LB Re Financing No. 3 Ltd v Excalibur Funding No.1 Plc & Ors

This Part 8 claim by the administrators of LB Re Financing No 3 Ltd (“LB3”) raises issues as to the true construction of Conditions in a securitisation Trust Deed dated 23^rd May 2008 constituting €2,166,541,000 Class A Notes and €722,181,000 Class B Notes, both originally due April 2054.

The claimant LB3 is the holder of the Class B Notes. The first defendant Excalibur Funding No. 1 plc (“Excalibur”) is the Issuer of the Notes. The second defendant Deutsche Bundesbank (“DBB”) is the holder of the vast majority of the Class A Notes. The third defendant US Bank Trustees Ltd (“USBT”) is the trustee under the Trust Deed.

The dispute between the parties is, in summary, whether an Event of Default under Condition 10(a)(iv) (“Inadequate Par Coverage”) had occurred and was continuing when, with effect on 8^th February 2011, USBT as Trustee served on the Noteholders a Notice of Acceleration, the consequence of which, if valid, was that both classes of Notes became immediately due and repayable. Both DBB and USBT assert that such an Event of Default had occurred and was then continuing. LB3 denies that any Event of Default had occurred, and separately denies that, if it had, it was continuing on 8^th February 2011.

There is no issue of fact between the parties. The question which I have identified turns entirely upon the construction of the Conditions, and in particular Condition 10(a)(iv).

Excalibur is a single purpose vehicle formed solely to act as issuer of the Notes, which constituted a securitisation of a portfolio of underlying collateral debt obligations (“CDOs”) acquired with the proceeds of the Notes from a number of Lehman group companies, and which consisted mainly of real estate, loans and mortgage backed securities relating mainly to commercial properties in a number of European countries. I will refer to the programme as “the Excalibur Securitisation” and to the underlying CDOs as “the Excalibur Portfolio”.

LB3 was the initial purchaser of all the Notes issued by Excalibur. The Class A Notes were acquired at par and the Class B Notes at a discount. LB3 then transferred the Class A Notes to Lehman Brothers Bankhaus AG (“LBB”) under a repurchase financing arrangement in exchange for the advance of monies to LB3. LBB in turn pledged the vast majority of the Class A Notes as security to the DBB, as part of the pool of collateral securing that bank’s funding to LBB. Following LBB’s insolvency in late 2008, DBB took possession of those Class A Notes.

In the absence of an Event of Default, the Conditions contained no provision whereby the Class A Noteholders could require early redemption (that is before 2054). Nonetheless, since the maturity dates of the underlying CDOs were mainly much shorter, it was envisaged that rolling pro rata redemption upon the maturity of the underlying CDOs would occur over time, pursuant to the mechanism which I will shortly describe in detail, subject to the qualified right of the Class B Noteholders to require that the maturity proceeds of CDOs be deployed in the acquisition of substitute collateral.

Pending redemption, the Class A Notes paid a floating rate coupon quarterly. It is unnecessary to describe the rights of the Class B Noteholders in detail. It is sufficient to say that, in substance, the Class B Noteholders held the equity interest in the Excalibur Portfolio subject to the prior rights of the Class A Noteholders to interest and, upon maturity or earlier redemption, to repayment of principal.

Viewed in the round therefore, the Excalibur Securitisation was a structure devised and put into place initially entirely within the Lehman group (save for the inclusion of USBT’s predecessor as Trustee), for the purpose of enabling the group to use the Class A Notes as security for its Euro borrowing. The first arms’ length transaction in relation to the Securitisation occurred upon the pledge of the bulk of the Class A Notes by LBB to DBB, as I have described.

Condition 3 contained an elaborate scheme for the receipt, handling and distribution by Excalibur as Issuer of all monies, both of a capital and income nature, derived from the Excalibur Portfolio. It required the Issuer to establish no less than ten separate accounts, of which four are relevant for present purposes, namely the Collection Account, the Principal Account, the Interest Account and the Payment Account: see Condition 3(j). Broadly speaking, all receipts are to be credited to the Collection Account. Thereafter receipts of an income nature are to be transferred to the Interest Account and receipts of a capital nature (including the proceeds of any realisation or enforcement of the CDOs) to the Principal Account. Thereafter, payments are to be made out of those accounts in accordance with detailed provisions as to priority on Payment Dates defined as being the 28^th January, 28^th April, 28^th July and 28^th October in each year, as well as on the Maturity Date (28^th April 2054) or any earlier Redemption Date (as defined). Separate schemes of priority, generally known as waterfalls, are set out in relation to the Principal Account and the Interest Account,. The interest waterfall is set out in Condition 3(c)(i) and the principal waterfall in Condition 3(c)(ii). All payments out are to be routed through the Payment Account. Pending payment, the Trustee is authorised to invest the balances standing to the credit of any of the Accounts in what are described as Eligible Investments (essentially short-term commercial paper and money market debt), but upon terms designed to ensure that those investments can be realised prior to the next Payment Date.

