QUEEN’S BENCH, COURT OF APPEAL & HOUSE OF LORDS

CHARTER REINSURANCE CO. LTD., RESPONDENTS

AND

FAGAN, APPELLANT

Authoritative report at:  [1997] A.C. 313

DATES:  (Q.B.D.)  1995 May 22, 23, 24, 25; June 23

(C.A.)  1995 Sept. 12, 13, 14; Oct. 25

(H.L.)  1996 Feb. 21, 22, 26; May 22

JUDGES:  (Q.B.D.)  Mance, J.

(C.A.)  Nourse, Staughton and Simon Brown L.JJ.

(H.L.)  Lord Goff of Chieveley, Lord Griffiths, Lord Browne-Wilkinson, Lord Mustill and Lord Hoffmann

COUNSEL:  (Q.B.D.)  Sydney Kentridge Q.C., John Rowland and Andrew Neish for the insurers.

Gordon Pollock Q.C., Robert Hildyard Q.C. and Stephen Ruttle for the reinsurers.

(C.A.) Gordon Pollock Q.C., Robert Hildyard Q.C. and Stephen Ruttle for the reinsurers.

Sydney Kentridge Q.C., John Rowland and Andrew Neish for the insurers.

(H.L.)  Sydney Kentridge Q.C., John Rowland and Andrew Neish for the insurers.

Gordon Pollock Q.C., Robert Hildyard Q.C. and Stephen Ruttle for the reinsurers.

SOLICITORS:  Ince & Co; Davies Arnold Cooper.

Queen’s Bench Division

28 June 1995.

MANCE J.  delivered the following judgment.

Introduction

This case concerns the effect of three excess of loss reinsurances incorporating ultimate net loss clauses in standard form in circumstances where the reinsured is insolvent. The issue is short and I am asked by both parties to determine it under R.S.C., Ord. 14A. Neither party has formally adduced any expert evidence, although I have been shown various sets of standard clause. The reinsurances contain arbitration agreements,  [*318]  but these have evidently been waived, and did not contain “honourable engagement” clauses.

The plaintiff is the Charter Reinsurance Co. Ltd. (“Charter”) in respect of which provisional liquidators were appointed by order of Blackburn J. on 23 June 1994. The defendant, Mr. Fagan, represents Syndicates 540 and/or 542 (“the syndicates”) whose members reinsured Charter for relatively small percentages under the relevant reinsurances. The issue is whether Charter is entitled to recover under the reinsurances in respect of claims received from its original insureds (or reinsureds) to the extent that Charter has either agreed or been held liable to meet the same, in circumstances where, due to its insolvency, it has not “made payment . . . by way of transfer of funds” to its original insureds or reinsureds. Since insurance and reinsurance claims may be settled in account through market clearing offices, I shall refer in this judgment to disbursement or other satisfaction, to identify what I perceive to be the real point. The syndicates submit that their liability to indemnify Charter is conditional upon disbursement or other satisfaction of such claims by Charter.

The issue has been raised though not decided in at least two previous cases in recent years involving reinsurances incorporating ultimate net loss clauses reinsuring other insolvent companies. It arose in Home and Overseas Insurance Co. Ltd. v. Mentor Insurance Co. (U.K.) Ltd. [1990] 1 W.L.R. 153 and in a subsequent arbitration (the award in which was not put before me), in both of which cases I was myself counsel. It was considered in In re A Company No. 0013734 of 1991 [1992] 2 Lloyd’s Rep. 415, a decision of Mr. Roger Kaye Q.C. sitting as a deputy judge of the Chancery Division. The syndicates’ skeleton argument records that the point is of fundamental market importance, that it involves very large sums of money and that it has wide repercussions; and also that the hearing of this summons has been expedited as a test case and that the Department of Trade and Industry is taking a keen interest in the outcome. One can understand why all of this should be so. Viewed generally the reinsurances appear to be in very typical wordings; the ultimate net loss clauses are in a form originating it appears in the 1930s; the point raised by the syndicates goes to the root of the application of the reinsurances on an excess of loss basis, particularly, though not exclusively in circumstances where the reinsured may be in financial difficulty, and it has potential regulatory and accounting significance. Insurance company insolvency is no new phenomenon, but its prevalence in recent years and the scale of liabilities and potential reinsurance recoveries currently involved are probably unprecedented. In the light of these matters, this judgment is given in open court.

