Barings Plc & Anor v Coopers & Lybrand (a firm)& Ors

SECTION paragraph number

The Barings group collapsed in February 1995. The immediate cause was unauthorised trading conducted on the Singapore International Monetary Exchange by Mr Nicholas Leeson (“Leeson”), the General Manager of Baring Futures (Singapore) Ltd (“BFS”). Leeson’s trading on his 88888 account incurred losses amounting to £791 million. These losses had been funded by other members of the Barings group, in ignorance of the real use to which the funds were being put. It was BFS’ inability to repay its debts to the other group companies that brought down the group.

In these proceedings BFS and two of its parent companies sued their auditors, Deloitte & Touche (Singapore) (“D&T”), Coopers & Lybrand (London) and Coopers & Lybrand (Singapore). The claims against the two Coopers firms were settled soon after the trial began and in an earlier judgment I struck out the claims by the Barings parent companies against D&T.

Accordingly this judgment relates to the claim brought by BFS against D&T, who were the auditors of BFS in 1992 and 1993. I have concluded that their audits were negligent in two respects:

A number of other allegations against D&T were either dropped in the course of the trial or I have found them not to have been proved.

My findings mean that D&T are liable to BFS for the losses suffered as a result of Leeson’s unauthorised trading from November 1992 onwards.

However I have also had to take into account a very high level of fault by BFS and those to whom the board of BFS delegated their functions. Significant elements of this fault were the failure of BFS’ local management to supervise Leeson’s running of the BFS back office, the failure of management in London and Japan to understand his “switching” trading or question the high profits he reported and the lack of adequate risk monitoring. Perhaps the most notable element of fault was BSL’s provision to Leeson from March 1994 onwards of funding which was not reconciled to trades or customers or supported by credible explanations, without any serious attempt to investigate the reasons why it was needed. In fact this funding was margining Leeson’s unauthorised trading.

In the light of this fault I have concluded that the damages to which BFS is entitled from November 1992 to April 1994 should be reduced by a proportion which rises over time from 50% to 80%. From the end of April 1994 onwards (that is, for the period during which by far the largest part of BFS’ loss was suffered), BFS’ fault was so great as to be the only cause of the losses suffered, and D&T have no further liability from that date.

The detailed calculation of BFS’ damages is to be agreed between the parties (or, if necessary, decided at a further hearing) in the light of the conclusions I have set out above and two further points of principle upon which I have ruled.

These proceedings arise out of the collapse of the Barings Group of companies in February 1995. Administration orders were made in respect of the principal companies in the Group in this court on 26 February 1995, and they have subsequently been placed in compulsory insolvent liquidation. The claimants are three of the Barings Group companies acting by their liquidators. The first is Barings plc (“Plc”), the holding company of the Barings Group. The second is Bishopscourt (BS) Ltd, at all material times called BSL, a company incorporated in the Cayman Islands but based in London and registered as an overseas company in England & Wales. BSL was the securities trading arm of the Barings Group, and was the indirect subsidiary of Plc through Baring Brothers & Co Ltd (“BB&Co”), the banking subsidiary of the Group. The third claimant is BFS. BFS was incorporated in Singapore and was the indirect subsidiary of BSL. It was the futures trading company of the Group on the Singapore International Monetary Exchange (“SIMEX”). It was by means of an account in the name of BFS that Leeson entered into a number of transactions in financial derivatives, the losses sustained from which caused the Barings Group to collapse.

The court has been concerned with two main actions. The first is that in which Plc and BSL are claimants, “the Plc action”, commenced by two writs issued and served in 1996 and 1998 which have been consolidated. In the Plc action the claimants claimed against the partners in the accountancy firm known as D&T, carrying on business in Singapore, part of a world-wide group of partnerships known as Deloitte & Touche. The claim against them is in negligence arising out of their audit of the Barings Group consolidation balances of BFS and BFS statutory accounts for the year to 30 September 1992 (“the 1992 audit”) and for the 15 month period to 31 December 1993 (“the 1993 audit”). In summary it is alleged in the claim that in breach of duty the two audits were so conducted as to fail to uncover the unauthorised trading of Leeson. The claim is for the consequential damage suffered.

In the Plc action the claimants also claimed on the same grounds similar relief against Coopers & Lybrand (London), the London-based accountants’ partnership (“C&LL”) and Coopers & Lybrand (Singapore) (“C&LS”) (together “the Coopers firms”). The breach of duty alleged against C&LS arose from their audit of consolidation schedules and statutory accounts of BFS for the year to 31 December 1994 (“the 1994 audit”). The breach of duty alleged against C&LL arose from their audit of Plc’s and BSL’s statutory accounts for 1992, 1993 and 1994 and of the Group accounts for those years. Included in BSL’s claim by assignment was a claim by Baring Securities (London) Limited (“BSLL”), which took over BSL’s proprietary securities trading operations during 1993, in respect of C&LL’s audit of its accounts for 1993 and 1994.

In the Plc action D&T counterclaim against Plc and BSL for deceit and breach of duty, seeking damages for any amount that they are adjudged liable to pay to Plc or BSL or to BFS. D&T have also issued third party proceedings against ING Baring Securities (Japan) Ltd (at all material times called Baring Securities (Japan) Ltd (“BSJ”)) and BFS, claiming contribution to any liability to Plc and BSL in respect of BSJ’s and BFS’ breach of duties owed to Plc and BSL. BSJ was the Japanese securities trading arm of the Barings Group which dealt in securities and, in particular, futures and options on the Tokyo and Osaka markets. C&LL issued third party proceedings against nine individual Barings officers, all of which proceedings have been settled or discontinued. C&LS issued third party proceedings against BFS.

The claimant in the second set of proceedings (“the BFS action”) is BFS. In those proceedings BFS claims similar relief to that claimed in the Plc action on similar grounds against D&T in respect of the 1992 and 1993 audits of its statutory accounts and against C&LS in respect of the 1994 audit of those accounts, though the damage is differently quantified. In those proceedings D&T brought third party claims or Part 20 claims against BSJ, Plc and BSL for contribution arising from breaches of duties owed by those parties to BFS. C&LS brought third party claims on similar grounds against Plc and BSL.

The trial of the combined action began on 2 October 2001. During its second week, it was announced that the claimants in both actions had agreed terms to settle their claims against the Coopers firms. As a result I adjourned the proceedings on 9 October 2001, pending approval of the settlement by creditors and the Companies Court. Both approvals were in due course given.

The settlement of the claims against the Coopers firms had a number of important consequences for the claims brought against D&T:

D&T’s response to the settlement involving the other parties was threefold. First, D&T brought Part 20 proceedings against the Coopers firms, which is the reason for their continued participation in this trial. Secondly, D&T applied in the Plc/BSL action to strike out Plc’s and BSL’s claims against them.

I heard the strike-out application between 23 October and 2 November 2001 and handed down my judgment on 23 November 2001. In it, I struck out Plc’s and BSL’s claims against D&T on the grounds that (i) neither claimant had pleaded that D&T had in contemplation, when giving their audit opinions, the transactions as a result of which that claimant suffered loss; (ii) BSL’s claims in relation to the Dollar Funding were also barred by the rule in Johnson v Gore Wood [2001] 2 WLR 72; and (iii) Plc’s claims for bonuses paid out failed because Plc did not meet their cost. Accordingly I dismissed Plc’s and BSL’s claims against D&T.

Plc and BSL have obtained leave to appeal against that decision but have not pursued their appeal pending the outcome of this trial.

The third limb of D&T’s response to the settlement was to apply, successfully, for the hearing of a preliminary issue. D&T contended that they had a complete defence to BFS’ claim, based on circuity of action and/or set-off. They said that this arose as a result of two representation letters, written to them by Mr Jones in the course of the 1992 and 1993 audits, which D&T alleged were recklessly fraudulent.

I heard the preliminary issue between 28 January and 8 March 2002, hearing evidence from Mr Mah, Mr Jones and Mr Mason (Mr Bax’s evidence on the issue was not challenged and so he was not called for cross-examination). References to the oral evidence of Mr Jones in this judgment are to his evidence given in the course of that hearing. He has not been called as a witness in the main trial.

I handed down my judgment on the preliminary issue on 20 March 2002. I concluded on the evidence adduced on the application that Mr Jones had not been recklessly fraudulent in signing the representation letters and therefore D&T did not have the complete defence for which they contended.

The trial recommenced on 7 May 2002. Opening submissions occupied a further 30 hearing days, ending on 27 June 2002. I heard 55 days of factual and expert evidence, ending on 12 December 2002. Closing submissions lasted from 17 February to 24 March 2003.

The time taken during the trial for witness evidence and closing submissions was reduced by an agreed order I made on 23 September 2002, hiving off to a later date D&T’s contribution claims against C&LL and C&LS. This order was made after opening submissions but before any witnesses for the Coopers firms had been called. The effect of the order was that in this judgment, which concludes Phase 1, I determine the liability of D&T to BFS, the extent of BFS’ contributory negligence, and whether in principle Plc, BSL and BSJ are liable to contribute to D&T’s liability. Whether the Coopers firms should contribute to that liability and the actual apportionment between all contributing parties is to be decided, if necessary, in Phase 2. The thinking was that Phase 2 would not be necessary if I were to decide Phase 1 entirely in favour of D&T, in which event the hive-off would have saved several weeks of court time.

This means that in the trial so far I have not heard the witnesses called by C&LL or C&LS, nor any evidence from D&T witnesses relating to D&T’s claims against them. Counsel for the Coopers firms did cross-examine those witnesses called in Phase 1 whom they felt had given evidence which was relevant to D&T’s contribution claims against them. They also addressed me in closing submissions on issues of fact or law that had arisen in Phase 1 and might impact on the claims against them. However I do not deal in this judgment with those issues or otherwise with the allegations made against the Coopers firms, save tangentially in relation to one of the issues relating to damages.

The facts surrounding the collapse of the Barings Group of companies have been exhaustively investigated already. The primary investigations were those of the report of the Board of Banking Supervision to the House of Commons which reported on the 13 July 1995 (“The BoBS Report”) and the report of inspectors to the Minister of Finance in Singapore into the collapse of BFS (“the SIR report”), also published in 1995. In addition the facts have been investigated by the Securities and Futures Authority, a committee of the House of Commons, the Accountants’ Joint Disciplinary Tribunal, the Public Accountants Board of Singapore and in the course of interviews of a number of Barings directors and employees by the Administrators of the claimants. Company Directors’ Disqualification Act (“CDDA”) proceedings were taken against various Barings directors, some of which were compromised and some defended. In respect of defended cases, Jonathan Parker J (as he then was) gave judgment on 1 December 1998 in a judgment reported as Re Barings plc (No. 5) [1999] 1 BCLC 433. His judgment was upheld in the Court of Appeal on 25 February 2000. Extensive materials from all these reports, investigations and hearings were before this court.

This factual description is broadly chronological but, in order to assist the reader, I attach as Schedule A to this judgment a brief chronology of the most important events described below. At Schedule B is a list of the abbreviations used, and descriptions of the persons mentioned, in this judgment. At Schedule J is a list of the exchange rates used in the proceedings.

A structure chart of the Barings group is attached to this judgment as Schedule C. I describe below the relevant companies.

Plc was the holding company of the Barings group. It was based in London and did not trade on its own account. Its operating subsidiaries (held via Baring Group Holdings Limited) were BB&Co, a merchant bank specialising in corporate finance and debt trading, and Baring Asset Management Limited (“BAM”), which provided fund and asset management services.

In 1984 BB&Co acquired the Far Eastern stockbroking business of Henderson Crosthwaite, headed by Christopher Heath. This became BSL. It was Barings’ first significant involvement in the securities broking business.

BSL was until the second half of 1992 run by Christopher Heath (chairman and chief executive officer) and Andrew Baylis (deputy chairman), quite separately from BB&Co’s banking business. BSL conducted securities and futures trading, both on behalf of customers and on its own account. It had strong positions in the Asian equity markets, particularly in Japan. In the late 1980s it made significant profits dealing in Japanese equity warrants, and it opened new offices and expanded into new markets. However its management structure did not keep up with its expansion and remained very limited.

As explained below (see paragraph 39 and following), BSL underwent radical changes from 1992 onwards under the leadership of Peter Norris. These culminated in BSL’s consolidation with BB&Co (described below) into a new, non-statutory entity. By so doing, these companies obtained certain regulatory advantages which saved costs. The resulting combined “company” was known as Baring Investment Bank (“BIB”).

From 1 November 1993, as a consequence of this consolidation of BB&Co and BSL, BSL used a subsidiary, BSLL, as the vehicle for its proprietary trading. Proprietary, or house, trading is trading on behalf of a Group company as principal. Customer, or agency, trading is trading on behalf of third party customers.

BSJ was BSL’s securities trading subsidiary in Japan. Its activities included:

BSJ transacted business in Japan on the Tokyo and Osaka exchanges.

BSS was BSL’s stockbroking (indirect) subsidiary in Singapore, trading securities listed on the Singapore and Malaysian stock exchanges. It was set up in 1987 by James Bax, who later became managing director of BFS and Regional Managing Director, Asia Pacific, for BSL and then BIB. Initially it used a local broker to carry out the trading but carried out its own settlements. Mr Bax recruited Simon Jones from Associated Merchandising Company in 1989 to take control of the settlement and funding process. BSS obtained membership of the Singapore Stock Exchange in October 1992, allowing it to trade directly on that Exchange (Footnote: ⁵).

At the time of the crash BSS employed 115 staff (Footnote: ⁶).

BFS was BSL’s (indirect) subsidiary and broker in Singapore for the transaction of futures and options business in Singapore on SIMEX. It was active from 1 June 1992 until the collapse in February 1995. Its history and activities are described at paragraph 163 and following below.

At the time of the crash BFS employed 23 staff (Footnote: ⁷). Its apparent turnover for the year ended December 1994 was approximately S$26.5m.

In 1992 BSL’s business suffered a severe downturn. BB&Co management asked Peter Norris to carry out a review of BSL’s operations. Implementation of this review resulted in the replacement of the existing management and the introduction of new management control systems.

In pursuance of the former step, Mr Norris took over as chief operating officer of BSL in September 1992, with Mr Heath stepping down as chief executive officer and becoming chairman only. In March 1993, Mr Heath and a number of other senior executives (including the entire management of the derivatives business (Footnote: ⁹)) were asked to leave the Barings Group altogether, and Mr Norris became chief executive officer of BSL.

As part of Mr Norris’ improvement of BSL’s reporting and control systems, he introduced from the end of 1992 (Footnote: ¹⁰) the concept of matrix management. This was a management structure under which each type of trading business activity had a global head, responsible for the conduct of that activity throughout the world, and across the various companies in the Group. A trader reported to the head of his particular trading activity (“his product head”), via a product reporting line. However (in theory at least) he also reported to local office management, which was responsible for local office infrastructure, including computer systems, controls and settlements (Footnote: ¹¹). This latter group of functions was referred to from time to time as “operations”.

Of the operations functions, settlements seems, at least from January 1994, to have had a separate reporting line. After that date, the head of settlements in a subsidiary reported up a reporting line to Mr Gamby, the head of settlements at BSL and later BIB, who in turn reported to Mr Barnett (Chief Operating Officer at BIB). I should express two qualifications to this. Firstly, Mr Gamby denied, before the Singapore inspectors and in the CDDA proceedings, any responsibility for settlements at BFS. This denial was rejected by Jonathan Parker J in Re Barings plc (No. 5). Secondly, Mr Bowser denied before me that, while he was BSL Derivatives Controller and so head of Settlements at BSL between March 1992 and November 1993, he had any responsibility for settlements outside BSL. I consider this further at paragraph 1004 below.

The matrix management system was designed to transcend corporate boundaries to unify the management of each type of business worldwide. Thus an aspect of the business of trading subsidiary A might be managed by an employee or director of subsidiary B reporting to an employee or director of subsidiary C, and ultimately to the appropriate product or operational head.

In late 1992 Barings management decided that the investment banking businesses of BB&Co and BSL should be combined into a single entity. As a first step towards such consolidation, BB&Co and BSL were “solo-consolidated” from November 1993. This meant that BSL was treated for regulatory purposes as a branch of BB&Co, and had to be wholly funded by BB&Co. Thereafter, funds advanced to BSL by BB&Co would not fall within the large exposure reporting requirements and maximum exposure limits of the Bank of England, so removing constraints on BB&Co’s ability to fund BSL.

In December 1993 Barings management took the next step towards consolidation, by establishing a new umbrella organisation comprising the merchant banking and securities businesses of BB&Co and BSL, which came to be known as BIB. BIB’s structure was put in place over the period May to November 1994. It was not an incorporated body, although from October 1994 it had titular directors.

BIB was divided into four main groups: Equity Broking and Trading (including the agency derivatives business), Banking (including the Financial Products Group, described below, and therefore the proprietary derivatives business), Corporate Finance, and Emerging Markets Corporate Finance. Mr Norris became Chief Executive of BIB. The Head of the Banking Group was George Maclean. The Deputy Head of the Banking Group was Ian Hopkins.

The consolidation was not complete by the time of the collapse, at which time BB&Co and BSL still remained separate companies operating in different offices, but there was a common BIB management structure and treasury.

The management structure within the Barings Group consisted primarily of various managerial committees, with the statutory boards of directors of companies within the Group performing only a limited function. Thus the board of Plc (under the chairmanship of Mr Peter Baring) met only infrequently. The committee structure within Barings is of limited relevance to this judgment, but in outline it was as follows (Footnote: ¹²):

Management decisions in Plc were taken by an executive committee of the board known as EXCO. EXCO met weekly, and was the senior committee of the Group.

Below EXCO in the Group management structure were the respective management committees of BB&Co and BSL. The BB&Co Management Committee played an active role in dealing with strategic and operational matters; the BSL Management Committee fulfilled a much more limited role.

From May 1994, the BB&Co Management Committee ceased to function and was merged into MANCO. The BSL Management Committee continued to exist until the time of the collapse, but its activities were confined to Equity Broking and Trading.

MANCO met weekly, on Mondays, and was the principal forum for management discussion and executive decision-making within BIB.

Prior to the formation of ALCO in November 1994 (as to which, see below), the control of market risk in BB&Co was the responsibility of its Treasury and Trading Committee (“the Treasury Committee”). Membership of the Treasury Committee as at 31 December 1993 included Mr Maclean (Chairman), Mr Hopkins, Mr Baker and Mr Hawes. The Treasury Committee met daily.

Within BSL, a similar function to that of the Treasury Committee was performed by BSL’s Risk Committee (“the Risk Committee”). Its role was setting limits for, and monitoring the results of, trading and sales activities which resulted in market and credit risk. The members of the Risk Committee included Mr Norris, Mr Hawes (Head of Treasury), and Mr Broadhurst (Finance Director). The Risk Committee met daily. Once a week it discussed credit issues, and on those occasions Mr Maclean and Mr Hopkins attended. As from 1 January 1994 either Mr Baker or Ms Walz attended meetings.

In November 1994, as part of the process of forming BIB, the Treasury Committee (of BB&Co) and the Risk Committee (of BSL) ceased to exist and were replaced by ALCO. ALCO met daily. Its members included Mr Norris, Mr Baker, Mr Hopkins and Mr Hawes.

In August 1994 GTR was formed, headed by Mr Hopkins. GTR combined the Treasury, Risk and Credit units of BIB.

In 1992 BB&Co established the Debt Financial Products group, to set up a debt trading operation. It was headed by Mr Ron Baker, who joined Barings in April 1992, from Bankers Trust International. Ms Mary Walz, who had been one of his colleagues at Bankers Trust, joined Barings at about the same time to work with Mr Baker. Mr Baker was very successful in establishing that operation and acquired a high reputation in BB&Co.

BSL had conducted agency and proprietary trading in derivatives since before 1990. Following the departure of Mr Heath and his colleagues, Mr Norris took charge of the derivatives business, with Mr Oades in a co-ordinating role (Footnote: ¹³). However this was only a short-term solution, and in mid-1993 Mr Norris asked Mr Baker to take charge of the derivatives business. According to Mr Norris, Mr Baker was unwilling to take on the agency side but agreed to take on the proprietary trading activities. He carried out an information-gathering visit to his new empire, including Tokyo, in November and December 1993 (Footnote: ¹⁴).

Thus FPG (Footnote: ¹⁵) was established as of 1 January 1994, headed by Mr Baker. It covered his old BB&Co debt operation but also all BSL’s proprietary derivatives businesses, including those carried on in Japan and Singapore. These latter were included within a sub-group, the Structured Products Group (“SPG”), headed by Ms Walz, who reported to Mr Baker. Leeson was at all material times General Manager of BFS, part of Barings’ derivatives trading business. So from 1 January 1994, under the matrix management system described above, Mr Baker was Leeson’s product manager in relation to Leeson’s proprietary trading activities.

Whilst the proprietary derivatives businesses of BSJ and BFS fell within FPG, until 1 January 1995 their agency businesses remained within a different group, Futures and Options Sales (“F&O Sales”), the head of which (as product manager) was Mr Killian. However, according to Mr Norris, Mr Baker realised during early 1994 the advantages to be gained from running the agency and proprietary businesses together. Despite some resistance from Mr Killian, on 1 January 1995 the agency business was also included in FPG, under Mr Baker.

Hence prior to 1 January 1995 the product manager in relation to Leeson’s agency trading activities was Mr Killian; thereafter until the collapse it was Mr Baker (Footnote: ¹⁶).

BB&Co had a long-established internal audit function, which by 1994 was headed by Ash Lewis. BSL’s Internal Audit Department was not formed until late 1992. By 1994 it consisted of three auditors, headed by Ian Manson. One of them was James Baker, who gave evidence before me. The last internal audit which the Department conducted was of the South Asian Region (including BFS in Singapore). That took place in mid-1994 and is described below (see paragraph 441). Immediately afterwards, it was amalgamated into a single BIB Internal Audit Department, which was formed under Ms Lewis.

Attached as Schedule B is a list of Dramatis Personae of the people and companies mentioned in this judgment, together where appropriate with the abbreviations used to describe them. However it may be helpful to describe here the individuals who feature most importantly, with their career details and relevant reporting lines. A chart of the reporting lines is attached as Schedule D.

An asterisk in the list below denotes that the individual gave evidence before me. The convention adopted during the trial was that witness statements of witnesses who were not called to give evidence could not be relied upon by the party which put in the statement, but could be used by other parties (Footnote: ¹⁷).

Leeson himself was not called as a witness, though BFS and D&T had both put in brief witness statements from him. I also had available transcripts of the interviews he gave to the Serious Fraud Office (“SFO”) while awaiting extradition from Germany to Singapore in August and September 1995, the transcript of a BBC interview and his book “Rogue Trader”.

Peter Norris*: Director BB&Co, running Asian corporate finance from Hong Kong 1987-1992. Chief Operating Officer, BSL, September 1992-March 1993. Chief Executive Officer, BSL, March 1993-November 1994. Chief Executive Officer, BIB, November 1994-February 1995, reporting to Andrew Tuckey, who became the Chairman of BIB. He was disqualified in the Company Directors’ Disqualification Act (“CDDA”) proceedings consequent upon the collapse of the Barings Group from being concerned in the management of a company for four years.

Ron Baker: Director, BB&Co, April 1992, running Debt Financial Products group (having joined from Bankers Trust). Head of FPG (BB&Co, then BIB), 1 January 1994-1995. Responsible for Japanese/Singapore proprietary business from January 1994, and for agency business from January 1995. Until August 1994, he reported to Ian Hopkins, Head of Treasury and Trading at BB&Co, and thereafter direct to George Maclean. He was disqualified in the CDDA proceedings for six years.

Mary Walz: Director, BB&Co, July 1992-December 1993, under Mr Baker in the Debt Financial Products Group. She headed SPG, later renamed Equity Financial Products, which included equity products, within FPG, January 1994-1995, reporting to Mr Baker. She was disqualified in the CDDA proceedings for 2 years.

Tony Gamby: joined BB&Co October 1979. Head of Settlements, BB&Co, not responsible for BSL, until January 1994, when he became Settlements Director, BIB, and director of BB&Co. He then had the Heads of Futures and Options Settlements, Equities Settlements and Derivatives Settlements reporting to him. He was disqualified in the CDDA proceedings for five years.

Gordon Bowser*: Derivatives Controller, BSL, responsible for managing the futures and options settlements department, March 1992-November 1993, reporting to Mr Norris, having joined from the Merrill Lynch swaps derivatives department. Manager of costs and budget issues, FPG, November 1993-August 1994. Risk Manager North Asia in Hong Kong, August 1994-February 1995.

Brenda Granger: Manager London Futures and Options Settlements, BSL, July 1992-4, reporting to Mr Bowser. Manager Futures and Options Settlements, BIB, January 1994-1995, reporting to Mr Gamby.

Tony Hawes*: Director, BB&Co, 1987-end 1993. Deputy Head of Treasury, BB&Co, 1987-1993, reporting to Mr Hopkins. Seconded to BSL in October 1992 to work on solo consolidation. Treasurer, BSL, reporting to Mr Broadhurst, May 1993-August 1994. Group Treasurer, BIB, August 1994-February 1995, reporting to Mr Hopkins. He was disqualified in the CDDA proceedings for five years.

Geoffrey Broadhurst: accountant, Deloitte Haskins & Sells, 1983-1986. Financial Controller, BB&Co, 1987-February 1992. Finance Director, BSL, February 1992-March 1993. Group Finance Director, BSL, March 1993-August 1994, reporting to Mr Norris. Group Finance Director, BIB, August 1994-February 1995, reporting to Mr Barnett. He was disqualified in the CDDA proceedings for three years.

Mike Killian*: became General Manager, BSJ F&O Sales, in September 1988 (having joined from Chase Manhattan), with director status, but not a director of BSJ (Footnote: ¹⁸). Director BSL May 1992-January 1993. Director Baring Securities (Far East) Ltd April 1991-March 1994 (Footnote: ¹⁹). Director, BFS, December 1992-1995. Head of Pacific Rim agency group, in June 1993, reporting to Mr Norris. Head of Global Equity F&O Sales, BIB, in 1994. From January 1995 he reported to Mr Baker. Based in San Francisco from 1 April 1994.

Fernando Gueler*: joined BSJ in April 1988 (initially part-time) as a research analyst. Assistant to Mr Killian, summer 1989-early 1990. Arbitrage trader in warrants with Richard Johnston in Tokyo, early 1990-1991. Index arbitrage trader, 1991-1995. Head of derivatives traders, BSJ, April 1993. Head of FPG in Tokyo, 1 January 1995. Not a director of BSJ.

James Bax*: Managing Director of BSS, August 1987-October 1994 (having joined from Hoare Govett), reporting to Andrew Fraser (Deputy Chairman, Asia, of BSL). Managing Director of BFS, June 1988-February 1995. Director, BSL, 1989-1995. Regional Managing Director, Asia Pacific, for BSL and then BIB, early 1994-February 1995, reporting (in his operations capacity at least) to Mr Barnett. He was disqualified in the CDDA proceedings for four years.

Simon Jones*: accountant Deloitte Haskins & Sells, 1980-1982. Regional Finance Director, and then Divisional Vice-President, Associated Merchandising Company, 1982-1989. Finance Director of BSS, February 1990-February 1995, reporting to Mr Bax and then Mr Marais. Director, BFS, May 1990-February 1995. Director of Operations, South Asia Region, BSL and then BIB, early 1994-February 1995, reporting to Mr Bax.

Rachel Yong: Financial Manager at BSS and BFS, January 1993-February 1995, reporting to Mr Jones.

Nick Leeson: joined BB&Co from Morgan Stanley in July 1989 to work in Futures and Options Settlements. Member of the BSL Business Development Group, May 1991. Manager BFS, March 1992. General Manager BFS, June 1993-February 1995.

The method by which Barings staff were remunerated had an important impact upon the events which led to the collapse of the Barings group. In each year of account, half of the group profit (for 1993, as I held in the strike-out application, the Plc and BSL groups were treated separately) was set aside for the payment of bonuses to staff over and above their nominal salaries. From this fund there would be allotted to each business group, by senior management, a part of the fund, the size of which depended upon the relative success of that group in generating revenue in that year. This in turn would then be shared out among the members of that group in accordance with an assessment of the contribution of each to the success of the group.

Typically the remuneration of a member of a successful group would primarily consist of this bonus, which might be several times the size of his nominal salary. It will be seen that this system of remuneration had a corrosive effect on Barings management generally. It gave rise to competition between groups to include among their number particular “star” earners; obstruction of the free flow of information about trading between groups, and between groups and senior management; and the pampering and protection of “stars”, in particular Leeson, for fear that they might take their services elsewhere.

It is necessary to give some basic explanation of the futures and options trading conducted by BFS on SIMEX (Footnote: ²⁰), and of the products traded. A graphical walkthrough of BFS’ trading on SIMEX and its reporting procedures is attached at Schedule E.

SIMEX was an open outcry market, with a separate pit for each commodity traded. Only brokers who were members of the exchange could trade on SIMEX, using authorised staff who had passed a (not very difficult) (Footnote: ²¹) aptitude examination set by the exchange.

Trading was conducted by traders on the floor of the pit calling out and signalling to each other. A clearing member therefore had a team of front office staff, who worked on the floor of the exchange. They consisted of order fillers (who conducted the trading), linespersons (who took orders from customers and relayed them to the order fillers) and booth clerks (who handled administrative matters). At the member’s office was a separate team of back office staff, who handled settlement of trades with SIMEX and with customers.

Derivatives: a derivative is a contract or instrument which changes in value depending on price movements in another instrument or in an index. Futures contracts and options are derivatives.

Futures contracts: a financial futures contract is an agreement to buy or sell a standard quantity of a specific financial instrument at a predetermined future date and an agreed price. The main contracts traded on SIMEX were:

Options: an exchange-traded option gives the purchaser the right, but not the obligation, to buy (a “call option”) or sell (a “put option”) an underlying financial futures contract at a predetermined “strike price” on or before a specified future date. In return, the purchaser pays a premium to the seller (or writer) of the option.

Equivalents of all these contracts could also be traded on (and of course they originated from) the Japanese exchanges at Tokyo and Osaka. There were differences in the terms of the contracts between Singapore and Japan: for example, a Nikkei 225 tradeable futures unit on Osaka was defined as 1000 times the Nikkei 225 price, whereas a unit on SIMEX was only 500 times that price, so SIMEX units were half the size of Osaka units. A “tick”, which was the basic amount of the movement up or down in the price of a futures contract, was worth four times as much on the Osaka market as on SIMEX (¥2,500 on SIMEX and ¥10,000 on Osaka). Also the margining requirements of the Japanese exchanges were more onerous than those of SIMEX. This was a major reason for the move of business from the Japanese exchanges to SIMEX which was under way from 1992 onwards.

When a futures or options contract was traded on SIMEX, the exchange interposed itself between the parties. Thus a broker such as BFS, which was a clearing member of SIMEX, when entering into such a contract on behalf of a customer, contracted as a principal with SIMEX. In turn, SIMEX as principal contracted with the counterparty’s broker. Therefore, subject to any right to be indemnified by its customer, the broker was individually liable for due performance of that contract. Such was, of course, also the position where the broker contracted on its own behalf or on behalf of a related company, i.e. when trading “in-house”.

Without some form of protection, SIMEX would be exposed to loss were a party to a contract made on the exchange to fail to honour its bargain, since SIMEX would be liable to the counterparty. SIMEX (like all such exchanges) protected itself against such loss by requiring clearing members to pay to it “margin”, when entering into a futures contract, and “settlement variation”, to reflect movements in the market which increased the member’s exposure to loss in relation to that contract. As discussed below, SIMEX protected itself against market movements affecting exposure on options slightly differently.

SIMEX required every member to enter into a direct debit arrangement with its bank. This allowed SIMEX to withdraw from the member’s bank account sufficient funds to meet the member’s margin and settlement variation without action from the member. SIMEX did this early each morning, in respect of the margin and settlement variation requirement generated by the previous day’s trading. Pursuant to this rule, BFS instructed Citibank to allow SIMEX to withdraw funds as necessary from BFS’ accounts with Citibank.

SIMEX held margin pending the maturity or closing-out of the contract. Settlement variation might be increased in amount or repaid to the member, depending on movements in the market from day to day.

SIMEX rules required both that (a) a member of the exchange should post margin with SIMEX in relation to all positions held overnight and (b), where the member held a position on behalf of a customer, that customer should deposit margin with the member. Futures contracts and options were treated somewhat differently.

SIMEX required a member of SIMEX to deposit with SIMEX:

Initial and maintenance margins for futures were fixed by SIMEX at a set level per contract, and did not vary with the market (Footnote: ²⁵). Margin to be deposited with SIMEX had to be deposited with SIMEX by the start of the next trading day.

SIMEX rules required that each member’s customers should deposit with the member initial margin in respect of each trade entered into by the member on the customer’s behalf (Footnote: ²⁶). This had to be deposited within three business days of the trade date (Footnote: ²⁷). Initial margin was 25 per cent higher than the maintenance margin which members had to deposit with SIMEX.

