LEHMAN BROTHERS HOLDINGS SCOTTISH LP 3 v Case No: A3/2020/1787 [2021] EWCA Civ 1523

IN THE COURT OF APPEAL (CIVIL DIVISION)
ON APPEAL FROM THE HIGH COURT OF JUSTICE
BUSINESS AND PROPERTY COURTS OF ENGLAND AND WALES
COMPANIES COURT (Chancery Division)
[2021] EWCA Civ 1523Case No s: A3/2020/1787, A3/2020/1810, A3/2020/1811Venue [2020] EWHC 1681 (Ch), Royal Courts of Justice, Strand, London, WC2A 2LL
Between
LEHMAN BROTHERS HOLDINGS SCOTTISH LP 3Case No: A3/2020/1787Appellant(1) LEHMAN BROTHERS HOLDINGS PLC (IN ADMINISTRATION) (2) DEUTSCHE BANK A.G. (LONDON BRANCH) (3) THE JOINT ADMINISTRATORS OF LB HOLDINGS INTERMEDIATE 2 LIMITED (IN ADMINISTRATION) (1) THE JOINT LIQUIDATORS OF LB GP No. 1 LIMITED (IN LIQUIDATION) (2) DEUTSCHE BANK A.G. (LONDON BRANCH) - and – (1) THE JOINT ADMINISTRATORS OF LEHMAN BROTHERS HOLDINGS PLC (IN ADMINISTRATION) (2) LEHMAN BROTHERS HOLDINGS INCRespondents Case Nos: A3/2020/1810 and A3/2020/1811Appellants/ Respondents Respondents
Before
Lord Justice LewisonLord Justice HendersonLady Justice AsplinMark Phillips QC, William Willson and Edoardo Lupi (instructed by Weil, Gotshal & Manges (London) LLP ) for the AppellantAdrian Beltrami QC and Adam Kramer QC (instructed by Hogan Lovells International LLP ) for the Respondent (1)Sonia Tolaney QC, Richard Fisher QC and Tim Goldfarb (instructed by Alston & Bird (City) LLP ) for the Respondent (2)Peter Arden QC and Rosanna Foskett (instructed by Dentons UK and Middle East LLP ) for the Respondent (3)Lexa Hilliard QC and Tom Roscoe (instructed by Charles Russell Speechlys LLP ) for the Appellant (1)Sonia Tolaney QC, Mr Richard Fisher QC and Tim Goldfarb (instructed by Alston & Bird (City) LLP ) for the Appellant (2)Mark Phillips QC, William Willson and Edoardo Lupi (instructed by Weil, Gotshal & Manges (London) LLP ) for the Respondent (2)Date 20 October 2021
[1]Introduction The main issues on this appeal concern the order of priority for payment of subordinated debt which came into existence for regulatory purposes. There are four relevant claims which the judge (Marcus Smith J) labelled Claim A, Claim B, Claim C and Claim D. His judgment is at [2020] EWHC 1681 (Ch).[2]The issues arise in the context of the distributing administration of LB Holdings Intermediate 2 Ltd, (LBHI2) a UK company in the Lehman Brothers group. The participants and the claims[3]Both Claims A and B are claims against LBHI2. Claim A arises under two long-term subordinated loan facility agreements and one short-term subordinated facility agreement. Lehman Brothers Holdings plc (PLC) was the lender and LBHI2 was the borrower. Claim B arises out of floating rate subordinated loan notes issued by LBHI2 pursuant to an offering circular. These notes are held by Lehman Brothers Holdings Scottish LP3 (SLP3).[4]Claims C and D, by contrast, are both claims against PLC. They can be satisfied out of what PLC receives under Claim A. Claim C arises under two long-term subordinated loan facility agreements and one short-term subordinated loan facility agreement. Lehman Brothers UK Holdings Ltd was the lender and PLC was the borrower. Because of assignments, the creditor now is Lehman Brothers Holdings Inc (LBHI), which is the ultimate holding company of the group. Claim D arises out of subordinated loan notes issued by PLC pursuant to offering circulars. This claim is now advanced by LB GP No 1 Ltd (GP1) and Deutsche Bank. The flow of funds is illustrated diagrammatically in an appendix to this judgment.[5]It is common ground that although the administrators will have substantial funds to distribute to subordinated creditors (somewhere between £800 million and £1 billion, some of which has already been distributed), it is not enough to satisfy all claims by subordinated creditors. It is therefore necessary to resolve the priority as between them. The waterfall[6]In Re Nortel GmbH [2013] UKSC 52 , [2014] AC 209 at [39] Lord Neuberger said:
“In a liquidation of a company and in an administration (where there is no question of trying to save the company or its business), the effect of insolvency legislation … as interpreted and extended by the courts, is that the order of priority for payment out of the company's assets is, in summary terms, as follows: (1) Fixed charge creditors; (2) Expenses of the insolvency proceedings; (3) Preferential creditors; (4) Floating charge creditors; (5) Unsecured provable debts; (6) Statutory interest; (7) Non-provable liabilities; and (8) Shareholders.”
[7]It is this ordering of priorities that is called the “waterfall”. As Lord Neuberger subsequently made clear, however, in Re Lehman Brothers International (Europe) (In administration) (No 4) (“ Waterfall 1” ) [2017] UKSC 38 , [2018] AC 465 at [17], his description was not intended to be a quasi-statutory statement of immutable legal principle. As can be seen from this description of the waterfall, shareholders come last in the queue.[8]The actual decision in that case was that subordinated debt, which ranked for regulatory purposes as capital, was payable after non-provable liabilities, but before any return to shareholders. Regulatory capital[9]For many companies, the amount of its capital is the nominal amount for which shareholders subscribe. Shares may be ordinary shares or preference shares. The terms on which the preference shares are issued may entitle the holders of preference shares to priority over ordinary shareholders in the event of an insolvency. In the case of banks, however, certain loans may also qualify as capital for regulatory purposes.[10]David Richards J explained the regulatory regime in detail in Waterfall 1 at first instance: [2014] EWHC 704 (Ch) , [2015] Ch 1 . It is not necessary to repeat that description. It is sufficient to note, for the purposes of these appeals, that the regulators permitted three types (or tiers) of capital, which David Richards J described as follows:
“[43] The characteristics of tier 1 capital were that it was able to absorb losses, it was permanent, it ranked for repayment upon winding up, administration or similar procedures after all other debts and liabilities and it had no fixed costs, such as an obligation to pay dividends or interest. The most common example of tier 1 capital is ordinary share capital. [44] Tier 2 capital was capital which did not meet the requirements of permanency and lack of fixed costs which were required for tier 1 capital. There were two types of tier 2 capital. Upper tier 2 capital was capital which was perpetual but which carried servicing costs which could not be waived at the firm's option. It specifically included cumulative preference shares. Lower tier 2 capital was capital which was either not perpetual or had fixed servicing costs that could not generally be waived or deferred. It was required generally to have an original maturity of at least five years and specifically included medium- to long-term subordinated debt. [45] Tier 3 capital was described by GENPRU as forms of capital conforming less well to the characteristics of tier 1 capital. It specifically included subordinated debt of short maturity. [46] Subordination was a characteristic of all three tiers of capital.”
[11]Such subordinated debt could take the form of a loan facility agreement or subordinated loan notes. Priority for payment as between these various forms of regulatory capital would be in reverse order (i.e. Tier 3 before Tier 2; and Tier 2 before Tier 1). Thus the very structure of regulatory capital shows that there can be relative priority as between the various tiers. It follows that the mere fact that something is intended to be regulatory capital does not necessarily tell you anything about its place in the queue. Subordination[12]In principle, debts other than preferential debts rank equally between themselves and, after the preferential debts, must be paid in full unless the assets are insufficient for meeting them, in which case they abate in equal proportions between themselves: Insolvency (England and Wales) Rules 2016 (“IR”) rule 14.12. This is the familiar pari passu principle.[13]The pari passu principle (which ultimately derives from the maxim that “equality is equity”) is an important principle in insolvency law. In Re Golden Key Ltd [2009] EWCA Civ 636 Arden LJ put it this way:
“[5] Pari passu provisions are commonly found in debentures. A provision for pari passu repayment can, however, be implied if it is clear that the debenture holders are to stand on an equal footing (see for example Murray v Scott (1883-4) 9 App Cas 519). Accordingly, where a document on its true interpretation provides for the distribution of assets to a group of persons as between whom no distinction is to be drawn, the court will imply a requirement to make distributions proportionately even though the words “equally” or “ pari passu ” are not used. Such an implication is not, however, possible where the document evinces an intention that the distribution should be on some other basis. [6] Given its importance, the concept of pari passu distribution can be taken to be part of the background to the issue of the [commercial paper] that would have been known to the parties. …The concept of pari passu distribution may also be a factor which makes one interpretation more plausible than another.”
[14]Mr Phillips QC, for SLP3, argued that in order to displace the pari passu principle there must be “clear and unequivocal language to the contrary.” I do not consider that Arden LJ went that far. What she said was that the instrument in question must evince a contrary intention; and in deciding whether it did, the pari passu principle “may be” a factor for preferring one interpretation over another. Moreover, Arden LJ was dealing with the rights of persons holding under the same title (or deriving rights from the same contract) as between themselves, rather than the interaction between different instruments.[15]Mr Phillips also emphasised the centrality of proof in the context of the insolvency regime. IR 14.2 provides that a debt (including a contingent debt) is a provable debt. IR 14.3 provides that a creditor who wishes to recover a debt “must submit a proof”. In the case of a contingent debt, IR 14.14 requires the office-holder to estimate its value. It follows that in order to recover a contingent debt from an insolvent estate, the creditor must submit a proof. I do not consider that that is controversial. The question is when a particular creditor is entitled to prove and whether he is entitled to be paid in priority to or pari passu with another creditor. I agree with Mr Beltrami QC, for PLC, that that is a question of interpretation of the various contractual instruments involved, rather than a question for the rules. The rules are there to give effect to an order of priority that has been contractually agreed as between subordinated debts.[16]The very purpose of a subordination clause is to exclude the pari passu principle as regards the particular creditor in order to postpone or downgrade that creditor’s entitlement to be paid; or, as Lord Neuberger put it in Waterfall 1 in the Supreme Court ( [2017] UKSC 38 , [2018] AC 465 at [64]) to position them:
“ …. at the end of the queue - and, in the event of an insolvency, at the bottom of the waterfall.”
