him that the agreements implemented the oral agreements.
There was discussion of the Said Letter of Authority and what it permitted Mr Said to do.
Mr Meidal told him that the provision stating that DBAG had no duty to enquire as to the nature of the relationship between SHI and Mr Said or as to any restriction upon his activities did not extend to limitations set out in the Letter itself.
Mr Meidal told him that Mr Said was restricted to trading in spot, tom next, forward foreign exchange transactions and currency options.
There was discussion about the Structured Options wording in the FXPBA;
Mr Meidal said that this referred to options with a condition in them.
Mr Meidal said that it provided an additional protection to SHI as Mr Said would need permission from DBAG to conduct such trades.
Mr Meidal assured him that DBAG would be vigilant and diligent in relation to monitoring Mr Said’s trading.
Furthermore, after all the contractual documents had been signed, he had a further conversation with Mr Meidal in which the latter confirmed all previous points which had been agreed.
These oral agreements are dependent upon the evidence of Mr Vik alone. Although originally it was alleged that the agreements between SHI and DBAG were made between Mr Vik on the one hand and Mr Meidal and Mr Brügelmann on the other, or, in other formulations, Mr Meidal and/or Mr Brügelmann, Mr Vik’s evidence, in his statements, was to the effect that they were made with Mr Meidal alone. That change in SHI’s case, in DBAG’s submission, calls for comment. SHI asked the court to draw inferences from the fact that DBAG did not call its ex-employee, Mr Meidal, to give evidence. Mr Said has, likewise, not been called by either party although SHI maintains that it reached agreement with Mr Said on the self-same matters as it later agreed with DBAG. Of course it is possible for the Court to draw inferences of one kind or another from the presence or absence of witnesses but at the end of the day it is the evidence which is presented to the Court upon which it must primarily base its decision. There are a number of difficulties with Mr Vik’s evidence about the agreements reached:
It is inconsistent with the written contracts which Mr Vik signed with DBAG.
It is inconsistent with, as well as being unsupported by, any contemporary documents.
It is inconsistent with Mr Said’s evidence on deposition.
It is internally inconsistent and has developed incrementally in the pleadings, in deposition and in witness statements.
It is implausible as the long list of items agreed, as set out above, itself suggests. For there to be agreements on all such matters, without any documentary support in the shape of email exchanges between Mr Vik, Mr Said and DBAG, is inherently unlikely.
It is inconsistent with Mr Brügelmann’s evidence, inasmuch as he knew nothing of any such agreements, and would have been expected to know of them, if Mr Meidal had committed DBAG in the manner suggested. Nor did any members of DBS CRM (Credit Department) nor any members of DBAG’s FXPB department have any knowledge of any such arrangement, which, if made, they should undoubtedly have been aware of, as it would have impacted on their roles.
It is inconsistent with the parties’ conduct after November 2006 and in particular the actions and inaction of Mr Vik in paying the margin calls without protest and closing out Mr Said’s trading with substantial premium, way in excess of US$35m.
The oral agreements alleged following execution of the written contract are improbable because of the lack of rationale for them and give rise to the suggestion by DBAG that they are a fabrication with a view to avoiding any argument based upon the entire agreement Clause and the “no oral modification” Clause in section 9(a) and (b) of the FX ISDA.
Mr Vik’s version of events was that he had reached agreement with Mr Said on four fundamental issues by September 20th and that these matters were then agreed with “the bank” thereafter, by which he meant Mr Meidal of DBS, acting on its behalf and on behalf of DBAG also. He said that four terms were specifically agreed with Mr Said which were absolutely critical:
First the capital for his trading would be US$35 million and no more and that was the maximum amount that SHI could therefore lose.
Secondly Mr Said would only enter into plain vanilla trades such as FX forwards and options, although he could combine them in any way he saw fit, provided that the other limits of the remit were complied with.
Thirdly trading would take place in a separate segregated account and be separately collateralised from SHI’s other non-FX trading.
Fourthly, his remuneration would be 10% of his net realised profits.
These matters were, he said, agreed by about 20th September when he gave Mr Said the name of Mr Meidal as his contact at DBS and asked him to liaise with him to discuss the necessary arrangements for a prime brokerage account and the documentation necessary to enable him to commence trading. As appears in relation to the different types of limitation alleged, this cannot be the case. Whilst Mr Vik’s evidence on deposition was that he had first been introduced to Mr Said in May 2006 at a meal at the golf club, and that he vaguely knew Mr Said before this, it was not until September 2006 that they met at Mrs Vik’s house in Greenwich, Connecticut. Mr Said, in his deposition, recalled making a deal with Mr Vik shortly after the end of summer 2006 on a bench by the putting green in Mrs Vik’s estate. Mr Said also recalled sending Mr Vik a proposal in which he suggested that he would do all of
SHI’s FX trading but Mr Vik had rejected this on the basis that he wanted his FX trading to be separate from that of Mr Said. It was on 19th September that Mr Said emailed his former colleagues at CS telling them that he had begun working with Mr Vik in Greenwich.
The documents show that the arrangements reached at that stage between Mr Said and Mr Vik were somewhat loose. Mr Said considered that a Prime Brokerage arrangement was the optimum way for SHI to trade in FX. He therefore approached Mr Meidal on 20th October with that in mind and the matter progressed from there. It was not however until 30th October 2006 that Mr Said proposed to Mr Vik that he should be able to trade a portfolio himself. The email in question can only be read as showing that, prior to this time, what was envisaged was Mr Said sourcing opportunities for Mr Vik to make decisions about trading through a Prime Brokerage arrangement. The email read as follows:
“Alex – I would like to propose that in addition to my role of sourcing opportunities in the currency (and other macro) markets for you I also be able to trade a (obviously much smaller) portfolio directly. We had talked about that briefly before I started and as I look at the markets I think it makes a lot of sense. There will be opportunities that are of the kind you like – pretty long term and with genuine home run potential and I will continue to spend a lot time finding and analysing them. But – there is good money in exploiting smaller and somewhat more medium term (a few months, 35%) type opportunities. But – to get to those you have to be nimble, quick and a little flexible. Also – they do not not often make it onto your radar screen if you are busy with other things.
I think I can do both. In practical terms DB is all but ready with their prime broker set-up and I have pushed them pretty hard on the collateral side where they have now agreed to what I would consider very favourable terms. You do not need to move assets around for we can determine an amount that stays invested as it is, but is earmarked as collateral for “my” account (it cannot be pledged for anything else). Let me suggest that you allocate between 25 and 50mm$ of assets for this – big enough to make a difference to your bottom line over time if I am successful, but small enough not to go overboard – I want to grow into this. I would envisage keeping you closely in the loop on what I do and of course results (daily, weekly, as you wish) and of building a somewhat more diversified portfolio than the core bets we might put on for you at times, which by nature will be very concentrated. … Can we discuss this please?”
This email is inexplicable on Mr Vik’s version of events. There could not, by 20th September 2006, have been an agreement between Mr Vik and Mr Said for Mr Said to conduct a Prime Brokerage account himself at all, let alone to segregated collateral of
any particular size or any limitation on the trades to be done. Prior to 30th October 2006, what appears to have been envisaged was a Prime Brokerage account where Mr Vik would make the decisions on being fed information by Mr Said (“sourcing opportunities”). As appears subsequently in this judgment, the basis of the Prime Brokerage arrangement for Mr Said emerged over a period of time following this email even though Mr Said had, between 18th and 20th October, discussed a sample portfolio with DBAG in New York which had given rise to a 200% VaR figure of US$28 million and an NOP figure of US$50 million. It can be seen that the 30th October email has these figures in mind but it was not until some time in late November that agreement was reached with DBAG on the amount of collateral to be posted and, since the account was to be Mr Vik’s account at all times prior to 30th October, there could be no suggestion of any agreed limitation as to the type of trades. The only one of the four critical elements to which Mr Vik’s evidence referred which could possibly have been agreed by 20th September was Mr Said’s remuneration, fixed at 10% of net profits with the sum of US$30,000 to be paid monthly on account.
There is no record, email, diary entry or report of any business meeting between Mr Vik and Mr Meidal in the period September to December 2006 although there is reference in the bank’s documents to a proposed meeting which was scheduled to take place in November, which may in fact refer to the meeting between Mr Said and Mr Quezada of the FXPB team. It was the practice of DBS employees to make visit reports for the file and, as Mr Vik recognised, it was the bank’s practice to put in writing any agreement reached. Equally, there are no emails between the parties that make reference to any visit or to any agreement of the kind alleged by SHI and emails exist in relation to all other known meetings. If agreements of the kind alleged had been reached, I do not see how they could have been left without a formal signed agreement recording them or at the very least some exchanges between the parties which set them out or referred to them in one way or another. There is a complete absence of any such records.
SHI complains about the absence of disclosure by DBAG of records of DBS’ telephone calls during the relevant period and DBAG makes the same point in relation to SHI’s telephone records. (There were in fact no recorded lines at DBS before July 2007 and all calls thereafter were recorded and retained). SHI complains about the absence of any notes of Mr Meidal, as well as his absence from the witness box. SHI, whose disclosure in some respects was, to my mind, lamentable, suggested that DBAG had failed to disclose documents in DBS’ possession showing Mr Meidal’s schedules for the relevant time. Mr Vik maintained that, despite having three personal assistants who organised his timetable, he had no diaries or schedules which would record where he was at any time. That I cannot accept. None of this helps but the fact remains that there is not a single document passing between the parties or any internal document which evidences any exchange between the parties in relation to the alleged agreements.
Furthermore, Mr Meidal was a relationship manager at DBS. He would not have seen himself as able to commit DBAG and its FXPB desk to anything which varied the terms of the FXPBA, just as Mr Brügelmann would not. As appears from the documents, Mr Vik entrusted Mr Said to negotiate the terms of the FXPBA and did not do so himself. On his own evidence Mr Vik paid cursory attention to the contractual documents, signing them when requested by Mr Said who had asked for
five to ten minutes of his time for that purpose. The idea that in these circumstances Mr Vik, “who was a man for big ideas and large scale strategy”, would descend into the type of detailed discussions necessary to conclude an agreement with Mr Meidal on the list of terms alleged, is far fetched.
Moreover, as a result of late disclosure by DBAG, long after statements had been exchanged, it appeared that Mr Vik and Mr Meidal had actually met at the Frieze Art Fair in London in October 2006 yet, despite Mr Vik’s recall of the other meetings at his house, he had no previous recall of this at all until the document was produced showing that hospitality event, whereupon, for the first time in cross-examination he said that they would have discussed business there and the terms for FXPB trading. The fact that there were records of this (albeit discovered late) highlights the absence of any documentary record of a business meeting of the kind suggested by Mr Vik, in circumstances where every other business meeting with Mr Vik is documented in one way or another.
It is of course no surprise that a witness does not recall events which occurred over six years earlier, let alone with any precision and Mr Vik, in cross-examination, said that his memory was “fuzzy” for that time period. His evidence about these agreements however bears all the hallmarks of being fabricated in order to make a case and, even in the absence of evidence from Mr Meidal, I reject it.
10(a) The Capital Limitation Agreement
There is here alleged to be both an agreement between Mr Vik and Mr Said (the Said Contract) on the one hand and an agreement between Mr Vik and Mr Meidal for DBAG on the other (the Capital Limitation Agreement). The agreement between Mr
Vik and Mr Said, it appears, must coincide with that made between Mr Vik and Mr Meidal because, at least on one formulation, it relates to Mr Said’s authority to transact business for SHI. If there was to be any agreement of this kind with DBAG relating to the operation of the FXPBA and the FX ISDA and the trading under it, the personnel to agree to it would be expected to be found in FXPB and not in DBS. It was the personnel in New York and New Jersey who handled the FXPB account and would therefore be in a position to monitor it. It was these persons with whom Mr Said was in regular contact during the operation of the FXPBA and with whom he was in contact in relation to its terms and operation beforehand. DBS was always once removed from its operation.
There are a number of different formulations of the agreement with Mr Meidal. One formulation is an agreement that SHI’s liability for any losses incurred in respect of Mr Said’s FX trading was limited to US$35 million and DBAG would have no recourse to SHI in excess of that. A second formulation is that Mr Said’s authority was limited to trades which required SHI to pay or would require SHI to pay no more than US$35 million in losses and/or which required SHI to pay no more than US$35 million by way of margin. The formulations thus put forward a limitation on liability on the one hand and a limitation on trading on the other. The trading limit was subsequently reformulated as a limit on Mr Said’s ability to enter into and remain in trades which had that effect. This involved the idea that Mr Said could commit SHI and DBAG to a Counterparty Transaction and Agent Transaction with authority but that at some point, if the margin limits were exceeded, whether by reason of movements in the market in the value of trades or otherwise, there would no longer be authority to remain in them. Finally, as put forward at trial, this later restriction was effectively excised on the authority case, so that a trade was authorised if, at the time it was concluded, it would not have the proscribed effect, even if subsequently it combined with other trades in the portfolio to have that impact. The US$35m limit then ceased to be such a hard and fast bright line. The notion of a trade concluded with authority subsequently becoming unauthorised by reason of market movements or withdrawal of cash from the account creates such obvious difficulties in the context of the Counterparty Transactions binding on DBAG that its impracticality makes agreement to it incredible in and of itself.
The nature of the agreement was elaborated in Further Information, when SHI was asked for an explanation as to the impact of losses and profits. SHI then put forward the case that the US$35 million was reduced by the net realised losses incurred by Mr Said’s trading but that any net realised profits could restore the limit to US$35 million once again but could not increase the limit beyond that. Further Information also qualified the position to say that the trading limit operated by reference to the maximum entitlement of DBAG to margin, in accordance with the contractual documents.
The complexity of this alleged agreement is revealed by the term which SHI alleges should be incorporated in the agreement in its rectification case. It reads as follows:
“i) The collateral which can support Mr Said’s FX transactions is limited to the US$35 million limit. The Bank shall not permit Mr Said (on behalf of SHI) to enter into any FX Transaction if that FX Transaction would cause the Value at Risk for transactions entered into by Mr Said on behalf of SHI multiplied by the Independent Amount Ratio plus the Bank’s Exposure in respect of transactions entered into by Mr Said on behalf of SHI to exceed the US$35 million limit. If the Value at Risk multiplied by the Independent Amount Ratio plus the Bank’s Exposure is, at any time, greater than the US$35 million limit the Bank will ensure that sufficient open FX transactions are closed out such that the Value at Risk multiplied by the Independent Amount Ratio plus the Bank’s Exposure falls below the US$35 million limit. The US$35 million limit is (i) US$35 million, minus (ii) any net realised losses on transactions entered into by Mr Said on behalf of SHI (net realised losses being realised losses on transactions entered into by Mr Said on behalf of SHI less any amount of realised profit Mr Said had made on behalf of SHI that remained available to be used as collateral in support of his transactions (but not such as to increase the limit above US$35 million).
ii) SHI’s liability to the Bank in respect of FX transactions entered into through Mr Said is limited to the sum of US$35 million and the Bank’s recourse against SHI with respect to FX transactions entered into through Mr Said is limited to US$35 million.”
The manner in which this is framed presents difficulties. How DBAG is expected to prevent Mr Said from entering into FX transactions which have the effect in question, when most FXPB transactions are concluded on an automated basis, is unclear. SHI appears to rely on the fact that, until trades are matched between the Counterparty and the Agent, there may not be any binding transaction but there are obvious difficulties in an FX Prime Broker monitoring trades as they are being concluded in order to avoid the margin being exceeded, rather than calculating margin after they are effected. There are even more obvious difficulties in requiring DBAG to “ensure that sufficient open FX transactions are closed out” when the US$35 million margin is exceeded. How this is expected to tie in with a US$35 million recourse limit as between DBAG and SHI is also uncertain, given liabilities that may already have been concluded as between DBAG and the Counterparty by Mr Said.
There is a further restriction alleged, inasmuch as it is said that Mr Said did not have authority to “trade on credit” by which is meant that he had no authority to enter into trades for which margin was not required by DBAG in accordance with its contractual entitlement. Of course, SHI traded on DBAG’s credit, in its name, with Counterparties and, at all times, traded on margin arrangements with DBAG, as security for potential debts. Under the FXPB SHI did not borrow money from DBAG as such and all the collateral provided by way of margin simply represented a contractual figure, which constituted a protection for itself against SHI’s default, based on DBAG’s assessment of any diminution in the MTM valuation of the trades and the estimated costs of closing them down, bearing in mind the degree of illiquidity and “slippage” which might occur.
As explained by SHI or Mr Vik at one point, “trading on credit” was said to mean trading in circumstances where a debtor-creditor relationship was created between the parties on a debt over and above the capital provided in the Pledged Account. It was said that DBAG only had the right to make margin calls if a debtor-creditor relationship existed between DBAG and SHI which was not permitted. The agreement therefore was a variant of the Capital Limitation Agreement and tallied with some of the arguments SHI advanced on construction at one time or another. As part of this agreement and argument, it was said that if DBAG required capital in excess of US$35m for SHI’s FX trading, SHI had no obligation to provide it, but could, in its discretion, opt to do so. If it did not do so, then it was said that DBAG was obliged not to allow the trade in question.
None of this optionality makes any sense in the light of SHI’s express contractual obligations to pay contractual debts and margin:
To pay sums due on trades under the Trade Confirmations on the due date under Clause 2(a) of the FX ISDA.
To pay interest on sums due under Clause 2(e).
To pay the Delivery Amount on demand made by DBAG under Paragraph 2(a) of the FX CSA.
To provide additional collateral under Part 1(l)(iii) of the FX Schedule. v) To pay interest under paragraph 9(a) of the FX CSA.
This agreement, in whatever form it is put, is inconsistent with the Said Letter of Authority, the FXPBA, the FX ISDA and its Schedule and CSA, in which any limit of this kind would necessarily have appeared, had it been agreed.
The Said Letter of Authority set out the authority given to him by SHI to trade in FX and Options Transactions with DBAG but without any financial limit expressed.
The Said Letter of Authority authorised Mr Said to sign and deliver various types of documentation including ISDA Master Agreements, CSA and security interests and other credit support documentation, again without any financial limit.
The Said Letter of Authority expressly exempted DBAG from any duty to enquire further as to any restrictions upon Mr Said’s activities.
Justice Kapnick held that the Said Letter of Authority constituted a complete defence to SHI’s allegations relating to unauthorised trades.
The FXPBA which governs SHI’s authority to enter into Counterparty Transactions on behalf of DBAG and thereby committed DBAG to conclude Agent Transactions with SHI on an identical basis, identified types of transactions which SHI was authorised to conclude but imposed no financial limits by reference to any collateral required.
The FXPBA made provision for financial limits in the shape of the Net Daily Settlement amount and the maximum Counterparty Net Open Position but nothing by reference to the capital allocated by SHI. The contractual limits referred to were capable of amendment by DBAG on twenty business days’ notice.
Paragraph 5 of Annex B of the FXPBA envisaged the possibility of a Ceiling Limit in the FX ISDA, but none appeared there.
The FXPBA specifically provided for the posting of collateral by SHI with respect to its obligations under the FX ISDA, in accordance with the CSA, as opposed to making any reference to a limited liability of US$35 million.
Clause 9(a) of the FX ISDA contains an entire agreement Clause which, as a matter of English law which governs it, takes effect as a binding agreement that the full contractual terms are to be found in it and nowhere else (see Inntrepreneur Pub Co v East Crown Limited [2000] 2 Lloyds Rep 611 at paragraph 7).
There are a series of provisions in the FX ISDA, Schedule and CSA which entitle DBAG to demand additional collateral, as set out earlier in this judgment and three provisions which entitle DBAG to terminate on the basis of inadequate provision of collateral, including paragraph 11(h)(v) of the CSA which provides for an additional termination event irrespective of whether or not additional margin has been requested. These provisions are directly inconsistent with any restriction on SHI’s exposure to the sum of US$35 million or to Mr Said’s authority to conclude trades which breach that limit.
The history by which the FXPBA came into existence does not fit with Mr Vik’s version of events. As appears from the email from Mr Said to Mr Vik on 30th October 2006, referred to above, it is clear that although Mr Said was seeking to set up an FXPBA from September 20th onwards, the idea was that he was to “source opportunities” for Mr Vik and to execute deals on his instructions. On a sample portfolio of spot trades presented to DBAG in New York, Mr Said had obtained a 200% VaR margin figure of US$28 million as against a US$50 million NOP figure. In that email, in which he proposed to Mr Vik that he be permitted to trade a smaller portfolio for his own account (about which they had, he said, talked briefly before he started) he suggested that assets need not be moved around but that “between 25 and 50 mm$” could be earmarked as collateral for his account, so that it could not be pledged for anything else. The suggestion of an allocation of a figure of this size was to be “big enough to make a difference to your bottom line over time if I am successful, but small enough not to go overboard”. He said he would be keeping Mr Vik closely in the loop on what he did and supplying daily or weekly results as Mr Vik wished.
On 13th and 18th October 2006 emails from DBS to DBAG talked in terms of US$50 million collateral and by 15th November Mr Brügelmann, in an email, stated that he did not know the size of the collateral that SHI wished to post. On 16th November 2006 Mr Meidal appears to have been under the impression that US$50 million was indeed the figure that would be pledged to support Mr Said’s trading.
Despite this aspect being a centrepiece of the agreements which Mr Vik says he made with Mr Said prior to 20th September and with Mr Meidal subsequently, the fixing of US$35 million as the figure for collateral seems to have occurred very late on. It was on 28th November that Mr Said emailed Mr Vik to say he had the documents for the FXPB account and that he needed five to ten minutes of Mr Vik’s time that day for him to sign them. The Said Letter of Authority, the Amendment Authority, the FXPBA, the FX ISDA, the Schedule, the CSA and the Pledge Agreement were all dated 28th November 2006 and, in all cases, save the Said Letter of Authority, that date appeared alongside Mr Vik’s signature and appears to have been entered by him. None of these agreements make any reference to the figure of US$35 million as will have been noted above. They refer to no figure at all and the CSA is framed in terms of the Allocated Portion of the Pledged Account, as calculated by DBAG in its sole discretion and notified by it to SHI from time to time. The Pledged Account was named in the Pledge Agreement as account 2011084 with DBS.
On 29th November 2006 Mr Meidal emailed Mr Vik, referring first to a sub-account at DBS to which Mr Vik had sent US$90 million as margin for his Futures Transactions. “The account currently holds the EUR/NOK and the NOK/SEK FX positions and the Argentinean bond positions traded by Klaus (Said).” The email continued:
“As per the legal documents we sent to you, we are suggesting to open a new sub-account for Sebastian Holdings. Klaus would have a Limited Power over this account. We propose that the collateral for the PB FX line would be booked on this new account as well as other potential trades made by Klaus, including the two existing Argentin[e]an bond positions. The reason is to clearly separate Klaus' P&L from other trades.
Klaus assumes that he needs approx. USD 75 million, USD 35 million for the FX line and USD 40 million for other trades mainly in fixed income (incl. the existing Argentin[e]an bond positions).
Please confirm if you agree to transfer USD 75 million or alternatively the entire balance held in the existing sub-account to the new-sub account.”
Mr Vik’s response that day was short and to the point – “ok”.
On December 7th DBS signed the TPMCA which referred to the sums in the Pledged Account as totalling on that date pledged assets with a current loanable value of US$62,933,152. DBS undertook to monitor the loanable value of the pledged assets and to advise DBAG if it fell below US$35 million.
Contrary therefore to SHI’s case, it appears that, as between SHI and DBAG, it was Mr Said who put forward the figure of US$35 million for his FX trading, although he was also looking for US$40 million for other trades and Mr Vik gave his approval to this in his one word 29th November email response to Mr Meidal. US$35 million became the Allocated Portion notified by DBAG to SHI, which never changed throughout the life of the FXPBA. Whilst the basis of margin was changed by agreement between DBAG and Mr Said on behalf of SHI, there was no further earmarking of assets in the Pledged Account to increase the US$35 million to anything larger. The figure was then included in the TPMCA between DBS and DBAG, to which SHI was not even a party.
Mr Said’s evidence in deposition on this point was limited. He stated that there were two things that formed the collateral pool in his mind. Primarily there was US$35 million but “[d]efinitionally whenever you have money … in an account like this, whenever you have made money in this, built a profit in the account, that automatically also forms collateral. So in my mind it was the US$35 million plus whatever profit was in there.” In his re-examination by SHI’s attorneys in his deposition, Mr Said was referred to paragraph 7 of his affidavit of 21st May 2009 in the New York proceedings in which he referred to the sample margin calculation effected by DBAG on the hypothetical portfolio of trades. In that and the two following paragraphs:
He referred to the figure of US$35 million which SHI had agreed to allocate for his trading, stating there that DBAG understood that his trading had to be separate and isolated from SHI’s assets and that SHI was only willing to expose a specific sum in respect of it.
He said that all the trades he did were based on the US$35 million pledged amount and that he understood at all times, as did the bank, that his trading was limited to the specific amount of collateral and no more.
In deposition however, he said that the only discussion he recalled with Mr Vik was after he got the number from Deutsche Bank for the hypothetical portfolio. He said he was not totally certain but he believed that the US$35 million number was his and that he came up with it based on the US$28 million figure mentioned by DBAG “to give it a little bit of leeway”: “I told him that’s what we should put in the account”. When asked further about what he had said in paragraphs 7-9, he said that the affidavit imputed a lot of knowledge on his part as to the structure of SHI which he did not have and that, as he recalled it, it was a very simple discussion where the specimen portfolio suggested US$28m and he thought that US$35m was plenty.
He went on to say that all he was saying in his affidavit was his trading was limited to the specific amount of collateral “which of course is always the case. You can only trade to the degree of your collateral”.
He then spoke of what would have occurred had DBAG reported that the collateral requirements of his trades exceeded US$35 million and the accrued profits and said that in those circumstances he could not have continued to trade, whatever he had wanted to do. “That’s the definition of a collateral call. You either put in collateral, you put in more money or you cut positions. So that would have been impossible. If I think back, however, to the general interaction, the general approach that Alex and I had towards this trading, what I think I would likely have done is thought real carefully how strongly I felt about the positions. If I felt very very strongly, which on some of them may have been possible, may have, I would have gone to Alex to ask and he would have been the arbiter”.
When asked subsequently in the deposition about whether he would have been in the EDTs when the “perfect storm” of October 2008 occurred, if DBAG had correctly reported collateral positions earlier, he said that he would have been in those trades if Mr Vik had agreed to add substantial amounts of collateral, which, given the amounts involved, he was doubtful about.
Yet again, later he said that if there were MTM losses of US$40 or US$50 or US$60 million, he would probably have run with that because he was comfortable that they were only marked to market losses, the market was stable and he had made money elsewhere. If he had seen US$150 or US$200 million, that would have been no longer manageable, but with accurate reporting he would have gone to Mr Vik to see if he wanted to add more collateral.
It is plain from this that Mr Said understood that the US$35 million collateral placed some practical limitation upon the extent of his trading, but did not consider it more than this. Despite the earlier references in his affidavit, which he referred to as drafted by others, he plainly did not have any understanding that there was any trading limit to his authority, nor that the bank’s recourse would be limited in any way by the amount of margin provided. If he had entered into an agreement of the kind suggested with Mr Vik, that would have made its way into the contractual documents which he negotiated with DBAG, since it would, so far as he was concerned, have been a basic restriction on his trading which had to be set out. It was not.
So far from understanding that he was subject to a US$35 million margin limit there are references to him in May and June 2008 telling Counterparties at GS and CS that he had margin available of US$70 million, based, it would seem, on a misunderstanding of 200% VaR. Such an understanding is wholly inconsistent with any agreement between him and Mr Vik as to a fixed collateral limit of US$35 million, let alone a trading limit at that level. (As appears elsewhere, Mr Said concerned himself little with margin requirements in the context of his trading in any event and Mr Brügelmann at the outset and Mr Walsh in October 2008 were asked by him to explain what the constituent figures on the GEM website meant in the context of the overall margin figures.)
The evidence shows Mr Vik’s admitted expectation that the bank would wish to record agreements in writing, his readiness to leave the negotiation of the contractual documents to Mr Said, his lack of checking that the written agreements correlated with the oral agreements he alleged and the perfunctory approach he took towards signing those documents. I have already referred to the form of the original allegations, namely that Mr Vik made these oral agreements with Mr Meidal and/or Mr Brügelmann and Mr Brügelmann’s subsequent disappearance from the picture, as set out in Mr Vik’s evidence, and to the absence of any documentary support for the contents of the oral agreements as alleged, of the meetings in Connecticut or of the telephone calls relied on.
When it came to cross-examination Mr Vik was much less positive than his written statements. In reference to his discussions with Mr Said, when asked whether US$35 million was discussed and agreed with Mr Said as the maximum amount he could lose, Mr Vik replied in the affirmative and then said he would qualify that a little bit inasmuch as “that was the amount allocated to him, so that was all he could trade on and, as a result, that is sort of self-evident that he couldn’t lose more than that. So that is probably a better answer.” He went on to say that the chances of him losing more than US$35 million were remote but he had not had a conversation in which Mr Said had said that there was no way he could lose more than that. He did not remember any specific conversation about what effect Mr Said’s profits would have on the collateral available to support his trading. On then saying that he did have an agreement with DBAG that accrued profits were not available to Mr Said, save to offset losses, he then said he did not remember having a conversation with Mr Meidal or anyone else at the bank on that subject. He stated in cross-examination that his memory was fuzzy in relation to the two meetings which he had said, in his statement, had taken place between him and Mr Meidal in Greenwich, with Mr Said present for part of them, and then suggested that there must have been discussion about the FXPB arrangements when he met Mr Meidal at the Frieze Art Fair in London in October 2008, which was nowhere referred to in his statements and the existence of which had only come to light in late disclosure by DBAG. He agreed that perhaps he had forgotten that he had attended the Frieze Art Fair in London with Mr Meidal and said that a lot of his lengthy witness statements was recollection refreshed by documents seen.
In circumstances in which it is clear from the documents that it was Mr Said who negotiated and reviewed the terms of the FXPBA, Schedule and CSA, with DBAG and DBS, it would be an extraordinary feature of the arrangements if some separate agreement was reached between DBAG and Mr Vik which was not recorded in writing. The emails make clear that Mr Said talked to Mr Quezada and others about FXPB, that Mr Brügelmann was in correspondence with Mr Said about the question of margin, with Mr Meidal copied in and Mr Lay, PWM’s relevant credit officer, also involved. Mr Said sought to obtain the lowest margining terms he could, thereby maximising leverage on existing assets and negotiating DBAG and DBS down from their initial demands to that which appeared in the CSA. There were a number of other people directly involved in the setting up of the FXPB arrangements at both DBAG and DBS.
It is simply not possible for there to have been an agreement of the kind which Mr Vik alleges between himself and Mr Meidal and for neither Mr Said nor Mr Brügelmann or any of these others to have been aware of it, as none of them were. If they had been aware of it, there would be some record. The monitoring of the account would have involved the calculation of the margin in advance of each trade being concluded, and the prevention of trades where they brought about an increase in the margin requirement over and above the limit. CRM would have had to know of this, as would Mr Brügelmann and personnel at FXPB if such activity was to take place.
Both Mr Said and Mr Brügelmann understood, as is obviously the case, that margining by a bank can create a practical limitation on the amount of trading that is carried out. If a bank requires more collateral from its client, the client has the two options to which Mr Said referred, namely to put up more margin or to reduce the trading positions to bring the overall position within the existing collateral provided. It is in this context that not only Mr Brügelmann but other personnel within DBAG and DBS referred to the US$35 million as “risk capital”, “a risk budget”, a “risk limit” or even a “trading limit”. What is absolutely clear, as Mr Brügelmann said in evidence, is that there is no way in which it is possible to guarantee that exposure on FX transactions can be kept within a specified limit. It is in the very nature of FX transactions that the market may move and may do so violently, as happened in October 2008 following the collapse of Lehman Brothers. A trader may conclude transactions which, on a bank’s calculation, for example, give rise to a margin requirement of US$32 million but movements in the market of the order of 10-20% would inevitably take the margin requirement beyond US$35 million, at which point the customer would have to decide which of the two steps referred to above he would wish to pursue. If particular types of trade were complex, illiquid and difficult to mark to market, and if circumstances constituted a one in a century event (such as the market movements in October 2008), a VaR calculation based on a 95% confidence level might well prove seriously inadequate since the circumstances would fall into the remaining 5% not taken into account. A bank’s margin calculations could prove to be insufficient and if all the trades were closed down at this point, losses in excess of US$35 million could be realised.
The probability or possibility of huge losses occurring in such circumstances was remote and was seen to be remote by Mr Said himself, as his deposition recognises. The nature of the EDTs, as appears elsewhere, was that they combined a good probability of relatively small profits against a low probability of very large losses. Mr Said knew that each EDT with daily or weekly fixings over a period of a year could theoretically accumulate very large losses running into hundreds of millions of dollars but never considered that the market was likely to move in such a way that this would happen. As long as the market remained within certain parameters he expected
to make tens of millions of dollars profit on most if not all of these trades. Had they been margined with accurate MTM and VaR in accordance with the contractual methodology, DBAG would have been looking for additional collateral before the events of October 2008 and with that practical limitation on trading, Mr Vik would have been faced with the decision to which Mr Said referred in his deposition. There could and would however have been no guarantee that SHI’s exposure was limited to the US$35 million figure, even though the expectation would be that losses would be unlikely to exceed it by much, if at all.
It is in this context that Mr Vik’s alleged oral agreement has to be seen. SHI says it is a small step beyond the practical limitation of margin for Mr Meidal, on behalf of DBAG, to agree that there would be no recourse to SHI for any losses which exceeded that US$35 million limit in circumstances where DBAG was in a position to keep tabs on the transactions, mark them to market and call for appropriate levels of margin. There could, however, be no business rationale for such an agreement to be reached on the part of anyone at DBAG. It would have been an odd agreement to be made by anyone, but particularly by a relationship manager at DBS, as opposed to a senior figure involved in CRM or FXPB. Nor is it a limit which Mr Vik would have been likely to request since he could not have expected any bank to agree to it. As he well knew and recognised in cross-examination, an agreement as to the amount of collateral is very different from an agreement to a limit to exposure. And he himself, in an email to Mr Said of 7 January 2008, stated he did not think in terms of equity capital allocations and did not want that to be a restriction on Mr Said’s activities if he had strong ideas. A term of this kind could only be seen by an individual at a bank as novel and important and one which would require discussion with others and memorialisation. That did not occur.
Whilst I referred at the beginning of this judgment to the possibility that Mr Vik might have deceived himself into thinking that some such agreement had been made, I do not consider that this can be the case. It is inconceivable that Mr Vik could have thought that he had entered into such an agreement with Mr Meidal which was not recorded anywhere in writing and which ran so counter to ordinary concepts of collateral and liability. The agreement is an invention on his part in circumstances where, by the time the margin calls were made, he was aware from Mr Said that DBAG had not been capable of marking the trades to market and therefore capable of seeking appropriate margin. From 16th October onwards he was requesting details of DBAG’s margin calculations over the preceding months, whilst stating his inability to understand how DBAG had allowed losses of this size to occur when DBAG was assuring him that proper margining had been carried out. DBAG refused to admit what had actually happened and Mr Vik commenced proceedings in New York.
Paragraph 30 of the New York Complaint alleged an agreement to pledge US$35 million and an agreement that the maximum exposure of SHI in respect of the FXPBA trading would be limited to the same amount, going on to set out DBAG’s failure to calculate VaR and MTM for the EDTs. SHI has sought, in this action, to make its point good about an exposure limit by reference to the terms of the FXPBA, FX ISDA, Schedule and CSA, by means of contrived construction arguments. The weakness of such arguments no doubt explains the alternative reliance upon oral agreements based on Mr Vik’s evidence.
Yet, if Mr Vik had, for one moment in October 2008, considered that he had previously agreed with DBAG that SHI’s exposure was limited to US$35 million, there is not the slightest chance that he would have decided on and agreed to an orderly close out of the trades concluded by Mr Said or ever have paid US$511 million by way of margin calls on Mr Said’s FX trading, in order to achieve this. Indeed, on being told of potential margin calls by Mr Said, who told him on 10th October that closing out the trades would cost hundreds of millions of US$, on being told by DBAG through Mr Said of a likely first call of US$40-60 million and on being faced with the first margin call of US$98 million approximately, the very first thing that Mr Vik would have said was that there was an agreement under which SHI’s liability to DBAG was limited to US$35 million. Mr Vik is forceful and demanding. He is not a man who shuns confrontation or who would meekly respond to a bank’s demand for money and pay up, if he thought there was a good basis for refusing. There is not a whisper of any semblance of this oral agreement until January 2009, after payment of the margin calls without protest. I conclude that what Mr Vik has done is to seize upon the bank’s failure to effect margin calculations, to seek to make capital of it and to fabricate an oral agreement with an individual who was once employed by DBS and who may now be sympathetic to his position but who was not, as he knew by the time of his statements, to be called as a witness by DBAG.