If the initial margin deposited with the member fell below the level of maintenance margin referred to above, the member had to ensure that the customer topped up his margin to the initial margin level, again within three business days. However BFS adopted a more conservative practice, requiring customers to maintain margin at the initial level throughout (Footnote: ²⁸).

Thus the initial margin which a member obtained from customers was higher than the maintenance margin deposited with SIMEX. It follows that the margin paid or owed to a member by its customers should have exceeded the margin which the member had to deposit with the exchange in relation to its customers’ trades by about 25%. Therefore, once the customer had paid his margin in full, the member would hold a surplus.

SIMEX revalued, or “marked-to-market”, each futures contract open at the end of a trading day and calculated the member’s potential profit or loss on it as a result of market movements. The profits and losses on all open contracts were aggregated. If the market had moved adversely to the member’s positions overall, the member was required to pay the net loss as “settlement variation” (sometimes referred to as variation margin). This protected SIMEX against the member’s increased exposure to loss. If the market had moved in favour of the member’s positions overall, its settlement variation requirement would reduce, reflecting the member’s reduced exposure to loss. In that event SIMEX was required (on demand by the member) to repay the excess settlement variation to the member.

While a member’s contract with its customer no doubt required him to put up margin when required, unpaid margin was not treated in a member’s accounts as a debt owed by the customer. Rather, margin was security for the customer’s trading. Once paid, it was treated as a debt owed by the member to the customer, until consumed by realised losses or a call for settlement variation. Settlement variation, which amounted to notional realised profits or losses on the customer’s positions, did create a debt from customer to member, which the member could meet by using the customer’s surplus margin (if any) (Footnote: ²⁹).

Similarly margin deposited by a member with SIMEX was treated as security for the member’s trading on SIMEX, and therefore as a debt owed by SIMEX to the member. Indeed SIMEX paid interest on the member’s balances, for which in the normal way a member accounted to its customers (it is relevant in this case that BFS accounted to BSJ and BNP for interest on their margins which it received from SIMEX. By contrast, as a result of some internal understanding of which I have seen no direct evidence, it did not account to BSL for such interest until October 1994 (Footnote: ³⁰)). Again, realised losses or settlement variation created a debt from the member to SIMEX, which SIMEX could use the member’s margin surplus to meet.

Only the seller of an option had to deposit margin. The buyer had to pay the premium but thereafter incurred no further risk.

Unlike its rules as to futures, SIMEX did not fix in advance the margin on options. Instead, it calculated the margin on options using a software package, the Standard Portfolio Analysis of Risk margining system (“SPAN”). A member could buy the SPAN software and download from Compuserve or the internet risk arrays published by SIMEX. The member could then feed into its computer details of its positions and the software would calculate the margin due on them.

SPAN’s calculation took into account two elements. Firstly, it effectively marked the option to market, calculating the loss or profit if the option were closed out immediately. Secondly, it added a risk element, to cover the largest loss the position was reasonably likely to incur as a result of market movements over the next 24 hours. As a result, the initial margin on the sale of an option was usually equal to or more than the premium received (Footnote: ³²). The member selling the option had to pay the margin to SIMEX on the day after the option was sold, whereas the member did not receive the premium until the day after that (Footnote: ³³).

Margin was calculated daily on the member’s whole option portfolio and the margin required on each option varied with movements in the market. SIMEX demanded further margin, or repaid that held, as the market moved against or in favour of the member’s position. Therefore margin fulfilled the function of settlement variation for futures, and SIMEX did not levy settlement variation on options (Footnote: ³⁴).

SIMEX required its members to use SPAN in calculating the margin they demanded from customers for option trades. I explain at paragraph 116 below how BFS’ accounting system treated options in its customer accounts.

Each member had two accounts with SIMEX, a house account (for trading on its own account or that of related companies) and a customer account (for trading on behalf of unrelated third parties). BFS’ proprietary trading for BSL or BSJ should therefore have been recorded in BFS’ house account (Footnote: ³⁵). In fact some of Leeson’s arbitrage activity conducted on behalf of BSJ, known as “switching” (see paragraph 212 below), was recorded in BFS’ customer account (Footnote: ³⁶). On the other hand, agency trades for BSL’s customers should have been recorded in the customer account (as they were, but so, wrongly, were trades on the 88888 account, to which Leeson booked his unauthorised trades).

A member maintained accounts in its own accounting system on behalf of customers. BFS only had four customers, BSL, BSLL, BSJ and (from October 1992) Banque Nationale de Paris in Tokyo (“BNP”) (Footnote: ³⁷), but it maintained a large number of sub-accounts, one for each of BSL’s and BSJ’s own customers.

As part of the task of establishing BFS’ office in early 1992, Leeson and Mr Dickel selected a computer system to maintain BFS’ day-to-day trading and accounting records. On the recommendation of Mr Killian (Footnote: ³⁹), and with the concurrence of Mr Jones and Mr Bowser (who visited Singapore in April 1992), they chose a system known as CONTAC. This was a popular system among Singapore derivatives traders and Mr Killian had used it at Chase Manhattan. CONTAC was used only to service BFS’ activities and was maintained wholly by BFS’ settlements staff.

At the close of each trading day, BFS received from SIMEX:

CONTAC generated a number of reports, including a daily activity statement for each account. This showed the trades done that day for the account, the cumulative futures and options positions on the account and the realised and unrealised profits and losses. The daily activity statement ended with a summary which calculated the equity balance on the account and the margin due from or to the customer, as explained below. The last page of a sample daily activity statement is attached as Schedule I to this judgment (Footnote: ⁴⁰). A “walkthrough” of BFS’ systems, from customer order to reporting to the customer is attached at Schedule E (Footnote: ⁴¹).

BFS’ financial and statutory accounting records were maintained on the separate SUN system, in common with the other Barings Singapore companies. At the end of each month, BFS staff updated the SUN system by manually transferring relevant trading balances from CONTAC to SUN.

The Daily Activity Statement produced by CONTAC for each account ended in a calculation of an equity balance and the margin due from or to the customer. A sample of such a calculation is at Schedule I. The elements of the calculation were as follows:

When BFS entered into a futures contract for a customer, the initial margin required went into (ix) on the trade date as margin required. When the customer paid that margin, it was entered into (ii) as cash received from the customer. That increased the equity balance in (vii). So, other things being equal, it balanced out the margin requirement in (ix), resulting in a zero margin call.

SIMEX marked any futures position to market, giving rise to an unrealised profit or loss, on the trade date and thereafter until the position was closed out or matured (Footnote: ⁴³). As such, this was entered into (vi). It thus increased or decreased the equity balance in (vii). If a loss decreased the equity balance, since the margin requirement in (ix) stayed constant, the result would be a margin deficiency in (x) and a margin call. By contrast, if a gain increased the equity balance, the result would be a margin excess to be repaid to the client. A margin call or repayment, when paid by or to the customer, resulted in a credit or debit to (ii) (cash paid/received). This balanced out the unrealised profit or loss resulting from the marking-to-market, returning the margin excess/deficiency in (x) to zero.

Where a customer bought or sold an option, the daily activity statement for the trade date showed the premium on the sale of an option in line (iii) as a realised profit (for the seller of the option) or loss (for the buyer).

If a buyer of an option exercised it at a profit, at that point his profit was reflected as a realised profit in line (iii). This increased his equity balance and, other things being equal, triggered a repayment to him of an equivalent sum from the gross accumulated cash held by the broker on the buyer’s behalf. Until that point, changes in the value of the option caused by market movements were not reflected as unrealised profit or loss in line (vi), as happened with futures positions.

However, as described at paragraphs 102 to 105 above, the seller of an option had to put up margin. This effectively marked the option to market and covered SIMEX against potential further losses over the next 24 hours. It was usually slightly more than the amount of the premium received by the seller.

As mentioned above, BFS entered the option premium at line (iii) (realised profit/loss) on the trade date, thus increasing the equity balance at line (vii). It entered the margin requirement at line (ix) (initial margin required). The result was a margin call for the amount by which the margin requirement exceeded the premium. When the customer paid the margin call, this was entered at line (ii) (cash paid or received). It therefore increased the customer’s equity balance at line (vii) and reduced the margin excess/deficiency to zero (Footnote: ⁴⁴).

As on the purchase of an option, subsequent changes in the value of the option were not reflected directly as unrealised profit or loss in the daily activity statement. Until exercise, the only effect of such changes on the financial balance was on the margin which the customer had to deposit with BFS. A fall in the value of the option increased the margin required at line (ix) and triggered a margin call. Once paid, the margin was entered into line (ii) as cash received and increased the equity balance in line (vii) (though of course it only balanced out the increased margin requirement in line (ix)).

If the option was exercised against the customer, the resulting payment to the buyer was at that stage reflected in the daily activity statement in line (iii) as a realised loss (and therefore was paid from the margin which was included in the equity balance against such an eventuality).

This treatment of options was significant in Leeson’s concealment of his fraud. Leeson conducted his unauthorised trading through a BFS account numbered 88888. In order to conceal his unauthorised trading on the 88888 account, which was consistently loss-making, Leeson needed to bring the balance on the account to nil over any relevant reporting date (balances had to be reported at month-ends, for the purpose of preparing monthly management accounts for BFS; and at accounting reference dates, auditors required confirmations by clients of any positive or negative balance shown on any account). As explained below, one means by which Leeson did this was to sell options on the 88888 account, funding the margins required out of excess margins held, BFS’ overdraft and, from February 1994, out of the Dollar Funding. On the sale of an option, the 88888 account was credited by CONTAC with the premium. But until it was exercised, that was the only effect of the sale of the option on the 88888 account equity balance. The equity balance:

Thus in the short term, Leeson could use the sale of options to reduce or cancel out the debit balance on the 88888 account. Had anyone looked at the daily or monthly activity statement for the account, he would have seen a huge and growing margin requirement at lines (ix), (x) and (xi) of the calculation described at paragraph 113 above. But the equity balance, which was the figure reflected in overall CONTAC balances, was, when it mattered, nil (enabling Leeson to avoid having the auditors require the balance on the 88888 account to be confirmed by a customer). Only if an option was exercised against Leeson did CONTAC at that point debit the loss to the account, as a realised loss.

Most of BFS’ agency trading was for accounts maintained by BSL in London, although BSJ also had some domestic Japanese customers of its own. Reduced to its simplest terms, the funding flow from BSL to BFS in relation to customer trades ought to have been as follows: -

BSL thus served as a conduit for margin between the third party customer and BFS. Where BSL had to pay margin to BFS before it received it from its customers, BSL obtained funding from the Baring Securities Group Treasury. The same was also true where a customer had deposited collateral (for example shares or gilts) with BSL which secured its trading, reducing the margin BSL called from the customer (Footnote: ⁴⁷). Subject to those cases, BSL ought to have stood in a net even position in relation to the inflow and outflow of margin between the third party customers and BFS.

BFS’ “house” trades on trading books domiciled in London were booked through BSL until late 1993 and BSLL from late 1993 onwards. House trades on Japanese books were booked through BSJ. BSL/BSLL (which were funded by Group Treasury) and BSJ (which had its own sources of funding) paid the required margin to BFS.

As observed by Jonathan Parker J in Re Barings (No. 5), any investment house needs to reconcile funding of margin requirements to known trading positions. In relation to customer trades, such a reconciliation is necessary because the broker needs to know whose customer moneys it holds, and from whom it needs to call margin and in what amounts. Further, in cases where margin is to be paid to the exchange by the broker before the requisite funds have been collected from the customer, the broker needs to be able to satisfy itself that the credit risk (i.e. the risk of the customer defaulting) is acceptable.

Further particular factors made such a reconciliation necessary in the case of BSL’s customer trades, viz:

In relation to house trades, BSL/BSLL needed to be able to calculate the cost of its house positions, and to apportion such costs accurately. Apart from the obvious requirement for this in order to manage the companies efficiently, failure to do this would risk a breach of section 221 of the Companies Act 1985 (duty to keep accounting records).

In relation both to customer and to house trades, BSL needed to be satisfied that the margin calls made by SIMEX had been correctly calculated by reference to recorded trades. And in terms of internal control, it was necessary to ensure that funds called for margin were duly applied for that purpose.

The BSL Futures and Options Settlements Department was responsible for allocating and reporting to BSL’s customers the trades conducted on their behalves by internal or external brokers, and acting as a conduit between the customers and the brokers for margins and trading proceeds in respect of those trades. Its computer system for the recording and settlement of trades was known as First Futures. When proprietary trading began to be booked to BSLL in 1993, that was also done on the First Futures system.

It is clear that for a period up to and including 1991 the Settlements Department suffered from the inadequate attention paid to operational and control matters by the Heath management. There were in 1991 serious problems in reconciling customer margins, with many large outstanding unreconciled balances. BSL had bought the First Futures system in 1989, and it took some time to develop it to the point at which it met BSL’s needs in a satisfactory manner. Thus it could not calculate customer margins on business transacted on the Japanese exchanges until mid-1992. That function had to be performed on an unreliable add-on system known as the KP System, with all trade details having to be entered a second time.

However following the recruitment of Mr Bowser and Miss Granger in spring 1992, and further development work on First Futures, these problems had been largely overcome by late 1992 (Footnote: ⁴⁹). First Futures was by then able to calculate Japanese margins and customer reconciliations were up to date. However Mr Bowser agreed that the Settlements Department remained thinly-manned and under pressure: whereas it was out of control in 1991, by the end of 1992 it was “barely in control” (Footnote: ⁵⁰).

Mr Bowser’s main requirement in being asked to approve BFS’ computer system was that it should have the capacity to send trade and margin details directly into BSL’s First Futures system by an electronic feed. This avoided BSL staff having to key in such details a second time and therefore avoided the errors which are inevitable with manual keying (Footnote: ⁵²).

Accordingly, at the end of each trading day BFS transmitted information as to the trades it had conducted and its margin requirements direct from CONTAC to First Futures. The main reports which BFS sent to London were:

This was an electronic feed containing information as to all trades conducted for BSL and BSLL (Footnote: ⁵³). As will be described, Leeson arranged for trades booked to the 88888 account to be omitted from this.

This was an electronic feed containing BFS’ calculation of the initial and maintenance margin required from each of BSL’s customers (and from BSLL) in respect of the customer’s open positions at the close of the previous trading day (Footnote: ⁵⁴). As discussed at paragraphs 150 to 152 below, BSL could not check the accuracy of this, at least for portfolios including options, since it did not have the appropriate SPAN software. However, apart from the omission of details of transactions on the 88888 account from the trade feed, there is no evidence that during the period in question the information passed by the trade feed or the margin feed was incorrect or miscalculated. In particular there was no evidence of any BSL client complaining that margin calls were incorrect.

The feed was downloaded each day into First Futures, which used the information to calculate margin calls from BSL’s customers. The feed was not used to calculate BSL’s margin payments to BFS. They were handled through the Funding Spreadsheet described below.

The margin feed included the margin required on the 88888 account (Footnote: ⁵⁵), but First Futures did not recognise 88888 as a valid account and so ignored the data. Nevertheless the daily feed, if printed out, was a short one-page document and clearly showed the large sums payable for margin to SIMEX in relation to the 88888 account (Footnote: ⁵⁶). Furthermore First Futures allowed the user to view two separate screens, one showing the margin balances received from CONTAC and one showing the accounts recognised by First Futures. However the BSL settlements clerks did not need to look at either screen when carrying out the daily download, and no one at BSL seems to have compared the two (Footnote: ⁵⁷).

This was an electronic feed containing closing prices for all SIMEX contracts. It was downloaded into First Futures and used to mark-to-market open positions.

This was also an electronic feed (Footnote: ⁵⁹). It was effectively a daily activity statement for all accounts in the names of BSL and BSLL. Leeson arranged for the 88888 account to be omitted from this report (Footnote: ⁶⁰).

The London Gross Report listed, for each BSL and BSLL account (there were seven), all trades done and positions held. For agency trades, it identified the sub-account to which each trade was booked (Footnote: ⁶¹). At the end of the listing of the trades on each account was an account summary equivalent to that on the daily activity statement (see paragraph 113 above), calculating the Equity Balance due to, and the margin to be paid by, or due to, BSL or BSLL.

However the figure for margin required was shown as a gross figure for each BSL or BSLL “omnibus” account. It was not broken down between individual customer sub-accounts (for the BSL customer accounts) or trades. There was no obvious correlation between, on the one hand, the trades listed or the figures in the account summary giving rise to the equity balance and, on the other, the margin required.

The Dollar Funding (see paragraph 325 and following below) was shown in two separate accounts, with no underlying trades listed.

This was a document prepared by BFS showing the margin due from or to BSL and BSLL. It was sent daily from BFS to BSL. It was not broken down by reference to BSL’s individual customers. Again, the Dollar Funding was shown as two separate accounts, BSL Segregated Account (USD) Collateral Account, and BSLL House Account (USD) Collateral Account, though the Funding Spreadsheet was not used to call for funds on those accounts.

The initial margin figures on the Funding Spreadsheet did not include the margin requirement on the 88888 account (Footnote: ⁶³).

BSL Settlements Department processed the Trade Feed, updating the trade data in First Futures and also checking that First Futures recognised all the account information in the feed. They then reconciled the London Gross report to the trade information in First Futures (and thus checked that First Futures had correctly processed the Trade Feed).

The Settlements Department also checked that the figures in the Funding Spreadsheet agreed with the figures for margin required in the London Gross Report. Based on these figures, they transferred margin if required from the Baring Group Treasury to Citibank, BFS’ bank, in time for the next trading day. Payments of margin to BFS were authorised by staff in the Settlements Department, usually Miss Granger and one other (Footnote: ⁶⁵).

First Futures updated BSL’s customers’ positions to reflect trades done on their behalf and market movements. It used the information in the margin feed (after combining it with margin calls from other exchanges on which the customer might have been trading through BSL) and the price feed to calculate the margin and settlement variation on each account. This enabled the Settlements Department to request any amounts due from BSL’s customers (Footnote: ⁶⁶).

The BSL Settlements Department did not themselves calculate the margin required for SIMEX transactions, but left this task wholly to BFS. This contrasts with the position for BSL’s trading on other exchanges: at least by March 1993 BSL was able to verify margin calls by BSJ and Baring Securities Hong Kong and by the external brokers it used for trades on LIFFE and LTOM (Footnote: ⁶⁷). To do this for the Hong Kong exchange, LIFFE and LTOM, BSL had installed a PC using the same SPAN software as was used by SIMEX and risk arrays downloaded from the relevant exchange over the internet. It had developed an interface between that PC and First Futures. BSL was thus able to download trade details from First Futures onto the PC, calculate the margin owed to the relevant exchange, and upload the margin calculation into the customer accounts on First Futures.

BSL could have installed such a PC-SPAN system when BFS was activated, to calculate margins on SIMEX, but it did not. According to Mr Bowser’s evidence, this was because at that time the priority for BSL’s IT department was development of a system allowing First Futures to calculate margins on the Japanese exchanges: as I have mentioned, BSJ could not calculate margins at the customer level and BSL’s existing system for calculating the margin due from individual customers was highly unreliable. BFS could and did calculate margins at the customer level. Therefore installation of SPAN software in London to calculate SIMEX margins for what was at the time expected to be a low-volume office did not seem essential. The expedient solution was to rely on BFS’ margin data without checking it. It was no doubt also relevant that at the time SIMEX, which had introduced SPAN only in March 1992, was the only exchange in the Far East to use it and BSL did not use SPAN in London until September 1992 (Footnote: ⁶⁸). Unfortunately this decision was not revisited the following year, by which time SPAN software had been installed in London and BFS’ trading volumes had grown significantly (Footnote: ⁶⁹).

As a result the Settlements Department could not check that the margin requested by the Funding Spreadsheet:

It follows from the above that there was some limited scope for Leeson to have tampered with the Funding Spreadsheet and the London Gross report so as to ask for more margin than his trades justified, without the Settlements Department noticing. Indeed, he did abuse the Funding Spreadsheet in January and February 1994, by failing to give credit for dollar collateral sent to BFS on behalf of customers (see paragraph 325 below). However after that point, when Leeson needed extra margin from February 1994 onwards to finance his unauthorised trading, he requested it, and was sent it, outside the standard system (see paragraph 330 below).

Therefore, whether or not it was reasonable for BSL to leave calculation of margin wholly to BFS, until January 1994 no loss was suffered as a result of this aspect of the standard system. If BSL had recalculated the margin on the trade data sent by BFS, it would have found BFS’ calculation to be correct. As far as is known, the allocation to individual customers was correct.

D&T pointed out that the Settlements Department could have compared the margin requested (in the Funding Spreadsheet and the London Gross report) with the margin feed. As described above, the margin feed set out all the margin requested, by customer account and sub-account, including the 88888 account (Footnote: ⁷²). This comparison would have made little sense as a routine reconciliation, as both sets of data came from BFS and it would have amounted to checking that BFS was saying the same thing twice. However once the Dollar Funding started in 1994, such a comparison would immediately have brought to light the 88888 account as the explanation for the funding requirement which so puzzled Barings personnel for over a year.

A similar, though less automated, process applied to BSJ. At the end of each trading day, BFS confirmed to BSJ, by fax or telephone, all trades carried out on the instructions of BSJ traders, whether they were booked to BSL or BSJ accounts. Trades which were for BSJ accounts were also reported to BSJ by faxing the daily activity sheets for the BSJ accounts, with a covering page indicating the excess or deficit on the accounts.

BSJ’s settlements department reconciled all trades done on BSJ’s behalf. In relation to futures, it also calculated the margin required, to verify BFS’ margin calculation. There is no clear evidence on the point, but it appears that BSJ may not have had SPAN. If so, it could not recalculate margin in relation to the options which Mr Brindle traded on SIMEX on behalf of BSJ until August 1994 (after that date, Mr Brindle’s book was moved to BSLL). BSJ then remitted the margin to BFS (Footnote: ⁷⁴).

BSJ sent to BSL a daily Risk Position Summary Report, which covered all proprietary trades carried out on SIMEX and the Japanese exchanges, and whether booked to BSL/BSLL or BSJ (Footnote: ⁷⁵).

It does not appear that Leeson, at any stage, attempted to fund his unauthorised trading by obtaining payments for margin from BSJ which exceeded those properly due. He did obtain funding from BSJ to the extent that BSJ did not require payment of balances due to BSJ and its customers. At the time of the collapse the surplus due from BFS to BSJ and not repaid was ¥20 billion (£133 million): see paragraphs 323 and 324 below.

An error account was an account maintained by a broker to deal with trades which were executed in error. For example, the broker might have bought more futures than the client instructed it to, or might have bought when the client had instructed it to sell. Once the error had been identified, the broker booked such a trade into its error account. It then closed out the trade as quickly as possible, leaving a profit or loss in the error account. The balance on the account would usually appear in the broker’s profit and loss account. It did not in the case of BFS, as BSL met the cost of errors.

This is to be contrasted with an unallocated trades account. That was an account to which were booked trades that had not been done in error, but which for some reason could not be booked, at least yet, to a client account: for example, the client account might be in the course of being opened.

One particular use for an unallocated trades account was to book give-up trades. A customer might instruct BFS to execute a trade for it, but instruct BFS then to “give up” the trade to another broker for settlement and clearing. Such a customer might not have an account with BFS at all. Therefore if the trade had to be booked to an account pending being given up to the other broker, it would be booked to an unallocated trades account.

BSL was originally a niche broker in Far Eastern Securities. It achieved dramatic success in equities, Japanese warrants and convertible bonds on the back of a bull market in Japan which began in the early 1980s and continued until the crash of the Nikkei index at the end of 1989.

BFS was incorporated as an indirect subsidiary of BSL on 17 September 1986, and obtained a seat on SIMEX as a corporate non-clearing member. It was dormant until June 1992. Until then, when Barings companies needed trades executed on SIMEX, they used the Singapore branch of Chase Manhattan as an external broker.

Mr Killian joined Barings in the summer of 1988. He was the first F&O specialist hired by Barings and was employed to establish an F&O business for the Barings group at BSJ in Japan. Barings had an established client base dealing in Japanese equities and wanted the facility to offer their clients Japanese futures and options derivatives to hedge their positions in the event that the market fell.

When Mr Killian joined Barings he already had some 7 years of experience in the F&O brokerage business, all of which he obtained on SIMEX. He had set up an F&O brokerage business for Continental Illinois in 1981 and had managed its SIMEX agency business in Singapore until 1986 when he joined Chase Manhattan as the General Manager of its SIMEX brokerage business. He had been in charge of both the front and back offices of these businesses, had passed the SIMEX trading exam and had traded on the SIMEX floor. He had also served as a director of SIMEX and had been a member of the SIMEX Rules Committee and the vice-chairman of the SIMEX Education and Marketing Committee.

Mr Killian was hired by Barings because of his expertise in futures and options, to provide the managerial input and guidance necessary to establish and develop a F&O sales business for BSJ and the Barings group. Of all the employees within the Barings group, Mr Killian was the person most experienced in activities on futures trading floors.

Mr Gueler had joined Barings as a part time analyst for BSJ in April 1988. By mid-1989, he had been co-opted by Mr Killian and was executing client orders for the futures and options agency group. In 1990, BSJ set up a warrant arbitrage and index arbitrage trading operation. Mr Gueler’s computer and mathematical skills meant that he was a suitable person to assist in setting up the systems necessary to conduct these relatively sophisticated trading strategies and Mr Gueler therefore joined this group. Once the head of the arbitrage group, Mr Johnston, returned to London in 1992, Mr Gueler was the most senior proprietary derivatives trader at BSJ.

Mr Gueler assumed full management responsibility for BSJ’s index arbitrage activity in 1991 and 1992. At Mr Martin’s request, he also took responsibility for the Osaka options volatility book being run by Mr Brindle in April 1993. Thus, by early spring 1993, Mr Gueler was in charge of all proprietary Nikkei 225 trading in Japan. As Mr Gueler explained in his witness statement: “The futures, options and arbitrage derivative traders in Japan were put into one group and I was put in charge of them” (Footnote: ⁷⁶).

Management responsibility for front office activities within the BSL group was organised pursuant to the matrix system, on a product rather than a geographical basis. Hence, although he was based in Tokyo, Mr Gueler was made responsible for the Osaka-based options volatility book because it formed part of the same product (proprietary trading in Japanese derivatives), though run out of a different location. In June 1993, Mr Gueler also assumed management responsibility for the proprietary derivatives business undertaken by Baring Futures (Hong Kong) Limited in Hong Kong.

In 1993, Mr Killian and Mr Gueler both reported to Mr Norris. Because proprietary trading involved taking risk on behalf of the bank, Mr Gueler also reported to the Risk Committee in London.

Mr Killian was the chief proponent of the activation of Barings’ derivatives business in Singapore and of BFS becoming a full clearing member of SIMEX. By the beginning of 1992, a substantial number of Mr Killian’s F&O agency customers were expressing a preference to have their business transacted on SIMEX rather than the Japanese exchanges. Futures and options on the Nikkei 225 contract could be traded on SIMEX as well as the Japanese exchanges and margin costs and commissions were lower on SIMEX. By 1992, a substantial proportion of Barings’ agency business was being executed by Chase Manhattan on SIMEX. Mr Killian’s perception was that this flow of business to SIMEX away from the Japanese exchanges would continue. All the big houses were active on SIMEX and Mr Killian recognised that it was important for Barings to keep up with the competition.

Following discussions between Mr Killian and the then senior management in London (Mr Heath, Mr Baylis and Mr Martin), Mr Killian was asked to put forward a proposal for the activation of BFS. On 19 February 1992, Mr Killian made a written presentation to Barings management in London which recommended the activation of BFS. His proposal was that BFS would operate as an execution facility for the established agency business that Mr Killian had already built up in Japan. There was no intention that BFS would be engaged in proprietary trading. As Mr Killian explained in his evidence “My original concept was that we would activate our dormant membership at SIMEX and we would perform the clearing function so that we would save money by clearing the trades ourselves, as well as when margin monies were being placed by clients they were actually being placed by us so we were allowed to earn interest income. But I was not interested at the time in replacing... the execution staff of Chase Manhattan, because they were very good ” (Footnote: ⁷⁷).

Mr Killian envisaged BFS as a “lean operation” (Footnote: ⁷⁸), so structured and staffed as to keep both start-up and running costs as low as possible. 1991 and 1992 were difficult years in financial terms for BSL and Mr Killian recognised that it was important to keep these costs low in order for his proposal to be approved. The total staff requirement was six plus a secretary and the use of the BSS accountant, Ms Yong. Mr Killian did not make any suggestion as to what the role of the existing local management of BSS should be.

Mr Killian asked senior management to approve his proposal, and to direct him to begin the application process and to begin looking for initial staff. The London management committee gave approval for the activation of BFS at its meeting on 26 February 1992. Mr Dickel was allocated the practical task of setting up BFS’ operations in Singapore and London.

Once the plan had been approved, Mr Killian was actively involved in the start-up of BFS, and he was appointed a director in December 1992. Mr Killian accepted that he knew BFS was relying on him to assist Mr Dickel and Leeson in the process of setting up the BFS business.

Although experienced in the practical task of setting up overseas offices, Mr Dickel had no knowledge or experience of the futures and options business. Similarly, as Mr Killian knew, Mr Bax’s and Mr Jones’s experience at BSS was confined to the equities business conducted on the Singapore Stock Exchange. Mr Dickel was dependent upon Mr Killian to inform and guide him in relation to futures and options and the requirements of such a business. Mr Killian accepted that he was in constant contact with Mr Dickel and advised him on the best way to structure the BFS operation.

Mr Killian’s priority was to establish a clearing and settlements capacity, to clear the Barings trades that would be executed in the SIMEX pits by Chase Manhattan and given up to BFS. This strategy was reflected in the application that was made to SIMEX to upgrade BFS’ status to that of a clearing member, which stated “we will initially operate by executing through other brokers. It is envisaged that we will have our own salaried brokers in the pit 6 to 9 months from now” (Footnote: ⁷⁹).

Mr Killian recommended that the CONTAC system be used by the BFS back office (Footnote: ⁸⁰). Mr Killian had no recollection of being involved in the establishment of the settlements procedures in London in relation to the SIMEX business. However, he did understand, from an e-mail he received from BSL’ systems specialist, Mr Dixon, on 10 March 1992, that BSL in London would be having the SPAN margining capability installed to enable it to calculate or re-calculate the SIMEX margin calls from trade data.

Consistent with the intention to establish BFS as a settlements function for the first 6 to 9 months of its operations, Mr Killian’s initial focus was to identify a suitable settlements specialist who could run the back office in Singapore.

Leeson had joined BSL in London on 31 July 1989 as a clerk in Futures and Options Settlements. By 1992 he had developed a reputation in BSL as an expert in this field. As well as helping to sort out BSL’s own back office, he was used by BSL as a trouble-shooter to sort out the settlements systems of BSL’s company in Indonesia. In March 1992 Leeson was just 25 (Footnote: ⁸¹).

While still working for BSL, Leeson applied to the SFA for admission to its register of securities representatives. On the application he answered in the negative a question asking if there were any unsatisfied judgment debts against him. That answer was false, as he had had a judgment entered against him, for £639 in February 1991, by a bank.

The SFA wrote to BSL in March 1992, informing them of the non-disclosure of the judgment. They wrote again in August 1992, by which time Leeson was in Singapore. Ms Thomas, the head of BSL’s Compliance Department, replied that BSL did not propose to continue with the application. BSL did not inform Mr Bax of these events, so that he did not disclose it to SIMEX when he applied for Leeson to become an associated person at SIMEX.

Leeson had accompanied Mr Dickel on his visit to Singapore in March 1992 to assess the feasibility of the proposed activation of BFS. During that trip, Mr Dickel telephoned Mr Killian in Tokyo and explained that Leeson had asked whether he could be considered for the job of settlements manager at BFS.

Following “extensive” (Footnote: ⁸²) discussions involving Messrs Dickel, Killian, Martin (BSL’s then finance director) and Jones, Leeson was appointed to run the SIMEX operation on 19 March 1992 for a trial period of 6 months. He was given the title Derivatives Operations Manager and posted to Singapore to (in the words of Mr Martin’s fax of 24 March 1992) “...head up our SIMEX operation and also act as floor manager” (Footnote: ⁸³).

Mr Killian’s contribution to those discussions was informed by his familiarity with Leeson as a settlements “firefighter”. In 1991, Leeson had been sent to Japan to help resolve BSJ’s futures and options settlement problems. Mr Killian’s resultant appeal to Mr Martin that “we need 10 more Nicks...” (Footnote: ⁸⁴) showed that Mr Killian was impressed by Leeson as a settlements operator. However, Leeson was not a trader. He was 25 years old and, at the time he was posted to BFS, the entirety of his experience and training had been derived from his role in settlements. Both Mr Killian and Mr Gueler knew that.

Since BFS was initially intended to function as a SIMEX clearing operation for BSJ and its clients, Leeson’s lack of prior trading experience did not deter Mr Killian from recommending that he be appointed to head up the back office of the Singapore operation. However, as Mr Killian accepted in his evidence, Leeson was not qualified for the position of SIMEX floor manager. As Mr Killian put it: “... if there were people to manage, yes, [Leeson] would not have been a good floor manager” (Footnote: ⁸⁵).