[17]In Waterfall 1 , the Supreme Court confirmed that there was nothing objectionable about that. The various instruments that we are called upon to interpret are independent agreements. Given that a creditor cannot, by agreement with the debtor, advance his position in the queue from where it would otherwise have been, the question that arises in relation to each instrument is how far back in the queue the creditor has agreed to stand.[18]In Waterfall 1 Lord Neuberger was dealing with the order in which creditors of an insolvent company were to be paid. His quoted observation, however, says nothing about the priority, as between themselves, of those creditors at the end of the queue. Shareholders are, of course, even further behind in the queue.[19]Since there is nothing objectionable about a creditor agreeing to occupy a place in the queue which is further back than the place that he would otherwise have occupied, there is equally nothing objectionable in one creditor who might otherwise have shared pari passu with another agreeing that he will occupy a place even further back than that other. Nor, in my judgment, is there anything objectionable in a creditor who is entitled to a provable debt agreeing not to prove until some particular event has occurred. In other words creditors who would otherwise have occupied the same place in the waterfall may agree to subordinate their claims to other such claims. The question is whether the various instruments evince an intention that distribution should be on some basis other than pari passu .[20]The judge found at [61] (3) (c) that the regulators were indifferent to the relative priority between different subordinated debt instruments. He also found that those who put the instruments in place took the view that the relative priority between subordinated debt instruments was, in practice, a question that was not going to arise; because the possible insolvency of the Lehman group was simply not something that was contemplated.[21]The judge identified three methods of achieving subordination by consent. They were: i) Trust subordination; ii) Contingent debt subordination; and iii) Simple contractual subordination.[22]Trust subordination need not concern us. As regards the other two methods the judge said at [111]: “Contingent debt subordination involves the creation of a debt that is payable only in the event of a given contingency being satisfied. By modifying the nature of the obligation owed by the debtor to the creditor, the operation of the rules of legal subordination changes. It is not that the rules themselves are changed, but they apply differently because of the contingent nature of the debt.” (Original emphasis)[23]Having referred to Lord Neuberger’s judgment in Waterfall 1 in which he held that a creditor whose debt had been subordinated only by means of a contingency would nevertheless be entitled to prove for his debt before the contingency had been satisfied, he continued: “In short, contingent debt subordination, by its nature, does not necessarily achieve the kind of total subordination of one obligation or category of obligation below another obligation or category of obligation that may be desired in some cases, notably in the case of regulatory capital.” (Original emphasis)[24]At [115] the judge dealt with what he called simple contractual subordination. That kind of subordination alters the order in which the legal waterfall operates; and only operates in the case of an insolvency process. At [117] he said: “Like contingent debt subordination, simple contractual subordination can only serve to demote the creditor's interests in relation to other creditors. But, whereas this is achieved through the articulation of a contingency in the case of contingent debt subordination, simple contractual subordination simply enables the creditor to select at what point in the waterfall his or her interests will be considered, provided:(1) That point is lower than the priority accorded to the creditor by the legal waterfall; and(2) All the creditor is doing is selecting a point in the legal waterfall, without re-writing that order. In other words, all a creditor is doing is waiving his or her right to be paid at a certain priority and electing to sit at a priority lower than would be the case according to the rules of legal subordination.” (Original emphasis)[25]The main difference between the two methods, in the judge’s view, related to when the creditor was entitled to prove for his debt. The judge perceived a difference of view about the operation of methods of subordination as between this court in Waterfall 1 and the Supreme Court in the same case, both of which considered the same form of sub-debt agreement as is in issue on this appeal. In this court ( [2015] EWCA Civ 485 , [2016] Ch 50 ) I described the subordination provisions at [38]. I said that clause 5 (1) of the sub-debt agreement meant that the right to repayment was contingent on satisfaction of clause 5 (1) (b) and, if appropriate, clause 5 (1) (a) as well. In the Supreme Court Lord Neuberger did not refer to this part of my judgment. The real issue was when the subordinated creditor could lodge a proof. I said that the contingent creditor could lodge a proof; and on that point the Supreme Court at [69] to [70] clearly disagreed. I had based my decision on the terms of clause 5 of the agreement and held that clause 7 of the agreement did not bear on the question, because the contingent debt would be valued at nil. It was on that point that the Supreme Court disagreed. Lord Neuberger’s judgment at [68] and [69] is clearly focussed on the effect of clause 7. He held that the effect of clause 7 was to prevent a right to prove because (contrary to the view I had taken) the lodging of a proof could adversely affect senior creditors if the value of the contingent claim was greater than nil. But I do not see that the disagreement goes any further than that. In their skeleton argument SLP3 said that the judge was right to see a conflict; and PLC that he was wrong to do so.[26]Both SLP3 and PLC agree, however, that in the instruments we have to consider contingent debt subordination was not a free-standing alternative to simple contractual subordination. The subordination provisions in each instrument must be read as a whole. Both were intended to achieve the same objective; namely to identify the creditor’s place in the queue. Given this common ground, I need not enter any further into this debate. Principles of interpretation[27]The general principles of contractual interpretation are well-settled; and are set out in Wood v Capita Insurance Services Ltd [2017] UKSC 24 , [2017] AC 1173 . I will not attempt to distil or paraphrase that learning. As Lord Hodge said at [9], the legal profession has sufficient judicial statements of that nature.[28]There are, however, a few more particular points that I should emphasise. First, the instruments that we are asked to interpret have clearly been drafted by skilled professionals and are at the high end of sophistication. Textual analysis is therefore likely to be the principal method of analysis: Wood at [13]. Second, the loan agreements giving rise to Claim A and C were in a standard form required by the FSA. All the other instruments discussed on this appeal were either standard forms imposed by the FSA; or were derived from such forms or at the very least had an underlying regulatory function. The notes giving rise to Claims B and D were, at least in principle, tradable financial instruments. In either case, background has a very limited part to play; and, once again, the primary tool of interpretation is textual analysis: Re Lehman Brothers (No 8) [2016] EWHC 2417 (Ch) , [2017] All ER Comm 275; BNY Mellon Corporate Trustee Services Ltd v LBG Capital No 1 plc [2016] UKSC 29 , [2016] Bus LR 725 . Third, in the case of an instrument that has been amended, the amendment must be interpreted in the light of the whole instrument as amended: Portsmouth City FC Ltd v Sellar Properties (Portsmouth) Ltd [2004] EWCA Civ 760 ; Stena Line Ltd v MNRPF Trustees Ltd [2011] EWCA Civ 543 , [2011] Pens LR 223 . Fourth, the court may admit expert evidence of market practice in order to explain shorthand terms used in that market; or to enable the court to understand the factual background: Crema v Cenkos Securities PLC [2010] EWCA Civ 1444 , [2011] 1 WLR 2066 .[29]There is one other point that I should make at this stage. In considering the form of instrument in Claim C, the point was raised that it might not have been compliant with FSA requirements. The judge explored that point at [322] to [326]. He came to the conclusion that it might well have been intended to be regulatory capital, but that it was not compliant with FSA rules. He continued at [327]:
“I raise this point so that I can dismiss it for the purposes of this Judgment. Even assuming the form of the PLC Sub-Debt Agreements was deficient, the most that this could have done was rendered what the Lehman Group intended as regulatory capital not regulatory capital. Obviously, that would or might be serious in the regulatory context, but any such deficiency would sound only in the regulatory context and would not otherwise affect the obligations arising under the PLC Sub-Debt Agreements. In short, the rules regarding regulatory capital, and their potential breach, have no bearing on the questions of subordination that I must address.”
[30]In my judgment this puts the point too high. In Digby v General Accident Fire and Life Assurance Corporation Ltd [1943] AC 121 insurers issued a policy of motor insurance. Legislation made it an offence to drive a car without third party insurance. Lord Wright said:
“The policy might on its true construction fail to comply with the statute, but prima facie at least it may be assumed that it is intended to comply with the law.”
[31]Even where there is no criminal offence lurking in the background, a statutory scheme is still a potential aid to interpretation: City of London v Leaseholders of Great Arthur House [2021] EWCA Civ 431 , [2021] L & TR 13. In my judgment, the same goes for regulatory background. Of course, ultimately, the meaning of a contract depends on its own terms (see Office of Telecommunications v Floe Telecom Ltd [2009] EWCA Civ 47 , [2009] Bus LR 1116 ) but I do not think that it can be said that the regulatory background has “no bearing” on questions of interpretation.[32]I will deal first with Claims A and B, the latter in its amended form. There are two aspects to these claims. The first is a question of interpretation of the two sets of subordination provisions. The second, which is contingent on the answer to the first question, is whether SLP3 is entitled to rectification of the subordination provisions in the loan notes (i.e. Claim B). The subordination provisions relating to Claims A and B Claim A[33]The relevant subordination provisions relating to Claim A are in paragraph 5 of section C of the loan agreement:
“(1) Notwithstanding the provisions of paragraph 4, the rights of the Lender in respect of the Subordinated Liabilities are subordinated to the Senior Liabilities and accordingly payment of any amount (whether principal, interest or otherwise) of the Subordinated Liabilities is conditional upon – (a) (if an order has not been made or an effective resolution passed for the Insolvency of the Borrower and, being a partnership, the Borrower has not been dissolved) the Borrower being in compliance with not less than 100% of its Financial Resources Requirement immediately after payment by the Borrower and accordingly no such amount which would otherwise fall due for payment shall be payable except to the extent that – i. paragraph 4(3) has been complied with; and ii. the Borrower could make such payment and still be in compliance with such Financial Resources Requirement; and (b) the Borrower being "solvent" at the time of, and immediately after, the payment by the Borrower and accordingly no such amount which would otherwise fall due for payment shall be payable except to the extent that the Borrower could make such payment and still be "solvent". (2) For the purposes of sub-paragraph (1)(b) above, the Borrower shall be "solvent" if it is able to pay its Liabilities (other than the Subordinated Liabilities) in full disregarding– (a) obligations which are not payable or capable of being established or determined in the Insolvency of the Borrower, and (b) the Excluded Liabilities.”
[34]A number of these terms are defined in paragraph 1 (1) of section C: i) “Lender” is a reference to PLC. ii) “Subordinated Liabilities” is defined to mean “all Liabilities to the Lender in respect of the Loan or each Advance made under this Agreement and all interest payable thereon”. iii) “Liabilities” means “all present and future sums, liabilities and obligations payable or owing by the Borrower (whether actual or contingent, jointly or severally or otherwise howsoever)”. iv) “Borrower” refers to LBHI2. v) “Senior Liabilities” means “all Liabilities except the Subordinated Liabilities and Excluded Liabilities”. vi) “Excluded Liabilities” means “Liabilities which are expressed to be and, in the opinion of the Insolvency Officer of the Borrower, do, rank junior to the Subordinated Liabilities in any Insolvency of the Borrower”.[35]The Insolvency Officer includes the administrators. It is common ground (in relation to all the instruments we have to consider) that the Insolvency Officer will be guided by the court’s conclusions on priority; although the manner in which the Insolvency Officer gives effect to the court’s interpretation will be a matter for them. Claim B[36]The relevant subordination provisions relating to Claim B have been substantially amended since the notes were first issued. The notes as amended are contained in a clean print out which gives no clue about which parts have or have not been amended. In their amended form they are as follows:
“Status and subordination (a) The [LBHI2 Sub-Notes] constitute direct, unsecured and subordinated obligations of the Issuer and the rights and claims of the Noteholders against the Issuer rank pari passu without any preference among themselves. The rights of the Noteholders against the Issuer in respect of the Notes are subordinated in right of payment to the Senior Creditors (as defined below) and accordingly payment of principal and interest (including Arrears of Interest as defined below) in respect of the Notes is (subject as provided below) conditional upon the Issuer being solvent at the time of, and immediately after, such payment, and accordingly no such amount which would otherwise fall due for payment shall be payable except to the extent that the Issuer could make such payment and still be solvent immediately thereafter. The conditionality referred to above shall not apply where an order is made by a competent court, or a resolution passed, for the winding-up or dissolution of the Issuer (except for the purposes of a reconstruction, amalgamation, reorganisation, merger or consolidation on terms previously approved in writing by an Extraordinary Resolution of the Noteholders). If any time an order is made by a competent court, or a resolution passed, for the winding-up or dissolution of the Issuer (except for the purposes of a reconstruction, amalgamation, reorganisation, merger or consolidation on terms previously approved in writing by an Extraordinary Resolution of the Noteholders), there shall be payable by the Issuer in respect of each Note (in lieu of any other payment by the Issuer) such amount, if any, as would have been payable to the Noteholder, if, on the day prior to the commencement of the winding-up and thereafter, such Noteholder were the holder of one of a class of preference shares in the capital of the Issuer having a preferential right to a return of assets in the winding-up of the Issuer over: (i) the holders of all other classes of issued shares in each case for the time being in the capital of the Issuer; and (ii) the Notional Holders, on the assumption that such preference share was entitled to receive, on a return of assets in such winding-up, an amount equal to the principal amount of such Note together with Arrears of Interest (if any) and any accrued interest (other than Arrears of Interest). For the purposes of the above provisions: “Notional Holder” means any creditor of the Issuer whose claims against the Issuer on a winding-up are quantified as though they held a Notional Share. “Notional Share” means any notional and unissued shares in the capital of the Issuer which have a preferential right to a return of assets in the winding- up of the Issuer over the holders of all other classes of issued shares for the time being in the capital of the Issuer but not further or otherwise. The Notes are intended to have a right to a return of assets in the winding-up or dissolution of the Issuer in priority to the rights of the holders of any securities of the Issuer which qualify (or, save where their non-qualification is due only to any applicable limitation on the amount of such capital, would qualify) as Upper Tier 2 Capital or Tier 1 Capital (within the respective meanings given to such terms in the General Prudential Sourcebook published by the Financial Services Authority, as amended, supplemented or replaced from time to time). (b) For the purposes of Condition 3(a) above, the Issuer shall be 'solvent' if (i) it is able to pay its debts as they fall due and (ii) its Assets exceed its Liabilities (each as defined below) (other than its Liabilities to persons who are not Senior Creditors). A report as to the solvency of the Issuer by two directors of the Issuer or, if the Issuer is dissolved or being wound up, its liquidator, shall, in the absence of proven error, be treated and accepted by the Issuer and the Noteholders as correct and sufficient evidence thereof. For the purposes of the above provisions: “Senior Creditors” means creditors of the Issuer (i) who are unsubordinated creditors of the Issuer or (ii) who are subordinated creditors of the Issuer other than those with whose claims the claims of the Noteholders are expressed to rank pari passu and those whose claims rank, or are expressed to rank, pari passu with, or junior to, the claims of the Noteholders; “Assets” means the unconsolidated gross assets of the Issuer and “Liabilities” means the unconsolidated gross liabilities of the Issuer, all as shown by the latest published audited balance sheet of the Issuer, but adjusted for contingencies and for subsequent events, all in such manner as two directors of the Issuer, its auditors or its liquidator (as the case may be) may determine.”
Interpretation of Claims A and B[37]Although the judge started by considering the interpretation of Claim A, followed by Claim B in its unamended form, I consider that the better starting point is to consider Claim B in its amended form. The effect of Claim B in its unamended form is, in my judgment, only relevant to the claim for rectification. As I have said, the notes created by Claim B are tradable financial instruments. In their amended form the Notes do not identify the changes that had been made; and it would be quite wrong for their interpretation to be influenced by a previous iteration which might be unknown to a subsequent holder of the notes: compare Barnardo’s v Buckinghamshire [2018] UKSC 55 , [2019] ICR 495 at [26]. Although Mr Phillips submitted that it was not the (subjective) intention that the Notes would ever be transferred outside the Lehman group, that cannot affect the legal nature of the instrument or the proper approach to its interpretation.[38]Mr Phillips argued that the extent of the subordination was to be found in the first part of the condition which subordinated the debt to Senior Creditors (as defined). The conditions attached to payment could not alter the extent of the subordination. The difficulty with this argument is that it gives no weight to the conditions for payment, which must have been intended to have some legal effect. Although the conditions ought not, if possible, to be interpreted so as to result in an inconsistency with the first part of condition 3, they can (and in my judgment do) shed some light on how the condition is to be interpreted, taken as a whole.

