Lehman Brothers International (Europe) v Exxonmobil Financial Services BV

These issues go to whether the Default Valuation Notice was given within the contractual window, which (as I have held) opened on 15 September 2008 when the Default Notice was given. Reference is made to what is said above in the introduction to the discussion of the issues.

Specifically, the issues concern when the window closed. The notice was received at 6.02 pm London time on 22 September 2008, which LBIE accepts was in time for the North American securities, because the time difference means that the markets there closed later in the day. However, LBIE says that the notice was not served in time for the non-North American securities, which account for the considerable majority.

EMFS says that the notice was given in time for all the securities, but if not, it accepts LBIE’s determination of the Appropriate Market by reference to the individual securities.

The first question is whether it was open to EMFS to determine a single Appropriate Market, that is “the global market” alternatively “the US market”, for all the equities and bonds. This is Issue 7, and LBIE says the answer is no. It depends on the construction of the contractual provisions as to “Appropriate Market” in the light of the expert evidence.

If the answer is no, it is common ground that LBIE’s security by security analysis applies.

If LBIE is not right about Issue 7, then, after Issue 8 (whether, on its true construction, the Default Valuation Notice conveyed a determination of the “global market” and that was permissible) and Issue 9(a) (LBIE’s case for an objective determination), there come the issues as to the Appropriate Market or Markets that EMFS would have determined if it had complied with the terms of the GMRA (Issues 9(b) and 10).

The remaining issues under this head are as to what Appropriate Market should be determined (Issue 10), and when close of business in the Appropriate Market occurred, and whether the DVN was given before the DVT (Issues 11 and 12).

In relation to the “would have” issue, at the court’s request the parties provided further written submissions after oral closings, which I have taken fully into account.

In determining the actual deadline for the service of the DVN, it is common ground that account has to be taken of the provisions as to “Appropriate Market” together with paragraph 14(b) of the GMRA which deals with the time at which notices are effective at the place of service. The latter provisions have been identified above.

LBIE accepts that the consequence of its case that the Appropriate Market has to be ascertained on a security by security basis is that the actual deadline for giving the DVN (taking account of paragraph 14) depends upon the time zones in the Appropriate Market together with that in the place where the DVN is to be given (London).

Where, as here, multiple securities in different time zones are comprised in the repo, working out the deadline on this basis becomes complicated. I need not go into the details, which were set out in LBIE’s closing submissions. But to take a particular example which was raised at trial, LBIE says that if the Appropriate Market is in Japan (which is 9 hours ahead of London) and the place for service is in London, then close of business in the market will be (on LBIE’s case) 3.00 pm in Japan, which is 6.00 am in London. LBIE says that as 6.00 am in London is not during working hours, the notice would need to be served in London before the close of business in London on the previous working day, which, in this case was the previous Friday, that is 19 September 2008.

I do not necessarily accept this analysis, which involves the concept of opening as well as close of business (and as a matter of detail, the time difference between London and Tokyo in September is 8 not 9 hours). But the analysis was not specifically challenged by EMFS, which did however rely on the complexity as supporting its own case. It says that LBIE’s construction is unworkable.

LBIE responds that a non-Defaulting Party can give more than one DVN so that, for example, securities sold in Japan could be the subject of a first DVN and securities sold in the United States could be the subject of a second DVN, provided that both were served before the deadline of close of business in London.

On the other hand, LBIE was asked during the trial whether in fact it ever received more than one DVN. No clear answer was given, and it is a reasonable inference that it did not.

EMFS’s contention is that it was open to it under the GMRA to determine a single “Appropriate Market” for the securities, being the global market, alternatively the US market. On this approach, it says that the complexities attendant on LBIE’s case are avoided.

LBIE responds that even on EMFS’s case the ‘Global Market’ and the US Market both close at 7.00 pm in New York. As that is midnight in London, it is by that time too late to serve a DVN in London so that the DVN must be served at least 5 hours earlier (assuming, as EMFS contends, a 7.00pm close of business in London), i.e. before 2.00pm in New York.

In short, EMFS’s response is that though a few hours may be shaved off the available time, this is not comparable to the complexity inherent in LBIE’s construction.

I need not decide the detailed points raised. The key point, to stay with the example of Japanese securities, is that if LBIE is correct that Appropriate Market has to be determined security by security, then the Default Valuation Notice given at 6.02 pm London time on 22 September 2008 fell well outside the window in respect of those securities. This is not in dispute.

It is sufficient to say that in my view, the exercise of giving a Default Valuation Notice is relatively straightforward on EMFS’s construction. There can be no question, however, that in the case of a portfolio of the diverse nature involved in this case, LBIE’s construction gives rise to a relatively complicated exercise to ascertain the deadline (or deadlines) for giving the notice.

EMFS says that the conclusion “that the DVN (or at least the first of multiple DVNs) needed to have reached LBIE by 5pm on Friday, 19 September 2008 … would come as an unwelcome and potentially calamitous shock to any reasonable industry user of the GMRA, conscious of being afforded ‘five dealing days’ to lodge the DVN”. It says that LBIE’s analysis is barely workable, and that where the analysis of commercial contracts leads to a conclusion that flouts business common sense, the analysis must be made to yield to business common sense.

My view is as follows. Clearly the court will seek to avoid a conclusion that “flouts business common sense”. Nevertheless, ultimately the question is one of construing the contract, and applying the terms.

There is some indication that timing issues on this aspect of GRMA 2000 have been recognised and addressed by those responsible for its drafting. The Guidance Notes for use with the Global Master Repurchase Agreement (2011 version) state that:

“The procedure for the calculation of the close-out amount has been amended in the 2011 Version, amongst other things, to provide more flexibility to the non-defaulting party as to the default valuation time. The non-defaulting party calculates the close-out amount by reference to an actual sale or purchase price or, if the non-defaulting party chooses, the market value of the securities, in either case at any time "on or about the Early Termination Date" (as opposed to the requirement under the 2000 Version that this be during the five dealing days following the occurrence of the Event of Default).”

This however was not the subject of comment by the parties, and no more need be said.

Paragraph 10(e)(i) of the GMRA defines the time window for giving a Default Valuation Notice:

“If between the occurrence of the relevant Event of Default and the Default Valuation Time the non-Defaulting Party gives to the Defaulting Party a written notice (a “Default Valuation Notice”) which – …”.

The requirements as in (A), (B), and (C) are then set out as summarised above. (No issue arises as to the form of the DVN.)

Paragraph 10(d)(ii) provides that:

“the “Default Valuation Time” means, in relation to an Event of Default, the close of business in the Appropriate Market on the fifth dealing day after the day on which the Event of Default occurs ...”.

Paragraph 10(d)(i) provides that:

“the “Appropriate Market” means, in relation to Securities of any description, the market which is the most appropriate market for Securities of that description, as determined by the non-Defaulting Party”.

The words “as determined by the non-Defaulting Party” give that party a discretion, and an issue arises as to how this applies in present circumstances.

In its closing submissions, EMFS took the questions in a somewhat different order than in the agreed list of issues, but I do not think it made any practical difference. I shall stick to the order in the list.

Each party’s case on the “global market” in equities and bonds was largely based on their expert evidence, which as developed by the parties in their submissions was as follows.

Among other things, Dr Ellis (EMFS’s expert) pointed out that listed equities are often not confined to a single stock exchange, with cross-listing on multiple exchanges. An investment bank such as JP Morgan, when instructed to liquidate a portfolio, will apply a best execution policy that assesses the trading opportunities on primary, secondary and OTC markets.

A point of particular importance in this context, Dr Ellis said, concerns the volumes of equities traded off-exchange, in over-the-counter (“OTC”) markets. The significance of OTC markets is that they have no firm opening or closing times, with trading possible whenever willing participants are available, globally and on a 24-hour basis.

EMFS submits that the question, however, is not whether a factual finding of a global market can be made, but whether a non-Defaulting Party like EMFS could rationally form the view that the Appropriate Market for the portfolio was a global one.

EMFS says that no-one suggests that traders in practice regularly engage in the exercise said to be mandated by the GMRA of making a determination of the “most appropriate market” for a given set of securities. The fact that Mr Ruiz may not have come across a determination of a global “Appropriate Market” is therefore a factor which carries no real weight.

EMFS says that the equities could be traded across the globe via a range of primary and secondary exchanges, as well as OTC: 132 could be traded in North America, 131 in Europe, 58 in Asia Pacific and 3 in South America. Thus, 35 equities were cross-listed or eligible for trading in three world regions and 91 in two world regions, with only 37 listed in just a single region. 137 of the securities were listed or eligible for trading in between two and seven countries. Only 26 securities were confined to a single country.

It would have been impractical, EMFS says, to determine an Appropriate Market separately for each security, e.g. because some securities were issued by a company in country A, listed in country B (and possibly cross-listed in country C and D) and the securities may have been held in custody in country E, quite apart from the fact that the security may have been denominated in the currency of country F.

As to bonds, EMFS points to common ground between the experts that bonds traded mainly OTC, where there is a single, 24-hour global market, subject only to country-specific restrictions. As Ms Nguyen stated, “a bond could, in theory, trade at any time or in any place, subject to regulatory requirements”. The existence of a market does not necessarily presuppose a uniform level of liquidity, but merely the possibility of trading. The characteristics of many of the bonds were international, so that it is beside the point that Ms Nguyen never encountered the concept of a global market during her career in banks.