When Leeson was appointed floor manager in March 1992, the only activity that it was envisaged BFS would be undertaking during the first 6 to 9 months of its operations was the acceptance of give-up trades from Chase Manhattan (trades executed by Chase Manhattan but handed over for clearing by the BFS back office). At that time it was not anticipated that Leeson would be managing any BFS floor staff, let alone that Leeson would be actually executing SIMEX trades himself. Accordingly Mr Martin’s fax did not identify a product reporting line for Leeson, as in charge of BFS’ market traders, because it was not relevant at that time.

Instead, a few days before Leeson’s appointment, Mr Dickel had explained to Mr Jones “Nick will report to you but Gordon Bowser will also have an overall input as global derivatives operations head” (Footnote: ⁸⁶). Leeson had been posted to BFS to run the back office, so reporting lines to Mr Jones and Mr Bowser made obvious sense.

Mr Bax and Mr Jones were unhappy about this joint reporting arrangement. Mr Bax wrote to Mr Fraser on 25 March to protest that Mr Jones should be solely responsible for BFS’ operations side. However Mr Bax told me that he discussed his concerns with Mr Martin and he was reassured by the proposed structure of the business: it was to be an execution-only service, with all trades settled in London or Japan. BSS had established a similar execution-only office in Thailand, which was effectively run from Singapore. On that basis, Mr Bax accepted the arrangement.

According to Mr Bowser, Mr Bax and Mr Jones achieved their objectives by another route. When Mr Bowser visited Singapore shortly after he joined BSL in March 1992, they made it clear to him that Leeson would not report to him (Footnote: ⁸⁷). They pointed out that functions such as assessment could only be done locally. Mr Bowser said that Mr Martin impliedly accepted that and as a result Leeson never reported to Mr Bowser. Mr Bax disagreed with that account. He said he would not have contradicted Mr Martin’s decision. His understanding was that Mr Bowser was responsible for setting up and managing the BFS settlements function, and that Leeson reported to him (Footnote: ⁸⁸).

BFS commenced its SIMEX operations in June 1992. Leeson had arrived in Singapore, following his honeymoon, in early April 1992. Mr Killian played an “active role” (Footnote: ⁸⁹) during the period leading up to the start of BFS’ operations. He was in regular contact with Leeson by telephone and visited Singapore in May 1992. According to Leeson in his interview by the SFO, Mr Killian was “relied on heavily” (Footnote: ⁹⁰) in the process of establishing the operations of BFS, not least because of his previous experience as a director of SIMEX.

During June 1992, BFS cleared trades that had been executed on its behalf by Chase Manhattan, as originally envisaged. That arrangement had originally been expected to continue until the end of 1992 at least. But Leeson clearly had a vision for BFS which was different from that which Mr Killian had set out in his planning papers. Leeson started recruiting front-office staff to execute trades and on 24 June 1992 distributed a memorandum announcing that BFS’ own floor operation on SIMEX would start on 1 July 1992.

Mr Killian was presumably consulted as to this change of plan as, during his visit to Singapore in May, Mr Killian participated in an interview of Eric Chang, who was recruited as the first member of BFS’ floor team. Mr Chang had been an employee of Chase Manhattan and had worked with Mr Killian there. Connie Tan, another recruit from Chase Manhattan, was hired as a result of a referral from BSJ’s Tokyo office. Chang, Tan and the other traders employed by BFS seem to have been engaged in straightforward execution-only trading on the instructions of BSL, BSJ and their customers. Only Mr Leaning seems to have had any involvement in discretionary arbitrage trading, in a very junior capacity under Leeson. None of these traders was called to give evidence before me.

In his fax of 24 June 1992 Leeson did not identify himself as one of the BFS floor traders. Leeson sat the SIMEX Institute of Banking and Finance, Futures and Trading test two days later, on 26 June 1992 (Footnote: ⁹¹). This gave him access to the SIMEX floor. Mr Killian saw that as a useful marketing tool, allowing Leeson to speak to clients on the telephone from the booth and to show visitors around. When BFS began its floor operation, Leeson had not yet received his badge or participated in the SIMEX briefing and simulated pit trading session which he was scheduled to attend on 25 August 1992.

In his fax of 24 June 1992, Leeson proposed that the BFS floor staff would service the agency team in Japan. For the time being, the proprietary traders should continue to place their orders with Chase Manhattan. This arrangement meant that Mr Gueler’s group of proprietary traders would still pay commission to an external broker, which was less satisfactory than using BFS as its in-house broker. When the house traders tried to direct their orders to the BFS floor staff (rather than to Chase), they found that customer orders were given a higher priority.

This situation was unsatisfactory to the BSJ proprietary traders. At Mr Killian’s suggestion, Mr Gueler spoke to Leeson in an attempt to resolve the problem. Whether as a consequence of this or not, Leeson himself became engaged in the execution of the orders of proprietary traders on SIMEX. Once he was active on the floor of SIMEX, both the F&O sales team and the proprietary traders at BSJ began using Leeson to execute their orders on SIMEX.

As Mr Killian accepted, Leeson required permission from a higher authority to execute trades on the floor of SIMEX, because it involved an expansion of his role (Footnote: ⁹²). There is no evidence that anyone senior to Leeson gave such permission. Mr Bax identified Mr Killian as one of the individuals with authority to do so, but both Mr Killian and Mr Gueler denied that they had done so. Mr Killian had not anticipated Leeson becoming an active floor trader, at least for some time. In re-examination, he confirmed that Leeson was not at that point a “suitable floor trader. He was not experienced yet” (Footnote: ⁹³). At the time, he would have seen Leeson as developing that experience, little by little, under Eric Chang’s tutelage.

Nonetheless, by July 1992 Leeson was out of the back office and actively engaged in executing trades on the floor of SIMEX. Mr Killian accepted that he was aware of that (Footnote: ⁹⁴). As Mr Killian accepted he knew, Leeson was not subject to any direct supervision from anyone in relation to his activities on the trading floor of SIMEX.

On 3 July 1992, Leeson opened an account, 88888, in the CONTAC system, and booked his first SIMEX trade to the account on the same day. The 88888 account was described on the CONTAC new account form as a BFS error account (for a description of error accounts, see paragraph 160 above) (Footnote: ⁹⁵). When Leeson notified it to SIMEX on 26 August 1992, he described it initially as a BSL error account, but then changed the description to merely a BSL account (Footnote: ⁹⁶). As described at paragraph 300 below, on 8 July 1992 he arranged for reports of trading activity using the 88888 account to be omitted from the trade feed to London and the London Gross Report.

D&T put to Mr Bowser Leeson’s account (in his book, “Rogue Trader”) that he opened the 88888 account, and then had it removed from the trade feed, at Mr Bowser’s suggestion (Footnote: ⁹⁷). According to Leeson, this was to avoid London having to deal with a large number of small errors. He said that later on Mr Bowser asked Leeson to revert to reporting all errors to London, but Leeson did not then or subsequently arrange for data from the 88888 account to be included in the trade feed.

Mr Bowser did not recall any such conversation and thought it unlikely that he would have made such a request, given that the errors did not cause any extra work in London (Footnote: ⁹⁸). However there was some later documentary evidence that Mr Bowser was aware of Leeson having an error account which was not reported to London (Footnote: ⁹⁹). It is perhaps more likely that Leeson opened an account at Mr Bowser’s request which was not to be reported to London, and later realised that it could be used for unauthorised trading, than that he opened such an illicit account within eight days of BFS starting floor trading. It also seems unlikely that, if Leeson’s intentions had been fraudulent from the outset, he should have removed the 88888 account from the trade feed but not from the margin feed.

BFS’ agency customers were mostly customers of BSL in London, although BSJ also had some domestic Japanese customers. BFS had only one outside customer of its own, BNP, from October 1992. Proprietary trading was on behalf of both BSL (later BSLL) and BSJ (Footnote: ¹⁰⁰). For both agency and proprietary trading, the trading was mostly initiated by BSJ traders (or Leeson).

Leeson’s rise to prominence began in late 1992 and early 1993 when he was involved in rolling over both client and house futures positions on SIMEX. Positions needed to be rolled over because a Nikkei futures contract only had a life of three months and matured at the quarter end in March, June, September or December. Consequently, if a client or house book wanted to maintain an existing futures position it had to be rolled over to the next contract period before expiry. As Mr Killian explained “a roll... involves effectively closing out the near contract and purchasing a far contract to replace it” (Footnote: ¹⁰¹). Market makers on SIMEX offered rolls as a single trade whereby they would buy (say) September futures at x and sell an equivalent number of December futures at y. The difference between x and y would be the spread and represent the cost to the customer in rolling over his position. There was, therefore, a prevailing price for rolls on SIMEX which could be bought as a single trade without the necessity for the order filler to take any directional risk (that is, risk that the market would move during the roll operation) or execution risk (that is, the risk of an error in implementing the order).

When roll orders were given to Leeson, he was apparently able to achieve prices which were often more favourable than the prices being offered in the spread market. The evidence of Dr Fitzgerald was that the most cogent explanation (from BSJ’s perspective) for Leeson’s good prices would have been that he was “legging” into the roll (Footnote: ¹⁰²). In other words, Leeson would execute one leg of the roll order and then wait before executing the other leg in the hope that the market would move to improve the spread. Mr Killian’s evidence was that he understood Leeson to be beating the spread price by carrying out the two trades independently. He had done the same himself as a trader (Footnote: ¹⁰³). This was legging, but only in the sense it was not possible to carry out the two trades completely simultaneously.

When its floor operation commenced in July 1992, BFS’ authority was limited to the execution of orders on behalf of other companies within the Barings group and their clients. Nobody at BFS had authority to take directional risk, or to initiate house positions or to exercise discretion in the execution of customer orders without prior approval from a compliance officer. Mr Killian knew this, and was not aware of Ms Thomas or any other compliance officer having given permission for Leeson to accept discretionary orders from customers. Mr Killian had not budgeted for a specific compliance officer for BFS in his business plan or proposal and none had been appointed.

The fact that BFS was intended to be an agency-only business meant that there was no segregation of duties problem arising out of Leeson’s position in charge of both front and back offices. As Mr Killian explained to the Singapore inspectors and confirmed in his evidence “...it’s typical to have General Managers who are responsible for the floor and back office... It was an ordinary thing” and “... when I was in Singapore I stood on the floor and I would take orders from clients and execute the orders and I still retained the title of general manager and I was responsible for the back office...” (Footnote: ¹⁰⁴)

During the course of 1993, Leeson metamorphosed from an execution-only broker to a trader. A broker executes an order at the instigation of another, at a stated price, “at best” or “at market”, exercising discretion only in his judgement as to how precisely to execute the order. A trader has a discretion as to what trades to execute and at what price. Unlike a broker, a trader initiates trades and effectively puts the bank’s money at risk. As a result, as Mr Killian and Mr Norris accepted (Footnote: ¹⁰⁵), trading of a proprietary nature gives rise to more complex risk management concerns than pure execution activity. It requires transparency and usually involves the individual trader being given limits to control his activity.

Leeson began trading in the above sense in early 1993, and slowly this activity grew. Leeson was given more and more discretion by the BSJ traders to move their futures hedges from Osaka to SIMEX. It was apparent (to Mr Killian at least) that by early 1993, Leeson was effectively trading; initiating house positions and exercising discretion (Footnote: ¹⁰⁶). Mr Gueler admitted that it was he and the members of his proprietary desk at BSJ who conferred discretion on Leeson (Footnote: ¹⁰⁷). However, as Mr Killian knew, Leeson had not been given any authority by higher management to engage in that activity and had not received approval from any compliance officer entitling him to exercise discretion in the execution of orders (as the BSL Compliance Manual required) (Footnote: ¹⁰⁸).

By spring 1993, Leeson was apparently beginning to make a sizeable contribution to the profitability of the house books run by the BSJ proprietary traders in Japan. Mr Gueler’s April memorandum to Mr Norris noted “Our SIMEX execution and trading ability is superb and extremely valuable to us. This is due to the great efforts made by Nick Leeson and his team.” (Footnote: ¹⁰⁹), though in his evidence, Mr Gueler stated that the word “trading” did not mean proprietary trading (Footnote: ¹¹⁰). Mr Killian recalled being told by Mr Gueler’s group that they had managed to add £100,000 to the value of the house books on a single day as a result of Leeson legging the execution of proprietary positions that were being moved from Osaka to SIMEX (Footnote: ¹¹¹).

Mr Killian appreciated that Leeson’s involvement in proprietary trading gave rise to a potential conflict of interest. He raised this concern with Mr Baylis and Mr Martin in London but was told to “relax” because the house was just a customer of BFS and Leeson was simply executing the house orders (Footnote: ¹¹²). At that time, which must have been before the end of March 1993 when Mr Baylis and Mr Martin left, Mr Killian regarded this response as satisfactory “because [Leeson] was executing in an agency capacity only, not as a proprietary trader” (Footnote: ¹¹³). It is also true that Mr Killian’s group stood to benefit (by way of internal commission) from the increasing volume of trades due to Leeson servicing the house books.

The trading activity at which Leeson was apparently so successful was arbitraging between SIMEX and the Osaka exchange. This was known as “switching”. Arbitrage of this nature (sometimes called “price anomaly arbitrage” (Footnote: ¹¹⁴)) aims to capture price differences (price anomalies) in the same product on two different markets. The trader sells the product on the market on which the price is higher and buys an equal quantity of the same product on the other market where the price is lower. As long as he effects the two transactions (the two “legs” of the arbitrage) as nearly as possible simultaneously, the strategy is one of low market risk. The price differential is “captured” once the two markets converge and the price of the product is the same on both markets, whichever way the markets may have moved in the meantime.

In price anomaly arbitrage the positions held on the two exchanges must be equal and opposite. If they are not, the unmatched (or “open”) position will be directly vulnerable to market movements. So, during the period while only one of the two legs of the arbitrage is in place, there is an exposure to market movement, and therefore directional risk.

Instead of effecting both the legs of the arbitrage as near as possible simultaneously, a trader can deliberately create an open position, either by delaying putting down one of the two legs (“legging in”), or by “lifting” one of the legs after it has been put in place. Both these variants involve straightforward directional trading, that is, gambling on the directional movement of the market. In both cases, a properly-run bank or broker would limit the extent of the positions which the trader was permitted to have open, either in the course of a day’s trading or overnight.

The origins of Leeson’s switching business lay in the hedging strategy pursued by BSJ’s proprietary traders in Japan. As already mentioned, Adrian Brindle was trading Nikkei options on a volatility book. As a hedging strategy, to offset the risks involved in options trading, BSJ dealt in futures, both on the Japanese markets and (through BFS) on SIMEX. In the course of executing these hedging trades for BSJ, Leeson purported to show that he could arbitrage profitably between Osaka and SIMEX (Footnote: ¹¹⁵).

There was nothing new or unique about switching. In a fax sent by BSJ to Mr Baylis in September 1992, “Osaka SIMEX arbitrage” had been identified by Mr Brindle and Mr Fou as one of the strategies in which BSJ was engaged. Arbitrage trading between two exchanges was not sophisticated. Mr Gueler called it “downstairs trading requiring little or no technical knowledge” (Footnote: ¹¹⁶). According to Mr Gueler, Leeson was a “barrow boy” just buying cheap and selling dear (Footnote: ¹¹⁷). Although Leeson’s perceived ability to make proprietary profits by reason of his access to customer orders may have meant that his proprietary trading was not as “pure” as that being conducted by the traders at BSJ, it was not beyond the competence of Mr Gueler to comprehend. Indeed, Mr Gueler said he had tried it himself but had got his “fingers burnt” (Footnote: ¹¹⁸).

Mr Gueler attributed Leeson’s success to his presence on the floor in SIMEX, with a dedicated phone line to Osaka. This allowed Leeson to spot price anomalies more quickly than Mr Gueler had been able to do trading from a screen in Tokyo (Footnote: ¹¹⁹). As I will later describe, Leeson’s apparent success was actually largely the result of concealed alterations to trades using the 88888 account, to its disadvantage and thus that of BFS.

When it started, Leeson’s switching was accounted for as part of BSJ’s volatility book. Initially it was restricted to dealing in Nikkei 225 futures contracts. It seems from a fax from Mr Oades to Mr Norris dated 17 June 1993 (Footnote: ¹²⁰) that the Nikkei switching business was already established by that date, and that the possibility of expanding into JGBs was under consideration (as happened in 1994), though apparently no one in London told Mr Baker about it before he went out on his due diligence at the end of 1993 (Footnote: ¹²¹). Mr Killian told me that he knew as early as late 1992 that Leeson was involved in switching, but understood him only to be carrying out the directions of BSJ traders. Mr Killian only became aware in mid to late 1993 that Leeson was taking the initiative in the trading (Footnote: ¹²²).

A development of the switching business, and what was perceived to be a significant element in the success of the business, was liquidity arbitrage. This is described by BOBS in the following terms (Footnote: ¹²³):

“if BFS received a large order on SIMEX and the liquidity (Footnote: ¹²⁴) was not available to execute it immediately, BFS would offer to execute the customer’s order by taking on the other side itself and laying off the risk with the equivalent number of contracts on the more liquid Japanese exchange. BFS would charge the customer an extra price tick (Footnote: ¹²⁵) for doing this..... and would seek to unwind it later on when both markets had liquidity.”

Mr Gueler’s evidence was that he believed Leeson to be engaging in a combination of price-anomaly arbitrage, liquidity arbitrage and directional trading.

By mid-1993, Leeson and the BSJ house traders in Japan had decided that Leeson should run the switching strategy from Singapore. This resulted from an informal group decision made by the BSJ proprietary traders and Leeson in the summer of 1993. From that point, as Leeson put it in his interviews by the Serious Fraud Office, “I was the trader. I was the decision-maker” (Footnote: ¹²⁶). As Mr Gueler accepted, from August 1993 onwards “Nick [Leeson] would call the shot”; the “go/no go decision” would come from Singapore (Footnote: ¹²⁷). Rather than simply executing orders on SIMEX at BSJ’s request, Leeson initiated trades, exercised discretion and gave the orders to BSJ to execute trades in Japan. Leeson thus clearly crossed the line between an order-filler and a trader.

As Mr Norris accepted, this transition amounted to a change in roles, importance and status for Leeson (Footnote: ¹²⁸). For BSJ to confer authority on Leeson to conduct discretionary arbitrage was, as both Mr Norris and Mr Killian agreed, to give him “a book in all but name” (Footnote: ¹²⁹). Mr Gueler agreed too but then had second thoughts later in his evidence, suggesting that, throughout 1993 and until late spring/early summer of 1994, Leeson’s ability to initiate trades was limited by parameters set by BSJ on the morning of each day (Footnote: ¹³⁰). However this is contrary to both Leeson’s account to the SFO quoted above and Mr Killian’s evidence. It is also inconsistent with Mr Gueler’s evidence that there was no difference between what Leeson was doing before and after he got his own trading book (Footnote: ¹³¹). I find Mr Gueler’s first thoughts to have been correct.

Mr Gueler’s group of proprietary traders had not asked Mr Killian’s permission for Leeson effectively to run a proprietary switching book from Singapore; nor was Mr Killian aware of Leeson having been authorised to conduct the switching business by anyone in senior management (Footnote: ¹³²). Mr Norris said he would have expected such a change in Leeson’s role to be reported to him but he was not made aware of it (Footnote: ¹³³).

Mr Killian had learnt that Leeson was running his own book in all but name by the autumn of 1993. He discovered this from his conversations with the proprietary traders and his own execution sales staff in Japan (Footnote: ¹³⁴). By the autumn of 1993, Mr Gueler said that Leeson had become “principal” among the team running the Osaka volatility book and that Leeson’s switching was contributing the most profit (Footnote: ¹³⁵). Mr Killian recognised that, by the end of 1993, Leeson was “clearly now doing proprietary trading and he’s part of the proprietary traders...” (Footnote: ¹³⁶).

Mr Killian told me that he appreciated that the switching business (and Leeson’s role in it) represented a fundamental departure from the execution-only role that BFS had been set up to conduct and that it was “essential” that the questions of risk limits and risk management should be considered (Footnote: ¹³⁷). Moreover, Mr Killian knew that the change in Leeson’s role had serious implications for the agency business. Leeson now faced a potential conflict of interest and there was a risk of complaints of front-running and other similar abuses from customers. Mr Killian also accepted that he knew that Leeson’s unsegregated role, in running both the front and back offices of BFS, was no longer appropriate now that he was engaged in proprietary trading (Footnote: ¹³⁸).

Leeson’s switching activity in Nikkei 225s was established by June 1993, but no separate figures for its profitability are available until September 1994. However it seems that the switching began to make striking apparent profits in November and December 1993. On the basis of comments in contemporary documents and in evidence to the Singapore inspectors, D&T have provided figures which treat Leeson’s switching as representing 60% of the profitability of Mr Brindle’s volatility book in those two months. This gives a figure for Leeson’s income of £3.5 million for November and December (Footnote: ¹³⁹). This was included in Mr Brindle’s book and therefore reflected in BSJ’s financial results, not BFS’.

1993 was a year of substantial change within the BSL group. Throughout 1993 the BSJ proprietary traders were not subject to any clear or active management supervision. Mr Gueler was distracted by his responsibilities for Hong Kong and, as the result of an “undesirable interregnum” in senior management, the derivatives group had a succession of different bosses to report to in London (Footnote: ¹⁴⁰). As Ms Walz put it, before she and Mr Baker took over, the derivatives business was “tossed around a whole lot of people” (Footnote: ¹⁴¹). Mr Norris himself had no expertise or experience of proprietary trading and was ambivalent as to whether Barings should persevere with a derivatives business at all. Mr Oades, whom Mr Norris attempted to groom as a potential senior manager, was regarded by Mr Gueler as an unsuitable manager with no understanding of the derivatives business (Footnote: ¹⁴²). In the result, as Mr Killian explained “... the young guys like [Mr Gueler] and [Leeson] and Mr Fuchs really did not have a sponsor/mentor for them for the proprietary activities out of London” (Footnote: ¹⁴³). Mr Gueler said that he and his group did not trust senior management and felt neglected and “thoroughly undervalued” in 1993 (Footnote: ¹⁴⁴).

When Mr Baker visited Japan at the end of 1993 he described the proprietary trading group as being “very unmanaged and very unloved” and operating in a “climate of concealment” (Footnote: ¹⁴⁵), which was a perception Mr Killian confirmed was probably accurate (Footnote: ¹⁴⁶). This was the climate which prevailed in Japan throughout the period when Leeson’s switching business was born.

Even though the switching was Leeson’s business, BSJ performed an essential role in it. BSJ appointed dedicated trading assistants to execute trades in Japan at Leeson’s behest and received details of the corresponding trades on SIMEX from Leeson each day. This was done by members of Mr Gueler’s group and under his supervision.

Because in 1993 Leeson’s switching was booked to Mr Brindle’s volatility book, its profits were hidden within BSJ’s books and were not reported to either London or Singapore in a transparent manner. Mr Brindle kept an unofficial spreadsheet recording the contribution of Leeson’s switching, but this was not known to anyone outside BSJ until 1994.

Once Leeson had been given his own books (see paragraphs 271 and 272 below), BSJ included Leeson’s positions and profit in the official traders’ spreadsheets. These were saved in a shared directory on a computer, and were used by BSJ risk control to produce the Daily Risk Report for the Risk Committee and Financial Control in London. All the FPG managers and traders, including Mr Gueler, Mr Baker and Ms Walz, received a copy of the Equity Financial Products Daily Income Report from early 1994 and a copy of the group weekly income report which set out details of Leeson’s profits.

Therefore Mr Gueler received details of all Leeson’s switching trades (including prices and volumes) from 1993 through to the end of 1994. Mr Baker and Ms Walz received details of Leeson’s profits from early 1994, as Mr Brindle began adding manuscript remarks to his daily report, signifying the level of Leeson’s contribution to his book’s income, and Leeson received his own books for JGB and Euroyen trading.

Mr Norris’ tenure as the senior manager with overall responsibility for the derivatives business was a temporary one, assumed by default following the departure of Mr Martin, Mr Baylis and Ms Gibson in March 1993. During this managerial “vacuum” (Footnote: ¹⁴⁷), Mr Norris was looking for a replacement.

In the second half of 1993, Mr Norris attempted to persuade Mr Baker (who had joined BB&Co from Bankers Trust in 1992) to take control of the derivatives business as part of his Debt Financial Products Group. Although Mr Baker had no experience of exchange traded derivatives, he agreed to look at the business. At that stage, Mr Baker was only interested in the proprietary trading and he did not want responsibility for the agency business run by Mr Killian.

In September 1993 Mr Baker was sent copies of spreadsheets on the derivatives books recording the reported results from January to August 1993 which Mr Norris described as “most of the relevant information on the Derivatives Unit here” (Footnote: ¹⁴⁸). On 26 October 1993, Mr Baker circulated a memorandum entitled “Organisation of a Financial Products Group”, which included the BSJ derivatives business within a sub-group named Equity Financial Products to be headed by Ms Walz. Two days later Mr Norris issued a memorandum to all staff in the BSL group announcing that Mr Baker would be assuming responsibility “for managing our existing derivative trading and stock borrowing businesses” (Footnote: ¹⁴⁹).

Mr Baker did not visit Japan until late 1993 and early 1994. It was only then, from his discussions with the traders in Japan, that Mr Baker “discovered” the switching business and the importance of Leeson to the apparent profitability of BSJ’s Japan trading books (Footnote: ¹⁵⁰). Mr Killian said that Mr Baker “spied” the opportunity to take control of this business which, since it was concealed within the BSJ books, had not been claimed by anyone else in London (Footnote: ¹⁵¹). It would not have been apparent from the spreadsheets Mr Baker had received from Mr Norris in September because Leeson’s trading was still buried in the Osaka volatility book. Until Mr Baker discovered Leeson’s trading, only BSJ was aware of the level of Leeson’s apparent contribution to the results of the Japanese house books. It was, as Mr Killian described it “... a bit of a home grown build up of business opportunity” (Footnote: ¹⁵²).

Mr Baker’s first meeting with Mr Killian was over lunch in Tokyo in November 1993. According to Mr Killian “My conflict began with Ron Baker on the first day I met him” (Footnote: ¹⁵³). Mr Baker told Mr Killian that he intended to expand the proprietary business and, as part of his vision, it would be important for proprietary traders to “take advantage of the client information curve” (Footnote: ¹⁵⁴). Mr Killian asked Mr Baker whether he was talking about front-running, which Mr Baker denied. But after the collapse, Mr Killian told Mr Baker “When you told me when we first met that you were going to trade the client information curve, I knew you were going to build a front running scam as part of your vision and I was not going to be part of that” (Footnote: ¹⁵⁵).

Mr Killian understood the phrase “taking advantage of the client information curve” to mean that Mr Baker intended the proprietary traders to make trading decisions in the light of information available to them through Barings also executing customer orders (Footnote: ¹⁵⁶). Mr Killian was “stunned” by what he was told by Mr Baker because of the detrimental impact such a strategy would have upon the agency business that Mr Killian had sought to establish (Footnote: ¹⁵⁷).

The absence of any form of Chinese wall in Singapore and the strategy of taking advantage of the client information curve gave rise to a continuing compliance concern which, as far as Mr Killian was concerned, was never adequately addressed by either Mr Baker or Mr Norris (Footnote: ¹⁵⁸). Ms Thomas in the compliance department was never made aware of these matters by Mr Killian or by anybody else within Barings. She said later that, had she looked at the matter, she would have regarded the practice of taking advantage of the client information curve as being “completely out of order...” (Footnote: ¹⁵⁹). Mr Killian accepted that “that is an issue that should have been raised with her” (Footnote: ¹⁶⁰) but it was not.

Shortly after his November 1993 meeting with Mr Baker, Mr Killian learned that Mr Baker had invited Leeson to attend an off-site meeting of the proprietary traders in Hong Kong. Mr Killian was angry and upset by this development. He felt that Leeson would now be openly treated as a proprietary trader and be encouraged by Mr Baker to front-run or abuse client orders by taking advantage of the client information curve (Footnote: ¹⁶¹).

It was this invitation to Leeson which prompted Mr Killian to complain to Mr Norris about Leeson’s role as a proprietary trader. Mr Killian told me that he raised his concerns with senior management in London and suggested that Leeson be given a separate booth on SIMEX. Neither they nor Mr Killian took any steps to change Leeson’s role or responsibilities. Mr Killian accepted that he “should have called [Mr Jones] as well” but he did not (Footnote: ¹⁶²).

Mr Norris in his evidence analysed Mr Killian’s expressed concerns as being twofold. First, Mr Killian was concerned that his clients should continue to receive a good service. Secondly, Mr Killian was worried that the agency group would not get any financial benefit from Leeson’s activities if he was to be a fully-fledged member of Mr Baker’s FPG (Footnote: ¹⁶³). Mr Killian accepted that he raised his concerns in general terms and that this characterisation was essentially accurate (Footnote: ¹⁶⁴).

Mr Norris did not agree with Mr Killian’s view as to the need to separate the availability of customer and house information at the point at which trading and execution decisions are taken. Mr Norris considered it to be both unnecessary and impracticable to have any form of physical separation between Leeson and order-fillers on the SIMEX floor (Footnote: ¹⁶⁵). Mr Killian, who was by far the more experienced, disagreed and maintained that Leeson should have been given a separate booth. However, Mr Killian never ensured that this happened and said that, whenever he raised these issues with Mr Norris, the latter simply “glazed”; it was like “Oh Jeez, here we go again. It was old news” (Footnote: ¹⁶⁶).

Mr Killian’s concerns were never dealt with by Mr Norris satisfactorily and Mr Killian’s continued complaints earned him the nickname of “the bad penny”. Although Mr Norris denied ever referring to Mr Killian in those terms (Footnote: ¹⁶⁷), Mr Killian said that “...when the chairman says that to you, of your company, it is one of those things I tend to remember” (Footnote: ¹⁶⁸).

In the last 2 months of 1993, Leeson’s reported switching profits increased markedly. Having contributed (according to Mr Gueler) an estimated US$200,000 to the income of the volatility book for the first 10 months of 1993, Leeson was apparently able to generate switching profits of about £3.5m in the last two months. By the turn of the year, Leeson’s switching was consistently generating some 60% of the income of the Osaka volatility book and Mr Gueler regarded Leeson as “exceptional” (Footnote: ¹⁶⁹).

Mr Killian was aware, at least from conversations in the office, of the fact that approximately 75% of the proprietary profits for the last 2 months of 1993 were emanating from switching (Footnote: ¹⁷⁰). Mr Killian recalled one acquaintance telling him, in late 1993, that the only way Leeson could be making a lot of money from switching was by “doing one side of the trade and then perhaps waiting before completing the other side, taking the chance on a few more ticks and profits” (Footnote: ¹⁷¹). Mr Killian said he understood Leeson could only have generated profits of such magnitude by taking directional risk (Footnote: ¹⁷²).

By the end of 1993 Leeson had already become the single most profitable trader in Mr Gueler’s group. On occasions when Leeson was absent for any reason, in Mr Hawes’ words BFS’ “profitability dropped to virtually nothing” (Footnote: ¹⁷³). Mr Gueler accepted that Leeson’s results should have been investigated at this stage though, he said, not by him (Footnote: ¹⁷⁴). Yet no one at BSJ or BSL understood or sought to investigate Leeson’s trading activities or the reason for the dramatic increase in his profits. For his part, Mr Gueler admitted that he found Leeson’s profits “puzzling” from the end of 1993 onwards but that he had no idea how Leeson was making them (Footnote: ¹⁷⁵). Mr Gueler took no steps to try to discover, either from Leeson or elsewhere how these profits were being made.

When Mr Baker discovered the switching business at the beginning of 1994 he believed that Leeson’s profits were so substantial that he must be taking risk. According to Mr Gueler, Mr Baker was sceptical about Leeson’s profits and found them “incredible” (Footnote: ¹⁷⁶). But nobody investigated whether the explanation was directional trading and what risks Leeson was taking. When called upon by Mr Baker (at an off-site in Bruges in December 1993) to explain how Leeson made his money, Mr Gueler said that “I had difficulty explaining it then” (Footnote: ¹⁷⁷).

Mr Killian recommended that Leeson’s bonus for 1993 should be S$600,000. This suggestion was based on a bonus pool for the Pacific Rim agency group (from which Leeson’s bonus was originally to be paid) that included £1.4 million of internal commission payable by the house traders to the agency group. Mr Killian had reached an agreement with Mr Gueler during 1993 that such a commission would be payable and rationalised it to senior management in London on the basis that “95% of Nick[’s] trading day is spent servicing the house traders” (Footnote: ¹⁷⁸).

When this suggestion was vetoed by London, Mr Killian initially agreed a compromise with Mr Baker that half of Leeson’s bonus should be paid from the Pacific Rim Agency Group and the other half would be paid by Mr Baker’s FPG. In the event, however, all of Leeson’s bonus was paid by Mr Baker’s group because Mr Killian regarded it as unfair that the agency group should have to bear the overhead of Leeson’s proprietary trading (i.e. SIMEX clearing fees and BFS costs) without the benefit of any internal commission.