The judge considered that there were two free-standing subordination regimes:

[39]one “simple contractual subordination” and the other “contingent subordination”. It is common ground that he was wrong to chop up the clause in that way; and in my view the parties were correct about that. He interpreted the word “accordingly” following the statement that the debt was subordinated to Senior Creditors as meaning “and also”. In agreement with both parties, I consider that the more natural interpretation of that word is “in consequence” or “therefore”. So I consider that the conditions in 3(a) and 3(b) are spelling out the consequences of the statement that the debt is subordinated to Senior Creditors.[40]Before the judge SLP3 argued that the conditionality in the first part of condition 3 (a) continued to apply. It was only disapplied in the event that a court made a winding up order; and the order made in the present case was an administration order rather than a winding up order. The judge rejected that argument. He held that the relevant distinction was between one set of provisions that applied in the case of a solvent LBHI2 and another that applies in the event of an insolvent LBHI2. There is no appeal against that part of his decision.[41]The first and second paragraphs of condition 3 apply in mutually inconsistent scenarios: outside insolvency and inside insolvency. They use completely different techniques for determining the ranking of the claim. Importantly, it is only what happened in insolvency that was of any concern to the regulators.[42]The immediately relevant part of condition 3 (a), then, is the second paragraph. That provides that if a winding up order (or, as the judge held, an administration order) is made, the Noteholders are entitled to be repaid as if they had been the holders of preference shares having a preferential right to a return of assets in the insolvency over all other shareholders (or Notional Holders). It is not suggested that the Notes were in fact transformed into preference shares. Of course they were not. But the basis of payment introduces a deeming provision which is introduced for the purpose of determining the placing of the Notes in the queue. It can have no other purpose. The hypothesis created by a deeming provision must be carried through to the extent that it fulfils that purpose (for a recent example see Fowler v HMRC [2020] UKSC 22 , [2021] 1 WLR 2227).[43]The return of capital to shareholders (including holders of preference shares) occupies a place in the queue after all other creditors (whether subordinated creditors or not). The effect of this, in my judgment, is that liability under the Notes is postponed to all other liabilities, whether subordinated or not; unless they, too, have been relegated to the place in the queue occupied by shareholders. Mr Phillips pointed to the assumption required by the Notes; namely that the hypothetical holder of a preference share is entitled to an amount equal to the principal amount of the Notes together with arrears of interest. That, he said, was unknown in an insolvent administration or winding up. But in my judgment the effect of the assumption is to do no more than to make it clear that the holders of the Notes are to be paid in priority to real holders of preference shares. The deeming mechanism therefore sets both a ceiling and a floor. The Noteholders are entitled to be paid after all other creditors (including subordinated creditors) but before real preference shareholders. Mr Phillips also pointed to the inclusion of Notional Holders and the statement that the Noteholders under Claim B were to have priority “over” such Notional Holders. I do not consider that that affects the overall interpretation. What it means is that if there are Notional Holders (i.e. other creditors whose claims are quantified as if they held preference shares) then Claim B is to take priority over them. In other words Claim B has not agreed to stand so far back in the queue as to rank pari passu with other creditors whose entitlements are to be quantified as if they were preference shareholders. It does not, in my judgment, give Claim B equal ranking with another claim which is not quantified in that way.[44]As I have said, Mr Phillips argued that the effective subordination provision was found (and found only) in the second sentence of condition 3 (a) which stated that the Notes were subordinated in right of payment to the Senior Creditors. Senior Creditors was a defined expression. They are creditors of the Issuer other than (i) unsubordinated creditors and (ii) subordinated creditors other than those whose claims are expressed to rank pari passu and those whose claims rank pari passu with or junior to the claims of the Noteholders. Thus Claim B expressly contemplates that it may be subordinated, not only to unsubordinated creditors; but also to some subordinated creditors. In other words, Claim B departs from the pari passu principle even as between subordinated creditors. I understood Mr Phillips to agree that there can be relative priorities as between subordinated creditors. So it is necessary to decide where Claim A fits into the definition of Senior Creditors. The claimants under Claim A are creditors of the borrower; but they are clearly not unsubordinated creditors. In order to see whether their claims are expressed to rank pari passu or do rank pari passu with the claims of the Noteholders, it is necessary to see what the Noteholders’ claims are. That can only be deduced from the terms of the Notes themselves. Those terms, in my judgment, state that the Noteholders’ claims are claims to be paid out on the hypothetical basis that they are preference shareholders entitled to a full return of capital and interest. And that basis, as I have said, does no more than confirm that the Noteholders are to be paid in priority to real preference shareholders (and Notional Holders). I do not consider that the terms of the sub-debt agreements which constitute Claim A mean that Claim A is either expressed to rank or does rank pari passu with or junior to the Noteholders’ claim as described in the Notes. It follows, in my judgment, that Claim A does fall within the definition of “Senior Creditors” in the Notes, with the consequence that there is no internal disharmony in the condition.[45]On the other hand the fact that Claim B has agreed to take its place in the queue along with (even though just ahead of) preference shareholders (who have a fixed place in the queue after debt) necessarily means that it has expressed itself to be junior to all forms of debt which is and is treated as such.[46]Mr Phillips also relied on the statement in italics in the amended notes; namely the statement that the Noteholders were intended to have priority over Upper Tier 2 Capital or Tier 1 Capital. It is common ground that that statement is not an operative provision of the Notes, and cannot override its contractual language. On the other hand, a statement of that kind is relevant to the interpretation of the operative conditions. Tier 1 capital, as we have seen is permanent capital (such as share capital). Upper Tier 2 capital may consist of shares but may also be permanent capital but this time with servicing costs. Preference shares would ordinarily be Upper Tier 2 capital. By selecting a place in the queue by deeming Claim B to be that of a preference shareholder, the obvious inference is that the drafter intended Claim B to be the equivalent of Upper Tier 2 capital rather than Lower Tier 2. If there are Notional Holders (as defined) they would, for the purposes of ranking also count as Upper Tier 2 capital by virtue of the deeming provision even though they were in fact Lower Tier 2 capital.[47]Debt arising under Claim A is neither of these; it is Lower Tier 2 capital. It may be that the effect of the operative provisions (as opposed to the expression of intention in the explanatory note) is that Claim B does not have priority over all Lower Tier 2 capital (as opposed to some Lower Tier 2 capital) but that does not, in my judgment, result in a different interpretation of the subordination provisions.[48]For these reasons I consider that Claim B is subordinated to Claim A, which remains a claim to be treated as a creditor’s claim. Rectification[49]SLP3’s claim to rectification proceeds by the following steps: i) In their unamended form the Notes were not subordinated to Claim A; ii) The purpose of the amendment was (and was only) an intention to defer the payment of interest. If the amendment effected anything else, that was contrary to the relevant intention; iii) As interpreted by the judge, the effect of the amendment was to subordinate Claim B to Claim A, which was not the position before the amendments; iv) Therefore the effect of the amendment, to the extent that it did anything other than defer interest, was that it did so as the result of a common mistake; v) The Notes should therefore be rectified to eliminate that mistake.[50]The judge agreed with the first step of this argument (although not in the way that SLP3 put the case); but held that the claim to rectification failed on the facts. He found that there was no relevant common intention.[51]In considering the first step in the argument, the judge held that in its unamended form, Claim B took priority over Claim A. SLP3, however, had argued that the two claims ranked pari passu ; and revives that argument on appeal. PLC, supported by Deutsche Bank, also challenges the judge’s conclusion; and argues that Claim B was always subordinated to Claim A. Thus no party supports the judge’s conclusion.[52]The flaw in the judge’s reasoning can be seen clearly in part of the Table 2 at [185] of his judgment. The relevant part of that table deals with what he called stage 4 (d) in analysing unamended Claim B. What he said was this:
“According to Claim B, SLP3's rights are subordinated to "Senior Creditors". Senior Creditors are relevantly defined as subordinated creditors other than those whose claims rank or are expressed to rank pari passu with or junior to SLP3's rights. In other words, excluded from this definition of Senior Creditors are those whose claims rank or are expressed to rank pari passu with or junior to SLP3's rights. This, in turn, requires reference back to the provisions in Claim A to see how PLC's claims rank and/or are expressed to rank. PLC's claims do not fall within this definition. They are (on their own terms) subordinated and are not expressed to rank either pari passu or junior to SLP3's rights. PLC is not, therefore, a Senior Creditor. SLP3's rights are not subordinated.”
[53]The judge’s logic is the wrong way round. If PLC’s claims do not fall within the carve-out from the definition of Senior Creditors (as I consider, and as the judge held), the consequence is that they are Senior Creditors, not the reverse. The last sentence of the quoted passage is therefore wrong even on its own terms. So I need to consider the question afresh.[54]I deal this time first with Claim A. Condition 5 begins with the general statement that “the rights of the Lender in respect of the Subordinated Liabilities are subordinated to the Senior Liabilities”. It then proceeds to set out two conditions. The first of those conditions applies only if an order for winding up or administration has not been made. Accordingly, it does not apply in this case. The second condition, which does apply, is that payment is conditional upon:
“the Borrower being "solvent" at the time of, and immediately after, the payment by the Borrower and accordingly no such amount which would otherwise fall due for payment shall be payable except to the extent that the Borrower could make such payment and still be "solvent".”
[55]The expression “solvent” is defined. The Borrower is solvent if it is able to pay its Liabilities (other than Subordinated Liabilities) disregarding Excluded Liabilities. Liabilities is widely defined, and would include liability to repay subordinated debts, such as debt incurred under the Notes. Subordinated Liabilities are limited to liabilities under the loan agreement itself. Debt incurred under the Notes does not fall within that definition. So the question turns on the definition of “Excluded Liabilities.” If debt incurred under the Notes falls within that definition, then the Borrower will be solvent even if it cannot redeem the Notes. If, on the other hand, debt incurred under the Notes does not fall within the definition, then Claim A cannot be paid unless the Borrower can redeem the Notes. The relevant part of that definition is:
“Liabilities which are expressed to be and … do rank junior to the Subordinated Liabilities”
[56]It is not possible to decide whether a liability is “expressed to be” junior to debt incurred under Claim A without looking at the terms of the instrument creating that liability. It is not suggested that any special form of words is needed in order to express juniority. Thus we must look at the Notes in their unamended form, interpreted as a whole.[57]Condition 3(a) of the Notes was not affected by the amendment. Under that condition, repayment was conditional on the Borrower being solvent at the time of any payment and immediately thereafter. Condition 3(b) was also not affected by the amendment; and prescribed when the Borrower would be solvent. It (relevantly) provided:
“For the purposes of Condition 3(a) above, the Issuer shall be 'solvent' if (i) it is able to pay its debts as they fall due and (ii) its Assets exceed its Liabilities (each as defined below) (other than its Liabilities to persons who are not Senior Creditors).”
[58]As the judge correctly held, these are cumulative conditions, both of which need to be satisfied. The judge also held, correctly, that solvency is to be judged from time to time as at the date of any payment. But the judge did not go on to consider what the first limb of the definition of solvency actually meant.[59]PLC submit that limb (i) of this definition is a simple “cash flow” condition which precludes payment under the Notes if any debt remains unpaid. There is no exclusion, in this limb, of other subordinated debts. It is a much broader definition than that contained in Claim A. Accordingly, the effect of this condition is that debt due under the unamended Notes is subordinated to the lowest possible level of debts due. Necessarily, that is a lower level than debt due under Claim A, which does not contain the same condition precluding payment. The terms of the Notes do, therefore, contain an express ranking junior to Subordinated Liabilities (as defined in Claim A).[60]Mr Phillips challenges this interpretation. He argues that “debts” in condition 3 (a) is limited to “debts (other than liabilities to persons who are not senior creditors)”. There are a number of reasons for this. First it avoids asymmetry between the cash flow test of solvency in limb (i) on the one hand, and the balance sheet test in limb (ii) on the other. Second, it would have the effect that if a particular debt were expressly subordinated to the Notes, then for so long as the issuer was solvent, the Noteholders could not be paid out before the holders of the junior debt. Third, he says, the solvency condition only requires the issuer to be able to pay debts “as they fall due”. If a debt has not yet fallen due, then the Issuer need not be able to pay it in order to be solvent under the cash flow limb of the test.[61]I do not regard the first of these reasons as a strong point. There are quite clearly two different solvency tests, expressed in different language. That they may produce different results is not entirely surprising. The second point is more formidable. It may be, that in order to avoid absurdity, “debts” would have to be interpreted as excluding debts which are expressly junior to the Notes, but that does not in my judgment justify a further incursion into the ordinary meaning of the word. As Lord Wensleydale put it in Grey v Pearson (1857) HL Cas 61:
“in construing … all written instruments, the grammatical and ordinary sense of the words is to be adhered to, unless that would lead to some absurdity, or some repugnance or inconsistency with the rest of the instrument, in which case the grammatical and ordinary sense of the words may be, modified so as to avoid that absurdity or inconsistency, but no farther.”