EMFS says that Ms Nguyen “ducks” the need to identify an actual market in bonds in favour of classifying the world by reference to broad time zones. She adopts complex evaluation with five factors. She never convincingly explained why it should be assumed that a particular country will be the most appropriate market for a bond, in circumstances where she herself acknowledges that (leaving aside specific legal restrictions) bonds can be bought and sold anywhere in the world.

As with Mr Ruiz, there was no suggestion in either of Ms Nguyen’s reports that she had ever had to make a determination of Appropriate Market for the purposes of the GMRA.

For similar reasons, EMFS says that a determination of a US Appropriate Market would also have been permissible for the equities. About a third of the equities had their primary exchange in the USA, while approximately two-thirds of the remainder had an exchange open in the USA or were traded in OTC marketplaces in the USA. The remaining 39 equities could, in principle, have been traded OTC in the USA.

Mr Ruiz says the concept of a “market” in the definition of Appropriate Market must mean an exchange. EMFS says that though he says that a participant in the equities market would typically use a price on the primary exchange to value an equity, this assumes that the non-Defaulting Party will be a “seasoned user of Bloomberg”. But not all parties to the GMRA will be investment bankers.

Mr Ruiz’ analysis of Appropriate Market, EMFS says, in fact calls for the exercise of judgment in various ways, and gives rise to an exercise of considerable complexity on which different parties would have reached different outcomes.

EMFS says that a portfolio trade was a legitimate possibility (albeit not one which JP Morgan was prepared to agree to on its own account) and would likely have been in the US and have been denominated and settled in US dollars.

As to bonds, EMFS says that for factual reasons it is agreed that it is unnecessary for the court to make any finding on whether the US was an Appropriate Market.

EMFS invites the court to reject LBIE’s approach to Appropriate Market as unworkable, but it is sufficient for the court to hold that a rational person could permissibly have adopted a simpler and more practical approach.

Turning to LBIE’s case, it relies on Mr Ruiz’s evidence that he has never encountered or heard of the concept of the “global market”. Likewise, Ms Nguyen said that the concept of a “global market” for bonds is generally nonsensical.

LBIE says that in the Joint Memorandum, Dr Ellis appeared to be rowing back from his position, saying “This does not mean that there is an official, regulated “global market”. Rather the exchange linkages, technology developments and new venues ... form what is effectively a global market”.

LBIE says that the lack of reality in Dr Ellis’s “Global Market” case was demonstrated by the absence of any pricing produced by EMFS from the Global Market or produced at the time when the “Global Market” is said to close.

LBIE relies on that Dr Ellis’s reliance on OTC trading as demonstrating the existence of a “global market” in equities was suspect. The mere existence of OTC trading does not support his proposition. Also, his figures for the levels of OTC trading did not support his position.

He was “forced to accept in cross-examination that despite the number and volume of his exhibits, he had not produced a single page referring to the “Global Market”. Mr Sanders was forced to accept that there was not a single reference in the chronological bundles where he, JP Morgan or anyone else had referred to the idea of a “Global Market””.

As to EMFS’s alternative case, for the vast majority of the securities, there were higher levels of liquidity in markets other than the United States, and for many of them, there was almost no liquidity in the United States. 31 of the 163 Outstanding Equities had no listing at all in “North America”. For three of the bonds, their sale in the United States was actually prohibited during their primary market period.

The issue the court has to decide is a specific one, namely whether the “global market” if there is one falls within the GMRA paragraph 10(d)(i) definition of “the “Appropriate Market””. This term “means, in relation to Securities of any description, the market which is the most appropriate market for Securities of that description, as determined by the non-Defaulting Party”.

EMFS contends that the definition of Appropriate Market is deliberately non-prescriptive. Not only does it leave the choice of market to the innocent party, it also affords a wide discretion as to how, in the case of a portfolio of securities, the securities are classified: an Appropriate Market is to be determined for “Securities of any description”. It is thus for EMFS to decide, when assigning an Appropriate Market, whether to group the securities together on the basis of a fairly high-level description – e.g. simply grouping them collectively as equities or bonds – or on the basis of a much more granular description.

EMFS contends that the word “description” in the GMRA, upon which LBIE places much weight, is not a term of art. Where the GMRA seeks to encapsulate the concept of a security that represents the fungible equivalent of another security, the label it uses is the word “equivalent”, which is defined at para 2(t) of the GMRA. Accordingly, if the GMRA had been intended to provide that a separate Appropriate Market had to be designated for each individual security or its fungible equivalent, the definition of Appropriate Market would simply have required the identification of the most appropriate market “for each Security or its equivalent”.

However, EMFS submits, not only does the definition of Appropriate Market refer to “Securities”, in the plural presupposing that an Appropriate Market can be established on a collective basis for a group of securities, but it uses deliberately broad-textured language (“of any description”). The broad language used is fitting for the nature of the exercise that is required. It is not an exercise whereby the court must decide a single right or wrong answer. Rather, the definition of Appropriate Market ensures certainty by allocating the decision to EMFS.

It was thus for EMFS to decide whether to make a designation for the portfolio as a whole, to do so for sub-groups of securities, or to descend to the level of finding an Appropriate Market for each individual security in turn. EMFS’s determination was constrained by considerations of good faith and rationality. However, there would be nothing irrational when confronted with a portfolio of 181 securities in circumstances of extreme market stress, and in a context of some urgency as a result of the clock ticking on the five dealing days, about seeking to simplify the exercise as much as possible. On the contrary, it would be entirely rational to avoid a multiplicity of different DVTs, with the attendant risk of confusion, by taking a fairly high level approach to the question of Appropriate Market. This would also maximise the time for service of the DVN, which is self-evidently in the non-Defaulting Party’s rational commercial interests. EMFS contends that it was certainly within its discretion to choose one Appropriate Market for the equities and one for the bonds.

There were discussions at the time, ultimately abortive, about EMFS selling the equities to JP Morgan. If that was a possibility, it is hard to see (EMFS says) how it could have been impermissible for EMFS also to determine the Appropriate Market on a broadly portfolio-wide basis.

Moreover, there was no evidence of any description that any of the many counterparties around the world faced with the failure of Lehmans did anything other than serve a single DVN. This is not surprising, where paragraph 10(e)(i) refers to the non-Defaulting Party serving “a written notice (a ‘Default Valuation Notice’)”, i.e. in the singular. The provision goes on to identify a limited scenario where the non-Defaulting Party may serve more than one DVN which does not apply here. There is no suggestion anywhere else in paragraph 10(e)(i) that multiple DVNs may be served.

Even if it is nevertheless assumed that multiple DVNs might be permissible outside that narrow situation, EMFS says that a non-Defaulting Party is likely to have quite enough on its plate during the five dealing day window than to contemplate serving a succession of different DVNs.

EMFS submits that LBIE’s asserted need to identify a separate Appropriate Market for each of the 181 securities becomes all the more unsustainable when the difficult and diffuse nature of the exercise proposed by LBIE (to be carried out within the five dealing day window) is understood. (This has been described above.)

On its part, LBIE submits that its case is a simple one. In general each equity has a primary exchange and each bond has a home market, depending on its issuer, currency and so on. The time at which each exchange and market closes is well known and closing prices markets are easily obtainable. In consequence it is straightforward to determine for each security the time of close of business in the Appropriate Market for the purposes of paragraph 10(d)(ii), and in respect of bonds, who are market makers or regular dealers in the Appropriate Market and therefore qualified to provide quotations for the purposes of paragraph 10(e)(i)(B), and market prices as at close of business in the Appropriate Market in order to arrive at a valuation at the time of close of business in that market for the purposes of paragraph 10(e)(ii).

LBIE submits that paragraph 10(d) of the GMRA plainly envisages that the most appropriate market will be determined individually for each security of a particular description. That is consistent with the normal understanding of a description of a security, which involves identifying the issuer, terms, rights, and other familiar details.

LBIE submits that it is fundamental to the workings of a repo agreement under the GMRA that on the repurchase the original buyer is not obliged to return the actual securities which it purchased, but must return securities which are identical to those which it purchased. This requirement of fungibility is essential to the character of a repo. The concept of the “description” of a security is used in the GMRA to give effect to this fundamental need for securities to be fungible at the most detailed level.

Thus, by paragraph 2(t), “Securities” are “equivalent to” other Securities for the purposes of this Agreement if they are: (i) of the same issuer, (ii) part of the same issue, and (iii) of an identical type, nominal value, description and (except where otherwise stated) amount as those other Securities...” (emphasis added).

LBIE says that it is a consequence of these obligations that in the event of a default the non-Defaulting Party can quantify its loss by selling or obtaining quotations for Securities which are fungible with the Equivalent Securities. The word “description” is used in the GMRA to refer to the various detailed and precise identifying characteristics of a particular security.

It follows, LBIE says, that it is not possible under paragraph 10(d)(i) to determine the most appropriate market on a portfolio basis, or by classifying securities into extremely high level groupings such as “equities” and “bonds”. Paragraph 10 like the GMRA overall works on a security-by-security basis. Any other approach results in securities being valued at times and in markets which are not appropriate to them.

Although it is correct for EMFS to say that the GMRA provided that EMFS was to make the relevant determination, LBIE submits that the scope of that discretion is to be ascertained by the wording of the GMRA. EMFS was obliged to determine, for securities of a particular description, “the most appropriate market” for securities of that description. Neither of EMFS’s claimed determinations that the most appropriate market for all the securities were the global market or the US market can be valid under the GMRA.