The disagreement over Leeson’s 1993 bonus marked the beginning of a turf war between Mr Killian’s F&O sales group and Mr Baker’s FPG as to Leeson’s position and the extent to which each group was for bonus purposes entitled to credit from a share of his apparent income. Leeson was now perceived as a “star” and a “natural trader” whom Mr Baker felt should be brought within the fold of his newly formed group of derivative traders (Footnote: ¹⁷⁹). Mr Killian, on the other hand, was reluctant to deprive the F&O sales group of the reflected glory created by Leeson’s apparent success and a share of his revenue. Mr Killian considered that he was entitled to some benefit from Leeson’s proprietary profits and was still maintaining as late as June 1994 that the agency bonus pool should receive a credit from the proprietary profits being generated by Leeson (Footnote: ¹⁸⁰).

This quarrel between Mr Killian and Mr Baker reflected a critical ambiguity in relation to Leeson, namely whether he was within the sphere of Mr Baker’s or Mr Killian’s managerial responsibility. It dragged on until October 1994, when it was finally decided that the agency business would be merged with FPG as a single product under Mr Baker’s leadership with effect from 1 January 1995.

Despite the fact that Mr Killian was and remained Leeson’s direct superior in relation to agency business until the end of 1994, his reaction to the fact that Leeson became increasingly aligned with the proprietary traders appears to have been the complete abdication of any responsibility for BFS and Leeson. At the end of March 1994, Mr Killian moved from Japan to the United States. Thereafter his only contribution to the management of Leeson’s business appears to have consisted of sarcastic observations on the levels of Leeson’s reported profits. According to Mr Baker, Mr Killian became cynical and lost interest in Barings (Footnote: ¹⁸¹). Mr Gueler said that Mr Killian “just turned a blind eye to where [Leeson’s] commissions would have been coming from” (Footnote: ¹⁸²).

The result of Mr Killian’s insouciance was the absence of any active or effective management supervision over Leeson’s agency business in 1994. Nor was there any effective communication between the managers of the agency team and FPG. This meant that no one in a managerial position developed an overall understanding of Leeson’s activities. This was a critical failure in the events leading up to the collapse of Barings, for which Mr Norris acknowledged he was ultimately at fault (Footnote: ¹⁸³). Mr Gueler agreed that this serious failure was “absolutely at the heart of what went wrong” (Footnote: ¹⁸⁴). In particular, it enabled Leeson to indulge in myths of secret clients with fantastic strategies to explain to his proprietary managers the very large number of trades he was conducting (in reality for account 88888), while no one knew who these clients were or the volume of business they were conducting.

Mr Killian was not prepared to recognise the lack of communication as a failure (Footnote: ¹⁸⁵), but Mr Gueler was: “... [Mr Killian] and I did not communicate with each other, okay. That was a fault okay, maybe... That is what really the problem is, you know it, everybody knows it, that is where the fault lies” (Footnote: ¹⁸⁶). Mr Baker too was willing to accept that “... if I had seen both sides of that puzzle, it would have helped me to be able to pick up what was going on” (Footnote: ¹⁸⁷).

With effect from 1 January 1994, BSJ’s proprietary trading (which included Leeson’s switching) was brought within FPG under Mr Baker with Ms Walz, as head of Equity Financial Products, reporting to him. Mr Gueler became the Head of Equity Derivatives and, from January 1995, Regional Manager of FPG in Japan. Mr Norris explained that Mr Gueler (despite his denials) continued to be Leeson’s most immediate source of authority (Footnote: ¹⁸⁸). I find that the appointment of Mr Baker and Ms Walz did not alter Mr Gueler’s responsibilities but simply meant that, for the first time, he had clear senior management.

During 1994, Leeson was to become fully integrated into Mr Gueler’s section of FPG. Leeson’s apparent success was such that he was invited to the FPG off-site meetings in December 1993 and January 1994. Leeson was regarded as a “miracle worker”. He was referred to as the “Michael Jordan of the floor” after the American basketball star (Footnote: ¹⁸⁹). Leeson’s “legend” in BSL continued to grow; stories as to how much money he was making became bigger and his reputation grew. He felt confident enough to adopt “superman” as his e-mail address (Footnote: ¹⁹⁰).

By the beginning of 1994, Mr Killian’s impression was that Leeson had become bored of agency execution and that he was doing only a limited amount of client business (Footnote: ¹⁹¹). This is consistent with Mr Killian’s comment, in a letter quoted above in the context of Leeson’s bonus, that “95% of Nick[‘s] trading day is spent servicing the house traders” (Footnote: ¹⁹²).

BSJ (and thus the BSL group) was a major beneficiary of Leeson’s trading. Leeson’s reported profits (generated at the expense of adjustments and loss making cross-trades with account 88888) were in large part attributed to BSJ. The reported revenue generated by Leeson in 1994 exceeded £36m. Of this, more than £12m was attributed to Nikkei 225 switching and was included in BSJ’s accounts. From late 1994, some of Leeson’s profits from JGB switching were also booked to BSJ.

The members of the FPG (which included the proprietary traders at BSJ) also benefited personally from the apparent success of Leeson’s trading. The switching revenues generated by Leeson contributed to the pool from which their bonuses were paid. For example, for 1993, Mr Gueler’s bonus was £275,000, which was 4 times greater than the bonus he received for 1992. After the amazingly successful FPG results for 1994 (to which Leeson had personally contributed almost 60%), Mr Gueler’s bonus would have been £390,000 but for the collapse. Mr Baker would have received a bonus of £880,000 on top of his salary of £100,000. Mr Gueler was also invited by Mr Norris to become a director of BIB in recognition of “the significance of his managerial role and responsibilities”, in part in relation to the role he had in Leeson’s switching business (Footnote: ¹⁹³).

As the senior proprietary trader and head of the FPG in Japan, Mr Gueler retained managerial responsibility for all of the Japanese trading books including the volatility book of which, until September 1994, Leeson’s Nikkei switching formed part. Leeson continued to report his daily trades and positions to BSJ and Mr Baker and Ms Walz both relied upon Mr Gueler to keep them informed of the switching business between SIMEX and Osaka Stock Exchange. As Ms Walz told the Singapore inspectors, “a direct reporting relationship for trading risk” existed between Leeson and Mr Gueler and she relied upon Mr Gueler to monitor Leeson (Footnote: ¹⁹⁴).

Unlike the traders at BSJ, Ms Walz and Mr Baker had no experience of exchange traded derivatives of the sort that BSJ and Leeson were engaged in. As Mr Baker has said, it was “clear to me after one day [observing Leeson at SIMEX] that I was never going to understand, if I stayed there for a month, exactly what took place” (Footnote: ¹⁹⁵). As Ms Walz said “[Mr Gueler] clearly had experience in the markets in Japan that I didn’t have...” (Footnote: ¹⁹⁶). Hence, they were heavily dependent upon Mr Gueler to understand and monitor Leeson’s business. Mr Gueler accepted that senior management was dependent upon his specialist derivatives experience (Footnote: ¹⁹⁷).

According to D&T’s Income Tables (Footnote: ¹⁹⁸), Leeson’s apparent revenues from Nikkei 225 switching for 1994 amounted to £12.546 million (firm figures are available only from 23 September 1994, when Leeson was given his own book. D&T’s estimate is if anything conservative, given that the Internal Audit Report quotes a figure for the first seven months of 1994 of £11 million (US$17 million)). Revenue from JGB and Euroyen switching (which began in April and May 1994) amounted to a further £23.931 million, giving an overall total for the switching business in 1994 of £36.478 million. This is to be compared with the total revenue for the whole of FPG in 1994 of £62 million, of the entire Bank Group of £84.4 million, and the whole of BIB of £406.6 million. Thus Leeson’s reported switching revenues amounted to 59% of FPG’s revenue, 43% of the entire Bank Group’s and 9% of BIB’s during 1994.

As to profits, SPG reported profits for 1994 which exceeded the entire Bank Group’s operating profits for the year and amounted to some 17.8 per cent of the Baring Group operating profits for the year. The apparent contribution of Leeson’s switching business to overall Group profits for 1994 was at least 12.5 per cent of the Baring Group operating profits (Footnote: ¹⁹⁹).

Leeson’s reported income from his different activities was consistently high throughout 1994 and showed some striking peaks (Footnote: ²⁰⁰):

Mr Norris’ evidence (Footnote: ²⁰²) was that until early 1994 he had seen Leeson as one of a team, performing a valuable role in enhancing, by his skill at executions, profits on trading run from Japan. However it became clear in early 1994 that Leeson was initiating trades and managing risk himself (including an element of intra-day directional trading). Therefore Mr Norris obtained agreement that Leeson should be subject to his own risk limits and that his revenue should be split out from Mr Brindle’s volatility book.

Implementation of this agreement was delayed by opposition from Mr Baker and Miss Walz, who, according to Mr Norris, feared that the latter change would impact on their group’s profits. According to Ms Walz at the time, the reason was fear that an audit of the book could “send the wrong signals” to the BSJ traders and demotivate Mr Brindle (Footnote: ²⁰³). Whichever is correct, Leeson’s profits from Nikkei switching remained subsumed within BSJ’s volatility book until September 1994. However Mr Brindle did begin to add brief manuscript remarks to his daily report to London and Tokyo signifying the level of Leeson’s contribution to the book’s income.

Mr Norris said that, while this debate was continuing, Leeson’s proprietary switching business was seen as falling within existing risk limits. Mr Brindle managed risk on the Nikkei switching within the limit on his volatility book. BB&Co managed risk on the JGB switching, when that started in April 1994, within its overall limit for JGB-linked trading (as was stated in the Risk Committee minute recording the start of JGB switching (Footnote: ²⁰⁴)). However it was clearly appropriate that the trader managing the risk should be subject to his own, appropriately-set limits: Mr Brindle’s limits, based on maximum loss, were difficult to apply to an arbitrage business, for which a limit based on a maximum number of open contracts was appropriate. Hence the imposition in June 1994 of formal risk limits on Leeson’s switching business (Footnote: ²⁰⁵).

In the first week of January 1994, the BSL risk committee was told that Mr Gueler and Ms Walz were to make a presentation of proposed trading limits to the committee and that the “Osaka business may be expanded to include JGB arbitrage” (Footnote: ²⁰⁶). On 4 February Mr Baker asked Mr Gueler and Ms Walz to report to him on how the SIMEX switching business could be expanded to include JGBs and Euroyen and the intra-day and overnight trading limits that were required. They replied by a fax of 27 February 1994 (Footnote: ²⁰⁷), in which they described Leeson’s business as “intra-day trading... that exploits and relieves price and liquidity anomalies that occur between the Osaka and Simex exchanges.” They attributed BFS’ success to “our pole position in the information flow: we see huge client orders (ours and theirs), we know well the locals and their trading habits, and have an intelligent line into Osaka. We know all the bids and offers. It is difficult to be specific about how we put money at risk when there are so many safety nets at our fingertips”. Mr Gueler was recorded as suggesting a “Max Loss limit applying to Singapore’s activities as a discipline”. No specific limits were recommended although the report did identify the need to “define the terms” on which the switching business was to progress.

On 16 March 1994 Ms Walz and Mr Baker sent a memorandum to the BSL risk committee proposing an intra-day limit of 150 futures for the volatility book and stating that “Mr Gueler will manage the intra day risks up to a total maximum delta of 175 futures across all the books. He may, after consent from [Ms Walz or Mr Baker], go to a maximum of 300 futures across all the books” (Footnote: ²⁰⁸). The limits authorised by the risk committee were sent to BSJ on or shortly after 16 March 1994 and Mr Gueler faxed a copy on to Leeson on 28 March 1994. In evidence, Mr Gueler could not explain why he did so (Footnote: ²⁰⁹). The likely explanation is that he did so as Leeson’s product manager.

On 6 April 1994, Mr Baker reported to the BSL risk committee that “Fernando Gueler’s team has started arbitraging JGB contracts between Simex and Tokyo...” (Footnote: ²¹⁰). The relevant member of Mr Gueler’s team was Leeson, who was given a trading book for this purpose. It seems that this new activity was reported to the risk committee as a fait accompli which Mr Baker and Mr Gueler had allowed to start without any prior authority and without any specific limits in place.

On 20 May 1994, Leeson commenced switching in Euroyen and was given a further book for this purpose. This new activity does not appear to have been minuted (let alone sanctioned) by the Risk Committee.

As noted above, Leeson’s commencement of JGB switching in April 1994 was marked by a second startling explosion in the magnitude of his reported profits, with revenues in his first week of over £1.8m from JGB switching alone.

Mr Gueler admitted that Leeson’s JGB profits were “really fantastic from the word go”. He described Leeson’s reported profits in April 1994 as “very amazing, something very spectacular” and agreed that they “beggar[ed] belief” (Footnote: ²¹¹). Mr Gueler accepted that Leeson’s JGB revenues were reported to him on a daily basis (Footnote: ²¹²).

On 16 June 1994, the BSL Risk Committee approved risk limits for the switching business proposed by Ms Walz and Mr Gueler. They were “intra-day” limits only, meaning that positions had to be “flat” overnight: in other words, Leeson was not authorised to carry an open position overnight. The risk limits applied to open (exposed) positions as they occurred from time to time during the day on the two exchanges; no gross limits were imposed (Footnote: ²¹³). Thus, for example, Leeson was not authorised to hold an open position at any time during the day in excess of 200 Nikkei 225 futures contracts (expressed in Osaka terms), or 400 Nikkei 225s (expressed in SIMEX terms). He was not authorised to trade in options (other than for third party customers) (Footnote: ²¹⁴).

Despite his failure to detect unauthorised trading by the Hong Kong traders in 1993, Barings continued to rely upon Mr Gueler to monitor and control Leeson’s proprietary trading. Ash Lewis made a note of a discussion she had with Ms Walz in June 1994 which records “...[Ms Walz] does not really understand what it is that [Leeson] does. However, the daily contact for [Leeson] is [Mr Gueler] in Tokyo. He probably does understand what [Leeson] is doing. The problem here is [Ms Walz’s] assessment of [Mr Gueler’s] ability to control and manage” (Footnote: ²¹⁵).

Mr Gueler admitted that Ms Walz’s and Mr Baker’s concerns as to his management skills were justified (Footnote: ²¹⁶). Nevertheless, the minutes of the BSL Risk Committee meeting of 16 June 1994 recorded that “It was agreed Fernando Gueler would be responsible for controlling the use of these limits...” (Footnote: ²¹⁷). Ms Walz’s fax to the individual traders on 23 June 1994 stated “All risk continues to report into Fernando in Tokyo, who in turn reports to me in London” (Footnote: ²¹⁸).

Mr Gueler regarded the absence of a risk controller in Singapore as a serious weakness in the system but said that Mr Baker’s attitude was that the business should be developed whether or not the controls were in place (Footnote: ²¹⁹). According to Mr Gueler, Mr Norris told him that any further pressure on the attempt to improve controls could cause the system to “fall apart” (Footnote: ²²⁰). Consequently, as Mr Gueler admitted, there was no one else better placed than him to assume this responsibility and he “dutifully” did so (Footnote: ²²¹).

Mr Baker told the Singapore inspectors: “Nick would talk every day to Fernando. People seem to forget Fernando’s role in this. He had been there for two or three years and his role was to talk to Nick every day about his intra-day risk. That was his managerial responsibility. He was in the same time zone as Nick, he was on the other end of the market phone and Nick in the pit, and Fernando’s job was to monitor the intra-day risk” (Footnote: ²²²).

Not only was Mr Gueler explicitly responsible for controlling Leeson’s use of his intra-day limits but BSJ also checked that Leeson’s overnight positions were flat, calculated his daily profit and loss and reported on these matters to London. According to Mr Norris the risk control function in Japan (Mr Clarke and Mr Sue) had a responsibility “to monitor end of day positions on the books managed by Mr Gueler and his team [which included the SIMEX activity], to verify prices and other aspects relevant to marking the positions to market, and to follow up on anomalies” (Footnote: ²²³). The fact that Leeson’s position was apparently flat at the end of the day was not the end of the matter. As Mr Norris explained “... an exceptional result, either way, in relation to the size of the book or the nature of the commodity being transacted, would be something I would have expected people to follow up on” (Footnote: ²²⁴).

The responsibilities on the part of BSJ for Leeson’s business were described in the internal audit report of the Tokyo FPG (Footnote: ²²⁵), which was distributed to Mr Gueler and other members of senior management in November 1994. Mr Norris said that this report gave a good description of how he understood matters were organised (Footnote: ²²⁶). The report said [emphasis supplied]:

“The Equity Derivatives Group in the Tokyo office, managed by Fernando Mr Gueler, is part of the Structured Products Group, headed by Mary Ms Walz... Nick Leeson, who trades futures and options on Simex, also reports to Fernando...

The Tokyo Risk department is responsible for the daily collation and reporting of trading positions for business executed by both Equity Derivatives and Nick Leeson in Singapore. Tokyo Risk department are primarily responsible for monitoring all trades against the risk limits as approved by the Risk Committee, independently verifying revaluation prices which are supplied by the front office, and reporting risk positions and MOF [Ministry of Finance] capital utilised in a summary report for the daily London Risk Committee meeting. The department was headed by Kevin Clarke until 1 November when he transferred to the Equity Derivatives trading team. Kevin has been replaced by his assistant, Vincent Sue...

.... detailed risk monitoring of the Simex business is undertaken by Tokyo Risk department and in addition Fernando is kept informed of all positions and p&l on a daily basis.”

That report was based, at least in part, on conversations between the internal auditors and Mr Gueler. Despite seeing a draft copy of the report in order to make management comments, there was no suggestion from Mr Gueler at the time that it was inaccurate. In evidence all Mr Gueler could say was that the report was wrong (Footnote: ²²⁷).

Leeson was given his own trading book for his Nikkei switching as from 23 September 1994. From that date his Nikkei revenue was reported separately from that of Mr Brindle’s volatility book.

As set out above, Leeson’s reported profitability during 1994 reached very high levels. Mr Gueler admitted that Leeson had “many exceptional results” and that someone within Barings should have investigated them (Footnote: ²²⁸) but no one did so.

Other than the risk control role performed by Mr Clarke and Mr Sue from Japan, there was no independent middle office or risk controller dedicated to Leeson’s switching business. Mr Gueler accepted that the absence of any form of local monitoring in Singapore meant that the system was “defective” and that an “important control” over Leeson’s switching business was missing (Footnote: ²²⁹).

By October 1994, market volatility had become heavily depressed and, as Leeson described it in a conversation with Mr Gueler, the market on SIMEX was “very boring” (Footnote: ²³⁰). Nonetheless, when Leeson returned from holiday in the middle of October, he reported a record profit of over £4m for the week.

For several months before October, Mr Gueler had become increasingly concerned about the size of Leeson’s profits. He told me that he had “many concerns” about Leeson’s business and, as the profits grew, these concerns intensified (Footnote: ²³¹). By October 1994, Mr Gueler was so uncomfortable with Leeson’s levels of profitability that he believed Leeson must have been abusing customer orders to generate such profits (Footnote: ²³²).

Mr Gueler explained that Leeson’s position and combination of customer and proprietary business gave rise to a compliance-related concern. He said this had been “growing through [him] for several months” because it seemed to him that everyone else was “turning a blind eye to the fact that [Leeson] is taking a large spread from the client” and Barings might get sued as a result (Footnote: ²³³).

Through the summer of 1994 Mr Gueler had been concerned at the fact that Leeson was making 2 ticks on every Nikkei trade. By October Leeson was taking 3-4 ticks from his Nikkei trades which Mr Gueler said “seemed too big” (Footnote: ²³⁴). Mr Gueler said that he had not appreciated that Leeson was taking 7 ticks per JGB trade by October 1994 but agreed that, if he had noticed this as he should have done, he would have been even more concerned (Footnote: ²³⁵). Leeson’s October results were so exceptional that Mr Gueler telephoned Ms Walz and, in his own words, told her that he was “concerned about how [Leeson] is making this money” and that he could not “easily sleep at night” through worry about how the money was being made. He said that he believed Leeson’s trading should be carefully looked at “before it all ends up in tears” (Footnote: ²³⁶). Such was the level of Mr Gueler’s concern that, as he communicated to Mr Baker and Ms Walz, he “wanted nothing more to do with [Leeson]” (Footnote: ²³⁷).

Although Mr Gueler said that his concerns were “very serious”, he produced no analysis to support them, did not instigate an investigation and did not mention them to either Mr Killian or Mr Norris (Footnote: ²³⁸). In retrospect, Mr Gueler now says that he wished he had communicated his concerns to Mr Norris and he accepted that he should have produced some “concrete” evidence to make Mr Baker and Ms Walz take his concerns seriously (Footnote: ²³⁹). In the event, Mr Baker and Ms Walz just brushed Mr Gueler’s concerns aside.

Despite being put on notice that the senior derivatives trader in the group had serious concerns over the propriety of Leeson’s activities, neither Mr Baker nor Ms Walz did anything to investigate or analyse Leeson’s trading. Nor did Mr Baker take any steps to ensure that Leeson’s role was properly segregated, even though he admitted to the Singapore inspectors that he knew Leeson’s position was “fundamentally wrong in principle” (Footnote: ²⁴⁰).

During January 1995 Leeson reported revenues amounting to £11.278 million, which represented 68 per cent of the Bank Group’s total revenue for that month of £16.5 million (Footnote: ²⁴¹). Again, there were some amazing apparent results, including £4.9m from two days’ trading on 24 and 25 January. His total reported income for the seven weeks in which he traded in 1995 was over £17m.

I have described at paragraph 200 above Leeson’s opening of the 88888 account on 3 July 1992. Leeson used the 88888 account (Footnote: ²⁴²):

By (ii) and (iii) above, Leeson caused his trading results to appear far better than they in fact were, by creating losses in the 88888 account. Since there is no evidence that Leeson himself benefited directly from his unauthorised trading, his motive for conducting it appears to have been largely the desire to appear a successful trader, with the rewards in prestige and bonuses that that position would bring.

Leeson traded options legitimately on behalf of BSL and BSJ (including on Mr Brindle’s volatility book) and their customers. He was not authorised to trade options otherwise. However from October 1992 he was a consistent seller of options on the Nikkei 225 futures contract, booking them to the 88888 account.

This may in part have been speculative trading which Leeson hoped would be profitable. Certainly in 1994 and 1995 his strategy was to sell straddles, by which he effectively gambled upon the markets moving less than was generally expected – which was why the Kobe earthquake (which caused a sharp fall in Nikkei futures and a rise in JGBs) caused him substantial losses.

However it seems clear that in large part the reason for Leeson’s sale of options was to conceal his losses on the 88888 account. Leeson could not use the sale of options to fund his other trading: as explained at paragraph 102 above, the seller of an option in fact had to pay out more in margin than he received in premium (at least until the market moved in favour of the option and the margin requirement dropped). But in the short term the premiums did allow him to clear the debit balance on the 88888 account at month-ends and other reporting dates. The method for achieving this is described at paragraph 122 above.

The cumulative losses resulting from trading booked to the 88888 account were:

I consider below, first, the means which Leeson adopted to conceal his unauthorised trading and, secondly, the reasons why he was able to succeed in doing so.

Using his knowledge of the BSL back office, on 8 July 1992 Leeson persuaded Dr Wong, the responsible director of the company which provided the CONTAC system, to adjust the CONTAC computer system so as to omit information recorded on the 88888 account from the trade feed and the London Gross Report to BSL (Footnote: ²⁴⁴). Possibly as a result of an oversight, it was not omitted from the margin feed but, because BSL’s First Futures system did not have an account under that number, the system ignored the information as to the 88888 account, which reached it by way of that feed.

The CONTAC system permitted changes to be made to the trading data received by it from SIMEX, without keeping any record of the changes. Accordingly Leeson was able on a regular basis to instruct his staff to make entries in CONTAC which facilitated and concealed the unauthorised transactions involving the 88888 account described above at paragraphs 293 and 294. The relevant entries for which the staff were responsible were as follows:

As I have explained above (see paragraph 112), BFS’ financial and statutory accounting records were maintained on the separate SUN system. At the end of each month, BFS staff manually transferred the relevant trading balances from CONTAC to SUN.

If the debit balance on the 88888 account had appeared on the SUN accounts, it would have been evident that Leeson was using the account for unauthorised trading. Therefore he had to reduce the balance to zero at month-ends. At financial year-ends the need was even greater, due to the presence of the external auditors.

Leeson used a number of techniques to achieve this. The main ones are described below.

At 30 September 1992 the deficit on account 88888 stood at ¥664 million (S$9.2 million). Leeson eliminated this from the month- and year-end balances by arranging for BSL to send to him ¥670m, which was received into BFS’ bank account on 1 October. BSL entered the payment in its books on that date, but Leeson caused it to be credited to the 88888 account in BFS’ books as received on 30 September. The result was that the deficit in the 88888 account was cancelled out, leaving a small credit balance of ¥5.6m for which Leeson forged a confirmation from BSL. This transaction was the subject of one of the allegations of negligence against D&T and will be examined in much more detail below.

Leeson used an exactly similar technique to eliminate the balance on the 88888 account at 31 December 1992, though the losses at that date were lower than they had been in September. Leeson arranged for BSL to remit ¥395 million to BFS for value 4 January 1993, which was credited to the 88888 account on 31 December (Footnote: ²⁴⁵).

In most months from February 1993 onwards, Leeson caused false accounting entries to be made in BFS’ books to cancel out any balance on the 88888 account. This was vividly described by Miss Hassan in her evidence (Footnote: ²⁴⁶). Initially Leeson put through the entries himself, but from 1994 he gave Miss Hassan a standing instruction to make an internal journal entry after doing the closing for the last day of the month (and therefore after calculating the balance on the 88888 account). The entry credited to the 88888 account an amount equal to the balance on that account (thereby reducing the balance to zero) and debited the same amount to the “bank funds receivable” account in CONTAC. Leeson directed the BFS staff to print out the report of end-of-month balances from CONTAC for entry into the SUN system after these fictitious entries had been made, and then to reverse them the following day.

The “funds to be reverse” entry which I describe below in connection with the 1994 audit reflected such a false accounting entry, which reduced the balance on the 88888 account to zero at the October 1994 month-end.

Leeson used the sale of options to conceal losses at some month- and year-ends, including December 1993. As explained at paragraph 116 above, the option premium received on the sale of an option was treated by CONTAC as an asset and could be credited to the 88888 account so as to reduce or cancel out the accumulated losses on that account. Only if the option was exercised by the buyer did it result in a debit to the equity balance.

Premiums from the sale of options were credited to the 88888 account to eliminate deficits at the end of April, May, August and December 1993 (and therefore at the time of the 1993 audit), and March and June 1994.

As at 31 December 1994 the deficit on the 88888 account, even after crediting option premiums to it, was over ¥7.778 billion (£49.9 million). Leeson concealed this by causing a false option trade to be booked to the 88888 account, crediting to it a premium of ¥7.778 billion. The small debit left on the account was cancelled as an “out trade resolution”.

Leeson sought to deal with the distorting effect this fictitious trade had on BFS’ settlement variation balance with SIMEX by instructing Linda Hassan to ignore the false trade when preparing the reconciliation of that balance at 30 December 1994. However this meant that the settlement variation balance shown on the Broker Reconciliation sheet was different from that in BFS’ books. It was this discrepancy which was noticed by C&LS in conducting their year-end audit and which Leeson tried to explain by his story of the “SLK receivable”, described at paragraph 462 below.

As General Manager of BFS, Leeson was at all material times in day to day practical control both of BFS’ trading activities - its “front office” - and of its administrative activities - its “back office”. He was BFS’ head trader and he was also in charge of the paperwork, including the recording and settling of trades. This allowed him to alter the records of his own trading and to create records of fictitious trades. In particular, it enabled him to arrange for there to be a nil balance on the 88888 account in CONTAC at month- and year-ends, when data was transferred to the SUN system for inclusion in management and statutory accounts.

As well as Leeson combining responsibilities which ideally should have been segregated, he was, as will be discussed below, effectively unsupervised. The executive directors of BFS, Mr Bax and Mr Jones, did not supervise either his trading or settlement activities. At most, Mr Jones provided financial and administrative back-up. Mr Killian and Mr Gueler, whom Mr Jones believed to be supervising Leeson’s trading activities, were not doing so. Responsibility over control of margins and settlements, which both Mr Jones and the auditors believed to be exercised by BSL’s settlements department in London, was denied by Mr Bowser and wholly negated by the Dollar Funding which I describe shortly.

The final ingredient was BFS’ inexperienced staff who did what Leeson told them to without question and, apparently, in all innocence. The evidence of all the BFS staff called as witnesses in these proceedings (the only experienced trader employed by BFS, Eric Chang, was not called as a witness) was to the same effect: they were new to futures trading, they believed that, if Leeson told them to record the entries as he did, it must be all right, and they never suspected that anything improper was occurring. I discuss the actions of the BFS staff at paragraphs 301 above and 996 to 1003 below.

The combination of these three factors allowed Leeson to conceal his unauthorised trading for nearly three years.

As explained at paragraph 93 and following above, BFS had to deposit margin with SIMEX in respect of all its trading on the exchange, authorised and unauthorised, so that all trading losses had to be paid as they arose.

From the start of BFS’ operations, BFS had available six sources of funding (Footnote: ²⁴⁷):

Of these, Leeson was able to use (ii), (iii), (v) and (vi), at least, to fund his unauthorised trading. Accordingly, as of December 1994 Leeson had available approximately US$150m to fund his unauthorised trading (the sum of (ii), (iii) and (vi) above), even if one ignores the possible use of treasury bills, any profits on options trading and the Dollar Funding described below.

Until early September 1993, the difference between initial and maintenance margins on authorised trading was sufficient to fund the losses on Leeson’s unauthorised trading (save for a point in April 1993, when Leeson had to draw on the other sources indicated). However from September 1993, Leeson’s losses always exceeded this difference (Footnote: ²⁵¹).

For a few months Leeson was able to carry on by using the other sources of funds described above, supplemented by use of BFS’ overdraft with Citibank. This had been subject to a limit of S$1.65 million when established, but by February 1994 the limit had been increased to US$50 million. However that was a daylight facility, available only against pre-advices of incoming funds during the day. Leeson’s use of the overdraft to fund continuing negative balances on the 88888 account in late 1993 and early 1994 was a breach of its terms and led to a letter of complaint from the Citibank manager to Mr Jones on 18 February 1994 (Footnote: ²⁵²).

Accordingly in early 1994 Leeson had to find a further source of funds for the ever-growing losses on the 88888 account. His solution was to request from BSL, and later also from BSLL, the Dollar Funding which I discuss in detail below. Leeson continued to use the excess of margins and other sources of funds identified above, but requested Dollar Funding to meet the difference between those funds and the amount required for margin on BFS’ positions, including the 88888 account (Footnote: ²⁵³). At the time of the collapse the Dollar Funding amounted to US$639 million (£403 million) (Footnote: ²⁵⁴).

One final source of funding which I should mention for completeness, but which became important only after December 1994, came from balances held on behalf of BSJ. These consisted of margin on trades conducted by BFS on behalf of BSJ and profit on those trades which BFS did not repay to BSJ. In relation to the switching business, there was an equal and opposite loss-making trade on a Japanese exchange for every profitable trade on SIMEX. When BFS failed to pay over the profits on the SIMEX trades, which occurred especially in January and February 1995, BSJ ended up borrowing heavily from its banks in order to fund its own loss-making positions (Footnote: ²⁵⁵). At the date of the collapse the unrepaid surplus owed by BFS to BSJ amounted to ¥20 billion (£133 million) (Footnote: ²⁵⁶).

By comparison with these figures, the total amounts held by BFS on behalf of Barings companies, legitimately and illegitimately (i.e. margins on positions, Dollar Funding and withheld profits), at the time of the collapse were:

As discussed above, by early 1994 Leeson’s losses on the 88888 account meant that he needed a new source of funding. He obtained it by taking advantage of a recent innovation in BSL’s margining systems.

Until late 1993 all customers had had to put up margin in yen for trades in yen-denominated instruments on SIMEX and the Japanese exchanges. In 1993 SIMEX began accepting margin in US dollars and some customers pressed BSL to allow them to put up margin in dollars. Accordingly in November 1993 BSL began to accept margin in dollars from customers for SIMEX trading and to forward that margin in dollars to BFS (Footnote: ²⁵⁸).

To accommodate this process, BSL opened on 5 January 1994 a separate ledger account for SIMEX US dollar collateral, named BSINGCOLL (Barings Singapore Collateral). The opening balance on the account was US$160,843, which was the net amount of dollars held as of that date as a result of the US dollar margin transactions up to then.

Until 1 March 1994, all except one of the US dollar payments made by BSL to BFS and entered on the BSINGCOLL account seem to have been initiated by BSL and to have related to customers who wished to put up margin in US dollars. However when requesting yen margin through the funding spreadsheet, BFS did not give credit for US dollars paid by BSL (Footnote: ²⁵⁹). Thus Leeson obtained the margin twice over, at the cost (since BSL did give its customers credit for the dollars) of BSL. Leeson was able to use the excess to fund his unauthorised trading (Footnote: ²⁶⁰). As of 28 February 1994, the net amount involved was at least US$25 million (Footnote: ²⁶¹).