10(b) The Pledged Account Limit (the PAL)
In addition to the Capital Limitation Agreement of US$35 million in respect of Mr Said’s FX trading, SHI also alleged a Pledged Account Limit in respect of SHI’s FX trading as a whole (namely that of Mr Said and Mr Vik together). SHI’s case as to this agreement has gone through several versions. The alleged limit is a limit on the total losses that SHI could suffer on Mr Said’s FX trading through DBAG and Mr Vik’s FX trading, once it commenced through DBAG in 2008. This too is a fiction.
Although the argument overlaps with the argument that Mr Vik’s FX trading with DBAG in 2008 was subject to the margin arrangements of the FX ISDA, Schedule and CSA, I here determine only whether there was an agreement that SHI could not lose more that the PAL limit on Mr Said’s and Mr Vik’s trading, if aggregated together.
Many of the same points apply in relation to this alleged limit as for the Capital Limitation Agreement, since it is said to have been made at the same time with Mr Meidal and features as an alternative trading or authority limit on Mr Said’s trading alone. I need not repeat the points about the absence of evidence of meetings or of documentary support for any such limitation, nor of its inconsistency with the written agreements.
It is on its face, in its wider form, a highly surprising allegation. Although, as originally contemplated, Mr Vik anticipated making the decisions which Mr Said would execute under the FXPBA, the account was always intended to be separate from other trading and once Mr Said was given control over the FXPB account with DBAG, Mr Vik continued to conduct his trading in FX through DBS, which, he accepted in cross-examination, was not subject to any trading limit of any kind, and was secured against his general assets placed with DBS. Although he had authority to use it, it was not then anticipated that Mr Vik would use that FXPB account to trade through DBAG nor trade with DBAG as he later did in 2008 outside that account.
The idea that, from the outset in 2006, it was agreed that he and Mr Said would together be limited in their FX trading with DBAG, whether under a Prime Brokerage arrangement or otherwise, to a joint trading limit savours of the absurd. Mr Vik would thus be limiting his own trade for no good reason and doing so by reference to the losses which Mr Said might make, thus restricting his own trading scope considerably. If Mr Said could incur losses up to US$35m, Mr Vik would only be able to incur losses to the extent of the net profits that SHI happened to make and the balance on the Pledged Account. As in fact Mr Vik withdrew US$30m of Mr Said’s profits at the end of 2007 and another US$75m approximately in mid 2008, this would have reduced his room to trade considerably in 2008, well below what he actually did.
Yet, as owner of SHI, there were substantial assets available as collateral for his trading outside the Allocated Portion, both in the Pledged Account and elsewhere with DBS in 2006-2007 and in DBAG from early 2008. Neither Mr Vik nor DBAG could ever have thought that he, the owner of SHI, was not permitted to incur FX losses in trading above and beyond the balance of the Pledged Account plus SHI’s FX trading profit. It is inconceivable that he would have agreed to any such limit, nor agreed that his trading be limited by the amount of collateral supplied for an entirely different and specifically segregated purpose, as he testified, namely for Mr Said’s trading, even with the addition of Mr Said’s trading profits.
Nor, once he started trading FX with DBAG in 2008, could he ever have thought that his large directional trading in FX in 2008 was subject to such a small collateral requirement, let alone such a trading limit. The size of his positions made that impossible, as Mr Vik virtually accepted in cross- examination.
The PAL is represented by the Pledged Account balance plus SHI’s net realised FX profits. It is right that paragraph 10 of the Schedule to the FX ISDA provides that payments of profits on transactions under the FXPBA should be credited to the Credit Support Balance, and that the CSA provides that the Credit Support Balance on any Valuation Date is the Allocated Portion. The CSA refers specifically to the Allocated Portion but this supports Mr Said’s view, expressed in deposition, that it acted as additional collateral for his trading above and beyond the figure of US$35m. As a limit for his trading or as a limit on his authority, it can fare no better than the argument for a US$35m trading limit or limit on his authority but, as a limit on Mr Vik’s ability to trade, it makes no sense at all. Mr Vik accepted that recourse for SHI’s debts other than FX trading was not limited to collateral held by DBAG or DBS.
Following the transfer of most of SHI’s assets from DBS to DBAG in early 2008, specifically for the purpose of providing collateral for Mr Vik’s trading in Equities, F&O, swaps and FX at DBAG on the GPF platform at DBAG, there was no shortage of assets there to support that trading. What would be the point of a trading limit of this kind, from the perspective of Mr Vik, SHI or DBAG?
Despite Mr Vik’s evidence at paragraph 95(b) of his first witness statement that it was expressly agreed between himself and Mr Meidal that the bank’s recourse against SHI in relation to its FX trading would be limited to the amount of capital secured in favour of the bank in the separate and segregated account held by DBS, plus any built-up profits made on FX trading, Mr Vik did not maintain this point in crossexamination. There was no suggestion that the PAL applied whilst Mr Vik was trading FX through DBS. He agreed that it would not apply if and when he commenced trading through CM Beatrice, which was anticipated to happen from May 2008 onwards and actually could have begun at the very time the margin calls were taking place, had Mr Vik not withdrawn from the arrangements made. Mr Vik then said that during the period in which he traded FX through DBAG he did not consider whether the PAL limit did or did not apply. On the following day he said that he did not rely on the PAL and did not even consider it until 2009 when it was brought to his attention by his legal team. The question of the PAL was, he said “really a legal question” as to what agreements governed. He was referred to his affidavit in the New York proceedings of 19th February 2009 where he said that the accounts of SHI opened at the bank’s offices in London (meaning the accounts opened for his trading of Equities, Futures and FX) were unrelated in any way to Mr Said’s FXPB account in New York. He agreed that the argument that his FX trading was subject to the PAL had not occurred to him at that stage. He was then referred to his further affidavit of 6th April 2009 wherein he said that the London accounts had no relationship to the New York FXPB account or Mr Said’s trades at all and agreed that he thought of Mr Said’s accounts as being in New York whilst his FX trading was through the London account. What became plain was that the PAL was a point first thought of by Mr Vik’s English lawyers in the course of oral submission in the Vik Millahue proceeding on 8th March 2010. It thus becomes apparent that what Mr Vik said in his first witness statement about the PAL was simply an invention of evidence in an effort to support a lawyer’s point. There was no evidence from anyone else upon which SHI could rely in support of this alleged oral agreement. Mr Said said nothing that would assist and Mr Brügelmann knew nothing of it, nor did any other DBAG witness. There is no documentary record or reference to the PAL anywhere in the contemporaneous material, whether prior to or following the execution of the FXPB documents at the end of November 2006.
DBAG are justified in pointing to this as a good example of Mr Vik fabricating evidence in support of a lawyer’s contrivance. The extent of that contrivance was also revealed by the various formulations of the point by SHI on 24th March 2011, 22nd June 2012, 3rd August 2012 and 1st March 2013 in SHI’s pleadings, the second of which was reflected in Mr Vik’s first statement and in the term to be inserted in the FX ISDA by way of rectification. SHI makes none of these formulations good as oral agreements.
10(c) The Oral Agreements as to the types of trade
Whilst SHI’s closing submissions state that it is no part of SHI’s case that Mr Said’s authority was more limited than the express terms of the Said Letter of Authority and/or the FXPBA, the construction it puts upon those documents is said to tally with oral agreements made between Mr Vik and Mr Said and between Mr Vik and a representative or representatives of DBAG, as confirmed by the latter. In its pleadings, SHI referred to “the Said Authority Agreement” which it said was contained in and/or evidenced by the two instruments referred to. In Further Information served on 28th July 2011, SHI stated that the Said Authority Agreement was made between September and November 2006 during discussions between Mr Vik, Mr Meidal and Mr Brügelmann and between Mr Said and other representatives acting on behalf of the bank. SHI referred to earlier pleaded references to discussions
which included telephone conversations in October, November and/or December 2006 and two meetings between Mr Vik, Mr Meidal and Mr Brügelmann in about November 2006 at Mrs Vik’s house in Greenwich, Connecticut. The discussions were also said to have taken place at the time of the signing of the Said Letter of Authority. The further information then stated that the agreement was made between Mr Vik and Mr Meidal and/or Mr Brügelmann and/or the individuals at DBAG with whom Mr Meidal and/or Mr Brügelmann liaised in relation to SHI matters and/or who were otherwise concerned with SHI matters on behalf of DBAG and/or was contained in and/or evidenced by the Said Letter of Authority and the FXPBA.
SHI also alleged the existence of “the Said Contract” resulting from discussions between Mr Said and Mr Vik in which it was said to be an express and/or implied term that Mr Said would comply with the limits of authority granted to him by SHI, as expressed in the two instruments in question.
When Mr Vik’s first statement was served, his evidence was that meetings and discussions took place between himself and Mr Said in September and October at his wife’s house. There they discussed a portfolio that would comprise “plain vanilla” trades such as FX forwards and options combined in such a way as to provide appropriate strategies, diversification and hedges which would generate good profits for SHI, but with an exposure limited to the amount of capital that SHI was prepared to provide to support the portfolio. It was clearly understood, Mr Vik said, that Mr Said’s mandate was to be specifically limited to “plain vanilla” transactions and to the management of the amount of capital given to him by SHI in a segregated account to support his trading. It was there, prior to September 20th, that the four critical terms referred to above were agreed.
As set out above, the evidence of Mr Vik in his witness statement was that the “Said Authority Agreement” was made solely with Mr Meidal by reference to the Said Letter of Authority and the FXPBA. In its closing submissions, SHI states that it has always accepted that it is bound by the terms of the written instruments but it is plain from the pleadings and from Mr Vik’s witness statements that it was being said that there was both a prior and a subsequent agreement between Mr Vik and Mr Meidal that Mr Said’s trading authority was limited to “vanilla trades” and that this was what the instruments meant and were agreed to mean.
Furthermore, not only did SHI allege that Mr Said’s authority was so limited (paragraphs 47 and 47a of the RRRADC) but in paragraph 48 SHI contended that it was an express or implied term of the Said Authority Agreement and/or the FXPBA that DBAG would not accept, authorise or permit Mr Said to enter into any trades on behalf of SHI and/or DBAG other than the FX Transactions and Options described in the Said Letter of Authority and/or the FXPBA.
There is, self-evidently, a significant difference between an allegation of limits on Mr Said’s authority and an allegation of an obligation on DBAG to refuse to permit any trade which went beyond such authority.
As I have already said earlier in this judgment, SHI set out the authority given to Mr Said to trade with DBAG in the Said Letter of Authority. It was signed by SHI and by Mr Said and addressed to DBAG. DBAG did not assume any obligations under it at all. It was the FXPBA which set out the authority given by DBAG to SHI to act as
its agent in effecting the Counterparty Transactions, as described, and by which DBAG agreed to enter into contemporaneous offsetting transactions with SHI as Agent Transactions. Neither instrument placed any obligation on DBAG to refuse any type of transaction proposed by Mr Said although DBAG could be at risk if Mr Said traded outside the limits of the authority given to him to transact business on behalf of SHI.
In the Said Letter of Authority, SHI expressly recognised that DBAG would rely upon it “in connection with FX and Options Transactions” which it had authorised Mr Said to conclude in the first paragraph. Furthermore, in the second paragraph it specifically accepted that the conclusion of such transactions and related documentation would be binding upon it, agreeing that DBAG had no duty to enquire beyond that. Any restriction suggested by SHI not encompassed in the words of the first paragraph and the definition of FX and Options Transactions is therefore directly inconsistent with the terms of this Letter. If the Letter, on its proper construction, does not restrict Mr Said to trading in vanilla options, there can be no other basis for any such restriction.
The FXPBA described the trades which are referred to as FX Transactions in the same way as the Said Letter of Authority and also included currency options in the same way as that Letter. It also however expressly referred to “Structured Options”, defining them in terms which, on any view, included non-vanilla options. The only options, which are not Structured Options, according to the terms of the FXPBA, are those which are put or call options without special features and single barrier options. Any other option is a Structured Option. In his deposition, Mr Vik said that his understanding of what “plain vanilla” FX trades meant was “spot, tom next, forwards and currency options”. He said in cross-examination that if asked in 2006 what the phrase “plain vanilla” meant that is what he would have replied but he would not be able to tell by looking at a trade whether it fell into that category or not. This answer is extremely odd because he would be specifying a limit that he did not understand.
At all events, by the terms of the FXPBA, SHI accepted that Mr Said would have DBAG’s authority to enter into Counterparty Transactions which consisted of Structured Options provided that DBAG gave its consent. It further agreed that
DBAG should conclude offsetting Agent Transactions with it on the same terms. Thus SHI agreed to be bound by Structured Options proposed by Mr Said to DBAG, should DBAG agree to them. Any restriction above and beyond the terms of the FXPBA in relation to types of trade is therefore inconsistent with the FXPBA. In particular, any limitation to vanilla trades is directly contrary to the authority given to conclude Structured Options with DBAG’s consent.
Furthermore, an obligation on the part of DBAG not to permit Mr Said to conduct anything other than vanilla trades goes well beyond the ambit of the Said Letter of Authority and the FXPBA which essentially set out the authority given by SHI and DBAG respectively. There is not a hint of any restriction preventing either SHI or DBAG from authorising further trades which fall outside the express authority given.
Indeed the absence of any duty to enquire, as it appears in the Said Letter of Authority, has been construed by the New York court as excusing DBAG from the need to enquire whether Mr Said had any authority above and beyond that given for the type of trades referred to in the Letter itself. Whilst that is not necessarily a conclusion to which this Court would have come because of the reference to FX and Options Transactions in the relevant sentence, the idea that the parties could not authorise Mr Said to bind them to other transactions runs counter to the wording of the instruments, the objective purpose for which it can be seen they were given, and the permissive authority which they set out. The test under New York law for implying such a term is not met. The parties would not have considered such a term to be included and it cannot be said to be implicit in the agreement, taken as a whole.
I have already referred to the Said Contract made between Mr Vik and Mr Said and Mr Vik’s witness statement which referred to the express restriction to vanilla trades being agreed between them prior to 20th September 2006. His evidence in crossexamination was that the phrase “plain vanilla trades” was used in discussions with Mr Said in contradistinction to “structured products” which was a concept with which Mr Vik said he was familiar in 2006. He said that he doubted if he had discussed the meaning of the phrase “plain vanilla” with Mr Said but that its meaning was well understood, as he had described it in his deposition. He understood that Mr Said, in his deposition, had said that the distinction was not obvious because different people had different ideas of what a vanilla trade was but his own understanding corresponded with the terminology used in the Said Letter of Authority and the FXPBA.
In his deposition, Mr Said had said that no limitation as to any type of trading was imposed upon him by Mr Vik and that he had no recall of any discussion with Mr Meidal as to the type of trading he anticipated doing. He would have had second thoughts about taking on the engagement, if that had been the case. The very object of him being taken on was to increase the width of the products traded by SHI. As appears elsewhere in this judgment, his evidence on deposition was that he described the nature of types of Structured Options (the exotic trades – the EDTs and OCTs) with Mr Vik before he entered into them. He discussed the risk characteristics of each of the different types of trades with Mr Vik and specifically recalled discussing the risk on the first EDT prior to entering into it in February 2008, at some point after discussing it with Mr Chapin of CS when in his car driving down to Newport. He would discuss matters with Mr Vik when they were in the office together and by telephone and email. Anything new would be passed by Mr Vik with an explanation of the basic pay-off characteristics and the risk but once he had done a structure two or three times he would not refer back to Mr Vik again. The email exchanges in 2008 particularly show this to be the case. There is no way in which Mr Vik could have read these emails without appreciating that Mr Said was engaged in trades which would not, on any view, be considered vanilla.
The documents both prior to and subsequent to the FXPBA do not bear out Mr Vik’s evidence. Although neither Mr Said nor Mr Meidal gave evidence before the court and there is a vacuum with regard to the records of telephone calls between Mr Vik and Mr Meidal, both on Mr Vik’s side and that of DBS, it would be highly surprising if there was not some documentary record in DBS of an agreement made by Mr Meidal which differed from the written agreements in the manner suggested. Exchanges between Mr Said and Mr Meidal and Mr Meidal and those responsible for preparing the contractual documents reveal no understanding on anybody’s part that there was a limitation to vanilla trades. Mr Brügelmann had never heard of such a limitation and, as Mr Meidal’s assistant, he could be expected to know of it if any such agreement was reached. I accept his evidence that he knew of no such limitation
and that, had any such agreement been reached, Mr Meidal would have told him. He did not see how Mr Meidal could have gone off on trips to the USA, made agreements with Mr Vik and come back without recording or telling anyone of the agreements reached. It was standard practice for DBS employees to make visit reports for the file but there were none for any trip at the relevant time. Similarly, if agreements were reached on the telephone as Mr Vik suggested, a report should have been made and Mr Brügelmann would surely have been informed, regardless of that. Mr Vik customarily dealt with Mr Brügelmann by email, sent on his Blackberry because he was constantly travelling and Mr Brügelmann thought that he dealt with Mr Meidal in the same way. Yet of such oral agreements above and beyond the written agreements, there was no trace.
It was on 20th September 2006 that Mr Said emailed Mr Meidal saying that he would be originating and executing trades and investments in the currency and fixed income markets for Mr Vik and that he wanted to set up a Prime Brokerage Agreement which allowed him to trade with one collateral posting and which gave him good back office and reporting functionality. He asked to be put in touch with the right people in the currency/Prime Brokerage area. On 22nd September Mr Meidal emailed Mr Vik proposing an interim arrangement until a Prime Brokerage Agreement for fixed income and FX trading was finalised. A limited power of attorney was signed, giving Mr Said authority to trade with a position limit of EUR 300 million in the interim. In this interim period and initially under the FXPBA, Mr Said referred every decision to Mr Vik for his confirmation before concluding a trade.
On 2nd October 2006 Mr Quezada sent Mr Meidal and Mr Brügelmann an FXPB Pitch Book and Web Report PDF for Mr Said to review. That Pitch Book, which is referred to elsewhere in this judgment, specifically stated that FXPB could accommodate exotic options. In an exchange on 10th October between Mr Brügelmann and Mr Said about “limits”, Mr Brügelmann asked Mr Said if he would prefer his trading limits to be expressed in terms of NOP or VaR. He said that either was possible but if Mr Said planned to do exotic options, it would have to be VaR. Mr Said’s response was to say that he thought VaR might be better as it captured the right risks.
On 18th October Mr Said sent an email to both Mr Meidal and a representative of FXPB in New York in relation to a meeting in New York to discuss the issue of NOP as against VaR in more detail. In preparation for it he produced a sample portfolio and said in the email that he would like to see what DBAG’s models produced as VaR or NOP exposure and collateral requirements against that portfolio. The sample portfolio contained five spot positions randomly selected, according to Mr Said’s deposition. No options were included but Mr Said specifically said in the email that he would like to see and understand how option structures factored into the sample, including exotics which he understood could not be margined under an NOP system.
In an email a year later to Mr Quezada and Ms Greenberg of FXPB, Mr Said, in seeking to persuade them to take in a Structured Option, referred to a discussion about the FXPBA and taking in “one off (non-standard)” trades wherein Mr Quezada had told him that this would not present a problem but requested that he should not swamp them with such trades. He went on to say:
“Frankly a PB agreement where I can only do spot and simple options is of no use to me and that was discussed at the outset of our discussions.”
Mr Quezada’s recollection, as set out in his deposition, was that this was discussed at the time when the account was first set up and Mr Said had said that he would trade exotics at some point.
As already pointed out, the FXPBA which Mr Said negotiated on behalf of SHI specifically permitted trading in Structured Options. Mr Said reviewed drafts of it, approved the wording and presented it to Mr Vik for signature. In an email of 6th November 2006, Mr Said referred to a draft of the FXPBA that he had received on which he had a few questions which he would discuss with Mr Quezada but he found it essentially simple and uncontroversial. It is presumably to this discussion that Mr Said referred in his email a year later.
Following signature of the FXPBA and other contractual documents on 28th November 2006, Mr Said commenced trading under it, obtaining Mr Vik’s approval to trading decisions in the first two or three months. Thereafter he made his own trading decisions and reported to Mr Vik weekly on a Friday about his activities that week. The emails generally included an update on the realised profits and loss position on the account, his views of the market and information about trades that he had concluded, with varying degrees of detail.
In the overall period in which he traded, Mr Said concluded ninety-four trades which could be described as “exotic” – forty-one EDTs and fifty-three OCTs. By contrast, he entered into seven hundred and twelve vanilla trades. The first of the OCTs and non-vanilla trades was concluded on 8th February 2007 and the first EDTs on 19th February 2008. If there was a restriction to vanilla trading in either the Said Contract or the Said Authority Agreement, Mr Said breached this no more than two months or so following the signing of the FXPBA. Moreover, as appears from the email exchanges between him and Mr Vik, he showed no reluctance to tell Mr Vik about such trades and Mr Vik raised no objection to them. Some of these trades were described in some detail, both OCTs and EDTs, and in particular in 2008 Mr Said kept Mr Vik informed of the progress of his EDTs, having explained their basic format and offered on more than one occasion to “walk through” the terms with him.
In his end of the year report on his profits, Mr Said specifically referred to Structured Options “net of several timer options and correlation swaps”. Mr Vik accepted in evidence that he would have read this email, upon which Mr Said was claiming his 10% profit share and would have noticed this reference. A clearer reference to the conclusion of non-vanilla trades would be hard to find.
It is plain beyond peradventure that Mr Said did not see himself as being under any restriction which limited him to entering into vanilla trades only.
If Mr Vik had indeed entered into the Said Contract and the Said Authority Agreement that is alleged, there cannot be the slightest doubt that Mr Vik would have raised the point with Mr Said and with DBS or DBAG. He never did so. He had brought in Mr Said with a view to increasing the breadth of SHI’s trading. In his statement he said that Mr Said would be engaging in much more active and diverse
trading in the execution of complex and diverse strategies but he drew the distinction, he said in evidence, between complex and diverse strategies with vanilla trades in the shape of straightforward options on the one hand and complex trades, complex options and structured products on the other.
Mr Vik’s own approach to trading was that of a risk taker, as he himself had told Mr Said. He looked to “win big because we lose big sometimes as well”. Moreover, Mr Said’s trading was not even limited to FX. Prior to the conclusion of the FXPBA he had concluded Fixed Income investments in Argentinean bonds and the early exchanges with DBAG show that, even after the US$35m figure was fixed for Mr Said’s FX trading (and became the Allocated Portion in the CSA), a further US$40 million was being discussed as capital for FI trading for him. As previously mentioned, a figure of something under US$70 million was in fact put into the Pledged Account. In mid-2007 Mr Said concluded a further Argentinean bond transaction and CD swaps with notional values of US$300 million. In August 2007, also, when vain efforts were being made to set up an FIPB, Mr Vik was keen to obtain facilities that would enable both him and Mr Said to trade “[a]ll products of any kind”.
Although it appears that Mr Said’s initial FX trading was in vanilla transactions and that he increasingly concluded more exotic transactions as time went by, it is clear from the evidence that whatever reservations he originally expressed, whether to Mr Geisker or Mr Quezada about exotic trades, once he had been introduced to the EDTs by Mr Chapin of CS in February 2008 he became increasingly enamoured of them as the only way to make money in what he perceived to be essentially a directionless market.
When all went wrong in October 2008, despite DBAG’s futile attempts to conceal the problems it had in booking, valuing and margining the EDTs (of which Mr Said was only too aware), it was straightforward in referring to the EDTs themselves when making margin calls. By this stage, Mr Said had produced details of his trades to Mr Vik for him to review in order to work out a way forward. Indeed Mr Vik had asked Mr Said for (and it is to be assumed that he had received) five year charts on the currency pairs in Mr Said’s outstanding trades. Over the course of the weekend of October 10th/12th and the week that followed when the margin calls were made, Mr Said and Mr Vik were in constant touch about those trades and the state of the market. It cannot seriously be suggested that Mr Vik was unaware of the essential details of the EDTs at this point (although there is good evidence that he was aware long before). Without protest, albeit asking from 16th October onwards for details of DBAG’s prior margining of Mr Said’s trades, Mr Vik authorised payment of over US$500 million by way of collateral and premium to close out trades. At no stage did he suggest that Mr Said had exceeded his authority in entering into non-vanilla trades. Nor did he make this point at his meeting with senior DBAG personnel on 30th
October. Nor was the point made in the New York Complaint served in January 2009. The expression “vanilla trades” appears nowhere in any of the documents at all until the Defence and Counterclaim served in this action by SHI on 21st March 2011.
SHI relied upon a few of DBAG’s internal documents and data in support of its allegation but those documents will not bear the weight which SHI seeks to put upon them. A client profile created for Mr Said’s FX trading by Ms Greenberg of FXPB listed the products as “Spot, Fwd, Swap, NDF, Precious Metals, Vanilla Options”.
This was an online form which had to be completed as a matter of administration. Ms Greenberg’s evidence was that she could not recall why she had selected those products and did not recall spending much time considering or selecting them. The degree of care employed is revealed by the email which refers to the attached client profile for “Floyd Currency Trading LLC”, rather than SHI. The purpose of the email profile was, according to Mr Giery, to set out the trade booking codes to be used in DBAG’s systems for trades done. The document refers to file transmission by the client submitting trades manually online and this, as Mr Giery’s evidence suggested, may in fact be the reason for the products listed as these would be capable of being submitted by the client through TRM, whereas Structured Options not only required DBAG’s consent under the FXPBA but DBAG’s involvement in booking. It is clear from other documents sent by Ms Greenberg at around the same time and afterwards that she had no understanding that there was a limitation on SHI’s ability to conduct non-vanilla trades. This document cannot therefore reflect some communication from Mr Meidal of an agreement of the kind suggested.
Although SHI made a point about structured data which suggested that the permitted products for SHI on the initial GEM set-up were 13 month cash trades which would include swaps, spot and forward trades, this was later amended in February to include 13 month one-bar and vanilla options and Ms Greenberg at the time of effecting those amendments observed that the list of permitted products had been updated several times but still did not include all the products which SHI was entitled to trade. The structured data in fact always limited trades inaccurately to 13 months. Once again the explanation for the set-up may be that these particular products were those which could be manually submitted online by the client and fed into GEM web reporting, as two FXPB witnesses suggested. The structured data should have been based upon a form called a CIF or a MIF and that form for SHI simply referred to options, without any restriction to plain vanilla trades.
These snippets do not take SHI’s case any further. The suggestion of any oral agreement to the effect suggested cannot stand in the light of the written agreements, Mr Said’s deposition evidence, the documentary records, the commercial realities and the universal understanding of DBAG personnel. It is another invention of Mr Vik’s or his lawyers.
None of the alleged oral agreements referred to in sections 10(a)-(c) were made prior to execution of the written contracts, at the time of execution or agreed or confirmed afterwards. The need for multiple expressions of agreement or confirmation results from the terms of the “entire agreement” sub-clause and the “no oral modification” sub-clause in Clause 9 of the FX ISDA. Mr Vik’s evidence represented a crude attempt to escape the effect of these provisions and fails because of the absence of any agreement to the effect suggested at any time.
10(d) Common Understanding, Mutual Assumption, Estoppel by Convention, Acquiescence and Rectification
428. Each of the alternative ways by which SHI sought to introduce limitations for the types of trade to be conducted under the FXPBA or for financial limits to the transactions to be conducted, by reference to some form of trading limits, as a result of exchanges between Mr Vik and Mr Meidal all fall to the ground because of my rejection of Mr Vik’s evidence. There were no collateral agreements as alleged and
no factual basis for asserting any common assumption, mutual understanding or acquiescence in an assumed state of fact or law which could give rise to any form of estoppel by convention whether considered as a matter of New York law or English law. Clause 9 of the FX ISDA militates against any such allegation in any event but as I have found that Mr Vik’s evidence is a fiction, the precise effect of that Clause does not matter. There was, likewise, no continuing common intention or outward expression of accord consonant with the allegations made and so there is no room for rectification of the written contracts.
10(e) The Collateral Warning Agreement
This allegation first surfaced on 3rd August 2012 following the exchange of witness statements. It was alleged that in telephone conversations between October and December 2006 and during the two meetings previously referred to between Mr Vik and Messrs Meidal and Brügelmann in November 2006 it was agreed that whenever the collateral requirements of SHI’s trading with DBAG were approaching the upper limits of the collateral then available for trading, DBAG would inform Mr Vik.
This allegation is a late entrant onto the scene in circumstances where I have already rejected the evidence of Mr Vik relating to meetings and telephone calls with Mr Meidal or Mr Brügelmann in 2006 relating to limits on Mr Said’s authority, trading limits and limits as to the types of trade which SHI could conduct. There is no evidence of such meetings or telephone calls and the fact that this point surfaced long after the earlier allegations which I have already rejected as fabrications by Mr Vik militates against an acceptance of this as a further collateral contractual obligation undertaken by DBAG.
The emergence of this allegation as a plea is no doubt attributable to Mr Brügelmann’s statement as to what had taken place on 14th November 2007 in the meeting in New York with Mr Vik and Mr Orme-Smith about the GPF account to be opened with DBAG for Mr Vik’s own trading. In that witness statement Mr Brügelmann referred to Mr Vik’s declaration at the meeting that he never wanted to be in a margin call situation. Mr Brügelmann said that was something that Mr Vik had explained to him on previous occasions in relation to the DBS accounts. Mr Vik then asked him as a courtesy to monitor the GPF account and to warn him when the value of the account was within 5-10% of a margin call though no specific agreement was made to that effect. The rationale for this is relatively clear. Mr Vik used Mr Brügelmann as his “go-to” man who executed his instructions for purchase and sales of assets for his DBS accounts and was to continue to do so for Mr Vik’s trading on the GPF platform with DBAG. He thus had sight of all that was going on with Mr Vik’s day to day trading. He was authorised to execute trades on behalf of SHI in respect of Mr Vik’s trading on the GPF platform.
In his first witness statement Mr Vik had said that he wished to be informed immediately if there was any chance that the positions taken by Mr Said were too large to be supported by the US$35 million allocated as collateral to his trading and that he was told by Mr Meidal and Mr Brügelmann repeatedly that DBAG had good collateral management systems and would therefore give him such information. He said that Mr Brügelmann was supposed to maintain overall supervision of SHI’s trading in London and New York and he relied on him to do so. There was no specificity as to the time when such assurances were given or such an agreement was made.
In his second statement, following the amendment by SHI to plead the Collateral Warning Agreement, he said that there was an agreement between SHI and DBAG that it would notify SHI whenever there was a risk that its trading could no longer be fully collateralised by the existing collateral. He said that he did not discuss a precise number with Mr Meidal or Mr Brügelmann but he thought that “an accurate reflection of what was intended” was that a warning should be given when SHI was about 10% away from reaching the US$35 million figure. He said he had no recall of the specific conversation with Mr Brügelmann to which the latter had referred on 14th November 2007 but there had been many conversations about this.
Mr Brügelmann’s further statement pointed out the difference between Mr Vik’s own trading at DBS and on the GPF account with DBAG on the one hand and the trading of Mr Said on the FXPB account on the other. On the former, Mr Brügelmann supplied daily figures to Mr Vik whereas on the latter Mr Said made his own reports to Mr Vik and it was only for the occasions that Mr Vik met with Mr Brügelmann that the latter made enquiries and obtained some figures about Mr Said’s trading on that account. (Mr Brügelmann only met Mr Vik three times in 2007 and three times in 2008 and those appear to be the only occasions upon which there was any discussion of Mr Said’s trading, following much lengthier discussion on other subjects with Mr Brügelmann and others). So far as he was concerned, he did not have access to information in relation to the FXPB account without asking for it whereas he had been given specific authority by Mr Vik to access DBX for the GPF figures and saw them daily in order to report to Mr Vik. Whereas he was asked, as a courtesy, to notify Mr Vik if he was getting close to a margin call on his own trading (which Mr Brügelmann executed on his instructions), because Mr Said reported directly to Mr Vik on his trading Mr Brügelmann was not asked to provide similar information on Mr Said’s account. There was never any specific agreement to give a notification about potential margin calls on Mr Said’s account.
In cross-examination Mr Brügelmann said that, at the meeting of 14th November in New York Mr Vik had said that he did not want to be in a margin call situation, which was something he had said on previous occasions. Mr Brügelmann said that as a matter of good client service he said he would keep an eye on the account to which he had access and tell him when there might be a potential call on its way. In January 2008, Mr Brügelmann observed to Ms Hart in a telephone call that he needed to make contact with Mr Vik once SHI were “5% close” to a margin call because when he and Mr Orme-Smith had seen Mr Vik, Mr Vik said he did not want to get into a margin call situation and said he wanted to be told before. Likewise, in August 2008 Mr Brügelmann emailed Mr Orme-Smith referring to the meeting with Mr Vik in which he had asked to be kept abreast of his margin situation well ahead of any margin call. There is also a reference in Mr Brügelmann’s email following a meeting with Mr Vik on 7th May 2008 when Mr Vik had a series of meetings with different DBAG employees before speaking alone with Mr Brügelmann at the end. Following this meeting Mr Brügelmann sent an email to Mr Orme-Smith referring to Mr Vik stating again that he did not wish to get close to any margin calls, at which “a smile went over Clare’s [Claire’s] face”. Claire Davies was involved in the discussion relating to GPF, being a member of Hedge Funds Credit who had some expertise to bring to bear
in that area. She would not have been present in the discussion between Mr Vik and Mr Brügelmann about the FXPB account when Mr Brügelmann produced some figures he had obtained from GEM, in accordance with his normal practice for meetings with Mr Vik. All the other persons involved in these discussions on DBAG’s side, apart from Mr Brügelmann, were persons whose role was limited to GPF, whilst Mr Brügelmann’s prime functions related to Mr Vik’s own trading, not that of Mr Said. I find that Mr Brügelmann was wrong in his evidence under crossexamination when he said that there was discussion of the FXPB account in the presence of Mr Orme-Smith and Claire Davies.
Mr Vik’s own evidence, when cross-examined, was that he could not remember any specific discussion with Mr Meidal about a collateral warning during the set up of the FXPB account for Mr Said’s trading. He considered that he had a general unwritten agreement in relation to all his accounts.
I reject Mr Vik’s evidence that there was any understanding at all relating to the FXPB account. Mr Brügelmann and Mr Vik had a close relationship in relation to Mr Vik’s own trading because of the daily instructions that Mr Vik would give to Mr Brügelmann and the daily reports that Mr Brügelmann would give to Mr Vik. In that context Mr Brügelmann was entirely up to speed with what was going on in Mr Vik’s trading accounts but this was not true of the position with regard to Mr Said’s trading, as Mr Vik must have known. Mr Vik received very little reporting from Mr Brügelmann on Mr Said’s FXPB account, being reliant upon Mr Said himself to make reports as he did weekly, religiously at the end of week. In consequence, Mr Vik knew more about Mr Said’s trading than Mr Brügelmann ever did.
Once again the reaction of Mr Vik to the events of October 2008 is significant in this context. At no point did he ever suggest that Mr Brügelmann should have supplied a warning about Mr Said’s margin situation. By October 10th, at the latest, Mr Vik knew from email exchanges with Mr Said that the cost of closing out the account would run into hundreds of millions of dollars. This meant that margin limits had long since been exceeded and yet there was not a murmur of complaint from Mr Vik that Mr Brügelmann had failed to tell him of this at any time beforehand.
Mr Brügelmann’s conduct is likewise consistent with his evidence that the discussion related only to Mr Vik’s trading account. On 3rd September he did give Mr Vik a warning about his proximity to a margin call on his trading accounts in an email which is the cause of a different complaint by Mr Vik, namely that the warning was wrongly based upon an understanding that Mr Vik’s FX trading was governed by the Equities PBA and not the FXPBA.
The fact is that the conduct both of Mr Brügelmann and Mr Vik is consistent with what the documentary records show, namely that Mr Vik asked for a margin call warning in respect of his own trading and Mr Brügelmann told him that he would give such warning.
There is no indication anywhere in the documents that Mr Vik ever asked for a margin warning in relation to Mr Said’s FXPB account which was always regarded as something distinct from Mr Vik’s own trading. I find that Mr Brügelmann, when he spoke of giving a warning, did so only in relation to the accounts where he had immediate visibility, namely those where Mr Vik used him to execute trades on his
behalf and to which he had authorised access. Neither he nor Mr Meidal had authorised access to the FXPB account. Any agreement on this, as for the Capital Limitation Agreement, would be expected to be made with FXPB personnel not DBS personnel.