In order to ensure an efficient and orderly distribution from the Accounts on each Payment Date, the Conditions made provision for calculation of the amounts payable to be made on a Determination Date (defined as three business days prior to each Payment Date) by the Collateral Administrator (Bank of America National Association (London Branch)), and for the transfer of all payments thus identified from (among others) the Principal Account and the Interest Account to the Payment Account one business day before the Payment Date: see Condition 3(e).

Condition 3(c)(i) sets out no less than seventeen successive priorities for distribution of sums in the Interest Account (described as Interest Proceeds), in stages (A) to (Q). They may be sufficiently summarised as providing first for the payment of fees and expenses, secondly for ongoing hedging costs, thirdly for payment of interest due under the Class A Notes, fourthly for termination hedging payments and fifthly for payments of interest due under the Class B Notes.

There is however inserted into that abbreviated list, immediately after payment of interest due under the Class A Notes, a provision in the following terms:

“(G)

in the event any of the Coverage Tests is not satisfied on the related Determination Date to the redemption of the Class A Notes, to the extent necessary to cause the Coverage Tests to be met if recalculated following such redemption;”

There are two Coverage Tests, namely the Class A Par Value Test and the Class A Interest Coverage Test. For present purposes the latter may be disregarded. I shall have to describe the nature and effect of this stage in the interest waterfall, and the Class A Par Value Test, in due course.

This is set out in Condition 3(c)(ii) and consists of seven successive stages, lettered (A) to (G). Stages (A) and (C) require any shortfall in payment of fees, expenses and hedging payments from the Interest Account to be made good from the Principal Account. Stage (B) is a freestanding provision for payment of certain fees from the Principal Account. Stage (D) requires any shortfall in interest payments on the Class A Notes to be made good from the Principal Account.

Stage (E) provides for the balance to be applied in the redemption first of the Class A Notes and secondly of the Class B Notes, subject to a prior right of the holders of two thirds of the Class B Notes to direct that the amount be reinvested in the purchase of substitute CDOs. That right is hedged about with a raft of conditions, the most important of which is that such purchases should not adversely affect the rating of the Class A Notes. The final two stages of the principal waterfall are of no relevance.

Condition 7(d) inserts an additional stage into the principal waterfall in the event that either of the Coverage Tests is not met, between stages (C) and (D), which requires the balance to be used in redeeming Class A Notes to the extent necessary to satisfy those tests. One of its effects, where applicable, is to disable the right of the Class B Noteholders to require the balance to be reinvested in the purchase of substitute CDOs under stage (E). More generally it serves to reconstitute the Class A Par Value Ratio in much the same way as stage (G) in the interest waterfall.

Since it is contemplated by Condition 3(e) that the Collateral Administrator will be able precisely to calculate all distributions from the Principal Account on each Determination Date, it is I think clear that any election by the Class B Noteholders in favour of the purchase of substitute CDOs has to be made in advance of the Determination Date and that, in the absence of such an election, the net amounts in the Principal Account (after payments under stages (A) to (D)) are to be applied towards the redemption of the Class A Notes, until redeemed in full.

A feature of the principal waterfall is that there can be no accumulation of monies in the Principal Account after each Payment Date (save perhaps for the purchase of substitute CDOs if the Class B Noteholders so elect). It will therefore be cleared out on a quarterly basis so that, on any Determination Date, it will consist of the proceeds of the realisation of CDOs during the immediately preceding quarter, less three business days. There is, in short, no scope for the accumulation of cash in the Principal Account other than within each quarter.

The Class A Par Value Test is defined as meaning:

“a test that shall be satisfied if as at any Measurement Date the Class A Par Value Ratio is at least 110 per cent.”

The Class A Par Value Ratio is defined as meaning:

“as at any Measurement Date, the ratio (expressed as a percentage) obtained by dividing (a) the Par Coverage Numerator by (b) the sum of the aggregate Principal Amount Outstanding of the Class A Notes and the amount of any unreimbursed Interest Advances.”

The Par Coverage Numerator is defined as meaning:

“on any particular Measurement Date, the sum in Euro as at such date of the following:

(a)

the aggregate of the Principal Balances of the Collateral Debt Obligations (other than any Defaulted Obligation or Discount Obligation);

(b)

with respect to each Discount Obligation, an amount equal to the purchase price (excluding accrued interest thereon) paid by the Issuer upon the acquisition of such Discount Obligation;

(c)

the aggregate of the Balances standing to the credit of the Principal Account; and

(d)

with respect to each Defaulted Obligation, an amount equal to the S&P Recovery Rate in respect of such Collateral Debt Obligation.”