As regards the evidence, I do not accept LBIE’s dismissive view of Dr Ellis’ opinion. Of course Mr Ruiz and Ms Nguyen are right to point to the absence of closing times, and the like, in the case of a “global market” as opposed, for example, to a domestic stock exchange, as well as to the absence of a single global regulator, and other matters.

But I consider that Dr Ellis was justified in drawing attention to the exchange linkages, technological developments, and new venues (he mentioned ECNs, that is electronic communications networks, and so-called “dark pools”), as well as OTC trading, and other instruments such as depositary receipts that enable trading away from exchanges. Although it is not possible or necessary to reach any conclusions as to the proportion of equity trading that takes place on an OTC basis, I am satisfied that it is substantial. All the bonds were traded on an OTC basis. Further, the possibility existed of disposing of the entire portfolio to a single buyer on what the experts referred to as a “risk” bid or price. Such a price may well be (as LBIE points out) based on pricing on primary exchanges, but from the seller’s perspective it is a single deal.

I accept therefore that it is meaningful to speak in a general sense of a global market in securities. However, this in itself is a conclusion of limited scope, because the question for decision is one of construction of a specific agreement.

EMFS says that the court should take account of the practical inconvenience of LBIE’s position as regards the practical task of giving a Default Valuation Notice in what may be adverse commercial circumstances. There is much force in these points, as the discussion above shows. On the other hand, they should not be exaggerated. As LBIE says, JP Morgan had no difficulty in identifying the markets for the disposal of the securities within a matter of hours of being instructed.

I have sympathy with EMFS’s complaint that LBIE has propounded its case on Appropriate Market with the sole purpose of showing that the Default Valuation Notice was late though most of the securities had already been sold. But as explained below, the term has an important application in the GMRA 2000 where the non-Defaulting Party wishes to rely on quotations to determine the Default Market Value of securities. In any event, the court cannot tailor its construction of the terms of a standard agreement to meet what on these facts may seem to be a hard case.

The issue is whether, on a true construction of the GMRA, it was open to EMFS to determine a single Appropriate Market for all the securities, being the “global”, alternatively the US market, or whether LBIE is correct to say that under the GMRA, the Appropriate Market must be ascertained on a security by security basis.

In support of its submission that a “global market” can fall within paragraph 10(d)(i) of the GMRA, EMFS emphasises the reference to “Securities of any description” as being deliberately non-prescriptive.

However, I agree with LBIE that this takes these words out of context. Appropriate Market in relation to securities of any description, is defined as meaning the market which is the most appropriate market for securities of that description. It is the reference to securities of that description that plainly in my view contemplates an exercise on a security by security, rather than a global, basis.

This construction is supported by the fact that the obligation of the purchaser under a repo agreement in GMRA form is not to return the actual securities which it purchased, but equivalent securities. By paragraph 2(t), securities are “equivalent to” other securities if they are of the same issuer, part of the same issue, and of an identical type, nominal value, description and amount as those other securities.

Further, the paragraph 10(d)(i) definition does not refer to an appropriate market, or even the appropriate market, but stipulates the most appropriate market for securities of that description.

The significance of this can be shown as follows. Although the defined term “Appropriate Market” arises in this case in the context of the deadline for serving a DVN, it has another (and important) role in the GMRA. Paragraph 10(e)(i)(B) permits the non-Defaulting Party to rely on quotations to determine the Default Market Value of securities. It requires:

“… quotations in respect of Securities of the relevant description from two or more market makers or regular dealers in the Appropriate Market in a commercially reasonable size (as determined by the non-Defaulting Party) …” (emphasis added)

The purpose, as LBIE says, is clearly to promote a proper determination of the Default Market Value by requiring that quotations are received from broker-dealers and the like in the market which is most appropriate for the securities in question. That clearly means the individual securities, not the portfolio on a global basis. This applies equally to bonds as to equities.

So while I agree with EMFS that it is meaningful to speak of a “global market” in securities in a general sense, it cannot in my view be a “market” for the purposes of paragraph 10(d)(i) of the GMRA.

Similarly, at least on the evidence before the court in this case, I am satisfied that the US market cannot be the Appropriate Market for the non-North American securities. LBIE made its case good in that regard by reference to the strikingly small percentages in a comparison between (a) the total volume of each Outstanding Equity traded in the United States over the 5 trading days from 16-22 September 2008, taken from Bloomberg, with (b) the aggregate of that United States volume and the volume traded on the primary exchange identified by LBIE. For three of the bonds, their sale in the United States was prohibited during their primary market period.

I further agree with LBIE that this view gains support from the fact that JP Morgan dealt with sales of the securities in regional markets on a security by security basis, as shown by the pre-trade analysis sent by the bank on 16 September 2008, and subsequently. The course of the sale process shows that the securities were divided up into particular regions for sale purposes. The existence of the “Country” column in the DVN demonstrates a similar approach. The volume of information provided in the DVN on a security by security basis confirms this.

As to EMFS’s submissions based on a sale of the whole portfolio, such a sale could (I consider) found a determination of Default Market Value under paragraph 10(e)(i)(A) of the GMRA. However, I agree with LBIE that it does not shed light on the construction of the term “Appropriate Market”, since such a sale is not a sale in a market.

EMFS points out that paragraph 10(e)(ii) refers to the giving of a Default Valuation Notice, i.e. in the singular, whereas LBIE accepts that its construction may mean (as explained above) that more than one DVN may needs to be given. However, the use of the singular in this context does not rule out the giving of more than one notice, if that is the course the non-Defaulting Party decides to take.

I have taken account of the other points made by EMFS detailed above, particularly the significance of OTC markets and the fact that there are primary and secondary exchanges for many securities. These points may be valid in themselves, but do not in my view affect the basic point of construction.

It is common ground that if LBIE is correct in its argument on Issue 7 that the GMRA requires a security-by-security determination of the most appropriate market then no further issues as to Appropriate Markets arise: specifically, Issues 8-10 do not arise: this is stated in the Agreed List of Issues.

The reason for this is that, if LBIE is right on Issue 7, any determination not made on a security-by-security basis would be invalid as not in accordance with the contractual framework. The question whether a determination of the global market or the US market for all the securities would be “rational”, does not arise on this basis, and I would not have upheld a case that a determination could be “rational” if it was impermissible under the contractual provisions. (The contractual discretion issue also arises in connection with the ascertainment of the Net Value of the securities, and is dealt with in that context below, including discussion of the Socimer case, and other authority.)

It follows that I accept LBIE’s case that the DVN was served in time in respect of the North American securities, but was not served in time for the non-North American securities. (This is subject to my conclusion as regards the UK and Irish securities, where the timing may work out differently.)

Though they do not arise on my findings, I shall express my conclusions on Issues 8-10 so far as these are not covered above, since they were subject of some argument.

EMFS accepts that Mr Sanders and his colleagues did not think in September 2008 about choosing an Appropriate Market for the securities: but says that the question would be whether EMFS had communicated its determination as to the Appropriate Market to LBIE, and this would fall to be ascertained objectively.

EMFS contends that, on a proper construction, the DVN would have conveyed to the reasonable reader a selection of a global market, both for bonds and for equities:

LBIE submits that:

I accept LBIE’s case in this regard. It is plain that the DVN did not convey a determination of a “global” Appropriate Market. Indeed, I think that the reverse is true.

This conclusion is consistent with how these points developed over the course of the dispute. Neither the global market point nor the US market point was raised at the outset. When EMFS first expressed its position on this issue in a letter of 6 January 2014, it said (understandably but wrongly) that the Default Valuation Time was intended to apply to the securities in the aggregate, and was not divisible according to the markets in which individual securities might be traded. On 11 June 2014, its solicitors said that EMFS had determined the Default Valuation Time by reference to close of business in the United States. EMFS’s primary case on the “Global Market” only appeared when it served its Defence in February 2015.

The points that I need determine are factual ones. Mr Sanders says that, if asked, he would have chosen a global market for both the equities and for the bonds. In response to the fact that JP Morgan had put the portfolio into “depots” which reflected different exchanges and therefore different markets, Mr Sanders said that these exchanges were relevant only to the settlement of the securities, and did not prevent there being a single global market for the purposes of selling the securities.

I do not accept his evidence that the exchanges were relevant only to settlement. I agree with LBIE that the standard settlement instructions (SSIs) prepared by EMFS at JP Morgan’s request proceeded on a security-by-security basis, using both ISIN and SEDOL numbers to identify the particular security as well as the position size. As LBIE says, this approach is the opposite of a portfolio-wide approach to the securities. The SSIs were divided up into 18 separate country “markets”, in each of which a local bank account of JP Morgan was authorised to receive the proceeds of the sale in local currency before remitting it back to EMFS’s US dollar account with JP Morgan in London, thereby demonstrating that the bank intended to sell the securities in the currency of that particular market and, it can further be inferred, on an exchange within that particular market.

Also, good witness though Mr Sanders in general was, had he considered the question of Appropriate Market in September 2008, I find that he would not have assigned a global Appropriate Market to the portfolio of equities and the portfolio of bonds. That would not have reflected his experience over the previous few days, in which under a conventional liquidation exercise carried out by JP Morgan in consultation with EMFS, the securities were being sold into particular markets. For the same reason, had he been told that a global market was not allowed, I find that he would not in the alternative have determined a US market.