There is an issue between the parties as to whether or not Mr Brügelmann undertook a binding obligation on the part of DBAG as a collateral agreement. There was no formal agreement in the shape of a written contract nor even an exchange of correspondence or emails. On Mr Brügelmann’s evidence there were a number of conversations in a business context but the effect of the binding agreement alleged by SHI would be a significant alteration in the obligations of DBAG under the FXPBA, the FX ISDA and the CSA annexed. Mr Brügelmann would not have considered himself able to vary or modify the CSA and it would be hard to see how Mr Vik could have thought that he could, since he was a relationship manager at DBS. What Mr Brügelmann was doing was managing the relationship with Mr Vik in an effort to assist but there was no intention on his part to change DBAG’s entitlement to call for additional collateral in circumstances where the contracts allowed it. Nor could Mr Vik have thought that anything that Mr Brügelmann offered to do in this context by way of warning could affect DBAG’s right to demand margin in accordance with the terms of the CSA.
The Collateral Warning Agreement is inconsistent with the contractual rights of DBAG to call for margin, whether or not it is alleged that the giving of the warning was a pre-condition to the right to call for margin or it is said that it is a breach of contract not to give such a warning before making such a call. The alleged oral agreement touches on central rights of DBAG to protection as provided in the FXPBA and the FX ISDA.
The nature of the alleged duty is not clear. SHI accepts that if it was not practicable to give a warning because of sudden large market movements, there would be no liability. As appears hereafter, in circumstances where DBAG did not know what its margin entitlement was because Mr Said persuaded DBAG to take in EDTs which it could not value or margin on its systems, DBAG could not give any warning and the agreement/waiver of any duty to report MTM and margin carried with it the waiver of any duty to warn.
The agreement to warn is alleged to be collateral to the FXPBA although it relates to giving a warning in respect of margin calls which would be made under the FX CSA. The FX ISDA includes Section 9(a) and (b), the entire agreement and no oral modification provisions. Furthermore, Clauses 12.6 and 29.1 of the Equities PBA referred to the absence of any representation or warranty made outside the terms of the agreement which constitutes the entire agreement and understanding of the parties with respect to its subject matter.
Whilst the entire agreements clauses prevent any assurance given in the course of negotiations from having legal effect, they are not directly applicable to subsequent variations of the contract. The no oral modification Clause is effective under New York law only where modification remains executory and is not the subject of partial performance in the sense of action which is unequivocally referable to the alleged modification. In English law as applicable to the FX ISDA and 2008 Agreements, the
terms take full effect with the clearly expressed intention that the parties’ agreement is to be found in written contracts and nowhere else.
The 3rd September email is, in reality, explicable only by reference to the conversations between Mr Vik and Mr Brügelmann, insofar as a warning relates to Mr Vik’s own trading on the GPF platform under the Equities PBA. There is however no conduct which is capable of amounting to performance of any amendment to the FXPBA and in circumstances where the parties have agreed to formalities of the kind which appear in all these agreements, in order for contractual rights to be affected, I conclude that Mr Brügelmann did not enter into a binding contractual commitment on behalf of DBAG to give a margin warning which could have any impact upon DBAG’s right to call for margin under either the FXPBA or the Equities PBA or the ISDA agreements which related to them. Mr Vik expected any commitment made by the bank to be made in writing and this was not. He knew that what he was getting was assistance from his relationship manager without commitment on DBAG’s part. There was, within the meaning of the authorities, no animus contrahendi or intention to create legal relations in the sense of a subjective intention on either party to enter into a binding contract.
As a matter of fact, DBAG’s practice set three different levels at which action was or was not taken. Initial margin was set at 200% of 5 day VaR for Mr Said’s account- the maintenance level. The call level was 150% and the close out level was 100%. The practice was not to call for margin at the 200% level but only at the 150% level, so that, in effect there was latitude given to the customer as against the margin requirements agreed and which were supposed to appear on the GEM web reports. An email of 18th October 2006 to Mr Said spelt out the practice of DBAG to call for margin at the 150% level and to close out at 100%. The effect would be that the customer would be notified of a deficiency in margin at a level greater than that at which a call could be made. As to causation, I need say nothing further but it is plain that if DBAG had issued a margin warning to Mr Vik, as opposed to Mr Said, about Mr Said’s FX trading, Mr Vik would have had a discussion with Mr Said about his margin requirements and the trading and the FXPB account. If a margin call had been made, then Mr Said would have had to discuss the matter with Mr Vik. There is no reason why DBAG would not have observed its ordinary practice in that connection in calling for margin at the 150% level.
In Annex 3 appear the margin figures as calculated by the forensic accountants. As mentioned elsewhere I conclude that with the profits Mr Said was making Mr Vik would not have stopped Mr Said’s trading or required him to reduce his positions whilst trading appeared profitable and until the margin figures hit US$150-$200 million in September 2008.
11. The Meaning of Currency Options and Structured Options, the Said Letter of Authority and the FXPBA
The Said Letter of Authority referred to currency options, whilst the FXPBA referred to currency options and Structured Options. In neither instrument was the reference to currency options made to a capitalised term, so that they were not defined by reference to the 1998 ISDA FX and Currency Option Definitions (the “ISDA Definitions”), as published by ISDA and EMTA. The FXPBA contained its own definition of Structured Options.
There is a dispute between the parties as to the extent of Mr Said’s authority to bind SHI to various types of transaction concluded by him in the name of SHI. DBAG has accepted that it is bound by Mr Said’s actions in relation to the Counterparties under the terms of the FXPBA while SHI maintains that Mr Said had no authority to bind it to various offsetting transactions under the FXPBA because they fell outside the terms of the authority given thereby and in the Said Letter of Authority. Clearly the authority given to Mr Said under the FXPBA to bind DBAG to the Counterparties and to bind SHI to DBAG was identical, but the parties have taken different stances in relation to that. The issue is less a matter of construction between the parties and more a matter of application of the terms used in the agreements to the transactions which are disputed – the EDTs and the OCTs. Were they currency options or Structured Options within the meaning of the FXPBA and Said Letter of Authority?
The FX ISDA Schedule and the Amendment Agreement of 28 November 2006 both stated that they governed FX Transactions and Currency Option Transactions, as defined in the ISDA Definitions, unless otherwise agreed between the parties, but that is of no direct assistance in construing the FXPBA and Said Letter of Authority, save insofar as the ISDA Definitions represent market understanding. Each confirmation expressly stated that the transaction recorded therein was subject to ISDA terms and Definitions, but that does not help in determining whether it was truly a currency option or Structured Option, although it shows that the parties wanted them treated as an FX Transaction or Currency Option, within those definitions for the purpose of incorporating ISDA terms.
The ISDA Definitions contain the following:
“Section 1.5. Currency Option Transaction "Currency Option Transaction" means a transaction entitling Buyer, upon exercise, to purchase from Seller at the Strike Price a specified quantity of Call Currency and to sell to Seller at the Strike Price a specified quantity of Put Currency.
…
Section 1.12. FX Transaction "FX Transaction" means a transaction providing for the purchase of an agreed amount in one currency by one party to such transaction in exchange for the sale by it of an agreed amount in another currency to the other party to such transaction.”
The definition of Currency Option Transaction is wide and would encompass vanilla and exotic options with any number of additional terms and conditions attaching to them. In 2005, ISDA produced a Barrier Option Supplement to the ISDA Definitions which made it clear that the definition of Currency Option included Barrier and Binary Options. The Volatility Swaps Supplement specified that these are to be treated as a type of FX Transaction. These definitions exist to facilitate better trading and to provide a general framework to document trades, where standard ISDA templates can be used and tailored for non-vanilla trades. The question still remains as to what the market sees as being currency options and what falls within the definition of Structured Options as set out in the FXPBA.
It is necessary to determine what is meant by a currency option and a Structured Option by reference to market understanding, as opposed simply to legal analysis of the terms of the trades done, as recorded in Trade Confirmations, which incorporated the ISDA Definitions. I had reports from two experts and heard evidence from each about market understanding – from Mr Malik on the one hand, the expert engaged by DBAG, and Professor Wystup on the other, the expert engaged by SHI. I will say more about these experts later.
The disputed transactions which, in SHI’s submission, fell outside the ambit of Mr Said’s authority are as follows: the Target Profit Forwards and Pivot Target Profit Forwards, the Forward Setting Currency Options, the Knock Out Currency Options, the Double Knock Out Currency Options, the Knock Out Timing Options, the Digital Currency Options, the Dual Range Digital Currency Options and the Fade-in Forwards. Both experts agreed that the Correlation Swaps did not fall within the definition of currency options or Structured Options in the agreement.
As a matter of legal analysis, an option, in simple terms, gives the holder the right but not the obligation to do something, often to purchase or sell, those being respectively referred to as a call option and a put option. A currency option, in its ordinary form, is therefore an agreement between two parties which entitles one party to purchase from or sell a specified currency to the other. The ISDA Definitions are set out above for a Foreign Exchange Transaction and a Currency Option Transaction. The ISDA Definitions list call and put options as Currency Option Transactions entitling the buyer or seller, as the case may be, to purchase or sell the relevant call or put currency.
It is again plain that, as a matter of freedom of contract, parties can negotiate and agree terms and conditions for options as they think fit, so that the basic form of option may become more complex. There may be conditions attached to the exercise of an option or an option may be deemed to be exercised or treated as exercised automatically in given circumstances. The parties can agree that no currency need be delivered and that, on exercise, automatic exercise or deemed exercise there is simply a pay-off which represents the profit or loss made on the transaction by reference to a base currency. The parties may also agree a series of put and call options over a period of time with daily or weekly exercise and a resulting pay-off which is the net result of this combination of options. Such transactions may limit the total profit payable in respect of any series of options.
The experts helpfully produced an “Agreed FX Primer” which set out an overview of the FX market and of FX products including, for the most part, agreed descriptions of the different types of product in issue in this action, although they were unable to agree on a description of Fade-in Forwards. Reference can be made to this document or to the experts’ reports in relation to each of the types of disputed transaction and I will not attempt to set them out here. I will however, later, set out their description of Target Forward Products (as Professor Wystup preferred to refer to them) including Target Profit Forwards (TPFs) and Pivot TPFs. Whatever else the term Structured Option may mean (a term which Professor Wystup said he had not come across before this litigation) it must connote a degree of complexity or structure to it. In a letter from ISDA in December 2009 (and therefore over a year after the events in issue in this litigation) the following appeared:
“Currency options can take many different forms. There are "plain vanilla" options where one party (the "buyer") pays a premium to the other (the "seller") in order to have the right at expiration or during a specified period to exercise the option into a spot foreign exchange transaction for a specified currency pair. There are also "structured" options, which have different features that allow the option holder to achieve different results, the most common of which are known as "barrier options" (i.e., knock-in and knock-out options) and "binary options" (i.e., one touch and no touch options). These structured options look at the spot exchange rate for a specified currency at expiration or during a specified period in order to establish whether the option is exercised or the option holder is entitled to a specified payment.”
In the FXPBA, Structured Options have their own defined meaning, namely an option other than one which is a put or call option that does not have special features or a single barrier option. Self-evidently therefore, any currency option which does have special features or more than one barrier is a Structured Option for the purposes of the FXPBA. This does not however help with regard to the question as to whether or not what is being considered is truly a currency option in the first place.
A further term which is used in the market is the expression “Exotic Options”. In Mr Malik’s view the term is synonymous with Structured Options as used in the market, namely more complex options with special features of one kind or another which take them outside the range of what are considered as vanilla options. Over time, the experts agreed, the market’s perception changes as to how exotic a product is.
In the Agreed FX Primer, Target Forward Products (TFP) are described in the following manner:
“2.16.1. In a TFP, the investor agrees to buy a specified amount of a currency at an agreed rate (the strike price) on a number of dates (“fixing dates”) which can be daily, weekly or monthly, depending on the contract.
2.16.2 On each fixing date a settlement amount is determined.
The settlement amount for a purchased TFP is calculated as (spot – strike) x notional per fixing date. If this amount is positive, i.e. spot is greater than strike, it is a gain to the TFP investor. Otherwise it is a loss (i.e. a gain to the TFP seller).
2.16.3 The positive and negative amounts accrue with each fixing date and are typically cash settled on the final settlement date. The terms of the trade do not contain any cap on the investor’s potential losses).
2.16.4 If the sum of positive settlement amounts reaches or exceeds the pre-specified target profit, the TFP terminates. The sum of positive settlement amounts (which is capped by the target profit in the SHI set of transactions) and the sum of negative settlement amounts are either paid by the seller and buyer respectively on the settlement date or are cash settled by the payment of a net amount by either the seller (if positive amounts exceed negative amounts) or the buyer (if the opposite is the case).
2.16.5 As long as the sum of positive settlement amounts stays below the target profit, the trade continues to exist and settlement amounts continue to be determined and accrued. As is the case when the TFP terminates due to the target profit being met, the accrued positive and negative settlement amounts are paid or cash settled on the settlement date.
2.16.6 The investor in a TFP faces a limited upside (gain is capped at target profit) and a potentially unlimited loss. Typically the TFPs are “zero-cost” structures with a strike price that allows investors to buy the notional amount at a betterthan-market outright forward rate.
2.16.7 At inception, the strike price of a TFP transaction is typically in the money (in that the strike is better than the outright forward rate). The investor hopes the TFP will remain in the money and the target profit will be reached within a short period of time.
2.16.8 However, if the FX rate was to move below the strike price (and stayed below) soon after inception, the likelihood of achieving the target profit becomes low. In this scenario, the investor would accumulate losses.
2.16.9 It is possible that the investor will not make a profit even if the target profit is reached. This is because the sum of the negative amounts may exceed the sum of positive amounts on the date when the target profit is reached and the transaction terminates. A TFP is therefore a means by which an investor can express a combination of a directional and low volatility view, as if the market moves in his favour the trade will terminate early and he will receive a profit.”
463. The Agreed FX Primer describes “Pivot” TFPs in the following way:
“2.17.1 In a pivot TFP, the payment by each party is by reference to three prices (the “low strike price”, the “high strike price” and a “pivot”), instead of just one, as in a TFP. A pivot TFP pays out to an investor if the FX rate remains within a particular range and close to a pre- determined pivot.
2.17.2 The investor receives a settlement amount if the FX rate is between the high strike price and the low strike price and has to pay a settlement amount if the FX rate is outside of this range (see diagram below).
2.17.3 As in a TFP, the positive amounts and negative amounts (to the investor) accrue throughout the life of the transaction and if the sum of the positive amounts reaches the target profit, the transaction terminates.
2.17.4 In a pivot TFP, the settlement amounts are determined by reference to the pivot. On each fixing date, if the fixing is above the pivot, the settlement amount is calculated as (high strike – spot) x notional per fixing date. If the FX rate is below the pivot, the settlement amount is calculated as (spot – low strike) x notional per fixing date.
…
2.17.9 … The terms of the trade contain no cap on the amount of the investor’s losses which are dependent upon the amount of the average fixing, while the profit is capped at the target profit. A pivot TFP is therefore a means for an investor to express a view on the volatility of the FX rate. Here, the investor benefits if the FX rate fixings are close to the pivot, however it doesn’t matter if the FX rate fixings are above or below the pivot. An investor in the pivot TFP depicted above would be of the view that the volatility of USDJPY would remain low and the trade would terminate early as the target profit would be reached.
2.17.10 If this view was incorrect and the FX rate moved beyond the high or low strike prices, so the trade did not terminate early the likelihood of achieving the target profit would become low. In this scenario, the investor would accumulate losses. This would place the investor in a similar position to being out-of the money on a “sold” vanilla option on each fixing date.”
11(a) The Expert Evidence
The fundamental difference between Mr Malik and Professor Wystup can be expressed relatively shortly. Professor Wystup’s view was that there were two distinctive features of currency options, without which a transaction could not properly be so termed. The first was an element of optionality, namely a discretion on the part of the holder whether to exercise the right to purchase or sell, as the case may be (a call or put option) without any obligation to do so. The second was that an option could not have a negative payoff and result in loss. Mr Malik’s view was that the market considered that currency options could consist of a series of bought and sold options, that any seller of an option had an obligation to fulfil the sale (if the buyer chose to exercise the option) and that a sold option always contained the possibility of a negative pay-off. If the market price moved below the strike price on the sold option, and a deal was cash settled, the seller would suffer a loss representing the market differential. His view was that each of the disputed EDTs and OCTs (other than the correlation swaps) was taken by the market as representing a series of embedded bought and sold options, in whatever language the trade confirmations
were expressed, and that the final pay-off represented the aggregation of the sums payable on automatic or deemed exercise of the options on the fixing dates, subject to the condition agreed as to a maximum profit – the target feature. Professor Wystup considered that the combination in a single contract of a portfolio of options could not be considered a “currency option” although it might be considered a “Structured Product” or an “FX derivative”.
Mr Malik expressed the opinion that market participants used the terms FX options and currency options interchangeably and would regard the EDTs and OCTs to be a form of FX transaction. They would also consider them to be Currency Options as defined in the ISDA Definitions with the exception of the correlation swaps which would be “Foreign Exchange Transactions” but not Currency Options. With this exception, all of the EDTs and OCTs would be Structured Options, both within the meaning of the term as ordinarily used in the market and within the definition of the FXPBA, save for Knock Out Options which have only one barrier and therefore are simply currency options.
Professor Wystup took the position that currency options meant simply vanilla options and did not include anything other than straightforward put and call options. He divided exotic options into three categories, namely first, second and third generation exotics. None of the EDTs or OCTs were in his view vanilla (as is common ground) but, as he put it, on a “very generous” reading of currency options, Digital Options and Knock-out Options fell within the meaning of the phrase and
Knock-out Timing Options were on the borderline. Everything else fell outside it. When looking at Structured Options, he offered three potential meanings, the broadest of which would encompass “market standard products” – i.e. first generation exotic options, in his classification. He was insistent that currency options and Structured Options must have a non-negative pay-off whilst recognising that there were some grey areas imported by market standards in relation to volumes, exercisability and risk profiles.
Professor Wystup had never traded a TPF (which he accepted had been in the market since 2004). Since November 2003 he had been the managing director of Math Finance AG which he described as “a global network of financial engineers which advises market participants on FX options, software and trading”. He is an honorary professor of quantitative finance at the Frankfurt School of Finance & Management and an associate fellow in the Finance Group of Warwick Business School at the University of Warwick. He is also a lecturer in Financial Engineering Programmes at the National University of Singapore. He describes himself as having worked in the FX options industry since 1991 and, following internships, worked for three years between 1996 and 1999 as a consultant and foreign exchange options quantitative analyst at a private bank in Frankfurt. Between 1999 and 2004 he worked at a Frankfurt bank as a quantitative specialist, risk manager and product developer in the Global Foreign Exchange Options Department where he was responsible for developing and implementing FX derivatives models, generating tailor-made structures for the bank’s clients and internal and external consulting for all FX options matters. The reality was that he had never worked as a trader or in a front office role and had no experience of FXPB. He said that he had experience on the FX structuring desk dealing with a lot of salespeople who in turn dealt with clients in buying and selling FX and that he was involved in training.
Mr Malik, in the relevant period of 2007-2009, was head of DCRM Solutions at Barclays Capital. He ran the front office trading and structuring team that risk managed, traded and distributed contingent credit risk arising from Interest Rate/FX, commodity and emerging markets derivative deals in the rates market world. During his time he helped manage counterparty risk arising from vanilla swaps and options, exotics, TARNs, FX2 and choosers (TARN is another term used for a TPF). From 2001-2007 he had set up his own fund business in India to manage money for individuals and trained investment banking professionals in, amongst other things, FX and credit products. He had extensive experience prior to that, going back to 1990 as managing director in fixed income credit trading and structured credit.
Professor Wystup’s opinions were hampered by the fact that in August 2011 he had been interviewed by Freshfields, who were acting for DBAG, and had expressed to them, at that stage, views which were diametrically opposed to those he gave to the Court.
Despite efforts to explain this away, the fact remains that, when first sent copies of trade confirmations for two TPFs, he concluded that they fell under the category of options which he considered a broad term. Moreover he considered that it would be highly impractical to create a full list of what fell within the definition of options and that it would be more usual for prohibited trades to be specifically listed in a contract or for there to be an explicit statement that only vanilla trades were allowed. This was, he said, not only a matter of common sense but he said that he had never seen anyone approach the issue differently and it was not something that, so far as he was aware, had ever been questioned. Unless therefore there was explicit exclusion in an agreement, a TPF would be an option. A currency option was a very general term and any book written on the subject would cover a lot more than vanilla options.
On 10th August 2011, when looking for an expert who might be able to give evidence on the issues which arise for my decision in this area, Freshfields contacted Professor Wystup, enclosing an extract of part of the Defence and Counterclaim which set out SHI’s argument that Mr Said had exceeded his authority in entering into EDTs at paragraphs 85 and 86. In the Defence and Counterclaim, it was said that each of the EDTs was not an FX transaction because they each contained a Target Knock-out event feature which meant that they were “exotic FX derivative transactions” and not FX or Option transactions. Moreover, it was alleged that the EDTs did not give SHI a right to exercise as opposed to an obligation, and their character was such, by reference to pricing/valuation, trading, risk management and documentation that they were not vanilla transactions. Also enclosed were the FXPBA and the Said Letter of Authority together with two trade Confirmations from CS of an AUD/NZD Target Redemption Forward with a strike at 1.2440, with daily observations for a year and a USD/CAD Pivot Target Accrual Forward with a low strike, a high strike and a pivot, running for a year, with daily fixings. Additionally the declaration of an expert Mr Jananayagam, put forward by SHI in the New York proceedings, was sent to Professor Wystup, in which he expressed the view that the EDTs could not be categorised as “currency options” as such terms were used in the FXPBA and Said Letter of Authority. Mr Jananayagam expressed the view that an option was a financial instrument which gave the right but not an obligation to the option buyer to purchase or sell an asset at a pre-agreed price on a specified date. A currency option therefore allowed the buyer the right but not the obligation to purchase or sell one currency in exchange for another and no such right was given to either SHI or DBAG in the EDTs.
With the benefit of these documents sent to him on 10th August, Professor Wystup gave his views in a telephone consultation on 18th August with representatives of DBAG, Freshfields and DBAG’s New York attorneys. Notes were taken by a representative of each of those entities and expressed clearly what Professor Wystup was saying at the time. I have already expressed in outline the record of his view as noted by the Freshfields’ representative which concluded by disagreeing with Mr Jananayagam’s affirmation and the views expressed in it. The notes of the representative of DBAG (a member of its legal department) referred to Professor Wystup’s view that a market view of the difference between vanilla and exotic options depended upon the person being asked. A trader would merely distinguish between a vanilla option and a structured option. His own experience was gained within a group of colleagues within the bank and the range of corporate and institutional clients with whom he dealt. He was specifically asked in the call if he would be able to testify that TPFs would be covered by an ISDA to which his answer was yes. He said he was familiar with the Target Redemption Forwards but had not seen the TPF before with its “extra knock-ins and knock-outs” but he considered those as merely tweaks from the ones he had seen. He regarded Target Redemption Forwards as a type of Forward and a TPF was an FX Forward or an Option and would be treated as such unless specifically excluded in an agreement. It would be hard to list products and keep them up to date in an agreement, so the general practice would be to include everything but list exceptions. If only vanilla trades were intended, that is what would be expressly said. When referred to Mr Jananayagam’s views, he said he did not agree. The EDTs were not vanilla put or call options but currency option was a general term and in any book that you opened, you would see a whole variety of options and not just vanilla. Based on his experience the market would regard these transactions as options or FX forwards.
The notes of another representative of Freshfields were to the same effect. Market views would depend on the identity of the person asked as to whether something was vanilla or exotic. An experienced options trader would consider all as options and would specify if he wanted to trade vanilla alone. Options included the whole product range including “super-derivatives”. He said he had no familiarity with FXPB and no experience of the terms in which authority was given to traders. He had come across TPFs but not the pivot trades which were more recent but both were properly described as FX options and would be covered unless specifically excluded in an agreement. A TPF was a type of forward, and because of developments, it would be difficult to list a set of authorised products in a Master Agreement and keep it up to date. Hence such an agreement would tend to be general and only list the exceptions that were not permitted. He had not seen it done differently. He specifically disagreed with Mr Jananayagam’s declaration that exotic derivatives, including knock-outs, would not fall into the category of options or forwards. He saw currency option as a very general term which would include vanilla trades but was not exclusive to it. Any book that you opened regarding options would provide you with a wide variety of them. When asked if he had ever had to give an expert report on whether a pivot or TPF would commonly be regarded as an FX forward or an option, he said that he never expected anyone to doubt it.
In a follow up email the same day, Professor Wystup made a suggestion to Freshfields. He said that if further evidence was needed for the scope of the terms “forwards” and “options” in Prime Brokerage Agreements/Master Agreements/ISDA Agreements, his consulting company could carry out a survey of the number of banks known to them and take statements. He concluded by saying “I think it is clear anyway, but having some supporting material could support your claim”.
Professor Wystup’s evidence to the court was that, at the time he expressed these views, he had not taken much trouble to review the documents he had been sent, had not spoken to others in the market and had not sufficiently thought through the issues as he had by the time he came to provide his reports and give evidence. In a statement, he said that the views expressed in the telephone consultation were his “initial views, expressed off-the-cuff” and gave “a completely bank-focused approach”. His responses were formulated on the call without prior consideration. He used labels loosely and now considered that there was a significant difference between an internal bank trading mandate and a client-facing contractual trading authority and he was only considering the former in the telephone call.
Later in the witness statement he said that he had used the term “option” in a very broad and loose way and had not settled exclusively either on “forward” or “option” as a classification for the TFPs. The reference to “option” was likely to have been because his peer group often addressed pivots and TFPs when talking or writing about options and he was using the word in the way that some of his peer group did use it. This of course is, in fact, very much to the point. If others in the market consider these types of trades as options, then later reflection and analysis may not in fact assist.
Whilst professing a consciousness of his duties to the Court, Professor Wystup, both in his reports and in his oral evidence, was a very unsatisfactory expert. He was partisan and argumentative and sought to argue questions of construction of the contracts rather than focussing on the question of how the market understood the position. His views, expressed in the telephone call, were close to those of Mr Malik, at a time when he was anticipating that he might give expert evidence on behalf of DBAG. When consulted by SHI, he changed his views to give diametrically the opposite opinion without any qualifications or reservations and without any explanation of any kind until he was faced with the notes of what he had said in the telephone call in August 2011. I did not therefore find him an expert upon whose market views I could rely.
By contrast, Mr Malik came across as a man who knew the market at first hand in a trading context as well as in the context of talking to back offices and clients. As he said, people in the market did not tend to analyse the transactions in question but simply thought of them as currency options. There were indicia as to why this might be so but the rationale for the view was not something that would be discussed much or expressed in the market. The Trade Confirmations were usually expressed in terms of “pay-off” but the underlying basis of the disputed transactions was the series of call and put options for each of the fixing days, which he described as “embedded” within the terms of the deals. The concepts for structured options had been developed from more straightforward options and strategies of trading straightforward call and put options. This was done in order to sell products on the market that met particular business needs and in particular the desire of hedge funds and traders to speculate in currency without payment of premium – “zero cost options”. Combining vanilla options provided some degree of flexibility for investors (Call Spreads, Risk Reversals, Straddles, Strangles and Butterflies) whether for hedging purposes or otherwise, but the strategies provided by the use of such options might not capture the market view of the investor. Customised products therefore were brought to the market which came to be known as “exotic” currency options or “structured” options, according to Mr Malik. The latter phrase was usually used to describe transactions which had the combination of options or embedded options which Mr Malik said were inherent in the EDTs.
The vast majority of trading on the FX market which is the world’s largest and most liquid financial market with a daily currency turnover of nearly US$4 trillion, is speculative. One of the attractions of buying options is that for a fraction of investment of option premium, investors are able to assume risk on the whole investment but, with zero cost options, the attractions become even more obvious. If the underlying asset moves in favour of the holder, substantial sums can be earned. Moreover, although options fall to be exercised, if an option is “in the money” there would be no reason not to exercise it and in practice exercise always occurs absent an operational error or an oversight on the part of the buyer. Where there is a combination of options, the party in the money is always treated as having exercised the option.
Mr Malik’s evidence was that TPFs were very popular vehicles for investors who wished to express a highly customised view on a currency pair. Although profits were limited by the target feature, the investor was compensated for the limits on the upside by not being required to pay an upfront option premium and by a strike price that allowed the investor to buy or sell the currency pair at a better than market strike price. The agreement to buy and sell the currency was structured through the purchase and sale of embedded options by the bank and the investor to one another. Both TPFs and Pivot TPFs can be described as transactions that provide the holder with a higher probability of a relatively smaller gain and a lower probability of a relatively larger loss. The experts agreed that the maximum loss was readily capable of calculation by multiplying the daily notional by the number of fixings. In practice, the loss would not be expected to be anything like that amount but would all depend upon the market movements.
If the example is taken of a TPF with a daily notional of 100 on a USD/JPY trade, where the holder makes a profit if the FX rate settles at above 100 and makes a loss if it settles below, with a target profit of 5, the pay-off to the investor on any day represents the result of the investor purchasing a call option (which benefits if the FX rate rises above the strike price of 100) and the pay-off to the bank on each day can be regarded as the investor selling the bank a put option at the same strike price of 100 so that the position loses money if the FX rate falls below that level. Thus the purchase of a call option and the sale of a put option at the same strike price is embedded within a TPF structure on each Observation Date or fixing date. As Mr Malik explained, the FX market has evolved so that the transaction documentation (the term sheets and trade confirmations) simply document the pay-off rather than documenting a series of combinations of options for every day of the duration of the transaction. Each option is taken to be exercised if it is in the money on any Observation Date so that profits and losses accrue over the period of the transaction, subject to hitting the
target profit, in which case the transaction knocks out. That represents a condition which applies to the body of options taken as a whole and consequently to each and every option, when considered in conjunction with the others.
In a Pivot TPF, the payment by each party is by reference to three strike prices, as previously set out. The pay-off in a Pivot TPF is also structured by the use of embedded options. The investor and the bank purchase and sell to one another a series of embedded daily call and put options at different strike prices. The target feature once again operates as a condition of each and every option transaction, taken in conjunction with the other. If, for example, the FX rate is 101 and the low strike price is 98, the pivot 100 and the high strike price 102, the pay-off represents the following four options for each fixing. The investor sells an out of the money put option at the low strike price of 98. The investor purchases one in the money call option at the low strike price of 98 (the two in combination operate as a synthetic forward with a strike price of 98). Additionally, the investor sells two call options at the pivot of 100. Holders are again taken to have exercised their rights under these embedded options if the options are in the money. The effect of this combination of options is that the investor makes a profit if the FX price is between 98 and 102 (above the low strike price and below the high strike price) and makes no money at all if the FX rate equates with the high or low strike prices. As soon as the rate moves outside the high or low strike, the investor starts to accumulate losses. Thus the purchaser of a TPF is expressing a market view about volatility, hoping and expecting that the currency stays within the high and low strike prices. The closer the FX rate to the pivot, the more profit is earned until the target figure is reached. Generally, the longer the transaction goes on, the more chance there is of a change in the rate which would take the FX rate beyond the strike level with accumulation of losses in consequence. Again, as with TPFs, investors are compensated for the loss of potential upside by avoiding the need to pay any upfront premiums as well as receiving a better strike price at the outset. In the event that the investor has to pay out on a pivot TPF, the exposure is equivalent to that of the sale of vanilla options.
The term sheets and trade confirmations set out the basic details of trade date, notional amount, high strike, low strike and pivot, together with the observation dates and the duration of the transaction by reference to the settlement date. The target cap level and the target knock-out event are specified and an explanation given as to the sums owing by way of settlement depending on where the FX rate settles as against the pivot and strike prices. Each transaction specifically incorporates the provisions of the ISDA Definitions whilst the trade confirmations provide that, in the event of any inconsistency, the terms of the confirmation will prevail. The confirmation supplements, forms part of and is subject to any ISDA Master Agreement between the parties. The standard non-reliance Clause appears in the confirmation whereby the investor represents that it is acting for its own account, has made its own independent decisions as to whether to enter into the transaction or not and exercised its own judgment as to the suitability of it, without reliance upon any communication from the other party as investment advice or as a recommendation to enter into the trade. The investor represents that it is capable of assessing the merits of and understanding the terms, conditions and risks of the transaction and accepts those risks.
Although therefore SHI, in the person of Professor Wystup, points to the absence of any express language referring to the series of options exercisable on each observation
date, the incorporation of the ISDA Definitions and the reference to strike prices reveals the basis upon which these deals were originally constructed and the existence of the “embedded options”.
Mr Malik’s evidence was that, in his experience, which was substantial in the area, market participants would classify TPFs and Pivot TPFs as currency options. He produced trade publications and research documents which spoke about the matter confirming that view. Articles by representatives of BNP Paribas, HSBC, ICICI Bank and Unicredit in such publications as SuperDerivatives and FX Week reveal this market understanding to which Mr Malik testified. He referred to ISDA surveys in 2004 and 2008. He said that quants, pricing people who had worked with him and clients he had spoken to all referred to them as options, particularly those he advised in 2007-2008, often in relation to deals done in 2006 or earlier.
The pay-offs were determined by reference to an underlying currency pair or pairs. They were conditional and explicable by reference to a series of options, even though they were more usually labelled by reference to their conditional pay-off features than as options per se. The existence of the negative pay-off was a function of selling options, rather than buying them. Of course even a bought option has a negative payoff in the sense of premium paid in the ordinary way, but the sale of an option will always give rise to a negative pay-off after the event (if exercised) because the seller is obliged to sell when the market has risen from the agreed price on the exercise of the option by the buyer. Whether a straight line TPF was seen by some as a strip of cash flows or a strip of forwards, that was not the issue because the market participants classified them as currency options. Moreover the effect of incorporating the ISDA definitions was to treat these as currency options within those definitions, as between SHI, the Counterparty and DBAG.
Professor Wystup’s position in reality depended not upon market understanding on which he was probably ill-equipped to speak in terms of trading experience but on his analysis of what an option truly was and his construction of the contracts. Neither was a matter for an expert unless the analysis was one which was shared in the market as the basis for classifying trades. Professor Wystup divided products into options, option-like derivatives (which did not have optionality but did have a non-negative pay-off) and non-option like derivatives which had no optionality and could have negative pay-offs. He further classified what he described as vanilla options on the one hand and first generation, second generation and third generation Exotic options and structured products, refusing to accept that Structured Options was a recognised term in the market. He accepted that no-one in the market talked of trading first generation, second generation or third generation exotic options but maintained that TPFs and Pivot TPFs were traded as such and not as currency options. Exotic options were not currency options unless they fulfilled what Professor Wystup saw as the indispensible requirements of an option, namely the granting of a right to be exercised (not an obligation) and providing for a non-negative pay-off.
Whilst Professor Wystup said that many people in the market used terms loosely and carelessly as he had done on occasion and that his current analytical view was the correct way to view these products, even he recognised that one touch exotic options were treated in the market as options, without any exercise and where the buyer had no discretion. Furthermore, there is no doubt that the grantor of a call option is obliged to sell if the option is exercised and is at risk of a negative pay-off (a loss) in
doing so. Professor Wystup accepted in cross-examination that as soon as there was a leg of a transaction where the investor was short, there was the possibility of a negative pay out and a combination of a bought call and a sold put option could give rise to that.
The other pillar in Professor Wystup’s analysis therefore was that single products were to be distinguished from structures and strategies of trading and if an instrument included a series of options, whether linked with other terms and conditions or not, it could not itself be seen as an option but was a “structured product”. This was a phrase which Mr Malik saw as describing a combination of a cash product with a derivative, whether in the context of FX, Interest rate or Credit transactions. In my judgment, it is a term which covers a multitude of products and is so imprecise that it is of little value. Professor Wystup said a package of options was not an option and that an instrument was not the equivalent of its component parts, nor of the products to which it could be decomposed, nor of the products that could dynamically replicate its risks. He said it was highly artificial to explain the EDTs and OCTs in terms of options even if they could be broken down and represented in this way. It is hard however to see how an option can cease to be an option when combined with other options in a package and, when authority is given to trade in options in the plural, why there is a problem with structures which put together a series of options. Even on its own terms, Professor Wystup’s analysis did not hold together.
Professor Wystup said he could not give evidence of a general understanding in the market that currency options meant only vanilla options. He did not maintain that currency options could not include exotic options but concluded that this was what was meant by the phrase when used in the FXPBA. He said that there was no general market understanding, though there were people in the market who regarded TPFs as options and it was reasonable for them to be booked in an options book by an options trader at an options desk. He refused to accept that structured options were seen by the market as synonymous with exotic options despite the market literature to that effect.
Professor Wystup’s view founders on a number of obvious points:
The question is one of market understanding not analysis of the constituent elements in any transaction.