The expression “Principal Amount Outstanding”, as used in the Class A Par Value Ratio, is defined as meaning:

“in relation to a Note of any Class on any date (i) the initial principal amount thereof (for the avoidance of doubt, being its initial face value), less (ii) the aggregate of all principal payments in respect of a Note of the relevant Class that have become due and payable and have been paid since the date of issuance of such Note.”

The purpose of the Class A Par Value Test, the special stage (G) in the interest waterfall and the additional stage in the principal waterfall inserted by Condition 7(d), both of which are triggered when that test is not met, is to divert, to the extent necessary to enable that test to be met in the future, a sufficient income stream away from the Class B Noteholders and towards the redemption of Class A Notes. The effect of redemption is, of course, to reduce the Principal Amount Outstanding of the Class A Notes so as thereby to reduce the size of the denominator in the Class A Par Value Ratio. The effect of that diversion of income and principal to redemption of Class A Notes is therefore designed over time to restore the collateralisation ratio between the Portfolio and the Class A Notes to 110 per cent, at the expense of the Class B Noteholders. Nonetheless, failing the Class A Par Value Test does not constitute an Event of Default under the Conditions.

It is necessary to look a little more closely at the detailed definition of Par Coverage Numerator, set out above. Clause (a) brings into account at face value all CDOs, other than those which were originally acquired at a discount or those which have since defaulted. Clause (b) values CDOs purchased at a discount to their original purchase price. Clause (d) values defaulted CDOs at the recovery rate certified in respect of them by the rating agency Standard & Poor’s Rating Services.

Clause (c) brings into account all monies standing to the credit of the Principal Account as at the Measurement Date. This will consist mainly of the proceeds of the redemption or enforcement of CDOs within the Portfolio. The Measurement Date is defined as having three limbs. Ordinarily it is the same as each Determination Date (i.e. three business days before a quarterly Payment Date). The definition also includes (although not relevantly for present purposes) any day on which a substitution or acquisition of a CDO occurs, or any business day requested on prior notice by S&P.

The effect of clause (c) is therefore that, on a Measurement Date falling just before a quarterly Payment Date, CDOs which have been realised during the preceding quarter will not fall entirely out of the account, because their proceeds will have been paid into the Principal Account in the meantime.

Condition 10(a) defines the following seven Events of Default:

Condition 10(a)(iv) is as follows:

“Inadequate Par Coverage: on any Measurement Date, the Par Coverage Numerator (without regard to clauses (c) of such definition and the provisos to such definition) falls below 100 per cent. of the Principal Amount Outstanding of the Class A Notes”

Condition 10(c) headed “Acceleration” provides that, if an Event of Default has occurred and is continuing, the Trustee may, and must if directed by the holders of the Class A Notes, give notice to the Issuer that all the Notes are immediately due and payable. Condition 11 then contains provisions for enforcement.

The Inadequate Par Coverage test differs from the Class A Par Value Test in four significant respects. First, its failure constitutes an Event of Default. Secondly, it is set at 100 per cent. rather than 110 per cent. Thirdly, the unreimbursed Interest Advances are not included within the denominator. Fourthly, and most importantly, the Balances standing to the credit of the Principal Account are, on the face of it, excluded from the Par Coverage Numerator. I use the phrase “on the face of it” because at the heart of LB3’s case is the submission that this literal exclusion of the Principal Account Balances from the definition of the Par Coverage Numerator produces such commercially absurd results that the court must strive for some other interpretation.

There are two obvious mistakes in Condition 10(a)(iv). First, the phrase “clauses (c)” contains the obvious typographical error of using the plural to describe a single clause. Secondly, there are no provisos to the definition of Par Coverage Numerator. I shall address the significance of those obvious mistakes in due course.

The first Determination Date in 2011 fell on 25^th January, three business days before the Payment Date on 28^th January. 25^th January was also therefore a Measurement Date for the purposes of Condition 10(a)(iv). There had by that date accumulated in the Principal Account since the previous Payment Date (28^th October 2010) no less than €140,595,961 from the realisation of CDOs. Of that sum, the Collateral Administrator certified on 25^th January that €127,739,207.72 was due to be paid on 28^th January to the Class A Noteholders by way of redemption, after meeting all prior payment obligations under the principal waterfall.

Following a literal reading of Condition 10(a)(iv) the Collateral Administrator duly excluded the whole of the €140 million odd from its calculation of the Par Coverage Numerator, with the result that it amounted to only 99.5% of the Principal Amount Outstanding of Class A Notes as at that date. If the €140 million odd had been included, the ratio would have been 107.8%. If, alternatively, the €127 million odd due to be paid by way of redemption three days later had been deducted from the Principal Amount Outstanding of the Class A Notes, the ratio would have been 107.7%. Furthermore, if a further calculation had been carried out as at 28^th January, the ratio would indeed have been 107.7%, there having occurred no other relevant changes than the payment of the €127 million odd by way of redemption.