[62]So far as the third point is concerned, in my judgment it begs the question which needs to be resolved, rather than answering it.[63]In my judgment the solvency condition in unamended Claim B is an expression of juniority to Claim A. It follows, therefore, that the fundamental premise underlying the rectification claim, namely that the amendment altered the priority as between Claim A and Claim B is wrong. From this it follows that there was no relevant mistake and that, accordingly, the rectification claim does not arise.[64]But even if that is wrong, the rectification claim faces further insuperable hurdles on the facts. First, the claim seeks to excise words which were deliberately inserted in order to cater for a perceived tax difficulty. That entailed removing the solvency condition and substituting the mechanism of a deemed preference share. The mechanism of a deemed preference share was deliberately inserted; and had nothing to do with the deferral of interest. It is clear from the email of 12 June 2008 (from Mr Grant of Allen & Overy to Ms Dolby at Lehman) that the words now sought to be excised were deliberately inserted to deal with “tax sensitivities.” The amendments were designed “to ensure these sensitivities are met.” The fact of the insertion was thus drawn to her attention. The amendments themselves were attached to his email, and highlighted in blue. Although Mr Phillips suggested in oral argument that the perceived tax difficulty was illusory, that was not an argument advanced before the judge; and there is neither a finding nor evidence to support it. So the effect of rectification would be to remove from the Notes deliberate wording which was designed to deal with a particular problem, leaving that particular perceived problem in the air. Thus the second step in the argument (i.e. that the intention was “only” to defer interest) is flatly contrary to the facts.[65]Second, what may have gone wrong is no more than an uncontemplated knock-on effect of the words deliberately inserted. The legal effect of the deemed preference share mechanism, in its own terms, was exactly what was intended. What may have gone unnoticed was how that mechanism would interact commercially with other instruments of which Mr Grant knew nothing.[66]Third, what must be established by convincing evidence is a positive intention (manifested by outward accord) not to change the relative ranking of Claim B. The absence of discussion about a particular change is one strand in the evidence which might lead the fact-finder to conclude that the parties had a positive intention not to make that particular change. Having considered all the evidence, the judge declined to make that finding. That is not a surprising conclusion because: i) The judge had already found that no thought had been given to the relative ranking of subordinated debt, because the possible insolvency of the Lehman group was simply not something that was contemplated. ii) None of the decision-makers were called to give evidence. iii) There was no expression of intention on the part of the Noteholders other than agreement to the change; and therefore nothing that could rank as an outward manifestation of accord.[67]The highest that SLP3 could put the case was that if the knock-on effect had been appreciated, and if it had been drawn to the attention of the decision-makers, there might have been some alteration to the amendments. What alterations there would have been is entirely speculative; and in view of the fact that the wording now sought to be excised was deliberately inserted to cope with a tax problem, it seems inherently unlikely that that wording would have been completely excised, leaving the perceived tax problem unresolved.[68]The judge’s findings of fact are unassailable. Conclusion on Claims A and B[69]Claim A ranks in priority to Claim B and always has done. But even if it did not do so before the amendment, it does now, and the Notes cannot be rectified in the manner claimed. Claims C and D Claim C[70]Claims C and D are both claims against PLC. Claim C arises out of sub-debt agreements made by PLC. The subordination provisions are in condition 5 and read as follows:
“(1) Notwithstanding the provisions of paragraph 4, the rights of the Lender in respect of the Subordinated Liabilities are subordinated to the Senior Liabilities and accordingly payment of any amount (whether principal, interest or otherwise) of the Subordinated Liabilities is conditional upon – (a) (if an order has not been made or an effective resolution passed for the Insolvency of the Borrower and, being a partnership, the Borrower has not been dissolved) the Borrower being in compliance with not less than 120% of its Financial Resources Requirement immediately after payment by the Borrower and accordingly no such amount which would otherwise fall due for payment shall be payable except to the extent that – i. paragraph 4(3) has been complied with; and ii. the Borrower could make such payment and still be in compliance with such Financial Resources Requirement; and (b) the Borrower being "solvent" at the time of, and immediately after, the payment by the Borrower and accordingly no such amount which would otherwise fall due for payment shall be payable except to the extent that the Borrower could make such payment and still be "solvent". (2) For the purposes of sub-paragraph (1)(b) above, the Borrower shall be "solvent" if it is able to pay its Liabilities (other than the Subordinated Liabilities) in full disregarding– (a) obligations which are not payable or capable of being established or determined in the Insolvency of the Borrower, and (b) the Excluded Liabilities.”
[71]A number of these expressions are defined:(1) "Liabilities" is defined as "all present and future sums, liabilities and obligations payable or owing by the Borrower (whether actual or contingent, jointly or severally or otherwise howsoever)".(2) "Senior Liabilities" are "all Liabilities except the Subordinated Liabilities and Excluded Liabilities".(3) "Subordinated Liabilities" are "all Liabilities to the Lender in respect of each Advance made under the Agreement and all interest payable thereon".(4) "Excluded Liabilities" are "Liabilities which are expressed to be and, in the opinion of the Insolvency Officer, do, rank junior to the Subordinated Liabilities in any Insolvency of the Borrower Officer". Claim D[72]Claim D arises out of sub-notes issued by PLC. The relevant subordination provisions read as follows:
“(a) The [PLC Sub-Notes] constitute direct, unsecured and subordinated obligations of the Issuer and the rights and claims of the Noteholders against the Issuer rank pari passu without any preference among themselves. The rights of the Noteholders in respect of the Notes are subordinated to the Senior Liabilities and accordingly payment of any amount (whether principal, interest or otherwise) in respect of the Notes is conditional upon: (i) (if an order has not been made or an effective resolution passed for the Insolvency of the Issuer) the Issuer being in compliance with not less than 100 per cent of its Financial Resources Requirement immediately after such payment, and accordingly no such amount which would otherwise fall due for payment shall be payable except to the extent that (a) Condition 3(d) or Condition 3(g), as the case may be, has been complied with; and (b) the Issuer could make such payment and still be in compliance with such Financial Resources Requirements; and (ii) the Issuer being solvent at the time of, and immediately after, such payment, and accordingly no such amount which would otherwise fall due for payment shall be payable except to the extent that the Issuer could make such payment and still be solvent. (b) For the purposes of Condition 3(a) above, the Issuer shall be "solvent" if it is able to pay its Liabilities (other than the Subordinated Liabilities) in full disregarding (i) obligations which are not payable or capable of being established or determined in the Insolvency of the Issuer, and (ii) the Excluded Liabilities.”
[73]These conditions also contain defined terms:(1) “Senior Liabilities" is defined as "all Liabilities except the Subordinated Liabilities and Excluded Liabilities".(2) "Liabilities" are "all present and future sums, liabilities and obligations payable or owing by the Issuer (whether actual or contingent, jointly or severally or otherwise howsoever)".(3) "Subordinated Liabilities" are "all Liabilities to Noteholders in respect of the Notes and all other Liabilities of the Issuer which rank or are expressed to rank pari passu with the Notes".(4) "Excluded Liabilities" are "Liabilities which are expressed to be and, in the opinion of the Insolvency Officer do, rank junior to the Subordinated Liabilities in any Insolvency of the Issuer".[74]FSA rules required loans intended to rank as regulatory capital to take a particular form unless a departure from that form was sanctioned by the FSA. Claim D did not follow the standard form; and therefore FSA sanction was required. FSA sanction (if given) is published on the FSA website and is therefore reasonably available to any subscriber to the Notes. Although Ms Hilliard QC submitted that what the FSA said was no more than its subjective opinion, I consider that it potentially has greater weight than that. It is part of the regulatory background; and hence is potentially relevant (though not, of course, determinative) of what the instrument means. The relevant part of the FSA waiver modified the application of rule 10-63 of IPRU (NV) to impose requirements. The relevant modified requirement reads:
“(A) the degree of subordination of the loan capital is no less than that provided for by form 10-6;”
[75]Form 10-6 is the standard form that was used for Claim C. The waiver then expressly permitted the definition of “Subordinated Liabilities” now found in Claim D.[76]Although I have said that the regulatory background (and the FSA waiver) are potentially relevant to interpretation, I do not consider that on the facts we have to consider they are of any real moment. We are concerned with relative ranking as between debts which are all subordinated debts. The judge found that the FSA was indifferent to relative ranking as between subordinated creditors; and that its only concern was that subordinated debt should rank after unsubordinated debt. Accordingly, in my judgment, where the waiver refers to the “degree of subordination” being “no less than” form 10-6, what it was concerned with was subordination vis-à-vis unsubordinated creditors; and not relative ranking as between subordinated creditors. Relative priority of Claims C and D[77]The judge concluded, in effect, that there was a logical impasse between Claim C and Claim D. Each claim was, on the proper interpretation of the instrument creating that claim, subordinate to the other. That, in turn, meant that the subordination provisions could not be applied as between the two instruments. The solution was that the two claims ranked pari passu with each other. Ms Hilliard, for LB GPI Ltd (who are the holders of Claim D), supported by Ms Tolaney QC, for Deutsche Bank, challenges that conclusion. Mr Phillips QC (this time for LBHI) supports it.[78]At [342] the judge noted a “significant difference” between the definition of Subordinated Liabilities in Claim C and the same definition in Claim D. He went on to say at [356]:
“However, as I noted in paragraph 342, the definition of the term Subordinated Liability in Claim D is not on all fours with the definition of that term in Claim C. The question is whether this makes a difference. The question is this: Is Claim C a Liability of the Issuer ranking or expressed to rank pari passu with the Notes? Given that Claim C is subordinated debt and is not expressed to rank pari passu with the PLC Sub-Notes, this definitional difference makes no difference to the outcome. Claim C is not a Subordinated Liability as understood and defined by Claim D.”
[79]Ms Hilliard argued that this is where the judge went wrong. The difference in definition made all the difference. Mr Phillips, on the other hand, said that the definitions were “materially” the same. That is the point we need to decide.[80]As I have said, it is not possible for a creditor to advance his position in the queue. All he can do is to agree how far behind what would otherwise have been his position he is prepared to wait. The question, then, is to consider in relation to each instrument how far down the queue the particular creditor agreed to stand.[81]Starting with Claim D, it is necessary to decide how far back in the queue Claim D has agreed to stand. The way to decide it can be summarised as follows: i) Is Claim C a Liability? ii) If, yes, is Claim C a Subordinated Liability? If yes, then it is not a Senior Liability. If no: iii) Is Claim Can Excluded Liability? If yes, then it is not a Senior Liability. If no, then it is a Senior Liability.[82]The claim is subordinated to the Senior Liabilities as defined. They are all liabilities except Subordinated Liabilities and Excluded Liabilities. Claim C is clearly a Liability as defined. If Claim C falls within either of the excepted categories, then it is excluded from the definition of Senior Liabilities and Claim D is not subordinated to Claim C. Taking first the definition of Subordinated Liabilities, they are claims which either rank or are expressed to rank pari passu with Claim D. If, therefore, a claim were to rank pari passu with Claim D, then it falls within the definition of Subordinated Liabilities and hence outside the definition of Senior Liabilities. In that event, Claim D is not subordinated to it. The clear intention behind the exclusion is that Claim D has not agreed to stand further back in the queue than claims which rank pari passu with it. In relation to such claims it will share in a distribution pari passu .[83]Going next to the definition of Excluded Liabilities, the question here is whether Claim C is expressed to be junior to Subordinated Liabilities (i.e. junior to all liabilities which are expressed to rank or which do rank pari passu with Claim D). Claim C expresses itself as subordinated to (i.e. junior to) Subordinated Liabilities and Excluded Liabilities. Since Subordinated Liabilities in Claim C are limited to advances under the same agreement, I need consider that no further. Excluded Liabilities in both Claim C and Claim D are liabilities which are expressed to be and, in the opinion of the Insolvency Officer, do, rank junior to the Subordinated Liabilities. If Claim C expressed to rank (and do rank) pari passu with Claim D (and therefore not junior to Claim D) then it does not fall within the definition of Excluded Liabilities in Claim D. In that event, they are senior creditors. Claim C expresses itself to be junior to all claims except those which are themselves junior to Claim C. If a claim would otherwise rank pari passu with Claim C, Claim C has subordinated itself to that claim. Claim C does, therefore, express itself to be junior to the Subordinated Liabilities as defined in Claim D. It is therefore within Claim D’s definition of Excluded Liabilities. It follows that Claim D has not subordinated itself to Claim C.[84]The same process applies in reverse to Claim C. There are two relevant parts of the subordination provisions. First, the right to repayment is subordinated to the “Senior Liabilities” as defined. They are “all Liabilities except the Subordinated Liabilities and Excluded Liabilities.” The rights under Claim D cannot be Subordinated Liabilities, because they do not arise out of the same loan agreement. If they are not Excluded Liabilities, they must be Senior Liabilities. Second, the solvency condition precludes payment unless the borrower is able to pay “its Liabilities (other than the Subordinated Liabilities) in full disregarding … Excluded Liabilities”. Once again, the question boils down to whether the rights under Claim D fall within the definition of Excluded Liabilities.[85]Excluded Liabilities are defined thus:
“Liabilities which are expressed to be and, in the opinion of the Insolvency Officer, do, rank junior to the Subordinated Liabilities in any Insolvency of the Borrower.”