An option transaction can plainly include any number of different terms and conditions above and beyond a straightforward put or call option, without ceasing to be an option.
The FXPBA specifically refers to Structured Options as having “special features” other than a “single barrier”. Barrier options with more than one barrier are self-evidently possible candidates as Structured Options.
A currency option does not cease to be such if it appears in the same instrument as another currency option. Even Professor Wystup was prepared to some extent to recognise that combining two currency options in one transaction could not negate their character as options.
A non-negative pay-off is not a constituent element or essential criterion for an option. The sale of an option when the market has moved inevitably involves a loss, if exercised.
The authority given to Mr Said and SHI was to trade in currency options in the plural and Structured Options are plainly regarded as a class of such options under the FXPBA for which DBAG’s prior consent is required. It cannot therefore be said that Mr Said’s authority did not extend to trading currency options in combination or as part of a structure with special features, terms and conditions. To the contrary, this is exactly what appears to have been envisaged.
The trade confirmations specifically referred to the ISDA Definitions effectively providing that these transactions be treated as governed by ISDA terms.
If the market regards EDTs and OCTs as currency options, structured options or exotic options, no questions of decomposability arise. The reason however that they are accepted in the market as being currency options is because their pay-off characteristics derive from the combination of options, combined with the target knock out feature and the acceptance that each option is taken to be exercised when it is in the money.
Once this is recognised, it is clear that there is optionality within the transactions albeit that the exercise of in the money options is taken as read in each situation, whether on the part of the investor or the bank.
There is no difficulty about the FX rate being “the underlying” for any of the EDTs or OCTs (save for the correlation swaps). The transactions are entirely sensitive to the movement of the FX rate in question.
As Professor Wystup recognises, all zero cost options have negative MTM at inception since otherwise a bank would make a loss at once. The absence of premium accounts for this and it cannot be a requirement of an option that premium be paid.
The EDTs and OCTs were traded with various counterparty banks as currency options and the evidence of Mr Said, in his depositions, and indeed in the email exchanges with Mr Vik and others, shows that this is how he considered them too.
Professor Wystup’s classification fits neither the contract nor market understanding. Professor Wystup’s position is untenable in the light of the contractual documents which do not limit currency options to vanilla options only nor to some class of standardised market product of first generation exotics. His views are inconsistent with the literature about the market, including the Foreign Exchange Committee of Federal Reserve Bank of New York’s Annual Survey of 2008, the ISDA 2004 Operations Survey, the ISDA 2008 Survey and the Bank for International Settlements Triennial Survey, quite apart from Mr Said’s evidence and Professor Wystup’s own views expressed to Freshfields. Currency options are not limited in the way in which he suggested and in the end he was prepared to accept that it was a term which could
have a narrow meaning of vanilla puts and calls or it could have a broader meaning where exotic options were included. He said it was all a question of context.
The context here is the FXPBA and the Said Letter of Authority and of a general market understanding of which Mr Malik gave evidence and whose evidence I unhesitatingly accept. TPFs were traded from 2004 onwards and much more commonly in the years 2007/2008. They were regarded by those trading in the market as currency options and as Currency Options within the meaning of the ISDA Definitions. In particular they were seen as structured options or exotic options.
While I regard the TPFs and Pivot TPFs as little more than bets on currency (hence Mr Said’s description of them as “range bets”) the market developed these products out of options and considers them to be currency options which can be used as speculation – as a bet on the market movements of currencies.
Not only was Professor Wystup’s initial view, as expressed to Freshfields, consistent with this, but so also was an article which he had jointly authored in which he described target redemption products under the heading of “Options”. Further a course offered by his consulting company again referred to target redemption forwards in the context of foreign exchange options.
The market view is what counts and I am satisfied that Mr Malik’s evidence, which coincided with that of Professor Wystup when first approached by Freshfields, reflected the market view. Professor Wystup’s reasoning did not hold together and his approach in evidence was not to give evidence of market understanding, but of his analysis and construction.
Whilst the focus of the evidence was on the TFPs and Pivot TFPs, the effect of the evidence covered the range of EDTs and OCTs. I accept therefore Mr Malik’s market view in respect of all the EDTs and OCTs and not just the TPFs and pivot TPFs, so that all (save the Correlation Swaps) fall to be treated as currency options which Mr Said was entitled to trade under the FXPBA and Said Letter of Authority.
11(b) Clause 2(iii) of the FXPBA
I have decided that there were no oral agreements of the kind alleged by SHI so that Mr Said did not act in breach of the authority given to him in other respects, whether by reference to any supposed financial trading limits (the Capital Limitation Agreement or the PAL), “trading on credit” or limits on the type of business to be concluded. There remains a further point taken by SHI in its closing submissions where it is said that thirty-one of the EDTs were unauthorised because there was no compliance with the terms of Clause 2(iii) of the FXPBA which required any Structured Option proposed by SHI as a Counterparty Transaction to be approved by DBAG at the time of its execution by SHI, failing which DBAG would not be responsible for the Counterparty Transaction in question.
This point goes nowhere. Mr Said’s evidence was to the effect that all the “offline trades” were the subject of prior approval by Mr Quezada or Mr Walsh but even if that had not been the case, DBAG accepted the trades by processing them and settling them, thus accepting liability on them to the Counterparties. The Counterparties have been paid in respect of the transactions in question. DBAG was entitled to waive the
need for prior approval, should it so wish and, if it did so, SHI would be bound by the off setting transaction which it had agreed to conclude with DBAG in respect of any such Structured Option accepted by DBAG.
12. The VaR Parameters
Although a great deal of time and energy was spent on the subject of VaR during the course of the proceedings, I need not dwell on it to the same extent as many of the issues which arise because, at the end of the day, there was a considerable measure of agreement between the parties and their respective experts on the subject. Those experts were impressive and had undoubted expertise in the subject matter.
I have set out in other sections of this judgment the early history by which Mr Said came to agree VaR terms with DBAG in November 2006 and how that was enshrined in the FX CSA as 2 x 5 day VaR. I have also set out elsewhere:
the difficulties which the ARCS VaR system had in coping with complex trades which included the EDTs and many of the OCTs and
the problems which arose in respect of the MTM on vanilla trades as reported in GEM as a result of feed issues from ARCS VaR.
There was however a further issue which ran parallel to those issues from March 2008 onwards in relation to the perceived inadequacy of the agreed VaR parameters in capturing the risk in SHI’s portfolio. Mr Gunewardena, Ms Serafini, Ms Liau and Ms Miranda all joined DBAG as a team from GS in 2006, and Mr Spokoyny joined them in 2007 as a risk officer. He was set to work to build a new margining and risk system for the FXPB platform in circumstances where DBAG calculated Initial Margin for its FXPB clients by two different methodologies, namely NOP (Net Open Position) and VaR. The majority were on NOP margining. Variation margin reflected the MTM of the transactions or assets in question under either methodology. Mr Spokoyny and Ms Serafini introduced a new stress-based margining methodology which was implemented for a limited number of clients in 2008.
Margin requirements for NOP clients were calculated by DBAG’s GEM system itself, both for IM and variation margin (MTM). For VaR clients, the potential future exposure and MTM of positions and the aggregate MTM of the portfolio were separately calculated by the ARCS VaR system and then fed into GEM. GEM would then subtract the potential future exposure from the aggregate current exposure of the portfolio to derive the 5 or 10 day VaR. Any VaR multiplier that had been agreed with a particular client would then be applied by GEM to give an overall initial margin requirement for the client. Additionally, however, a liquidity add-on could be calculated using a template spreadsheet that sat above, and took information from GEM. Such calculations were then added to the margin shown on the GEM system and details would be sent to CMV who would input this into their system to generate the margin requirement. Trade details, MTMs and portfolio level VaR requirements and the overall margin requirement (including both initial margin and variation margin) were reported to clients using a web-based reporting interface linked to GEM. Liquidity add-on did not appear there and, if agreed, would appear in margin statements from CMV which were usually sent only if there was a margin call or if requested by the client.
On joining DBAG, Mr Spokoyny began a general exercise to review the margin terms which DBAG had in place with all its FXPB clients and the types of assets that were posted by those clients as collateral, whilst also considering the counterparty risk posed to DBAG by those clients. As part of the review he designed a series of stress tests which worked by applying hypothetical “shocks” in exchange rates, volatility and interest rates to a client’s portfolio and to specific currency pairs: e.g. dramatic changes caused by large moves in the oil price and similar dramatic economic events. The object was to ascertain the magnitude of losses that any client might suffer if any of these hypothetical shock events occurred. The hypothetical losses thus calculated were then compared to the level of collateral required from each client under existing margin terms and if there was a significant shortfall, this was referred to within DBAG as “Gap Risk”. If a Gap Risk was calculated, it was intended that the FXPB team should seek to negotiate amendments in margin terms, or increase the level of fees, or ask the client to reduce risk in the portfolio or, in the worst case scenario where DBAG would not want to continue acting as prime broker, terminate the FXPB relationship. The results of the stress tests that were carried out were not typically shared with clients.
Neither Ms Serafini nor Mr Spokoyny considered that the VaR simulation used by ARCS VaR protected DBAG sufficiently against counterparty risk for VaR margined clients and steps were generally being taken to negotiate a change from VaR-based margining to NOP-based margining with clients who traded options. In 2007 and 2008, Mr Spokoyny believed that a more conservative confidence level should be used and that the Monte Carlo simulation used by ARCS VaR should use historical data for longer time periods and should incorporate “jumps” or pre-determined movements in the exchange rate for developed market currencies as well as emerging market currencies (which it already did). He also held the view that ARCS VaR did not accurately account for the increased liquidation costs that arose if a client held particularly large positions.
In August 2007, a potential Gap Risk was identified in relation to SHI but prioritisation of other accounts meant that Mr Spokoyny did not complete his stress tests and report on SHI until 11th March 2008, when he identified significant Gap Risk. There was thereafter discussion as to whether the counterparty risk on the client should be borne by PWM, by FXPB or GPF where most of SHI’s assets with DBAG and DBS were by then located. Mr Spokoyny continued to carry out stress tests on the portfolio as it appeared in GEM which included some resurrecting fader options. His evidence was that they appeared in groups of four entries which appeared to him to be offsetting put and call options. He did not consider that they added materially to the risk in the portfolio. Mr Spokoyny’s evidence was that he did not know that each set of four entries represented an inadequate proxy for an EDT with a large notional value when all the daily fixings were taken into account. The stress testing which he carried out therefore failed to take account of the true nature of these transactions and the full risk presented by them, even where they did appear in GEM. Where they remained unbooked, there was nothing Mr Spokoyny would have been able to see and therefore no reason for him to know of their unbooked existence. It was not until October 2008 that he came to understand the true position.
Between late March and early May of 2008 there was a series of internal discussions about this Gap Risk, involving Mr Spokoyny, Ms Liau, Mr Giery and Mr Quezada of
FXPB, Mr Brügelmann of PWM/DBS, and Mr Lay and Mr Halfmann of PWM CRM.
By the time of the first of the telephone calls on this subject on 28th March 2008, Mr Spokoyny had identified a Gap Risk which could be as much as US$100 million, without reference to the terms of the EDTs (of which four had been concluded by this date). In the call, Mr Spokoyny suggested NOP terms by reference to the different currency tiers or an amendment by increasing the VaR multiplier and changing 5 day terms to 10 day terms. There was a suggestion that Mr Said created strategies of a kind that did not lend itself well to the VaR calculation. Matters were left on the basis that FXPB should come up with figures that should be discussed.
An analysis carried out by Mr Spokoyny subsequently showed a Gap Risk of US$95.5 million on his stress testing. On NOP margining, his calculation showed that Gap Risk would not exist.
On the 9th and 10th April there were exchanges of emails in relation to this Gap Risk. Mr Lay’s immediate reaction was to say that the VaR FX limit for SHI was suspended and that no further exposure could be entered into by the client which provoked the immediate response from Ms Liau that SHI was not in breach of its existing margin terms. Mr Brügelmann, with the aim of overcoming the issue, asked for further details of Mr Spokoyny’s analysis and sent an email to Mr Orme-Smith of GPF requesting that he block US US$100 million in SHI’s GPF account to secure any potential shortfall until new margin terms had been agreed. Mr Orme-Smith responded “Noted – thanks”. CRM personnel regarded this as an effective blocking and Mr Brügelmann plainly regarded this as a practical solution, even if the ring fencing was not legally enforceable. SHI was never told about this and waxed indignant about it at the trial. Nothing ever came of this ring fencing and GPF Risk objected when the matter came to its attention and Mr Orme-Smith, on 21st July 2008, stated that his email was never intended to be an agreement to such ring fencing and that he had never been in a position to achieve that in any event. In the meantime on 15th April 2008 Mr Spokoyny sent Ms Stingelin details of his stress methodology with a Trade Details Report downloaded for GEM, which included eight entries for “Resurrecting Fader Options” (four each for EDTs 1 and 2). Each entry included N/A in the MTM column which should, in my judgment, have alerted Mr Spokoyny to a potential problem, but, on his evidence, did not. His review of Trade Details in GEM did not lead him to think that there were trades which materially added to the risk of which he was failing to take account. Had he investigated the position then, as he did in October 2008, the problem would have been evident to him.
On the 2nd May 2008 there was another conference call with Mr Giery, Mr Quezada, Mr Brügelmann, Mr Lay, Mr Halfmann and Ms Stingelin who was the Head of Private Client FX. The purpose was to discuss a new prime brokerage account that SHI wanted to set up for Mr Vik for FX and/or FI and the margin terms for the existing FXPB account. In the course of this call, Mr Brügelmann explained that there could not be separate FXPB accounts for the same legal entity with different margining methodologies. Whilst Ms Serafini and Mr Spokoyny saw the solution as being the application of the NOP methodology to Mr Said’s trading account, Mr Giery and Mr Quezada explained that Mr Said was concluding structured options and pivot faders that they did not think could be margined by NOP but could be captured in ARCS VaR. The suggestion was made of trade level margining by a methodology approved by CRM. Mr Spokoyny expressed his disquiet that the calculations performed by ARCS VaR were less than transparent because one could not see each and every trade being modelled but only a composite number produced as the margin figure. He explained that the ARCS VaR system could not readily be changed in order to give DBAG what he considered an adequate protection against the counterparty risk and this would affect all other clients and involve significant investment in hardware and time. In that call, Mr Quezada referred to pivot accrual faders which presented difficulties in booking, saying his only recourse was to go to a quant expert in the sales desk who was reluctant to help. He expressed the view that if these types of trade could not be booked in an automated fashion, FXPB would probably not accept them. Mr Giery said that they could not book some of the stuff that Mr Said was doing at the moment and that it would take somebody in the Structured Options Group to do it and they would then have to value it daily using DB Analytics spreadsheets which FXPB did not have.
The end result of the telephone call appeared to be a consensus that, if Mr Said wanted to trade in such “funky stuff” he should only do so as direct trades with the DBAG trade desk which could use the DB Analytics tool, and that such trades should not be accepted by FXPB for give up even though this would mean a “tough conversation with Klaus because he saw it as a window of opportunity to make money”.
It was then suggested and agreed that Mr Brügelmann, in his scheduled meeting with Mr Vik on May 7th, should tell him that there was a need to change Mr Said’s margining to NOP, and for NOP to be used for any FX and FI trading by Mr Vik. Mr Brügelmann said he would seek to spin an argument on the basis of concentration of risks, liquidating risks and correlation breakdown. The VaR solution was not however completely out of the question if there was development which enabled subaccounts to function on the system with different margining methodologies. Mr Quezada wished to go down that route because of the limitation that NOP might impose upon the types of trade which Mr Said could conduct, even though he had recognised earlier in the call that there were still issues about how to book the trades, whether VaR could capture the full risk and that the easy solution was to insist that he conducted the “funky structures” in direct trades with the DBAG trade desk or with CS or MS under separate ISDA agreements with them.
Following this call, Mr Walsh and Mr Quezada spoke to Mr Said on 5th May in a telephone call and Mr Brügelmann met with Mr Vik on 7th May. I have set out elsewhere in this judgment details of the telephone conversation of 5th May in which Mr Said persuaded Mr Quezada and Mr Walsh to take in the EDTs on the basis that he did not require the MTM to be reported, that there was nothing for DBAG to do save enter cash settlements at knockout and stated that margining was a matter for DBAG not for him. In the meeting on 7th May, Mr Brügelmann told Mr Vik of the issues in relation to setting up an FXPB and FIPB account for him, and the potential deficiency in the existing margining arrangements for Mr Said. I have made findings about exactly what was discussed on that occasion but the idea of using a separate legal entity for Mr Vik’s FX Prime Brokerage potentially overcame the difficulties of sub-accounts with different margining methodologies. Mr Vik refused to discuss the potential shortfall in the margin calculations and insisted that Mr Brügelmann talk to Mr Said about it. The overall effect of these exchanges was that Mr Said’s trading would continue on VaR (including EDTs) whilst DBAG, for its own protection,
wanted to amend the VaR parameters to ensure a larger measure of protection and would discuss them with Mr Said.
On 14th May Mr Brügelmann sent an email to Mr Quezada, Ms Liau and Mr Spokoyny saying that he had spoken to Mr Said who was expecting Mr Quezada to call to discuss the VaR issues.
By 19th August Mr Spokoyny had conducted further stress tests which indicated that the Gap Risk had reduced to about US$40 million but the EDTs had still not been taken into account because of their absence from GEM and Mr Spokoyny’s lack of understanding as to what the resurrecting fader entries truly represented. In a call of that date, Mr Spokoyny explained that Mr Said’s net open position in his FXPB account had reduced from over US$1 billion earlier in the year to some US$600 million at that point and that the current margin requirement was US$23.5 million when he thought the current risk was actually more like US$50-60 million. On discussing what changes might be required to the VaR parameters, Mr Brügelmann said that he had spoken to Mr Said who had told him that he would not agree a move to NOP margining. Mr Spokoyny said that the 95% confidence element of the VaR calculation could not be changed but the multiplier could be altered and a liquidity add-on employed. He then carried out calculations thereafter and worked out what such an appropriate multiplier might be.
12(a) The Changed Parameters
It was on 8th September 2008 that Mr Quezada and Mr Spokoyny met with Mr Said at DBAG’s offices in New York. Details of this meeting appear in section 15 of this judgment but in essence Mr Spokoyny explained the need for an increase in the margin terms and how his stress tests worked while Mr Said stated his view that these stress tests gave rise to conservative results. Mr Said was told that there were still problems in booking the trades but that DBAG would get to it shortly.
With no agreement reached on 8th September, Mr Quezada and Mr Spokoyny spoke on the telephone to Mr Said on 30th September, suggesting amended VaR terms of 3 x 10 day VaR + liquidity add-on, which, as set out in section 15 of this judgment, Mr Said was not prepared to accept, seeking to minimise the margin to be provided. Mr Said said that DBAG probably had no risk on the account because Mr Vik stood behind it.
On 6th October Mr Spokoyny emailed Mr Said to inform him that he had received approval within DBAG to amend the margin terms to 2.5 x 10 VaR + liquidity add-on which would have the effect of increasing Mr Said’s current collateral requirement from approximately US$21 million to US$40 million. Mr Spokoyny sent with the email an attachment setting out the liquidity add-on methodology. Within a matter of minutes Mr Said replied, accepting that offer with the words “That seems fair. I can live with that.”
I have decided elsewhere that Mr Said had authority to vary the margin terms and conclude therefore that the new margin parameters were applicable at the time when the first margin call was made. In any event, SHI accepted Mr Said’s agreement as binding on it when paying the margin calls between October 13th and October 22nd, thereby ratifying any excess of authority.
12(b) The Computer Models
There is one further range of issues concerning VaR which relates to the different computer models built by the parties’ respective experts Mr Millar and Dr Drudge. The complexities of some of the structured options, including all of the EDTs, were such that DBAG’s ARCS VaR system could not capture the risk. In order to calculate DBAG’s contractual entitlement, each of the parties instructed an expert to construct a model which could calculate margin in accordance with the FX CSA which required VaR to be calculated in accordance with the methodology determined “[by DBAG] in its discretion which it customarily uses with its counterparties”. The models built by Dr Drudge and Mr Millar gave rise to comparable but different results.
Whilst, in the light of my other findings, it does not matter which of these two models produces more exactly the contractual margin requirement, out of deference to the work done on the subject I set out my conclusions on this matter.
Although at one time DBAG put forward methods of calculating Initial Margin, in the course of 2012 it recognised that this was not a VaR methodology at all and, during his cross-examination, Mr Millar accepted that zero VaR could not constitute part of a VaR methodology either.
In overview, the purpose of the methodology utilised by DBAG was to estimate, using recent market data, the potential change in the value of a portfolio of financial instruments over a specified time period and within a specified confidence level of 95%. The specified time period was usually two days, five days or ten days. The “Disclosed Methodology” of which this is an overview, was the only VaR methodology used by DBAG in respect of counterparties (i.e. customers of DBAG who traded with DBAG’s FX Trade Desk and/or under a FXPB relationship with it) who were margined according to VaR methodology between 2006-2008. The methodology was calibrated to calculate the potential change over a five day period and consisted of the following components:
market data extractions/transformation (used to generate the data to be used by the Monte Carlo engine),
the Monte Carlo simulation engine (used to generate one thousand paths for the FX spot rates used by the pricing engine),
the Pricing engine (using the trade data, pricing functions and simulated risk factors),
the P&L vector construction module (which took the difference in MTM between the MTM of the trade value under one of the one thousand Monte Carlo scenarios and the original value of the trade (i.e. the current actual MTM of the trade)),
the VaR calculation module.
The component parts of the methodology are illustrated in Annex 4.
The market data used by ARCS VaR to value a portfolio of FX trades consisted of FX spot rates, LIBOR interest rate curves and at-the-money FX implied volatility curves. The market data were then used to produce the necessary inputs for the Monte Carlo engine. The last ninety days of London close of business daily spot FX rate changes with EUR as the Reference Currency were used to generate a linear correlation matrix which was then used to control the evolution and joint behaviour of spot rate changes in one thousand scenarios so that FX rates that tended to move in line with each other tended to maintain the same behaviour in the Monte Carlo simulation. The last ninety days of London close of business daily spot FX rate changes were also used to calculate the historical volatility of each currency pair. The standard deviation of the returns was used in the Monte Carlo engine to determine how much an individual FX spot rate can change.
The Monte Carlo engine within the VaR process generated one thousand new sets of FX spot rates. Based on the starting FX spot rates, a new set of spot rates was predicted. Each FX rate could move up or down each new day in a random Brownian motion with the size of the random move related to the standard deviation of the FX rate but with the FX rate changes of all the currency pairs constrained by the correlation. Occasionally, a “jump” could occur. This occurred with a degree of probability to which I shall return and did not need to obey the derived correlations. There was an additional rule in the model which did not allow more than one jump to occur for a currency pair.
In this way one thousand predictions were made for what the FX spot rates would be in five days time. The conditions used to generate these (the historical standard deviation, correlations and pre-defined jumps) were known on any day but the final predicted rates were slightly different every time they were generated and the system recorded the predicted rates for a number of different periods.
The methodology assumed that all other market data (implied volatilities and interest rates) did not change over the simulation period.
The change in value of the portfolio of trades was calculated for each of the one thousand five day scenarios (i.e. the difference between the scenario valuation and the original valuation). These were ordered in terms of descending loss (i.e. the biggest potential loss was the first). The 95% five day VaR is the fiftieth largest potential loss of the one thousand losses calculated.
Attention should be focused on a number of elements in the disclosed methodologies, namely the Monte Carlo engine, the jumps and the approach to implied volatilities.
In their joint memorandum Mr Millar and Dr Drudge set out the areas of agreement and disagreement between them in relation to their respective VaR methodologies. Mr Millar used DBAG’s proprietary Monte Carlo VaR methodology (the Disclosed Methodology or DM) whilst Dr Drudge used an alternative approach, commonly used in the Financial Services industry, which was referred to as the Prism Methodology or PM. Mr Millar and Dr Drudge agreed that the use of either a Monte Carlo simulation as used in the DM or an historical simulation as used in the PM was an acceptable theoretical basis for the calculation of VaR and both were in general use in the Financial Services industry. Assessment of the relative qualities of the DM and the
PM and of their results is necessary in the light of the portfolio of trades and the market conditions.
Mr Millar and Dr Drudge agreed that the VaR results calculated by them were consistent with their respective methodologies and where there were differences in the results, they identified the primary factors that they believed caused those differences, as set out below:
“a. The existence of jumps in the (Monte Carlo based) Disclosed Methodology. Jumps do not form part of the historical simulation used by Dr Drudge to calculate VaR. The jumps have the effect of increasing the VaR results where specific currency pairs are included within the portfolio.
Differences in the historical time period used to generate predictions of potential future losses. The Disclosed Methodology uses a shorter period of historical data (90 business days) than the Prism Methodology does (250 business days). Mr Millar and Dr Drudge agree that generally where a shorter period of historical data is used, more recent changes in market data have greater impact on the VaR results.
Differences in the way that potential losses over a five day period are computed between the Disclosed Methodology and the Prism Methodology. The Prism Methodology calculates losses over a one day period and then scales these figures to five days whereas the Disclosed Methodology uses an estimate of the variation in the underlying variables over a five day period and uses these to calculate losses over this period. In the case of SHI's FX portfolio the risk profile is such that the effect of using a five day computational method increased the VaR results.
Differences in the treatment of potential changes in implied volatility. The Disclosed Methodology does not stress implied volatility, but the Prism Methodology does. Including such a stress within the methodology should generally have the effect of increasing the VaR results under the Prism Methodology.”
The effect of these different factors was that the VaR results produced on SHI’s FX portfolio using the DM produced larger VaR figures on a portfolio level than those produced using the PM.
The Joint Memorandum continued:
“2.9.1 Mr Millar and Dr Drudge agree that implied volatility is a factor affecting valuation of TPFs, particularly at deal inception. In addition, when considering the impact of stressing the implied volatility parameter at an individual trade level, Mr Millar and Dr Drudge agree that there can be a significant impact on the overall VaR calculation for the individual trade, as explained in further detail in paragraph 2.9.4 below. However Mr Millar and Dr Drudge also agree that the impact of stressing the implied volatility parameter at portfolio level will not always be material, depending on the particular constituent trades contained in the portfolio and market conditions, as set out below.
Mr Millar and Dr Drudge agree that their VaR results indicate that the approach set out in the Disclosed Methodology which does not include the simulation of changes in implied volatility contains other features, as set out above in paragraph 2.8.2, which result in higher VaR calculations than Dr Drudge produces using the Prism Methodology when applied to SHI's FX portfolio. The Prism Methodology does include the simulation of changes in implied volatility.
On this basis Mr Millar and Dr Drudge agree that when applied to SHI's FX portfolio (for example as produced for each of the Alternative Scenarios by Mr Millar), the Disclosed Methodology does not produce unreasonably low VaR results as a result of not simulating changes in implied volatility. Mr Millar and Dr Drudge's area of disagreement regarding commercially reasonable VaR estimates is detailed in section 3.
When considering the theoretical impact of stressing the implied volatility parameter at portfolio level, Mr Millar and Dr Drudge agree that the following factors are likely to be relevant:
The relative sensitivities of the individual trades to other risk factors being stressed by the model, in this case FX spot rates, driven by the current economics of the trade in the portfolio. For example a TPF that is out of the money would be much less sensitive to movements in implied volatility in comparison to a TPF that is not out of the money, for example a TPF that had been recently traded.
The relative variability of each risk factor. For example if as a consequence of a particularly quiet historical observation period implied volatility was not expected to move very much then it would not be expected that VaR would be highly affected, and conversely if implied volatility had experienced large moves over the historical observation period, this would have a greater impact on the VaR.
The degree of diversification across all of the trades in the portfolio and how this generally decreases the marginal impact that risk factors have at trade-level. For example whereas the valuation and risk of a certain trade might be strongly affected by a certain risk factor (i.e, an FX spot rate or implied volatility), if the proportion of trades in the portfolio affected by that risk factor is small, then at portfolio level the risk factor may not be significant at all.
The inter-relationship between these aspects of the VaR methodology. For example in the case of EDTs the impact of including jumps in an FX spot rate may mean that, as the sensitivity to implied volatility decreases the more the trade is out of the money, and as VaR scenarios are likely to include jumps, the number of VaR scenarios where implied volatility would have a significant impact may be minimal for currency pairs where the Disclosed Methodology prescribes jumps.
2.9.5 As a result of this, the potential impact of shifting the implied volatility parameter on VaR results over time at the level of any particular portfolio will depend on the trades which constitute that portfolio over time. See the comments made by Dr Drudge below in relation to his conclusions regarding commercial reasonableness following examination of the VaR results.”
The areas of disagreement therefore were comparatively few. Dr Drudge agreed that the numbers produced by the DM for SHI’s FX portfolio as it was constituted over time were not unreasonable inasmuch as there was little difference in the VaR amounts computed using the DM and the PM. However in Dr Drudge’s view the DM would not be a commercially reasonable choice for all portfolios composed of trades of the types represented in SHI’s FX portfolio because it could have led to either systematic underestimates or over-estimates of VaR for instruments of the types listed in the portfolio. Mr Millar was unaware of any agreed yardstick to measure commercial reasonableness and pointed to the degree of latitude which existed in the margin calculations made by different banks.
Mr Millar considered that, for DBAG to move from the DM to the PM would represent a considerable challenge in terms of time and expense in creating a new model. Operational considerations would come into play in such circumstances. Dr Drudge in cross-examination accepted that it would be a significant decision for the bank to change to the PM from its existing system. Practicalities would arise with the impact on all portfolios managed by DBAG to be taken into account.
Backtesting is the principal performance indicator for a VaR methodology but the method must be theoretically sound and qualify as a VaR methodology. In Dr Drudge’s first report in reply he said that “the principal concern when assessing the reasonableness or otherwise of a particular approach to calculating VaR on a derivative portfolio is how it performs in back testing, in other words whether breaches of the VaR amount occur with approximately the correct frequency over the relevant VaR time horizon.” On backtesting for the whole portfolio, the “actual” outcomes exceeded the DM VaR figures 7.5% of the time as opposed to 10.42% of the time on the PM. Looking at the EDT portfolio alone, the difference between “actual” and the DM was 6% whereas the difference between “actual” and the PM was 15%. On a backtesting basis therefore the DM produced more accurate results.
The reason for this was that, in the events that happened, the VaR was substantially more determined by changes in spot rate than by implied volatility.
Dr Drudge agreed that, in the market conditions that were actually observed during the period, there was no systematic understatement of VaR for the TPFs on a portfolio basis by the DM.
Essentially Dr Drudge’s criticisms related to the jumps and the approach to implied volatility. DBAG’s ARCS system included jumps for emerging market currencies – in fact for everything except Tier 1 currencies, those which were considered the most stable. In fact the only currency pair with jumps that applied to the EDTs were the BRL trades, though jumps appeared on other traded currencies as well.
The point put to Mr Millar in cross-examination was that the jump up and jump down rate used in the DM was 7.24%. The mean size of the jump was 15% and the standard deviation was 7.5% of that. If the figures were taken as meaning literally 7.24 jumps per annum up and down, the probability of a jump in any five day period was 13.9% in either direction. With a generation of one thousand different paths in the DM, nearly 28% of those would experience a jump up or down which meant that jumps would occur more often than 5% of the time in any five day period. That, it was said, was not consistent with a 95% confidence level VaR. Mr Millar rejected that criticism.
The VaR number is the estimated loss in value of a portfolio that is unlikely to be exceeded over a specified time period, given a specified confidence level, here 95%. It is an estimate because it inherently requires assumptions. It does not guarantee a maximum loss figure and there is about a 5% chance that the actual loss will be greater than the VaR number produced over a five day period. It is not an attempt to capture all possible losses and the assumption is that in about 5% of cases it will not. This calculation of VaR at this confidence level is by definition the 95th percentile worst loss – the fiftieth worse loss in a thousand scenarios. The purpose of the jumps was however to make allowance for the risk associated with currencies exposed to unlikely but relatively severe events not reflected in the ninety or two hundred and fifty days historical data used in most VaR calculations. In the DM, as utilised by DBAG in its systems and by Mr Millar, a flat rate judgemental assumption was made for all currencies in particular tiers. It was an assumption which was made that fell into the 95% calculation but was not factoring in an unlikely risk that would necessarily occur in less than 5% of cases as such.
Mr Millar pointed out that a jump of the size in question actually occurred in October 2008 and was similar to the jumps in 1999 and 2002 with BRL. It could not therefore be said it was inappropriate and backtesting of the model showed that the assumption had proved correct. It was a subjective judgment about assumptions to be put into the calculation to get 95% VaR over five days. The assumptions that were put into the model were different to the actual assessment of the confidence level at 95%. If the judgment had been based on short term historical movement only, jumps would not be present but it was inherently reasonable to take into account events beyond the ninety day or two hundred and fifty day period and to allow for them in some way, as the jumps did. Backtesting showed the assumptions to be well founded for the SHI portfolio in the present case.
Banks used tools of this kind to build in assumptions about event risks and liquidity risks in order to factor them into VaR models, although they might not call them jumps or make the same assumptions. It was not however right to say that this was an external element that should be outside the VaR calculation in the same way as a liquidity add-on or a fat tail add-on which were specifically designed to take account of matters that the VaR calculation did not. Whilst what was being built into the 95% VaR figure was an event which might take place less frequently than in 5% of cases, that was not uncommonly done. There was a great deal of latitude in the definition of VaR when computing a 95% VaR figure and very often the jumps or tools used were achieved by reverse engineering. In fact, the BRL EDTs suffered greater losses in their lifetime than the VaR calculations for them under the DM. The DM therefore did not in fact take account of the worst possible loss scenario. The assumption was built in in order to get to the 95% confidence level and was not designed just for extreme events but also for lesser events covering factors like difficulties in closing out positions or liquidating portfolios. Mr Millar did not think it fair to say that the reasonableness of the parameters should be assessed solely by reference to the frequency of extreme events.
The DM took no account of the sensitivity of the portfolio to movements in implied volatility per se, whereas the PM did. There was compensation for the absence of the volatility parameter by the use of the jumps which made assumptions about changes in spot rate, though these were not directly related to volatility as such. If however implied volatility shift were to be inserted on top of that, there would be an element of double counting. The effect of the differences between the DM and the PM and their approach to implied volatility and the jumps depended upon the constitution of the portfolio in question and, in the case of SHI’s portfolio, when taken as a whole, it was more sensitive to movements in spot rates than to movements in implied volatility. Where jumps had the effect of offsetting the absence of implied volatility shifts, that might or might not be random but the effects on this portfolio had an offsetting effect in practice. Mr Millar made the point that he did not consider the sensitivity of this portfolio to vega to be very high and he could not build in implied volatility into the DM without making assumptions that, to his knowledge, did not replicate anything that DBAG would have done.
He accepted that all models were subject to a degree of imperfection in one respect or another, including as they did various assumptions and attempts to produce VaR calculations of risk to the bank. He did not consider that the DM was rendered unreasonable by exclusion of the implied volatility parameter that appeared in the PM. There was compensation by the provision of jumps, by the longer period during which the DM considered the change in FX rate (five days as opposed to one day scaled up to five) and the shorter time period of ninety days, as opposed to two hundred and fifty days which made the DM more reactive to short term changes in spot rates. When performance of the model as a whole was taken against “actual”, the DM achieved results which were closer to reality than the PM. The main differences between the PM and the DM appeared in late July to late September where the jumps made most of the difference and where the DM was closer to reality than the PM.
Dr Drudge accepted that the numbers produced by the DM were not unreasonable but targeted the jumps as an area for criticism. He accepted that the DM with jumps produced higher figures than the PM on 75 observations at trade level and that, (without jumps) the DM still gave higher figures than the PM at portfolio level. Dr Drudge accepted that the PM, without jumps, did not allow for events which might affect the spot rate but which had not been seen within its relevant historical time series, such as particular geo-political risk events, sovereign risk events and the like which did not occur with great frequency but could nonetheless take place. He accepted that jumps did seek to take these events into account and accepted that the range of movements in the spot rate under the DM, with the jumps for the USD/BRL, was in line with the range of movements in the spot rate that was observed in the market in September and October 2008. In his view however, jumps or similar tools, whether the result of stress tests or otherwise, were more commonly added on top of VaR rather than blended into the VaR methodology. His essential criticism was therefore that the jumps in the DM were to be found in the VaR calculation as opposed to being outside it as an add-on. He would not criticise the view that risk in emerging market currencies was not adequately reflected by a particular period of historical data and banks were entitled to make different assumptions of risk in producing their numbers. For him, the question for BRL was whether it was reasonable to put in a once in ten year event or a once in five year event, as he saw it, into a five day 95% VaR.