The result of the calculation actually carried out by the Collateral Administrator was that an Event of Default was declared and, at the direction of the Class A Noteholders, USBT served an Acceleration Notice on 8^th February 2011. On the Measurement Date which occurred on 25^th April 2011, the Inadequate Par Coverage ratio had, even on the Collateral Administrator’s calculations, been restored to 101.5%, largely because a much smaller amount (€88.7 million odd) had in the intervening period arrived in the Principal Account, such that its exclusion from the Par Coverage Numerator was insufficient to tip the ratio below 100%.

LB3’s central submission was that, in the words of Lord Hoffmann in the ICS case, “something must have gone wrong with the language” of Condition 10(a)(iv). Mr Pascoe QC for LB3 submitted that it was commercially absurd to exclude a sum of cash (or equivalent) from the calculation of the Par Value of the assets representing the security, while at the same time ignoring that the same amount was known to be going to be paid by way of reducing the Principal Amount Outstanding of the Class A Notes in three days time. Whatever reason there might be for excluding assets of uncertain value, there could be no rational purpose in excluding cash (or equivalent Eligible Investments). Nor could there be any rational basis for rendering the passing of the Inadequate Par Coverage test dependent upon the unpredictable and (for collateralisation ratio purposes) irrelevant variations in the amounts of CDO realisation proceeds building up during any particular quarter, pending payment to the Class A Noteholders by way of redemption.

Mr Pascoe offered, in his skeleton argument, two alternative solutions by way of addressing the alleged absurdity. The first was to treat the whole of the phrase in parenthesis “without regard to clauses (c) of such definition and the provisos to such definition” as having been included from some earlier securitisation as the result of an obvious drafting mistake. The second was to treat money certified on a Measurement Date as money payable three days later by way of redemption as if it was indeed “paid” for the purposes of reducing the Principal Amount Outstanding which is the denominator for the ratio.

In oral submissions Mr Pascoe added a third, and apparently preferred, solution, based upon a forensic examination of an earlier securitisation (“the Duncannon Securitisation”) from which he submitted that it could be seen that the offending phrase had been drawn, in which clause (c) of the definition of Par Coverage Numerator referred to Discount Obligations. Thus he said, the court could identify an alternative construction of Condition 10(a)(iv) by reading “clauses (c)” as meaning “clause (b)” which is the relevant clause in the definition of Par Coverage Numerator in the Excalibur Conditions relating to Discount Obligations. He added that, even if I were not persuaded that the exclusion of the Principal Balances was commercially absurd, the adoption of the alternative relating to Discount Obligations was commercially obviously preferable, since Discount Obligations are, by comparison with cash, a type of impaired asset prima facie worthy of consideration for exclusion from a collateralisation ratio.

LB3’s final submission was that, even if it failed on the question whether an Event of Default had occurred on 25^th January, it was no longer continuing by 8^th February, since the payment of €127 million odd on 28^th January by way of redemption of the Class A Notes had by then restored the Inadequate Par Coverage ratio to an amount substantially in excess of 100%.

DBB’s case, supported by USBT, may be summarised as follows. First, there was nothing commercially absurd in excluding Balances in the Principal Account from the Par Coverage Numerator. Looking at the matter at the time when the structure was created, it could not be predicted what balances there would be on any future Measurement Date nor, more importantly, to what priority purposes under the principal waterfall those balances would fall to be applied. They might need to be used to meet substantial interest shortfalls or hedging liabilities, rather than for the purposes of redemption.

Secondly, by no process of construction (as opposed to rectification) could the unambiguous phrase “without regard to clauses (c) of such definition” simply be ignored. Nor was there any legitimate basis to treat the reference to (c) as an obviously mistaken reference to (b). Furthermore, recourse to a forensic exploration of the drafting history of prior Lehman securitisations was wholly outwith any legitimate process of interpretation.

Next, Mr Zacaroli QC for DBB submitted that there were insuperable difficulties in reformulating the definition of Principal Amount Outstanding by treating as paid redemption amounts due to be paid only three days later, not least because this would produce an obviously inappropriate measure of double counting when that definition was (as it must be) used for the purpose of the Class A Par Value Test.

Finally, Mr Zacaroli submitted that an Event of Default constituted by failing the Inadequate Par Coverage test on a particular Measurement Date necessarily continued until the next Measurement Date, and could not be treated as having been remedied merely because a calculation not provided for in the Conditions would have shown it to have been passed three days later. Such a flexible approach to the continuation of that type of Default would, he submitted, give rise to a commercially unacceptable degree of uncertainty in the working of the Conditions.