[86]Unless the rights under Claim D are “expressed to be junior” to the rights under Claim C, they will not be Excluded Liabilities. In that event they must be paid in order to satisfy the solvency condition. As in the case of the contest between Claim A and Claim B, it is necessary to look at Claim D to see what is expressed.[87]In Claim D, the rights under the Notes are subordinated to Senior Liabilities. But Senior Liabilities do not include Subordinated Liabilities or Excluded Liabilities. If, therefore, Claim C falls within the definitions of “Subordinated Liabilities” or “Excluded Liabilities” in Claim D, then Claim D is not subordinated to it. And if Claim D is not subordinated to Claim C, then it does not express itself as being junior to Claim C (which is what Claim C requires). The definition of Subordinated Liabilities in Claim D is:
“all Liabilities to Noteholders in respect of the Notes and all other Liabilities of the Issuer which rank or are expressed to rank pari passu with the Notes”
[88]The clear thrust of this definition is that Claim D is not subordinated to claims which have an equal ranking with Claim D. It must follow that the reasonable reader of that provision would understand that Claim D was not to be subordinated to a claim that had a junior ranking. Whatever else may be said about the clarity of the drafting, it is not possible to regard Claim D as “expressing” itself to be “junior” to Claim C. In addition, the definition of “Excluded Liabilities” in Claim D has a bearing on this question. That definition is:
“Liabilities which are expressed to be and, in the opinion of the Insolvency Officer do, rank junior to the Subordinated Liabilities in any Insolvency of the Issuer”
[89]This directs the reader’s attention back to Claim C. The liabilities in Claim C express themselves to be junior to Subordinated Liabilities (which do not include Claim D). They therefore fall within the definition of Excluded Liabilities in Claim D. Since they fall within that definition, Claim D does not subordinate itself to Claim C.[90]This is, perhaps, a convoluted way of arriving at a conclusion which can be more shortly expressed. There are three possible categories of claim by unsecured creditors: senior claims, pari passu claims and junior claims. Claim D subordinates itself to claims which are senior to it. It does not subordinate itself to claims which rank pari passu with it. It takes its place in the queue alongside other creditors whose claims rank pari passu with it. Claim C on the other hand has agreed to stand even further back in the queue. It has agreed to subordinate itself to claims other than those that are junior to it. In other words, it has agreed to stand in the queue behind creditors whose claims would otherwise rank pari passu with Claim C.[91]Mr Phillips said that there was no commercial reason why Claim D would have been intended to be paid before Claim C. There are, I think, at least two answers to that. First, it is not for the party who relies on the words actually used to establish that those words effect a sensible commercial purpose. It should be assumed, at least as a starting point, that parties understood the purpose which the words they chose effected: City Alliance Ltd v Oxford Forecasting Services Ltd [2001] 1 All ER (Comm) 233. Second, the Notes which give rise to Claim D, unlike the Sub-Debt agreement which give rise to Claim C, were intended to raise money from external investors outside the Lehman group. It would be commercial reason enough to repay external investors before internal ones. Conclusion on Claims C and D[92]It follows, in my judgment, that Claim D must be paid in priority to Claim C. Partial discharge by payment[93]C laim C arises out of the PLC Sub-Debt Agreements, made on 30 July 2004 and 31 October 2005, which were originally between Lehman Brothers UK Holdings Limited (LB Holdings) (as lender) and PLC (as borrower). PLC entered administration on 15 September 2008. PLC's obligations under the Sub-Debt Agreements were guaranteed by LBHI pursuant to a guarantee (“the LBHI Guarantee”). LB Holdings thus had a claim against LBHI pursuant to the LBHI Guarantee, which it pursued, and in respect of which it has (in part) been paid. That part payment is said to be about 36% of the money that was owing to LB Holdings under the PLC Sub-Debt Agreements. That payment was made pursuant to the Settlement Agreement made on 24 October 2011 which the judge dealt with at [274] and following.[94]LBHI is now the assignee of LB Holdings’ claim. It claims to be entitled to prove in PLC’s administration for the whole of the sum originally advanced by the PLC Sub-Debt Agreements, without giving credit for the payment which (in its previous capacity as surety) it made to LBHIs.[95]Under normal circumstances where a surety pays part of the debt guaranteed he is entitled to an indemnity from the principal debtor against that part of the debt. In this case, however, it is common ground that the extraordinarily verbose language of clause 8.02 of the Settlement Agreement released any such claim by LBH1 against PLC.[96]The question, then is whether the payment made under the Settlement Agreement, coupled with the release of LBH1’s right to indemnity from PLC has reduced the amount that LB Holdings was (and LBHI now is) entitled to recover in PLC’s insolvency in its capacity as creditor in respect of the PLC Sub-Debt.[97]Instinctively, one considers that the answer ought to be “yes”. If the surety has paid part of the debt guaranteed, surely the creditor is only entitled to the balance. Chitty on Contracts (34 rd ed) para 45-088 states:
“Where a surety enters the contract at the request of the principal debtor, it is clear that payment of the debt by the surety discharges that debt as between the creditor and principal debtor.”
[98]It is, I think, common ground that if the surety discharges “the debt” (i.e. the whole debt) then that proposition is correct. But what if the surety only pays an amount equal to part of the debt? Why should there be any difference in principle? Chitty goes on to state at para 45-089:
“Clearly, payment by a surety of amounts owed under the guarantee discharges the surety either wholly or pro tanto.”
[99]Mr Phillips submits that this statement is wrong.[100]In MS Fashions Ltd v BCCI the simplified facts were these. Two company directors each signed as a “principal debtor” an agreement with the bank under which, as guarantee for repayment of loans by the bank to his company, the bank could withdraw money from his deposit account with that bank towards satisfaction of his company's debts. Before it was wound up BCCI made demands against one of the companies for repayment of over £1 million. The director of that company had a deposit with BCCI of some £426,000 at the time of the making of the winding-up order. That sum was also charged with repayment of the indebtedness to BCCI of the company. In the other case the company owed about £3.3 million and the director’s deposit account had a credit balance of about £4.5 million. The case came to this court twice, following BCCI’s insolvency. The first time ([1992] BCC 571) was an interlocutory appeal against a refusal of leave to begin proceedings against BCCI for a declaration that the amount of the debt which the company owed BCCI had been reduced by insolvency set-off. Scott LJ said at 575:
“It is plain enough that payment by the surety, whether in whole or in part, as the case may be, not only releases the surety but also discharges or reduces, as the case may be, the liability to the creditor of the principal debtor.”
[101]He went on to hold that insolvency set-off had the same effect. Woolf LJ said at 577:
“The rule operates as a matter of law upon the company going into liquidation. At that stage an account is required to be taken of what is due from each party to the other and there is then to be a set off between them. Once there has been that set off, to the extent of the amount which is set off, the company has been paid. That means that not only is the guarantor or joint principal discharged to the extent of the set off, but so is any other debtor who is liable in relation to the same sum.”
[102]The second round of the proceedings came before Hoffmann LJ at first instance and then to this court: [1993] Ch 425. Hoffmann LJ posed the question at 430 thus:
“A bank advances money to a company. Repayment is guaranteed by a director who has a deposit account with the bank. As between himself and the bank, the director is expressed to be a principal debtor. On the insolvency of the bank, can the director set off his claim for return of his deposit against his liability to pay the company's debt, so that the debt is wholly or pro tanto extinguished? Or can the bank claim the whole debt from the company and leave the director to prove in the liquidation for his deposit?”
[103]Dealing with the “principal debtor” clause Hoffmann LJ said at 436:
“In my judgment the “principal debtor” clauses have the effect of creating primary liability for the purposes of the rule that the debt is not contingent upon demand. … It is true that for some purposes the courts will look to the underlying reality of the suretyship relationship rather than the formal agreement that liability is to be as principal debtor. But this is only for the purpose of protecting the surety's equitable rights against the principal debtor and giving effect to such consequences as may affect the creditor, such as the surety's right to take over securities and the rule against double proof. Otherwise there is no reason why creditor and surety should not make whatever terms they choose.”
[104]One of the arguments was that set-off by the surety would “pro tanto extinguish the liability of the [principal debtor].” BCCI argued that although set-off might operate between BCCI and the director, it did not amount to payment of the debt by the principal debtor. Hoffmann LJ rejected that argument. He said at 439:
“This, I think ignores the fact that the director's set-off operates in respect of the same debt as that owed by the company. If, as I think it must be, the set-off is equivalent to payment by the director (see Ex parte Barnett; In re Deveze (1874) LR 9 Ch App 293) then I think it must operate also to extinguish to the same extent the debt owed by the company. I will therefore declare that the indebtedness of each of the companies as at the date of the winding up has been extinguished or reduced by the amount which on that date was standing to the credit of the directors on their respective deposit accounts.”
[105]An appeal against that order was dismissed. In this court Dillon LJ (with whom Nolan and Steyn LJJ agreed) also took the view that the “principal debtor” clause dispensed with any need for a demand. He went on to say:
“If there is set-off between Mr. Amir and Mr. Ahmed and B.C.C.I. that must automatically reduce or extinguish the indebtedness to B.C.C.I. of the companies. … It operates to reduce or extinguish the liability of the guarantor and necessarily therefore operates as in effect a payment by him to be set against the liability of the principal debtor. A creditor cannot sue the principal debtor for an amount of the debt which the creditor has already received from a guarantor.”
[106]So Hoffmann LJ’s order stood. At both stages of MS Fashions it was only part of the overall debt to BCCI that was in issue.[107]Mr Phillips explained that case as turning on the existence of the “principal debtor” clause. The existence of that clause did not feature at all in the interlocutory appeal; and the significance that both Hoffmann LJ and Dillon LJ attributed to it did not, as I read the judgments, go further than to dispense with the need for a demand. None of the textbooks that we were shown explain MS Fashions on the ground that Mr Phillips advanced. In Lehman Brothers Commodity Services Inc v Crédit Agricole Corporate Investment Bank [2011] EWHC 1390 (Comm) , [2012] 1 All ER (Comm) 254 Field J cited MS Fashions at [29] as authority for the entirely general proposition that:
“In the absence of an express agreement to the contrary, where an obligee is owed money by A and can look to B in respect of the same debt, a set-off by B reduces pro tanto the debt owed by A.”
[108]Andrews and Millett on The Law of Guarantees (7 th ed) discuss “principal debtor” clauses at para 1-015 (citing MS Fashions ) and para 7-006; and do not suggest that it makes any difference to the question whether a partial payment by a surety discharges a debt pro tanto . At para 6-002 they discuss McGuinness v Norwich and Peterborough BS [2011] EWCA Civ 1286 , [2012] 2 BCLC 233 . That was a case of a guarantee which also contained a “principal debtor” clause. At first instance Briggs J held that the effect of that clause was that the surety made the principal debtor’s debt his own. Andrews and Millett comment:
“Again with respect, that takes the object and purpose of the standard principal debtor clause too far. Its purpose is not to make the surety a principal debtor so that his liability is joint and several with the principal debtor and not truly accessory. Its purpose, as explained by Dillon LJ in MS Fashions v BCCI … (in a passage both cited and applied by Briggs J) was simply to dispense with the need for a demand on the surety.”
[109]I do not consider that the “principal debtor” clause will bear the weight that Mr Phillips attributed to it.[110]Milverton Group Ltd v Warner World Ltd [1995] 2 EGLR 28 concerned a guarantee of rent payable by an assignee of a lease. The simplified facts were as follows. L1 granted a lease to T1. T1 assigned to T2. The licence to assign contained covenants by S1 that T2 would pay the rent and perform the covenants in the lease; and that S1 would make good any losses. T2 assigned to T3. The licence to assign contained surety covenants by S2 and S3 in the same terms as that given by S1. The reversion was assigned to L2. When T3 became insolvent, L2 demanded the Michaelmas rent of £19,500 from T1 and S2 and S3. S2 and S3 paid £50,000 in consideration of a release from liability. L2 then began proceedings against T1 for the Michaelmas rent and subsequent rent. S1 paid L2 £10,000 in consideration of a release from liability. This court held that all the sums paid by S1, S2 and S3 could be set off against the rent and therefore reduced the amount to which L2 was entitled. There was a clause equivalent to a principal debtor clause; but it played no part in the court’s reasoning.[111]Glidewell LJ (with whom Kennedy LJ agreed) said at 30B:
“I would express the proposition which produces this result as follows. If a lessor is entitled to be paid a sum by way of rent for a particular period, and the original lessee, an assignee and a surety have all covenanted to pay that rent, the lessor may recover it from any one of them (in the case of the surety, if the assignee has defaulted). If the lessor does however recover that sum from any one of the three, the rent has then been paid. The other two persons who were liable cease to be liable to pay that rent though of course they are still liable for any future rent under their respective covenants.”
[112]Hoffmann LJ said at 31F:
“For the purpose of deciding whether money owed by more than one person has been paid, I do not think that it is possible for the creditor and one of the debtors to characterise a payment in return for a release as anything other than a part performance of the obligation. If this were possible, a creditor could pick off his debtors one by one and recover in total more than the whole debt. For the payment to count as part discharge of the common obligation, it is sufficient for the payment to be referable to the guarantee.”
[113]As Megarry J put it in Re Hawkins [1972] Ch 714 (approved by Hoffmann LJ in Milverton ):
“Rent is rent, a fine is a fine, a debt is a debt, and interest is interest, whoever pays it.”