The following exchange took place in cross-examination:
“Q. The terminological or definitional difference, can I suggest, between you is that you are saying it may be the ninety-fifth worst output of this model or this disclosed methodology, but because it reflects within the assumptions used in the disclosed methodology more extreme assumptions for size and frequency of particular emerging market currency spot rates changing, it is not the ninety-fifth worst outcome during a normal market period?
A. Yes, I think that … I think that is about right. Ultimately you can put anything into the 95th percentile if you choose to, right.”
Ultimately, in cross-examination, Dr Drudge agreed that, for the portfolio as it was constituted, given the backtesting numbers that Mr Millar had produced, the bank was acting commercially reasonably in not changing its VaR methodology simply because it might be possible to come up with one that was perhaps even better in its backtesting performance.
My conclusions on the basis of this evidence are as follows, in the light of the contractual provisions requiring calculations, valuations and determinations to be made “in good faith and in a commercially reasonable manner” and for VaR to be determined by DBAG “in accordance with its methodology determined in its discretion which it customarily uses with its counterparties”.
DBAG’s ARCS system had jumps in it and that was not commercially unreasonable and represented a genuine attempt by DBAG to take account of events which could properly be taken into account in assessing a five or ten day VaR at 95% confidence level. Mr Millar was therefore right to include them in his DM model.
Although Dr Drudge considered that jumps were inappropriate because they catered for occasions falling within the 5% of occasions outside the 95% VaR, that is a restrictive approach with regard to the calculation of the 95% confidence level. There is much latitude given in the assumptions to be fed into a 95% VaR calculation and it is not uncommon to see such approximations made as a matter of reverse engineering in order to take into account events which fall outside the period of historical data which feed into the engine, but are recognised as being events which can and do occur (in the case of the BRL in 1998, 2002 and 2008). There is nothing objectionable about building in such events into the 95th percentile as opposed to treating them as an additional add-on outside the VaR calculation.
Jumps cater not only for extreme events but also for other factors such as the difficulty in closing out positions or liquidating portfolios where there are levels of less extreme stress. Jumps attempt to capture that as well as the more extreme situation.
The DM, which does not include the simulation of changes in implied volatility, has other features which take this into account in one way or another, namely the jumps, the use of a ninety day historical data period and the computing of potential loss over a five day time period as opposed to scaling up the outputs of a one day period. The jumps, in particular, compensate for the absence of implied volatility in relation to currencies other than Tier 1 currencies.
Backtesting shows that, for the SHI portfolio, the DM produced results closer to the actual than the PM, in particular in September and October 2008. There was no systematic under-estimate or over-estimate of VaR.
In consequence, I conclude that the calculations produced by Mr Millar’s DM model represent VaR, calculated in accordance with the methodology which DBAG customarily used with its counterparties and that the results were commercially reasonable and therefore represent DBAG’s contractual entitlement to margin in the relevant periods (subject only to any minor alterations necessary to take account of the Dual Currency Range Trade).
13. The problems created by the OCTs and the EDTs for DBAG's systems
SHI has devoted a great deal of time, energy and print to its submission on the vagaries of DBAG’s systems and in particular to the inconsistency between such limited print outs as exist in respect of the period 2007-2008 and what appeared on the system when access to DBAG’s GEM Web Reporting was restored on May 10th 2012. There are undoubtedly unexplained anomalies, despite the best efforts of Deloitte, DBAG’s forensic accountants, to resolve the questions.
It is also the case, as appeared clearly from the evidence, that there were problems with DBAG’s systems at the time. The exact nature of these difficulties does not, in my judgment, ultimately matter, because what Mr Said came to appreciate in 2007 and 2008 was that Structured Options, within the meaning of the FXPBA, created real problems for DBAG’s systems. As appears hereafter, he understood that various OCTs and all EDTs either did not appear at all in GEM Web Reporting or appeared sporadically, popping up now and again and that the MTMs, where they did appear, produced spurious numbers so that portfolio VaR margining, based upon such MTMs, was chaotic.
Since he made the point that he did not rely upon GEM as a risk management tool, was not interested in MTM valuations of his Structured Options and regarded margin requirements as simply a matter for DBAG, none of this concerned him save in so far as the appearance of MTM figures, when they did appear, might distort the overall figure for his trades or might produce a portfolio margin figure greater than DBAG’s contractual entitlement. In consequence he wanted the EDTs removed, when they appeared, or the MTM zeroed out. As he knew, the inevitable result of the nonappearance of the EDTs in the GEM web reports to which he had access and on which he was told that margining was based meant that he was getting “freebies” in respect of that element of the margin calculation which should have reflected the EDTs.
Nonetheless, in order to gain some understanding of the events as they unfolded, it may be necessary to set out something of DBAG’s systems. I do so by reference to SHI’s closing submissions in the following paragraphs.
“Booking
297. Although some simpler trades could be processed either automatically or semi-automatically after being entered by the client in TRM, complex trades (such as the EDTs and OCTs at issue in this litigation) had to be booked manually in RMS. Of course, for that approach to do any good, such bookings had then to flow through into the relevant systems to ensure that they were valued, margined and reported. In theory, the initial manual entry of trades proceeded as follows:
For “indirect” trades, i.e. those traded by Mr Said with third-party banks, FX PB (usually Mr Walsh) would book both legs onto the system. Where Mr Walsh was unfamiliar with a trade type, his general practice was to try to book it in the first instance and then ask for help if he had problems (but to ask first if he really had no idea). He would often book a trade as “pending” and then ask someone to check it. Trades saved as pending would be visible to other users but not flow through [DBAG’s] systems, and so not appear in GEM.
For “direct” trades, i.e. those entered into with [DBAG’s] trading desk:
Simple trades were booked into RMS by the salesman (such as Mr Geisker). Complex trades were booked into RMS by the trader (such as Mr Chin), who then sent the salesman an email containing its basic details, from which Mr Geisker would then create a Generic Sales Ticket.
In the usual scheme of things, trades booked by the desk to be given up to SHI’s FX PB account would be booked as a pair of offsetting trades between the trading desk and FX PB, although sometimes when FX PB could not book a trade it would ask the desk to book a trade directly to the client account.
Confirmation
Confirmations of options and complex trades (for both the client and the counterparty) were generated by the FX Options Operations team. Evidence on that process was given in particular by Mr Manrique. As he explained, once a trade had been booked in RMS it moved through a series of queues, through which many simple trades moved automatically, but complex trades such as OCTs and EDTs had to be manually processed. Until a trade passed the relevant check at each stage, it did not move on to the next queue. The procedure was designed to check that each trade was properly booked in [DBAG’s] systems, confirmed and processed.
First, trades entered the “New Queue”, where they remained until their RMS bookings had been verified by Mr Manrique’s team.
After being so verified, trades moved into the “Production Queue”, where they remained until confirmations had been produced for them. Generally speaking, the more complex the option, the more manual was the production of the relevant confirmation.
After their confirmations had been generated, trades entered the “Dispatch Queue”. All confirmations other than those automatically generated for some vanilla put and call options were checked by a second member of the team (i.e., other than the person who had produced the confirmation).
After confirmations had been sent, trades entered the “Return Queue” until they had been confirmed by the counterparty/client.
These queues were monitored regularly by Mr
Manrique’s team, and in particular his manager, Ms Oglivie, in order to identify and escalate trades – generally only exotics – that had become stuck, and were thus not being properly handled. Trades that had not been booked into RMS at all, of course, had no way of making it onto Mr Manrique’s radar.
Knock-out/settlement
300. If a trade knocked out, it was the job of the person who had booked the trade (whether within the trading desk or FX PB) to change its status on RMS. That led to it being placed in the “Knock-Out Queue.” A similar process occurred when a trade settled, although Mr Manrique was uncertain how far the triggering of settlement and calculation of the amount took place manually. In each case it was then the job of FX Operations to ensure that both parties agreed on the outcome.
Valuation and margining
VaR, on an FX PB client’s portfolio, was calculated (or at least was meant to be calculated) in a system known as ARCS, which took a feed of open positions from RMS and in turn fed the VaR calculated on that portfolio to GEM, which reported the figure it received and included it in its margin calculations.
The system that calculated MTM valuations of a client’s individual trades depended on the client’s margin basis:
NOP-based clients’ positions were valued directly in GEM, which could value and margin only a very limited set of trades: spots, forwards, non-deliverable forwards, swaps, Euro options and single-barrier Euro options. Of those, only the latter (termed Knockout Currency Options in the experts’ lists) fall within the sets of trade types referred to in this litigation as OCTs (and then at the simplest end).
For clients margined, like SHI, according to VaR, MTM valuations of individual trades were provided (or at least were meant to be provided) by ARCS. ARCS could value a wider range of trade types than GEM, but still only a limited set. That fact was known within FX PB, although knowledge of precisely which trades ARCS could handle was rather more patchy, as the events described below demonstrate.
ARCS’s trade-type limitations applied to its calculation of VaR as well as to that of MTM on individual trades. [DBAG] admits that it failed to report accurate MTM valuations of, and ARCS was unable accurately to margin, the trades known in this litigation as EDTs and OCTs.”
Mr Said was given access to the GEM Web Reporting System which was accessible through a web browser, displaying a number of reports generated by DBAG’s GEM system which was much more extensive and contained various internal administrative facilities and functionalities, to which various DBAG personnel had access but Mr Said did not. SHI’s access to GEM Web Reporting was withdrawn on 6th November 2008 but was restored in May 2012 for the purposes of this litigation. In its closing submissions SHI stated that, at least in terms of trade population, the material which emanated from DBAG’s systems from May 2012 onwards (the 2012 Reports) was not
reliable evidence of the information that was in fact available to Mr Said through GEM web reporting during the period of his trading.
In May 2008 Mr Said listed the reports that he considered of main importance to him, when asked which reports he currently used on the GEM Web Reporting system. Included amongst those were the Open P&L and the Collateral Summary Screens.
There was another screen which showed open trades on a per trade basis, namely the Trade Detail (Outstanding Trades) Report, which appears to have taken a slightly different format from the Open P&L Report, at least according to the Presentation sent to Mr Said on 4th December 2006. However each included a cash section dealing with swaps and forwards and an options section with columns setting out the currency pairs involved, the trade date and the current MTM value (the 2012 Reports are much the same as one another rather than taking different formats from each other). When Mr Said complained about the problems he experienced with MTM in 2007-2008, he referred to the Open P&L report, not the Trade Details (Outstanding Trades) report. Alongside the Collateral Summary report there was also a Margin Status report but it was the Collateral Summary report which Mr Said used and which Mr Giery said was the primary tool used by DBAG to monitor Mr Said’s collateral position. The 2012 Reports do not include historical sets of either the Collateral Summary or the Margin Status reports because they cannot be generated retrospectively. There are however reports which set out the history of certain figures which appear in those reports.
The aggregate of MTM open positions was referred to in the Collateral Summary report as the Available CMV Amount, which was calculated in ARCS VaR and fed from there to GEM rather than by GEM itself calculating the sum of the individual MTM figures it had received from ARCS VaR. During the period in which Mr Said traded OCTs and EDTs, there were significant differences between the Available CMV Amount as set out in the Collateral Summary report and the sum of MTM values on individual trades which were simultaneously reported by GEM.
SHI’s closing submissions in respect of reporting of trades include the following:
“343. Setting aside the valuation issues experienced by Mr Said throughout the duration of the FX PB relationship, described in more detail below, Web Reporting was designed to, and did, report the relevant trade details for swaps, forwards, cash trades, and vanilla and single barrier options (known in the experts’ lists in this litigation as Knockout Currency Options). As to more exotic options, however, the position was considerably more complicated.
There were certain trade types that could be booked in RMS but did not feed through to GEM, and thus did not appear in GEM reports or Web Reporting at all. As would become clear in October 2008, these certainly included the DBA Security trade types used to book EDTs by [DBAG’s] trading desk, but also, for example, correlation swaps, which do not appear in any of the 2012 Reports.
Other exotic trade types did feed through to GEM after a fashion. However:
GEM and Web Reporting could not properly report the trade details for those trades, not least in that their reports did not include fields for all of the relevant information.
Further, at least some such trades did not (when booked and open) appear in the relevant reports with any regularity. For example, as described in more detail in section D14, trades booked as “Resurrecting Fader Options” only appeared at random intervals in the Open P&L report used by Mr Said.
Yet further, at least some such trades, when they appeared at all, were ascribed an MTM value of “N/A”, or zero; and when, conversely, an MTM value was reported, it was for at least some such trade types not remotely accurate.
344. The evidence in respect of the problems to which these system failures gave rise in the course of Mr Said’s trading is discussed further below. However, it is important to note that, even now, the nature and even extent of some of those problems remains a considerable mystery. For example, while (halfway through the trial) [DBAG] produced a letter from
Deloitte that purported to address the zeroing out of the Available CMV Amount in historical reports, no explanation has been given of why (as further described below) certain trades would pop up in Mr Said’s Open P&L reports at apparently random intervals. Even more fundamentally, as noted above, there is very little reliable evidence of quite what data (or even simply which trades) were included in the reports available to Mr Said in any given report on any given date, and it does not seem that the 2012 Reports in particular can be relied on for that purpose. [Deloitte’s maintained that zero appeared wherever there was a timing issue with the feed from ARCS to GEM on MTM whereas n/a appeared where there was no feed at all from ARCS VaR to GEM.]
…
Daily monitoring of the collateral position on Mr Said’s FX PB account was the responsibility of the CMV team. The team, which was distributed across three time zones to ensure 24-hour monitoring, used GEM for that purpose, and was thus plainly reliant on the trades having been booked properly into RMS and fed properly into GEM. As Mr Gehlfuss explained:
The Net Equity figure calculated by GEM was automatically compared to three multiples of (what was intended to be) the VaR on Mr Said’s portfolio:
a “maintain level” of 200% of VaR, corresponding to the Independent Amount Ratio of 200% defined in ¶11(h)(i)(B) of the FX CSA;
a “call level” of 150% of VaR; and
a “close-out level” of 100% of VaR, corresponding to the Close-Out Ratio of 100% defined in ¶11(h)(i)(A) of the FX CSA.
The result of this automatic comparison was displayed using a “traffic-light system” where “green” indicated the client had provided sufficient collateral to cover their exposure (i.e. greater than the maintain level); “yellow” indicate the collateral was below the maintain level but above the call level; “amber” indicated below call level but above close out level; and “red” was below the close-out level.
Should the Net Equity drop below the call level, CMV would, in line with the policy of full collateralisation, issue a margin call to the people specified in the relevant “Notes” field for the client in GEM: in the case of SHI, Mr Said and Mr Vik, copying various Bank/DB Suisse employees. Other than consulting with DB Suisse CRM as to whether the TPMC[A] could be increased, the issuance of such a call was a purely mechanical exercise devoid of any qualitative judgment.”
As mentioned elsewhere in this judgment, and referred to in the above passage in SHI’s closing submissions, there was an ongoing problem with the MTM valuation on the GEM Web reports in relation to the MTM valuation of vanilla trades. The continuing problem was that ARCS VaR was not properly feeding MTM through to GEM so that Mr Said was complaining about the inaccuracy of the MTM figures in the Open P&L Report and about some vanilla trades not appearing until some time after they had been concluded. This problem surfaced in January, April, June and July 2007 and on 31st July 2007 Mr Said expressed his exasperation in an email to Mr Giery. He referred to erroneous P&L numbers appearing which were the same as the day before and the absence of the previous day’s deals on the system because they had not fed through. Mr Giery responded that the question had been escalated to senior IT management and that all agreed that this was unacceptable. He explained that both VaR and individual trade P&L was calculated in a separate risk engine and that the individual trade P&Ls had to be consistent with his overall VaR number (which was of course a portfolio figure). He told Mr Said that the problem was that the feed from that separate risk engine into the reporting engine had failed several times in the previous month. Though he did not name the engines concerned, it is clear that he was referring to ARCS VaR as the separate “risk engine” in which both
VaR and MTM was calculated and to GEM which was the “reporting engine”.
Although SHI suggested that this was not clear from the email, in my judgment it was. When the email referred to “[y]our VaR and the individual trade PnL is calculated in a separate risk engine” and talked of a “feed from that engine into the reporting engine”, it is plain that Mr Said must have understood that VaR and MTM (both elements in the margining calculation) were the subject of calculation by the same “risk engine”.
The problem was system-wide and was not unique to reporting for SHI’s portfolio. As SHI was margined on VaR, trade level valuations in respect of SHI’s trading through the FXPB Account were calculated by ARCS VaR and those calculations were carried out a number of times each day. The valuations were then sent to GEM in batches and were used to populate the Open P&L report and the Trade Details (Outstanding Trades) Report that could be viewed through the web reporting interface. New trade MTMs would only appear in the reports once they had been sent to GEM by ARCS VaR. The delay in the feed between ARCS VaR and GEM meant that the valuations displayed on web reporting were not updating as frequently as they should have and there was a considerable delay between the time when a trade was booked and the time when the trade details and MTM values for the trade appeared in the Web report. In addition, the rates used by ARCS VaR and GEM to recalculate MTM valuations for FXPB transactions throughout each day were updated at different intervals. This meant that for clients such as SHI who were margined using VaR, the open P&L report in GEM would have displayed a different revaluation rate (the GEM rate) to the one that had actually been used by ARCS VaR to calculate the MTM values which actually appeared in the report.
In order to resolve this issue in the interim, before a permanent fix could be achieved by IT, a separate P&L Reporting Account was created (referred to from time to time as a “dummy” reporting account) in GEM based on NOP but which would be accessible only to DBAG personnel and not to clients such as SHI. It would show MTMs calculated by GEM and not by ARCS VaR and would therefore provide a work-around solution to the feed problem from one to the other. As Mr Said would not be able to access this newly created report, manual spreadsheets would be prepared on a daily basis of the MTM (referred to by Mr Said as his P&L) of these vanilla trades at close of business, reflecting the trades shown on GEM and the NOP MTMs thus calculated. On 1st August 2007 Mr Said was told by Mr Quezada and Mr Giery of this separate account reflecting the MTM of open trades on NOP based calculations “calculated in real time”, with Mr Giery enclosing “a cut of your current Open PnL out of our internal risk engine”. The attached spreadsheet which was the first of the “manual” spreadsheets produced set out the trades as they appeared in the Trade Details (Outstanding Trades) Report available to Mr Said and then corrected the MTMs contained within it by reference to the NOP-based MTM system within RMS, using a pricing module to convert euro figures into USD. The spreadsheet itself refers to the RMS mark to market figure for each trade with two main headings – “CASH” and “OPTIONS”.
This dummy P&L Reporting Account was not however immune from problems despite avoiding the feed issues from ARCS VaR to GEM. The new account was set up as a parent of the ordinary reporting accounts, so that the trade population of the two should have been identical but, even on the first spreadsheet, Mr Said identified missing trades and not only the Correlation Swaps which were not recorded in GEM
at that point. So also it might be expected that one-off exotic transactions booked in RMS but not in GEM would not appear in the spreadsheets produced by DBAG. The evidence of Mr Giery and Mr Walsh was that although the MTM figures for cash transactions came from GEM/RMS to give the unrealised P&L, the MTM for the options still came from ARCS VaR up until mid-November 2007 but thereafter, according to Mr Walsh who took responsibility for provision of the spreadsheets, those sections were taken from the P&L Reporting Account and therefore from GEM/RMS. The manual spreadsheets did not however ever include trades which had been booked in RMS but not in GEM, regardless of the source of MTM figures.
As indicated elsewhere, the problem on these vanilla trade MTMs resurfaced from time to time with the provision of manual spreadsheets between 1st August 2007 and 26th November 2007, 5th to 7th February 2008, 19th to 22nd February 2008, 12th May to 7th July 2008, 21st July to 27th August 2008 and again on 19th September 2008. This problem was wholly unconnected with the issues relating to the reporting of OCTs which, with the exception of Knock-Out Currency Options, Digital Currency options and FW Setting Currency Options, either did not appear in GEM Web Reporting or the manual spreadsheets or appeared only sporadically.
I do not intend to recite the whole history of reporting of OCTs and EDTs from February 2007 till October 2008 (in the manner that SHI’s closing submissions do). It will suffice to refer to some examples and the key points. Of the 804 trades executed by Mr Said in his FXPB account, there were 53 OCTs and 41 EDTs, as subsequently classified by SHI and DBAG. DBAG, since February 2012, has accepted that its systems could not accurately value or margin the EDTs and, for the purposes of this action, has also since accepted that the same holds good for the OCTs, with the exception of 29 knock-out currency options. Some 8% of trades were therefore not accurately valued or margined.
The first OCT which presented a difficulty was OCT4, described as a knock-out timing option or an e-timer, concluded by Mr Said on 25th April 2007, about which Mr Said consulted Mr Vik beforehand (see the section of this judgment rehearsing Mr Said’s FX trading). Mr Walsh asked for Mr Geisker’s assistance in booking this, although he appears subsequently to have been able to book OCT12 which was an identical form of transaction, both being direct trades with DBAG.
On 12th June 2007, Mr Said wished to conduct a trade with CS, referred to by it as a Gated Range Accrual (OCT7). Mr Walsh referred the matter to Mr Giery, making reference to OCT4 as a similar trade which had been done directly with DBAG. Mr Giery’s response was to raise three points. The first was to check whether it could be supported; the second to check whether it could be included in VaR and the third was to ensure that Mr Said knew that it could not be reported properly, if it could be booked. The initial reply with regard to VaR was that, if the trade was not covered by ARCS VaR, it would be possible to book trade level margin in Sentry but a subsequent response in relation to OCT4 was that trade level margin could be provided for each. The figures would not therefore form part of the overall VaR calculation for the portfolio and as SHI’s FX trading was not set up to feed through to Sentry, any trade level margining would not have been included in SHI’s overall web based report of margin calculations at all. Mr Quezada gave instructions to take the trade in and to inform PWM, presumably with a view to CRM conducting trade margining, but in fact only Mr Meidal and Mr Brügelmann were contacted and the
matter never got to Mr Lay or Mr Halfmann for such assessment of any margin. It does not appear that any margin was therefore actually calculated for these trades at all. Mr Walsh could not recall whether he did or did not tell Mr Said that the trade could not be properly reported and he himself, in his evidence, said that he did not understand either the details of margining or the systems under which they took place.
Some three weeks later another Gated Accrual Trade was concluded by Mr Said (OCT11) which again was the subject of exchanges between Mr Vik and Mr Said, as referred to elsewhere in this judgment. Mr Walsh appears to have taken this trade in without further ado on the basis that OCT7 had been accepted. OCT12 was another extinguishing timer option in DBAG’s sales desk parlance and was a direct trade concluded with Mr Said on 7th August. Mr Geisker booked OCT12 directly to SHI’s account as he had done with OCT4 and Mr Avery confirmed that, as an extinguishing timer, it should be covered by VaR. In an email from Mr Quezada to Mr Said, he stated that the MTM had been confirmed with David Geisker. In DBAG’s parlance for the trial, OCTs 4, 7, 11 and 12 were all Knock-Out Timing Options.
In the daily run of reports generated by Deloitte in 2012, these four Knock-Out Timing Options appeared on only four apparently random dates in June and August 2007 in the Open P&L Report, though they do appear more often in the Trade Detail (Outstanding Trades) Report, but usually with a zero or n/a figure for MTM. If this reflects the position in 2007, SHI is right in saying that they “popped up” with “sporadic” appearances in the same way that EDTs subsequently did.
At the end of June 2007 Mr Said was seeking to conclude two Correlation Swaps which, it is accepted by the experts, are not currency options but which Mr Said expressly referred to Mr Vik on 25th June, asking if Mr Vik had any questions about the trades. The email explained the details and I have rejected Mr Vik’s evidence that he did not understand what he was being told. He did and must be taken to have approved the conclusion of the trade which was in two parts. Between them, Mr Walsh, Mr Giery, Mr Quezada and Mr Thaung exchanged emails and a hard coded trade level margin of 8% of notional appears to have resulted. The correlation swaps do not seem to have appeared in GEM at all if the 2012 Reports accurately reflect the position in 2007 in this respect. Nor did they appear in any of the disclosed manual P&L spreadsheets prepared by Mr Walsh. Any trade level margining would, for the reasons set out above, not have fed into GEM for SHI. SHI accepts that not only was it likely that Mr Walsh told Mr Said that the Gated Range Accrual concluded in June 2007 would not be reported properly (as Mr Giery instructed Mr Walsh to tell him) but that Mr Said got used to the fact in 2007 that various types of OCTs would not appear in GEM at all.
On 18th October 2007 Mr Said concluded transactions that are described as Fade-In Forwards but are probably better described as Pivot Accrual forwards. CS referred to them as Pivot Accruals and DBAG referred to them as Resurrecting Fader call options. These transactions constituted OCTs 16 and 18 (with CS) and OCTs 17 and 19 with DBAG. As described by Mr Malik (and it will be recalled that the experts could not agree upon a description) each of the transactions that are described as Fade-in Forwards are made up of two components, each of which in turn is a combination of put and call options. One component consists of the purchase of a call option and the sale of a put option at the same strike with a barrier (a pivot) at a higher level. The other component has the same barrier or pivot but consists of the purchase of a put option and the selling of a call option at a strike higher than the pivot. Together they constitute a Pivot Accrual Forward. There is no knock-out feature in this and the investor with this combination of options accrues a gain for each fixing date when the FX price is within the range of the high and low strike prices, either side of the barrier or pivot. The final payout is determined by reference to the accumulated gains on each fixing. This is therefore effectively a bet by the investor that the FX price will stay within the range of the strike prices. Professor Wystup, consistent with his view, saw these transactions as a combination of two forwards, rather than options, but that explanation does not account for the profits and losses which accrue by reference to the pivot or barrier and the high and low strike prices. Only Mr Malik’s explanation adequately fits the pay-off and the form of the transaction thus demonstrates the nature of the embedded options within it. It was to these transactions that Mr Said was later to make reference when first concluding the TPFs and telling Mr Walsh of DBAG about them.
On 17th October Mr Walsh asked Mr Avery whether DBAG would be able to take in OCT16 and OCT18, the Pivot Accruals which Mr Said wished to conclude with CS. He asked Mr Avery if the risk could be captured for this trade, setting out the essential details of it as a one month EUR/NOK Pivot Accrual with 23 fixing dates, the upper strike, the lower strike and the pivot level. He chased for a response the following day. Although there is no record of any direct response, it appears from an exchange between Mr Quezada and Mr Walsh on 20th and 22nd October that Mr Avery confirmed that the risk could be captured because it was a “Resurrecting Fader option” even though the subject matter of the email exchange appears as an American Reverse Knock-Out (OCT20). OCT20 could not be described as a Resurrecting Fader – it was simply a Knock-Out Currency Option and it appears that Mr Walsh was confusing the issue but was in fact referring to the Pivot Accruals/Fade-in Forwards. As will be seen, when a list of trade types which could be valued in ARCS VaR was forwarded by Mr Avery on 30th November, it included FX Resurrecting Fader options although, as was to appear in a further email exchange relating to TPFs, the system could not cope with Resurrecting Faders which had a large number of observation dates or fixings. As OCT16 and OCT18 had 23 such fixings, it is not clear whether that in itself constituted a problem.
So far as OCTs 17 and 19 were concerned, which were direct trades with DBAG, it seems that either Mr Geisker or Mr Chin booked them directly into RMS since Mr Walsh asked for an RMS number for them. They were booked as FX Resurrecting Fader options which should have fed through to ARCS VaR and GEM. The confirmation which Mr Geisker sent to Mr Said who forwarded it to Mr Walsh described the transactions as Resurrecting Faders. Mr Walsh referred, in a conversation with Mr Giery, to the booking of what were presumably the CS Resurrecting Faders as “a nightmare”.
If regard is had to the 2012 versions of the Open P&L reports, these Pivot Accruals/Fade-In Forwards appeared only very occasionally in the GEM Web Reports. OCTs 16/18 (the CS transactions) appear only on 12th November 2007 and OCTs 17/19 do not appear at all on that date. In the manual spreadsheets produced by Mr Walsh there is no reference to these transactions at all.
On 28th November 2007, Mr Said made a provisional agreement with CS to trade OCTs 23 and 24, a Forward Volatility Agreement in CS parlance, or a FW Setting Currency Option in DBAG parlance. In an instant messaging chat with Mr Chapin of CS, Mr Said said he needed to hear back from DBAG on booking, but that CS would have to let him out in the extremely unlikely event that DBAG would not handle the trade, since they had done Pivot Swaps, Gated Accruals and Correlation Swaps. He told Mr Chapin that if DBAG gave him “any stick”, he would take out a bigger stick, with Mr Chapin commenting that “you are good at that”. He was having difficulty making contact with Mr Quezada.
Mr Walsh sought assistance in an email to Mr Quezada, Mr Giery and Ms Greenberg, asking whether the trade could be taken in but got no response from the first two and a reply from Ms Greenberg that she was unable to take the decision, any more than Mr Walsh was.
About half an hour to an hour later, Mr Said and Mr Chapin continued their instant messaging chat with Mr Chapin asking whether DBAG was “still holding out”. Mr Said responded that the “guy is off desk” (meaning Mr Quezada) but he was trying again and then later said that DBAG was being a pain. He continued by saying that there was no need to worry as he would “browbeat them into [this]”, despite their apparent new appreciation that the exotics were not enumerated as such in the FXPBA, which, as he put it, would be a stupid way of doing it. He said he would get it sorted.
An email from Mr Said to DBAG stated that he was surprised that there was any issue about the trade because, when he had discussed the FXPBA at the outset, he had talked with Mr Quezada about one off (non-standard) trades and had been told they would not be a problem as long as DBAG was not swamped with them. He said that there had since been a few of them but never more than two or three outstanding at any time and none were ever an issue as Structured Options were provided for in the FXPBA. He then said that a prime brokerage agreement where he could only do spot and simple options was of no use to him and that was discussed at the outset. The particular trade was time critical and he had been waiting for a response but the issue was deeper because it went to the basic understanding of DBAG’s approach to Prime Brokerage and the agreement between them.
It appears that a little later he sent a further email to Mr Quezada, Ms Greenberg and Mr Walsh saying “I’ve had enough of this. Our agreement provides for structured options – this is a structured option. There is plenty of precedent. We have tons of collateral. Your unresponsiveness is making it very difficult. I am doing this trade before it runs away. If you decide NOT to book it, I will unwind it at my cost and we will then deal with the relationship impact. I would have preferred to discuss this but since no-one answers my phone calls that is difficult.”
SHI recognises that this was typical of Mr Said’s approach in such situations. This was, as he said, a Structured Option for which, as he knew, he needed DBAG’s consent but he was prepared to put pressure on to get it dealt with urgently so that the trade could be done and to threaten the destruction of the relationship, not just with him, but more importantly, with Mr Vik.
In his further instant messaging chat with Mr Chapin, Mr Said said that he needed DBAG to “play ball” and that he had begun to “draft ‘Klaus’ emails” which Mr Chapin regarded as “fun” and wished to receive a blind copy of the email which is presumably the one to which I have just referred. Mr Said later reported that “the browbeating worked” and sought a term sheet for “the dimwits” and “goons” at DBAG.
DBAG agreed to take in the trade as can be seen from the email from Mr Walsh to Mr Quezada and Mr Brügelmann saying that Mr Said was told that it would be taken in. In the email Mr Walsh said that credit approval had not yet been obtained because of the absence of full trade details but once those details were available they would be sent to PWM CRM to determine the necessary margin. Mr Walsh in his evidence said he did not know why it was necessary to do that but he thought that PWM could determine trade level margining. It must have been Mr Quezada who raised this but Mr Said understood the position perfectly well. In an email from him to Mr Walsh he enclosed the term sheet for the Forward Volatility Agreement stating that the notional size of the underlying straddle equated to a P&L exposure of €1 million per one percentage point move in the forward volatility curve and that “your margin guys may want to know that”.
It is thus plain that, in this particular instance, DBAG accepted the Forward Volatility Agreement as a Structured Option on the basis of applying trade level margining. That of course DBAG was entitled to do. DBAG was entitled to decline to take in any Structured Option under the FXPBA but if it decided to do so it could impose such terms as Mr Said was prepared to agree. If a trade was not capable of being margined in ARCS VaR, for whatever reason, it was open to DBAG to say that the only basis upon which the trade would be accepted would be if SHI would accept trade level margining. There would then of course be an issue as to how the trade level margin would be added into the VaR figure for the rest of the portfolio but, assuming that could be achieved, and however cumbersome it might be, it should not have been beyond the wit of man to achieve it. There was no reason why that should not have been agreed between them. What, if anything, happened to the trade level margining that was imposed here, is unclear, since it would ordinarily feature in Sentry which did not have any connection to GEM. As appears from Mr Said’s Timeline, he anticipated that Structured Options would have to be dealt with separately from the automated systems, in terms of MTM and margin and no doubt that was the basis upon which the original discussion took place about DBAG not being swamped with such one offs. It was for DBAG to work out how to aggregate VaR on the portfolio with trade level margin on individual trades which had been the subject of acceptance by DBAG on that basis.
There were then further exchanges between Ms Greenberg, Mr Lay and Mr Said, with the former obtaining further information for margining purposes but it remains unclear whether any trade margin figure was ever notified to Mr Said, what figure was calculated and how it was supposed to tie in and be implemented alongside the portfolio VaR.
Mr Said kept up the pressure with a further email to Mr Quezada asking him to call him in order to discuss, now that the trade had been booked, how this was going to work in future. He said he needed to know what DBAG could provide and whether the service level had changed. It may be that it was this which caused Mr Quezada to enquire of Mr Avery as to the list of products which ARCS VaR could handle. A list was sent to him on 30th November, so that, subject to identification of any particular trade alongside the names used by DBAG, Mr Quezada was then in a position to appreciate what could and what could not be the subject of DBAG’s VaR system. As already mentioned, different banks tended to use different names for some of the products.
In December 2007 Mr Said concluded OCT26, the first Double Knock-out Option also referred to as a Window Double Knock-Out. This did not appear in the list of trade types sent to Mr Quezada on 30th November as a trade which could be valued in ARCS VaR. Mr Walsh confirmed to GS, with whom the trade was done, that it would be taken in when GS questioned whether or not it was within the terms of the Counterparty Agreement. Mr Walsh assumed in evidence that he must have obtained approval from Mr Quezada or Mr Giery since this was a new type of option so far as he was concerned. There is a record of a telephone conversation between Mr Walsh and Mr Said on 2nd January in which Mr Said referred to a conversation which he had with Mr Walsh in which Mr Said had expressed surprise that DBAG’s system could not handle the trade and had been told by Mr Walsh that he would book it. Mr Said pointed out in the January conversation that he could not see it anywhere and usually, even if it was not valued, it was visible as being booked. Mr Walsh confirmed that it had been booked, saying in evidence that he must have checked in RMS, as the trade did not show in GEM Web Reporting. There is no record in any of the 2012 Reports, nor in the manual spreadsheets sent to Mr Said of this trade at any stage. In the telephone conversation Mr Walsh said that he could “manually get this one for you” and on being told that the Window Barrier should knock out on the 4th and that it became a simple option at that point, Mr Walsh said that he would “get you a mark on it” as it was just a vanilla transaction. It was not of course just a vanilla transaction and there is no suggestion that Mr Walsh ever did give Mr Said a mark on it, unless by that he meant the cash settlement figure on knock-out. The absence of entries on the GEM Web Reporting system was evident to Mr Said.
Towards the end of 2007, when asking DBAG for details of OCT4 and OCT12, Mr Said referred to them as being “offline”. When Mr Said was putting together his 2007 year end P&L, he sent emails to Mr Chapin of CS and Mr Geisker of the DBAG sales desk requesting details of the terms of OCT11 and OCT12 with regard to dates, premium paid and accruals received, referring to them as “offline deals not readily available [from] any system”. Mr Said obviously knew, at December 2007, that GEM web reporting did not include the details of the trades. Had he thought that GEM included them at all, he would have asked Mr Walsh for details. As he said in his Timeline in October 2008, he knew that “Structured Options” was basically a catchall of all options that the automated deal capture and MTM systems of DBAG could not handle and which were subject to case-by-case approval by DBAG.
As SHI accepted, in its closing submissions, when he referred to “an offline deal” he meant that the trade details were not available to him from any of the DBAG systems.