The reinsurance contracts

In the last analysis the issue before me is one of construction of each of the three reinsurances, although the ultimate net loss clause is in general use in excess of loss reinsurances. The three reinsurances were underwritten on a “losses occurring” basis to cover losses occurring during 1989 in the case of the first two and during 1990 in the case of the third. The first two reinsurances, to which both syndicates were party, were evidently part of  [*319]  a structure of layered reinsurances and were closely linked by number, date and drafting. In summary, the first gave cover of £2m. (or U.S.$4m. or Can.$4m.) excess of £3m. (or U.S.$6m. or Can.$6m.) each and every loss, while the second gave cover at a higher level of £2á5m. (or U.S.$5m. or Can.$5m.) excess of £7á5m. (or U.S.$15m. or Can.$15m.) each and every loss. They can be taken together since their wording is effectively identical, and accordingly I will set out only relevant extracts from the first:

“Reinsuring clause

“This Reinsurance is to pay all losses howsoever and wheresoever arising during the period of this Reinsurance on any Interest under Policies and/or Contracts of Insurance and/or Reinsurance underwritten by the Reinsured in their Whole Account. Subject however to the following terms and conditions.

“Liability clause

“The Reinsurers shall only be liable if and when the Ultimate net Loss sustained by the Reinsured in respect of interest coming within the scope of the Reinsuring Clause exceeds £3,000,000 or U.S. or Can.$6,000,000 each and every loss and/or Catastrophe and/or Calamity and/or Occurrence and/or Series of Occurrences arising out of one event and the Reinsurers shall thereupon become liable for the amount in excess thereof in each and every loss, but their liability hereunder is limited to £2,000,000 or U.S. or Can.$4,000,000 each and every loss and/or Catastrophe and/or Calamity and/or Occurrence and/or Series of Occurrences arising out of one event.

“Warranted Reinsurers hereon to have benefit of Specific Reinsurances as per Schedule attached.

“Ultimate net loss clause

“The term ‘Net Loss’ shall mean the sum actually paid by the Reinsured in settlement of losses or liability after making deductions for all recoveries, all salvages and all claims upon other Reinsurances whether collected or not and shall include all adjustment expenses arising from the settlement of claims other than the salaries of employees and the office expenses of the Reinsured.

“All Salvages, Recoveries or Payments recovered or received subsequent to a loss settlement under this Reinsurance shall be applied as if recovered or received prior to the aforesaid settlement and all necessary adjustments shall be made by the parties hereto. Provided always that nothing in this clause shall be construed to mean that losses under this Reinsurance are not recoverable until the Reinsured’s Ultimate net Loss has been ascertained.

“Notwithstanding anything contained herein to the contrary, it is understood and agreed that recoveries under all Underlying Excess Reinsurance Treaties and/or Contracts (as far as applicable) are for the sole benefit of the Reinsured and shall not be taken into account in computing the Ultimate net Loss or Losses in excess of which this Reinsurance attaches nor in any way prejudice the Reinsured’s right of recovery hereunder.  [*320]

“Period of reinsurance clause

“This Reinsurance covers Losses Occurring during the period commencing with 1 January 1989 and ending with 31 December 1989 both days inclusive, Local Standard time at the place where the loss occurs.

“If this Reinsurance should expire whilst a Loss and/or Catastrophe and/or Calamity and/or Occurrence and/or Series of Occurrences arising out of one event is in progress, it is agreed that subject to the other conditions of this reinsurance, the Reinsurers shall pay their proportion of the entire loss or damage provided that the Loss and/or Catastrophe and/or Calamity and/or Occurrence and/or Series of Occurrences arising out of one event commenced before the time of expiration of this Reinsurance.

“Currency clause

“Losses (if any) paid by the Reinsured in currencies other than Sterling, shall be converted into Sterling at the rate of exchange ruling at the date of the settlement of loss or losses by the Reinsured other than losses paid in U.S. or Can. Dollars which will be paid in those currencies.

“Losses discovered or claims made clause

“It is understood and agreed that as regards losses arising under Policies and/or Contracts covering on a ‘losses discovered’ or ‘claims made’ basis, that is to say policies and/or contracts in which the date of discovery of the loss or the date when the claim is made determines under which policy or contract the loss is collectible such losses are covered hereunder irrespective of the date on which the loss occurs provided that the date of the discovery of the loss, in respect of Policies and/or Contracts on a ‘losses discovered’ basis or the date the claim is made in respect of Policies and/or Contracts on a ‘claims made’ basis, falls within the period of this Reinsurance.

“For the purpose of the foregoing the date of the first discovery of a loss occurrence or the date a claim is first made shall be the date applicable to the entire loss and the Reinsurers shall be liable for their proportion of the entire loss irrespective of the expiry date of this Reinsurance provided that such date falls within the period of this Reinsurance.