BFS’ failure to give credit for the collateral went unnoticed by BSL until 23 June 1994, when, as described below, Mr Railton raised the point with Leeson but then appears to have let it lapse.

In early 1994, Leeson began to request from BSL additional funding, in US dollars, which was not linked to individual customers (“the Dollar Funding”) (Footnote: ²⁶²). When the trial started, the date of the first such request was agreed to be 12 January 1994. However the issue was the subject of further analysis as the trial proceeded. That suggested that until 9 February all payments of US dollars from BSL to BFS were attributable to BSL passing on to BFS collateral received in US dollars from BSL’s customers.

The first US dollar payment which was not of that type, that is, was paid from BSL funds, was a payment of US$32 million on 9 February. That seems to have been linked to a fax of the same date from Mr Jones to Mr Hawes, asking for an immediate credit line of US$50 million. This was required to cover late payments due to “teething problems with the settlement system, plus increased margin calls re Chinese New Year holidays imposed by SIMEX” (Footnote: ²⁶³).

Before me, Mr Hawes did not recall having done anything in response to this fax other than to discuss the daylight overdraft facility with Citibank and to reply to Mr Jones that there was a US$70m facility in place. However this payment was not put to Mr Hawes, and it seems unlikely that Mr Leeson would have made a separate and unrelated request for US$32 million on the same date as Mr Jones’ fax. The existing daylight facility would not have met the difficulty referred to in Mr Jones’ fax if Mr Jones told Mr Hawes that the funds were needed for more than one day. Furthermore, unlike the later Dollar Funding payments requested by Leeson, the US$32 million was deducted from BSL’s normal yen margin due on that day. The balance of probability is therefore that on the 9 February 1994 Leeson persuaded Mr Jones to make a request for US$32 million, in order to make good a shortfall of funds available to Leeson to finance his unauthorised trading.

The US$32 million was repaid by BFS eight days later as part of a somewhat larger sum.

Mr Jones was involved again, on 1 March 1994, when he and Leeson sent a fax to Mr Hawes (Footnote: ²⁶⁴). They posited an option transaction for a client between SIMEX and Osaka which would generate an overdraft at Citibank in Singapore of US$50 million for one day, and asked to discuss it as soon as possible. Whether as a result of that or not, on 2 March BSL paid BFS US$50 million, which was repaid the following day.

Accordingly Mr Jones seems to have been involved in persuading Mr Hawes to make the dollar payments on 9 February and 2 March. However BSL also paid BFS US$40 million on 1 March, for which BFS was able to offer no explanation. It was not matched to any client receipt by BSL and it was not offset against the normal yen margin payment. This therefore seems to have been the first “classic” Dollar Funding payment, in the sense of a straightforward request for money by Leeson alone, without Mr Jones’ involvement and not linked to any client.

There is no clear evidence, but it seems likely from the above circumstances that Leeson used Mr Jones to accustom BSL to paying out unreconciled cash, and then Leeson felt able to start making further calls for cash on his own. During the rest of March, Leeson made regular calls for round sums of dollars, on top of the margin in yen requested through the funding spreadsheet, and not-quite-so-regular repayments. The first such request which survives is an e-mail, dated 10 March 1994 (Footnote: ²⁶⁵). It reads in full as follows:

“Brenda, Due to the holiday in Singapore on Monday, SIMEX have asked for additional margin. Kindly send us USD 16,000,000 to our a/c with Citibank for good value 11 March 1994.

Kindly confirm.

Can you check also whether you paid us USD 14,000,000 for value 9 March 1994 since there was no such funds received on that date.

Regards, Linda.”

This pattern became standard and continued on a regular, later daily, basis, until the collapse (Footnote: ²⁶⁶). The reasons given for individual requests were brief – “margin call” or “Advance margin call – House” – but frequently no reason at all was given. On the other hand, on many days BFS’ e-mail was a notification that BFS would be returning funds to BSL. The later e-mails, at least, were addressed to eight members of staff in the BSL settlements department (including Ms Granger and Mr Railton), and to one member of staff in Group Treasury. They arranged for the monies requested to be transferred to BFS’ account with Citibank in Singapore.

As described below, at first the Dollar Funding was treated as if it were funding customer margins, but in time came to be split between house and customer margins. None of the purported customer element of the Dollar Funding was reconciled by BSL to its own customers, so no attempt was made to recover it from customers. Instead, the customer element was simply booked by BSL in the BSINGCOLL account, as a convenient account (Footnote: ²⁶⁷). At the end of December 1994, the BSINGCOLL account recorded net payments totalling US$ 201 million (Footnote: ²⁶⁸). By the date of the collapse, the figure was US$ 486 million (Footnote: ²⁶⁹).

The house element was remitted to BFS through BSLL (although the funds were ultimately provided by BB&Co) and recorded in BSLL’s books in a BSSHSECOLL account. At the end of December 1994, the BSSHSECOLL account stood at US$ 5 million (Footnote: ²⁷⁰). By the time of the collapse, the figure was US$141 million (Footnote: ²⁷¹).

The Dollar Funding appeared clearly in one accounting record produced by BSL. During the “solo consolidation” of BB&Co and BSL in early 1993 (described at paragraph 44 above), a working group chaired by Mr Hawes prepared a “solo balance sheet”. This covered BSL’s core business which would result from the intended amalgamation. When the solo balance sheet was prepared in early 1993, it identified a difference between the fiduciary assets held by BSL in connection with customer business and BSL’s corresponding liabilities: BSL had more fiduciary assets (cash deposited with exchanges and at the bank) than it had received from customers.

In order to make the fiduciary assets and liabilities balance in BSL’s solo balance sheet, and as a temporary expedient, in April 1993 the working group inserted a liability identified as “loan contra” into the fiduciary liabilities section of the balance sheet (shown on line K2). A balancing asset described as “customer loans” (shown on line P4 of the balance sheet) was inserted in the banking assets section. Thus, the difference was described as the “K2/P4 line” (Footnote: ²⁷²).

An initial study, described below, concluded that the K2/P4 line resulted from delays in the receipt of margin from customers. Therefore BSL was effectively making loans to customers by “topping up” their margin requirements. Hence the balance was known within BSL as the “top-up” balance. As at 31 December 1993 it was £21.7 million.

Mr Hawes in his evidence before me suggested an explanation for the K2/P4 balance in 1993 which had not until that point occurred either to him or to the others who considered the question in 1993 to 1995. This was that the balance sheet reflected margin and cash collateral paid by customers to BSL, but not non-cash collateral such as shares or treasury notes (Footnote: ²⁷³). Customers who had deposited non-cash collateral against some or all of their margin requirements would not be required to put up margin in cash to that extent, and would not be treated by Mrs Granger as debtors when they did not do so. Yet, to the extent that the non-cash collateral was not deposited with an exchange but was retained by BSL, there would be an imbalance between the cash received from customers and that paid to exchanges on customers’ behalf. It is true that in 1993 the non-cash collateral exceeded the K2/P4 line but that simply indicates that not all the collateral deposited by customers with BSL was being actively used as security against which to trade futures and options (Footnote: ²⁷⁴).

It is indeed surprising that this did not occur to those considering the issue at the time or subsequently until Mr Hawes gave his evidence, and was not even mentioned in his witness statement. In any event, the “non-cash collateral” represented securities held by BSL from a number of different clients whose individual debit or credit position may well have varied widely. There was no evidence of any kind why the deposits were made and in particular whether they were actually made as security for margin payments made by BSL on behalf of clients.

Whatever the explanation for the K2/P4 line in 1993, when the Dollar Funding began in March 1994 the element of it purporting to relate to customer margin was directly reflected in the K2/P4 line: as the Dollar Funding did not relate to genuine customer trades, equivalent funds were not received from customers. As the Dollar Funding increased, so did the shortfall shown on the K2/P4 line. On 30 June 1994 it was £120 million. As at 24 February 1995 it stood at £306 million (Footnote: ²⁷⁵). Attached at Schedule H to this judgment is a graph showing the growth in the K2/P4 line over the course of 1994 (Footnote: ²⁷⁶).

In a sense Mr Hawes’ recent possible explanation of the K2/P4 line in 1993 is irrelevant to assessing his actions in 1994, as he was not then aware of it. But if his explanation is correct, and if BSL had investigated the K2/P4 line further at that time, they need only have compared the total margin required from customers (a total certainly available from First Futures) with the total margin paid by customers plus any debtors. The difference would represent the extent to which customers were trading against the security of non-cash collateral. Comparing that difference with the K2/P4 line would demonstrate that by March 1994 non-cash collateral explained only part, and a decreasing part, of the K2/P4 line. The remainder was the Dollar Funding.

Unfortunately the solo balance sheet for BSL had a restricted circulation (Messrs Hawes and Sacranie and their assistants). Mr Hawes told me that there were other balance sheets produced in a different form for use by management, one for the solo consolidated group and one for BSL alone (and no doubt others). They had a wider circulation. They showed the item as part of “advances (up to 7 days)”. Only in the case of the balance sheet for BSL might that trigger the question why a securities company was lending so much money. Mr Hawes thought Mr Norris and Mr Maclean saw the BSL balance sheet but was not sure. ALCO saw only the consolidated group balance sheet, into which the BB&Co, BSL and Treasury pages were consolidated, and the K2/P4 figures were subsumed in an omnibus figure for loans. One can understand that it would raise few questions if a banking group were making loans (Footnote: ²⁷⁷).

BSL’s credit committee did not review the item at all. According to Mr Hawes, this was because the customers who were allegedly receiving the benefit of the sums being lent had not yet been identified – as of course they never were, which was precisely the problem (Footnote: ²⁷⁸).

The story of BSL’s investigation of the Dollar Funding is complicated by the three strands running through it. The first strand in time is the investigation in April and May 1993 into the K2/P4 line, thrown up by the solo consolidation exercise. The second is the enquiry as to the reasons for the Dollar Funding, which Leeson began requesting on 1 March 1994. The third is a series of attempts beginning in May 1994, to obtain a reasonably accurate division of how much had been, or was being, advanced for the purpose of financing house trading and how much to finance customer trading. The principal reason for these latter enquiries was concern that Barings’ large exposure reporting to the Bank of England did not reveal accurately the exposure of BSL to SIMEX as a result of margin deposited for house trading.

As I have explained, the solo consolidation working group created the K2/P4 line in the BSL solo balance sheet as a temporary expedient in April 1993, to make the totals for fiduciary assets and liabilities balance. The working group commissioned Miss Henderson, BSL’s financial controller, and Miss Thomas, the compliance officer, to establish the reasons for the difference between those totals.

Miss Henderson and Miss Thomas concluded in May 1993 that the difference was attributable to delays in receiving margin from customers, delays in transferring commission from customer to office account, and other minor factors. The working group asked Miss Henderson to carry out further work on the issue but she was made redundant not long afterwards. Although Mr Hawes believed that Miss Henderson’s successor was continuing the exercise, that proved not to be the case and no further investigation of the reason for the K2/P4 line took place for over a year.

On 24/25 February 1994, Mr Hawes paid a one-day visit to Singapore and met Mr Jones and Leeson. His evidence in his witness statement, confirmed in cross-examination, was that he did then discuss the need for the Dollar Funding, and was given similar explanations to those which Leeson gave him later in 1994, and which I consider below (Footnote: ²⁷⁹). Since Mr Hawes gave evidence, the parties have carried out the further analysis of the US dollar payments to BFS in January and February 1994 to which I have referred above. It now seems that the only request for Dollar Funding made prior to Mr Hawes’ visit was Mr Jones’ fax of 9 February. Therefore it seems likely that Mr Hawes during his visit discussed only the factors that had prompted that fax, but not the pattern of Dollar Funding which became established in March.

Whether or not Mr Hawes discussed other reasons for the need for Dollar Funding in February 1994, he does seem to have discussed with Mr Jones and Leeson the time required to send yen from Tokyo to Singapore. Leeson said that this required BFS to put up margin itself against BSJ’s positions until the yen arrived. Mr Hawes formed the view that this was a potential reason for BFS requiring funds from BSL and for the K2/P4 line, though he did not put the latter point expressly to Mr Jones and Leeson (Footnote: ²⁸⁰).

I have described above the start of the Dollar Funding on 1 March 1994 and have quoted the earliest surviving request on 10 March. It appears from Mrs Granger’s evidence to the Singapore inspectors that she discussed with Leeson the reasons for the Dollar Funding and he regularly gave her the explanation that it was needed to fund advance margin called by SIMEX because of a holiday. In any event, the Settlements Department and Treasury seem at first to have accepted whatever reasons Leeson gave without having asked any questions. Their reaction was similar to that of Mr Bowser to Leeson’s requests for ¥670m and ¥395m in September and December 1992, i.e. unquestioning payment.

Mrs Granger was not called to give evidence. She told the Singapore inspectors that she began to discuss the Dollar Funding with Leeson, Mr Hawes and those working for Mr Hawes in the summer of 1994. The only specific conversation with Leeson of which she gave evidence was, she said, between the third and fourth quarters of 1994. In the course of that conversation, Leeson told her that the Dollar Funding was required to fund BSJ’s house trading: she said that she could not remember specifically “if he just said advance margin calls or just Tokyo was unable to move funds quickly enough” (Footnote: ²⁸¹). Leeson told her that BFS had no credit lines of its own, so it came to London for funds. He described London as “the cash cow”.

At the time of this conversation or earlier, Mrs Granger was clear that the Dollar Funding was an inter-company loan to BFS. Her evidence to the inspectors was that she knew from the start, and told all those from Credit and Treasury who asked her, that she had no debtors among her agency customers. Therefore loans to them could not account for the K2/P4 balance. She believed that she had explained that to Mr Hawes and that he had approved the loans, as she had told him of Leeson’s explanations of why he needed the money and debated with Mr Hawes which account it should come from. Mrs Granger’s concern was, not that the funding was being called, but that it was being treated as customer funds: although BFS called BSJ a customer, actually it was house and so she felt that the money for their trading should come out of the house account (Footnote: ²⁸²).

Mrs Granger told the inspectors that she did not know that BSJ was actually funding its own positions, but that Treasury must have known (Footnote: ²⁸³).

Mr Railton’s evidence before me supported Mrs Granger’s account. He said that within the Settlements Department the Dollar Funding was regarded as inter-branch funding, meeting a number of costs of BFS’ doing business. Treasury and those above Mr Railton in Settlements had agreed the Funding should be paid. No one in Settlements checked whether BSJ was actually funding its own positions or SIMEX was making advance margin calls, until Mr Railton went out to Singapore in February 1995 (Footnote: ²⁸⁴).

Mr Hawes’ evidence was that he knew that BSJ was meeting its funding obligations (Footnote: ²⁸⁵) (and from October 1994 at least BSL had reports from BSJ of the funds they placed with BFS). Therefore, throughout 1994, he saw the Dollar Funding as 80-90% an agency problem and correctly treated as customer funds. But he suspected that, at the edges, the Dollar Funding was partly providing short-term funding for BSJ positions, due to the delay in transmitting yen from BSJ to BFS, and that that was misreported due to sloppy management and book-keeping by BFS (Footnote: ²⁸⁶). He agreed in evidence that such factors would not result in the BSJ element amounting to a figure as large as US$50m, though that is the figure for BSJ funding shown in the first funding breakdown supplied by Leeson to BSL, as of 14 June 1994 (Footnote: ²⁸⁷).

One of the many extraordinary features of this story is that Mrs Granger and Mr Hawes worked very near to each other, and clearly discussed the existence of the Dollar Funding repeatedly over a period of a year, yet they seem to have held completely different and unreconcileable views as to the reasons for it. Mrs Granger believed it to be an inter-company loan to fund BSJ house business. Mr Hawes knew that not to be the case, other than to a minor extent, and told me that such a loan would have been structured quite differently, as funding direct from Treasury (Footnote: ²⁸⁸). For his part, Mr Hawes believed the Dollar Funding to be mostly funding agency business. This was so even though Mrs Granger had told him that the Funding could not be in relation to BSL agency clients and BFS’ one significant non-BSL agency client, BNP, could not possibly have accounted for the volumes of Funding seen by mid-1994. Mr Hawes told me that, although “concerned”, he believed that Mrs Granger’s department did know the reasons for the Dollar Funding requests:

“Q. What reconciliation did you understand the settlements department was doing, at that stage [before June 1994]?

A. I do not think that is a question I ever asked myself, what reconciliations they were doing. I trusted them to be doing their job and to be – and to know for what purpose the dollars were required.

Q. So that is what you mean when you say that you did not have any idea about how little they really knew?

A.

Yes.” (Footnote: ²⁸⁹)

Yet surely even the briefest of conversations with Mrs Granger would have established that the settlements department did not know what the Funding was for, beyond an explanation which Mr Hawes knew to be incorrect.

For the first ten months Dollar Funding requests were framed as requests to pay margin to BFS’ segregated (i.e. customer) account in Singapore. The funds to meet these requests were transferred out of a BSL segregated (i.e. customer) account in London and paid to BFS. Thus the requests were treated as if they related to agency business.

In April 1994, when reviewing the large exposure report which was about to be submitted to the Bank of England, Mr Hawes realised that it understated the Group’s exposure to SIMEX on house business. This was because BFS placed the margin for such business in its customer account with SIMEX, and described it as customer margin in its “enclosure B” reports to BSL. Such margin was therefore ignored in preparing the Group’s large exposure report. This issue could not be resolved in time for the March 1994 report. Mr Barnett had to explain the problem to the Bank of England in a letter of 29 April 1994, and promise to establish a system which would assess accurately the exposure to SIMEX.

The result was firm instructions from Mr Cooke in BSL’s regulatory department to Leeson on 5 May 1994, instructing him to distinguish clearly in BFS’ enclosure B reports between margins for outside customers and those for Barings companies (Footnote: ²⁹⁰). Leeson’s response was from then on to split the money and other securities deposited with SIMEX in the customer account between house and customer, on the basis that the deposits of non-cash securities covered the house exposure (Footnote: ²⁹¹). This was at best an inaccurate indication of the extent to which the customer account included house exposure (Footnote: ²⁹²).

According to Mr Railton’s evidence, Treasury asked Settlements in June 1994 to try to break down the Dollar Funding (Footnote: ²⁹³). Mr Railton had a further concern, which was whether BSL was receiving credit from BFS for genuine customer dollar collateral which, as described above, BSL had paid to BFS. At least some of the dollar funding was attributed by Leeson to initial margin, and so should have been deducted from the yen margin required (Footnote: ²⁹⁴). Hence on 15 June 1994 Mr Railton e-mailed BFS, asking for a breakdown of the US$ collateral held by BFS and quoting the amount of that collateral which represented US$ paid to BSL as collateral by customers (Footnote: ²⁹⁵).

Possibly as a result of this query, BFS sent on 16 June 1994 its first purported breakdown of the dollar funding. This purported to allocate the total funding provided (including the amount requested) between advance margin, funding of particular customers, etc (Footnote: ²⁹⁶). Such breakdowns continued to be sent intermittently after that, and from November 1994 they became regular. From November, BSL seem to have attempted to monitor them by comparing each daily funding breakdown with that for the preceding day (Footnote: ²⁹⁷). However although one or two discrepancies were raised with Leeson, the fact that the funding in each category tended to stay the same from day to day and was for round figures which were unlikely to be accurate passed almost without comment until January 1995 (Footnote: ²⁹⁸).

On 20 and 23 June 1994, Mr Railton sent further messages to BFS, directed at the issue of why BSL was not receiving credit for the dollars deposited as collateral. Mr Railton could not recall receiving any satisfactory reply to these, but his enquiry seems to have lapsed at that point.

Meanwhile Mr Hawes began to make his own efforts to understand the K2/P4 balance and the Dollar Funding. His initial evidence was that in June 1994 he asked Miss O’Donoghue, Head of the Credit Department, to investigate the K2/P4 balance. She traced it to Singapore and the Dollar Funding, and she asked Leeson for the reasons for the funding. (Footnote: ²⁹⁹) Mr Hawes said that she passed these reasons on to him, but they differed little from the reasons Leeson had given to him in February.

Miss O’Donoghue’s evidence to the Singapore inspectors was that Mr Hawes simply asked her to split Leeson’s first funding breakdown between house and customer business. In order to do this, she spoke to Leeson on either 17 or 29 June 1994 and obtained a split. She passed that, but not Leeson’s explanations for the funding, to Mr Hawes. She made no connection with the K2/P4 line, which she had not seen since moving to Credit in September 1993 (Footnote: ³⁰⁰). Miss O’Donoghue’s evidence was put to Mr Hawes in cross-examination and he adhered to his own recollection, but without great conviction (Footnote: ³⁰¹).

Mr Hawes told the Singapore inspectors that it was only in June 1994, when Miss O’Donoghue made her call to Leeson, that he made the connection between the increasing K2/P4 balance and the Dollar Funding. Before me, as I explain at paragraphs 850 to 852 below, he said that he made the connection in March or April 1994 (Footnote: ³⁰²). In the light of Miss O’Donoghue’s evidence, I think that must be correct. Nor does there seem any reason for Mr Hawes, who had full knowledge of both the K2/P4 line and the Dollar Funding payments as they were made, to take four or five months to make the connection. Accordingly I accept D&T’s submission that Mr Hawes knew of that connection in or before April 1994. His later request to Miss O’Donoghue was not a request to establish the reason for the K2/P4 line but was rather a request to split the Dollar Funding breakdown supplied by Leeson between client and house for regulatory purposes.

Mr Hawes’ evidence was that, having received Miss O’Donoghue’s account (and on the assumption that his version of her task was correct), he realised that investigation in Singapore was needed. Treasury had only limited resources and so Mr Broadhurst suggested that Mr Hawes should ask the BSL internal audit team, who were about to audit BFS, to look at the issue. On 22 June 1994 Mr Hawes therefore told Mr James Baker, the auditor, of the high level of long-standing funding and asked him to investigate it (Footnote: ³⁰³). At the time it seemed to him a rational decision to use this means of investigating the issue in a non-confrontational way, given his lack of resources, though he accepted that by the end of June 1994 the K2/P4 balance had increased sharply over the last three months (to £120m).

Mr James Baker’s evidence, which was supported by his contemporaneous meeting note (Footnote: ³⁰⁴), contradicts this account. Mr James Baker said that Mr Hawes raised with him only the difficulties caused by requests for funding received at short notice, which were not reconciled at the time of the call. Mr Hawes did not tell him of the size of the balance (Footnote: ³⁰⁵).

I prefer Mr Baker’s account of this conversation. It is consistent with his contemporary note and with his subsequent failure to investigate the issue of funding when in Singapore. Such a failure would be quite inexplicable for an internal auditor who had been warned that the company he was about to audit held over £100 million of unreconciled funds.

Meanwhile, the regulatory department told Mr Hawes in August 1994 that BFS had not properly implemented Mr Cooke’s request to separate out the customer margin into genuine customer margin and margin deposited on behalf of Barings companies, for large exposure purposes. Therefore on 8 September 1994 Mr Hawes repeated the request and queried why the house element was rising so sharply. Leeson’s response simply referred to the large positions taken, by BSL and BSJ. Before me, Mr Hawes could not recall how the September large exposures report (Footnote: ³⁰⁶) was prepared. He told the Singapore inspectors that Leeson explained in September 1994 that the margins for house trading had been covered by the deposited securities and so had not been mentioned in the margin reports. Now house exposure had exceeded the securities, but the BFS staff had become accustomed to not needing to report house margin (Footnote: ³⁰⁷).

Independently of Mr Hawes, the Settlements Department was asking for breakdowns of the Dollar Funding on 21 September, and received the usual uninformative list of round numbers, on 22 and again on 29 September. They queried why the numbers did not agree with the figures in the London Gross report, but not why many of the figures did not change over the six days (Footnote: ³⁰⁸).

The internal audit report was not published until October 1994, though Mr Hawes started seeing drafts in mid-September. A recommendation in an earlier draft of the report, that margin calls by BFS should be reconciled with those made by SIMEX on BFS, was omitted from the final version. Nor did the report deal with the explanation for the Dollar Funding. It limited itself to suggestions as to the use of letters of credit to decrease the proportion of margin deposited by BFS in cash and to recommending a comprehensive review of BFS’ funding by Group Treasury.

Mr Hawes gave evidence that he discussed with James Baker the reasons for the report’s failure to deal with funding. James Baker told him that he had been dissuaded by Leeson, with support from Mr Broadhurst (Footnote: ³⁰⁹), from including the recommendation as to reconciling margin calls. Even in Mr Hawes’ account, there was no discussion of the reasons for the Dollar Funding. James Baker’s evidence was that he just told Mr Hawes he had not been able to make any progress on funding. Mr Hawes told me that he discussed with Mr Hopkins sending James Baker out to Singapore again, but decided not to as he was himself due to go there in two or three weeks (Footnote: ³¹⁰).

Mr Hawes visited Singapore at the beginning of October 1994, following a visit to Tokyo. He missed Leeson, who claimed to have been detained in London, and neither Miss Hassan nor Mr Jones was able to provide any help in understanding the funding issue (Footnote: ³¹¹). Therefore Mr Hawes met Leeson in London on his return.

They discussed the reasons for the Dollar Funding. The reasons for the Funding which Leeson gave were the same reasons as (according to Mr Hawes) he had given previously. Mr Hawes admitted that he was “only just about convinced” by these explanations (Footnote: ³¹²). They were general and unsupported. In summary they were (Footnote: ³¹³):

However Mr Hawes did not probe any of Leeson’s explanations, then or later in 1994 (Footnote: ³¹⁴). He said that this was because he hoped to obtain approval of his proposal for a regional treasurer in the Far East, who would be able to investigate the funding question on the ground. Unfortunately this did not come about, though accounts differed as to whether the reason was a veto by Mr Norris on cost grounds or the resignation of the candidate, Mr Ashley.

Mr Hawes admitted in cross-examination that the focus of his meeting with Leeson was in fact not the reasons for the Dollar Funding. Rather, it was the need for the requests to be split between house and customer, so as to allow for accurate large exposure reporting (Footnote: ³¹⁵). The result of the meeting was that from November 1994 BFS did break down the Dollar Funding requests between customer and house business. However the breakdown was often 50:50, and the figures were often round numbers. Moreover, it frequently happened that Leeson would give one figure in the morning for the funds he required, but when the Dollar Funding request arrived later in the day the figure would be different (Footnote: ³¹⁶). Mr Hawes said that he looked at the requests only occasionally, regarding them as no more than an approximation to the truth, and so did not notice that they tended to be split 50:50. He distrusted the information coming from BFS, but felt only investigation in Singapore could resolve the issue (Footnote: ³¹⁷). It followed from this that Mr Hawes should have realised that the large exposure reports continued to be inaccurate. But he took no further steps to improve the quality of information coming from Singapore for inclusion in them until 27 January 1995.

Mr Hawes told me that his next attempt to investigate came in December 1994 when Jeremy Stunt, an accountant in the Finance Department, was carrying out an investigation into financing costs generally. Mr Hawes said that he asked Mr Stunt also to investigate the reasons for the Dollar Funding, but Mr Stunt did not carry on to do that. Mr Hawes’ evidence on this point conflicts with that of the others involved: Mr Railton, who assisted Mr Stunt in the project, said that it was always limited to the cost of funding (Footnote: ³¹⁸), and Mr Stunt’s evidence to the Singapore inspectors and his SFA witness statement supports that. Even Mr Hawes admitted that the cost of funding was the main focus. He told me that even the limited exercise which Mr Stunt carried out did show that the money was there in Singapore deposited on SIMEX, which gave him some comfort (Footnote: ³¹⁹). That comfort ought to have been limited, since it did not follow that money deposited as SIMEX margin would return from SIMEX to BFS in due course.

In January 1995, Mr Hawes sent to BFS the project specification for a “Singapore Project”, one object of which was “identification of the causes of the very large and consistent requirement for funding from London under the heading ‘Loans to Futures and Options Customers’”. He told me that he intended the project to be carried out by Mr Ashley, whom he and Mr Hopkins were hoping to take on in a project role after the idea of a Regional Treasurer had fallen through (Footnote: ³²⁰). When Mr Ashley resigned the next day, Mr Hawes had no one to conduct the investigation. Mr Bax simply passed the project to Leeson, who produced an obfuscatory explanation which Mr Jones co-signed without carrying out any investigation himself (Footnote: ³²¹).

Mr Stunt’s exercise did have the effect that, in the course of assisting with it, Mr Railton started examining and comparing the funding requests from Singapore and asking BFS some pertinent questions. He realised that the numbers on the requests did not move and could not be accurate. His enquiries expanded during January 1995 to the point where the BSL Settlements Department realised that, at the least, the BFS Settlements Department was preparing the requests extremely badly (Footnote: ³²²). The result was that Mr Railton was sent out to BFS at the start of February 1995 to help out in the BFS back office and to try to improve the quality of the information being sent, in particular the funding requests.

Leeson did his best to keep Mr Railton occupied with other projects during his three weeks in Singapore. However despite this Mr Railton established that the Dollar Funding requests were meaningless and that there was a large shortfall between (i) the funds deposited by BFS at SIMEX and the banks and (ii) the figure shown by BFS’ accounts as owing to BSL, BSLL, BSJ and BNP. His questioning about this shortfall was a major factor causing Leeson to flee to Kuala Lumpur.

It is striking that, while Mr Hawes and others were making these ineffectual efforts to understand the Dollar Funding, no one thought to look at a print-out of the Margin Feed. As explained above at paragraphs 137 to 139, this set out the margin required for each BFS account. Had anyone looked they would have found that this included margin required for 88888 account transactions which were not being reported on the trade feed or the London Gross Report.

Mr Railton told me that he did not use the feed to answer queries about Dollar Funding because he did not see one as relating to the other. Even though the Dollar Funding was described in Leeson’s breakdowns as partly initial margin, if BSJ was paying some and BSL paying some it would not help to find out the total BFS margin from the feed. Furthermore advance margin would not have appeared on the margin feed at all. But he accepted that the feed should have been explored (Footnote: ³²³).

Mr Bowser told me that he and others in the Settlements Department did not realise that the margin feed could be viewed. Nor did they realise that the feed set out all BFS’ margin requirements, including margin payable by BSJ and BNP, and therefore contained more information than was in the customer statements obtained from First Futures (Footnote: ³²⁴). It is not clear whether this was also Mrs Granger’s perception: she admitted to the Singapore inspectors that she probably knew that the margin feed split the margin between accounts and she could not say why no one looked at it when trying to reconcile the dollar funding (Footnote: ³²⁵).

The BOBS inquiry and the pleadings in this action assumed that the first Dollar Funding payment was one on 12 January 1994. As I have explained at paragraphs 325 to 329 above, it now appears that the position was more complicated than that. Most of the payments in January and February 1994 were genuine payments of customer funds but, because BSL paid twice over, still resulted in Leeson obtaining funds which he could use for unauthorised trading. The same applies to the two payments in which Mr Jones seems to have been involved. The first “classic” Leeson-initiated Dollar Funding request was on 1 March 1994, and there were no further BSL-initiated payments after that date.

The duplication of customer payments in January and February and Mr Jones’ intervention on 9 February 1994 made available to Leeson for his unauthorised trading a balance of at least US$25 million at the end of February. The “classic” Leeson-initiated Dollar Funding then began on 1 March. Including both the January and February payments and the Leeson-initiated Dollar Funding from March onwards, the total available to Leeson was US$125 million on 15 March. The total then dipped, but recovered to around US$100 million in mid-April. It again dipped until mid-May, but from 20 May onwards was well in excess of US$100 million. From 8 September 1994 and for the remainder of the year the total remained over US$150 million. As at 30 December 1994 the figure was US$205 million. It then increased rapidly in January, and especially February, 1995 (it more than doubled between 15 and 24 February) and at the collapse was US$639 million (£403 million) (Footnote: ³²⁶).

Leeson was in charge of both trading and settlements at BFS. It is a notable, though no doubt completely understandable, aspect of the Barings collapse that, according to their evidence since the collapse, none of those who might have been thought to have had immediate responsibility for Leeson in either capacity admit to having had any such responsibility and no one else has put themselves forward as responsible.

It was never envisaged that anyone in Singapore would supervise Leeson on the trading side. D&T contended in these proceedings that Leeson had a direct reporting line to, and was supervised by, Mr Killian, in relation to his agency trading (from June 1992 until 1 January 1995) and by Mr Gueler (throughout). Mr Killian, Mr Gueler and BSJ denied this, save that Mr Gueler was expressly given responsibility for monitoring Leeson’s intra-day risk limits between June and October 1994. For the reasons I give at paragraphs 1034 to 1040 below, I conclude that D&T are correct, but that the issue of who precisely should have supervised Leeson’s trading does not affect the outcome of the case.

It is accepted that Mr Baker took overall responsibility for Leeson’s proprietary trading from January 1994 onwards, and for Leeson’s agency trading from January 1995.

As I describe at paragraph 975 and following below, virtually the whole of the rest of the Barings Group seems to have believed that Mr Jones was supervising BFS’ settlements and finance sides, save for Mr Jones himself and the BFS staff. Mr Jones played no part in supervising the settlements side. On the finance side he signed the year-end accounts and dealt with minor accounting and administrative matters, but did little more than that. Miss Yong’s role in BFS, when she was appointed in 1993, was similarly limited.