There are emails on 8th January 2008 recording that Mr Said called Mr Walsh, asking if DBAG could take in a Double Digital option and explaining that on the expiry date both of the two currency pairs had to be above agreed barriers to result in a fixed cash flow. Mr Walsh said he was unsure whether there was a product type for this in RMS and asked Mr Giery whether it could be taken in. Mr Giery told him to speak to middle office to see if it could be booked. Mr Walsh asked a member of the middle office team in London FXPB, who told him that the product type was a “Threefactor Discrete Double No Touch” option which had been booked in London in the past for
another client. On Mr Walsh forwarding this to Mr Giery, the latter asked if the trade was captured by VaR for that client and Mr Walsh then asked Mr Avery if VaR could capture the risk. He was told that VaR could not cover the product and there was no suitable proxy. Mr Giery then told Mr Walsh to ask how London had dealt with the matter for the London client and the answer came back that these types of trade were taken in if the client was buying (and therefore paying premium) or selling by closing out an open position. As SHI was buying the option and the maximum loss was the premium which would be paid two days after the trade date, Mr Giery authorised the acceptance of it. There were, in all, five Dual Currency Range Digital options concluded by Mr Said (OCTs 30, 37, 39, 50 and 51), none of which were properly recorded in RMS, GEM and Ethos, let alone in ARCS VaR. As was evident to Mr Said, they were not on the GEM Web Reporting nor in the manual P&L spreadsheets sent to him.
Against this background Mr Said entered into his first EDTs on 19th February 2008 with CS. It was Mr Chapin who introduced him to the concept of TPFs, TARFs or TARNs as they were referred to by various entities. He actively marketed the idea to Mr Said who then approached Mr Geisker and Mr Bergad of DBAG asking whether DBAG could price transactions of this kind. He saw them as similar to the Pivot Accruals of October 2007 but with a target knock-out feature which in practice made for wider ranges between the high and low strikes. This was the essential basis upon which Mr Said put the matter to Mr Walsh on 19th February seeking DBAG’s consent and saying that the cap on the maximum profit made the deals cheaper. Mr Said explained further, when Mr Walsh suggested that the only real risk was the premium which he was paying, that there was more risk than that because it was a Pivot Accrual and he was “short of some options here”. In other words he recognised that the TPFs involved the selling of options which meant that he could lose money on them. He said that it was a very broad range trade but if there was a massive market move that did not revert, he would lose money unless he hedged, which he would do. There was less risk than the Pivot Accruals because the knock-out assumed that the maturity was shorter. Mr Walsh asked for the trade details as he needed to clear this with “business” meaning Mr Quezada or Mr Giery.
The same day Mr Walsh sent Mr Quezada and Mr Giery indicative terms which had been produced by DBAG’s sales desk, rather than CS, saying that the details might differ slightly but that Mr Said wanted to know if DBAG could take in the trades. Mr Walsh told Mr Quezada that Mr Said had done one of these trades back in October as well (referring no doubt to the Pivot Accruals). The indicative details set out the pivot, the high and low strike prices, the daily fixing, the target profit, the capped knock-out and the one year duration with 256 fixings. Mr Quezada’s response was to say that, if it had been taken in before, Mr Walsh could go ahead but he should check with Graham Avery if VaR could capture the risk. Mr Walsh then asked Mr Avery that question, stating that it was an FX Resurrecting Fader option, as if the trade was identical to OCTs 17 and 19, without referring to the target profit knock-out feature. Mr Walsh had no recall of getting any reply from Mr Avery and there was no document evidencing a response. Mr Walsh thought that he must have proceeded on the basis of Mr Quezada’s go ahead, even though he had no answer back from Mr Avery since, if he had received a response saying that the trades could not be captured by VaR, he would have referred the matter to Mr Quezada and Mr Giery once more.
It seems unlikely that Mr Walsh did get any response from Mr Avery, because of the terms of later responses to questions asked.
Mr Walsh proceeded to book these trades, telling Ms Ng that he would book this trade (which was basically the same as one he did a while ago but had some “weird shit”) in the way he thought it should be booked and then sign off on it. When asked by her what kind of an option it was, he said that all that mattered was that the exercise/expiry was agreed and it did not matter how he booked it. Although in evidence Mr Walsh said this was an exaggeration, it is plain that he thought that it was the cash settlements which were crucial, rather than the details of the trade itself. In evidence he said he did not think he could book the knock-out elements of the TPFs on a Resurrecting Fader booking which was the form that he adopted for these TPFs and which was inappropriate for a number of reasons. It seems that, generally speaking, when Mr Walsh booked pivot TPFs he booked them as four Resurrecting Fader Options and when he booked non-pivot TPFs he booked them as two Resurrecting Fader Options. Being booked in this way meant that, insofar as they fed through to ARCS VaR on GEM, they did so incorrectly. Some did not feed through at all, because, it was said at the time, of the huge number of fixings.
Although in 2007 one OCT was apparently the subject of agreement to trade level margining by CRM or PWM Credit, there is no evidence of any real thought being given to acceptance of the EDTs on terms that trade level margining be applied by either of those two departments without reference to the Trade Desk. DBAG would have been entitled to accept the Structured Options on terms requiring trade level margining if it chose to do so, or to refuse otherwise to take them in at all. Presumably, Mr Quezada considered that CRM or PWM Credit would have no familiarity with these complex types of option and that the best source of margining was Mr Hutchings/Mr Geisker’s Trade Desk which was familiar with them because it sold and traded them. All efforts were therefore directed to achieving trade level margining by reference to the Trade Desk in the course of 2008 when it was clear that GEM and ARCS VaR could not cope. As appears elsewhere, from Mr Said’s Timeline, he always expected Structured Options to be dealt with outside DBAG’s MTM system, because he knew it could not handle them.
By 25th February, problems with the booking had surfaced and Mr Manrique, who was part of the FX Options Operations Team in New Jersey, asked for copies of the term sheets for the trades and, on receiving them from Mr Walsh, entitled “Pivot Target Accrual Forward”, spoke to Mr Chin and other traders at DBAG and ascertained that TPFs could only be booked using a DB Analytics security trade type which would enable the individual making the booking to upload the details of the trade into RMS from an Excel spreadsheet. Once uploaded, RMS could generate a text file and the trader could attach a free-form file providing details describing the trade and explaining the pay out formula (a Generic Sales Ticket). The trades, he understood, required monitoring because the cash flows were “path dependent” and accrued on the fixing dates. Without access to the tools used by the traders to monitor the ongoing life-cycle of TPFs, daily cash flows would have to be tracked manually to determine whether the trade had knocked out and whether payments needed to be made. Although DBAG’s FX traders and the Trade Management Group had access to DB Analytics Securities, in 2008 FXPB client services did not. Mr Manrique could see from Ethos that Mr Walsh had booked the transactions into RMS as Resurrecting
Fader options as opposed to the method which he had ascertained from Mr Chin to be correct. His immediate reaction was to tell Mr Walsh that he believed that he should speak to SHI and claim the cancellation of those TPFs and he should not accept any more TPFs if they could not be properly booked. He forwarded the term sheets to Ms Ogilvie, to whom he was accountable and she sent them on to Mr Kim, the manger of the FXPB Operations Team. In her email, she referred to a conversation which she had had with Mr Kim, enclosed the term sheets and stated that they represented TPFs which required a GST to be booked because Ethos could not fully support the product type. If the trades were booked incorrectly, the correct confirmations could not be sent out to the client or the Counterparty. Mr Kim said he did not follow up on this email and told the Court that his concern was purely from an operational perspective, as was that of Ms Ogilvie and Mr Manrique. Ms Ogilvie’s concern was that she could not create confirmations which was her responsibility, because the information in Ethos/RMS would not be adequate for it and the trade required a GST. Mr Kim said he was not aware until sometime later that the TPFs had been booked as Resurrecting Faders which meant that Mr Walsh had booked four records in respect of each trade, as had been done for the Pivot Accruals/Fade-in Forwards.
On 28th February in an email Mr Said told Mr Walsh that he had done a third Pivot Target Accrual the previous day which was the same as the first two but with a different currency pair. He said that these had to be kept out of the live MTM module because the system could not handle it and was giving silly numbers. He attached a copy of the trade terms for EDT 3. Two days later he entered into EDT 4 which he explained was really no more than an increase on the previous deal that needed to be treated separately for booking purposes. In an instant messaging chat with Ms Ng, Mr Walsh said that he was booking Pivot Accruals. He did not think that DBAG could take them in but he still booked them because he did not want to say no to Mr Said. He told her that Mr Manrique was telling him that DBAG was having trouble confirming with the Counterparty because RMS could not handle the options type. He told Ms Ng that he was telling Mr Manrique to sign off on it but the reason that they had trouble issuing confirmations was because all the details of the trade could not be fed into RMS because there was not an appropriate field for them. He was having trouble booking them and kept changing things but the notional was US$768 million, as calculated by him. Mr Walsh did not apparently seek Mr Quezada’s approval on the basis that he had approved EDTs 1 and 2.
It seems that, by 10th March, Mr Walsh had not booked EDT 4 in any way. He confessed in a Bloomberg chat with Ms Ng that he got scared every time he saw Klaus calling and on being pressed by her to book the trade, said that he might do it that afternoon. What he did in fact was to contact the FXPB middle office in London once again by email, attaching the terms of trade of EDT 1 and EDT 3 and asking for advice, explaining that he had been told that a GST might be needed for booking. The response was to say that he should not be taking them in and that the only way properly to book the trades would be through a spreadsheet to which FXPB should not have access and for which it did not have the infrastructure. A “fudge” booking was a possibility but the author of the email said that he did not have time to look at the term sheet and find the best fix for the moment. Mr Walsh replied by saying that the trades had already been taken in and that the same trade type had been taken in in the previous year. He asked if there could be an attempt to fudge a booking the next day.
On 11th March in a Bloomberg chat with Ms Wu, his immediate superior, Mr Walsh said that he had been told by Mr Manrique that the four TPFs had to be cancelled because they could not be taken in. She said that Mr Said would have a fit and Mr Walsh recognised that it would be a huge problem since he would not accept it but Ms Ogilvie had discussed the trades with Mr Kim and it looked as though they had to be cancelled. Nonetheless, Mr Walsh said that it was not Mr Kim’s call and he was not too worried and he would be talking to Mr Quezada the next day: “The day [I] go to Steve to ask if we can take in a trade … that will never happen”. Mr Manrique recognised that cancellation was a matter for the front desk and not for the Operations Department and over the course of the next month appears to have repeatedly asked Mr Walsh to speak to Mr Quezada, obtaining assurances from Mr Walsh that he would do so.
On 27th March, Mr Byrne (the equivalent of Mr Kim in London) asked Mr Manrique why EDTs 1 and 2 were in the system with the note “should be cancelled” and was told that these were TPFs which had been booked incorrectly, that the situation had been escalated to Mr Walsh with advice that they could not be taken in and that he had referred the matter to the front desk but no response had emerged. Mr Byrne’s response was strongly to advise against FXPB taking on any trade that needed to be booked as a DBA Security Type with which Mr Manrique agreed. Mr Byrne chased for an update on this on 31st March and Mr Manrique emailed Mr Walsh asking when they would be cancelled, stating that they should be removed by the end of the week.
Mr Walsh was still having difficulty in talking to Mr Quezada about the matter though he told Ms Ng on 31st March that he was about to do so and she offered words of comfort that Mr Quezada would speak to Mr Said with him. From a later chat it appears that Mr Walsh could not bring himself to talk to Mr Quezada about it and so he was thinking of quitting that day because of the SHI issues. Despite further chasing by Mr Manrique on 4th April, saying that the EDTs needed to be cleared away on Monday 7th, by 4.30 pm that day Mr Walsh had still not spoken to Mr Quezada and was expressing the desire to Ms Ng to speak to Mr Said to find out if SHI could hold the trades directly with the Counterparty or whether he would have to cancel them. He hoped to be able to “bullshit” Mr Said by telling him that CS would not agree to the confirmations. He emailed Mr Said asking him to call, which he duly did. In the telephone conversation Mr Walsh asked if the trades could be concluded directly with CS or whether they had to be put through the Prime Brokerage system and Mr Said told him that he had no other way of trading them because he had no documentation set up with CS and everything he did had to go through the FXPBA. Mr Walsh said that there were problems with confirmation details and if it was possible to move the trades to CS those problems could be avoided. Mr Said then said that, touch wood, the first TPF would probably be gone within ten days and as the trades were all exactly the same, the documentation should be clear. Once the knock-out event occurred, the deal was finished and off the books so that all that would be left would be the unwind. If there was any trouble with the documentation, Mr Said offered to assist.
Following that conversation, Mr Walsh again had a Bloomberg chat with Ms Ng in which he repeated Mr Said’s comment that one of the trades was about to knock-out (one which he had not booked) which meant that he could then simply book the cash flow, although he could not do so for a week or more. Ms Ng told him to call Mr Quezada before the situation got worse to which he responded that he had checked with Mr Quezada on the trade date in respect of all of the trades, or at least he thought he had. She told him to call him up, tell him that he approved the trades but they could not be booked and ask him what to do. He referred then to a Bloomberg chat which he had with Mr Manrique earlier in the day when the latter had said there were two options. Either Mr Walsh got someone from the front office to book it or the deal could not be taken in. In that conversation he had told Mr Manrique of his hope that Mr Said might do the trade directly with CS but Mr Said’s comments had now ruled out that solution.
In consequence Mr Walsh then called Mr Quezada who, he knew, would have to sign off on any solution. What appears to have been agreed with Mr Quezada was that the Trade Desk should be asked to book the EDTs whilst, in the interim, Mr Quezada himself would sign off on the confirmations produced by the counterparty, which would solve the confirmation issue for Mr Manrique and Ms Ogilvie.
When EDT 3 knocked-out on 16th April, Mr Walsh booked a cash flow but Mr Manrique then said that there had to be a booked transaction to which such a cash flow could attach. Mr Walsh then gave the RMS numbers for the bookings as Resurrecting Faders thinking that, because this was the end of the trade, all issues relating to the original booking were no longer of any importance. FXPB in London however did not agree. Both EDT 3 and EDT 4 had knocked out but London was questioning why the trades still existed in the knock-out queue since they were incapable of being booked, save by DB Analytics. Although Mr Manrique suggested that the problem had been resolved by knock-out, that there were notes within RMS booking which allowed for a proper audit trail and the confirmations were being sent for the business manager to sign off, London wanted to know whether FXPB was going to accept TPFs in the future and how it was going to book them when the risk could not be captured by bookings as Resurrecting Faders. Mr Walsh’s response was to say that the Counterparty confirmations would be signed off by business to clear outstanding unconfirmed trades and in the future each trade would be examined and signed off by business (Mr Quezada/Mr Giery) on a one off basis before acceptance.
In relation to the direct trades, EDT 05 and EDT 06, Mr Walsh asked Mr Geisker to book the trades directly to SHI’s account, giving a Counterparty code. Mr Chin, rather than Mr Geisker, carried out the booking with DB Analytics, so that it appeared in RMS but, for reasons given elsewhere, this did not feed downstream. On these direct trades DBAG, like other banks selling the product to SHI, provided spreadsheets showing the accruing profits and losses on the fixing dates.As the DBAG systems could not, for the most part, cope with OCTs, so also they could not cope with EDTs. The trade details of TPFs could not be captured by the use of the Resurrecting Fader proxy and Mr Said knew from the outset (see the email dated 28th February 2008) that the system could not handle the TPFs and gave silly numbers for their MTM. Mr Walsh’s evidence was that Mr Said wanted the trades removed or the MTM zeroed out so that his overall MTM figure was not distorted by the inclusion of such silly numbers. Mr Said specifically asked him to remove from the manual spreadsheets the MTM values of the EDTs booked as Resurrecting Faders in RMS and also wanted him to ensure that MTM values for those transactions did not appear in the report showing MTMs for individual trades on GEM Web Reporting itself. Mr Walsh had no control over the latter, in so far as the booking of a trade as a Resurrecting Fader, whether in four component legs or otherwise, would give rise to whatever figures the programmes produced. If a trade was not booked into GEM, then of course no figures would appear. Mr Walsh did not have a precise recollection as to whether Mr Said wanted him to remove all references to the transaction from the Web reports and manual spreadsheets or whether he simply wanted him to ensure that the MTM valuations would appear as zero. His best recollection was that Mr Said requested one or the other in their conversations and that he did both from time to time. Because of the absence of appropriate fields in which to record the details of the TPFs and Pivot TPFs, any proxy booking as a Resurrecting Fader was obviously inadequate and, to the extent that a trade was booked as a Resurrecting Fader, common sense suggests that it would be likely to generate some MTM or margin figure. The only way to avoid that potential issue of production of silly numbers was not to book the trade at all until knock out and cash settlement. That appeared to be what Mr Said wanted when he requested removal of the trade or zeroing out of the false MTMs.
602. There were ongoing issues with the ARCS VaR feed into GEM and on 18th April Mr Said complained again about the revaluation rate as well as complaining that faders had appeared again in the Open P&L reports. On 15th and 22nd April Mr Said concluded EDT 5 and EDT 6 with DBAG and a similar transaction (EDT 7) on 24th April with CS. The direct trades with DBAG were booked into RMS by the Trade
Desk using DB Analytics and GSTs but this did not feed through to ARCS VaR or GEM at all. In consequence the existence of these direct trades (six in all) never featured in GEM Web Reporting at all, although DBAG’s Trade Desk sent Mr Said accrual sheets on each fixing date for them. On 28th April Mr Said wanted to conclude a Pivot TPF with MS and this gave rise to a Bloomberg chat between Ms Greenberg and Ms Wu, the latter of whom opened the conversation by saying she had spoken to Options (by which she presumably meant Mr Manrique) who said that this was a TPF and that it should not be taken in because it could not be booked without access to GSTs and DB Analytics which FXPB did not have. Unless the Trade Desk was prepared to book it, it could not be taken in but she had been told that a trade of this kind had been taken in by DBAG in the past. She thought that Mr Walsh just took it in without asking anybody but Options said that it was booked incorrectly and was a mess. Ms Wu and Ms Greenberg agreed that if it was a TPF, it would not be taken in. Internal emails within MS showed that DBAG was saying that the trade was too exotic for the Operations people to handle and reliance was being placed by MS on the fact that Mr Said had told them that he had entered into such transactions with other banks in the past and with the DBAG Trade Desk. Further discussion within DBAG on 29th April suggests that Ms Greenberg was talking to Mr Quezada or Mr Giery about it and Mr Walsh told Ms Wu that they had been taken in as one-offs in the past though they could not be properly booked in RMS and Options Operations was displeased about that. Mr Walsh said that Mr Quezada had approved them in the past even though Options had told them to cancel the trades. Mr Walsh said that Mr Said had sent him the CS TPF (EDT 7) the previous Friday and asked if he should say that it could not be taken in either but the decision was taken to await the result of Ms Greenberg’s discussions with Business. On the same day Mr Walsh emailed Mr Avery asking if ARCS VaR could margin/value the CS TPF executed on 24th April (EDT 7). There is no record of an email response but in a further Bloomberg chat, Mr Walsh stated that day that Mr Avery had said that the options were being valued and margined on the existing bookings (which were as Resurrecting Faders). Mr Walsh, in that chat, said that the only details that could not be booked were the knock-out levels which he thought would have the effect of reducing the risk. Discussions continued between Ms Greenberg, Ms Wu and Mr Walsh with reference to the need for Mr Kim, Mr Quezada and Mr Giery to make a decision as to whether the MS options would be taken in. Ms Greenberg commented that Mr Quezada did not know how to say no to Mr Said because TPFs had been taken in in the past. Further conversation between Mr Walsh and Ms Ng revealed that the trade could not be properly booked anywhere by FXPB and that the Trade Desk would have to book it which it was not thought it would ever be prepared to do. MS were told by Mr Walsh that DBAG would be unable to take the trade because it was not covered in the Counterparty Agreement but the evidence shows that the real reason was the perceived inability properly to book the trade. Mr Walsh said he would have sent this email on instructions from someone else and it appears that this was Mr Kim’s stance while Mr Quezada took a different view. It was unclear how it was that the MS trade was declined but the CS trade was accepted and Mr Walsh had no recollection of the reasons for that, though it may have reflected either disagreement or a compromise between Mr Quezada and Mr Kim. At all events Mr Walsh was asked to go down to Wall Street to ask for the assistance of Mr Chin to book the CS trade and to be trained in doing it himself. That never happened because the Trade Desk refused to help in booking the trade for FXPB in circumstances where it had quoted for the business with Mr Said but he had gone elsewhere with it to another bank. The Trade Desk did not want to assist DBAG’s competitors in doing business with SHI. The difficulties were then touched on during part of the conversation on 2nd May between (inter alia) Mr Quezada, Mr Giery, Mr Spokoyny and others to which reference is made in the previous section of this judgment. This all led to the 5th May telephone conversation between Mr Quezada, Mr Said and Mr Walsh, the details of which are set out in the section of this judgment dealing with Mr Said’s trading and his evidence relating thereto. In short, Mr Said told Mr Quezada and Mr Walsh that he was not concerned about getting MTMs on the TPFs and agreed to DBAG taking them in without being able to report that information. He told them there was nothing to do on the trades from an administrative point of view save on knock-out or maturity whilst margin issues were a matter for DBAG and, if it was ever concerned about that, SHI would over-collateralise. It was following that call that, on 6th May, Mr Walsh, at Mr Quezada’s instigation, sought confirmation from Mr Avery that EDT 1 (an indirect CS TPF) and EDT 6 (a direct trade with DBAG) were both being margined. The response was that neither was being valued in the system, the first because of the number of fixings and the second because it was booked on a generic sales ticket and was not valued by RMS or any of the downstream systems. Neither Mr Quezada nor Mr Walsh informed Mr Said of Mr Avery’s response but since Mr Said regarded margining as a matter for DBAG, and accepted that he was getting inaccurate MTM figures upon which he placed no reliance, there was no reason for them to do so. Mr Quezada however might have been expected to inform others who were considering the question of the adequacy of the VaR calculation at the time for Mr Said’s trades, in particular Mr Giery who sat next to him in the office and had been involved in those discussions (see the section of this judgment relating to The VaR Parameters).
610. On 15th May Mr Said emailed Mr Quezada setting out his understanding that the latter was more comfortable with direct trades with DBAG and asking if he could do such a trade. Mr Said was obviously conscious of DBAG’s booking, valuing and margining difficulties when making this request. Mr Quezada told Mr Said to go ahead, stating that it would give him “the fire power” that he needed to have the Trade
Desk open up the DB Analytics system to the PB team. Mr Quezada emailed Mr Hutchings of the Trade Desk referring to the Direct Trade and stating that Mr Said had said he needed to be able to conduct such trades with other banks through FXPBA and if FXPBA could not oblige, he would not trade with DBAG at all. Mr Said also put pressure on the desk, telling Mr Quezada that he had made it clear that unless the PB desk got a pricing tool, direct trades with DBAG would stop. He continued to apply pressure, according to emails of 21st May and 3rd June, and then concluded another TPF (EDT 11) with DBAG on 4th June. These efforts to gain access to DB Analytics came to nothing because, it appears, the Trade Desk maintained its stance that, if it were to assist the PB desk by providing the pricing tool and/or training a member of FXPB in its use, the effect would be to improve the position of its competitors vis-à-vis itself. Although Mr Quezada appears to have met with Mr Hutchings, although Mr Geisker made some suggestions for solving the “[K]laus problems” and although there was a suggestion of interim solutions pending a “longer term [RMS] solution”, nothing came of this. It was not until about 17th October 2008 after most or all of the margin calls had been made that indirect EDTs were booked as such in RMS using the DB Analytics system.On 21st May Mr Said asked if he could conclude a TPF with GS and subsequently sent an email to Mr Walsh saying that he had received Mr Quezada’s approval. GS was informed by Mr Walsh that such trades were accepted as Structured Options on a one-off basis. This 21st May trade with GS was only booked by Mr Walsh on knock-out on 23rd June.
From hereon this appears to have become the pattern of action or inaction by Mr Walsh. Trade confirmations would be received from Mr Said and the Counterparty and he would print off copies which would then sit on his desk but he would not book the trades in DBAG’s systems, unless there was some particular trigger to do so such as a Counterparty requiring confirmation of a trade. On knock-out, Mr Walsh would book the trade in his standard manner for Resurrecting Faders with a cash settlement. This kept the silly numbers out of the MTMs that Mr Said was receiving, which was what Mr Said wanted.On 28th May Mr Walsh provided Mr Quezada with a list of the exotic trade types that Mr Said had executed, including Correlation Swaps, Resurrecting Faders, Gated Range Accruals, Pivot Target Accruals, Dual Binary Options and Double No Touch Options. Mr Quezada asked him to check with Graham Avery to see if ARCS VaR captured all those trade types in order to frame a request for the use of DB Analytics in relation to those which could not be captured. Mr Walsh replied by saying that there were two trade types that could not be completely valued in accordance with the way they were booked because not all the details could be entered, whilst the others could be valued and margined by ARCS VaR. The two in question were the Gated Range Accrual and the Pivot Target Accrual. Whilst Mr Quezada and Mr Walsh hoped to resolve the booking and margining issue by use of DB Analytics, London FXPB returned to the issue on finding that EDT 02 had not been processed. It appears that CS had not produced the Counterparty confirmation. This led to FXPB in London taking the matter up with Mr Kim by asking him if he was aware of these TPF trades and asking how FXPB could agree to take them in. Mr Byrne in London commented: “This trade type is one of the most structured types of business we are currently supporting on the franchise side, and there is no way PB can accurately book or monitor this trade type. Apparently the business is signing off on these when we receive CS confirmation, although this does not alter the fact that the trade is not accurately captured in RMS.”At about the same time Mr Cook of IT responded to Mr Walsh’s email of 21st May in which he asked how to obtain access to DB Analytics in circumstances where FXPB had previously been fudging the booking of some complex options. Mr Walsh asked what had to be done in order to capture the risks and margin the trades. On 5th June the reply was received telling him that the trade types he was talking about were complex risk trades which were generally booked and managed by the Complex Risk Group and were booked externally through RMS in Excel and uploaded as free text entries into RMS. Questions of who would book the deals, who would manage the risk, who would provide valuations (which were generally done manually) and how margin would be done outside of RMS all arose. There was therefore a business process to be gone through as between PB and FX business managers. Mr Quezada’s response was to state in an email to Mr Cook, Mr Walsh, Mr Hutchings and Mr Geisker that FXPB wanted to have the ability to book the trades in the same way as the Trade Desk and that initially use could be made of the latter’s expertise with risk management effected by the Complex Risk Team on that desk. He said that valuations were not critical to the client and manual valuations would be sufficient. He understood that ARCS VaR could capture many of the structures but, if not, margin would have to be captured manually. He wanted to know who could arrange FXPB’s set up to have access to DB Analytics. Mr Kim, following receipt of Mr Byrne’s email, asked who had agreed to take the trades in, whereupon Mr Walsh told him that Mr Quezada had signed off on these trades after discussing them with SHI. He said that he and Mr Quezada were working on a way to book the trades in the future and Mr Quezada was aware of the risk/valuation problems but there was no choice but to leave the trades booked as they were because they did not yet have access to the necessary RMS functions for booking the trade types. Mr Kim’s response was to say that he needed to approve new trades, not Mr Quezada, and that he needed to explain to FX management why the trade had been taken in and to ask for the trade details to be sent to him. Mr Walsh then said he would send over the details but the trades were those which had been discussed in April and booked in February. Whilst saying that he would speak to Mr Quezada to explain that Mr Kim did not want to take the trades in, he asked him to speak directly to Mr Quezada as well. Mr Walsh then sent details of EDT 1 and EDT 2 alone to Mr Kim and none of the other eight that had by this stage been executed and three of which remained unbooked. FX Operations in London had picked up on EDT 09 with GS because of contact with them by the latter and there then followed email exchanges between London and Mr Kim about the continued acceptance of such trades and the need to “push back” and not accept them in the future, whilst trying to decide what was to be done with those which had already been accepted. It was agreed that all future trades should be handled in New Jersey as opposed to London in order to avoid confusion. 619. Mr Kim’s evidence was that, as at 4th June, EDT 09 with GS, concluded on 21st to 23rd May, had not yet been booked and he was told by Mr Beels, his equivalent in London, that the only way to book it was with DB Analytics to which FXPB had no access. He had ascertained from Ms Ogilvie or Mr Manrique that drafting confirmations for these TPFs would take 3-4 hours each. He was concerned about booking and about matching and settlement. He intended to speak to Mr Quezada but at the end of the day the issue of approving the trades was a matter for the business side, namely Mr Quezada. Operations would have to manage the result of any decision taken. His evidence was that he did not want to take these trades in at all and contemporaneous records of conversations of others suggested that he was furious at Mr Walsh and Mr Quezada’s decision, which effectively by-passed him. As recorded in an email, Mr Kim was told by Mr Walsh that each one of these trades was approved by Mr Quezada before execution. Mr Kim’s evidence was that after 4th
June he spoke with Mr Quezada in perhaps three different conversations in which he was told that SHI wanted to do the transactions, that FXPB had to take them in because the client wanted it, the Sales Desk did them and SHI was saying that if DBAG could trade them directly, SHI should be able to do so through FXPB. Mr Quezada said he was seeking to obtain access to DB Analytics and GST. He also assured him that SHI was aware that DBAG could not book these trades properly and that there were valuation and margining issues which went with that. Mr Kim’s response was to say that the numbers had to be limited whilst waiting for the DB Analytics solution, because the operations team could not handle many of these transactions which took a great deal of time on a manual basis. There was a discussion about charging US$1,200-1,500 per transaction for this reason, a figure which was agreed with SHI and subsequently charged. The transcript of a telephone call between Mr Kim and Mr Quezada on 4th June is consistent with Mr Kim’s evidence with reference to the pressure being exerted by Mr Said on the Trade Desk in telling that desk that if they wanted to sell him TPFs, they had to allow him to “trade away”, through FXPB, with other banks. He said that Mr Said had agreed to do whatever he could to minimise any of the difficulties and play within any constraints DBAG wanted to put upon him apart from telling him that he could not do the trades at all. If DBAG could not figure out a way to do the trades through FXPB, he would shut down the franchise and move all the business. The effect of what Mr Quezada was saying was that DBAG had little choice but to accept the business. Margin and risk was not much of an issue because SHI was over-collateralised and was willing to put up whatever additional margin was needed. “Some of [those] pivot things” did not get marked on VaR. Mr Quezada said he would produce something on paper and asked Mr Kim to tell him in response if he did not feel comfortable with it and tell him what else could be done to alleviate his concerns. Mr Kim’s concern, as expressed, was the pressure he was receiving from London. On 6th June Mr Quezada sent an email to Mr Hutchings and Mr Geisker of the Trade Desk and to Mr Kim and Mr Walsh. The email commenced with Mr Quezada saying that he needed “to put a full court press on this” (a basketball reference to an aggressive defensive tactic in which the members of the team cover their opponents over the full court).
The email continued:“Note that Sebastian Holdings is on VaR and all their positions are valued and risk managed for margin purposes out of [ARCS] [V]aR.”
He then incorporated the 30th November list of the trades that ARCS VaR could cover and then set out three points. In the first he asked how the Trade Desk currently margined the trades done by SHI with it, or for similar hedge fund clients. He spoke of the need to sit down with Credit to determine consistent/proper haircuts. The second point was to ask how valuations were being handled for SHI whilst the third point said that “[f]or starters” access was needed to “the book” from the PB side, by which was meant DB Analytics, as used by the Trade Desk.
On 12th June Mr Walsh emailed Mr Said and told him that, as he knew, DBAG was working on finding a way to book the fader options and wished to work with the Trade Desk and Dave Geisker to achieve that. He asked if Mr Said would mind the Trade Desk being told the financial details of the trades done through FXPB. Mr Said said he had no objection. Mr Quezada referred to this as a good first step in an email to Mr Walsh saying that the capture of risk and margin was the next step.
The provision of this consent made no difference and at no stage thereafter, despite whatever efforts were made, did the Trade Desk render assistance in booking the TPFs. There does not appear to have been any great sense of urgency about progressing this matter and Mr Kim’s evidence was that he anticipated it might take six months to gain access to DB Analytics which is why he wanted a limit on the volume of trades accepted. He told the Court that as long as it was no more than five trades in a week it would be acceptable. He simply wanted to limit the number of TPFs taken on so as not to put a strain on the Options confirmation team in doing the business manually, taking term sheets, matching them, issuing confirmations and settling the trades. As far as he was concerned, it was a matter for Mr Quezada to deal with the downstream points of valuation and margining, about which he had given him assurances. Mr Walsh appeared resigned to the position and booked EDT 09 on knockout on 25th June 2008 (it had been traded on 21st May 2008) a practice that he largely then adopted for other EDTs.
In an email of 21st July sent by Mr Said to Mr Walsh on his return from holiday, Mr Said told Mr Walsh that one Fader with spurious P&L kept appearing. Mr Walsh said that he had just booked more Pivot Accruals which would probably appear on the web P&L, so he was sending his usual manual spreadsheet, with them zeroed out. Some EDTs he did book – others he did not until knock out.
On 22nd July Mr Walsh and Mr Said held a recorded telephone conversation about the fact that spurious MTM numbers for the trades were appearing in the Web Reports which were ‘never even remotely right’. Mr Said was told that DBAG’s margining for TPFs was based on those numbers, which Mr Said recognised as “chaotic”. He said that more collateral could always be provided. Details of this conversation appear elsewhere in this judgment. Mr Said continued to trade EDTs seeking DBAG’s consent to do so but knowing that no solution had been found to the booking issue and that many bookings were not being made on GEM (nor appeared on the manual spreadsheets) until knock out, as compared with the Accrual Spreadsheets that he received on each fixture date. The first loss-making EDT was concluded on 22nd July (EDT 20). From 23rd July onwards Mr Said entered into twenty-one EDTs of which four knocked out (with aggregate profit of US$14 million). The total net losses on these twenty-one EDTs were of the order of US$560 million.
Mr Kim’s evidence was that he was not aware thereafter that Mr Walsh was not booking trades but thought they were being booked under a proxy. Mr Walsh notified him of three TPFs which were concluded in June and he raised no objection. He said that he did not give his approval but it was not needed as a result of his discussions with Mr Quezada. On 16th September however Mr Manrique asked Mr Kim to sign off on the use of Counterparty term sheets for generating Confirmations as a form of GST in respect of ten TPFs which were concluded between April and August, five of which had already knocked-out. At that stage, if not before, it must have been obvious to Mr Kim that these TPFs had not been booked in the system contemporaneously. The exact numbers of unbooked trades would of course not be known to anyone except Mr Walsh who kept a pile of confirmations on his desk until October. Once they were booked in RMS, they would be open to sight by anyone with access to the system and, as part of his role, would have been picked up by Mr Manrique if not his superiors in Operations.
Mr Manrique’s evidence was that he understood there to be a conflict between Business on the one hand which wanted to take the TPFs in (Mr Quezada) and Operations on the other (Mr Kim) which did not. He did not recall how that had been resolved but from June to October he saw a number of TPFs when they knocked-out and were booked the same day. He assumed that there had been resolution as the deals kept coming through and, because there were not large numbers involved, he did not think much of it. It was odd to have settlements appearing on the days when trades were booked with an earlier date specified as the date of the deal. It was obvious then to him that the deals had not been booked earlier. Until a trade was booked he knew nothing of its existence, whereas if it was booked inaccurately he would see it and might take action. If a Counterparty sent a confirmation, he would raise the matter with Mr Walsh and a trade would then be booked but otherwise, he could see that trades were being booked on cash settlement but had no knowledge of how many were in the pipeline until October.
Between 20th June and 6th October thirty EDTs were concluded by Mr Said, of which two were direct trades with DBAG. Eleven EDTs were booked between 20th June and 10th October on the date of knock-out. Within the same dates, Mr Manrique was informed of nine new EDTs by Counterparties. On 8th September Mr Walsh forwarded the confirmation of EDT 33 to Mr Quezada, saying it was “the one done today”. The same day, before meeting Mr Said, Mr Quezada received an accrual sheet from Mr Geisker telling him of EDT 31 concluded with DBAG which was then outstanding.
On October 7th/8th/9th, Mr Walsh booked nine EDTs for the first time. There were only two individuals who were likely to have been aware of the number of unbooked trades at that date, namely Mr Walsh and Mr Said. It would have been obvious to Mr Said from the GEM Web Report and from the manual spreadsheets that there were unbooked EDTs since he had his own trading records and would immediately have seen the absence of the EDTs in DBAG’s reports. This was of no concern to him because he was receiving accrual spreadsheets from each of the Counterparties on the indirect EDTs and from DBAG on the direct EDTs. He did not look to the MTM or margin calculations as a risk management tool. He specifically agreed to the absence of reporting of the former and implicitly to the latter, which he regarded in any event as a matter for DBAG, not himself.