“Aviation settlements clause (in respect of Aviation business)

“All loss settlements by the Reassured including compromise settlements and the establishment of Funds for the settlement of losses shall be binding upon the Reinsurers, providing such settlements are within the terms and conditions of the original policies and/or contracts (or as provided for in the Extra Contractual Obligations Clause hereof) and within the terms and conditions of this Contract. However, where the Reinsured acts in the capacity as a Leading Underwriter on the business protected by this Contract and is called upon to establish a Fund in respect of a loss or losses which may form the whole or part of a claim within the terms and conditions of this Contract (including provisional collections in respect of Claims  [*321]  which ultimately may not be recoverable hereon) such Fund shall only be binding upon the Reinsurers subject to prior notice being given and agreement obtained from Leading Lloyd’s Underwriter and Leading Company hereon if this Contract is affected by the establishment of such Fund.

“For the purposes of this clause the term ‘Leading Underwriter’ shall be understood to mean those underwriters involved in negotiating the establishment of the Fund.

“Claims clause

“In the event of loss which may cause a claim under this Policy, the Reinsured shall give immediate notice to the Reinsurers, but inadvertent error or omission of such notification shall not prejudice this Policy. All loss settlements by the Reinsured shall be binding upon the Reinsurers provided that such settlements are within the terms and conditions of the Original Policies and within the terms and conditions of this policy and the Reinsurers shall pay the amounts due from them upon reasonable evidence of the amounts paid being given by the Reinsured.

“The amount of the Reinsurers’ liability hereunder in respect of any loss or losses shall not be increased by reason of the inability of the Reinsured to collect from any other Reinsurers (whether specific or general) any amounts which may have become due from them whether such inability arises from the insolvency of such other Reinsurers or otherwise.”

The third reinsurance for 1990, to which Syndicate 540 alone was a party (presumably because it concerned aviation business), is different in important respects. It reads as follows:

“Article 1—Interest clause

“This Reinsurance is to cover the liability of the Reinsured arising under Policies and/or Contracts of Reinsurance in respect of the Reinsured’s Aviation Excess of Loss Underwritings, including the risks of War, Civil War and Allied Perils as and if in original.

“Notwithstanding the foregoing, this Reinsurance shall exclude:Aviation Losses emanating from other departments.Satellite business other than Liability.

“Article 3—Period clause

“This Reinsurance attaches and covers in respect of all losses occurring during the period 12 months commencing with the 1st January, 1990 and ending with the 31st December, 1990 both days inclusive.

“Article 4—Extended expiration clause

“Reinsurers agree that if this Reinsurance should expire whilst a loss, as defined herein, is in progress, then it is agreed that subject to the other conditions of this Reinsurance, Reinsurers shall be liable as if the Whole Loss had occurred during the currency of this Reinsurance.  [*322]

“Article 6—Territorial scope

“This Reinsurance shall apply in respect of all losses wheresoever occurring.

“Article 7—Reinsuring clause

“This Reinsurance is to Indemnify the Reinsured, subject to its provisions, for all losses which may be sustained by the Reinsured in excess of an Ultimate net Loss of £28,000,000 or U.S. or Can.$56,000,000 or equivalent in other currencies each and every loss, as defined herein, up to a further £2,500,000 or U.S. or Can.$5,000,000 each and every such loss.

“Notwithstanding the foregoing Reinsurers shall also be liable for any amount up to the sum reinsured hereunder by which the Reinsured’s Ultimate net Loss exceeds £6,000,000/$12,000,000 each and every loss etc. as defined after all possible recoveries have been made under the Reinsured’s underlying Excess Loss Programmes as Schedule

“Article 8—Original aviation loss warranty clause

“This Reinsurance shall only apply to a claim (or claims) where the ‘Total Original Incurred Aviation Loss’ (or Losses) is equal to or exceeds U.S.$50,000,000 or the equivalent in any other currency based upon the reserve in the London Aviation Market, as defined hereunder, at the time of the claim hereunder.

“Nothing in this clause shall mean that loss (or losses) are not recoverable hereon until the ‘Total Original Incurred Aviation Loss’ has been finally determined and paid. The Reinsured specifically agrees to return all claims monies received from Reinsurers hereon as soon as practicable, in respect of any loss (or losses) which (a) is finally determined at below U.S.$50,000,000 or equivalent in any other currency or (b) when the reserve, as defined hereunder, is reduced to below U.S.$50,000,000 or equivalent in any other currency, whichever the sooner.