Generally, the court’s task when addressing issues of construction is to ascertain the meaning which the instrument would convey to a reasonable person having all the background knowledge which would reasonably be available to the audience to whom the instrument is addressed. In the present case the instrument in question is the Trust Deed, to which the Conditions are appended as a Schedule. Although devised and initially put in place internally within the Lehman group, its function is to constitute and define the terms of the Notes, and the Class A Notes (in particular) were intended to be used by way of sale or (more likely) security for borrowing, such that the relevant audience for present purposes must be taken to include entities considering buying or lending upon the security of the Class A Notes.

Identification of the relevant audience is important, because it serves to identify the range of background facts relevant to interpretation. Although in principle the “matrix of fact … includes absolutely anything which would have affected the way in which the language of the document would have been understood by a reasonable man”, it is subject to the controlling requirement that it should have been “reasonably available to the parties”, and to the exclusion of an examination of the parties’ previous negotiations: see ICS Limited v. West Bromwich Building Society [1998] 1 WLR 896 at 912-3.

The first of those limitations was further explained by Lord Collins SCJ in Re Sigma Finance Corp [2009] UKSC 2 at paragraph 37, as follows:

“Where a security document secures a number of creditors who have advanced funds over a long period it would be quite wrong to take account of circumstances which are not known to all of them. In this type of case it is the wording of the instrument which is paramount. The instrument must be interpreted as a whole in the light of the commercial intention which may be inferred from the face of the instrument and from the nature of the debtor’s business.”

The concept of commercial absurdity has a long and distinguished history in the interpretation of business contracts, beginning with Lord Diplock’s famous observation in Antaios Compania Naviera v. Salen Rederierna AB [1985] AC 191, at 201:

“If detailed semantic and syntactical analysis of words in a commercial contract is going to lead to a conclusion that flouts business common sense, it must be made to yield to business common sense.”

In this context, a distinction must be made between absurdity and irrationality on the one hand, and apparent unfairness or one-sidedness on the other. The former may compel the court to conclude that something must have gone wrong with the language, but it is no part of the court’s task to mend businessmen’s bargains: see in particular Chartbrook Ltd v Persimmon Homes Ltd [2009] 1AC 1101 per Lord Hoffmann at paragraph 20, and Ing Bank NV v. Ros Roca [2011] EWCA Civ353, per Rix LJ at paragraph 110.

Commercial absurdity may require the court to depart even from the apparently unambiguous natural meaning of a provision in an instrument, because “the law does not require judges to attribute to the parties an intention they plainly could not have had”: see per Lord Hoffmann in the ICS case at page 913. Questions of commercial common sense falling short of absurdity may however enable the court to choose between genuinely alternative meanings of an ambiguous provision. The greater the ambiguity, the more persuasive may be an argument based upon the apparently greater degree of common sense of one version over the other. As Longmore LJ put it in Barclays Bank plc v. HHY Luxembourg SARL [2010] EWCA Civ 1248, at paragraph 25:

“When alternative constructions are available one has to consider which is the more commercially sensible.”

At paragraph 26, he continued:

“The judge said that it did not flout common sense to say the clause provided for a very limited level of release, but that, with respect, is not quite the way to look at the matter. If a clause is capable of two meanings, as on any view this clause is, it is quite possible that neither meaning will flout common sense. In such circumstances, it is much more appropriate to adopt the more, rather than the less, commercial construction.”

I have not been persuaded that the exclusion of the Principal Account Balances from the Par Coverage Numerator for the purposes of the Inadequate Par Coverage test in Condition 10(a)(iv) can itself properly be described as commercially absurd, still less am I persuaded that it is appropriate simply to ignore the whole of the phrase in parenthesis “without regard to clauses (c) of such definition and provisos to such definition” as having been included as the result of an obvious mistake. My reasons follow.

First, something has indeed gone wrong with the language, but only in the sense that the quoted phrase misuses the plural in “clauses” and refers to nonexistent provisos. In the absence of proven commercial absurdity, (and the onus is in this respect squarely upon LB3), to ignore the whole phrase merely because of those two patent infelicities would be to throw out the baby with the bathwater.

Secondly, there was in my view good reason to exclude Principal Account Balances from wholesale incorporation within the Par Coverage Numerator, because of the propensity for the principal waterfall to lead to the payment out of cash from the Principal Account for purposes other than redemption of the Class A Notes, or the acquisition of substitute CDOs. The Principal Account was liable to be resorted to under the principal waterfall for meeting any interest shortfall arising under the interest waterfall, and for meeting ongoing hedging payments which could, in the case of currency or interest rate swaps, be substantial even though in fact no such hedges have, as I understand it, yet been put in place as part of the Excalibur Securitisation. To include the whole of the Principal Account Balances within the Par Coverage Numerator would risk inappropriately swelling the monetary amount of the collateral by reference to sums which were about to be used for a wholly different purpose.