At DBAG however, Mr Manrique was aware of the general practice which Mr Walsh was adopting when trades were booked on settlement with an earlier trade date. Mr Kim was certainly aware of the position with regard to a number of those trades in September and Mr Quezada, if he ever looked at any of the reports, would have realised that any new EDTs approved by him or known to him were not appearing when they should. It was Mr Said’s evidence on deposition that he obtained prior approval for every Structured Option from Mr Walsh or Mr Quezada before concluding them. When Mr Walsh took time off in July and his colleague Elizabeth Ngo took over his duties, he told her that Mr Said might trade EDTs and that, as there was no way currently to book them, she should leave them on his desk for him. She printed them out and kept them in a pile for his return.
Mr Walsh’s evidence was that both Mr Kim and Mr Quezada knew that he was not booking the trades as he was receiving them. They certainly knew that he could not book them properly and that the Resurrecting Fader was an inadequate proxy. He
said that Mr Quezada and Mr Kim both knew that trades were continuing to be executed but that there was no way to book the trades. Certainly by September or October he was confident that they knew this though he could not recall any specific conversation in which he told them that this was what was happening. He said however that he did not seek approval from Mr Quezada or Mr Kim before taking in any trades after late June but Mr Said copied Mr Quezada in on emails sent to Mr Walsh in September about EDTs that he was proposing to do or had just done.
Mr Walsh said that he did not know if Mr Quezada was correct in his deposition when he said that he was unaware of inaccurate bookings or failures to book as at September, that he assumed that all trades had been booked, that he was unaware that some trades were not valued or margined and assumed that all the numbers discussed by Mr Spokoyny with Mr Said at a meeting on 8th September at which he was present embraced all trades.
In my judgment it is clear that Mr Quezada knew that every EDT was either inaccurately booked or not booked at all and was aware that every one of them was neither valued nor margined correctly, if at all. My conclusions as to the 8th September meeting are to be found elsewhere in this judgment.
The evidence shows that the GEM system was capable of dealing only with swaps, forwards, cash trades, vanilla options and single barrier options, as Mr Giery’s statement said. The definition of “Structured Option” in the FXPBA tallies with this, inasmuch as it is expressed negatively by reference to options which are not put or call options without special features or single barrier options. Mr Said appreciated this as his October 2008 Timeline reveals. He refers there to “several non-standard Structured Options (defined in the PB agreement) as basically a catch-all of options that the automated deal capture and MTM system of DB could not handle.” So it was that, towards the end of 2007, when asking DBAG for details of OCT4 and OCT12, he referred to them as being “offline”. As SHI submitted, in its closing submissions, when he referred to “an offline deal” he meant that the trade details were not readily available to him from any of the DBAG systems. SHI accepts that not only was it likely that Mr Walsh told Mr Said that the Gated Range Accrual concluded in June 2007 would not be reported properly (as Mr Giery instructed Mr Walsh to tell him) but that Mr Said got used to the fact in 2007 that various types of OCTs would not appear in GEM at all.
SHI points to the Fade-In Forwards (or Resurrecting Fader call options) which constitute OCTs 16-19 and their appearance in GEM Web Reporting. The transactions with CS which constitute OCTs 16 and 18 appeared for one day in the Open P&L report on GEM but only for that day. OCTs 17 and 19, which were the direct transactions with DBAG, did not appear at all.
In SHI’s closing submissions a table appears with the following explanation.
“418. The following table summarises the degree to which each category of OCTs seems to have appeared in (1) the Open P&L report on Web Reporting and (2) the manual P&L spreadsheets sent to Mr Said, and whether MTM valuations were included. Where “partial” appears in the final column, that indicates that MTM values did generally appear in spreadsheets whose options section was based on the MTS Sebastian account, but not those based on the P&L Reporting account.
419. Of course, DBAG admits that those MTMs that were reported for OCTs other than Knock-Out Currency Options were not accurate.
Trade type | Open P&L reports | Manual spreadsheets | ||
appears | MTM | appears | MTM | |
Knock-Out Currency Opt | regular | yes | regular | yes |
Knock-Out Timing Opt | sporadic | yes | regular | yes |
Digital Currency Opt | regular | yes | regular | partial |
Correlation Swap | never | — | never | — |
Fade-In Forward | sporadic | yes | never | — |
Fw Setting Currency Opt | regular | yes | regular | yes |
Double Knockout Opt | never | — | never | — |
Dual Currency Range Digital Opt | never | — | never | — |
…”
Mr Said could not have failed to notice the absence of reference in the GEM web reports to some of his Structured Options, whether absent in their entirety or only sporadically included. He knew from the outset that, whether they appeared or not, DBAG’s MTM system could not handle them.
As appears earlier in this section of the judgment, as soon as Mr Said commenced trading EDTs, he asked that the figures for them be kept out of the live MTM module because they gave silly numbers and he was concerned that they might have the effect of distorting the figures for his other trades. Again he would have known that the EDTs, as with the OCTs (other than the Knock-Out Currency Options), were too complex for the GEM system to handle or for reporting of their MTM.
As appears from conversations between himself and Mr Walsh, the TPFs which were booked as Resurrecting Faders, as the nearest proxy, would appear occasionally in the Open P&L report within GEM Web Reporting. The 2012 Reports suggest that between 20th February and the margin calls, only on eleven days was there any reference to any TPF. On the Trade Details (Outstanding Trades) Reports on the GEM Web Reporting system however, they appeared somewhat more extensively. The direct TPFs concluded with DBAG did not appear at all, being booked, as it would appear, in RMS but with no downstream feed.
Mr Said, according to his email to Mr Walsh of 5th March, listed the Open P&L account as one of “main importance” to him, not including in his list the Trade Details (Outstanding Trades) Report on the web reporting system. He therefore saw that they were generally absent but complained to Mr Walsh when they “popped up”, with obviously wrong MTM, requesting Mr Walsh to remove them, which of course Mr Walsh could not do. All he could do was to abstain from booking TPFs until knock-
out and, where they were booked, make a manual alteration to the spreadsheets which he compiled and sent to Mr Said by email in the circumstances outlined above.
According to Deloitte, the source of the data for the manual spreadsheets compiled by Mr Walsh varied (as the evidence of Mr Giery and Mr Walsh suggested). Between August 2007 and October 2007 the data which represented the starting point for the spreadsheets came from the same account whose details were available to Mr Said through Web Reporting, save for one date when it appears to have been based on the P&L Reporting Account which was the internal account created with NOP MTMs. Between November 2007 and February 2008 four of the eight spreadsheet reports were based on the account viewable through Web Reporting system and four on the P&L Reporting Account for the corresponding dates. Between May 2008 and August 2008 the spreadsheets were based on the P&L Reporting Account for the relevant date. On eight particular days between August 2007 and August 2008 however, there were multiple versions of the same spreadsheet, some of which were based on the Web Reporting system and some on the P&L Reporting Account.
Again, according to Deloitte, the trades included in the P&L Reporting Account were a subset of those included in the Web Reporting accounts. Five OCTs (two Digital options, two Fade-in Forwards (Resurrecting Faders) and one Knock-Out Timing option) were in the Web Reporting system but not in the P&L Reporting Account. So far as EDTs are concerned, whilst it should be borne in mind that nine TPFs were booked between October 7th and 9th, there is only one TPF booked as a Resurrecting Fader option which appears in the Web Reporting system but does not appear in the P&L Reporting Account. Each TPF was booked with a set of two or four trade entries constituting the single transaction and given a single structure ID. A new version of each trade entry was generally created every working day to reflect the daily fixing schedule.
Although there are different date ranges for twenty-five TPFs, as between the Web Reporting system and the P&L Reporting Account, sixty-eight of those reflect entries in the P&L Reporting Account in respect of the period after 13th October, when the first margin call was made.
There are numerous instances of the P&L Reporting Account containing multiple versions of the same trade entry. 129,231 out of 207,135 entries are duplicative and are likely therefore to reflect the creation of entries for each fixing day. The manual spreadsheets compiled by Mr Walsh did not contain such multiple entries. It would appear that they were edited out, as one might expect.
The spreadsheets were produced at various different times during the course of a working day, whereas the figures in the GEM Web Reporting system and the figures in the P&L Reporting Account contained information as at the close of business each day. Some differences are therefore to be expected between the spreadsheet figures and those appearing on DBAG’s systems. The discrepancies between the GEM Web Reporting system, whether the Open P&L Account or the Trade Details (Outstanding Trades) Account and the P&L Reporting Account were unexplained but Deloitte did not consider that the fact that the P&L Reporting Account operated by reference to NOP methodology explained the difference.
The end of day MTM figures shown in the Trade Detail (Outstanding Trades) Report were fed into GEM from ARCS VaR. When this report is now run for historical dates, it displays either zero or N/A for MTM figures in respect of a number of positions held in the FXPB account. Deloitte’s investigations with DBAG have led to the conclusion that where there was no valuation feed to GEM, “N/A” was displayed: where there was a timing issue with the valuation feed to GEM (because of a delay causing the valuation to be received in respect of a particular trade version) the valuation showed as “zero”. Both these types of entry were therefore the product of the feed of valuations from ARCS VaR to GEM, not of any changes made to the underlying structured data held in GEM. The History of Collateral Summary report available in GEM Web Reporting shows the daily end of day Available CMV Amount but this shows as zero as a result of a system wide coding change which affected the historical reports for all clients margined on a VaR basis. The date of that change is unknown. In the Collateral Summary Report, however, the latest Current Credit Exposure appears, which is a calculation of the open MTM valuation of all positions in the account, which feeds from ARCS VaR and displays as the “Available CMV Amount”. Those end of day CCE calculations fed from ARCS VaR are retained in the GEM database and have been disclosed whereas the zero figures in the Historical Report are not recorded in the GEM database but are generated within the Report at the point when the Historical Report is run.
There are still unexplained anomalies in the figures produced by the 2012 Reports and the historical record of what was supposedly shown on the four 2007-2008 Web Reporting accounts which showed MTM and margin. At the end of the day I do not think that matters, in the light of the critical evidence. Taken overall, there is no doubt that the DBAG systems were not designed to cope with the Structured Options and could not do so. That Mr Said knew and understood, although he would not have understood the reasons for the faders “popping up” with spurious MTM numbers sporadically from time to time, nor why figures differed between accounts, if he noticed that they did. He did appreciate that the trade details of the EDTs and most of the OCTs could not be captured and that any MTM or margining was therefore inevitably askew. Mr Said knew that he was not getting accurate reports on the EDTs or their MTMs but his only concern was to ensure that they did not distort the MTM figures in respect of his other trades. It was the accrual spreadsheets, which he got from Counterparty banks on the indirect TPFs and from DBAG on the direct TPFs, on which he relied in assessing his position and risk in respect of the EDTs. As appears hereafter, he told Mr Walsh that he wanted the EDTs removed or the MTM zeroed out of the figures supplied to him by DBAG whether by way of web reporting or manual spreadsheet, which could only achieved by not booking them at all and removing them manually from the spreadsheets if they were booked. He did not look to MTM figures for risk management of his EDTs because he regarded them as accrual trades. Margin was a matter for DBAG to sort out for its own protection and he did not regard the GEM system as a risk management tool.
14. Mr Said’s Evidence in Affidavits, on Deposition and in his Timeline
As I have already indicated, neither party called Mr Said to give evidence. At the time of the relevant events he was, of course, SHI’s agent, employed by one of Mr Vik’s companies in order to get the advantage of a medical health package but engaged by SHI without any written contract to trade FX on its behalf. Under the
terms of his engagement he was to be remunerated by the receipt of 10% of the net annual profits of SHI resulting from his trading and during the first year he received an annual salary of some US$360,000, which was to be set off against his entitlement to net profits, if earned. He remained on Mr Vik’s company’s payroll until about June 2009, having told Mr Vik that he would work for him for free in an attempt to recoup the losses suffered as a result of his trading in 2008. Mr Said parted company with Mr Vik after swearing three affidavits in the New York action, the last of which was sworn on 21st May 2009 and is relied on by SHI under the Civil Evidence Act.
Each party relies upon different elements of Mr Said’s deposition in New York which took place on nine different dates. The first three sessions took place on January 30th, January 31st and February 1st 2012. The fourth and fifth occasions took place on 7th and 10th September 2012 and the sixth on March 11th 2013. On all these occasions Mr Said was questioned by attorneys for SHI. On the seventh, eighth and ninth occasions in March and April 2013 he was questioned by attorneys for DBAG and on that last day he was re-examined by SHI’s lawyers. Mr Said was not wholly consistent in these depositions.
Whilst there is, of course, no property in a witness, if either party was to call Mr Said, the natural expectation would be that it would be SHI as he was SHI’s agent and SHI blames DBAG for the losses caused by Mr Said’s trading. SHI however does not treat Mr Said as a witness of truth in relation to the trades which Mr Vik authorised him to conduct. DBAG contends that Mr Said did not tell the truth about his relationships with DBAG personnel on the vexed subject of the EDTs and in particular about his state of knowledge as to DBAG’s failure to book, value and margin those trades.
I have read the whole of Mr Said’s depositions in order to evaluate the parts relied on by each party and to determine what can and cannot be relied on as accurate. I have found that the surest guide to the accuracy of that testimony is, as might well be expected, the content of the contemporary documents, including in particular the emails passing between Mr Said, Mr Vik and DBAG personnel, transcripts of telephone conversations recorded by DBAG and the immediate reactions of those involved when issues came to a head in October 2008. Additionally, there are Bloomberg chats which throw light on the contemporary events. When these matters are considered in the light of the motivation which can readily be ascribed to those involved, I have not found it difficult to determine where the truth lies in what Mr Said has said on the critical issues between the parties.
One factor stands out. When, in October 2008, DBAG made substantial margin calls of SHI, Mr Said did not blame DBAG in any respect for the losses sustained in his trading, nor suggest that DBAG had hidden anything from him in its accounting, valuing and margining, nor that the margin calls were inappropriate. It is clear that he had substantial discussions with Mr Vik in October and that the margin calls made between 13th and 17th October were all paid. On more than one occasion Mr Said said that there was no-one to blame for the losses incurred other than himself and the one in one hundred years market conditions he had encountered. In the middle of that week Mr Vik started to ask questions about DBAG’s prior margining, but there was not a hint of any suggestion from him that Mr Said’s trades were unauthorised, that there had been an agreed limit of US$35 million for SHI’s liability, or that Mr Said had been left in ignorance about his trading positions and would have acted entirely differently if DBAG had provided him with the MTM and margin information of which complaint is now made. By that time, in fact by sometime in the preceding week, if not earlier, he knew what Mr Said knew about DBAG’s failures to value margin and report thereon.
Moreover Mr Said drafted a “Timeline of Dealings with Prime Brokerage”, running to five pages, covering the position from September 2006 to October 2008, which he said he produced some time in October 2008 for the benefit of SHI’s lawyers and which therefore might be expected to put the best face on his and SHI’s position. The contemporaneous documents are reflected more closely in that Timeline than in his affidavits in the New York proceedings. When asked in his depositions by SHI’s lawyers how it was that, in his depositions he was saying that he had told Mr Vik about the different types of Structured Options that he conducted before he did them, whereas in paragraph 23 of his third affidavit he had maintained the opposite, he said that, in his affidavit, he had “[m]ade a mistake … [d]idn’t read it carefully enough. … It’s just not correct. I don’t have a good … I think the military answer would be no excuse.” At the time of swearing those affidavits in April and May 2009, Mr Said was working with SHI’s lawyers and he explained in his deposition that he did not personally draft the affidavits and did not remember reviewing any emails or other documents in conjunction with them at all. What Mr Said was there saying was that he was effectively acting under orders when swearing the affidavits (“the military answer”) and that the affidavits were drafted for him, doubtless by SHI’s lawyers.
It is suggested by SHI that Mr Said’s third affidavit is accurate and that, where contradicted by him in his depositions, Mr Said is not to be believed. It is suggested that, following his affidavits, DBAG’s lawyer made contact with Mr Said, suggested that SHI was depicting him as a “rogue trader” and asked him if he was prepared to meet him. Mr Said refused to do that and subsequently instructed his own lawyer for the purpose of the depositions. I cannot see how this provides any sort of an explanation for the change in stance adopted by Mr Said in his depositions, as compared to the affidavit, nor any explanation for the differences between his affidavits and the earlier Timeline, which was originally claimed to be a privileged document but which was the subject of an order for disclosure by the New York court.
Mr Said still operates in the FX market. At the time of his affidavits, he still had good reason to support Mr Vik/SHI and signed his name to affidavits drafted in support of SHI’s case in New York. He no doubt wished to help Mr Vik and SHI to recoup some of the losses suffered from his trading and, whilst it should not be the case, it is the fact that often people are more prepared to sign affidavits drafted for them with less regard for the truth than they are to make untrue statements in face-to-face questioning. When, in the context of a video-taped face-to-face deposition, witnesses are referred to, or have read, documents which tell a different story from that set out in a statement in writing, they are, naturally, more reluctant to uphold a position contradicted by those documents or not reflective of reality.
Mr Said, at the time of the depositions, still had reason to support SHI and Mr Vik in seeking recovery from DBAG and casting blame upon it for his trading losses and failures to read the market. He may however also now be more highly motivated to protect his own reputation in the FX market, having severed his connection with Mr Vik. It is true therefore that he might be considered more reluctant now to agree that he had acted in breach of any authority given to him but, at the time of his affidavits,
the point was not being put in that way in the New York proceedings. Now however SHI criticises him not simply for a wrong market judgment in the extreme conditions of October 2008 but for exceeding the trading authority given to him.
Mr Said’s natural motivation would be to seek to justify himself and to seek to cast blame on others if he could do so. The only possible target is DBAG. It is clear that he is not going to be sued by SHI, Mr Vik or by DBAG because he is not good for the sums in issue in this action. He therefore had no particular reason at the time of the depositions to take DBAG’s part in the dispute and, as far as he was able, he did support SHI’s case. He changed his stance (from that in his affidavit) on Mr Vik’s knowledge and understanding of the Structured Options in the light of the documents which showed discussion between them on the subject.
In coming to the views that I have, I have therefore had close regard to the contemporary documents and actions of the persons involved, the commercial probabilities in the light of their relevant FX and business knowledge and the experience and personalities of those involved as shown by the contemporaneous reactions of others to them at the time.
Mr Said had experience of the FX market over a period of twenty years and was knowledgeable about how it worked. He was self-confident, decisive and forceful. He knew what he wanted to do and bent others to his will. He had been Global Head of Foreign Exchange and Money Markets at CS for five years from 2001 to 2006, having previously worked at JP Morgan, rising to Global Head of Foreign Exchange, Short Term Rates Trading and Precious Metals between 1999 and 2001. He knew how banks operated and knew how to put pressure on the employees of DBAG, including in particular Mr Quezada and Mr Walsh in the FXPB department. It is plain from Bloomberg chats that he felt able to exert leverage on them by threatening to withdraw SHI’s business or putting them in bad odour with Mr Vik, an important billionaire customer of DBAG. As appears hereafter, he was able to persuade them to take on trades that they were reluctant to accept. He was also a good negotiator and was prepared to gild the lily when talking to others and seeking to persuade them to adopt a course of action more amenable to himself. As any trader would, he kept his own records of trades and had his own pricing tool for vanilla transactions. He could see market movements in spot trades, forwards and volatility. He could check his own records against those of DBAG’s web reporting system at any given time and was able to point out errors in DBAG’s MTM calculations on both vanilla and exotic trades, when they did appear. He could see where EDTs were not booked even if he had been charged the fee for matching trade confirmations.
Mr Vik was a highly experienced businessman who traded in FX himself, essentially in directional trades, taking a longer term view of investment in the FX market. He was astute in financial affairs and, whilst not a man for descending into intricate detail, exhibited a ready grasp of the essentials of any trade proposition, including in particular the risk/reward equation. Described as “savvy”, he could readily understand the nature of varied financial transactions of considerable complexity with a keen eye for the numbers. Mr Brügelmann at DBS held Mr Vik in awe because of his extraordinary faculty for retaining details of his trading positions in his head and his ability to conduct his business on a blackberry, without detailed records in front of him. Mr Said valued Mr Vik’s opinions and judgment in the context of FX transactions, though Mr Said, with his years of experience, knew much more about
the different products and how they worked in the FX market. If Mr Said explained a trade to Mr Vik he had no doubt, and I have no doubt, that Mr Vik would readily understand the essential components of it. Mr Said was not a man who was slow coming forward and readily expressed his market views in Bloomberg chats and the nature of his trading activity and underlying trading philosophy.
Mr Said was described by some witnesses as a bully. It was clear from the contemporaneous exchanges between Mr Walsh and others, whether by email, telephone or instant messaging, that as a 23-25 year old, Mr Walsh found Mr Said overbearing and could not, given their relative positions, say no to him. He took guidance from Mr Quezada, on the FXPB Product Sales Desk, but Mr Quezada was likewise incapable of resisting Mr Said’s blandishments and his threats about taking the business elsewhere if DBAG would not accept the trades that he wanted to carry out. As appears from the contemporaneous documents, there were a number of personnel at DBAG who wished to refuse to accept the EDTs and/or OCTs that Mr Said wished to trade through FXPB because of the difficulties that were involved in booking, valuing and margining the trades but their objections were overborne by Mr Said’s threats to take the business elsewhere and in particular the loss of business that would accrue from his direct trades with DBAG’s Execution Desk in the person of Mr Geisker.
Whilst personnel in FXPB in London objected to these trades being taken in and Steven Kim, the Global Head of FX Prime Brokerage Operations in New Jersey and Mr Manrique in the FX Options Operations Team under Mr Kim also took the same view, the decision in the end rested with Ms Liau’s department and in particular Mr Quezada on the Business side (as opposed to the Operations side) of FXPB. If the decision was taken to take the trades in, the Operations department, having made known its difficulty in coping with them, in terms of booking, valuing or margining, had to go along with that and Mr Walsh, the most junior of the employees, was left to deal with Mr Said on the basis of the instructions given to him by Mr Quezada and by Mr Said. It was essentially Mr Walsh and Mr Quezada with whom Mr Said dealt at DBAG.
Mr Said, as is plain from the contemporary documents and the telephone conversations, kept his own records of trading. Like any other trader he knew what his positions in the market were. With his own calculator he could assess his MTMs on vanilla trades, for spot transactions, forwards and vanilla options, drawing on published market sources such as Bloomberg. The spot rate, the forward rate and the volatility rate were all there to be accessed. It is clear that he was able to tell DBAG when their assessments of MTM were wrong on such trades, which he did from time to time and with some vehemence.
It is clear from all the evidence that Mr Said began trading for SHI with essentially vanilla trades. For the first couple of months at the end of 2006 he sought Mr Vik’s approval before conducting any individual trades. By February 2007, confidence had grown between the two and Mr Said no longer needed to seek Mr Vik’s prior approval. He then began to branch out into non-vanilla trades in the shape of OCTs and between February 2007 and July 2008 he traded in fifty-three OCTs and fortyone EDTs (as they have subsequently been classified). This still represented a relatively small proportion (11.7%) of his total trading in terms of numbers of trades. Mr Said expressed his underlying trading philosophy for these non-vanilla trades on a
number of occasions, in email, Bloomberg chat and telephone conversations. Whilst there was a measure of volatility in the movement of currencies against one another, he saw the market as being essentially directionless in 2007/2008, particularly during the summer months, when he described the market as going “into hibernation”, an anomalous expression for the summer months. He maintained that the only way to make any real money in such periods was to “short volatility” – in other words to enter into trades which amounted to a bet that currencies would not move against each other beyond certain limits, whilst they might move within them. He described these as “range bets” or “range trades”. He was betting against increases in volatility beyond the specified range in the individual transactions. He considered some currencies to be mean reverting and although they might move out of the ranges specified from time to time, they would in the ordinary way come back within the limits of those ranges and under the terms of the trades he would make profits. If therefore a trader was prepared to ride out the storm of temporary volatility and the losses incurred when currencies did move beyond the ranges, profits were there to be made because they would come back in, but a trader had to be prepared to hold onto the trades and treat them as accrual trades with the profits building up and offsetting the occasional loss.
Mr Said thus perceived EDTs and other similar OCTs as good range bets. They were zero cost options with no money to be paid up front and, where pivots were involved with a knock out feature, the latter had the effect of limiting the total profit made but, in those circumstances enabling the trader to obtain a wider range within which the currency could move with greater or lesser degrees of profit, without touching the upper and lower strike prices where losses would begin to accrue. Such trades were not to be unwound because they were “buy and hold” trades and the unwinding of them would be prohibitively expensive. At almost all stages they would be likely to be out of the money on a mark to market basis, right from the moment of the trade being done. He regarded his position at SHI as giving him an advantage over banks or hedge funds because he did not have to respond to MTM movements in the way that they did.
It is clear that he fully understood how pivot TPFs worked and saw them as accrual trades. He told Mr Vik that “you have to watch them like a hawk” and if they moved towards the range, they had to be hedged with either directional TPFs or with other pivot trades with a wider range (i.e. a higher or lower strike which would enable profits to continue to be made as the market moved further away from the pivot on the earlier trade and beyond the strike set in that trade).
He knew in broad terms, because he said so, that the MTMs of TPFs were essentially negative at the outset and were generally negative throughout the duration of the TPF until knockout. He knew also that the MTM related to the spot rate, the forward rate, and implied volatility – information on all of which was available to him on Bloomberg. He also knew that losses could be substantial if the spot rate moved beyond the range and stayed there for any length of time. It is self-evident that any competent trader would know that the notional on a TPF, whilst expressed on the basis of one fixing, depended upon the number of fixings for the trade, whether weekly or daily. This is plain from any trade confirmation and the manner in which profits and losses accrued. Mr Said’s deposition evidence and the email exchanges show that he was well aware that profit was limited by the knock out feature to the “target” figure in the TPF, whereas losses, as with any sale of an option, were indefinite and theoretically without limit up to the total notional. It is clear that he knew the risks he was running in entering into these trades but considered that the substantial losses which ultimately occurred were extremely unlikely to eventuate. He referred to that possibility as involving “the world [going] to hell in a hand basket”. He therefore regarded the risk of losses of the magnitude which were ultimately suffered as so unlikely that it could be effectively discounted. He was proved wrong by the events of the autumn of 2008 which he described as a “perfect storm” in which he was caught up.
The Timeline was originally drafted by Mr Said for the benefit of SHI’s counsel and therefore was undoubtedly intended to help SHI’s case, but it is also to some extent self-serving on Mr Said’s part. Nonetheless a number of features appear clearly in the Timeline which contradict the case which SHI now makes:
Throughout 2007 Mr Said described his trading as successful with the build up of a positive profit figure which increased the margin amount available and the size of positions that could be taken. Mr Said considered the position heavily over-collateralised although, when Mr Vik withdrew US$30 million in cash from the account on 9th October 2007, available margin reduced once more. Overall profit for the fourteen months’ trading in 2006 and 2007 was, he said, about US$45 million so there was excess margin left in the account from built up profit even after that withdrawal of cash. He regarded the profit earned by his trading as increasing the margin available to him.
In 2007 Mr Said said he did several non-standard Structured Options, as defined in the FXPBA. He described the term Structured Options as “basically a catch-all of options” that the automated deal capture and MTM system of DBAG could not handle and which was subject to case by case approval by DBAG as Prime Broker. These were, after brief explanation, he said, accepted for give-up by DBAG. Mr Said said that he could not recall whether they had any major margin implications but he thought not because the margin situation was extremely benign throughout 2007 (and really until October 2008).
In 2008 Mr Said said he did more Structured Options and in February 2008 was shown by CS the Pivot Accrual structures. He found two which had very favourable risk reward ratios (a very high probability of knocking out early in most market scenarios given the short average life) and explained the transactions to DBAG, saying that they did not really involve exchange of cash flows until unwind, knock out or maturity. DBAG thought about it for a day or so and then said they were happy to accept these structures for give up and gave their approval for these Structured Options under the FXPBA.
Over the course of the year Mr Said said he did a meaningful number of these on a rolling basis and the market for much of the year was as he expected it to be – full of sharp moves, sometimes somewhat random, with plenty of volatility but ultimately no massive powerful trend. The portfolio performed very well despite the Bear Sterns crisis in March and the last big dollar weakening with the result that in early October realised profit on the option portfolio amounted to about US$65 million, having been around US$82
million in August but deteriorating since because of the weakening of the NOK.
The Timeline continued:
“DB as PB accepted all the trades as I did them and processed the knockouts and payments as they occurred. I did not receive mark to market on these structures form [sic] them however and I did not notice an appreciable impact of the options on the required margin calculations. The options also either did not show up at all in the online P+L or were there, but with nonsensical P+L numbers. There were some discussions with DB about their ability to handle these – I wanted to make sure they were in a position to support them so I initiated the conversation with Rafael Quezada. DB’s position I recall (from memory of phone calls) as follows:
we can support these structures
we want to support these structures
we should be able to mark them to market
But only our trading desk can and they don’t like doing it for deals not done with them.
They asked me quite directly to do some of these deals with the DB desk (which has not distinguished itself in terms of pricing whenever I gave them a chance) to “create some goodwill so we can work with the trading desk on the other structures. DB actually improved their pricing and I did several transactions with them.
That seemed to settle any residual issues DB might have had with booking or handling these options.
In terms of margin impact – it is not clear to me exactly how much INITIAL margin one of these structures should attract. But it seems form [sic] following the margin daily that DB may not have attributed any. In terms of variation margin – many of the options knocked out quickly and benignly without ever developing much mark to marked [sic] value – but some definitely did (I recall the very first euro Norwegian Krona option went right to the top of its ban and stayed there for a while before – as had been my view, retracing and knocking out with good profit. Again – from memory, I do not recall an impact on the margin calculation – and looking at it now, there should have been given the option must have a decent size MTM loss for a while which should have meaningfully decreased the margin capacity. It is true that there was built-up profit in the account which would have meaningfully INCREASED the margin capacity (see above). DB pointed this out on a call to me. However, much as he did in 2007, Alex did in the summer (I believe in July) withdraw 66mm$ from the account (a move I suggested to him given the cash was lying idle) which would have been substantially all the built up profit. Therefore in terms of margin capacity, we should have been back to the 35mm$ we started with (or in the general neighbourhood).
The portfolio of these options was actually very similar through-out much of 2008 – primarily eur/chf, $cad and eur nok with some currencies like aud/nzd eur/stg, eur$, $ yen stg/chf and $/brl added on occasion.
Throughout 2008 there were no margin calls form [sic] DB nor was MTM from the options represented in the P+L.”
It is relatively clear from this summary that although Mr Said refers to a settlement of any residual issues DB might have had with “booking or handling these options” following his discussion with Mr Quezada, there is a gap in his logic as to how this could be the case since only the Trade Desk was able to mark the transaction to market and it did not like doing it for indirect trades given up to DBAG as Prime Broker. Mr Said stated that although it was not clear exactly how much initial margin these options should attract, he appreciated, from following the margin daily, that DBAG might not have attributed any at all. He did not also recall any impact on the margin calculation at all when there should have been variation margin by reference to changing MTM. He said that there were no margin calls “nor was MTM from the options represented in the P&L” (by which he meant the reports of MTM).
Mr Said then referred to DBAG’s request in August of 2008 for a meeting to discuss margin in New York. He appreciated that the original terms were “simply too generous”. Discussions culminated in a confirmation on October 6th that new margin terms would only raise the required margin from US$21 million to US$40 million that day. In discussing this Mr Said said that “during our meeting on September 8th in NY they did reference the structures and said they were having some issues incorporating them into the margin calculation but would get to it shortly”. It is thus apparent that Mr Said knew that the Structured Options did not yet appear in the margin calculations. viii) The Said Timeline continued:
“The portfolio of options (still) did alright through September and early October (all but three accruing positive every day) with the exception of the 4 $ brl structures (really part of one trade but spread over time). $ brl had started to move up steadily and in the first week of October the move suddenly accelerated. I was aware that we were looking at negative accruals and what had to be a decent MTM loss (I believe I wrote to Alex about these options and the strategy given the illiquidity of the market). DB did not, to my knowledge react to any of this, nor was any negative MTM incorporated in the margin calculation.
What happened next, I believe in early October (week of Oct 6th I think) was that Morgan Stanley, which had dealt with three of the four structures in question, apparently approached DB about separately margining these. Suddenly I got several calls form [sic] DB now asking about these options and did I have a MTM on them or could I get it from the counterparts. I
think I told them I was pretty busy managing our risk in difficult markets and I wanted them to get the mtms – as they had initially said they could and would. This went on throughout the week of Oct 6 while I was discussing with Alex how to proceed on these options which were clearly showing a meaningful loss – but it was also not clear to me that cutting them out here was necessarily the right approach. Brl weakened steadily throughout the week, but we did not receive any margin calls from DB.”
Again it is clear from this that Mr Said appreciated that MTM on these Structured Options was not incorporated into the margin calculation and that it was only the approach from MS seeking separate margin from DBAG that led to the latter asking Mr Said for MTM or asking him to get it from the counterparty banks, though the matter was left on the basis that DBAG would get those figures from the counterparty banks themselves. During this week Mr Said was discussing matters with Mr Vik (as the emails show).
ix) Mr Said went on in the Timeline to refer to the margin calls saying that it appeared to him that it was only then that DB “had pieced together a picture of what they thought the MTM exposure was on all these structures”. He then referred to the market as being in “full blown crisis mode – extreme stress” and “a perfect storm” that “hit most of our positions”. He referred to DBAG’s first margin call as wrong because the bank clearly did not have the right MTM numbers, (including some positive exposures which made no sense) and the figures were too low given the market moves. He then referred to the collapse of currency markets that week and the multiplication of margin calls from DBAG where he said that the numbers they based their calls on were “extremely spurious – some accurate, some way off”.
670. Against the background of that Timeline drafted for SHI’s lawyers’ benefit, the affidavit of 21st May sworn by Mr Said in the New York proceedings falls to be considered. Paragraphs 8 and 9 of that affidavit read as follows:
“8. The structure of the collateral was also discussed. I explained to the Bank (and I understand that Mr Vik did as well) and the Bank understood that my trading had to be separate and isolated from other Sebastian Holdings’ assets and that Sebastian Holdings was only willing to expose a specific sum for my trading. The Bank recommended and agreed that this would be accomplished by Sebastian Holdings, in connection with the opening of the New York FX PB Account, pledging as collateral the equivalent sum of $35,000,000 in a newly opened separate account of Sebastian Holdings with the Bank in Geneva, Switzerland and that the Bank in Switzerland would issue a guarantee against such account, in such amount to the New York FX PB Account to support the FX trading in New York. This would also create a system of checks and balances for Sebastian Holdings as, for instance, Thomas Brugelmann could monitor the risk in the New York FX PB Account from the balance in the pledged account.
9. All of the trades I did in the New York FX PB Account were based on the $35,000,000 pledged by Sebastian Holdings in the Geneva account in Switzerland and the guarantee issued by the Bank in Switzerland to the New York FX PB Account. I understood at all times, as did the Bank, that my trading was limited to the specific amount of collateral and no more. Indeed, on two separate occasions, Sebastian Holdings transferred funds out of my account as such funds were not used to support my trading. There was never any discussion or agreement that any of Sebastian Holdings other accounts or assets would be available as collateral for my FX trading. In fact, in October 2008, Rafael Quezada of the Bank requested that I ask Sebastian Holdings to increase the pledge. From earlier communication with Mr Vik, I did not think that Sebastian Holdings would consider increasing the pledge and I never made such request of Mr Vik.” 671. The affidavit continued:
“15. Throughout my FX trading, I had continuing discussions with the Bank about its obligation to provide accurate reporting, either as part of the Bank's website to which I alone, not Mr. Vik, had access, or the daily reports that Bank personnel, including Matt Walsh, would periodically send only to me by e-mail. Several things should be noted: first, I often checked the "available" collateral on the Bank's website and found that I never got close to the limits. At no time before October 2008 did the Bank inform me that the Bank had failed to include any trades in the collateral calculations. At no time did I ever agree that the Bank had no duty to provide accurate reports. To the contrary, I was constantly assured, particularly by Quezada, that the Bank had a "good system" and that the Bank was capable of providing accurate reporting. Quezada and others at the Bank understood the Bank's obligations to provide accurate reporting and that such reporting was critical to monitor risk.
Indeed, I always made it clear to Quezada that the Bank should only "take in" the structured accrued pivot trades, which I started doing in 2008, if they could handle them and accurately value and put them into collateral calculations.
Quezada assured me that the Bank was able and happy to accept them and every trade was pre-approved and accepted by the Bank. Indeed, Quezada even asked me to do my best to do a few of these trades with the Bank rather than the other counterparties, which I agreed to do for him on a few occasions. The Bank was clearly eager for me to engage in the pivot trades and the Bank was able to value the pivot trades.
I never agreed to conduct pivot trades without their value being reported on the Bank's website. As the Bank well knew I did not have any authority to do so and reporting exposures was a prime obligation of the Bank as it well knew.