“The term ‘Total Original Incurred Aviation Loss’ shall mean the insured loss of the Hull (or Hulls) involved in the loss (or losses), whether insured in the Aviation Market or otherwise, before the deduction of any excess or co-insurance retained by the Original Insured, if any, plus the total of all liabilities to passengers and third parties (excluding Personal Accident Insurance and Workman’s Compensation Act payments) whether the amounts have been paid or promised to be paid or reserved whilst under negotiation or dispute with the original Claimant(s). Such amounts being determined by the Leading London Aviation Market Slip Insurers or Reinsurers in conjunction with their Adjusters, Attorneys or Solicitors and shall be inclusive of all loss costs and expenses of adjustment, defence or representation directly or indirectly attributable to the Original Loss (or Losses) coming within the definition of loss hereunder. In the event of no London Aviation Market reserve, the reserve for the purposes of this clause shall be that amount as advised to the Original Underwriters as an estimate of their liability by their Adjusters, Attorneys or Solicitors.  [*323]

“Notwithstanding anything contained herein to the contrary all loss reserves, as defined above, are to be agreed by the Leading Underwriter hereon.

“This clause shall be subject to the general terms, limits and conditions of this Reinsurance and the foregoing provisions shall take precedence in the event of any inconsistency.

“Article 9—Definition of loss clause

“For the purposes of this Reinsurance the term ‘each and every loss’ shall be understood to mean loss and/or accident and/or occurrence and/or catastrophe and/or calamity and/or series thereof arising out of one event.

“Article 10—Ultimate net loss clause

“The term ‘Ultimate net Loss’ shall mean the sum actually paid by the Reinsured in settlement of losses or liability after making deductions for all recoveries, all salvages and all claims upon other Reinsurances whether collected or not and shall include all adjustment expenses arising from the settlement of claims other than the salaries of employees and the office expenses of the Reinsured.

“All Salvages, Recoveries and Payments recovered or received subsequent to a loss settlement under this Reinsurance shall be regarded as if recovered or received prior to the said settlement and all necessary adjustment shall be undertaken by the parties hereto.

“Nothing, however, in this clause shall be construed to mean that losses are not recoverable from Reinsurers until the Ultimate net Loss to the Reinsured has been determined.

“Notwithstanding anything contained herein to the contrary, it is understood and agreed that recoveries under all underlying Reinsurances (as far as applicable) are for the sole benefit of the Reinsured and shall not be taken into account in computing the Ultimate net Loss or Losses in excess of which this Reinsurance attaches, nor in any way prejudice the Reinsured’s right of recovery hereunder.

“Article 14—Notification of loss and settlements clause

“In the event of a loss which may give rise to a claim hereunder the Reinsured shall give notice to the Reinsurers through the Broker for this Reinsurance as soon as practicable.

“All loss settlements by the Reinsured including compromise settlements and the establishment of funds for the settlement of losses shall be binding upon the Reinsurers, providing such settlements are within the terms and conditions of the original policies and/or contracts (or as provided for in the Extra Contractual Obligations Clause hereof) and within the terms and conditions of this Reinsurance.

“Article 15—Currency conversion clause

“All transactions hereunder shall be conducted in Sterling, United States Dollars and Canadian Dollars, the limit and retentions hereunder in United States and Canadian Dollars being the equivalent  [*324] of the relevant amounts as set forth in Article 7 exchanged at the rate of $2 = £1.

“For the purpose of this Reinsurance all claims paid by the Reinsured in currencies other than Sterling, United States Dollars or Canadian Dollars shall be converted into Sterling at the rates of exchange at which such items are entered in the Reinsured’s books.

“Where a loss involves payments in both United States and Canadian Dollars, the total number of dollars payable hereunder shall first be determined with the Reinsured’s loss in the two currencies added together. The amount so determined shall then be paid by the Reinsurers in the two currencies in the proportion that the United States and Canadian Dollar loss bears to the Total Ultimate net Loss.

“Where a loss involves payments in United States and/or Canadian Dollars and Sterling, the total amount payable hereunder shall first be determined in Sterling with the Reinsured’s Ultimate net Loss in United States and/or Canadian Dollars converted at the rate of $2 = £1. The amount due from Reinsurers having been thus determined the Reinsurers shall pay that amount in United States Dollars, Canadian Dollars and Sterling in the same proportion as the Dollar and Sterling Loss bears to the total Ultimate net Loss.

“Article 18—Reinsurance clause

“This Reinsurance shall be deemed to be subject to the same terms, clauses and conditions as the Original Policies as far as they may be applicable hereto, and shall pay as may be paid thereon, but subject nevertheless to the terms and conditions of this Reinsurance.