In the event that it is held that the DVN did not convey a determination of a global Appropriate Market, or that Mr Sanders would not have made such a determination in September 2008, or that such a determination would have been impermissible (and I have decided each of these points against EMFS), its alternative case is that it would have made a determination of a US Appropriate Market.

For the reasons set out above, this alternative case also fails.

EMFS accepts that if the court concludes that EMFS would not in fact have determined a global or US market, or that a selection of such markets would have been impermissible, then the only alternative is to adopt LBIE’s case as to what market would have been selected for the securities on a security-by-security analysis.

I record that EMFS stated in closing that as regards the bonds, it is unnecessary for the court to make any finding on whether a US Appropriate Market would have been permissible. This is because in respect of the 6 A1 bonds, LBIE says that 2 of them were North American bonds and, as regards the remaining 4, 2 were restricted from sale in the US. The quantum of the 2 remaining A1 bonds is immaterial in the context of the case as a whole (and LBIE’s valuation is more favourable to EMFS than the actual sales proceeds). In respect of the A2 bonds, it is accepted by EMFS that the DVN was invalid, and so the question of Appropriate Market is not engaged. In respect of the A3 bonds, LBIE treats these all as North American bonds in any event.

This raises another timing point. The dispute is as to what “close of business” means for the purposes of paragraph 10(d)(ii) of the GMRA, which as noted provides that “the “Default Valuation Time” means, in relation to an Event of Default, the close of business in the Appropriate Market on the fifth dealing day after the day on which that Event of Default occurs...”.

LBIE contends that “close of business” means the close of business for dealing. EMFS contends that it means the close of business “for financial institutions generally”.

LBIE’s contention is that:

If LBIE is correct that “close of business” in paragraph 10(d)(ii) means “close of business for dealing”, EMFS accepts that the relevant times are as specified by LBIE.

EMFS contends that:

My conclusion is as follows. I accept the submission of EMFS that LBIE’s approach does not in fact result in clarity as to timing, which is said to be its main virtue. Further, I refer to what I said above as to the meaning of the term “close of business” where the words appear in paragraph 14 of the GMRA. The same points are applicable here. As a matter of construction, the reference in the definition of “Default Valuation Time” to “the fifth dealing day” does not qualify the meaning of “close of business” on that day.

On a plain reading of the words used, I prefer the case of EMFS in this regard. There is no reason to construe “close of business in the Appropriate Market” as meaning “close of business for dealing in the Appropriate Market”. On that basis, so far as I am aware, the position as regards the UK and Irish securities is not in dispute.

It follows from the above that the DVN was given in time for the North American Securities, and the 4 equities and 1 bond for which the UK or Ireland is said by LBIE to have been the Appropriate Market (it has not been suggested by EMFS that any other securities fall to be dealt with in the same way).

Though potentially difficult valuation issues arise, the parties’ time estimate was insufficient to allow for valuation to be properly covered in closings. In any case, the parties do not ask the court to carry out a security-by-security valuation or to calculate values for any securities. Instead, the parties have agreed a document submitted after closings the purpose of which is to identify high-level points of principle which each party asks the court to decide. The valuation issues are dealt with in this judgment on that basis.

The agreed valuation document re-orders the issues to put them into a more logical sequence.

The issue arises on the premise that (as I have held) the Default Valuation Notice was not valid in respect of some or all of the securities. Paragraph 10(e)(ii) of the GMRA provides that if “by the Default Valuation Time the non-Defaulting Party has not given a Default Valuation Notice, the Default Market Value of the … [securities] shall be an amount equal to their Net Value at the Default Valuation Time”. Because as previously explained, EMFS did not conduct this valuation exercise, the question is how it should be done for the purposes of this case:

The issue arises in this way because paragraph 10(e)(ii) refers the issue to the “Net Value” definition in paragraph 10(d)(iv), which ascribes a value to the securities in “the amount which, in the reasonable opinion of the non-Defaulting Party, represents their fair market value” (italics added).

In full, “Net Value” means:

“...at any time, in relation to any Deliverable Securities or Receivable Securities, the amount which, in the reasonable opinion of the non-Defaulting Party, represents their fair market value, having regard to such pricing sources and methods (which may include, without limitation, available prices for Securities with similar maturities, terms and credit characteristics as the relevant Equivalent Securities or Equivalent Margin Securities) as the non-Defaulting Party considers appropriate, less, in the case of Receivable Securities, or plus, in the case of Deliverable Securities, all Transaction Costs which would be incurred in connection with the purchase or sale of such Securities”

LBIE’s case is that the securities should be ascribed their objectively reasonable fair market value, and that this is an exercise which the court should carry out.

LBIE’s alternative case is that the securities should be ascribed a fair market value in accordance with the opinion which EMFS (acting rationally) would have formed had it conducted the valuation exercise required by paragraph 10(e)(ii) of the GMRA in September 2008. This in effect the same as the case of EMFS as set out under (1) above.

Both parties’ cases were elaborated in submissions made at the court’s request after the hearing as to what has been described as the “counterfactual”, that is, what EMFS would have done at the time had LBIE rejected its Default Valuation Notice on grounds that it was late.

The question is as to the effect to be given to a contractual discretion. In cases of the present kind, where the decision arises under a contractual term which gives one of the parties a discretion as to valuation, it is agreed that the leading authority is Socimer International Bank Ltd v Standard Bank London Ltd [2008] EWCA Civ 116.

In that case, a forward sales transaction agreement required a bank to carry out a valuation of a portfolio at the date of termination, but gave it a discretion as to valuation. Mistaking the nature of its obligations, the bank did not conduct the valuation, and the question was as to the consequences of this failure. As Rix LJ pointed out at [60]:

“When a contract allocates only to one party a power to make decisions under the contract which may have an effect on both parties, at least two questions arise. One is, what if any are the limitations on the decision-maker's freedom of decision? The other is, what is to happen if the contractual power was not in fact exercised at the time when the relevant party was obliged to make a decision?”

These are two distinct questions. The first goes to the bounds of the discretion. The second goes to the correct approach to be applied by the court if the discretion is for some reason not exercised, and the issue is as to the legal consequences. The questions do not necessarily both arise, though they did in Socimer and do in the present case.

As regards the first question, applying authorities such as Abu Dhabi National Tanker Co v Product Star Shipping Ltd [1993] 1 Lloyd's Rep 397 at 404 (Leggatt LJ), Rix LJ said at [66] that:

“… a decision-maker's discretion will be limited, as a matter of necessary implication, by concepts of honesty, good faith, and genuineness, and the need for the absence of arbitrariness, capriciousness, perversity and irrationality. The concern is that the discretion should not be abused.”

An example of the application of this principle is found in Fondazione Enasarco v Lehman Brothers Finance S.A. [2015] EWHC 1307 (Ch), in which the test of rationality was applied in the context of the 1992 ISDA Master Agreement. Under that agreement, the obligation of the non-defaulting party is “reasonably” to determine its total losses and costs. Unlike in Socimer, the determination had been carried out, and the challenge (which failed) was as to the validity of the determination. David Richards J said at [53]:

“… the relevant authorities now quite clearly establish that in considering whether the non-defaulting party has “reasonably determined” its Loss, that party is not required to comply with some objective standard of care as in a claim for negligence, but, expressing it negatively, must not arrive at a determination which no reasonable non-defaulting party could come to. It is essentially a test of rationality, of the type developed in the quite different context of public law duties in Associated Provincial Picture Houses Ltd v Wednesbury Corporation [1948] 1 KB 223: see Australian & New Zealand Banking Group Ltd v Societe Generale [2000] 1 All ER (Comm) 682, Peregrine Fixed Income Ltd v Robinson Department Store Public Co Ltd [2000] CLC 1328.”

What is meant by “rationality” in these circumstances? In Hayes v Willoughby [2013] 1 WLR 935, the issue was as to the meaning of “purpose” in the tort of harassment. The distinction between “rationality” and “reasonableness” was explained by Lord Sumption at [14], and the same distinction is essentially applicable in the present case:

“Rationality is not the same as reasonableness. Reasonableness is an external, objective standard applied to the outcome of a person's thoughts or intentions. The question is whether a notional hypothetically reasonable person in his position would have engaged in the relevant conduct for the purpose of preventing or detecting crime. A test of rationality, by comparison, applies a minimum objective standard to the relevant person's mental processes. It imports a requirement of good faith, a requirement that there should be some logical connection between the evidence and the ostensible reasons for the decision, and (which will usually amount to the same thing) an absence of arbitrariness, of capriciousness or of reasoning so outrageous in its defiance of logic as to be perverse.”

Contrary to LBIE’s case, therefore, the court does not seek to establish an “objectively reasonable” fair market value to the securities. The court applies a rationality test to the discretionary decision. As was said in a different context, the application of an objective test would largely deprive the counterparty of the benefit of the discretion which the contract gives it (Mercuria Energy Trading PTE Ltd v Citibank NA [2015] EWHC 1481 (Comm), Phillips J, at [120]).

In the present case, there was no decision, and the court is concerned with the “counterfactual”. The process was explained in WestLB AG v Nomura Bank International plc [2012] EWCA Civ 495, which was also a financial case in which the decision maker ought to have, but did not, conduct a valuation exercise. At [58], Rix LJ said:

“… in the present case the decision maker is not the court, with or without expert or other evidence to assist it: the decision maker, with an absolute discretion, is Nomura (whether Nomura International or Nomura Bank). In circumstances where it ought to have, but has not conducted a valid valuation exercise, the question, as the judge rightly put to himself, is how would Nomura have decided the matter, on or at least as at 30 September 2008, had it made a valid determination, honestly and rationally: Socimer at [65]-[66].”