It is in my view not a sufficient answer to say that, in any typical month, redemption payments would be likely to exceed other prior payments (such as expenses, interest and hedging payments) by a very large margin. True it is that this was the case in January 2011, but it is dangerous to construe a structure designed to last for up to fifty years with the dubious benefit of hindsight gained from its operation during a single quarter.

The commerciality or otherwise of the Par Coverage tests in the Excalibur Securitisation must be viewed against the practical reality that they were included as a selling point by the Lehman group with a view to attracting buyers (or more likely lenders) to take Class A Notes by purchase or pledge. The inclusion of amounts which might not be used to swell the value of the collateral or reduce the value of the Class A obligations as part of a Par Coverage test designed to identify whether the collateral was worth less than the obligations in the Class A Notes would, on the face of it, hardly be a selling point. It would have been more realistic to provide for a limited exclusion of such parts of the Principal Account Balances as were not to be applied by way of redemption or the purchase of substitute CDOs, but it is difficult to describe as absurd the use of a broader brush represented by their total exclusion, in particular in the context of a securitisation which, on day one, had built into it a very substantial security cushion which would require serious erosion before the Par Coverage test was failed, even on the construction contended for by DBB.

It is plain that the exclusion of the Principal Account Balances was deliberate, from a comparison between clause 10(a)(iv) and the significantly different Class A Par Value Test which I have described above. The two tests performed different functions within the structure of the securitisation. The Class A Par Value Test was designed to trigger a repair rather than dismantling of the structure, and was set at a significantly higher percentage of collateralisation.

I am even less persuaded by Mr Pascoe’s oral submission that, by reference to the precedent of the Duncannon securitisation, it can be seen that the real mistake made by the drafter of Condition 10(a)(iv) was to refer to clause (c) rather than clause (b). It is an ingenious submission in the context of a case in which, in January 2011 there were, as it happened, no Discount Obligations the exclusion of which could lead to a reduction in the Par Coverage Numerator. But it is based upon a wholly illegitimate, after the event, forensic analysis of the drafting process, none of which could possibly have been known by the audience to whom (or to which) the Trust Deed and scheduled Conditions were addressed. It is a process which might be a valuable tool in a claim for rectification but it has, in my judgment, no role to play in relation to construction.

Even if reference is made to the Duncannon securitisation, the exclusions there applied related both to Discount Obligations and Principal Account Balances. Reference was made at the hearing to a further earlier Lehman securitisation in which Principal Account Balances were also excluded from the Par Coverage Numerator. The general impression created by looking at those securitisations, if admissible at all, is that, far from being a commercially absurd mistake, such exclusions had a distinct pre-history in securitisations constructed by the Lehman group.

I have by contrast found much more difficult the question whether it is commercially absurd to interpret the provisions for the identification of the Principal Amount Outstanding of the Class A Notes (i.e. the denominator under the Par Coverage test) in such a way as to prohibit a deduction of sums in the Principal Account already earmarked on the Measurement Date for payment by way of redemption three business days later. I have already explained how, by the Measurement Date, the provisions for determination of the amounts payable by way of redemption will always have thrown up a specific figure, and how that amount (if any) will always be available in the Principal Account for payment three days thereafter.

The difficulty is that the phrase “less (ii) the aggregate of all Principal Payments in respect of a Note of the relevant Class that have become due and payable and have been paid since the date of issue and of such Note” appears expressly and unambiguously to exclude payments identified as due and payable, where even a short period designed merely to enable the administrative arrangements for payment to be put in place has yet to expire.

There is, in the context of that amount (in January 2011 the €127 million odd rather than the €140 million odd) no risk that the money will be used for any purpose other than redemption. I find it difficult to envisage any rational basis for its exclusion, merely because of the three business day period given to the Issuer within which to make the necessary payment, among a host of other payments falling to be made at the same time.

Neither Mr Zacaroli nor Mr Miles QC for USBT could offer any rational explanation for the leaving out of account of a particular potentially very large redemption payment merely because of the three day period needed to organise its payment. Rather, Mr Zacaroli pointed to the difficulties and double counting which would arise if the court were simply to rewrite the definition of Principal Amount Outstanding, in the context of its use as altered for the purpose of the Class A Par Value Test. That test, it will be recalled, does include the Principal Account Balances as part of the Par Coverage Numerator, so that to treat the redemption payment as both included in the Numerator, and then deduct it from the denominator (the Principal Amount Outstanding), would indeed give rise to double counting and an obviously irrational distortion of the Par Value Ratio. Furthermore, Mr Zacaroli submitted, to address an apparent irrationality in the context of the Inadequate Par Coverage test in Condition 10(a)(iv) by rewriting one of the definitions employed would be to amend, rather than construe, the Conditions.