“Article 22—Special cancellation clause

“Either party shall have the right to terminate this Reinsurance immediately by giving the other party notice by Telex or Telegram and shall be deemed to be served upon despatch or where communication between the parties are interrupted upon attempted despatch. (i) If the performance of the whole or any part of this Reinsurance be prohibited or rendered impossible de jure or de facto in particular and without prejudice to the generality of the preceding words in consequence of any law or regulation which is or shall be in force in any country or territory or if any law or regulation shall prevent directly or indirectly the remittance of any payments due to or from either party. (ii) If the other party has become insolvent or unable to pay its debts or has lost the whole or any part of its paid up capital. (iii) If there is any material change in the management or control of the other party. (iv) If the country or territory in which the other party resides or has its head office or is incorporated shall be involved in armed hostilities with any other country whether war be declared or not or is partly or wholly occupied by another power or be in a state of civil war. (v) If the other party shall have failed to comply with any of the terms and conditions of this Reinsurance.

“All notices of termination served in accordance with any of the provisions of this Article shall be addressed to the party concerned at  [*325]  its head office or at any other address previously designed by the party.

“The Reinsurer shall remain liable for losses occurring up to and including the date of termination. Thereafter the liability of the Reinsurers shall cease outright other than so far as outstanding claims are concerned.

“In the event of this Reinsurance being terminated in accordance with the provisions of the Article the exact premium payable hereunder shall be based on the premium income figure for the period from inception to the effective date of termination and shall be subject to a minimum premium calculated on the amount specified in Article 17 pro rata for the effective period.”

The correct approach to construction

The parties’ submissions adopt different approaches to the issues of construction which arise. The syndicates say that the words of the contracts alone are so clear that any further thought about their implications or aid to their true construction is not only unnecessary but wholly inappropriate; they cite words of Lord Halsbury L.C. in Leader v. Duffey (1888) 13 App.Cas. 294 (a case concerning a marriage settlement) and Smith v. Cooke; Storey v. Cooke [1891] A.C. 297 (a case concerning a deed of assignment of a partnership business and property). Charter submit that construction should never be a wholly abstract or literal exercise, divorced from any consideration of context or practical implications. On the one hand, the court should not approach the construction of any contract with notions of principle or reasonableness conceived in the abstract and seek to force the provisions of a particular contract into that straitjacket: per Saville J. in Palm Shipping Inc. v. Kuwait Petroleum Corporation [1988] 1 Lloyd’s Rep. 500, citing an earlier dictum of Lord Goff of Chieveley to like effect in Société Anonyme Marocaine de l’Industrie du Raffinage v. Notos Maritime Corporation [1987] 1 Lloyd’s Rep. 503, 506. On the other hand, there is a wealth of authority, which is of particular relevance in the commercial context, that the court should seek to place itself in the same matrix as the parties were when contracting and to understand their general aim, objectively assessed, and that considerations of reasonableness or “commerciality” can play an important role in this exercise. Mr. Kentridge cited passages from Antaios Compania Naviera S.A. v. Salen Rederierna A.B. (The Antaios) [1985] A.C. 191, 200-201 and Wickman Machine Tool Sales Ltd. v. L. Schuler A.G. [1974] A.C. 235, the familiarity of which in no way detracts from their forcefulness on rereading in the present connection

For my part, I adopt and apply recent guidance given in the Court of Appeal in the unreported authority in Arbuthnott v. Fagan (unreported), 30 July 1993; Court of Appeal (Civil Division) Transcript No. 1024 of 1993 where a similar issue about the correct approach to construction arose. The context was the wording of the standard form of agency agreement prescribed by Lloyd’s byelaw prior to 1989; the language made it, according to the underwriting agencies, a condition precedent to the accrual of any cause of action against a particular agency in respect of a particular syndicate and year that the Name must first pay all calls made  [*326]  upon him or her for underwriting expenses or liabilities in respect of that syndicate and year. This defence failed. Sir Thomas Bingham M.R. said, on construction:

“Courts will never construe words in a vacuum. To a greater or lesser extent, depending on the subject matter, they will wish to be informed of what may variously be described as the context, the background, the factual matrix or the mischief. To seek to construe any instrument in ignorance or disregard of the circumstances which gave rise to it or the situation in which it is expected to take effect is in my view pedantic, sterile and productive of error. But that is not to say that an initial judgment of what an instrument was or should reasonably have been intended to achieve should be permitted to override the clear language of the instrument, since what an author says is usually the surest guide to what he means. To my mind construction is a composite exercise, neither uncompromisingly literal nor unswervingly purposive: the instrument must speak for itself, but it must do so in situ and not be transported to the laboratory for microscopic analysis.”