Accordingly, I accept EMFS’s submission that the securities should be ascribed a fair market value in accordance with the opinion which EMFS (acting rationally) would have formed had it conducted the valuation exercise required by paragraph 10(e)(ii) of the GMRA. This is largely a question of fact.

The present case illustrates another bound to the discretionary decision. As well as being rational, the decision has to be one which is permissible under the contract. The extent of a contractual discretion depends on its terms, and the test of rationality applies within those terms (see e.g. Braganza v BP Shipping Ltd [2015] 1 WLR 1661 at [32], Baroness Hale, and at [54], Lord Hodge).

This point arises in connection with the “Appropriate Market” issue (see above). LBIE raises it in connection with the Net Value issue as well. In Socimer, the relevant clause said that value “shall be determined on the date of termination by the Seller” (see at [10]). LBIE also points to the use of the words “in its sole and absolute discretion” earlier in the clause, and the fact that the agreement provided for the issue of a “binding save for manifest error” certificate. LBIE says that the court appears to have contrasted the “wide discretion” in the Socimer agreement with that in an earlier version of the GMRA (see [117]). Based on this, LBIE submits that the valuation standard is more precisely stated in the GMRA than that in the Socimer case, and is capable of objective application. Since the exercise was not done by EMFS, the logical solution is for the court to determine how fair market value should be arrived at.

I do not think that this submission is correct. Although the GMRA, unlike the contract in Socimer, spells out in paragraph 10(d)(iv) in some detail how the determination is to be made, the provision emphasises rather than restricts the scope of the discretion. As EMFS says, the terms of the earlier version of the GMRA are not the same as those applicable in this case, and nothing is to be derived from this. (LBIE cites a statement that I made in Forsta v Bank of New York Mellon (ibid) at [298] as to the valuation discretion under the GMRA, but that involved the suggestion by an expert witness that the counterparty had a discretion not to conduct a good faith valuation at all, and was not directed at the issue under discussion here.)

There was some debate between the parties in their post-hearing submissions as to how the test is to be applied, but in view of my factual conclusions below it is not necessary to go into the detail. It is sufficient to say that the test is one of rationality, applied in a commercial manner in accordance with the authority set out above.

LBIE did however submit that Braganza, ibid, is authority for the proposition that both limbs of the public law Wednesbury test (Associated Provincial Picture Houses Ltd v Wednesbury Corpn [1948] 1 KB 223) apply to the contractual discretion exercisable in this case, and thus that it is necessary, under the first limb of Wednesbury (which focuses on the decision-making process) for the contractual decision-maker not to leave out of account matters which it ought to take into account or to take into account matters which it ought to leave out of account.

However, as EMFS says, in Braganza at [32], Baroness Hale (for the majority) expressly left open the question of the extent to which procedural judicial review objections could arise in commercial contracts, stressing the employment context of the case. See to similar effect the observations of Lord Hodge, with whom Lord Kerr agreed at [54] and [55]. At [57], Lord Hodge said that the question in Braganza was whether a state of fact existed (i.e. the cause of death), which was the kind of question that courts were in a better position to review, as compared with questions like the grant of a performance-related bonus where there was “little scope for intensive scrutiny of the decision-making process”.

The contractual discretion in the present case is given to a commercial party to a contract with another commercial party on the wholesale financial markets where the decision is as to the valuation of securities in the case of default. The decision is one which can be (and may need to be) taken without delay, and in which the non-Defaulting Party is entitled to have regard to its own commercial interests. In this kind of situation, I do not agree with LBIE that Braganza requires the kind of analysis of the decision-making process that would be appropriate in the public law context. But it makes no difference to the result in the present case.

A final point is that LBIE says that there is no legal basis for a case that the securities should be valued in accordance with an opinion which EMFS “would have been entitled to form” (as distinct from “would have” formed). I do not understand EMFS to be arguing this point now.

Issue 16 is concerned with the point of principle. How it works out on the parties’ respective cases is explained in full under the relevant issues below, but in summary:

As explained above, Appendix 1 of the Default Valuation Notice which EMFS sent to LBIE on Monday, 22 September 2008 identified the 163 equities that had been sold, and valued them at their sale prices. The sale had mainly happened on the previous Thursday.

LBIE says that this cannot form the basis of the exercise required by paragraph 10(e)(ii), because it is not compliant with the GMRA to value the securities at a date or time other than the Default Valuation Time, and previous sales of identical securities by the non-Defaulting Party have no special status in the exercise, being no more relevant than previous trades by third parties in the same securities. In point of fact, LBIE says, the market had risen since the previous week. Accordingly, it is irrational and indefensible to use prices obtained in the previous week as market prices for the purpose of valuing liquid traded equities at close of business on Monday 22 September 2008. EMFS is, therefore, not entitled to say that it would validly have adopted under paragraph 10(e)(ii) the prices at which the A1 equities had been sold.

EMFS contends that it would validly have adopted under paragraph 10(e)(ii) the prices at which the A1 equities had sold. It was Mr Sanders’ unchallenged evidence that this is what he would have done. This would have been entirely rational in the circumstances prevailing in September 2008. Market conditions were stressed and involved huge uncertainty, particularly since the equity portfolio was weighted towards the financial sector. There had been a “drastic drop off in trading” on 22 September 2008 and a “drastic increase in the range of volatilities”. This meant that equities could not simply be valued by adopting the screen prices on 22 September 2008. Models did not perform well in distressed market conditions, and EMFS would have had to value the portfolio ex ante and to factor in the risk of adverse price movements over the time period needed to liquidate the portfolio. Rather than speculate about the size of the discount to be applied to screen prices, engage in the complexities inherent in a modelled approach, and to take upon itself the risks inherent in such an exercise, it would have been entirely rational for EMFS to ground its valuation in concrete data, being the recent sale prices realised by it for those very securities.

My conclusion is as follows. I accept the attraction of EMFS’s approach, not least since it avoids a windfall for either party. However, whilst the actual sale price on a day within the window provides an unimpeachable method of valuation if the Default Valuation Notice is given within the window (see above), where this does not happen, the exercise is one of ascertaining the fair market value of the securities at the Default Valuation Time. The parties are agreed that the Default Valuation Time was the close of business on the fifth dealing day, which is 22 September 2008. The fair market value of the securities on this day cannot be ascertained by reference to sale prices on an earlier day, for the straightforward reason that they may be different. Whilst I accept that the markets were stressed in the aftermath of the Lehman collapse, the evidence is that the situation was not such as to render the contractual provisions unworkable in this respect. I do not accept therefore that it would be either contractually permissible or rational to take the actual sale price.

LBIE’s case based on Mr Ruiz’s valuation methodology is that for most purposes, the closing price on the primary exchange on the correct day is all that is needed or used to value an equity holding. These are the prices which are widely published as the closing prices, which make up the many indices which depend on equity prices and which market participants use to value their books.

LBIE accepts that in determining the fair market value of the equities it is appropriate to make a market impact adjustment to the published closing prices to reflect the fact that if the securities had actually been sold in the closing auction which occurs at the end of the day on equity markets, those sales would have (very slightly it says) impacted the closing prices. Mr Ruiz considers that there would have been no issue as to whether those equities could have been sold in the closing auction on 22 September. He calculates the market impact adjustment using a methodology which he says is standard in the equities market and which (a) assigns an equity to one of four geographic zones, (b) calculates the proportion of the ADV the position size represented, (c) allocates those proportions to industry standard liquidity bands, and (d) assigns a market impact adjustment to each liquidity band.

This results he says in a market impact adjustment of 58 bps (0.58%). LBIE submits that Mr Ruiz’s approach is both the only reasonable approach before the court and the approach which JP Morgan would have recommended if they had been asked for advice by EMFS.

EMFS’s case is based on the valuation methodology of Dr Ellis. This assumes that no more than a pre-determined percentage of the Average Daily Volume of trading in a particular equity (“ADV%”) would be sold each day, reflecting the fact that the higher the ADV% represented by a holding of a security, the greater the market impact of selling it and the more unpredictable the size of the market impact becomes.

Dr Ellis has assumed that in the stressed market conditions of September 2008, 10% of ADV can be sold each day with zero market impact. On the basis of this assumption, he has calculated the number of days it will take (a “hedging period”) to sell down EMFS’S holding in each security, so as to minimise the market impact of selling a large holding on a single day. The “flipside” is that, while the market impact is managed, there is an increased opportunity cost (which reflects the risk that the price of the security will change between the time when the security is valued and the time when the security is sold) because of the greater time taken to liquidate the portfolio. The outcome of the methodology is an overall liquidity discount that can be applied to the closing screen prices for the securities.

Dr Ellis offered a “Method 1” which assumed sales starting on 22 September 2008, and a “Method 2” which assumed sales starting on 23 September 2008, on the basis that it would be too late to sell on 22 September at whatever price might have been achieved at the closing auction on that day. He regards Method 2 as the most reasonable approach, and as corrected by him during trial that is the case advanced by EMFS.

As corrected by him during trial, “Method 2” values the equities at US$185,598,181, which closely compares with the actual sales proceeds of US$186,315,103, though LBIE says that this confluence is because of the rise in the market over the intervening period.