[114]This seems to be intuitively correct; and both these cases are cited without criticism in support of the proposition in Chitty. But in any event, I consider that we are bound by that decision and by the decision of this court in MS Fashions . The declaration made by Hoffmann LJ and upheld by this court was that payment of part of the debt by the sureties discharged the principal debtor pro tanto . That must have been part of the ratio of the decision.[115]Nor, I think, is MS Fashions an outlier. In the first place, Milverton reached the same result. The proposition formulated by Glidewell LJ was expressed in very general terms. Hoffmann LJ’s discussion was predicated on the payment made by the sureties as having been “performance in part”. I do not consider that Milverton can be airbrushed out of existence either on the ground that there was a single set of obligations or that there was complete performance. Milverton was applied by the Singapore Court of Appeal in Royal Bank of Scotland NV v TT International Ltd [2012] SGCA 9 (referred to in Goode & Gullifer on Legal Problems of Credit and Security (6 th ed) para 8-18 footnote 127) where a security deposit provided by a surety went in reduction of claims made by a landlord in the context of voting in a scheme of arrangement. Although that court appears to have considered that there was some special rule in the context of landlord and tenant, I cannot see why that should be so.[116]Mr Phillips submitted that a surety could not be subrogated to the principal creditor’s rights against the principal debtor unless the debt remained in being. It therefore could not have been discharged either wholly or partly by a payment made by the surety. I disagree. In Banque Financière de la Cité v Parc (Battersea) Ltd [1999] 1 AC 221 Lord Hoffmann explained at 236: “In a case in which the whole of the secured debt is repaid, the charge is not kept alive at all. It is discharged and ceases to exist. In a case like the present, in which part of the secured debt is repaid, the charge remains alive only to secure the remainder of the debt for the benefit of the original chargee. Nothing can affect his rights and there is no question of competition between him and the party claiming subrogation. … When judges say that the charge is "kept alive" for the benefit of the plaintiff, what they mean is that his legal relations with a defendant who would otherwise be unjustly enriched are regulated as if the benefit of the charge had been assigned to him. It does not by any means follow that the plaintiff must for all purposes be treated as an actual assignee of the benefit of the charge and, in particular, that he would be so treated in relation to someone who would not be unjustly enriched.” (Original emphasis)[117]He went on to say that the same principle applied where only part of the debt had been paid. But in any event it is not necessarily the case that the surety’s right of indemnity from the principal debtor is limited to subrogation. It may be no more than a remedy to preclude unjust enrichment of the principal debtor at the surety’s expense. If so, there is no reason why the surety should have paid the whole of the debt before exercising his remedy. Moreover, unless the principal debtor’s liability has been partially discharged, there could not be any enrichment of him, unjust or otherwise.[118]In Davies v Humphreys (1840) 6 M & W 153 Parke B said at 167:
“… it is clear that each sum the plaintiff, the surety, paid, was paid in ease of the principal, and ought to have been paid in the first instance by him, and that the plaintiff had a right of action against him the instant he paid it, for so much money paid to his use. However convenient it might be to limit the number of actions in respect of one suretyship, there is no rule of law which requires the surety to pay the whole debt before he can call for reimbursement.”
[119]In Commercial Bank of Australia Ltd v Wilson [1893] AC 181 the principal debtor owed the bank £6,250. Sureties guaranteed the liability; but subsequently agreed with the bank that their liability should be limited to £3,000 which they deposited in a suspense account at the bank. The bank had power to apply that sum in partial discharge of the debt. When the principal debtor became bankrupt, the question was whether the bank was entitled to prove for the whole of the debt. The Privy Council clearly took the view that if the bank had had recourse to the £3,000 the debt would have been partially discharged; with the consequence that the bank would only have been entitled to prove for the balance. Lord Herschell LC said at 185:
“The bank no doubt had power when it thought it prudent to do so to appropriate that sum to the payment of the principal debt pro tanto, and as soon as they made such appropriation it would undoubtedly operate as payment . They never have made such appropriation. The question is whether prior to appropriation it operated as payment of the debt. Their Lordships are unable to see why it should do so. The money was put to a special account called a suspense account, presumably in the names of these guarantors who had paid it in. At all events it was ear-marked as a special account. Down to the time of appropriation by the bank of this amount their Lordships are unable to see anything which could discharge the principal debtor.” (Emphasis added)
[120]Second, it is consistent with the principle that where A is compelled to pay (and does pay) a sum for which B is primarily liable, he is entitled to recoupment from B. The basis of the claim is, in essence, unjust enrichment; but unless B’s debt is discharged there is no enrichment (unjust or otherwise). Even where the question of discharge is not explicitly discussed in the cases, it is the fundamental premise upon which recoupment rests. In Ibrahim v Barclays Bank plc [2012] EWCA Civ 640 , [2013] Ch 400 , in a judgment with which Rimer and McFarlane LJJ agreed, I discussed a number of such cases at [40] to [48] leading to the conclusion at [49] that:
“These authorities in my judgment justify the first two of Mr Goodall's propositions, namely: (i) payment by a third party to a creditor under legal compulsion on account of a debt owed by a debtor will automatically discharge the debtor's debt; (ii) that is the case even if the legal compulsion arises out of a contractual obligation voluntarily assumed by the third party.”
[121]Among the cases I discussed was Brook's Wharf and Bull Wharf Ltd v Goodman Bros [1937] 1 KB 534. In that case Goodman Brothers deposited imported furs in Brook's bonded warehouse. A number of them were stolen, without negligence on the part of the warehousemen. As bonded warehousemen Brook's were compelled to pay customs duties on the imported furs. They claimed recoupment from Goodman Brothers. The claim succeeded. On the question of discharge Lord Wright MR said, at 546:
“ The payment relieved the importer of his obligation . The plaintiffs were no doubt liable to pay the Customs, but, as between themselves and the defendants, the primary liability rested on the defendants. The liability of the plaintiffs as warehousemen was analogous to that of a surety . It was imposed in order to facilitate the collection of duties in a case like the present, where there might always be a question as to who stood in the position of importer. The defendants as actual importers have obtained the benefit of the payment made by the plaintiffs and they are thus discharged from the duties which otherwise would have been payable by them.” (Emphasis added)
[122]It is clear from this statement that his Lordship considered that payment by a surety discharged the liability of the principal debtor.[123]In Carter v Carter (1829) 5 Bing 406 a sub-tenant entered into a lease at a rent of £50 per annum. Under pressure from the head landlord he paid an amount on account of the ground rent and land tax. The intermediate landlord then distrained for the full amount of £25, representing 6 months’ rent under the sub-lease, which the sub-tenant had not paid. The sub-tenant succeeded in an action for excessive distress on the ground that, taking into account the payments he had made to the head landlord (which the intermediate landlord ought to have made), only £5 10s was due. Having referred to earlier cases, Best CJ said at 409:
“The payment of ground-rent by the occupier for the landlord, was holden not to constitute a cross demand, but to amount to payment of so much of the occupier's rent. Here, by the same means, all the Plaintiff's rent had been paid but 5l. 10s., notwithstanding which the Defendant distrains for 25l.; he is, therefore, clearly liable on the count which states the excessive distress in that way.”
[124]One significance of this case is that it was only a partial payment, yet it went in abatement of the landlord’s right to distrain. Thus far the cases seem to me to show (as I would intuitively expect) that part payment by a surety of a debt that he has guaranteed operates as a partial discharge of the underlying debt itself.[125]In Ulster Bank Ltd v Lambe [1966] NI 161 the principal debtor owed the bank £965. Two sureties guaranteed the indebtedness. Together they paid the bank £762 which the bank posted to a suspense account. The bank then sued the principal debtor for the full amount of £965. As far as I can see from the report, the principal debtor was solvent. The issue was whether the bank’s claim should have been limited to the balance. Lowry J held that the bank was entitled to judgment for the full amount. He said at 169:
“The true principle is that where the entire debt is guaranteed, with or without a limit, the creditor can sue the principal debtor, or claim in his bankruptcy, for the full amount of the debt, despite any payments on foot of a guarantee, whether they are made before or after the principal debtor’s bankruptcy, provided those payments in aggregate fall short of the amount of the debt. The benefit to the guarantor is that money received in excess of the full amount of the debt is held in trust for him.”
[126]Since the principal debtor in that case was not insolvent, the proposition as stated by Lowry J appears to be of general application. It is not, however, entirely clear what Lowry J meant by payments “on foot”. It may be that he attached significance to the fact that the payments made by the sureties were posted to a suspense account, rather than appropriated to and applied in reducing the principal debt. That would, at least, have been consistent with the decision in Commercial Bank of Australia . Nor, of course, did Lowry J have the benefit of the decisions of this court in MS Fashions or Milverton . What he appears to have done is to apply the rule against double proof in an insolvency to the rights and liabilities of a creditor, principal debtor and surety outside insolvency.[127]Mr Phillips points out (in common with Goode & Gullifer para 18-18) that neither Ulster Bank nor Sass (which I deal with later) was cited to this court in MS Fashions . Ulster Bank was, however, cited at first instance in MS Fashions together with Re Sass (although not referred to by Hoffmann LJ in his judgment). He must have thought that they were of no relevance to the issue he had to decide.[128]It is, of course, open to commentators to say that a decision of this court is wrong. But the rules of precedent mean that that is not a course available to us unless the case can be said to have been decided per incuriam (i.e. relevant binding authority was not cited) or there are conflicting decisions of this court, in which if it is the latter event we are free to choose between them. Since neither Ulster Bank nor Sass would have been binding on the court in MS Fashions , it is not possible to say that the decision was per incuriam . Nor have we been shown any conflicting binding authority. Moreover, Goode & Gulliver are dealing with the position in the context of insolvency, rather than the general law.[129]In The Public Trustee of Queensland v Octaviar Ltd [2009] QSC 202 McMurdo J considered the position at common law. It is important to note that the rule against double proof did not arise in that case; and to the extent that His Honour referred to cases dealing with that rule, it was only to distinguish them. At [75] he said:
“[The] proposition, which is that the indebtedness of the principal debtor is unaffected by payments by the surety, has some support in the authorities. But there is authority directly against it, and in my view it is wrong. I go first to the cases relied upon to support the proposition.”
[130]At [77] he referred to Ulster Bank and said:
“The references to the position in the event of the bankruptcy of the principal debtor show that Lowry J reasoned by analogy with the creditor’s rights in that context. As the following cases demonstrate, it is well established that in the event of the bankruptcy or winding up of the principal debtor, the creditor is entitled to prove for the whole of the debt without giving credit for payments by the guarantor, as long as the whole debt has not been discharged. That is a rule in relation to proving in insolvency, which is designed to preclude double proof in respect of the same debt.”
[131]At [78] he explained Sass in the same way. At [89] he distinguished between the partial discharge of a debt on the one hand, and the rule against double proof on the other. He quoted the observations of Powell J in McColl’s Wholesale Pty Ltd v State Bank of New South Wales [1984] 3 NSWLR 365:
“Prima facie, the surety’s right to an indemnity is converted into a right to prove in the winding up. However, because of the rule against double proof in the winding up of insolvent companies … the surety cannot prove in the winding up in respect of any amount which he has paid pursuant to his guarantee unless the creditor’s debt has been paid in full, or, in the case of a guarantee of part of the debt - as opposed to a limited guarantee… - the surety has paid to the creditor the full amount for which he is liable….”
[132]McMurdo J observed:
“That passage illustrates the point already discussed, which is that the cases concerned with the rule against double proof are not authority for the proposition for which Fortress and the administrators contend.”
[133]As His Honour correctly pointed out, a surety’s right to an indemnity arises precisely because he has discharged the liability of the principal debtor either in whole or in part. As he said at [90]:
“… it is not the case that the right to subrogation would be lost if the guarantor’s payment discharged the debt which had been owed by the principal. Rather, it is the benefit thereby derived by the debtor from the guarantor’s payment which entitles the guarantor to be indemnified by the debtor in the amount of any payment and to be subrogated to the creditor’s securities if the secured debt is paid in full.”
[134]Whether the position inside insolvency is the same is what I turn to next. The rule against double proof[135]A surety’s claim against the principal debtor under a pre-insolvency guarantee, once he has been called upon to pay, plainly falls within the wide definition of “debt” in IR 14.1; and on the face of it within the category of provable debts in IR 14.2. But as Lord Neuberger pointed out in Waterfall 1 at [13] the 1986 legislation is not a complete code. It sits alongside judge-made rules, one of which is the rule against double proof. Thus as he explained at [13]:
“And, as judge-made rules are ultimately part of the common law, there is no reason in principle why they cannot be developed, or indeed why new rules cannot be formulated. However, particularly in the light of the full and detailed nature of the current insolvency legislation and the need for certainty, any judge should think long and hard before extending or adapting an existing rule, and, even more, before formulating a new rule.”
[136]At its simplest, the rule against double proof is that an insolvent estate should not pay two (or more) dividends in respect of the same debt. This is the way that the rule is described in Rowlatt on Principal and Surety (6 th ed) at 11-01:
“The rule against double proof for what is in effect the same debt is in substance a rule against the receipt of two dividends.”
[137]As we have seen, a payment made by the surety is a payment of the same debt as that owed by the principal debtor (“A debt is a debt, whoever pays it”). The rule against double proof arises most commonly where a surety has paid part of a debt owed by the principal debtor, but the creditor has not recovered in full. In Re Kaupthing Singer & Friedlander Ltd (in administration) (No 2) [2011] UKSC 48 , [2012] 1 AC 804 Lord Walker summarised the rule and its operation at [11]:
“The rule prevents a double proof of what is in substance the same debt being made against the same estate, leading to the payment of a double dividend out of one estate. It is for that reason sometimes called the rule against double dividend. In the simplest case of suretyship (where the surety has neither given nor been provided with security, and has an unlimited liability) there is a triangle of rights and liabilities between the principal debtor (PD), the surety (S) and the creditor (C). PD has the primary obligation to C and a secondary obligation to indemnify S if and so far as S discharges PD's liability, but if PD is insolvent S may not enforce that right in competition with C. S has an obligation to C to answer for PD's liability, and the secondary right of obtaining an indemnity from PD. C can (after due notice) proceed against either or both of PD and S. If both PD and S are in insolvent liquidation, C can prove against each for 100p in the pound but may not recover more than 100p in the pound in all.”