Steyn L.J. said:

“I readily accept Mr. Eder’s submission that the starting point of the process of interpretation must be the language of the contract. But Mr. Eder went further and said that, if the meaning of the words is clear, as he submitted it is, the purpose of the contractual provisions cannot be allowed to influence the court’s interpretation. That involves approaching the process of interpretation in the fashion of a black-letter man. The argument assumes that interpretation is a purely linguistic or semantic process until an ambiguity is revealed. That is wrong. Dictionaries never solve concrete problems of construction. The meaning of words cannot be ascertained divorced from their context. And part of the contextual scene is the purpose of the provision. In the field of statutory interpretation the speeches of the House of Lords in Attorney-General v. Prince Ernest Augustus of Hanover [1957] A.C. 436 showed that the purpose of a statute, or part of a statute, is something to be taken into account in ascertaining the ordinary meaning of words in the statute: see Viscount Simonds’s speech, at p. 461, and Lord Somervell of Harrow’s speech, at p. 473. It is true that such a purpose may also be called in aid at a later stage in the process of interpretation if the language of the statute is ambiguous but it is important to bear in mind that the purpose of the statute is a permissible aid at all stages in the process of interpretation. In this respect a similar approach is applicable to the interpretation of a contractual text. That is why in Reardon Smith Line Ltd. v. Yngvar Hansen-Tangen (trading as H.E. Hansen-Tangen) [1976] 1 W.L.R. 989, 996 Lord Wilberforce, speaking for the majority of their Lordships, made plain that in construing a commercial contract it is always right that the court should take into account the purpose of a contract and that presupposes an appreciation of the contextual scene of the contract. Corbin on Contracts (1960), vol. 3, section 545, explains the role that the ascertainment of the purpose of a contract  [*327]  should play in the process of interpretation: ‘In order to determine purposes we are obliged to interpret their words in the document of agreement and their relevant words and acts extrinsic to that document. It may seem foolish, therefore, to say that the words of a contract should be interpreted in the light of the purposes that the parties meant to achieve, when we can turn on that light only by process of interpretation. Nevertheless, it is believed that such an admonition serves a useful purpose. As the evidence comes in and as interpretation is in process, the court may soon form a tentative conviction as to the principal purpose or purposes of the parties. As long as that conviction holds (and the court must be ready at all times to be moved by new evidence), further interpretation of the words of contract should be such as to attain that purpose, if reasonably possible.’ In the same section of this seminal work the author added that if the court is convinced that it knows the purpose of the contract, however vaguely expressed and poorly analysed, it should be loath to adopt any interpretation of the language that would produce a different result. In my judgment these observations accurately state the approach to be adopted. And in the present case the purpose of clause 9(c) is not in doubt.”

Steyn L.J.’s application of this approach is of note:

“The implications of the agents’ argument that clause 9(c) precludes the Names from suing the agents for negligence so long as a cash call in respect of syndicate and year account remains outstanding generates immediate scepticism. This is an invitation to adopt an interpretation which is at variance with the purpose of clause 9(c). This interpretation achieves something that is commercially unnecessary and different from the acknowledged purpose of clause 9(c). It amounts to saying that clause 9(c) has the coincidental or collateral effect that the agent is protected against actions in negligence while a cash call remains unpaid. Furthermore, as Mr. Boswood Q.C. said, the agents’ interpretation leads to the extraordinary result that if the agent ruins a Name by negligent underwriting, so that the Name cannot pay the cash call, the contract breaker or tortfeasor goes scot-free. And that result is inimical to the interests of policyholders and the Lloyd’s market since the claim against the agent may be an asset available to meet the policyholders’ claims. That is so uncommercial and unreasonable a result that words of the greatest precision would be required to achieve it. Clause 9(c) plainly comes nowhere near this.”

Finally, Hoffmann L.J. said:

“It seems to me legitimate to test the plausibility of a given construction by examining what the consequences would be. The construction for which the agents contend means that if they are going to be negligent, they should rather ruin their Names entirely than leave them with enough resources to pay their calls. In the latter case they will be exposed to an action for negligence whereas in the former case they will be immune. Mr. Eder said that his startling  [*328]  consequence had to be accepted in the interests of maintaining discipline at Lloyd’s and inducing the Names to pay their calls. But his argument cannot apply to those who have no money. And in cases of contumacious refusal to pay, it is hard to see why denial of the right to sue for negligence will be more effective than the undisputed right of Lloyd’s to obtain judgment for the unpaid calls.”