I have described above the relevant reporting lines within Barings and the uncertainties involved. I have also mentioned that Mr Heath and the entire management of the derivatives business left Barings in around March 1993. This included Mr Baylis, the only member of senior management who it seems may have had some direct knowledge of derivatives trading. Therefore after that time there was no one in the senior management of Barings who understood derivatives trading or what Leeson was apparently doing.

In this section I describe the steps taken by those to whom Leeson reported (or should have reported) and others to analyse his apparent trading profits. Any steps taken were of course ineffective until it was too late. I deal later in this judgment with the consequences of such failure by management.

I can deal briefly with BFS’ local management in Singapore. Both Mr Bax and Mr Jones gave evidence that they did not regard themselves as having any responsibility for supervising Leeson’s trading activities (Footnote: ³²⁷). They did not question (or, to be fair, profit directly from) his activities or the profits he made. Mr Jones was intended to be in overall charge of compliance at BFS (Footnote: ³²⁸) and of the settlement of trades between BFS and SIMEX, but in practice did nothing (Footnote: ³²⁹). It is one of the extraordinary aspects of this story that, at all times relevant to D&T’s audits, Mr Jones’ desk was on the same floor, a few yards away from the BFS settlements staff, and Mr Jones had responsibility for BFS’ settlements with SIMEX. Yet the settlements staff appear (whether knowingly or not it is not clear. It was not pleaded, nor was it put to them, that they were complicit in Leeson’s fraud) to have systematically falsified BFS’ records of its trades on SIMEX over a period of two and a half years without Mr Jones knowing anything about it.

The only occasions which Plc/BSL were able to put forward, even at the apogee of their case, as analyses undertaken by Leeson’s product managers as to how he was making his apparent substantial profits were the following (Footnote: ³³⁰):

Mr Norris accepted before me that none of these could be regarded as detailed analyses of Leeson’s profitability, such as he accepted that Mr Baker and Miss Walz should have made and such as he believed at the time they had made (Footnote: ³³⁷). He was not aware of any analysis actually done by Mr Gueler. In the case of Mr Baker’s alleged analysis at (iii), Mr Norris saw no document recording such an analysis. The only evidence for it is Mr Baker’s witness statement. Since BFS did not call him as a witness, that is of no evidential value.

By contrast, Mr Killian’s evidence to me was that in 1994, when rumours were circulating in Tokyo about Leeson earning $3m in a day from arbitrage, arbitrage traders in other firms whom Mr Killian knew told him that was impossible:

“I imagine I heard somebody say ‘Oh, Nick Leeson had a $3 million day’. I would say that to a friend of mine… and this person would say ‘No way, no way you can do that. Do not even listen to that type of story because it is not even possible’.” (Footnote: ³³⁸)

As a result Mr Killian attributed Leeson’s profits to high levels of directional trading, assisted by BFS’ level of customer orders. He presumed this was authorised by Mr Baker. But Mr Killian said that the Chinese wall between those involved on the agency and proprietary sides precluded him from asking. Therefore he did not know if this was the correct explanation, or what Leeson’s risk limits were. Mr Killian said that Leeson’s product managers should have gone to Singapore and, in his words, said “OK, now show me. Make $2m today. Let’s see how it’s done.” His proprietary product manager should have done that in order to understand such a high source of profits, and his agency product manager (and Mr Killian denied he had that responsibility) should have done so in order to ensure that the profits were not coming from unfair exploitation of customers (Footnote: ³³⁹).

Mr Gueler also questioned how Leeson could be making his profits, though his concern was that Leeson was charging customers too much for his liquidity arbitrage. As I have described, he went to Ms Walz with his concerns “about how [Leeson] is making this money” and told her that he could not “easily sleep at night” through worry about how the money was being made (see paragraph 284 and following above). He felt that Mr Baker and Ms Walz should understand how Leeson was making his profits. But, differing from Mr Norris, Mr Gueler did not feel that that was his job (Footnote: ³⁴⁰).

Finally there were the departments within Barings which might have been thought to have some responsibility for monitoring funding of, and risk from, the Group’s activities. BFS had no risk manager of its own (Rachel Yong was due to take over the role on 1 March 1995). The switching business fell within the purview of BSJ’s risk managers, Messrs Clarke and Sue, but Mr Norris’ evidence was that they fulfilled a statistical role, collecting and reporting trade data. For that purpose the risk manager relied on the data supplied by BFS and did not verify it independently or analyse the profits. Rather, Mr Norris said that it should have been Mr Gueler and the BSJ traders who understood Leeson’s trading and could have spotted anomalies. They failed to do so (Footnote: ³⁴¹).

The BSL Risk Committee does not seem to have carried out any review. Likewise the profits were discussed, but apparently not questioned, at MANCO and EXCO (Footnote: ³⁴²). Eventually, in 1995, ALCO and MANCO became concerned, but only about the size of Leeson’s positions and whether the bank could fund them, not about the genuineness of the profits (Footnote: ³⁴³).

Deloitte Haskins & Sells in London were for a number of years auditors to the Barings Group. As Barings established companies in Singapore, Deloitte Haskins & Sells, Singapore were appointed auditors to them (Footnote: ³⁴⁴).

In 1989, most of the component firms of Deloitte Haskins & Sells merged with the component firms of Touche Ross. This included the Singapore firm, who became D&T. An exception was the UK, where the UK practice of Deloitte Haskins & Sells merged with Coopers & Lybrand to form C&LL.

The result of this merger activity was that in 1990 C&LL audited the Barings companies in London and co-ordinated the group audits, while D&T audited the BSL subsidiaries in Singapore (including BFS). C&LS were appointed auditor of BB&Co’s Singapore branch in 1990. This situation continued until 1994, when C&LS succeeded in winning the poisoned chalice of appointment as auditors of the other Barings companies in Singapore.

BFS commenced active trading on SIMEX on 1 June 1992, though for the first month it used Chase Manhattan to execute trades for it (Footnote: ³⁴⁵). Its revenues, consisting of commission and interest, for the period of four months up to 30 September 1992 (the reporting date for the September 1992 audit) were S$3.83 million (£1.35 million) and its profit before tax was S$3.05 million (£1.08 million). Since all BFS’ trades were booked to clients, these results bore little relation to the results of BFS’ trading.

D&T had been appointed as BFS’ auditors by an exchange of letters dated 6 and 15 October 1986, when the company was first incorporated. D&T’s engagement letter set out their statutory functions as auditors and then stated:

“To enable us to express our opinion [on the statutory accounts], we shall make such tests and enquiries as we consider necessary. The nature and extent of our tests will vary according to our assessment of the company’s systems of internal accounting control. We shall report to the directors or to the appropriate level of management any material weaknesses in the company’s systems of internal accounting control which come to our notice and which we believe should be brought to their attention.

Our audit is designed, in accordance with normal practice, to enable us to express an opinion on the accounts. It should not be relied upon to disclose defalcations or other irregularities, although their disclosures, if they exist, may well result from the audit tests we undertake.

The foregoing does not cover maintaining the accounting records and the preparation of accounts, these being the responsibility of the company’s directors.” (Footnote: ³⁴⁶)

Neither the letter of appointment or D&T’s engagement letter mentioned D&T’s responsibilities under the Singapore Futures Trading Act or Futures Trading Regulations.

Chaly Mah had been the D&T partner responsible for the audits of the BSL Singapore companies since 1988, and therefore was responsible for the 1992 audit of BFS. In 1992 the audit juniors were Koo Mei Yen (“Miss Koo”) and Chang Suland, and the audit manager was Teo Pin Ghee (Footnote: ³⁴⁷).

In 1992 C&LL were the auditors to the BSL Group of which BFS formed part. C&LL sent to D&T on 21 July 1992 group audit instructions relating to the BSL subsidiaries in Singapore (BSS, BFS, Baring Management Services Pte Ltd and Barings Nominees Singapore Pte Ltd) and an audit procedures questionnaire. The audit was for the year ending 30 September 1992, although BFS had been active only for the last four months of that period.

D&T replied on 17 August 1992, sending to C&LL a completed Audit Planning Memorandum (“APM”) and audit time and costs budgets. The APM was prepared by Jenny Rodgerson of D&T, apparently after a visit to BSS and BFS (Footnote: ³⁴⁸).

While separate time and cost budgets were prepared for each company, and the APM described each company separately, issues such as risk assessment and systems and controls were dealt with jointly, and the comments on those issues were apparently directed wholly towards BSS. That was by far the larger company in terms of number of employees and turnover. In particular, the APM indicated that D&T would adopt a controls-based approach, which was actually the approach adopted for BSS, but not for BFS. The budgeted hours for the audit of BFS were 85, compared with 400 for that of BSS. The APM stated that D&T “have identified no specific risks in any audit areas”, and the covering fax reported that D&T “have not identified any significant audit/accounting issues that need to be brought to your [i.e. C&LL’s] attention” at that interim stage (Footnote: ³⁴⁹).

Despite the confusion in the APM, it is very clear when one looks at D&T’s working papers, and it was confirmed by the evidence of Mr Mah, that D&T adopted a substantive approach to the audit of BFS rather than a controls-based approach. A substantive approach involved obtaining assurance as to the correctness of the financial statements by direct verification of the balances in those statements and of the transactions and documents underlying them. If adopting a controls-based approach, an auditor obtains assurance by testing the extent and operation of the internal controls which should detect or prevent errors in the financial statements.

The procedure followed by D&T in carrying out the 1992 audit is described in detail in Schedule F to this judgment. D&T adopted a substantive approach to their audit, not because of any perceived audit risk or control weakness in BFS, but because they considered that it was the most cost-effective approach in the light of the simplicity of the business being conducted by BFS. They understood BFS to be an execution-only broker whose only customers were to be four associated Barings companies placing orders for execution on SIMEX on behalf of themselves or their customers.

In accordance with C&LL’s timetable, D&T sent to C&LL on 16 October 1992 the audited consolidation schedules for each company, an audit report for each company, an Audit Summary Memorandum and the completed Audit Procedures Questionnaire. The working papers, dated October 1992, for the Audit Summary Memorandum include the Risk Assessment: “No specific risk has been identified during the planning stage and the course of our audit.” (Footnote: ³⁵⁰)

The consolidation schedules were audited financial statements, prepared in accordance with UK accounting standards but using the materiality level adopted for the statutory accounts (Footnote: ³⁵¹). They were to be consolidated into the BSL Group accounts, and they needed to be sent relatively early in the audit process in order to allow time for C&LL to use them in preparing those Group accounts.

The audit report for BFS (Footnote: ³⁵²) was addressed to the directors of BSL. It stated:

“2.

Subject to the matters noted in paragraph 4 below, in our opinion all the consolidated schedules, as attached, were prepared in accordance with the [BSL] group accounting policies and consolidation instructions. The information in them is presented fairly in conformity with accounting practices generally accepted in the United Kingdom”.

3.

We intend to give an unqualified audit opinion on the local statutory accounts which will be based on the consolidation schedules, adjusted as necessary to comply with local legal or fiscal requirements or accounting practices generally accepted locally; and to include any adjustments considered necessary when the matters in paragraph 4 are resolved.

4.

Material unresolved matters

(a)

The bank confirmation for this year ended September 30 1992 is outstanding.

(b)

The confirmations for the intercompany balances are outstanding as at September 30 1992.

(c)

Local statutory accounts.”

The Auditing Procedures Questionnaire (Footnote: ³⁵³) did not distinguish between the various Singapore companies. It confirmed that D&T had carried out the audit properly. It included a positive answer to the question “Have you evaluated the adequacy of controls within the accounting system and identified whether reliance on their operation may be possible?” and a negative answer to the question “Are there any weaknesses in the company’s systems which are so significant that in your opinion they should be brought to our attention?”

D&T then carried out the further work necessary to submit the required regulatory reports to the Singapore regulators, MAS and SIMEX. This included the further checks necessary in order to complete Forms 17, 21, 22 and 23 prescribed by the Singapore Futures Trading Act, reviewing BFS’ correspondence with MAS and SIMEX, and completion of certain further checklists and programmes (Footnote: ³⁵⁴).

On 20 November 1992, D&T sent BFS’ statutory accounts in draft to Mr Jones, along with a draft representation letter addressed to D&T to cover the 1992 audit which they asked him to sign and return. They repeated the request on 10 December 1992.

On or about 17 December 1992, Mr Jones returned the signed representation letter. Apparently he had had it retyped on BFS letterhead, but he did not change the text from D&T’s draft. This representation letter and one in similar terms signed by Mr Jones in the course of the 1993 audit were the subject-matter of the preliminary issue which I have described at paragraphs 19 to 21 above.

On 31 December 1992, D&T issued their opinion on the statutory accounts of BFS. This stated that they had audited the financial statements in accordance with the relevant auditing standards and that, in their opinion, the financial statements of BFS were properly drawn up in accordance with statutory requirements and gave a true and fair view of the state of affairs of BFS. D&T also submitted their regulatory reports to MAS and SIMEX.

D&T’s fee for the 1992 audit was S$8,300 (around £2,900), though they had in fact spent over 50% more hours than they had budgeted for (Footnote: ³⁵⁵).

SIMEX carried out a regulatory audit in the spring of 1993 and wrote to BFS with their conclusions on 7 September 1993 (Footnote: ³⁵⁶). One of the points raised was that on two days in February 1993 BFS had called insufficient margin from BSL, through ignoring certain subsidiary accounts and illegitimately taking into account inter-market offsets. It had also taken no action when BSL failed to respond in due time to a margin call. SIMEX fined BFS S$23,000 (Footnote: ³⁵⁷).

In 1994 SIMEX conducted a more limited audit. It wrote to BFS on 16 January 1995, noting breach of the rules as to segregation and computation of customer funds. Further action was overtaken by the collapse of the Barings Group (Footnote: ³⁵⁸).

The 1993 audit covered the period 1 October 1992 to 31 December 1993 (BSL and BFS having changed their year-end to December). BFS’ revenues for the 15 months were S$27.9 million (£11.75 million) and its profit before tax S$22.6 million (£9.5 million) (Footnote: ³⁵⁹). As in 1992, since all BFS trades were booked to clients, these results bore little relation to the results of Leeson’s trading.

D&T’s 1993 audit of BFS followed a similar pattern to their 1992 audit. C&LL sent their audit instructions to D&T on 27 October 1993 and an Audit Procedures Questionnaire two days later. Since 1992, BSL had been “solo-consolidated” into the Plc group. Therefore the instructions were those for the audit of the Plc group and D&T were required to produce consolidation schedules for incorporation into the financial statements of the Plc group.

Mr Mah remained the partner in charge of the audit. Miss Koo was the audit senior and Philip Ng the audit junior. Teo Pin Ghee had left D&T and so Mr Mah acted as audit manager to the extent necessary (Footnote: ³⁶⁰).

On 1 November 1993 D&T submitted their time and cost budget for the audit. They estimated 134 hours in total for the audit, almost exactly the same as they had spent in 1992 (Footnote: ³⁶¹).

D&T sent their Audit Planning Memorandum (Footnote: ³⁶²) to C&LL on 1 December 1993. This again covered both BSS and BFS. It noted that:

“…management’s attitude towards control is good. Mr James Bax (Managing Director) and Mr Simon Jones (Finance Director) are actively involved in the day to day operations of the companies.

Pressure to ensure that proper controls are in place also comes from the regulatory bodies, SES [the Stock Exchange of Singapore] and SIMEX…

The head office in London oversees the local operations and monthly management reports are sent to the head office.” (Footnote: ³⁶³)

In the light of such factors D&T concluded that the engagement risk for the Baring Securities (Singapore) Group was “Normal”.

Planning work and substantive field work began in November 1993 and were completed in January 1994, save for the regulatory reports (Footnote: ³⁶⁴). As in 1992, D&T carried out a substantive audit. Again D&T relied upon the confirmation requests prepared by BFS to check customer balances, and verified that they received confirmations for all balances.

Mr Jones signed at D&T’s request a representation letter covering the 1993 audit on 27 January 1994. This was in identical form to that for 1992.

On 28 January 1994, D&T sent to C&LL audit reports addressed to the directors of Plc, matters for the attention of partners, consolidation schedules and completed audit questionnaires (Footnote: ³⁶⁵).

The audit report was in similar terms to that for 1992, save that the only outstanding matter was expressed to be bank confirmations. The Auditing Procedures Questionnaire was in 1993 specific to BFS, but gave the same reassuring answers as in 1992. As in 1992, the working papers, dated 15 January 1994, include a risk assessment: “No specific risk has been identified during the planning stage and the course of our audit” (Footnote: ³⁶⁶).

As in 1992, D&T conducted further tests necessary for the submission of regulatory reports. This included following up with BFS the points raised by SIMEX in its audit letter of 7 September 1993.

On 28 February 1994 D&T issued their opinion on the statutory accounts of BFS (Footnote: ³⁶⁷). This stated that they had audited the financial statements in accordance with the relevant auditing standards and that, in their opinion, the financial statements of BFS were properly drawn up in accordance with statutory requirements and gave a true and fair view of the state of affairs of BFS. D&T also submitted their regulatory reports to MAS and SIMEX.

D&T’s fee for the 1993 audit was S$17,400 (around £7,400), as compared with their estimate in November 1993 of S$9,500 (Footnote: ³⁶⁸).

In July 1994 the Internal Audit Department of BSL conducted an internal audit of BFS, as part of its review of the South Asia region. Prior to his departure for Singapore, James Baker, the internal auditor, spoke to, among others, Mr Hawes on 22 June 1994. Mr Hawes raised with him concerns as to the lack of segregation of Leeson’s functions, as James Baker recorded in his meeting note:

“Nick Leeson has too dominant a role looking after both trading (agency and proprietary) and settlements aspects of the business; there is no deputy to challenge him… TH believes that SJ basically leaves NL to his own devices. While he has no evidence to suggest that NL has indeed abused his position, the potential for his doing so needs examining.” (Footnote: ³⁶⁹)

Mr Sacranie was also concerned about separation of responsibilities (Footnote: ³⁷⁰). As discussed at paragraphs 371 to 373 above, Mr Hawes also raised BFS’ need for funding at short notice, though not the inability to reconcile or explain the Dollar Funding.

James Baker spent three weeks at BFS conducting the internal audit, from 19 July 1994 onwards.

After some delay in producing it, the final version of the Internal Audit Report was completed in October 1994. The Internal Audit Report (Footnote: ³⁷¹) made three key recommendations in relation to BFS, as follows:

(1)

"BF(S)’s back office should be reorganised so that the General Manager [i.e. Leeson] is no longer directly responsible for the back office" (Footnote: ³⁷²);

(2)

"BF(S)’s trading activities should be independently reviewed to ensure that regulations are followed and risk limits observed. A suitably experienced manager should be appointed to review the records, perform some tests of detail and discuss activity with BF(S)’s traders" (Footnote: ³⁷³); and

(3)

"London Group Treasury should perform a comprehensive review of BF(S)’s funding requirements" (Footnote: ³⁷⁴).

None of these recommendations made in the internal audit report was carried into effect.

So far as recommendation (1) above is concerned, the Executive Summary of the Internal Audit Report (Footnote: ³⁷⁵) stated that:

"there is significant general risk that the controls could be overridden by the General Manager. He is the key manager in the front and back office."

The main body of the report noted the mitigating factors which reduced the concentration of power in Leeson’s role (including reconciliation controls between BFS and its customers). It stated that:

“Given the lack of experienced and senior staff in the back office, we recognise that the General Manager must continue to take an active role in the detailed operations of both the front and the back office”, including double-checking the back office’s output and “roles which do not involve the settlement and recording of transactions, such as liaison with the SIMEX authorities and the arrangement of funding in conjunction with Group Treasury”.

But, after the recommendation noted above, it stated:

“Specifically, the General Manager should not: retain sole responsibility for the supervision of BF(S)’s back office team; retain cheque-signing or journal-passing powers; review and sign off [reconciliations].”

Later the report recommended:

“Other responsibilities relinquished by the General Manager should be taken on by the BS(S)’s Director of Finance and Operations. The main requirement is to ensure that the settlement and recording processes are adequately supervised including, for example:

Daily contact with the futures settlement supervisor who should be refer [sic] all significant matters arising in the office for discussion;

Daily review and sign off of SIMEX reconciliations;

Occasional review of daily checks of trades recorded on the system to SIMEX reports and trade tickets; and

Occasional review of standards maintained over standard procedures…

Management Response: Nick Leeson, Simon Jones

As agreed with Internal Audit these are not normal circumstances for BF(S) considering the current absence of third party customers. Should these emerge, the role of General Manager will obviously change. However, with immediate effect the General Manager will cease to perform the functions itemised… The Director of Finance (BS(S)/BS(F)) will ensure the adequate supervision of all settlement and recording processes.”

Despite this management response, no action was taken and Leeson remained until the collapse in control of both the front and the back offices at BFS.

As to recommendation (2) above, no such detailed review of trading activities was ever carried out. After an initial suggestion from Mr Jones that Mr Bowser should perform the role of risk manager for BFS, Miss Yong was appointed to the position, but had not started work by the time of the collapse.

As to recommendation (3) above, no investigation was made into Leeson’s funding requirements.

I deal later in this judgment with the evidence given before me as to why these recommendations were not implemented.

The 1994 audit, conducted by C&LS, covered the 12 months to 31 December 1994. BFS’ gross revenues for the year were S$30.587 million (£13.385 million) and its profit before tax S$22.041 million (£9.645 million) (Footnote: ³⁷⁶).

C&LS were appointed as auditors of BFS by an exchange of letters dated 11 July and 1 August 1994 (Footnote: ³⁷⁷). C&LS’ formal letter of engagement was not signed before the crash. The details of the 1994 audit were covered in opening, but not in evidence (being left for Phase 2) so what follows is subject to the hearing of oral evidence in due course.

C&LS had a preliminary meeting with Mr Bax, Mr Jones and Miss Yong in June 1994. During September 1994 they reviewed D&T’s working papers relating to the 1993 audit, and Seet Wee Teong (“Mr Seet”) of C&LS then met Mr Mah to discuss the audits (Footnote: ³⁷⁸).

In August 1994 Mr Seet and Khoo Kum Wing (“Mr Khoo”) of C&LS met the BSL internal auditors. Mr Seet’s evidence was that this was at the end of the internal auditors’ visit and they said that they had not noted any significant issues (Footnote: ³⁷⁹). James Baker does not accept that this interview took place at the end of the internal auditors’ visit. However C&LS’ notes of the interview in their audit working papers appear only to be consistent with this being the case.

C&LL sent group audit instructions to C&LS on 2 November 1994 (Footnote: ³⁸⁰). C&LS returned their Audit Strategy Memorandum (Footnote: ³⁸¹) to C&LL on 23 November 1994. This stated that “as this is our first year of audit and given the tight reporting deadline, full early work based on October management accounts would be done, except for trade and bank confirmations which would be done for year end balances”. C&LS noted that the control environment was satisfactory and stated that, in view of this and the high volume of trades, they would do a system-based audit. They stated, incorrectly, that “BFS does not engage in any House trading or discretionary trading for its clients.” (Footnote: ³⁸²) The section in the Audit Strategy Memorandum headed “Balance Sheet Review” compared the December 1993 balance sheet audited by D&T with that as at 30 September 1994 (Footnote: ³⁸³). The section showed that as at 30 September 1994 the figure for “Margin deposits with SIMEX and bank balance” exceeded that for “Margin deposits placed by customers” by S$55 million. But there is no indication that C&LS attached any significance to this.

C&LS carried out their audit testing in two stages, the first in October and November 1994 and the second in January and February 1995. They completed their fieldwork for the first stage by early December 1994 (Footnote: ³⁸⁴).

Part of C&LS’ early work was the topic of bank reconciliations. The 31 October 1994 bank reconciliation, prepared by BFS and reviewed by C&LS on 14 or 16 November 1994, included a significant reconciling item of ¥4.115 billion (S$62.5 million) which was described as “funds to be reverse”. This was in fact one of the false accounting entries routinely entered by BFS staff on Leeson’s instructions, to conceal the reduction of the balance on the 88888 account to zero at the month end. C&LS noted the item but did not investigate it further (Footnote: ³⁸⁵).

C&LS asked Rachel Yong for a copy of the internal audit report in December 1994 or January 1995. She replied that she was not aware of one, and one may not have been produced, but she would check with Mr Jones. As C&LS heard no more they assumed that no report had been produced (Footnote: ³⁸⁶).

Events after 31 December 1994, and therefore the work done by C&LS on the year-end accounts, are of marginal significance because of the cut-off on BFS’ claim described previously. I therefore just mention the most important dates:

Over the period 17 – 23 February 1995, Mr Railton, on secondment to BFS from BSL settlements, tried and failed to obtain an explanation from Leeson of a ¥14 billion discrepancy between the funds remitted to BFS by BSL, BSLL and BSJ and those held by BFS or deposited with SIMEX. On 23 February 1995 Mr Hughes of BSL sent to Mr Hawes (then in Tokyo), and Mr Hawes forwarded to Mr Jones, an e-mail expressing BSL’s inability to understand Leeson’s funding requests. Later that day, Leeson left a meeting with Mr Railton and Mr Jones which had been called to discuss the issue and fled with his wife to Malaysia.

That night Mr Hawes arrived in Singapore. He, Mr Railton and Mr Bax discovered the existence of the 88888 account and the losses on it almost immediately.

On 24 February 1995 Leeson faxed his resignation to BFS. Over the weekend of 25/26 February a team from Barings London and an official from the Bank of England arrived in Singapore. In London, negotiations to sell the Barings Group as a going concern continued through the weekend but failed, following which Plc, BB&Co and BSL were placed in administration. On 27 February SIMEX instructed BFS to stop trading and then, when it failed to pay its margin call that morning, suspended it. BFS was put into interim judicial management on the same day.

SIMEX appointed OCBC Bullion & Futures Ltd to close out BFS’ positions. This involved:

After taking into account the security and margin held by SIMEX and SIMEX’s costs of liquidating BFS’ positions, this process resulted in a payment by SIMEX to BFS of US$57 million (Footnote: ³⁹²). However BFS owed BSL, BSLL and BSJ a total of US$1,191 million (see paragraph 324 above), and its inability to repay them brought down the Group.

Just over a week later the majority of the assets and liabilities of the Barings Group (including BSJ) were purchased for a nominal sum by Internationale Nederlanden Groep NV (“ING”).

BFS alleged that D&T were negligent in the following respects:

The particulars of negligence set out for 1992 at (i), (ii), (vi), (viii), (ix), (x), (xi) and (xii) were repeated for 1993. In addition BFS alleged that:

By the conclusion of the trial the ambit of the audit case had substantially contracted. By then, the criticisms made were D&T’s failure:

In relation to the audit case, BFS relied primarily upon the evidence of Mr Swinson, the senior partner of Stoy Hayward. Its second audit expert, Mr Mason, is a former partner of Price Waterhouse in Singapore. He gave evidence only as to specifically Singaporean audit practice.

D&T relied on the evidence of Mr Spence, the London senior partner of Grant Thornton.

I have already described the events of D&T’s 1992 and 1993 audits of BFS (see paragraphs 408 to 425 and 428 to 440 above) and that in each year of account D&T elected to undertake a substantive audit. BFS do not criticise that choice. Indeed they contend that, having regard to the audit risks, both inherent in BFS’ business as a derivatives broker and arising as a result of the way in which BFS was managed during the years, and provided the audits in those years were planned and carried out with sufficient completeness, this was a wholly appropriate audit approach.

As explained at paragraph 415 above, a substantive audit involves direct verification of balances and of the transactions and documents underlying them, in order to be satisfied to the necessary standard that the client’s accounts present a true and fair view. By contrast, if following a compliance approach, an auditor tests the client’s internal controls which should detect or prevent errors in the financial statements, and thus gains the appropriate level of assurance which allows him to reduce the level of substantive testing required.

It is common ground that a substantive audit must test the completeness, as well as the accuracy, of the client’s financial statements. “Completeness” means that there are no unrecorded assets, liabilities or transactions. The main issue in this part of the case is the nature and extent of the audit procedures which D&T took or should have taken to be satisfied that BFS’ accounts for 1992 and 1993 presented a true and fair view.

As they were doing a substantive audit, D&T needed to confirm all the figures they were checking against external information. Their approach to this was to verify each item on BFS’ trial balance sheet. D&T therefore audited the account balances due to customers and the amounts due from SIMEX. Open positions resulting from transactions executed for clients on SIMEX as such did not feature on the balance sheet.

The normal means of verifying customer balances was by requesting the customers to confirm the balances. For SIMEX and bank balances the figures in the trial balance could be confirmed against the relevant SIMEX and bank statements.

In both 1992 and 1993 BFS prepared and sent out by fax confirmation requests to its five customers. There was one request for each omnibus account held by that customer: thus for BSL there was one request each for the segregated, the non-segregated and the house accounts. However in 1992 there was a separate (forged) request for confirmation of the 88888 account balance, and in 1993 a separate (genuine) request for the BSL hedge account. No satisfactory explanation was offered for why a separate confirmation of the balance on the 88888 account was thought necessary in the 1992 audit.

Each request set out figures taken from CONTAC for the account balance, initial margin, variation margin, and commission paid and outstanding, and listed all open positions on the account, all as of the accounting date. The request asked the customer to sign and return the request to BFS, to confirm the accuracy of the information. The requests were prepared and sent out by BFS staff, to be returned to BFS. BFS staff then passed the confirmations when received to D&T for use in the audit.

In 1993 BFS did not produce, and D&T did not require, an audit confirmation for the 88888 account, it is assumed because it had a nil account balance.

D&T took the confirmations received from customers and checked that the figures for “account balance” and “variation margin” on each account confirmation, all added together, equalled exactly the total figure for “client payable” in BFS’ trial balance. That total was itself checked against the client general ledger in CONTAC.

In relation to SIMEX, D&T confirmed the figures in the trial balance for amount payable to and receivable from SIMEX against CONTAC and against the SIMEX Combined Margins and Positions report (“the CMP report”). This was one of the reports produced by SIMEX daily for members, summarising their positions at the close of the day’s trading.

In 1992, Miss Koo, who did the field work for the audits, checked the open futures positions listed in the customer confirmations against the broker’s report from CONTAC. She did not check the option positions. In 1993 she checked a few, but far from all, of the open futures positions, and none of the option positions. Mr Mah in his evidence was unable to explain why she had done even this limited amount of checking, since D&T’s audit approach did not require her to verify open positions. Indeed, he said that it must have been BFS’ decision to seek confirmations of the open positions from customers in the first place, since D&T did not need them. Nor could he explain why Miss Koo had done some work on open positions for futures but not on those for options.

D&T did use the lists of open positions in 1993 for one purpose. C&LL required D&T, in their audit report to Plc and attached Matters for Attention of Partners, to provide details of any “commitments, contingent liabilities and off-balance sheet assets or liabilities” of BFS (Footnote: ³⁹³). D&T replied on 28 January 1994 that they were not aware of any such, except for “Outstanding Futures and Options contracts as listed on Appendix 2”. Appendix 2 consisted of the customer confirmations, including the lists of open positions.

In the course of their opening, counsel for D&T were able to demonstrate from D&T’s working papers that, following this course, D&T were able to confirm the figures in the balance sheets in the 1992 and 1993 accounts to the nearest yen. Counsel’s explanation is summarised at Schedule F to this judgment.

As of 30 September 1992, Mr Leeson had no open positions on the 88888 account. He had closed all positions, leaving a deficit which he had to conceal by the ¥670m window-dressing operation discussed below.

However as of 31 December 1993, Mr Leeson had sold considerable numbers of options, at least partly in order to bring the equity balance on the 88888 account down to zero (see paragraph 309 above). The result was that, as of that date, BFS had 22,109 short option positions, of which 8,560 were on account 88888 and not held on behalf of customers.

D&T did not detect these short option positions in their 1993 audit. The issue I now address is whether they were negligent in not doing so.

In analysing this issue, I set out first the parties’ cases and the relevant audit guidelines. I next consider D&T’s assessment of the audit risk posed by BFS, in the light of the risk of unauthorised trading, the lack of segregation of duties at BFS, and the various mitigating factors relied on by D&T. I then consider the effect of the audit risk on the audit procedures which D&T should have followed. Finally I deal with the question of whether those considerations required D&T to carry out the test on open positions for which BFS contended.

Among the allegations of negligence pleaded by BFS are that, in both 1992 and 1993, D&T failed to recognise that BFS’ internal controls were flawed by virtue of Leeson’s control of front and back offices; they failed adequately to take that into account in their audit; and they reported that BFS’ internal controls were adequate. However as the trial proceeded, BFS reacted to the evidence from both factual and expert witnesses that it was usual in small broking firms for one person to be in charge of both front and back offices. BFS ceased to contend that the failure to report the lack of segregation of duties was a self-contained allegation of negligence. Instead, it became an element of BFS’ case on audit risk.