The Bank was required to include all trades including the pivot trades, in their reporting and all trades, including the pivot trades, were supported only by the $35,000,000 in the pledged account of Sebastian Holdings in Geneva and the corresponding guarantee from the Bank in Geneva to the New York FX PB Account.
As this was my understanding as well as that of the Bank, I continued such trades in 2008. I did not notice any appreciable impact on the pledged collateral amount for these trades. The structures either did not show up on the website or sometimes were there but with nonsensical numbers which I pointed out to the Bank on several occasions in my efforts to make sure the reporting was correct. In all events, I engaged in such trades relying on the Bank's obligations to Sebastian Holdings as its prime broker and pursuant to the New York FX PB Agreement.
For example, I sent an e-mail to Matt Walsh alerting him that the numbers in the live mtm module relating to two earlier pivot trades did not seem accurate to me and I thought that these inaccurate numbers should be excluded from the real-time reporting system until they were corrected so as not to render all real-time information erroneous. This e-mail related only to those two trades and only about the real-time reporting. It was not an instruction to exclude pivot trades from being valued in the Bank's system. Communications like this were to make sure that, among other things, trades were properly matched and documented and correct information was being used. I again continued to rely on the Bank's assurances that they could value the trades and correctly report their calculations.
Never once during the many months of my pivot trades did the Bank ever suggest to me, nor to my knowledge, Mr. Vik, that the trades were in excess of the collateral limitation ($35,000,000) or that there was "inadequate security."
While in late August 2008 the Bank, in New York, did ask me, not Mr. Vik, to have discussions take place concerning what eventually resulted in their unilateral change of collateral calculation methodology, never once was I advised by Michael Spokoyny (or anyone else at the Bank) that he was aware of any deficiencies in collateral or what the Bank has come to now allege were "hundreds of millions of dollars of losses."
To the contrary, when the Bank and I (not Mr Vik) did meet, pivot trades were raised and I was assured by Spokoyny, as I had been in the past that the Bank was accurately valuing these trades and including them in their collateral calculations. All we discussed and eventually received was the Spokoyny email of October 6, 2008 unilaterally requesting that the methodology for calculating collateral requirements was to be modified with the result that the required collateral in the account of October 6, 2008 was to be increased by $5,000,000 ($35,000,000) to $40,000,000. Because my collateral was limited to $35,000,000, and Sebastian Holdings was not interested in increasing the pledge, this required me to reduce my trading positions. No mention was made of pivot trades or the fact that by that time the Bank may have known about losses amounting to hundreds of millions of dollars, all of which was unknown and unavailable to me and of course to Mr. Vik with whom, prior to mid-October 2008 I did not discuss my pivot trades.
I believe that it is only when the Bank thought it was going to receive a request from a counterparty (I believe Morgan Stanley) in October 2008 to post collateral for individual trades because of "mtm" (that is, mark to market calculations) done by such counterparty, that the Bank finally realized that it had to disclose to Sebastian Holdings what the Bank alone knew all along: that the losses had been and were becoming staggering and that the Bank had failed to comply with its calculation and reporting requirements to Sebastian Holdings under the New York FX PB Agreement and the prime brokerage relationship.
Even then, the Bank, recognizing and well aware of the $35,000,000 collateral limitation requested it be increased by only $5,000,000 to $40,000,000. The Bank, knowing I (and of course Mr. Vik) had no access to mark to market calculations, was the only party which could accurately calculate and report collateral requirements and it failed to do so.
Indeed, when I received the first purported "margin call" from the Bank on October 13, 2008 it was erroneous and the Bank knew that it was erroneous; understated by hundreds of millions of dollars. Had the Bank reported accurately, I would never have entered into the trades and I would have liquidated any trades on an earlier and more timely basis and Sebastian Holdings would have suffered substantially lesser, if any, losses and the wrongful margin calls would not have been satisfied.
My trading was supported only by the $35,000,000 guarantee issued by the Bank in Switzerland to the New York FX PB Account and neither the Bank nor any other party provided any other financing to support my trading activities. I did not have any authority to borrow from the Bank nor have I ever done any trades with the Bank on "margin".”
The difference between the Timeline on the one hand and Mr Said’s affidavit on the other in relation to his knowledge about the absence of MTM and margining of the Pivot Accrual trades is self evident. The affidavit makes out that DBAG knew throughout that it was not charging appropriate levels of collateral in respect of these trades whereas he and Mr Vik were completely ignorant of the position. The affidavit makes out that although he had continuing discussions with DBAG about reporting, he never agreed to conduct pivot trades without MTM reporting on DBAG’s GEM website, that he occasionally pointed out nonsensical numbers on the website to the bank in an effort to ensure the reporting was correct and that he relied upon DBAG accurately reporting MTM and margining accordingly.
What emerges from the contemporaneous exchanges between Mr Said and DBAG personnel in 2008 to which I refer elsewhere is an entirely different picture. Recognised in part by Mr Said in his depositions, the picture which emerges is an awareness on Mr Said’s part that DBAG could not book, value or margin the EDTs properly (nor many of the OCTs) because its systems were not capable of handling them.
Mr Said knew from the GEM website and from the manual spreadsheets supplied to him that the EDTs were not being booked accurately, that where they were being booked they were being booked as “Resurrecting Faders”, an inadequate proxy or placeholder for the transactions in question and that in the later stages they were not being booked at all until cash settlement.
Moreover Mr Said had gone to some lengths to tell Mr Quezada and Mr Walsh, in an effort to persuade them to accept the trades as Structured Options under the FXPBA, that DBAG did not have to do anything on the trade – they did not have to keep track of it or provide MTM and that the only time they had to get involved was when the trades knocked out or matured and cash settlements fell to be made.
He regarded margining as a problem for DBAG to work out for itself but inevitably, as an experienced trader, knew that if DBAG could not value the trades on an MTM basis, it could not margin them either.
He knew that the MTMs on EDTs which were booked as Resurrecting Faders were “spurious by tens of millions” on the website and asked for the MTMs to be “zeroed out” or removed from the reports and the manual spreadsheets which Mr Walsh sent him which otherwise showed a comprehensive list of his vanilla trades and the applicable MTM valuations. He also appreciated that margining on the basis of the valuations on the website would lead to numbers which were “chaotic”.
The reality therefore was that, in consequence in particular of telephone calls on 5th May and 22nd July 2008, Mr Said procured the agreement of DBAG to accept trades, the details of which could not be booked on GEM and which could not be valued or margined correctly, if at all. He was more than content to go along with that in order to use the prime brokerage to contract the EDTs and OCTs he desired. He told DBAG personnel that MTM was of no consequence to him because he regarded these as accrual trades or “buy and hold trades” and it was the cash accruals of unrealised profits and losses to which he had regard. As appears from his exchanges with individuals at counterparty banks and Mr Vik he knew that he was getting a “free ride” on margin because of DBAG’s inability to value the trades for a period of some 6-12 months and throughout the relevant period he was doing his best to keep the margin requirements down and was not in the slightest dependent upon either DBAG’s MTM valuations or its notification of margin requirements in conducting his own risk management.
Moreover, on odd occasions he obtained an MTM value from counterparty banks and any comparison of those with DBAG’s MTM figures or margining would inevitably have showed the inadequacy of DBAG’s figures (compare his comments made in the Timeline). The key point is that, as discussed directly between DBAG personnel and Mr Said, he knew, because he was told as much, that the FXPB department did not have access and never obtained access to the MTM tools used by the DBAG Complex Options Trade Desk, namely DB Analytics, for valuing the EDTs. He at one stage offered to help FXPB to gain access to DB Analytics by offering to do direct trades with the DBAG Trade Desk and suggesting that it should make the DB Analytics tool available to FXPB, failing which he would do no further trades with it. This all came to nothing, as he knew, since the Trade Desk was not, for whatever reason, prepared to make that system available to FXPB (whose personnel would have to be trained in the use of it). The reason for this was almost certainly the reluctance of the Trade Desk to assist FXPB in facilitating Mr Said’s trading such EDTs with other banks rather than the problem of preserving confidentiality, since Mr Said was content for details of his other trading to be disclosed to DBAG’s Trade Desk.
SHI, in its submissions, accepted that during the whole of 2007 and the first half of 2008 Mr Said knew that DBAG was not able to produce accurate MTM valuations for some OCTs and all his EDTs. The position remained the same throughout the rest of the period of Mr Said’s trading. There was nothing which could have altered his understanding on this point. Nor could he have, at any time, understood that the EDTs were being properly margined since such margining depended, so far as concerned variation margin, on DBAG’s systems producing valid MTM figures. This appears at least in part from his evidence on deposition. Elsewhere I set out the more significant exchanges between DBAG personnel and Mr Said which make his state of knowledge and agreement to the absence of proper booking, valuing and margining, clear, to the extent that he was not prepared to admit it on deposition.
In this part of the judgment I mention only in passing the resiling by Mr Said in his depositions from what he said in paragraph 8 of his third affidavit about SHI’s structure being established to separate and isolate his trading from other SHI assets and his comments about any limitation on his ability to trade by reference to the provision of the sum of US$35 million alone as capital. That too is dealt with elsewhere in this judgment. My focus here is on the issue of SHI’s booking, valuing and margining of the EDTs and OCTs.
Mr Said’s evidence on deposition was that all the EDTs or OCTs (which were given numbers for the purpose of this action) were the subject of prior approval by DBAG in the persons of Mr Quezada or Mr Walsh and SHI does not contend otherwise.
Mr Said’s deposition evidence contained a number of inconsistencies. In questioning by SHI’s attorney he started by saying that he had to run the Structured Options by DBAG to make sure that DBAG could accept the trades by booking them and managing them and that he would speak to Mr Walsh or Mr Quezada to obtain their agreement. He said he did not recall DBAG ever saying that they could not do these structured trades and in the end they were booked and executed as at first conceived. He said that in April 2008 he had emailed Mr Walsh to say that the FXPB system was throwing out numbers that were spurious by tens of millions in respect of the Pivot Accruals. He told Mr Walsh to exclude them from the MTM figures. He said in the deposition that he was not concerned over the current MTM of the Pivot Accruals (or faders as Mr Walsh called them) because they would accrue profit and loss over time anyway. He therefore told him to keep them out of the MTM figures in order to ensure that the other figures were straight. He said that he never told DBAG that the faders did not need to be valued or that accurate reporting of them was not necessary.
He said that obviously DBAG were not able to effect the MTM. That was always portrayed to him as an ongoing issue that would be solved but not that they could not do it. He then said that he was given an assurance by Mr Spokoyny at the meeting on September 8th when discussing VaR.
He said that in mid-May he knew that FXPB could not model the TPFs and that only DBAG’s Trade Desk could. He said that Mr Quezada wanted to be able to use that pricing model and had asked him to help put pressure on the Trade Desk to allow him access. This fix for the Pivot Accruals was always just round the corner. The MTM on the TPFs was therefore excluded from the website until they fixed the matter. Mr Said said in deposition that it never occurred to him that DBAG would not value the trades so far as its own risk went. When he sent an email to Mr Walsh saying that the numbers in the live MTM module relating to two pivot trades were not accurate and told him to exclude them from the system until they were correct this was not a general instruction to exclude pivot trades. He said he had no recall of discussing any incorrect booking of such items with Mr Walsh or Mr Quezada. The only initial question was whether they could take it and after a while they said they could.
He went on to say that he did not believe that Mr Quezada had ever said that they had not booked them but were going to find a way. He was merely saying that it was not simple and they would have to work out how to do it. If the bank could not accommodate the trades it had to say no. When discussing such options with Mr Quezada, the latter had told him that one off (non-standard) trades would not be a problem but that he should not swamp DBAG with them. He did not think that any trades were ever definitively declined but he was not 100% certain of that.
He had no recollection of being told by Mr Walsh that DBAG booked TPFs as Resurrecting Faders because there was no other way of booking them. Mr Walsh may however have told him that DBAG’s FXPB systems could only report trade details for
single barrier and vanilla options in addition to swaps, forwards and cash trades. He said he had no recall of being told that the Gated Range Accrual trade in June 2007 could not be reported properly by Mr Walsh though he recalled that there were certainly ongoing issues for reporting the structured trades on GEM which did not lend itself to them. GEM was not however in his mind a risk reporting tool at all so he did not look to it for that. He said he did not believe GEM included VaR reporting. It was only after the margin calls began that he actually learnt that DBAG had not captured the trades or valued them. He was surprised, he said, that DBAG did not have MTM information and had to use a valuation for the MS trades that came from MS, which he was told by Mr Quezada. He was asked then to call CS and get their valuation for trades with them.
He said there was no discussion in May 2008 as to whether the TPFs were being margined, nor at any time at all. He said he had no recall of being told that they were not being valued in the system. If a new trade could not be valued then he would not, he said, be able to engage in it. He said he would have been greatly surprised to hear at the end of May that they were still checking to see if the trades were captured in VaR. He conducted many more of those trades after the end of May and he would not have done one single further trade if he had been told that it had not been booked, valued or margined. It was only in October that he was told that trades had not been booked and he recalled going through all the trades that had been done with Mr Walsh at his request on October 7th. He did not think that Mr Walsh told him then that some had not been booked. He said he was never told that the trades had not been margined correctly.
He also said that he was always under the impression that DBAG’s VaR model could calculate MTM and margin. A sophisticated system was required to value MTMs on complex options and he relied on DBAG for all options. He was capable of calculating the MTM on spot transactions on his own calculator but the more complex and dynamic the positions the more crucial it was to have proper accurate MTM.
He said that in September 2008 he did not have the necessary MTM information from DBAG to be able to monitor the TPFs and see how VaR was affected. He never learnt whether DBAG had captured risk and margin. Inconsistently with his earlier statement in the depositions he said that in September 2008 at his meeting with Mr Spokoyny, the latter told him that they had not perfected booking but were getting there. Something along those lines had been said which was quite comforting. He considered margin a matter for negotiation and when in September DBAG came forward with a proposal to amend the existing margin methodology, he thought he would see if he could get away with less than was being sought.
Later in the depositions, however, he said that he knew all along that DBAG was not booking the TPFs “optimally” and that the MTMs were not reflected properly but he was always being assured that this was work in progress and that “we were getting there”. It was only in October he said that it dawned on him that DBAG did not have a handle at all on the trades, being ignorant of what they had, what the values were and how to deal with the trades. DBAG had received a margin call from MS and was struggling to reconcile the numbers. Mr Walsh asked him if he could get valuations from the counterparty banks and it was at about this time, perhaps a little earlier, that he realised that DBAG was unable to value the TPFs which meant that they had probably not charged margin on them. He had never given the margin situation much thought prior to that and it was never a problem. Because Mr Walsh asked for MTM information from counterparty banks, he deduced that Mr Walsh did not have it.
When asked questions by DBAG’s New York lawyers, he said that the first inkling that he got of real numbers was when the MS figures came through in respect of the collateral they were seeking on his trades from DBAG (October 3rd-October 6th 2008). He said that on and off he got MTM figures from CS but did not receive regular MTM figures from all the counterparties. He thought that they had the ability to provide those figures if he had asked for them. Every day or week each counterparty sent a report of the accrued profit and loss and he got detailed position reports from DBAG which included MTM information save on the Pivot Trades. He knew he was not getting MTM information on the Pivot Trades from DBAG at all.
He was then questioned about some of the exchanges to which I make reference elsewhere. He said that on 28th February 2008 he told Mr Walsh to keep the TPF MTMs out of the calculations because they were nonsense. The live MTM module could not handle it and was giving silly numbers. He understood that DBAG had issues with booking the TPFs but he would have been disinterested in the problems. It would be up to DBAG to say if they could not do the trades. It was up to them to figure out how they were to be booked. He understood that the TPFs were not the subject of MTM at the time of his conversation with Mr Quezada on 5th May 2008, but he personally looked at them as accrual trades. DBAG had an issue of margining from their own standpoint and he told Mr Quezada that if there was concern SHI would over-collateralise. In those circumstances Mr Quezada responded by talking about hard coding the margin. In answer to further questions he said he could not say whether Mr Quezada had ever told him that he had got the pricing tool from the Trade Desk.
He was referred to a telephone conversation with Mr Walsh on 22nd July 2008 when he had told Mr Walsh that the web MTM was total garbage. He told Mr Walsh that DBAG was margining the trades wrong, albeit not by a ton, but usually in SHI’s favour. He said that if DBAG was margining the trades based on the spurious fader numbers it was really chaotic and that if there ever was an issue of the faders not being margined correctly, SHI would just send more margin. He then said he did not recall any discussion of the problem at all at the September meeting with Mr Spokoyny and Mr Quezada.
On being asked about his exchanges with Mr Vik on 25th and 26th August, he said he had told Mr Vik that DBAG were getting there on margin and catching up on the levels of volatility in the market. On 26th August in a Bloomberg chat he had said that he thought that DBAG had finally figured out a way to actually margin all his pivot trades and were trying to break the news to him that the freebies were over. He said he was not sure that he knew that his pivot trades were not being margined but certainly had the impression that DBAG was taking too little margin overall. By October 9th or perhaps a little earlier than that he thought he must have realised that DBAG had not done MTM or margining on the TPFs at all and had discussed in principle with Mr Vik the free ride they had been getting for the previous six months from the perspective of MTM and collateral.
At one point in his depositions, Mr Said said that he had been left with the impression that the problems that DBAG had encountered with booking, valuing and margining
had been fixed. He could however point to nothing to justify such an impression and all the evidence points the other way.
What emerges from his deposition, when regard is had to the contemporary documents, is that, when faced with the transcripts of telephone conversations or emails, Mr Said could not gainsay their contents but still sought, so far as he could, to hold to the line that SHI has adopted in this action, namely that he did not fully understand that DBAG was incapable of margining the TPFs in accordance with its systems and methodology because it could neither accurately record the trade details in its system nor value the trades accordingly. That position is not sustainable in the light of his Timeline and the contemporary documents.
Mr Said did not suggest at any time that he used the GEM web reporting for risk management purposes. He accepted that he was uninterested in the MTM of the TPFs because he regarded them as accrual trades. Had he been interested in the MTM, he could have sought such figures from the counterparty banks or DBAG’s Trade Desk but he did not do so because he did not regard this as a relevant factor in his decision making. He considered margining to be an issue for DBAG to resolve for its own purposes and his objective, both in the original negotiation in 2006 and in August/September 2008 was to procure the lowest level of margin that he could.
It is plain that, if Mr Said was following the reported margin daily, he must have realised that DBAG was not attributing any initial margin at all to the TPFs. It is SHI’s case that he was following the margin in the Collateral Summary Report. He knew that the MTM from the Structured Options was not represented on GEM and that when the Faders appeared in it, the numbers were spurious. He knew also that variation margin (which was based on MTM) had to be wrong because Mr Walsh had told him that margining was based on the MTM figures produced on the GEM website, to which he had access.
When then Mr Said said in the Timeline that “it seems from following the margin daily that DBAG may not have attributed any initial margin” and that, from memory, he did not recall an impact on the margin calculation from MTM movements (which there should obviously have been as a result of changes in variation margin), this is not simply an ex post facto realisation.
In the Timeline he referred to the request in August 2008 for the meeting to discuss margin, which took place on 8th September 2008. He said he was unclear what DBAG had in mind but, as it turned out, the topic was the original terms that he had negotiated, which DBAG now felt were too generous. He asked for a proposal which resulted in DBAG putting forward a change in the VaR multiple and a liquidity addon. The discussion culminated, as he described it, in new margin terms of 2.5 x VaR and a liquidity add-on “which would have raised my required margin on oct 6th form [sic] 21mm$ to 40mm$!”. The relevant paragraph in the Timeline concludes with the reference to the meeting of 8th September where reference was made to the TPFs and the difficulty in incorporating them into the margin calculation which they “would get to shortly”. The terms of the paragraph, including the exclamation mark following the reference to the new proposal, show that Mr Said had been expecting to receive a much higher margin proposal because he thought that DBAG had at last found a way of margining the EDTs, whether by use of the Trade Desk DB Analytics tool or otherwise, only to discover that they still had not. Once again, that point emerges
more clearly when the terms of Mr Said’s recorded conversations with others are brought into the picture.
It is therefore clear from his deposition that on 5th May 2008 Mr Said accepted that DBAG was unable to make daily reports of MTM on the EDTS and to book them properly because he knew that the FXPB systems were incapable of accurately recording their terms. He knew that efforts were being made to find some alternative way of achieving this but that this had not occurred by September 8th and no-one had told him afterwards of any change in the situation. The impact of this on someone who was following the margin on the website, as against the particular trades done and the state of the market, would be apparent, as Mr Said recognised in his Timeline. As a trader he might well be expected to do this.
Mr Said could see the impact or lack of impact of market movement on the margin shown on GEM as well as the inaccurate numbers for MTM spilled out by the system when it produced any numbers at all for the faders, which he zeroed out or asked Mr Walsh to zero out on manual calculations, as appears elsewhere.
As to his discussions with Mr Vik, Mr Said said that he described every form of structured option that he traded under the FXPBA with Mr Vik in some form or fashion. Mr Said stated that his affidavit was wrong when it said that he did not discuss the pivot trades with Mr Vik as could be seen from the email reports that he sent. He said that the pivot trades were raised on numerous occasions because of the need to explain the P&L figures, particularly in June 2008. Mr Said recalled several conversations where he explained that the high profits gained in 2008 were due to the pivot trades which had knocked out rapidly, although there was still a lot of risk outstanding in others which was still extant. He said that his affidavit was not correct on this and on other points and “the military answer would be no excuse” by which he meant that he had effectively signed the affidavit under orders.
Mr Said, in his deposition, said that he had been brought in by Mr Vik to bring broader product breadth than Mr Vik’s own FX trading which was essentially directional with spot and forward trades. He said that he kept Mr Vik informed as he did new products that he had not done before. These were discussed. He specifically recalled discussing the first simple exotic which was a Knock-out on the NOK. He told Mr Vik what his view was and how it worked. There were fairly frequent exchanges of thoughts and ideas and he appreciated Mr Vik’s views.
He recalled a lot of discussion before the Gated Range Accrual trade in June 2007. He said he would discuss exotic options with Mr Vik, certainly the first time he did them and would do so directly in the office when he was there. He said that Mr Vik had no difficulty in understanding them. Anything new would be discussed with an explanation of the basic payoff characteristics and the risk but if he had done a Structured Option two or three times he would not refer that back to Mr Vik on further occasions. He discussed the risk of the TPFs with Mr Vik before doing his first trade in February 2008 after Mr Chapin of CS had approached him. He recalled speaking to Mr Vik whilst in the car driving down to Newport. He thought that the type of trade was so interesting that he sat down with Mr Vik and explained to him how it worked. It was later, in an email, that he referred to the unlimited risk involved in such TPFs “if the world goes to hell in a hand basket” but he considered that it would require such enormous moves for this to happen, given the way he structured
the trades, that it was unlikely. Mr Said said that he and Mr Vik discussed the TPFs on numerous occasions and he explained them by using examples. He particularly discussed the issue of accruals on these trades and the P&L which arose there from.
He discussed the risk characteristic of what he described as accrual trades, range trades, range bets or TPFs saying that they were known by a variety of different names in different banks. He said he tried to use words which explained the general nature of the transactions when talking to Mr Vik who was always interested in the risk characteristics although not the detail.
His weekly reports referred Mr Vik to his open positions and to his MTM with the exception of the “range trades” or “range bets” where there was no MTM and where he recorded only realised profit or loss. On occasions he referred to TPFs that might knock out soon and that had good accrued unrealised profits but he never included the accrued figures in the overall P&L numbers he reported prior to knockout.
Mr Said’s evidence about his discussions with Mr Vik on the subject of EDTs was in direct conflict with that of Mr Vik, whose evidence on this, as on many matters, I was unable to accept. The contemporary documents bore out what Mr Said had to say on this topic.
15. Mr Said’s Agreement to the Non Reporting of the EDTs, MTMs and Margin calculations which included them
The TPFs which constituted EDT 1 and EDT 2 were concluded with Mr Chapin of CS on 19th February 2008 but, prior to doing so, Mr Said spoke with Mr Walsh on the telephone (of which there is a transcript) explaining that these trades were variations on the Pivot Accrual trades of the previous year. These had been booked as Resurrecting Fader options – OCTs 16-19. The difference, as explained by Mr Said, was that TPFs had a cap on the maximum profit and, although there was a risk of loss, Mr Said said that only a massive move in the market which did not revert could cause that and he would hedge against that in any event. He said that these TPFs had less risk than the 2007 Pivot Accruals because of the knock out feature which meant that the trade would have shorter maturity and therefore less chance to move away from the pivot outside the range. In fact the Pivot Accruals, which were combinations of Resurrecting Fader options, had a different payoff formula and, because of his view of the unlikelihood of substantial losses occurring, in his telephone conversation with Mr Walsh, Mr Said downplayed the risk element. What was clear from the evidence of Mr Walsh and from the exchanges between him, Mr Quezada and Mr Said, was that not only was Mr Walsh scared of Mr Said but that he had limited understanding of the complexities of the TPFs and of the risks involved and had to have their basic structure explained to him in October 2008. Similarly Mr Quezada appears to have had a limited understanding though I am more dubious about this. Mr Said however fully understood the trades and was prepared to gloss over the potentially huge downside, of which he was well aware, in requesting DBAG to approve the EDTs.
Mr Walsh encountered difficulty in booking these trades, as he told Ms Ng. He booked them as “Resurrecting Faders” as he had for the previous Pivot Accrual trades, with the result that the GEM system, where it did produce figures, produced MTM figures for a different type of transaction. Mr Said executed EDT 3 on 27th February 2008 and the following day he sent Mr Walsh an email telling him that DBAG’s GEM system could not handle the TPFs and was producing “silly numbers”. He told Mr Walsh to “keep these out of the live MTM module”. This was an ongoing theme in his conversations with Mr Walsh. Mr Said appreciated that the GEM system could not cope with the trade details of these TPFs and knew that if the appropriate details of the transaction could not be fed into DBAG’s system, the numbers produced for an MTM valuation could not be accurate. He was concerned about the potentially distorting effect of those numbers on figures for other trades. He must therefore have appreciated the potential impact on valuation of the portfolio as a whole and the calculation of the portfolio margin in consequence. That appears from his Timeline to which I have referred in some detail in section 14 of this judgment.
As mentioned earlier, a problem had previously arisen with the reporting of vanilla FX positions, starting in July/August 2007. Mr Said told DBAG that the live MTM values on the GEM system were inaccurate, which he could tell from published sources. It was discovered that this was due to system-wide issues involving a “feed issue” between ARCS VaR and GEM (resulting in delay in the time between a trade being booked and its valuation appearing) and a lack of correspondence in the Reval rate which showed in GEM and that used by ARCS VaR to calculate the MTM. This problem led to the establishment of the “dummy” reporting account to which Mr Walsh, but not Mr Said, had access. This account used NOP margining as opposed to VaR margining and Mr Walsh would print off a spreadsheet from it which he would send to Mr Said electronically at the end of the day containing MTM values for the vanilla trades. This system of “manual” spreadsheets started on 1st August and continued on effectively a daily basis until 5th November 2007, by which time the problem seemed to have been solved though it recurred from time to time thereafter. In consequence other manual spreadsheets were sent on 27th November 2007, 7th February, 19th, 20th and 22nd February 2008 and between 12th May and 27th August 2008, with one further spreadsheet on 19th September 2008. These spreadsheets were supplied when Mr Said again identified problems with the valuation of straightforward FX positions.
The problem with the EDTs was distinct and incapable of solution in the ARCS VaR and GEM systems. In consequence, Mr Said did not want them to appear in GEM at all and whenever they did so as “Resurrecting Faders” he instructed Mr Walsh to remove them or zero them out which Mr Walsh could not and did not do. On the manual spreadsheets sent by Mr Walsh in consequence of the problem with straightforward FX transactions identified above, where the EDTs appear, they do so as FX Resurrecting Fader Opt, with their trade date and expiry date but with a Current Market Value (MTM) of zero. They either did not appear or their MTM was zeroed out by Mr Walsh. Mr Said wanted the website and manual spreadsheet to exclude the EDTs from MTM valuation and said so as can be seen from his emails of 28th February, 12th May, 3rd July, 21st July and 22nd July 2008. The spreadsheets of 27th August and 19th September 2008 continued to show the EDTs with zero MTM. What appears from all the exchanges is that Mr Said castigated DBAG for its inability to solve the problems relating to MTM valuation of straightforward trades, since he could get the figures right himself with published information and the use of his desk calculator. He did not however at any stage criticise DBAG or complain at the absence of MTM figures for the EDTs. On the contrary, he required their absence rather than the inclusion of “silly numbers” for them which served only to confuse and might have the effect of distorting the overall position on the other trades for
which appropriate MTM figures were (subject to the recurrence of the defect for straightforward trades referred to above) ordinarily available.
There are two key telephone conversations to which reference must be made, namely those of 5th May 2008 and 22nd July 2008, which illustrate clearly the basis upon which the EDTs were accepted by DBAG, with Mr Said stating in terms that there was nothing for FXPB to do save to cash settle the trades on knock out or maturity. He expressly accepted the inability of DBAG to value the trades. He expressed the view that margining was a matter for DBAG to work out for itself but that in any event SHI was over-collateralised and would always produce further margin if required.
By the time of the 5th May telephone conversation Mr Said, who had been referring to various OCT trades conducted from as far back as June 2007 as “offline trades”, was well aware of DBAG’s issues in booking and valuing any options which were not vanilla options. SHI at one point suggested that they were referred to in this way because they could not be booked online by him in the TRM system for Straight Through Processing (STP). However, Mr Said’s understanding of the booking and valuation problems was reflected to some extent in the Timeline to which I have already referred. He appreciated that Structured Options were the subject of special acceptance under the FXPBA and understood the inability of the system to cope with them. On 15th April 2008, by way of example, Mr Said emailed Mr Walsh, referring to a Dual Currency Range Digital Option (OCT39) saying “I realize they are also offline trades and not in the daily p+l”. By this time Mr Said was referring to TPFs as “fader options” or “faders” in his dealings with DBAG and on 18th April he referred to “the faders” being back in GEM web reporting. It was on 5th May that Mr Walsh told Mr Quezada that Mr Said referred to these trades as “fader options”.
Prior to this phone call, as appears earlier in this judgment, DBAG personnel had been having considerable discussions between themselves as to what was to be done with these trades which the systems could not handle. There was pressure from FXPB personnel in London and from the operations personnel in New Jersey, including Mr Kim and Mr Manrique, to cancel these FXPB bookings because of the inability of the system to handle them. What they appear to have envisaged as a possibility (although it was probably unrealistic) is that the deals would be rebooked as direct trades between SHI and the Counterparty bank, by agreement between them. However SHI had no ISDA agreements with such banks and would have had to negotiate and put up collateral to each of the banks concerned as opposed to using the current FXPB arrangements which enabled Mr Said to trade, using DBAG’s credit and only putting up margin to DBAG. Mr Said dismissed out of hand any suggestion made by Mr Walsh that he should conduct trades directly with other banks under new arrangements with them, rather than through the FXPB arrangements he currently had.
Equally, prior to this call on 5th May, Mr Said had sought to conclude further TPF trades with MS and CS at the end of April. By this time it had been ascertained by DBAG FXPB personnel (essentially Mr Walsh and Mr Quezada) that ARCS VaR could not capture the MTM or risk on the EDTs and that the only way that DBAG could book these trades was by using the DB Analytics tool that was used by the FX Trading Execution Desk for complex options. FXPB did not have access to this tool because of the Chinese wall between FXPB and the Trade Desk. So it was that Mr Quezada and Mr Walsh spoke to Mr Said on the telephone on 5th May, the transcript of which is available, to discuss what was to be done. Mr Said was told of the position with regard to the DB Analytics tool and of the problems created for valuing the trades, monitoring the risk and margining them.
15(a) The 5th May 2008 telephone call between Messrs Quezada, Walsh and Said
This telephone call is critical and, although Mr Quezada was at times inexact and confusing in what he said in the call, and was wrong in what he said about the VaR model, the following points emerge:
TPFs could not be booked properly in the FXPB systems, though Mr Quezada understood that they could be booked by the DBAG Trade Desk using their system of spreadsheets (DB Analytics). There was difficulty in obtaining the assistance of the Trade Desk in booking because of the need for the Chinese wall between FXPB and the Sales/Trade Desk since the latter should not be aware of trades done by the customer with parties other than DBAG. This was a cause for concern.
Mr Said’s immediate response was to say that he understood and accepted that the TPFs would not be PnL’d, by which he meant that he understood and agreed that DBAG would not be able to value the TPFs on a daily basis nor report MTM figures to him. SHI accepts that Mr Said expressly agreed to this. Mr Said explained that he understood that the TPFs were too complex for their MTM values to appear and that he regarded them as accrual trades anyway. that he was not interested in the MTM value in any event.
Mr Said then said that the only issue that DBAG could have would be an issue related to margining. He thus showed that he understood clearly, as was obvious, that a lack of proper MTM valuation would impact upon the margin calculation. A little later he stated that he understood that DBAG would “want to look at that” but that if DBAG was ever concerned about the issue, SHI would over-collateralise.
Mr Quezada’s response was to say that margining was an issue and that he and Mr Walsh needed to ensure that the right things were being done “on our side”, because these TPFs were novel to them.
Mr Said then insisted that there was nothing whatsoever for DBAG to do on the trade “from an admin point of view” leaving margining/VaR on one side. DBAG did not have to keep track of the trade as the counterparty did that and there were no payments for exchanges of currencies with the result that the only time the prime broker got involved was on knock out or maturity. “So in terms of you, Matt, and the rest of the gang doing the right thing on a daily basis, I still take issue with that a little because there are no things to do on a daily basis.”
At this point Mr Quezada repeated that the systems used by FXPB to book trades did not cover these trades, as opposed to the Trade Desk with its “accrual spreadsheets and DB Analytics and other tools that they use”. In consequence none of the trades fitted on the FXPB “side of the world” so they
were “sitting on these trades”. Although FXPB knew that the trades were fully offset with the counterparty which gave comfort and that VaR could handle the trades, booking them was a difficulty without the assistance of a quant on the trade side. On that front he was getting no co-operation so that, if he were to take in any more TPFs, he would just be sitting on the trades assuming they were all off-set and everything was fine “and then in 2009 we’re all sitting and saying “what happened?”” (a remarkably prescient observation, which could not have escaped Mr Said.)
Mr Said said he did not understand the issue about offsetting trades with a counterparty because there were trade confirmations from both SHI and the counterparty so that DBAG would be able to match them. He understood that DBAG would not know on any given day what the trade was worth on an MTM basis unless the VaR model calculated it and picked up on the point that Mr Quezada had said that it did.
Mr Quezada’s response was to say that his “guys” were telling him that it did but that the TPFs would not fit into their books.
Mr Said then enquired as to how it worked on a direct trade with the DBAG Trade Desk.
The response of Mr Quezada was that in those circumstances the Trade Desk would manage TPFs outside the system on their own spreadsheets with a hard coded margin supplied by DBAG and separate valuation statements being sent for the TPFs on a trade level – i.e. for individual Transactions.
Mr Said responded that this was not what occurred and that the DBAG Trade Desk did exactly what other counterparty banks did, which was simply to send daily accrual sheets and that he was not receiving MTM valuations for each trade from anyone.
Mr Walsh was asked what was lacking in the booking of the trades and said that there was not a “set trade type” that captured the details of the TPF for booking purposes, which was the same whether the trade was done with the DBAG Trade Desk or with a counterparty. FXPB did not have the
“spreadsheets and stuff” that Sales would use to book the trades so the trades could not be incorporated.
Mr Said said that although he would not be doing TPFs all the time and at that particular time would probably not be doing a new one, it depended on the state of the market. This was an excellent form of trade for his style of trading and he said it was something that he needed to have the ability to do.
Mr Quezada was responsive to this in an imprecise way but suggested that if a new instrument came across FXPB’s desk which “Trading” (the Trade Desk) traded and valued, he might get access to the tools they used but sometimes they responded that they did not “price” these instruments – by which he must have meant that the Trade Desk said that they did not give MTM values.
Mr Said responded that he was happy to assist FXPB in gaining access to “the spreadsheet so you can do it” and to threaten the Trade desk that if they would not assist in this way then he would do all his TPF trades in a private banking account with CS and never offer DBAG another deal of that kind.
Mr Quezada’s response was to say that he liked that approach but would let Mr Said know if it became necessary to use it.
In an instant message chat, whilst the 5th May call was actually taking place, Mr Walsh noted that “Klaus is convincing … I gotta give it to him”. It is plain that Mr Said was not only persuasive but exerted leverage in order to obtain DBAG’s consent to take in the EDTs on the basis discussed.
An interesting insight is gained into Mr Said’s view of this conversation by an internal email sent by Mr Chapin of CS which shows th