That is a powerful textual submission, but it does not address the apparent commercial absurdity of the outcome. Where something has gone wrong with the language, it is not in my judgment necessarily an objection to dealing with it in a way that avoids commercial absurdity that provisions have, apparently, to be rewritten, blue pencilled, or amplified so as to work rationally in particular circumstances.

In Chartbrook Limited v. Persimmon Homes Limited [2009] 1 AC 1101, one of the reasons why the majority of the Court of Appeal were not persuaded by the argument based upon irrationality which succeeded in the House of Lords was that it would require “rewriting” of the relevant clause: see per Lord Hoffmann at paragraph 26. But that proved no obstacle to the re-interpretation of that clause, once the argument based on irrationality was found by the House of Lords to be persuasive.

It is nonetheless the case that even a commercial absurdity argument must be confined to its proper role as a tool of interpretation rather than rewriting of an instrument. It is a tool designed to establish what the instrument really means. The questions for me are, first, whether the exclusion of the redemption payments from deduction from the Principal Amount Outstanding, when conducting the clause 10(a)(iv) Inadequate Par Coverage test, is irrational and secondly, if so, whether the relevant provisions, construed as a whole, can be interpreted so as to cure that irrationality.

It is in my view commercially absurd to imagine that the sophisticated framers of this complex and potentially long-lived structure would deliberately expose it to early destruction at the election of the Trustee or the Class A Noteholders merely because, during a particular quarter, a unusually large proportion of the collateral was turned into cash, and therefore left out of account on the Measurement Date in circumstances where, on that very date, it could be seen with certainty that a sufficient part of it would, three days later, be used for redemption in such a way as to restore the Inadequate Par Coverage ratio to a figure in excess of 100%. In commercial terms, the realisation of part of the collateral and the use of the proceeds for the purpose of redemption of Class A Notes is not something which impinges adversely upon the collateralisation ratio at all. It only does so under the present structure (assuming a literal reading of the definition of Principal Amount Outstanding) because the collateral realisations are held in cash for three business days before being distributed to the Class A Noteholders. That strikes me as one of those obviously un-commercial results that flow from a failure of the drafter, when immersed in the detail, to think through the consequences, or to see the wood from the trees: see Re Sigma (supra) per Lord Mance at paragraph 12. Viewed from the generality of the wood, the outcome which I have described is not one which the reasonable audience would think that the structure was, as a whole, meant to achieve. The obvious purpose of the Inadequate Par Coverage Event of Default was to give the Class A Noteholders the option to close down the securitisation by demanding immediate repayment where the value of the collateral had fallen short of the value of the Class A Note obligations. It was not to give them the same option merely because, during a particular quarter, a larger than usual part of the collateral had fallen due for realisation, and where the proceeds had been, or were about to be, used either to redeem the Class A Notes (or, for that matter, reinvested in substitute collateral).

It is reasonably clear from the agreed statistics that this commercially absurd consequence is indeed one that flows from DBB and USBT’s literal construction. Collateral realisations in the quarter ending in October 2010 were €95.4 million odd, and in the quarter ending in April 2011 €88.7 million odd. Neither of those were sufficient, on the literal construction, to cause the Par Coverage ratio to fall below 100%. By contrast, the collateral realisations in the quarter ending in January 2011 were €140.5 million odd, and that was sufficient to tip the ratio below 100%, by 0.5%. There is clearly displayed a direct correlation between the size of the collateral realisations in any quarter and the vulnerability of the structure to an Inadequate Par Coverage Event of Default.

The second question therefore is whether this absurdity can be remedied by construction. At this point it seems to me unhelpful to view Mr Pascoe’s remaining two alternatives completely in isolation from each other. The first, a reading of the definition of Principal Amount Outstanding so that it is interpreted differently for the purpose of the Inadequate Par Coverage test than for the purpose of the Class A Par Value Test is, however imaginative, a tortuous solution that would give very real pause for thought (in the mind of the relevant audience) as to whether it can really have been intended. The alternative solution is that an Event of Default constituted by a breach of the Inadequate Par Coverage test on a Measurement Date, which can be seen on that very day to be going to be remedied three business days later by the amount to be used for redemption of Class A Notes, is merely a fleeting breach which does not continue beyond the Payment Date. That is Mr Pascoe’s final submission.

This separate question, namely whether the Event of Default (if any) which occurred on 25^th January continued after 28^th January, was treated as a wholly self-contained issue by counsel in their submissions. In my view it needs to be addressed as part of the larger analysis of the question whether the absurdity which I have described can be resolved by any process of interpretation. If alternatives are available, then the court’s search should be for that which best accords with the terms of the transaction as a whole, and which does least violence to the plain meaning of its words, especially where they are free from ambiguity.