BFS contended that there was an appreciable risk of unauthorised trading, that it was likely that any unauthorised trading would be concealed and that manipulation of balances was a likely means of doing that. D&T’s audit procedures were essentially limited to checks of balances, and did not allow for those risks. In addition to checking balances, D&T should have checked that all open positions entered into by BFS were recorded in BFS’ books and authorised by customers. They should have done this by reconciling the open positions recorded by SIMEX against those confirmed by customers, as recommended by Mr Swinson. BFS also contended that D&T should have controlled the confirmation process and should have confirmed the 88888 account in 1993, despite it having a nil balance.

D&T contended that, as the evidence turned out at trial, no criticism was made of D&T’s assessment of risk. Unauthorised trading was a risk, but a low risk, and D&T were not required to devise any special procedures to guard against it. The lack of segregation should have caused no concern as long as BFS was providing an execution-only service, rather than itself initiating trades, and D&T had no reason to know in either 1992 or 1993 that Leeson was initiating trades.

D&T contended that the verification of balances which they carried out would have been expected to reveal unrecorded or unauthorised trading, since any trading affects the financial balances. Such an approach would be defeated only in a case of deliberate fraud involving manipulation of balances. That sort of fraud had, reasonably, not been identified as a specific risk. Therefore D&T were not obliged to devise specific procedures, such as that devised by Mr Swinson, to test for it. No relevant audit guideline or other professional practice guidance required such a test in the absence of any indication to the auditors that BFS’ financial statements were incomplete as a result of fraud, of which there was no indication.

The applicable audit guidelines are set out in the Statements of Auditing Guideline (“SAGs”) issued by the Singapore Society of Accountants (Footnote: ³⁹⁴). The SAGs are based upon the International Auditing Guidelines issued by the International Federation of Accountants.

SAG 4, “Basic principles governing an audit”, states at paragraph 17:

“Substantive procedures are designed to obtain evidence as to the completeness, accuracy and validity of the data produced by the accounting system.

“They are of two types:

•

tests of details of transactions and balances;

•

analysis of significant ratios and trends…”

SAG 7 deals with the evaluation of internal controls. Paragraph 10 states that:

“The auditor… needs reasonable assurance that transactions are properly recorded in the accounting records and that transactions have not been omitted. Internal controls… may contribute to the reasonable assurance the auditor seeks.”

SAG 7 makes clear that, where internal controls are unreliable, the auditor should perform substantive procedures instead. It makes this point particularly in the context of small businesses “where segregation of duties is limited and evidence of supervisory controls is lacking”. In such cases “the evidence necessary to support the auditor’s opinion on the financial information may have to be obtained largely through the performance of substantive procedures.” (paragraph 27).

SAG 9 deals with audit evidence. Paragraphs 3 and 4 provide:

“3.

The audit evidence should, in total, enable the auditor to form an opinion on the financial information…

“4.

The auditor’s judgement as to what is sufficient appropriate audit evidence is influenced by such factors as:

(a)

the degree of risk of misstatement. This risk may be affected by:

(i)

the nature of the item,

(ii)

the adequacy of internal control,

(iii)

the nature of the business carried on by the entity…”

Paragraph 6 of SAG 9 explains “completeness” in the context of audit evidence obtained from substantive procedures as meaning that “there are no unrecorded assets, liabilities or transactions”.

SAG 12 deals with Fraud and Error. It states:

“5.

…the auditor seeks reasonable assurance that fraud or error which may be material to the [financial] information has not occurred… The auditor therefore should plan his audit so that he has a reasonable expectation of detecting material misstatements in the financial information resulting from fraud or error…

“6.

Due to the inherent limitations of an audit there is a possibility that material misstatements of the financial information resulting from fraud and, to a lesser extent, error may not be detected. The subsequent discovery of [such material misstatements] does not, in itself, indicate that the auditor has failed to adhere to the basic principles governing an audit…

“8.

The risk of not detecting material misstatement resulting from fraud is greater than the risk of not detecting a material misstatement resulting from error, because fraud usually involves acts designed to conceal it… Unless the auditor’s examination reveals evidence to the contrary, he is entitled to accept representations as truthful and records and documents as genuine. However the auditor should plan and perform his audit with an attitude of professional scepticism, recognising that he may encounter conditions or events during his examination that would lead him to question whether fraud or error exist…

“11.

In planning and performing his examination, the auditor should take into consideration the risk of material misstatement of the financial information caused by fraud or error.

“13.

If circumstances indicate the possible existence of fraud or error, the auditor should consider the potential effect on the financial information. If the suspected fraud or error could have a material effect on the financial information, he should perform such modified or additional procedures as he determines to be appropriate.”

Audit risk is the risk that auditors may not detect a material misstatement in the audit client’s accounts and so may incorrectly give those accounts an unqualified audit certificate. According to SAG 27, this comprises three elements: inherent risk (the risk that the accounts may be misstated, which is the product of the client’s type of business, its environment and the nature of the accounts), control risk (the risk that misstatements may not be prevented or detected by the system of internal control) and detection risk (the risk that the auditor will not detect the misstatement).

An important part of the process of planning an audit is to identify the inherent risks affecting the client and to devise procedures which will afford the auditor a reasonable expectation of detecting material misstatements arising from those risks (see SAGs 4 and 12, quoted above, among others). BFS, as a futures broker, was subject to a number of inherent risks. Many, such as the risk of errors in implementing clients’ instructions, are not relevant to this judgment. In this section I deal with two inherent risks which were alleged to have been relevant to D&T’s audits of BFS: the risk of unauthorised trading, and the lack of segregation of duties (though the latter could also be categorised as a control risk which increased other risks).

An auditor’s assessment of audit risk is relevant to the audit evidence he will require in order to obtain reasonable assurance that the financial statements are not misstated: see SAG 9, paragraphs 3 and 4, quoted above.

In their 1992 Audit Planning Memorandum, in relation to both BFS and BSS, D&T noted that “generally, management’s attitude towards control is good.... Mr Simon Jones, a director, and Mr James Bax, the managing director, are actively involved in the management of the company and exercise direct management controls”. Mr Mah’s evidence was that this referred to BFS as well as BSS. D&T concluded that “Based on our review of the various accounts, we have identified no specific audit risks in any audit areas and as such we will be performing standard audit procedures to test the various balances” (Footnote: ³⁹⁵).

In 1993, the Audit Planning Memorandum referred to tight regulation by SIMEX. It noted, for both companies, that “the local operations is [sic] directly controlled by Mr James Bax and Mr Simon Jones, directors of the local companies who are closely involved in the day to day running of the companies… We have no reason to doubt their integrity”. D&T concluded that the engagement risk for the Baring Securities (Singapore) Group (that is, the assessment of whether the auditor should accept the client’s instructions) was “Normal” (Footnote: ³⁹⁶).

The Audit Planning Memorandum made similar comments under the heading “Understanding of the Control Environment”, noting that “management’s attitude towards control is good” and that London oversaw the local operations.

In both years D&T’s working papers for the Audit Summary Memorandum include the Risk Assessment: “No specific risk has been identified during the planning stage and the course of our audit.” In both years, their Audit Procedures Questionnaire stated that they had “evaluated the adequacy of controls within the accounting system and identified whether reliance on their operation may be possible” and stated that there were no significant weaknesses in the company’s systems.

An aspect of audit risk for the audit of any trading company is the risk that the auditor may be given, as accurate, company records which do not accurately record the transactions entered into by the company for the period being audited. Unauthorised trading by a company’s trader which binds the company is a paradigm example of circumstances where this may occur. Where a trader knowingly enters into a trade which binds his employer which he knows he was not authorised to do (the position is different where he expects it to be ratified later), he will wish to conceal the trade by falsifying the company’s records.

Mr Swinson contended, and Mr Mah, Mr Spence and Dr Fitzgerald (D&T’s expert on banking management) (Footnote: ³⁹⁷) accepted, that, where a company has authorised its employee to trade on a market in the company’s name, there will always be an inherent risk that the auditor will receive inaccurate trading records as a result of unauthorised trading. A futures broker is an obvious example of such a company. Thus at paragraph 6.66 of his report Mr Spence listed, as one of the three principal audit risks affecting a futures broker, “operational risk, including the risk of unrecorded or incorrectly recorded transactions, either caused by concealment of unauthorised trading or accounting breakdown.” He agreed in the course of cross-examination, as did Mr Mah, that unauthorised trading was an inherent risk for a futures broker and would usually involve concealment.

On the other hand, it was accepted by all the experts that BFS did not appear to pose any unusual risk of unauthorised trading over and above the basic inherent risk in any futures broker – or, as Mr Spence pointed out, any company which employs traders.

I have described above (see paragraph 193 to 199), how Leeson came to be in charge of the back office and the trading floor at BFS, and himself took an active, indeed a dominant, role as a floor trader. He was thus in a position to control both the execution of trades and how they were reported. This continued throughout the relevant period.

Mr Bax, Mr Norris, Mr Killian, Mr Swinson, Mr Spence and, to a degree, Dr Fitzgerald agreed that, for a small execution-only broker, it was common to have one person in charge of both front and back offices. There is no evidence that, when performing their audits, D&T knew BFS to be doing anything other than execution-only trading. Therefore BFS did not pursue their allegation that D&T were negligent in failing to draw attention in their audit reports to the lack of segregation.

Mr Swinson and Mr Spence agreed that lack of segregation in a broker was a control weakness which increased the inherent risk of unauthorised trading, and that the auditor had to take that increased risk into account in conducting his audit (Footnote: ³⁹⁸). However Mr Swinson agreed with Dr Fitzgerald that, for an execution-only broker, the risk posed by non-segregation was “fairly minimal” or “very modest” (Footnote: ³⁹⁹). Mr Swinson accepted that D&T might reasonably have perceived the audit risk as low, even taking into account the lack of segregation of duties (Footnote: ⁴⁰⁰). Mr Spence in cross-examination put it slightly differently, that the BFS audit was a normal risk for a futures broker, with mitigating factors (Footnote: ⁴⁰¹).

In principle, it seems to me that lack of segregation allows a trader, if fraudulent, to misreport his trades. The possibility of misreporting means that checks by the customer or by risk control for unauthorised trades might be ineffective. And by, for example, reporting trades he has not done, a trader could obtain margin for unauthorised trades. Therefore lack of segregation must increase the risk of unauthorised trading. However I have to take account of BFS’ own expert’s view that the increase in risk was “fairly minimal”.

D&T, and Mr Spence on their behalf, relied on a number of factors which mitigated the risks of unauthorised trading posed by the nature of BFS’ business, as increased (to the extent it was) by the lack of segregation. As Mr Spence put it in his report: “the inherent risks associated with a futures trader could reasonably be seen by the auditors to be considerably reduced in the case of BFS. (Footnote: ⁴⁰²)” These factors were that: Leeson merely reviewed the work of the settlements staff, rather than doing it himself; Mr Jones actively supervised those staff; BFS had no funds of its own but relied on funds provided by its customers, who would check the calls it made on them; BFS was an execution-only broker; and BFS had only four customers, three of whom were related companies. I shall deal with these in turn.

BFS’ front- and back-room staff formed separate teams, so below Leeson’s level trading and settlement functions were segregated. A number of witnesses, including Mr Norris, expressed their surprise that the BFS back-room staff obeyed Mr Leeson’s instructions without question over two and a half years. D&T reasonably contended that they could not have anticipated that. The presence and role of such staff at BFS meant that Leeson’s unsegregated role would have appeared less of a risk than it would have done if he had been performing both sides of BFS’ activities on his own.

On the other hand, all the settlements staff employed by BFS were extremely young and had no previous experience in the securities or futures industry. Miss Hassan, Miss Kader and Miss Sng gave evidence before me that they just did what Leeson told them to do, he was their boss and they assumed what he told them to do must be legitimate and normal in the industry. Mr Hawes gave evidence that, when in October 1994 he met Miss Hassan, who had been employed at BFS since July 1992, she seemed to have no knowledge of anything beyond the mechanical settlements tasks assigned to her. The staff’s lack of experience limited even their apparent value as a mitigating factor.

BFS did not allege that D&T were negligent in believing Mr Bax and Mr Jones to be performing their proper role as active committed managers. However, insofar as D&T rely on their supervision of Mr Leeson as a mitigating factor, it is appropriate to examine the evidence relating to that supervision.

Mr Jones’ evidence to me was clear. He told me that “no one, for a minute, thought I was involved on the settlements side of BFS.” He made no enquiries as to the operation of BFS’ control systems and had little idea what the settlements staff did. He said that he was not involved with BFS at all during 1992, until the end of the year accounts, and had less involvement in 1993. He acted as, at most, half an executive Finance Director. Once he began playing the other half of the role in February 1995, it took him less than a day to find things in BFS’ finances with which he was unhappy.

This accorded with the evidence of Miss Hassan, Miss Kader and Miss Sng, all of whom told me that they had no contact at all with Mr Jones.

Mr Jones’ evidence also accorded with that of Mr Hawes. Mr Hawes visited BFS for one day in February 1994. The most Mr Hawes could say in Mr Jones’ favour was that Mr Jones gave the impression of having a general overview of BFS’ activities and having a hand in controlling BFS. However on his return Mr Hawes asked Mr Norris whether he was happy for Leeson to be effectively running BFS alone (though Mr Norris did not recall that). Mr Hawes agreed that he knew that Mr Jones was not responsible for settlements or trading. He said that Mr Jones had no knowledge of the detail of BFS’ affairs, save in the area of cash management, and it was a waste of time to ask him about them. At least later in 1994, Mr Hawes knew that Mr Jones was not fulfilling the role of a finance director for BFS either. Mr Hawes passed these impressions on to James Baker in mid-1994 in advance of the internal audit. Mr Baker’s note of 22 June recorded that:

“Nick Leeson has too dominant a role looking after both trading (agency and proprietary) and settlements aspects of the business; there is no deputy to challenge him… TH believes that SJ basically leaves NL to his own devices. While he has no evidence to suggest that NL has indeed abused his position, the potential for his doing so needs examining.” (Footnote: ⁴⁰³)

Mr Hawes told the Singapore inspectors that his comments were based on his one-day trip in February 1994, and a few subsequent conversations on the phone with Leeson.

Likewise James Baker testified that, when he arrived at BFS in July 1994:

“It was quickly apparent from my work that Leeson had an all-embracing role within BFS which put him in charge of the trading of the company, for both house and clients, and in charge of the settlements department… Leeson had sole responsibility for the supervision of BFS’ back office team, he had cheque-signing and journal-passing powers, he was reviewing and signing off SIMEX deposit, variation margin and collateral reconciliations and he was reviewing and signing off bank reconciliations. It was also clear that Simon Jones had little, if any, role in the running of the settlements department of BFS. The indications that Leeson was in charge with little supervision by Jones, which I had received during my audit planning… were confirmed. I also found that, although Rachel Yong, the Financial Manager for BSS, prepared the accounts for BFS, she had no proper involvement in the accounts for BFS, but simply took whatever trading balances Leeson gave and produced BFS’ accounts from the figures given to her by Leeson. (Footnote: ⁴⁰⁴)”

Mr Jones’ lack of involvement in BFS’ affairs struck Mr Hawes in the course of a day’s visit, and struck James Baker “quickly”. It would be surprising if the perception of D&T’s audit team, who spent so much longer at BFS’ offices than Mr Hawes, was that Mr Jones was actively involved in supervising Leeson’s activities. Miss Koo spent 100 hours, and the other staff 33 hours, on the 1992 audit. Miss Koo spent 190 hours, and the other staff 79.5 hours, on the 1993 audit (Footnote: ⁴⁰⁵). Those figures ignore the time the more senior D&T staff spent on the BSS audit, when they would also have had contact with Mr Jones.

Mr Mah’s evidence was somewhat contradictory. In chief, he said that in discussions with Mr Jones he received the impression that Mr Jones was actively involved in BFS’ affairs. In cross-examination, he agreed that D&T did not understand Mr Jones to exercise day-to-day supervision over either trading or back office operations, but merely a finance and accounting function. In re-examination he said that Mr Jones did have day-to-day involvement in BFS’ operations, and that he had no reason to believe that Mr Jones was not supervising Leeson (Footnote: ⁴⁰⁶).

The problem here is that Mr Mah spent 4 hours on the 1992 BFS audit and 8 hours on that in 1993, no doubt mostly in his own office. His evidence as to D&T’s perception of the level of supervision of Leeson is of little value. Miss Koo, who spent far longer in BFS’ office and would have been better placed to give evidence on the point, was not called as a witness.

It seems to me likely that Mr Mah’s evidence in cross-examination was correct: D&T were aware in both 1992 and 1993 that Leeson was effectively unsupervised on a day-to-day basis by anyone in Singapore in his running of BFS’ trading and settlements. They may have believed Mr Jones to have performed at least some financial control function, although James Baker’s evidence suggests otherwise.

Accordingly I do not attach much weight to D&T’s perception that Mr Jones was supervising the back office as a factor which they were entitled to treat as reducing the audit risk. However they may have been entitled to take some comfort from Mr Jones’ role on the finance side.

D&T contended that they perceived BFS to have very limited funds of its own, but to be reliant on margin received from its customers, who would reconcile margin calls to the trades conducted. Therefore it had very little, if any, scope for unauthorised trading since it would not be able to fund it.

This was certainly a perception shared by BSL management, and indeed by Mr Bax. However they had considerably less knowledge of the workings of the SIMEX market than did Mr Mah.

As explained at paragraphs 93 to 106 and 318 above, BFS was required by SIMEX to obtain significantly more margin from its customers than it was required to deposit with SIMEX. Furthermore, while BFS was only obliged to call for further margin if a customer’s equity balance fell below its maintenance margin, BFS in fact required its customers to keep their margin with it topped up to the initial level. Therefore BFS always held, for futures trades, about 25% more margin than it had to deposit with SIMEX. As of 30 September 1992, the excess margin was US$16.7m and as of 31 December 1993 it was US$18m (Footnote: ⁴⁰⁸). We now know that during 1992 and 1993 Leeson financed his unauthorised trading largely from these excess customer margins, though by the end of 1993 he was also resorting to the use of BFS’ overdraft with Citibank (see paragraphs 317 to 324 above).

Mr Mah’s evidence was that D&T were aware that BFS would hold these excess margins. D&T’s workpapers confirm this, and also note that BFS intended to require customers to maintain margin at the initial level (Footnote: ⁴⁰⁹). Mr Mah’s only answer, when it was put to him that these were funds available for Leeson to use for unauthorised trading, was that they were customer funds which the customer could demand back at any time. That of course hardly answered the point.

Against this, Mr Gaisman for D&T pointed out to Mr Swinson in cross-examination that neither he in his first report nor Messrs Bax, Jones, Norris or Hawes, nor the Singapore inspectors, nor Jonathan Parker J in re Barings plc (No. 5), appreciated that BFS had access to the excess margins. Mr Gaisman put it to Mr Swinson, and Mr Swinson accepted, that in that light his criticism of D&T for not appreciating the point “does seem rather strong”. (Footnote: ⁴¹⁰) I can understand Mr Swinson’s embarrassment at having come to the point late, but neither he nor any of the others listed had the experience of SIMEX which Mr Mah had. Mr Spence agreed that D&T could not just assume that BFS had no funds of its own, but needed to evaluate the evidence (Footnote: ⁴¹¹). D&T knew that BFS did have access to this source of funds and cannot now rely on the fact that others, with less relevant knowledge than they, did not, as a factor reducing the audit risk.

A second source of funds for Leeson’s trading was BFS’ overdraft, described at paragraph 321 above. Leeson was not using this at the time of the 1992 audit. In 1993, his use of it was not apparent at the planning stage, and therefore not relevant to D&T’s initial assessment of audit risk, but was apparent by the time D&T received the trial balance sheet in January 1994. Mr Mah pointed out, correctly, that this was intended to be a “daylight” facility only, not to be drawn on overnight. But it was obvious from the 1993 balance sheet that the overdraft was being used on an overnight basis. It was shown as a current liability, to the extent of S$13.75m. Mr Spence agreed that it was obviously being used overnight, for funding trading, and that that undermined any assumption made by D&T that BFS had no resources of its own with which to trade.

Furthermore, BFS’ bank records, had D&T examined them, would have shown that BFS’ bank account had been overdrawn for most of November and December 1993. Therefore it was being used as a continuing source of funding, and certainly not as a daylight-only facility.

D&T contended that they were entitled to expect BSL to monitor its excess funds in BFS’ hands. But D&T had no evidence that BSL either did this or understood that there should be excess funds.

It might more strongly be argued that Mr Jones, as Finance Director, should have been monitoring the funds held by BFS and its overdraft, and ensuring that excess margins were on deposit at Citibank, not at SIMEX. Undoubtedly Mr Jones should have been doing that, and it was a natural assumption that he would do so. But it was only an assumption, and not tested. An effective finance function would also not negate the risks posed by Leeson running both trading and settlements. It might have prevented a trader using the excess margins to trade with, but it would not necessarily prevent him finding funds by other means. For example, he might report fictitious trades to clients and call margin in respect of them to finance his own trading; or build up extensive positions using the profits from temporarily profitable trading, which could expose the company to large losses if the market turned against him.

Accordingly BFS’ apparent lack of funds to use in unauthorised trading was of rather less weight as a mitigating factor than D&T contended.

Mr Swinson, Dr Fitzgerald and Mr Spence all agreed that, other things being equal, the risk of unauthorised trading was less for an execution-only broker than for a discretionary broker. This was because an execution-only broker had (i) limited resources and (ii) a customer watching over his shoulder.

I accept that BFS being an execution-only broker reduced the risk of unauthorised trading. However reason (i) is weakened by BFS’ access to excess customer margins, considered above. Reason (ii) assumes both that a customer is more efficient than internal risk control in monitoring limits and that the broker reports trades accurately – an assumption which is weakened where there is no segregation of duties.

It is difficult to see why this should significantly reduce the risk of unauthorised trading. Mr Spence in cross-examination said only that it helped somewhat, but did not radically reduce the risk.

In 1992 and 1993, there was an inherent risk that an employee of BFS might have engaged in unauthorised trading. But that risk was no higher than was present in any futures broker.

Mr Leeson’s position in charge of both front and back offices was usual for small brokers in Singapore and was agreed by Mr Swinson to increase the risk of unauthorised trading to only a minimal extent. Once the decision had been taken to carry out a substantive audit, the only impact of Leeson’s position was to require the auditors to be cautious about accepting representations from Leeson’s staff as audit evidence. However in a substantive audit they were anyway obliged to obtain external confirmation for the figures in the financial statements.

The mitigating factors relied on by Mr Mah were of less force than D&T contended. But my scepticism as to some of them does not mean that the audit was in any sense high risk. Mr Swinson agreed that D&T were justified in assessing the audit as low risk and Mr Spence said it was normal risk, in both cases with no specific risk of unauthorised trading. In the light of that evidence I cannot agree with BFS’ submission that D&T should have perceived there to be a substantial risk of unauthorised trading and its concealment.

Whether because of the simplicity of BFS’ balance sheet and its limited number of customers (D&T’s reason (Footnote: ⁴¹²)) or because SAG 7 required it where there was limited segregation of duties (BFS’ reason), the parties agreed that D&T was correct to carry out a substantive audit of BFS. This required D&T to obtain reasonable assurance that the figures in the financial statements were correct and complete, otherwise than from the operation of the company’s systems and controls.

The planning procedure in D&T’s audit manual required D&T, at the planning stage, to assess the audit risks to which BFS was subject and whether it was subject to any specific identified risks which required “particular attention in developing [their] audit plan” (Footnote: ⁴¹³). If D&T decided to rely on the company’s controls, they would test the controls and perform basic substantive tests. If performing a wholly substantive audit, then for each specific risk identified, D&T were obliged to perform “focused substantive tests” to address that risk. To address potential errors for which they had not identified specific risks, D&T were obliged to perform an intermediate level of substantive tests.

This approach was not criticised by BFS. Nor could Mr Swinson identify any specific risks which D&T should have identified at the planning stage. He hesitated as to how Leeson’s unsegregated role should have been categorised under the D&T approach, but did not suggest that either unauthorised trading or particular methods of concealing it should have been identified as a specific risk. Mr Spence agreed that unauthorised trading and its concealment were not, on the information available to D&T, specific risks which D&T should have identified and devised special procedures to detect.

Thus, when Mr Mah gave evidence in chief that:

“we had no reason to identify the possibility of BFS entering unrecorded [which he explained as including unauthorised] transactions as a risk, and did not do so. Accordingly there was no reason for us to devise procedures to test for such transactions” (Footnote: ⁴¹⁴),

he was referring to the identification of specific risks in the planning process required by D&T’s audit manual. When his comment is interpreted in that sense, Mr Swinson’s evidence was to similar effect.

Mr Swinson’s evidence was that D&T had to take the lack of segregation of duties into account in assessing the weight to be given to particular evidence, in particular explanations from Leeson. However he agreed with Mr Spence that in all other respects Leeson’s position was adequately dealt with by the decision to conduct a substantive audit. That position did not require the substantive audit procedures to be more rigorous, or the audit evidence stronger, than usual (Footnote: ⁴¹⁵). As I have noted, Mr Swinson agreed that BFS reasonably appeared to be a low risk audit, even after taking into account both the inherent risk of unauthorised trading and Leeson’s role.

BFS contended that nevertheless there was a risk of unauthorised trading and its concealment. D&T should have guarded against that risk by devising procedures that catered for it. However I accept D&T’s submission that it is not enough to identify the existence of a risk, without paying attention to its size. Otherwise an auditor would always be liable whenever a fraud occurs. By definition there would have been a risk of that fraud occurring and the auditor would have failed to devise procedures that detected it. That cannot be correct.

It is instead necessary to decide whether, in the circumstances of which the auditor should have known, the procedures which he followed gave him reasonable assurance that the financial statements were accurate. Only if a risk of which he should have known should have indicated to him that those procedures did not give such reasonable assurance was he required to carry out other procedures to cater for that risk. Clearly the size of the risk is relevant to that assessment. This seems to me the planning process laid down in the D&T manual and in my judgement was the correct approach.

BFS made three detailed criticisms of the tests which D&T conducted:

I consider each in turn.

As I have mentioned, BFS sent out the confirmation requests, and received them. They were in the form of specially-created faxes, rather than attaching reports from CONTAC. D&T subsequently checked the figures in which they were interested against CONTAC, but not others (such as most of the open positions listed in the faxes).

There was a rather arid debate between the experts as to whether Statement of Auditing Practice 1 (“SAP 1”), which governed the verification of debtor balances, applied to BFS’ balances with its customers and, if so, whether D&T were in breach of it. It does not matter greatly, since D&T were not in clear breach of SAP 1 even if it applied, but in my view SAP 1 did not apply, as the customers were creditors.

The point of substance is whether D&T should have retained control of the process of obtaining confirmations. This would have prevented Mr Leeson manufacturing a forged confirmation of the 88888 account in 1992. Mr Swinson’s evidence was that it was acceptable for BFS to send out the requests on faxes as it did, but that D&T should have had the returns directed to them. Mr Spence said that in an intra-group context it was usual to rely on inter-company correspondence, so it was acceptable for the returns to go to BFS. Mr Mason, BFS’ expert on Singapore practice, supported him. In the light of that evidence, I do not find D&T to have been at fault in allowing BFS to control the confirmation process.

As is clear from SAG 4, quoted at paragraph 496 above, an auditor is concerned to obtain reasonable assurance of the completeness of the information in the financial statements, particularly where he is conducting a substantive audit. Mr Swinson gave evidence that, when auditing a futures broker, this required an auditor to obtain reasonable assurance that the brokers’ records agreed both with the exchange and with customers (to the extent that the broker was not trading on its own account). This was true for open positions as well as balances. Mr Spence agreed with the requirement to obtain reasonable assurance of the completeness of the records, and I think with its application to open positions, although he disagreed as to what test should be done to achieve that assurance.

In any event, Mr Swinson’s basic proposition must be right, both as a matter of common sense and in the light of SAGs 4 and 7 above. A broker trades as a principal on SIMEX. If it has entered into trades on SIMEX which a customer refuses to acknowledge, the broker is liable for those trades. Conversely, if its customer has instructed it to effect certain trades but it has not done so, it is liable to the customer for breach of contract. An auditor has to obtain reasonable assurance that liabilities of this kind which would affect the financial statements as reflecting a true and fair view of the audit client’s affairs have been disclosed to him.

An aspect of this duty on an auditor is a duty under SAG 12 (quoted at paragraph 501 above) in an appropriate case to plan his audit so as to obtain reasonable assurance that no unauthorised trading has taken place, and to follow audit procedures that will give him that assurance.

Mr Swinson said that a test for completeness of the records of positions might be done in different ways for different company audits. But in BFS’ case, D&T had available an independent source of evidence in SIMEX. Mr Swinson suggested that D&T should have compared the total positions shown on the SIMEX CMP schedule with the total positions shown on the customer confirmations. This would have ensured that all positions entered into by BFS were authorised by customers. Given BFS’ small number of customers, such a test would have been quick and reasonably cheap to complete.

Mr Mah disagreed. He explained that D&T’s audit approach was to verify the balances owed to customers and owed to or by SIMEX. He accepted that an auditor needed to test for completeness of the data produced by the accounting system and reflected in the financial statements. He accepted that there was nothing in the audit guidelines to the effect that, in conducting his tests, an auditor need only look at balances and may ignore positions. But he contended that, in the case of BFS, substantive verification of the entire balance sheet was possible, and that provided D&T with reasonable assurance that the financial statements were complete.

This was because any transaction had an effect on the financial balances. I have explained at paragraphs 113 to 121 above how CONTAC dealt with futures and options trades. It will be evident from that that any transaction entered into on behalf of a customer had an immediate effect on the equity balance for that customer. Futures transactions gave rise to an immediate margin requirement which, when paid, increased the equity balance and to a continuing requirement for settlement variation which increased or reduced the equity balance. Option trades gave rise to an immediate realised profit or loss which affected the equity balance as the premium was debited (on a purchase) or credited (on a sale). On a sale, the seller had to put up margin which, when paid, increased the equity balance. Therefore, Mr Mah said, verifying the equity balance on an account verified whether or not there were open positions on that account: if the balance was nil, D&T could feel satisfied that there would be no open positions on it.

Mr Spence agreed with this logic. His evidence was that D&T complied with their duty, to obtain reasonable assurance that the financial records were complete and that no unauthorised trading had taken place, by doing a fully substantive audit which covered all balances. The customer confirmation for each account with a positive or negative financial balance established that all positions on that account were authorised. The fact that the confirmations added up to the exact total of amounts owed to customers shown in CONTAC meant that it was reasonable for D&T to believe that there were no other active accounts than those which had been confirmed. Mr Spence’s opinion was that D&T were not obliged to do more unless they were specifically put on enquiry as to the risk of fraud and thus of the manipulation of balances (Footnote: ⁴¹⁶).

BFS disagreed with this. First, it pointed out that, on a sale of an option, only the premium was reflected in the equity balance immediately. Margin was not reflected until paid, which might be late or never. However the important point is that any transaction had an effect on the equity balance on the trading day, whether or not it also had a further effect on subsequent days.

Secondly, I heard much debate as to the chances of an account balance being nil despite having open positions on it, other than through deliberate manipulation. Mr Swinson’s expert report originally took the point that, when one looked at the trading general ledger produced by CONTAC for 31 December 1993, there were a large number of BFS accounts which had nil balances but clearly had open positions, since they showed a figure for settlement variation. Under cross-examination this proved to be a misapprehension. In fact all the accounts to which Mr Swinson referred were inactive accounts with no open positions. The only account which had a nil balance but open positions was the 88888 account.

Mr Swinson’s misunderstanding of the description “variation margin” on the trading general ledger was perhaps understandable. But his contention that a test on balances alone was an inadequate test for open positions was clearly weakened by his acceptance that there were not, after all, other accounts with open positions which the balances test was unable to detect.

Despite this, Mr Swinson adhered to his view that, even in the absence of fraud, there was a realistic possibility that an account which had open positions might have a nil balance by chance. He said that nil was a number like any other, and the chances of a balance being nil by chance were no different to those of it being any other number. Mr Swinson’s position on this seemed to me unrealistic. His own logic indicated that he was wrong. The chances of a particular active account ending up at the accounting reference date with any particular balance, including a nil balance, are infinitesimal without deliberate manipulation.

Thirdly, BFS contended that D&T’s approach was inadequate because it was liable to be defeated by fraudulent manipulation of balances. Mr Swinson stated that auditors should be alive to the risk of accounts being manipulated to achieve a desired effect, whatever the type of company. Therefore BFS contended that D&T should have devised a test, such as Mr Swinson’s, which would not have been defeated in that way. I deal with this below.

It was put to Mr Swinson in cross-examination that D&T’s audit of balances included agreement of the figure for settlement variation between BFS’ records and those of SIMEX. That was the total of all margin, settlement variation on futures positions, option premiums and realised profit and loss. Mr Swinson accepted that the fact that BFS’ records agreed with SIMEX rendered it likely that BFS’ records reflected all transactions entered into by BFS on SIMEX (as of course they did). The converse would only be true in the event of deliberate fraud (Footnote: ⁴¹⁷).

Mr Swinson further agreed that the agreement of the balances in CONTAC to customer confirmations would (if the confirmations were prepared properly, which, in relation to balances, they were) detect both inadvertent and fraudulent unauthorised trades. The only exception would be if balances had been fraudulently manipulated to zero (Footnote: ⁴¹⁸).