In Toomey v. Eagle Star Insurance Co. Ltd. [1994] 1 Lloyd’s Rep. 516 Hobhouse L.J. giving the judgment of the court referred to Reardon Smith Line Ltd. v. Yngvar Hansen-Tangen (trading as H.E. Hansen-Tangen) [1976] 1 W.L.R. 989 for the relevance, in the context of a commercial contract, of its commercial purposes objectively ascertained, and then said, at p. 520:

“It is also necessary that the court should have regard to previous decisions of the courts upon the same or similar wording. Parties to a commercial contract are to be taken to have contracted against a background which includes the previous decisions upon the construction of similar contracts.”

There are features of previous decisions of the courts from which both parties in the present case seek to derive assistance, either by way of confirmation of their position or by way of contrast with the present case.

Mr. Kentridge further referred me to Black King Shipping Corporation v. Massie [1985] 1 Lloyd’s Rep. 437, 462-467 for authorities on the principles determining whether a stipulation will be treated as a condition precedent. Hirst J., at p. 462, cited words of Lord Denning M.R. in Attica Sea Carriers Corporation v. Ferrostaal Poseidon Bulk Reederei G.m.b.H. [1976] 1 Lloyd’s Rep. 250, 253:

“The parties can, by clear words, provide that complete performance of a particular stipulation can be a condition precedent. But, in the absence of clear words the court should look to see which of the rival interpretations gives the more reasonable result.”

Lord Denning M.R. then cited Lord Reid’s words in Wickman Machine Tool Sales Ltd. v. L. Schuler A.G. [1974] A.C. 235. A number of other authorities were cited by Hirst J. from different contexts to the present; many of them, such as Tarrabochia v. Hickie (1856) 1 H. & N. 183 to which Mr. Kentridge made particular reference, concern the question whether a particular term should be construed as a condition, warranty or innominate term. I will consider at a later stage to what extent these authorities may assist in the construction of the present contracts.

Previous English authorities on reinsurance

Whilst the present reinsurances and the issues of construction raised by their terms must be approached without preconceptions, it would be inappropriate to approach them without any awareness of previous English decisions in the field of reinsurance. Some words and phrases used in the present reinsurances have been the subject of authority. Others may be directed to problems which existed or were perceived as existing either when the reinsurances were made or when standard clauses incorporated into them were developed for general use.  [*329]  Charter relies upon two particular authorities: In re Eddystone Marine Insurance Co.; Ex parte Western Insurance Co. [1892] 2 Ch. 423 and In re Law Guarantee Trust and Accident Society Ltd; Liverpool Mortgage Insurance Co.’s Case [1914] 2 Ch. 617. They submit that these cases give guidance as to the objects which parties to reinsurance contracts may, at least prima facie, be expected to share, as well as identifying certain commercial considerations to which the courts have in the past attached significance. In these authorities, on the wordings there under review, courts rejected suggestions that a reinsurer’s duty to pay was conditional upon disbursement or satisfaction by its reinsured of incoming insurance claims. The syndicates point to (a) the fact that the wordings there under consideration were in very different form to the present, and (b) the different basis of reinsurance there involved (quota share instead of excess of loss). In their submission, the present wordings contrast starkly, and in all likelihood quite deliberately, with those the subject of any decided authority, and one reason for this is, they submit, the differing bases of the present reinsurances. The most obvious difference in wording, at the forefront of the syndicates’ case, is the incorporation in the present reinsurances of ultimate net loss clauses in standard wording only introduced in the 1930s. But the syndicates also submit that the first two reinsurances contain other indications of careful drafting designed to make prior disbursement or satisfaction a precondition to any reinsurance recovery, and that the wording of the third reinsurance, though less precise, has the like result. The object of these reinsurances is in their submission clear from the language itself, and should be taken at face value without preconceptions or further debate.

In the Eddystone case [1892] 2 Ch. 423 both parties were in liquidation. Whilst the two companies were going concerns, the former had reinsured the latter for the whole or part of its exposure as insurer of various vessels. In the test case tried the reinsurance was for half the exposure. The reinsurances were expressed, at p. 424, to be “subject to the same terms and conditions as the original policy or policies, and to pay as may be paid thereon.” Stirling J. recited the contention of the liquidator that those words made payment by the reinsured a condition precedent to payment by the reinsurer, and said, at p. 427:

“Now, a main object of reinsurance is to relieve the reinsured from a portion of the risk previously undertaken by him; and the result of giving effect to the liquidator’s contention would be that, before the reinsured obtains the benefit of his reinsurance, he must himself have paid on the original insurance, even though bankruptcy might be the result. I think that this could not be intended, and that such a construction ought not to be put on the language of the policy unless it is clearly called for. In my opinion the words do not clearly require to be so construed. They would be satisfied if they were held to amount simply to this—that the payment to be made on the reinsurance policy is to be regulated by that to be made on the original policy of insurance.”