According to LBIE, if Dr Ellis’s methodology is accepted, the only plausible course would be to accept his Method 1 (which uses closing prices on 22 September 2008) and to adjust his 10% ADV assumption to a more reasonable (but still, it says, implausible) 20% ADV. The result then is about US$400,000 below Mr Ruiz’s figure of US$201,818,374.

LBIE has two “threshold objections”. It says that EMFS would not in September 2008 have acted upon advice from Dr Ellis or upon his Methods. EMFS has advanced no evidence that it would have commissioned an expert report such as Dr Ellis’s report. Mr Sanders accepted that EMFS would have asked JP Morgan for advice for which, LBIE says, Mr Ruiz’s evidence is a good proxy. Further, LBIE submits that the court cannot place any reliance on Dr Ellis’s evidence, and particularly so in respect of the equities, in which he has no experience. Accordingly, it says, Dr Ellis’s Methods must be disregarded and Mr Ruiz’s valuation accepted.

Without prejudice to those points, LBIE says that the closing prices at the Default Valuation Time should be adjusted by a market impact adjustment of 58 bps as Mr Ruiz says and, conversely, should not be adjusted according to Dr Ellis’s Method 2 or his Method 1.

Mr Ruiz has taken the closing prices at the Default Valuation Time and made what LBIE says is a “conservative adjustment” of 58 bps for the market impact of selling the portfolio if it had been sold at that time. Because he assumes a sale of the equities into the closing auction at that time, Mr Ruiz does not need to factor in opportunity cost. This approach replicates the value which would have been obtained if EMFS had instructed JP Morgan to sell the portfolio at close of business on the fifth trading day. It is therefore precisely targeted at the exercise mandated by the GMRA of producing a fair market value for the portfolio at the Default Valuation Time.

Dr Ellis’s methods do not produce a fair market value at the Default Valuation Time, as required by the contract, because, in the case of Method 2, all the pricing is taken from projected prices the following day or later and, in the case of Method 1, part of the pricing is taken from such later projected prices. Neither method is a valuation at the Default Valuation Time, and instead constitute projected outcomes of a liquidation strategy for the portfolio.

As a liquidation strategy, LBIE submits that both methods are perverse and irrational. They are not conservative in approach. By assuming that sales are made at a very slow rate, they manufacture enormous opportunity costs (2.3% or 230 bps even on corrected Method 1). They achieve this result by ignoring the trade-off between market impact and opportunity cost which in any trading strategy is the epitome of irrationality.

Dr Ellis’s methods make no allowance for market impact cost at all. His assumption of sales restricted to a maximum of 10% ADV per day, LBIE says, flies in the face of the strategy advised by JP Morgan which adopted a fast rather than a slow rate of sales, and resulted in sales of 96% of the portfolio within one trading day.

Dr Ellis’s corrected Method 2 produces a discount to market prices at the Default Valuation Time on the relevant date of 8.6%, amounting to over US$ 17 million. This is an extraordinary and unrealistic discount, LBIE says, to apply to heavily traded liquid securities. The discount is plainly irrational, and a Method which produces such a discount cannot be rational.

In summary, LBIE says, “Dr Ellis’s methods are neither fish nor fowl: they do not produce a determination of “fair market value” at the Default Valuation time and as liquidation strategies they are unrealistic, perverse and irrational”.

LBIE says that the court should not infer that JP Morgan would have given the same advice as Dr Ellis. If EMFS wished to rely on evidence about that, it was for EMFS to adduce the evidence. Because Mr Ruiz has lengthy experience as a market practitioner, his evidence can be accepted as a proxy for the advice that another market practitioner such as JP Morgan would have given if asked for a conservative valuation at the Default Valuation Time.

LBIE therefore asks the Court to find that Mr Ruiz’s proposed adjustment should be accepted and Dr Ellis’s should be rejected. The answer to the “would have” question is that EMFS would have asked JP Morgan for advice and their advice would have corresponded to that given by Mr Ruiz. LBIE does not have to show, as EMFS suggests, that Mr Ruiz’s approach is the only rational approach. The court should adopt it because it is the only rational approach the parties have put before the court.

EMFS contends that it would have been rational to apply a discount to account for market liquidity, calculated by assuming the sale of the equities over the course of a hedging period whereby no more than a certain percentage of ADV is sold per day. Such an approach reduces the market impact but increases opportunity cost. Given that this approach previously represented LBIE’s pleaded case as to “the methodology that a trader would use”, LBIE cannot be heard to denounce it as irrational.

EMFS says that Dr Ellis’ inputs are not irrational. As to the assumption of 10% ADV being sold per day, LBIE had previously pleaded a sale of 20% ADV per day. In closing, LBIE described the difference between 10% and 20% ADV as a “minor adjustment”.

EMFS says that Dr Ellis’ assumption on his Method 2 that trading would not begin until the next dealing day is logically unimpeachable. If the market has closed for business, it will not be possible to sell any securities until the next day, or, at least, that would be a serious risk.

Contrary to LBIE’s case, the fact that the GMRA requires the securities to be valued at the DVT, not sold at the DVT, means that it is permissible to take into account, when valuing the equities at the DVT, the fact that it will not be possible to sell them until the next day and that prices may have moved in the meantime.

Dr Ellis has not ignored the trade-off between market impact and opportunity cost, but has chosen a method cautious on market impact in order to arrive at a more conservative valuation, protecting EMFS from risk.

LBIE’s case conflates (a) EMFS’S sales strategy in September 2008, when it was concerned to extract as much cash as quickly as possible from the collateral and (b) its valuation methodology, on the assumption that it has not filed a valid DVN. The latter would involve a cautious and conservative approach, designed to avoid EMFS being left bearing any risk.

EMFS disputes LBIE’s case that the only rational approach is that of Mr Ruiz. The assumption of a sale of the whole portfolio in a single closing auction on 22 September 2008 is not one that is required by the GMRA, which recognises that a range of methods and sources can be used, or one that would arise in the real world. It involves ignoring Mr Ruiz’ own rule of thumb against selling securities with an ADV% in excess of 50% in the closing auction. Moreover, there is good reason to believe that a sale of the whole portfolio in the closing auction would be infeasible.

LBIE’s case that Mr Ruiz’ approach is the only rational one is also undermined by the fact that his market impact figures are non-forensic, based on no standard text or independent source, guidance or commentary. They were also not adjustments he would have used as a trader, since he would then have used a software called WebBench to which he no longer had access. There is plainly room for reasonable disagreement about the size of the market adjustment to be applied, with greater unpredictability about the market adjustment as ADV% increased and as a result of market conditions in September 2008.

EMFS says that Mr Ruiz’ overall adjustment of just 0.58% to closing prices on 22 September 2008 cannot be said to be the only rational view. LBIE accepts that JP Morgan’s discount for buying the whole portfolio would have been at the very least more than 150 basis points.

LBIE’s criticism based on Dr Ellis’s lack of experience as an equities market trader makes sense only on the mistaken premise that the non-Defaulting Party is required to have the experience of an equity trader or use an equity trader’s advice for valuation. This is not a requirement under the GMRA and it is obvious that had EMFS asked Dr Ellis or someone of comparable experience or expertise to advise it on valuing the portfolio, that would have been rational and therefore within EMFS’S discretion.

In summary, EMFS says that the court should find that it would have determined the Net Value in accordance with Dr Ellis’ lower bound valuation of the equities.

As to LBIE’s suggestion that EMFS seeks to advance a new case to the effect that it would have obtained advice from JP Morgan which would have been to the same effect as the evidence of Dr Ellis, it is LBIE that has put forward a new case as to what EMFS would have done in September 2008. EMFS does not accept the premise that Mr Sanders would have asked JP Morgan to undertake a valuation of the equities. It was not for EMFS to adduce evidence, because no case had been advanced by either party on the pleadings to the effect that JP Morgan would have been asked for a valuation. LBIE has adduced no evidence of fact to make good its new case.

However, EMFS says that there are compelling reasons why the court should conclude, if it decides that the correct question is as to what advice JP Morgan would have given, that a conservative valuation would have reflected Dr Ellis’ lower bound valuation.

LBIE’s “threshold objections” go partly to its attack on the credibility of Dr Ellis, and its submission that the court must in the circumstances accept the evidence of Mr Ruiz. As stated elsewhere, I do not accept this proposition. It is correct that he does not have the expertise of Mr Ruiz in share trading. But as EMFS says, the non-Defaulting Party may not be an equity trader, and there is no requirement under the GMRA that an equity trader’s advice must be used for valuation.

It is useful to begin with the hypothetical situation against which the court must answer the question raised in Issue 17(b). In fact, the Default Valuation Notice given by EMFS on 22 September 2008 relied on the actual sale prices of the equities achieved by JP Morgan a few days earlier. There is no dispute that the sale was properly carried out, or that in the wake of the Lehman Brothers collapse, EMFS had to look to the collateral to recover its repo loan of USD$250m.

In point of fact, there was no objection to the notice by LBIE for some years, but the hypothetical is on the basis that LBIE did object at the time on the ground that the notice was late (by hours, rather than days or weeks, but nevertheless late), and that the hypothetical objection was (as I have held) justified in the case of the non-US securities. On that basis, the question is what EMFS would have done, acting within the contract, and the confines of rationality, if the objection had been taken at the time.