Accordingly it seems to me that Mr Mah was right that, subject to one proviso: (i) D&T in planning their audit were entitled to do so on the basis that any transaction would have an effect on the equity balance of the account to which it was booked; (ii) therefore the chances of an account having a nil balance but open positions were extremely low; (iii) therefore D&T were entitled to assume that, when they had obtained customer confirmations for account balances which added up exactly to the total in BFS’ books for sums due to customers, there were no other accounts which had open positions; (iv) therefore they had reasonable assurance that the financial statements they were auditing were complete, in the sense that they reflected all transactions entered into by BFS.

The one proviso was the possibility that, not only had there been unauthorised trading, but it had been concealed by fraudulent manipulation of the financial balances. This issue therefore comes down to whether D&T were negligent in not carrying out a test directed specifically to the risk of such manipulation.

D&T put forward a number of reasons indicating that the test devised by Mr Swinson (that is, agreeing open positions between SIMEX records and customer confirmations) was not a test which any reasonable auditor would have carried out in the circumstances of the BFS audit:

With less force, D&T pointed out that Mr Swinson’s test would not have been practicable in auditing a bigger broker with more customers, not all of whom would return confirmations. However in that case the auditor would carry out a controls-based, rather than a fully substantive, audit. Presumably he would devise tests that would check that those controls would produce complete records and detect unauthorised trades: that was not a question explored in evidence before me. But in this case D&T were carrying out a substantive audit precisely because BFS was a small operation, with few customers, and that was why it was verifying the balance sheet by looking for external evidence.

However I think the decisive point is Mr Spence’s evidence that D&T were not required to perform a test which was designed only to detect this particular form of concealment of fraud by unauthorised trading. I have, at paragraph 543 above, reached the conclusion that unauthorised trading, and deliberate concealment of that unauthorised trading, were inherent risks in any audit of a futures broker. However in the case of BFS, all the experts agreed that there did not appear to be any specific risk of unauthorised trading. Mr Swinson accepted that the risk appeared low. I have further concluded, at paragraphs 572 and 573 above, that the procedure followed by D&T would have detected any form of unauthorised trading save where it had been concealed by manipulation of the balances.

Given those conclusions, Mr Spence must be correct that the audit procedures which D&T followed gave them reasonable assurance that there had been no misstatement due to unrecorded or unauthorised trading, and D&T were not obliged to devise further procedures specifically directed to that risk. Still less were D&T obliged to devise procedures which had no other purpose than to detect one particular means by which fraudulent trading might be concealed. Mr Swinson’s test would have had no other purpose than that: he accepted that the only misstatement which his test would catch and D&T’s approach would not was one where the balances had been fraudulently manipulated to zero.

Accordingly I find that D&T were not negligent in failing to carry out Mr Swinson’s test for open positions in either 1992 or 1993. Given the considerations I have listed above, it is impossible to say that they fell below the standard described by Lord Bingham in Eckersley v Binnie, 18 ConLR 1, at p.80:

“[A professional man] must bring to any professional task he undertakes no less expertise, skill and care than any other ordinarily competent members of his profession would bring, but need bring no more. The standard is that of the reasonable average. The law does not require of a professional man that he be a paragon, combining the qualities of polymath and prophet.”

BFS’ contention on this issue would require Mr Mah to have been a prophet, rather than ordinarily competent.

There is a separate issue as to whether D&T should have insisted on a confirmation of the balance on the 88888 account in 1993, despite it having a nil balance. BFS’ argument here depended upon D&T appreciating that the 88888 account was a principal account, rather than one of the many sub-accounts of the omnibus accounts (for example the BSL segregated account). Otherwise there was no more reason for them to obtain a confirmation for it than for any of those sub-accounts. Mr Swinson accepted that, in the 1994 audit, it was understandable for C&LS not to confirm nil balances, where they believed the balances in question to be included in the omnibus accounts (Footnote: ⁴²⁰).

BFS relied on D&T having seen a confirmation for the 88888 account in 1992 (that forged by Leeson) and noted that in their 1993 working paper dealing with amounts due to related companies. Therefore, Mr Brindle argued, D&T knew it was a principal account. As such it should have been confirmed. Even though the balance on it was zero, that might be the result of fraudulent manipulation. It was also possible that the customer believed the balance to be something other than zero, but would not have the chance to correct BFS’ misapprehension if it was not asked.

I take BFS’ point that, logically, an account which was confirmed separately in 1992 cannot have been a sub-account of one of the other BSL accounts which was also confirmed. Otherwise its balance would have been counted twice. Therefore it must have been a separate, principal account.

However it follows from my conclusion on BFS’ wider case on testing for open positions that that does not take BFS far enough. I have concluded already that, in the absence of any specific identified risk of fraud involving manipulation of balances, D&T were entitled to assume that a nil balance would not have open positions. They were presented with confirmations of balances from customers which added up in total to the exact figure shown in CONTAC for the balance owed to customers. That gave them reasonable assurance that all other accounts must be nil and therefore had no open positions. In the absence of any warning bells, there can be no reason to require them to obtain a confirmation of any such accounts.

That accords with the evidence of Mr Mah and Mr Spence. It also accords with the evidence of Mr Mason, save that Mr Mason suggested that he would make an exception for accounts which appeared on a “lead schedule” (Footnote: ⁴²¹). I do not see any reason why either that factor (which bore strong signs of being manufactured to fit the present facts) or the related factor that a particular account had a balance the previous year and therefore ought to be the subject of a confirmation, should affect my conclusion.

As for BFS’ point that the customer might believe the balance to be something other than zero, I take Mr Spence’s point on that. He asked whether BFS suggested that D&T should have sought confirmations from third parties who were not shown as customers on CONTAC but who might have proved to be such and to have balances. I agree that BFS’ argument strays well beyond the limits of the speculation in which an auditor can be reasonably required to indulge when assessing whether he has reasonable assurance of the accuracy of a balance in the accounts.

I therefore conclude that D&T reasonably assessed the audit risk as normal for a futures broker, with no specific risk of unauthorised trading over and above that applicable to any futures broker. They checked the balances shown on CONTAC against both SIMEX’s records and customer confirmations. Given the level of risk which they had reasonably assessed to exist, that gave them reasonable assurance that (i) all transactions into which BFS had entered on SIMEX were reflected on CONTAC, (ii) BFS’ customers had authorised all transactions entered into on the accounts which were the subject of customer confirmations, (iii) those were the only active accounts and therefore (iv) the financial statements were complete.

D&T’s approach was liable to be defeated only by fraudulent manipulation of the balances on CONTAC. It was not alleged that D&T should have identified that as a specific risk. Therefore they were not required to devise and perform a procedure which had as its sole object the detection of such a fraud. Therefore D&T were not negligent in either the 1992 or the 1993 audit in failing to perform the test on open positions devised by Mr Swinson.

Nor were D&T negligent in following the procedure they did for obtaining customer confirmations nor in failing to obtain a customer confirmation for the 88888 account in 1993.

Even if D&T had carried out Mr Swinson’s test in 1992, it would not by itself have detected Leeson because he had closed out all open positions on the 88888 account as at the 31 September 1992. However that created a ¥670m “hole” in the accounts which it became necessary for Leeson to conceal from the auditors. The method that he used to do so was one which D&T’s audit approach, of confirming and reconciling balances only, was capable of uncovering.

As I have mentioned at paragraph 305 below, at the time of the September 1992 audit the 88888 account contained losses of ¥664 million (c.S$9.2 million). Leeson concealed them by the following steps (Footnote: ⁴²²):

The field work for the 1992 audit was conducted by Miss Koo. It appears that she investigated the ¥670m receipt. She did not give evidence either to this court or to any of the various bodies which have investigated the collapse of the Barings group. Therefore we can discern what Miss Koo did and saw only from the documents in the audit file, in particular from her markings on the documents in that file.

From the audit file, it appears that Miss Koo reviewed the bank reconciliation for 30 September (“the Bank Reconciliation”) on 7 October 1992, when auditing the figure in the trial balance for “Other deposits” (Footnote: ⁴³³). This figure included amounts on deposit with Citibank and at SIMEX, in both cases in BFS’ house and customer accounts. In the final balance sheet those amounts were shown as “Cash”.

From her manuscript notes on the relevant working papers, it appears that Miss Koo checked the cashbook figure for the Citibank customer account shown on the Bank Reconciliation against the Trading General Ledger Account balance summary in CONTAC. She verified the smaller reconciling item, which was settlement variation (which CONTAC took into account the day before SIMEX withdrew it from BFS’ bank using its direct debit mandate). Likewise she checked the figure for the bank balance against the Citibank statement. Miss Hassan had described the other reconciling item, of ¥670m, as “Funds due from BSL – Seg A/C”. Miss Koo clearly received an explanation for this item from BFS staff, because she annotated the Bank Reconciliation as follows:

“Baring Futures took into their books the receipt of funds based on customer advice that they will transfer funds value 30/9/92. However cash was not received as at 30/9/92.

“Sighted Citibank fax advising the receipt of $670m [sic] value 1/10/92”

It is not known which of BFS’ staff provided this explanation to Miss Koo.

I have described above the Citibank fax referred to in this note, which did indeed describe the receipt as (allowing for the notation) “value 1/10/92”.

Miss Koo saw a further relevant document, probably also on 7 October although she has not dated it. This was the daily activity statement for the 88888 account dated 30 September, showing the ¥670m as cash received, which I have mentioned above. Miss Koo wrote, against the ¥670m credit, “Cash received”. The statement shows the client as Baring Futures (Singapore) Ltd, with BFS’ address and “Attention: Nick Leeson”. The account number is shown as 88888.

Also on 7 October 1992 Miss Koo started her work on the item in the trial balance representing liabilities due to related companies. The evidence of Mr Spence was that Miss Koo would have started with the six confirmations which BFS had sent to, and received back from, BSL and BSJ (or apparently had done so in the case of the forged confirmation of the 88888 account). These were for four BSL accounts (the segregated, non-segregated, house and 88888 accounts), and the two accounts for BSJ Tokyo and BSJ Osaka. The totals of the figures for “account balance” and “variation margin” on each account confirmation, all added together, equalled exactly the total figure for “client payable” in BFS’ trial balance. That was itself checked against the client general ledger in CONTAC.

Accordingly BFS’ records showed that BFS owed related companies S$80.212m; and Miss Koo had confirmation that the related companies agreed that BFS owed them that amount. She also knew that, for example, BFS and BSL agreed that, of that S$80.212m, S$61.6m was in respect of the BSL segregated account (though Miss Koo did not check this figure against CONTAC). This meant that, if Miss Koo believed that BFS had credited the ¥670m to BSL’s segregated account on 30 September (and thus included the ¥670m in the S$61.6m), she might infer that BSL had also treated the payment as made on 30 September (and so done the same).

One of the consequences of the complete absence of evidence from Miss Koo is that we do not know in which order she dealt with the two topics of Cash at the bank and Liabilities to related companies. The work sheets for both are dated 7 October. However the fax from BSL providing the confirmation for the segregated account was sent in two versions, at 13.46 and 15.10 respectively London time on 7 October. Singapore was eight hours ahead of London. Therefore it seems unlikely that Miss Koo saw the BSL confirmation for the segregated account on 7 October. So she would have been unable to conclude her work on the issue of liabilities to related companies until the following day.

We do not know for certain when Miss Koo concluded her work on Cash. But she had all the documents necessary for that on 7 October. Therefore it must be likely that she concluded her work on the cash issue before she received the BSL confirmation for the segregated account and concluded her work on Liabilities to related companies.

BFS: BFS’ case was that conventions, described by Mr Swinson, governing the accounting treatment of receipts are clear. A mere promise to pay on a particular date does not justify booking the payment on that date unless it is actually received on that date. To do otherwise is to record an untruth. BFS however accepted an exception to this general rule in the case of payments between companies in the same group, in order to maintain consistency between the books of the paying and recipient companies. Thus BFS accepted that Miss Koo would have been justified in allowing BFS to treat the payment as having been received on 30 September 1992, and so in the period of the 1992 accounts, if she was satisfied that BSL was treating it as having been paid on that date.

BFS contended that the explanation given to Miss Koo was not to that effect, but was of a promise which had not been fulfilled. Timing differences were not an explanation: the Citibank fax made it clear that BSL intended the payment to be made on 1 October. Therefore it was clearly wrong to treat the payment as received on 30 September. It was of prime importance that payments were recorded in the correct financial year. In particular, auditors should be alert to the risk of window-dressing as a result of payments or receipts taking place close to accounting reference dates, and this payment was just the sort of transaction which should have been investigated with that in mind.

BFS contended that there was no evidence that D&T paid any attention to the BSL confirmation in resolving the anomalous treatment of the ¥670m payment. They could not have done so on 7 October, as the confirmation was not available until the next day. In any event, D&T should not have treated the confirmation as trumping the Bank Reconciliation or the evidence of receipt on 1 October. Instead they should have contacted BSL or C&LL to establish how the ¥670m had been treated in BSL’s books.

BFS suggested that Miss Koo sought to check the double-entry to the customer’s account and saw the daily activity statement for the 88888 account. She should then have investigated the obvious discrepancy between the payment being credited to that account and the explanation that the payment was to BSL’s segregated account. She should also have investigated the purpose of the 88888 account.

D&T: D&T contended that they had no ground when auditing the payment for suspecting fraud. Nor could they suspect that BSL would make such a payment on Leeson’s simple request, thus enabling him to practise such window-dressing. Mr Swinson’s conventions as to the recording of receipts were not firm, or even written, rules, but were just means to the end of ensuring the accounts gave a true and fair view. The key issue was to ensure that BFS’ treatment of the ¥670m payment was consistent with BSL’s.

The explanation apparently given to Miss Koo was consistent with BSL intending the payment to be made, and therefore booked, on 30 September. Problems caused by timing differences were commonplace, as Singapore was a working day ahead of London, so the explanation provided for the delayed receipt was plausible. The money was received in accordance with the explanation, from BFS’ parent and in cash. Given that BSL confirmed the inter-company balance, the only plausible explanation was that BSL had also booked the payment on 30 September.

Miss Koo looked at the daily activity statement in order to ensure that the payment was recorded in CONTAC for 30 September. Mr Swinson accepted that she could not be expected to remember that 88888 was not the number of the BSL segregated account. It was far too narrow a ground to hang a finding of negligence on, that Miss Koo should have noticed that the account holder was shown as BFS, given that in BFS’ books there was by that time another (error) account in BFS’ name.

The only factual witness evidence before me on this issue was that of Mr Mah, even though all the field work on both the 1992 and 1993 audits was done by Miss Koo. D&T explained that they were unable to call Miss Koo to give evidence: she declined to assist on the grounds that her current employer was not prepared to allow her to spend time on the case. BFS contended that this was no reason for not obtaining Miss Koo’s evidence in some form, if only by serving a witness statement, which might have been admitted in evidence, or by letter of request. D&T accepted that they had been in touch with Miss Koo. I concur with BFS’ criticism. As the following description shows, Miss Koo was plainly a material witness.

All the expert witnesses agreed that the issue of cut-off, or ensuring that transactions are recorded in the right accounting period, is one to which auditors should be alert. This is so without regard to the possibility of fraud or window-dressing: it is fundamental to ensuring that accounts give a true and fair view of a company’s affairs that a payment or a receipt is recorded in the correct period. This is made clear in the D&T International Audit Manual:

“In identifying specific risks, we need to identify how material mis-statements might occur. There are six types of possible mis-statements…:

…Cut-off: Transactions are recorded in accounts in the wrong period.” (Footnote: ⁴³⁴)

and again:

“The incidence of misstatements is greater for transactions recorded (or improperly omitted from recording) at or near the end of an accounting period (i.e. cutoff)” (Footnote: ⁴³⁵)

Window-dressing is a particular aspect of cut-off. It is not necessarily the product of fraud such as was encountered here: the company may just be seeking to improve its current year’s results. The Coopers & Lybrand manual devotes a section to window dressing (Footnote: ⁴³⁶), which it defines as:

“the process by which transactions are recorded before the year end but then reverse or mature soon after the balance sheet date. The purpose and substance of such transactions is to alter the appearance of the balance sheet.”

Among the examples given is “including in the bank balance monies received after the balance sheet date. This has the effect of inflating balances in the cash records.”

The manual advises that “when checking intra-group transactions, especially between group companies with different accounting dates, the auditor should be alert for entries intended to support window dressing.”

Likewise SAG 12 lists examples of conditions or events which increase the risk of fraud or error. Of the four categories, one is “Unusual transactions”, examples 1 and 2 of which are:

“1.

Unusual transactions, especially near the year end, that have a significant effect on earnings.

2.

Transactions with related parties.”

Accordingly D&T should have been alert to checking that transactions were recorded in the right accounting period, and the propriety of transactions just before or just after the end of the accounting period. That was so even though they had reasonably not identified any specific risks (including that of window-dressing) at the planning stage. The D&T audit manual states:

“The fact that we have not identified a specific risk relating to an account does not mean that a misstatement in that account cannot exist. Our audit plan is designed to ensure that, if a material misstatement exists, we are reasonably likely to detect it. Therefore, in developing our audit plan, we need to ensure that none of the six types of potential errors that may relate to each significant account balance or to the financial statements as a whole [one of which is cut-off] is overlooked.” (Footnote: ⁴³⁷)

In conducting its audit of BSS in 1992, D&T conducted an express “window dressing test”, which involved checking the bank accounts for five days either side of the year end and satisfying themselves that there were no significant unusual items. Mr Mah was unable to say why no such test was done for BFS (nor was it for BSS in 1993). There was a very similar test listed in an audit programme attached to the 1992 working papers for one of the BFS bank accounts (Footnote: ⁴³⁸):

“Select material payments from bank statement and ensure that these are recorded in the correct accounting period. Basis: amounts above $5,000.”

This was marked as outstanding on the work paper. While Mr Spence treated this programme in his report as applicable to all BFS’ accounts, D&T suggested that it only applied to one account, and that was not a trading account. Mr Spence also suggested it only applied to payments, not receipts.

Whether that was so or not, Mr Mah accepted that in this audit D&T looked at cut-off as being one of the areas of risk (Footnote: ⁴³⁹). Mr Spence accepted the importance of ensuring that transactions were recorded in the right accounting period and that, when considering the ¥670m, D&T were addressing cut-off (Footnote: ⁴⁴⁰). Therefore it can be no explanation, for example for Miss Koo not looking at the account details on the daily activity statement for the 88888 account, that she was just auditing the cash figure and needed to do no more than check that the cash had been received. She needed to check also that it had been recorded in the right accounting period.

Mr Swinson formulated what he described as the accounting conventions as to the recording of receipts. He derived them from the UK Statement of Standard Accounting Practice 2, which states that “revenue and profits are not anticipated, but are recognised by inclusion in the profit and loss account only when realised in the form either of cash or of other assets the ultimate cash realisation of which can be assessed with reasonable certainty.”

According to Mr Swinson’s conventions, a recipient of a payment should account for it on the date of receipt. Money is only treated as received if it has been received in cash or in a form convertible into cash without further action by the payer (for example, a cheque). Therefore the recipient of a payment which was promised for 30 September, but not received until 1 October, should account for it on 1 October.

If a payer instructed payment to be made for value on a particular date, the payer would book the payment on that date. Mr Swinson caused a certain amount of confusion about this, in that he stated at one point that a payer generally records payments in its cashbook when it issues instructions to make the payment. But in the light of his later evidence, and that of Mr Spence, it seems that a payer would only do that if it believed that payment was to be made on the same day. Thus Mr Swinson told me that, if BSL gave instructions on 30 September for payment to be made on 1 October, BSL would record the payment in its books on 1 October (as happened). Furthermore he said that it would not be reasonable for an auditor to assume that, because BSL must have given the instruction on 30 September, it must have booked the payment on the same date (Footnote: ⁴⁴¹).

Mr Mah and Mr Spence accepted Mr Swinson’s conventions as the general rules to be applied (Footnote: ⁴⁴²). However Mr Spence said that, since the conventions were not written down, Miss Koo would not necessarily have thought in terms of them. He contended that the ¥670m came within an exception to the conventions. He identified two possible exceptions. The first was that the payment was equivalent to an unbanked cheque, a suggestion which he put in his report but did not strenuously maintain before me. The second was that D&T might have assumed that BSL had intended the payment to be made on 30 September and the late receipt was due to the time difference between London and Singapore.

I accept that Mr Swinson’s conventions are not written down in the auditing guidelines. But they amount to little more than the expression of a common sense approach to accounting for receipts which I should have thought would be basic to any auditor’s approach. I was not shown any auditing textbooks but it seems fairly obvious, even to a layman, that a recipient of a payment should usually account for it on the date on which it is received. To do otherwise is to record an untruth. Likewise a payer should account for the payment on the date on which it is paid or intended to be paid. The experts agreed that, in an intra-group context, it is necessary that payer and payee should account for a payment on the same date and, depending on the method of payment, that may require a divergence from the usual rule. But the usual rule would be the auditor’s starting point.

Given the identified need to pay attention to the issue of cut-off, Mr Swinson’s view was that Miss Koo should have looked carefully at the transaction and the explanation she was given. This was so even though Miss Koo might not at that stage necessarily have considered window dressing as a possible explanation of the ¥670m item on the Bank Reconciliation. I agree with that. The credit seems to an outside eye a paradigm potential suspect for a window dressing transaction, given that it was between related companies, on the last day of the accounting period, and of a significant amount. Even Mr Mah admitted (Footnote: ⁴⁴³) that with hindsight the explanation given to Miss Koo raised obvious risks of window-dressing and double-counting. I accept, as did Mr Swinson, that at the time there was no evidence or suspicion of fraud. However, as I have recorded above, Mr Mah accepted that cut-off was a risk area and Mr Spence accepted (Footnote: ⁴⁴⁴) that, when considering the ¥670m, D&T were addressing cut-off. Miss Koo should have been concerned to check not only that the cash had been received, but that it was accounted for in the right accounting period. She should have investigated the accounting treatment of the receipt, and probed the explanation given, carefully.

In the light of the accounting evidence I have described above, the explanation that Miss Koo had been given indicated that BFS had accounted for the payment in the wrong accounting period. A payment received on 1 October should have been accounted for on 1 October. The only reason to depart from this rule would be if BSL had accounted for the payment on 30 September. In that case, for BFS to account for the receipt on 1 October would lead to inconsistent treatments of the payment in the group consolidated accounts. Mr Swinson accepted that in such circumstances BFS might account for the receipt on 30 September, but only after the question had been discussed with BSL or its auditors and such treatment had been agreed.

There was no reason to conclude from the explanation given that BSL had booked the payment on 30 September. Nor was there from the Citibank fax. Indeed, BFS contended that the statement on the fax that the payment was value 1 October (as Miss Koo noted on the Bank Reconciliation) should have indicated to her that BSL intended the payment to be made, and therefore booked, on 1 October. However I accept D&T’s contention that the fax signified no more than that Citibank Singapore credited the payment to BFS’ account as of 1 October, and did not denote that the payment was expressly made for value 1 October.

It seems from the expert evidence that there are only two circumstances in which BSL would have accounted for the payment on 30 September. These are if (i) BSL instructed its bank to make payment to BFS and believed wrongly that the payment would be made on 30 September, or (ii) BSL agreed with BFS that both would book the payment on that date. Given that both telexed instructions and telegraphic transfers are effectively instantaneous, an auditor would not assume (i) without further clear evidence. (ii) would clearly need to be confirmed with BSL.

Thus Miss Koo needed clear evidence that, for one or other of these two reasons, BSL had booked the payment on 30 September. She in fact saw two further pieces of evidence, the daily activity statement for the 88888 account and the customer confirmations.

Mr Mah stated that it was not necessary for Miss Koo to trace the credit into a customer account at all. She was just verifying the cash balance (Footnote: ⁴⁴⁵). Mr Swinson’s evidence was that Miss Koo should have verified the explanation given to her. That included checking that the payment had been booked into the BSL segregated account in CONTAC (Footnote: ⁴⁴⁶).

Mr Spence’s evidence was that Miss Koo needed to look for the double entry (that is, to find where in BFS’ books appeared the credit corresponding to the entry in the cashbook) only if the BSL confirmation had not satisfied her that the ¥670m had been recorded on the right date:

“I think even to get to that point she would have to be unhappy about the consistency point we discussed this morning, about the agreement of the balances. So to get to that point where she was looking for the double entry, I think she would have to have some reason to be unhappy with the overall position (Footnote: ⁴⁴⁷).

She might be unhappy with the position, and therefore need further evidence, because she did not regard the confirmation as sufficient evidence or because she had not dealt with the client confirmations at that point.

When Mr Spence was asked why Miss Koo should have looked at the daily activity statement at all if she were not looking for the double entry, he was unable to suggest an alternative explanation (Footnote: ⁴⁴⁸). In fact, Mr Spence’s own conditions requiring Miss Koo to look for the double entry seem to have been satisfied. The BSL segregated account confirmation was not received until late on the evening of 7 October (see paragraph 599 above). So it was not available to satisfy her that the ¥670m had been recorded on the right date. On that basis, it followed from Mr Spence’s evidence that he agreed with Mr Swinson that Miss Koo should have checked where in BFS’ books the ¥670m appeared.

The daily activity statement showed that the ¥670m had been paid into the 88888 account and that was an account in the name of BFS (there was one other account in BFS’ name, a legitimate error account, but that was opened only on 7 October 1992 (Footnote: ⁴⁴⁹)). Mr Swinson accepted that Miss Koo could not have been expected to remember that the BSL segregated account had a number other than 88888. But the account name on the statement made it clear that it was not the BSL segregated account. The statement was therefore inconsistent with the explanation that the payment had been from or to BSL’s segregated account. Mr Mah agreed that he would have found the statement most remarkable, for that reason (Footnote: ⁴⁵⁰).

Mr Mah and Mr Spence accepted that the explanation from BFS, the Citibank fax and the daily activity statement did not justify treating the ¥670m as received on 30 September (Footnote: ⁴⁵¹). However they said that clear evidence that BSL had treated the payment as made on 30 September, which justified BFS in doing the same, was provided by the second further piece of evidence which Miss Koo saw, the BSL confirmations. As I have said, it seems likely that Miss Koo only saw the complete set of confirmations after 7 October.

Mr Spence accepted that there is no evidence in the audit files that Miss Koo relied upon the BSL confirmation in this context. Mr Mah did not rely upon the argument in his February 2000 witness statement, though he denied in evidence that this was because he did not rely on the point at the time. We have no evidence from Miss Koo.

Without Miss Koo’s evidence, we shall never know, but it seems to me likely that Miss Koo did not in fact probe the Bank Reconciliation any further at this point. She regarded herself as carrying out an exercise to audit the figure for cash at bank. She was not given any express window dressing test to carry out. She had established that the money had been received into BFS’ bank and into CONTAC. At that point, she simply accepted the explanation given to her without giving any thought to cut-off or window dressing. She treated the client confirmations as a separate exercise, part of auditing the liabilities owed to related customers.

However, the relevant question is whether the effect of the confirmations was such as reasonably to satisfy the concerns which a reasonably competent auditor should have had, whether Miss Koo had them or not.

The confirmations showed that BSL agreed the amounts owed to it by BFS as set out in the confirmation requests. The requests set out overall totals for each omnibus account, so the confirmations did not provide express confirmation that the ¥670m was included in any particular account. Mr Spence inferred that Miss Koo must have believed that both BFS and BSL had treated the figure for the BSL segregated account, as shown on the confirmation for that account, as including the ¥670m. Mr Spence said that she must have concluded that both treated the payment as credited to that account on 30 September in order to be satisfied that the balance sheet gave a true and fair view (Footnote: ⁴⁵²). The question is whether, on what we can infer Miss Koo knew at the time, and in the absence of any direct evidence from her, she was justified in so concluding.

There are two objections to D&T’s contention that she was so justified. First, I have found that Miss Koo should have reached the point where she had a concern that the ¥670m had been treated as paid in the wrong accounting period, and needed clear evidence to allay that concern. She might have obtained that evidence by contacting C&LL, or raising the point with BSL. She did neither. The fact that BSL confirmed the balances for its accounts was consistent with the explanation that had been given by BFS. But the evidence of Mr Swinson, which I accept, was that it was not sufficiently clear evidence to rebut the concern that the payment was being recorded on a different date to that on which BSL had recorded it (Footnote: ⁴⁵³).

The second objection is even stronger. Miss Koo knew from the Bank Reconciliation that the payment of the ¥670m was from BSL, to its segregated account. She had seen the daily activity statement recording the entry of the ¥670m into BFS’ books. I have found that she should have observed who was shown as the account-holder, if not the account number. That was BFS. Had she noticed that, she could not have regarded confirmation of the balance on the BSL segregated account as confirming that the two companies had treated in the same way a payment which BFS had paid into an account in its own name. Nor could the (forged) confirmation on the 88888 account have provided any comfort, given that the payment was from BSL, and to its segregated account.

There was some debate during the hearing as to whether this conclusion would be different if Miss Koo did the work on the confirmations before that on the cash. In the light of my finding that Miss Koo would not have received all the confirmations until after 7 October, this point does not arise.

However it would not affect my conclusion even if the order were reversed. In that case, Miss Koo would have established first that BFS and BSL agreed the balances on BSL’s accounts. She would later have found a large payment which appeared to have been booked by BFS on the wrong date (according to the normal rules) and on a different date to that on which BSL would be assumed (according to those rules) to have booked it. Without further enquiry, it would not be open to her simply to assume that BSL must have booked it on 30 September. But she then would have seen the credit to the 88888 account. That should have made her realise that either the explanation which she noted on the Bank Reconciliation was wrong (in that BSL must have paid the money for credit to the 88888 account) or the confirmations for both the segregated account and the 88888 account were wrong.

My conclusion is therefore unaffected by the order in which Miss Koo carried out the two exercises.

I therefore conclude that:

Therefore Miss Koo should not have accepted an accounting treatment of the ¥670m in BFS’ books as having been received on 30 September. She or those supervising her should have required evidence from BSL or its auditors that BSL had treated the payment as made on 30 September. The acceptance by Miss Koo of the explanation apparently given to her as to the accounting treatment of this payment without further investigation was, in my judgement, negligent.

D&T contended that, even if I find Miss Koo to have been negligent in relation to the ¥670m, the claim against them based on this issue fails on causation grounds. They argued that the onus is on BFS to establish that proper performance of D&T’s duty would have prevented further losses. D&T said that, if they had challenged the treatment of the ¥670m, the minimum they were required to do was either to require the sum to be treated as received on 1 October, or to confirm that BSL treated it as paid on 30 September.

D&T contended that, if they had required the sum to be treated as received on 1 October, Leeson would have forged a confirmation from BSL for the previous balance less ¥670m. The resulting discrepancy between BFS’ and BSL’s records would not have been picked up on consolidation because the BSL financial control department was too incompetent and/or would have accepted whatever explanation Leeson gave them. In support of this, D&T pointed to an impressive contemporaneous example of incompetence in the BSL financial control department’s treatment of sums due between BFS, BSL and BSJ. D&T said that BSL would have been no more competent over this issue.

On the alternative hypothesis, that D&T sought confirmation from BSL that it had treated the payment as made on 30 September, D&T said that it is likely that Leeson would have forged a fax to that effect; or possibly persuaded Mr Bowser to provide such confirmation. In support of this, D&T pointed to Mr Bowser’s inexperience in the area of inter-company reconciliations and his sharing the relaxed mores of the BSL settlements department. They alleged that, when the possibility was put to him in cross-examination, his objection was one of practicality rather than propriety.

BFS accepted that the onus of proving causation was on it. It said that the correct test was, not the minimum that D&T could have done without being negligent, but what it is most likely that they would have done if they had not been negligent. Mr Swinson’s evidence was that D&T should have both taken the ¥670m out of cash and investigated the transaction leading to it. BFS also challenged the hypotheses that D&T would have accepted a revised confirmation and that Mr Bowser’s evidence gave any reason to suppose that he would have agreed to rewrite BSL’s records to assist Leeson.

In my view BFS is correct as to the test to be applied: see BBL v Eagle Star, [1997] AC 191, at page 221-2. However that issue is rendered irrelevant, and D&T’s hypotheses are answered, by BFS’ further contention. This was that, if D&T had not been negligent, they would have discovered that the ¥670m had been credited, not to the BSL segregated account, but to the 88888 account in BFS’ name. In the light of that, neither a forged balance confirmation from BSL nor a forged fax stating that BSL had treated the payment as made on 30 September would have answered the questions D&T should have asked.

That seems to me correct. If D&T had appreciated the significance of the daily activity statement, they would have pointed out to Leeson that it was inconsistent with the Bank Reconciliation. If he removed the inconsistency by crediting the ¥670m to the BSL segregated account instead, the balance on that account would then no longer match the BSL confirmation. This would prove that BSL had not treated the payment as made on 30 September. The debit balance on the 88888 account would also increase by ¥670m. Leeson might have produced a further forged confirmation for the increased balance, but it is difficult to see how D&T could reasonably have accepted it in those circumstances.