Mr Zacaroli and Mr Miles together advanced a powerful submission that an Event of Default under Condition 10(a)(iv) occurring on a particular Measurement Date could not sensibly be regarded as continuing for any shorter period than until the next Measurement Date. Otherwise, they submitted, the participants in the securitisation structure would be exposed to an unacceptable degree of uncertainty, there being no machinery whereby the Collateral Administrator was required or entitled to carry out the necessary calculations to ascertain whether the Inadequate Par Coverage ratio had been restored in the meantime, and a risk that, between two quarterly Measurement Dates, all sorts of changes to the relevant figures might occur, in addition to the simple disposal of an identified amount of cash from the Principal Account by way of partial redemption of the Class A Notes.

In circumstances where the question is whether a Par Coverage Default had continued for a significant period, for example well after a Payment Date, that submission appears to me to be unanswerable. But where, as here, the calculations required to be made by the Collateral Administrator on the Measurement Date demonstrate beyond the possibility of any doubt or uncertainty that the Par Coverage shortfall will in fact automatically be remedied three business days later on the Payment Date, I consider that those submissions lose most of their force. Their residual force lies only in the proposition that the phrase in Condition 10(c)(i) “occurs and is continuing” should be given a uniform and rigid single meaning in relation to any kind of default under Condition 10(a)(iv), rather than a more flexible interpretation which accommodates itself to particular facts.

By contrast with the definition of Principal Amount Outstanding, which in my view unambiguously prohibits the deduction of a redemption amount which, even if due, has yet to be paid, the expression “and is continuing” in Condition 10(c)(i) is much more open to ambiguity in its application to an Inadequate Par Coverage Event of Default. It is clear from Condition 10(a)(iv) itself that an Inadequate Par Coverage Event of Default can only occur on a Measurement Date. Thus for example, while it might be that a serious default in a CDO could in fact reduce the value of the Par Coverage Numerator at any time, that would not give rise to an Event of Default until the next Measurement Date. By contrast, Condition 10(a)(iv) says nothing in express terms about the duration of such a Default, and the force of the defendants’ submissions that it ought, generally, to persist until the next Measurement Date has to be derived (if it can be) from a wider appreciation of the workings of the structure as a whole. It is therefore one of those types of ambiguity capable of being resolved in favour of the more commercially realistic alternative.

Returning to the question how, if at all, the absurdity which I have identified is to be resolved as a matter of construction, I have not, after considerable hesitation, been persuaded that it should be resolved by an ambidextrous interpretation of the definition of Principal Amount Outstanding so that, purely in relation to the Inadequate Par Coverage test, it treats as paid redemption money in the Principal Account even though it will only be paid three days later. It is to my mind a strong thing to treat a single definition, plainly designed to be used for two different purposes, as meaning something different for each of those purposes, in the absence of any contextual direction to do so. The reasonable addressee would not think that this is what the definition meant.

By contrast, I have come to the conclusion that the commercial absurdity which I have identified can be ameliorated, although not perhaps entirely removed, by a resolution of the ambiguity in the phrase “and is continuing” in Condition 10(c)(i) in a way which permits a conclusion that an Inadequate Par Coverage Default which occurs on a particular Measurement Date in circumstances where it can be seen on that very day that it is automatically going to be remedied three days later on the Payment Date, does not continue beyond the Payment Date. It is precisely because it can clearly be seen from the calculations which the Collateral Administrator is required to provide on the Measurement Date that the Default is automatically going to be remedied three business days later that the uncertainties inherent in testing that question at some later date, and possibly in relation to additional complicating events, do not arise.

I have described this solution as ameliorating rather than curing the commercial absurdity in question because, of course, it leaves open the possibility that rapid action by the Class A Noteholders and Trustee during the three days between the Measurement Date and the Payment Date could lead to the service of a Notice of Acceleration before the default which occurred on the Measurement Date was speedily cured. If that creates its own mini-absurdity it is, happily, not one which need be addressed on the facts of this case. In practice, it is I think unlikely to occur anyway, for the following reasons. In the absence of a direction from the Class A Noteholders, the Trustee has only a discretion whether to serve a Notice of Acceleration. It would in my view be a surprising exercise of discretion to do so during the fleeting period between the occurrence of an Inadequate Par Coverage Event of Default and its automatic cure three days later. By contrast, the Class A Noteholders would find it difficult in practice to organise themselves (after receipt of the requisite calculations from the Collateral Administrator) in such a way as to direct the Trustee to serve a Notice of Acceleration in time. This may therefore be a theoretical rather than real defect in the solution to the absurdity.

The outcome of this necessarily complex analysis is that in my judgment there was an Event of Default on 25^th January 2011 because, on that date, the Par Coverage Numerator (stripped of the Balances sitting in the Principal Account) was less than 100% of the Principal Amount Outstanding, which had yet to be reduced by payment of the redemption money. Nonetheless that default did not continue beyond the payment of that redemption money on 28^th January 2011 so that there was not a continuing Event of Default on 8^th February 2011 when the purported Notice of Acceleration was served.