[138]He added at [12]:
“The effect of the rule is that so long as C has not been paid in full, S may not compete with C either directly by proving against PD for an indemnity, or indirectly by setting off his right to an indemnity against any separate debt owed by S to PD.”
[139]The rule thus has two main facets: i) It permits the creditor to prove for the whole of the original debt without giving credit for any part payment received from the surety; and ii) It precludes the surety from proving unless and until the creditor has recovered 100 p in the pound (either inside or outside the insolvency).[140]The principal question for us is whether these are rigid rules, or whether they may be qualified and, if so, in what circumstances.[141]Deutsche Bank argue that the purpose of the rule is to prevent the surety from competing with the creditor in the insolvency. Once the surety’s right to indemnity against the principal debtor has been released, there is no question of competition since the surety has given up his right to an indemnity from the principal debtor. There is nothing left for which he can prove. There can, therefore, be no question of double proof. Claim C is not a claim made by LBHI in its capacity as surety seeking an indemnity from the principal debtor. It is a claim made by LBHI in a completely different capacity; namely as assignee of LB Holdings . LBHI, on the other hand, argue that the rule is an independent rule, and has nothing to do with double proof. None of the cases which have thus far considered the rule against double proof have had to consider this unusual situation.[142]Mr Phillips took us through some of the early cases, from which it seems that what was at first a conclusion reached on the interpretation of the particular form of guarantee has hardened into a rule. In Midland Banking Co v Chambers (1869) LR 4 Ch App 398 in consideration of the bank advancing £300 to the principal debtor, Mr Thorpe guaranteed that overdraft at the bank up to the extent of £300. The guarantee was a continuing guarantee, which was not to be wholly or partially discharged by payment of any sums for the time being due on the balance of the account. It also provided that all “dividends, compositions and payments” were to be taken in gross and applied to any balance due to the bank. The customer then compounded with his creditors. There was a balance of £410 4s. 11d then due to the bank. Mr Thorpe then paid the bank the £300 which he had guaranteed, which he had in fact recovered under a counter-security that he had taken from the principal debtor. The trustee in bankruptcy argued that the creditor should only be entitled to prove for the outstanding balance, namely £110 4s 11d; but this court rejected that argument. Selwyn LJ said that the question was whether Mr Thorpe had waived his right to take the place of the creditor to the extent of what he had paid. He said at 400:
“The surety may, however, in his contract of suretyship agree to waive this right for the benefit of the creditor, and the question is, whether the surety did so in the present case. I am of opinion that the clause in the latter part of the guarantee was intended to exclude the surety from the right to have a share in the benefit of the proof, and to allow the creditor to receive the full amount of the dividend.”
[143]He went on to hold that the creditor’s right was unaffected by Mr Thorpe’s recovery under the counter-security. The reason for that was that the bankrupt estate could not make any claim against the creditor that the surety could not have made; and since the surety had given up his claim against the creditor, the estate could not advance it. Giffard LJ said at 402:
“The principle is this—if the surety had paid the £300 out of his own money he would have had the benefit of proof for that amount, and that benefit he has relinquished in favour of the Plaintiffs. The argument that it has been paid out of the debtor's estate is a fallacy; it was paid out of something which, having before the execution of the creditors' deed been dedicated to the purpose of indemnifying the surety, was not, at the time of the execution of that deed, part of the debtor's estate. Such a payment stands on the same footing as if it had been made by Thorpe out of his own moneys, and furnishes no ground for reducing the proof.”
[144]In Ex p National Provincial Bank of England In re Rees (1881) 18 Ch D 98 Mr Rees was a customer of the bank. He and Mr Powell, as surety for him, executed a joint and several bond for £1,000 in favour of the bank to secure the balance on his account. The bond limited the surety’s liability to £500; but also provided that it would be a continuing security “notwithstanding any settlement of account or other matter or thing whatsoever”. It also contained a proviso stating that if Mr Rees became bankrupt any dividend received by the bank should not, so far as concerned the surety, be taken as discharge of any of the principal monies to the extent of £500. Mr Rees was adjudicated bankrupt, owing the bank some £2,494-odd. Mr Powell subsequently paid the bank £500 plus some interest; and the bond was delivered up to him. The question was whether the bank was entitled to a dividend on the £500 that Mr Powell had paid. James LJ said that it was important that the condition of the bond was that it required payment of the whole £1,000 even though Mr Powell’s liability was limited to £500. He went on to say at 102-3:
“So in the present case the surety has chosen to contract himself out of that possible equity in the plainest and most distinct terms, and, as Lord Justice Knight Bruce said, he can have no right to intercept any dividend which would otherwise be payable to the principal creditor.”
[145]Cotton LJ said at 103:
“I think, especially having regard to the proviso, that this must be taken to be a bond, not for part of the debt, but to secure payment of the ultimate balance. The proviso clearly points out that that is so, that the surety is not to take advantage of any payments made from time to time by the principal debtor, but is to be liable, though not to a greater extent than £500. Therefore he is not entitled to hold the bank as accountable to him for any dividend they may receive from the principal debtor's estate, or to expunge part of their proof.”
[146]Both these cases seem to me to turn on the contractual arrangements between creditor and surety, in circumstances where the surety had retained the right of indemnity from the principal debtor. They did not consider the impact of any agreement between the surety and the principal debtor (because it did not arise on the facts). Nevertheless, the approach of these two cases does appear to have hardened into a rule.[147]The clearest exposition of the rule in its normal operation is Re Sass [1896] 2 QB 12 . Mr Stourton guaranteed Mr Sass’ indebtedness to the bank, up to a limit of £300. At the date of his bankruptcy, Mr Sass owed the bank £755. Following Mr Sass’ bankruptcy Mr Stourton paid £303 under the guarantee. The bank then proved in the bankruptcy for the whole of the £755; but the trustee rejected the proof on the ground that it ought to have been reduced by the £303 paid by Mr Stourton. Vaughan Williams J disagreed. He said at 14:
“I think that the common law right of the bank here was to sue the debtor for the whole amount that was due from him to them, irrespective of the sum which was paid by the surety, unless that sum amounted to 20s in the pound. When bankruptcy supervened the right of the principal creditor—the bank—was to prove for that amount, unless there was a surety and that surety was a surety for a part of the debt. In that case, if the surety is a surety for part of the debt, and the surety has paid that part, then by virtue of that payment the right of proof, which would have been the right of proof of the principal creditor, becomes pro tanto the right of proof by the surety. The surety has a right, having paid part of the debt in that way, to stand pro tanto in the shoes of the principal creditor; and even if the principal creditor has proved and has received the dividend, and the surety comes and repays the full amount, the principal creditor would then be trustee for the surety of the amount of the dividend which he had so received. In my judgment that right of the surety as against the principal creditor only arises in a case where the surety has paid the whole of the debt.”
[148]The first quoted sentence from the judgment of Vaughan Williams J certainly suggests that it is of general application. But he cited no authority for that proposition. Having looked at the authorities that were cited to him, it does not seem to me that any of them support that bald proposition. Moreover, despite the fact that Commercial Bank of Australia had been cited to him, he did not refer to it. In my judgment the common law (outside insolvency) is not as Vaughan Williams J stated it. It is as described in MS Fashions and Milverton . Sass is better seen as limited to an exposition of the rule against double proof in insolvency.[149]It is, however, clear from Sass that if the surety has only guaranteed part of the debt and has paid that part, he is entitled to prove (although that in itself sits uneasily with Vaughan Williams J’s unequivocal statement of the common law position). That is so even if the creditor has not recovered the full amount of the debt. The necessary corollary is that the payment by the surety will have discharged part of the debt. To that extent, therefore, there will be competition between the surety and the creditor for what can be salvaged out of the insolvent estate. That, therefore, is one qualification to the apparent rigidity of the rule.[150]This point is illustrated by re Bedell ex p Gilbey (1878) 8 Ch D 248. Mr Bedell’s creditors agreed a composition of his debts at 7s 6d in the pound, payable in three instalments. Mr Gilbey guaranteed the last of the three; and paid it. The composition failed, with the result that the creditors became entitled to their debts in full. In Mr Bedell’s bankruptcy Mr Gilbey opposed the lodging of a proof for the balance of the debt. He had, however, lodged his own proof for the amount that he had paid under the guarantee. James LJ said at 253:
“With regard to the main question, the right of a creditor to prove for the whole amount of his original debt, less what he has actually received, I cannot understand how, upon any legal principle, that right can be questioned. … Of course, the creditor who seeks to prove must give credit for whatever he has received, but it can make no difference that he has received it in respect of a composition. In whatever mode he has received it he must give credit for it.”
[151]Brett LJ said 254:
“If there were no surety it would be clear that the creditor would be entitled to prove in the bankruptcy for the whole amount of his original debt, less what he has received. But a surety is brought into the arrangement, and he undertakes to pay the third instalment of the composition in case the debtor does not. The surety has paid the third instalment, and it is said that the contract with him will be altered if the creditors are now allowed to prove for what remains unpaid of their original debts. It seems to me, however, that the surety must be taken to have contracted subject to the known rule of law by which, if the debtor fails to pay the composition, the creditors are remitted to their old debts. The surety has contracted, subject to that rule of law, that, if the debtor does not pay, he will do so, and I cannot see that his position or the consideration for which he undertook to pay will be in any way altered.”
[152]Thesiger LJ said at 255:
“So long as the creditors give credit for the sums which have been actually paid to them, whether by the debtor or by the surety, there is nothing of which the surety has a right to complain.”
[153]Re Sass was approved by this court in Re Fenton [1931] 1 Ch 85. In that case Mr Fenton was the guarantor of bank loans to an association of which he was a member. In the liquidation of the association and the bankruptcy of Mr Fenton, the association’s liquidator lodged a proof for sums due from Mr Fenton to the association. Mr Fenton’s trustee rejected the proof and claimed to set off sums which had been lent to the association and which Mr Fenton had guaranteed. The bank had also proved in Mr Fenton’s bankruptcy but had not received any payment. This court decided that since nothing had been paid to the bank by Mr Fenton or his trustee, the trustee was not entitled to set off Mr Fenton’s contingent liability against sums due from him to the association. As Lawrence LJ put it at 110-1:
“The effect of allowing a set-off in the circumstances is that the estate of the insolvent surety will obtain from the estate of the insolvent principal debtor 20s. in the pound on the guaranteed debts, although neither the surety nor his trustee has paid anything to the principal creditors under the guarantees and although the estate of the principal debtor, after such set-off, will still remain liable for the full amount of the guaranteed debts. That cannot be right.”
[154]He went on to say at 115:
“In the present case the surety has guaranteed the whole of the debts of the principal creditors, although he has limited his liability under each guarantee to a fixed amount. The effect of this is that the principal creditors have the right to prove against the estate of the principal debtor for the whole of their debts, and until they have received 20s. in the pound on those debts the surety cannot prove against the estate of the principal debtor, even although he may have paid the full amount for which he is liable under his guarantees…Even where the principal creditor has been paid in full partly by a dividend from the estate of the insolvent surety and partly by a dividend from the estate of the insolvent principal debtor, the trustee of the insolvent surety will not be allowed to prove against the estate of the principal debtor for the amount which the estate of the surety has contributed towards the payment of the debt, as it is only when the surety has paid the full amount of the debt that he will be subrogated to the rights of the principal creditor…”
[155]Romer LJ said 118-9:
“But I cannot agree that a surety who has not paid off the principal creditor can prove in the bankruptcy of the principal debtor so as to share in the distribution of his assets unless the principal creditor has renounced in some way his right to lodge a proof himself while preserving, of course, his rights against the surety. To allow such a sharing in the assets would be to subject the assets to two claims in respect of the same debt, and this is contrary to the well established rule in bankruptcy against double proof.”