The commercial effect of LBIE’s case is that hypothetical sale proceeds (slightly adjusted for market impact) remain the basis for the valuation, but that the date is moved forward to a hypothetical sale a few days after the sales that actually happened, so that LBIE in effect gets the benefit of the rise of the market in the meantime. LBIE says that this approach replicates the value which would have been obtained if EMFS had instructed JP Morgan to sell the portfolio at close of business on the fifth trading day: “It is therefore precisely targeted at the exercise mandated by the GMRA of producing a fair market value for the portfolio at the Default Valuation Time”.

I do not think that the analogy with the sale by JP Morgan is correct. In the stressed conditions following the collapse of Lehman Brothers, EMFS was anxious to sell the securities as quickly as possible. But in the hypothetical circumstance that LBIE rejected its valuation based on such sale, there is no necessary reason to impose a valuation based on an immediate sale on the fifth trading day. By rejecting the valuation based on the actual sale proceeds, LBIE necessarily opens the question up. As EMFS submits, its valid concern in such circumstances would have been to minimise the risk to itself, and this would not restrict it to any particular method.

In support of LBIE’s case, there is no dispute that in general terms, Mr Ruiz’s methodology provides a proper basis for valuation, or that it is standard market practice to value equities (and especially liquid equities) by reference to their published closing prices on the relevant valuation day. Further, his expertise in the field is not in doubt.

But it has not been suggested that this is the only basis for ascertaining the “fair market value” of shares. This is reinforced by the terms of GMRA. Under the contract, the exercise in determining the “fair market value” is a broad one. The non-Defaulting Party is entitled to have regard to such pricing sources and methods, which may include without limitation available prices for securities with similar maturities, terms and credit characteristics, as it considers appropriate.

Nor is there any dispute that EMFS would have been fully entitled to value these securities on a conservative basis. LBIE says that Mr Ruiz’s valuation is conservative, but the market impact adjustment he makes to the closing prices is 58 bps, or 0.58%, which the evidence suggests is minimal (JP Morgan’s discount for buying the whole portfolio would have been more than 150 bps, though this proposal was not pursued).

As to the hypothetical facts, it seems to me that the debate between the parties as to whether EMFS would have taken advice from JP Morgan, or experts such as Mr Ruiz, or Dr Ellis, is inconclusive. I do not accept the threshold criticism made by LBIE that “EMFS has advanced no evidence that it would have commissioned an expert report such as Dr Ellis’s report”. By the nature of the exercise the parties put to the court, the choice incorporated in Issue 17(b) is between two methodologies, and inevitably so.

Only tentative findings are possible. Upon the raising of the hypothetical objection, EMFS would no doubt have taken advice. However, if that advice had come from JP Morgan, as LBIE says it would, I do not accept that Mr Ruiz’s evidence is a “good proxy” for its content. On the evidence in the case as a whole, I consider that EMFS would more likely have been told that there are various ways of calculating the “fair market value” of equities within the ambit of the relevant provisions of the GMRA.

As EMFS says, it would not have been asking JP Morgan how to liquidate the portfolio as aggressively as possible, as was done in the previous week, but would have been seeking advice on valuation in circumstances where it would have had to proceed on the basis that any repayment might only ever come from the collateral.

As a matter of commercial sense, I agree with EMFS that it would have adopted the lowest legitimate valuation, provided that it was within the terms of the contract, and the confines of rationality. I have ruled out a valuation based on the actual sale price of the equities, which is what EMFS would undoubtedly have wished to do. The question for the court is between the valuation of Mr Ruiz incorporating a market impact adjustment of 58 bps, or that of Dr Ellis’s corrected Method 2.

As to the terms of the contract, LBIE argues in the post-hearing agreed summary of valuation issues that Dr Ellis’s approach is not a valuation, but a liquidation strategy. I do not think that this point was raised in quite this form before. In any case, I reject it. Both experts use a notional liquidation of the securities to arrive at a valuation. The difference is that Mr Ruiz assumes liquidation of all the securities in the relevant closing auction on the fifth trading day, whereas Dr Ellis assumes a liquidation over a period of days.

Although (as stated) Mr Ruiz’s methodology is in principle a proper one, he has to include a market impact adjustment. In other words, he does not suggest that closing prices alone can determine the issue. I accept the submission of EMFS that his adjustment of 58 bps is non-forensic, in the sense that it is not based on any stated standard, textbook, etc., and so involves a degree of subjectivity.

There may also be force in the question that EMFS raises as to the feasibility of a sale of this portfolio within a single closing auction. The sale of the shares in a single day was not achieved by JP Morgan.

In this regard, LBIE points out that over 85% of the equities by value was constituted by positions of particular equities which represented less than 25% of the ADV of those equities on the relevant exchange, and the weighted average of the ADV was 13.5%. As against that, EMFS says that the weighted average was higher than the average share of daily volume traded in the closing auction for certain markets, e.g. 4.2% (Japan) and 4.5% (US). Further, it is not in dispute that five of the equities had a % ADV higher than 50%, which Mr Ruiz himself would not have sought to sell in a single batch.

These points show that Mr Ruiz’s valuation methodology is not necessarily a straightforward analogue for what could in fact have happened at the time, had the DVN been rejected.

It is common ground that, when valuing equities, a trade-off must be struck between market impact and opportunity cost—or more simply, the depression of the price produced by a substantial sale as against the risk of a drop in prices over the so called “hedging period”. The main criticism of Dr Ellis is that his projected sale over a period gives rise to excessive opportunity cost. However, I agree with EMFS that under the “fair market value” provisions of the GMRA, the appropriate balance is and should be a matter for the non-Defaulting Party.

Further, there is nothing inherently unreasonable in the approach to valuation adopted by Dr Ellis. The striking thing about the case in this respect is that LBIE relied on the same approach, until its case was amended to reflect Mr Ruiz’s evidence. As EMFS says, given that this approach previously represented LBIE’s pleaded case as to “the methodology that a trader would use”, LBIE cannot be heard to describe it as irrational. This, in my view, makes LBIE’s case a difficult one on this point.

What Dr Ellis has done is to adjust the inputs, particularly lowering the notional daily sale to 10% from 20%, and commencing the exercise for Method 2 on 23 September 2016. Objection is taken to these adjustments on the basis that he ignores the required trade-off between market impact and opportunity cost to produce a discount to market prices at the Default Valuation Time on the relevant date of 8.6%. I accept that the discount is substantial, but the adjustments are not in themselves unreasonable. The result is close to the amount actually received for the shares, even if this reflects (as LBIE says) a rise in the market in the meantime. Neither the adjustment, nor the outcome, can be described as irrational. Nor, in my view, is commencing the calculation on 23 as opposed to 22 September 2008.

I do not therefore answer the “would have” question in favour of LBIE. I agree with EMFS that in the hypothetical situation posited, it would, and would have been entitled to, determine the Net Value in accordance with Dr Ellis’s lower bound valuation of the equities.

See above under Issue 16.

The facts as regards the valuation of the A3 bonds in the DVN are set out above. In summary, in respect of the five bonds for which JP Morgan had been unable to obtain any bids, EMFS valued them by reducing observed screen prices by 40%.

LBIE contends that the 40% discount was unreasonable and irrational:

EMFS contends that its valuation of the A3 bonds in the DVN was rational.

My conclusions are as follows. Both sides rely on conversations between JP Morgan and EMFS in the relevant week, i.e. up to 22 September. The bank was not asked for a valuation of the A3 bonds on that day, and as EMFS says, it said that it did not envy the task that EMFS had to perform.

However, whilst it is correct (as EMFS says) that on the morning of 19 September 2008 the bank mentioned a 40% discount, what was said was that the discount was “very, very issue specific”—i.e., dependent on the bond in question. The context of the remark was that the bank “had seen instruments trading anywhere from sort of a 10 to a 30% discount”.

Shortly afterwards, the transcript shows that Mr Ahsmann (EMFS’s Financial Operations Manager and, LBIE says, Mr Sanders’ senior) saying, “Big question is what is reasonable and 40% is definitely not reasonable”. It is correct as EMFS points out that this view was expressed during the liquidation process as distinct from the valuation process, but in my view it must have some relevance to valuation as well.

EMFS instructed the bank not to sell the remaining bonds at more than a 5% discount, but that afternoon the bank was instructed to discontinue trading. EMFS say that this was because of the lack of expressions of interest in the remaining bonds. However, I accept LBIE’s submission that it is also clear from the contemporary documents that senior management hoped that “the huge potential bailout being considered by the US government” could restore order to the market—the bailout is a reference to TARP (the Troubled Asset Relief Program).

On the afternoon of 22 September 2008, Mr Sanders says that the valuation of the bonds was “heavily debated”, and lists the factors that led to the 40% discount being applied. These were in effect the same points as EMFS relies on as set out above. He says that that the bonds were unsellable, and “were in that sense toxic”.

In this regard, though I am not asked to consider the bonds individually, two in particular have been cited by LBIE. Converium Holdings was part of the Berkshire Hathaway Group, which Mr Sanders agreed was possibly on a par with ExxonMobil (which is of course a prime name) itself. Statoilhydro (Statoil) is a Norwegian company, part owned by the Norwegian government, and rightly described as “quasi-sovereign” by Ms Nguyen. EMFS itself acknowledges that Statoil and Converium “might be expected to be creditworthy”, but this understates the position.

Whilst the evidence suggests that a 40% discount might be applied to toxic assets, it cannot in my view be right to describe bonds of this quality as in any sense “toxic”.