[156]He continued at 119:
“In the present case, if Fenton, not having paid the banks anything under his guarantee, were entitled to prove in the winding-up of the Association, or if, having paid them less than the amount due to them, he were to prove for the amount so paid, and the banks were also to prove in the winding-up of the Association for the full sum due to them, as they would be entitled to do, the estate of the Association would be subjected to more than one proof in respect of the same debt, and this is not permissible. The claim of Fenton against the Association in either case would be a claim in respect of the same debt as that claimed by the banks. … the only reason why Fenton is prevented from proving his claim is that his claim is in respect of the same debt as is that of the banks, and as between him and the banks the latter have the prior right of proof. I am accordingly of opinion that the claim of Fenton's trustee cannot be proved in the liquidation of the Association, there being no evidence that the banks have in any way renounced their right to prove .” (Emphasis added)
[157]Romer LJ does seem to have contemplated that if the creditor had renounced their right to prove, then Fenton would have been entitled to prove for his own debt. The reason, I infer, is because there would, in those circumstances, be no competition. That is another qualification of the apparent rigidity of the rule. Fenton was, of course, a case where nothing had been paid, so it does not directly bear on the situation in our case.[158]There is a further exception to the rule in the case of a negotiable instrument. Where the holder of a negotiable instrument has been partly paid by the drawer or an indorser before he has received a dividend in the acceptor’s insolvency, he is only entitled to prove for the balance; and the drawer and endorser are entitled to prove for what they have paid: Andrews and Millett para 13-008. This exception seems anomalous; and it is not clear why it exists.[159]Mr Phillips asserted (or conceded) a further exception. He said that where the guarantee contained a “principal debtor” clause, that made the surety primarily liable together with the principal debtor. In that event, the creditor would be restricted to proving for the unpaid balance; and the surety could prove alongside him for what he (the surety) had paid. The reason for this, he said, is that payment by the surety is equivalent to payment by the principal debtor. For the reasons that I have given, I do not consider that the effect of the principal debtor clause is as Mr Phillips contended. But if it were, then there would once again be competition between creditor and surety. That would represent another breach in the rule against double proof, as it has so far been formulated. As Ms Tolaney pointed out, it would also have the paradoxical consequence that the creditor with the stronger guarantee (i.e. one with the principal debtor clause) would be in a worse position than the creditor with the weaker guarantee (i.e. one without such a clause). The former would be in competition with the surety in seeking a dividend from the insolvent estate, while the latter would have the field to himself.[160]We have been referred to a number of cases from Australia and New Zealand ( Westpac Banking Corporation v Gollin & Co Ltd [1988] VR 397; Stotter v Equicorp Australia Ltd [2002] 2 NZLR 686 as well as The Public Trustee of Queensland v Octaviar Ltd [2009] QSC 202 to which I have referred) which consider the effect of part payment by a surety. They make for very interesting and thought-provoking reading but do not deal with the situation we have to consider. The last of these cases (to which I have already referred) was not directly concerned with the rule against double proof. The first two were; but since they reach opposite conclusions and neither is binding on us, I do not think that I need consider them further.[161]Rowlatt considers both Ulster Bank and Stotter at para 11-02 and says:
“The better view is that unsecured creditors of the principal debtor should not receive a windfall because the creditor could not prove for the full debt in the principal debtor’s bankruptcy, and, notwithstanding the part payment, the surety could not prove in the same bankruptcy because of the rule against double proof.”
[162]Phillips O’Donovan & Courtney on The Modern Contract of Guarantee (English edition 4 th ed) share that view at para 10-050. Andrews & Millett take the same view at 13-007:
“Where the principal is insolvent and the surety makes a part payment to the creditor before the creditor has been paid a dividend, the rule is that the surety has no right to prove, and the creditor does not have to give credit by reducing his proof by the amount received from the surety, so long as the creditor does not receive more than 100 pence in the pound. The creditor can prove for the full amount and the surety is barred from proving at all. It makes no difference that the payment was before or after the commencement of the insolvency.”
[163]Goode and Gullifer para 8-18 also take that view; and go on to explain:
“At first sight it seems surprising that, if the surety pays part of the debt the creditor should not have to give credit at least for sums received from the surety prior to the bankruptcy. But the rule has a sound policy base. It is a well settled principle of equity that until the creditor has received payment of the guaranteed debt in full the surety cannot prove in the insolvent debtor’s estate for a sum paid by him to the creditor. The reason for this is that he has, expressly or by implication, undertaken to be responsible for the full sum guaranteed, including whatever remains due to the creditor after receipt of dividends by him out of the bankrupt’s estate, and thus has no equity to prove for his right of reimbursement in competition with the creditor. If the creditor were required to give credit for a pre-bankruptcy part payment by the surety, neither of them could prove for the amount of such payment and the general body of creditors would thus be unjustly enriched. Similarly, sums received by the creditor before the bankruptcy from the realisation of security furnished by the surety are not deductible in computing the amount for which he can prove. A fortiori credit need not be given for sums received after bankruptcy and before proof, still less for receipts after proof.”
[164]Deutsche Bank do not dispute the existence of the rule against double proof; or, indeed, the policy underlying it. That policy is to prevent potentially competing claims between the creditor and the surety. What makes the difference in this case, they say, is that there is no possibility of a competing claim by the surety, because under the terms of the Settlement Agreement, LBHI in its capacity as surety released any right to indemnity from the principal debtor. This does not appear to have been an argument that the judge considered at all. Nor is it one that the textbook writers have considered. Nor is it a feature of any of the insolvency cases to which we were referred. We are, therefore, dealing with a novel situation.[165]Deutsche Bank support this argument by reference to paragraph 13-007 of Andrews & Millett in which the authors say:
“… where the surety does not need to rely on a right of subrogation in order to enforce his claim for an indemnity because he has an express or implied right to an indemnity, and he has made a part-payment, and the creditor has been paid the balance of the indebtedness by the principal or by a co-surety, then the surety may prove in the principal’s insolvency for the amount which he has paid to the creditor. The same consequence should follow if the creditor has been paid in full by a co-surety and the surety has made a payment to the co-surety in contribution. In either case there is no possibility of a double proof, because the creditor has been satisfied in full and would recover in excess of 100 pence in the pound.”
[166]They refer also to the judgment of Romer LJ in Fenton in which he said that if the creditor had renounced his right to prove, the surety would have been able to do so. That, they say, is consistent with the policy underlying the rule, namely the prevention of double dividends in respect of the same debt. If there is no such possibility, then the justification for the rule falls away.[167]I find it difficult to see how a judge-made rule such as the rule against double proof can alter the substantive rights of rival claimants. That is consistent with the broad proposition that, subject to certain exceptions, even the statutory insolvency code does not alter substantive rights; but merely regulates the manner in which they can be enforced (see Wight v Eckhardt Marine GmbH [2004] 1 AC 147 ). What is easier to understand, however, is a judge-made rule that regulates the procedure for giving effect to those rights. The circumstances in which a surety is allowed to prove suggest to me that it is indeed a procedural rule. Thus a surety may prove where he has paid the creditor in full; and he may also prove where he has guaranteed part of the debt (as opposed to the whole debt subject to a limit on liability). In the latter case, the rule against double proof does permit a surety to prove even though the creditor has not recovered the debt in full. Likewise in the case of a negotiable instrument, the creditor may only prove for the balance, and the drawer or indorser may prove for what he has paid.[168]That, of course, leaves the conundrum that, if I am right so far, outside the exceptional cases to which I have referred, the creditor is permitted to prove for the whole of the original debt despite the fact that it has been partially discharged. But I consider that Ms Tolaney is right to say that the creditor’s entitlement to prove for the whole of the original debt is a judge-made fiction, for the sole purposes of proof, that the whole of the original debt remains outstanding. It is tempered, in the ordinary case, by the creditor’s liability to repay the surety any surplus, over and above the original debt, that he recovers in the insolvency. The policy underlying the fiction is that the surety, rather than the creditor, has taken the risk of the principal debtor’s insolvency. Thus priority is given to the creditor to salvage what can be salvaged from the insolvent estate, leaving the surety without remedy against the estate unless and until the creditor has recovered in full (either from the estate or the surety or a combination of both). Only then can he participate in whatever crumbs are left.[169]The potential unfairness of the rigid application of the rule against double proof in the circumstances of our case can be illustrated by an example. Suppose that the principal debtor owes the creditor £1 million. The liability is guaranteed by a surety. The surety, on being called under the guarantee, pays £500,000. On ordinary principles, the surety would have a claim against the principal debtor for the £500,000 he has paid. But suppose that he releases that claim; so that the principal debtor owes him nothing. The principal debtor enters into some form of insolvency process. If there is nothing owing to the surety by the principal debtor, then there is nothing for which the surety can prove. The creditor, however, proves for the original £1 million. A dividend of 60 p in the pound is declared; and so the creditor is paid £600,000 in the insolvency. The creditor has now received £1.1 million: more than the original debt. The effect of that is that £100,000 has become unavailable for distribution among the principal debtor’s other creditors. It is the proving creditor rather than the other creditors who receives the windfall. But since the surety no longer has a right to be indemnified by the principal debtor (having released that right) it is hard to see what obligation (either in law or in equity) the creditor has to pay over the surplus £100,000 to the surety. Mr Phillips did not suggest a legal basis for such an obligation.[170]Mr Phillips sought to meet the point by referring to clause 2.04 of the Settlement Agreement. That, he said, had in effect substituted for the surety’s equitable right to receive payment from the creditor of any surplus a contractual right to the same effect. This was a point that was not argued before the judge (or indeed foreshadowed in his skeleton argument despite the effect of the release having been one of the mainstays of Deutsche Bank’s case). The convoluted drafting of the Settlement Agreement is difficult to understand; and Mr Phillips did not really develop the point. Moreover, it was not common ground that clause 2.04 had this effect. Since the Settlement Agreement is governed by the law of New York, the effect of the clause, if in dispute, would have to be proved by expert evidence. We were shown no such evidence, and the judge made no finding. Expert evidence on New York law was called at trial; but since this point was not raised, it was not covered by that evidence. I do not consider that it would be fair to Deutsche Bank to place reliance on this point, if indeed it is well-founded.[171]Even if my analysis of the rule thus far is wrong, the rule against double proof is a judge-made rule and may be developed by the judges when a new situation arises.[172]In my judgment, Deutsche Bank’s point is well-founded. Where(a) the surety has paid part of the debt owed by the principal debtor to the creditor and(b) the surety has given up any right to indemnity from the principal debtor, with the consequence that he has no entitlement to prove for anything, then the creditor must give credit for the payment in the insolvency of the principal debtor. It is the second condition that makes all the difference. There is nothing in the Insolvency Rules which deals with the rule against double proof. Consequently I do not consider that a modest development of the rule intrudes upon legislative competence. Result[173]For those reasons, I have reached the following conclusions: i) Claim A must be paid in priority to Claim B; ii) The claim to rectify Claim B fails; iii) Claim D must be paid before Claim C; iv) Claim C must be reduced, for the purposes of proof, by what the surety has paid.[174]I would therefore: i) Dismiss the appeal against paragraph 1 of the judge’s order; ii) Allow the appeal against paragraph 6 of the judge’s order; and iii) Allow the appeal against paragraph 7 of the judge’s order.

Lord Justice Henderson:

[175]I agree with both judgments.

Lady Justice Asplin:

[176]I am very grateful to Lewison LJ for his careful and comprehensive analysis of the relevant contractual provisions and authorities in this case, and I entirely agree with all of his conclusions.[177]I add a few words in relation to the novel point which has arisen as a result of the partial payment of Claim C under the LBHI Guarantee, coupled with the release of the right to an indemnity from the principal debtor pursuant to clause 8.02 of the Settlement Agreement, to which the surety would otherwise have been entitled.[178]I too consider, instinctively, that if a surety pays part of a guaranteed debt, and releases his right of indemnity from the principal debtor, the amount which the creditor is entitled to recover in the principal debtor’s insolvency must be reduced as a result of the payment and as a result, the creditor can only prove in the principal debtor’s estate for the lesser sum.[179]It seems to me that that instinct is borne out by the authorities to the extent that they address the point with which we are concerned. In my judgment, it is clear from MS Fashions and from the Milverton case, both of which are binding upon us, that a payment by a surety operates in respect of the same debt which is owed by the principal debtor and operates to extinguish or, in the case of part payment, reduce the debt itself. I am unable to accept Mr Phillips’ argument that MS Fashions should be distinguished on the basis that the guarantees contained “principal debtor” clauses. None of the judgments accords the principal debtor clause such importance and Mr Phillips’ approach is not borne out in the textbooks. Furthermore, I can see no reason why the broad proposition expressed by Glidewell LJ in the Milverton case (at 30B) should be restricted to landlord and tenant cases.[180]But what of the rule against double proof? The rule was explained succinctly by Lord Walker in Re Kaupthing Singer & Friedlander Ltd (in administration) (No 2) . The relevant passages are set out at [137] and [138] above. As Lord Walker explains, the effect of the rule is that if the creditor has not been paid in full, the surety, having paid a part of the debt, cannot compete with the creditor by proving in the principal debtor’s estate for an indemnity.[181]As Lewison LJ has explained at [139] above, the rule has two facets: it allows the creditor to prove for the whole of his debt without giving credit for what he has received from the surety in part payment; and it precludes the surety from proving, until one way or another, the creditor has received 100 pence in the pound.[182]I too consider that Ms Tolaney QC is right to describe the creditor’s right to prove for the entire debt despite having received part payment from the surety as a judge-made fiction which is procedural only. It operates to give the creditor the opportunity to recover what he can from the insolvent estate leaving the surety without a remedy unless the creditor recovers in full. The policy behind the fiction appears to be that the surety, who has guaranteed the whole debt, has an obligation to the creditor for the entirety of the liability and has taken the risk that the principal debtor might be unable to repay the original debt. Accordingly, the creditor should have priority to recover as much as he can from the principal debtor’s estate and the surety should not be able to compete with him. The harshness of the fiction is tempered by the fact that the creditor will be liable to repay the surety any surplus he has received over and above the original debt.[183]It seems to me, however, that there is no need for the fiction where there is no competition between the surety and the creditor in the principal debtor’s insolvency. In this case, there is no competition because the surety’s right to an indemnity has been released. He has nothing for which he can prove. No question of double proof arises. Accordingly, in my judgment, there is nothing, whether as a matter of policy or otherwise, to interfere with the operation of the principles in MS Fashions and the Milverton case. Part payment by the surety extinguishes part of the original debt and the amount for which the creditor can prove is, therefore, reduced. Appendix Lehman Brothers - Flow Of Funds

Cited in 4 later judgments