EMFS says that the bonds were illiquid, and Dr Ellis demonstrated that in the case of Statoil in which there were only 14 trades between January 2007 and 22 September 2008, the last being on 17 July 2007.

However, as LBIE says, Dr Ellis did not suggest that these bonds were thought to be distressed or requiring substantial discounts at that time. I agree that this evidence as to liquidity simply establishes that these bonds changed hands very rarely.

EMFS place strong reliance on the fact that JP Morgan had been unable to find a buyer for the A3 bonds, and had also been unable to elicit any firm bids for them. I agree that this is an important point.

However, there is a limit to how far the point goes. As LBIE says, EMFS did not test the market at discounts even approaching the 40% level. I have set out the facts above. EMFS says this is immaterial, and not how the bond market works. I do not agree with this. On the evidence, it is not all unlikely that buyers could have been found at discounts considerably less than 40%, given the quality of some of the paper. So whilst it is an important factor, I agree with LBIE that JP Morgan’s failure to sell or receive bids for the bonds is not conclusive.

The valuation of the A2 bonds showed a median discount of 15%, and EMFS says that given that the A3 bonds were more illiquid and generally had lower credit ratings, a higher discount was justified. In fact, in my view the credit ratings are not particularly insightful: for example, MBIA was rated as A- by S&P, but it is not in dispute that it was distressed. Nevertheless, I would accept EMFS’s general proposition, but nearly not to the extent of justifying a 40% discount applied to each of the A3 bonds including Statoil and Converium.

There is another significant factual matter. On 26 September 2008, so the Friday of the same week in which the DVN was given, the Credit Committee of EMFS met. The situation with Lehman Brothers was of course on the agenda. The committee received a presentation as to the “residual positions”—that is, the bonds which JP Morgan had been unable to sell.

The presentation showed in different columns the “Bloomberg Market Value” (the screen prices of the bonds), and the “Current Market Value”. The latter showed only a 20% discount on the A3 bonds, not a 40% discount. Mr Sanders suggested that, despite the heading, this was simply trying to show the committee that there was “hope” that the money would be recovered. I do not accept that, and agree with LBIE that EMFS’s Credit Committee was unlikely to be marking its books on a “hope” basis: if anything, it would be seeking to value the bonds as conservatively as was acceptable. Nor do I accept Mr Sanders’ suggestion that it was a mistake.

This is an internal, effectively contemporaneous valuation of the bonds, and I am satisfied that it shows the value that EMFS in fact ascribed to the bonds at the time.

It is correct, of course, that the test of rationality gives considerable leeway to the non-Defaulting Party. But it is nevertheless an important boundary, and if crossed, the other creditors wrongly lose out. Taken with the other evidence, I agree with LBIE that the current market value ascribed by EMFS to the A3 bonds on 26 September 2008 shows that the 40% discount in the DVN of 22 September cannot be justified on a rational basis.

This issue only arises in respect of the A3 bonds if the court concludes that the DVN was served validly and on time, as it has as regards these bonds (which LBIE treats as North American securities).

If the court finds that the DVN was served validly and on time, and also concludes that LBIE is correct to contend that EMFS’s 40% discount to the A3 bonds was invalid as being irrational, as it has, then LBIE’s case is that the A3 bonds fall to be valued as at the Default Valuation Time, 22 September 2008, and not earlier on 16 September 2008.

LBIE’s case is that the GMRA provides that if a DVN is not served in time, the valuation is carried out at the Default Valuation Time. There is no warrant for adopting a different approach if a DVN was served, but was invalid in whole or in relevant part.

EMFS contends that, if the DVN was in principle valid because it was served in time and by a proper method, but the court concludes that EMFS’s valuation of the A3 bonds in the DVN was invalid, then the court would have to consider what permissible Net Value EMFS would have given to the A3 bonds in the DVN, had it carried out a compliant valuation under paragraph 10(e)(i)(C) of the GMRA (and not under paragraph 10(e)(ii)). The practical significance of this is that any valuation under paragraph 10(e)(i)(C) would take place as at 16 September 2008 rather than 22 or 23 September 2008.

This point is taken by EMFS because on Dr Ellis’ figures, the A3 bonds were worth slightly less on 16 September 2008 than on 22 or 23 September 2008. It is not a good point for the reason given by LBIE. On the court’s findings, no valid notice was given in respect of the A3 bonds within the contractual window, so that they fall to be valued under paragraph 10(e)(ii) at the Default Valuation Time.

The answer is in substance the same as under Issue 14 above.

Ms Nguyen’s methodology starts from the observable price and makes adjustments for execution risk (where the observable price is not a traded price) and for the size of the position and for volatility (adopting whichever is the larger of these two adjustments). EMFS has a number of criticisms of her approach, including that as regards size and volatility discounts she applied only the higher of the two, and that she applied an undifferentiated liquidity discount of 5% (save for 5.67% for Lockheed Martin) against screen prices right across the A1, A2 and A3 bonds despite their very different characteristics.

Dr Ellis carried out a Discounted Cash Flow (DCF) analysis of the bonds. LBIE does not, and could not, take issue with that approach in itself, but it does take issue with the two adjustments which Dr Ellis added to it, namely his “intra-day volatility” discount, and his additional discount for liquidity which he set at 40% for his low valuation and 15% for his high valuation. LBIE says that the liquidity discount of 40% is simply copied from what EMFS did. It objects that this discount did not take account of the inherent characteristics of the bonds, and cannot therefore cannot be adopted wholesale into a DCF calculation which is aimed precisely at putting a value on the inherent characteristics of the bonds.

This is essentially the “would have” issue as it applies to the A3 bonds. In this regard, I refer to my conclusions under Issue 14 above. I refer particularly to the facts as to the meeting of the Credit Committee of EMFS on 26 September 2008, and the presentation as to the “Current Market Value” showing a 20% discount on the Appendix 3 bonds, not the 40% discount which it has sought to maintain. As stated earlier, this is an internal, effectively contemporaneous valuation of the bonds, and I am satisfied that it shows the value that EMFS in fact ascribed to the bonds at the time.

I would not accept Dr Ellis’s 40% liquidity discount in respect of the A3 bonds. I agree with LBIE that this is an attempt to transpose EMFS’s case into his valuation, without substantial validation.

My preliminary view, however, is that it is the EMFS valuation of 26 September 2008 which provides the best answer to the “would have” issue as it applies to the A3 bonds. I have expressed it as a preliminary view, because I am conscious that it does not adopt either alternative offered under Issue 20(b). I would ask the parties to consider this against the findings of the court as a whole. On this basis, they should be able to resolve this issue without further findings from the court.

Although LBIE maintains its submission that EMFS could not validly adopt the bid prices for the Appendix 2 bonds, it accepts that EMFS has “an apparently more defensible case in respect of these bonds”. This is because the factual case of EMFS (which I accept) is that only one quotation could be obtained by JP Morgan for these bonds, and this was “refreshed” by JPMS on 22 September 2008, so was current on that day.

That being so, I accept EMFS’s case based on the evidence of Mr Sanders that it would validly have adopted the quotations obtained by JP Morgan for the A2 bonds for the purposes of a valuation under paragraph 10(e)(ii). EMFS says, and I accept, that he would have had no reason to doubt that the quotations accurately reflected the price at which the A2 bonds would trade on 22 September 2008.

In these circumstances, there is no need to deal further with Issue 19(b) as to the valuation methodologies in respect of the Appendix 2 bonds, nor did the parties do so in the Agreed Document.

LBIE is content for this issue to be answered “yes”, in accordance with EMFS’s case. It does so because Ms Nguyen’s valuations of the Appendix 1 bonds are lower than the actual sale prices. (Issue 18(b), which deals with alternative valuations will then not arise.)

I record that LBIE adds that, “If LBIE’s case is accepted in relation to the Appendix II and Appendix III Bonds then, in order to be consistent, LBIE is willing to adopt Ms Nguyen’s conservative valuations of the Appendix I Bonds, which are lower than the sale prices, but it is not necessary for the Court to make any finding in this respect.”

Since it is common ground that the court is being asked to determine only the high-level points of principle on valuation, the net principal sum payable will be a matter for consideration after judgment. The parties will doubtless be able to agree this.

Issues as to interest and legal expenses will also be a matter for consideration after judgment.

As the parties no doubt anticipate, on the large number of issues placed before the court some points have been decided in favour of one party, and some in favour of the other. It is not practical or necessary to list all of these.

However, it may be helpful to indicate the result on some of the main points (making clear that nothing in the summary alters in any way the actual conclusions as expressed above). I accept LBIE’s case as to the validity of the 15 September 2008 Default Notice. I accept EMFS’s case that the Default Valuation Notice was validly served, save as to the Default Valuation Time. On the latter point, I accept LBIE’s case that it was not open to EMFS to determine a single Appropriate Market, ie a Global Market or the US Market for all of the securities. However, I accept EMFS’s case as to the close of business in this regard.

As regards valuation, and to the extent set out above, I accept EMFS’s case that the valuation needed to be rational, not objectively reasonable. I accept EMFS’s alternative case as to the A1 equities, and its case as to the A2 bonds. I accept LBIE’s case as to the A3 bonds to the extent that EMFS’s 22 September 2008 valuation was invalid.

Once the parties have had a chance to consider this judgment, consequential matters may arise. I am grateful to the parties for their assistance.