Charter relies on these words; it submits that a similar approach is applicable to the present wordings.  [*330]

In the Law Guarantee case [1914] 2 Ch. 617 the company in liquidation had, whilst a going concern, undertaken to debenture holders that principal and interest would duly be paid by the issuer of the debentures. It had then entered into a contract with a mortgage guarantee company for a yearly premium, whereby the latter company guaranteed it in turn to the extent of two-elevenths of its exposure. On the assumption that the original undertaking was an insurance, the latter contract was a reinsurance. Otherwise it was a simple indemnity. The mortgage guarantee company argued that its liability was limited to paying whatever dividend the Law Guarantee Co. might pay in its liquidation to the debenture holders. Whichever view of the contracts with debenture holders was appropriate, the court held that the company in liquidation was entitled to recover in full from the mortgage guarantee company. Dealing with the case as one of insurance and reinsurance, the court pointed out that in the Eddystone case no one appeared even to have suggested any contrary argument. All members of the court evidently regarded the consequences of the contrary argument as extreme and unacceptable. The force with which they expressed their views may be of some materiality and I quote some extracts. Buckley L.J. said, at p. 634:

“I am anxious to point out some consequences which would ensue if the view I have expressed were not the right one. The contrary view is one which makes it to the interest of the company that the society should be insolvent. For to such extent as the society cannot pay in full the argument is that the company are not liable to make payment. It is obvious that such a consideration cannot have entered into the contemplation of the parties in fixing the premium. The company has received a certain premium upon the terms that in an event it shall pay a certain sum. If its liability is reduced to 10s. in the pound, it has received payment of premium as the price of an obligation to pay 20s., but is, by reason of circumstances not material at all so far as the company is concerned, relieved of one half of the liability. If this were true a society whose credit was bad ought to pay a less premium than a society whose credit was good, because the obligation would in the former case result in a smaller liability. The fact is, I suppose, that as matter of business if the credit of the society were bad a larger, not a less, premium would be demanded as the price of the guarantee.”

Kennedy L.J. put the matter as follows, at p. 638:

“As Neville J. points out in the course of his judgment in this case [1913] 2 Ch. 612, if it be held that payment is a condition precedent to recovery, the contract may be of little value to the person to be indemnified, who may be unable to meet the claim in the first instance. If this is true in regard to a contract of indemnity in general, it is certainly true of contracts of insurance and reinsurance, which are essentially contracts of guarantee against loss by an event.”

He continued, at pp. 639-640:

“How the person who receives payment of a sum of money under a contract of insurance or reinsurance, or, I will add, of indemnity,  [*331]  deals with that sum is, in general and apart from special considerations, no concern of the party who, in fulfilment of his contract, has made the payment to him. Upon what grounds of equity or legal logic can it be argued that, because the law, on grounds of public policy, compels the creditor, the liability to whom is the event upon which the right of a bankrupt or of an insolvent company to payment of the sum covered by the contract arises, to be content with such share of the assets of the bankrupt or the company in liquidation as a pari passu distribution between creditors will give, those assets are not to include the payment due under the contract? If the discharge of the liability, against which the contract of insurance or reinsurance was the protection, is not (and I think that it is not) a condition precedent to the right to claim payment of the amount of the insurance or the reinsurance, upon what ground ought it to be held that, if the person entitled to that payment becomes bankrupt, or, if a company, goes into liquidation, the trustee of the bankrupt or the liquidator of the insolvent company does not have the right to the payment of that amount in full which the bankrupt before his bankruptcy or the company before its liquidation possessed because, when he has received the amount, the law intercepts the asset in his hands, and compels him to apply it in a particular way as part of the estate divisible amongst all the creditors?”

He continued:

“I do not think that it is true to say that in such a case as the present the party who is entitled makes a profit; he gets only the amount for the payment of which he bargained in the event which has occurred, and he has paid a premium based upon the risk, in such event, of that amount having to be paid. Buckley L.J. has dwelt in his judgment upon the strange consequences that, as pointed out by the appellants’ counsel in their argument before us, would ensue if the argument put forward in the present case on behalf of the company were held to prevail. I agree with him in these criticisms.”

The third member of the court, Scrutton J., said, at pp. 646-647:

“For the insurance company it was said that their contract, whatever it was called, was a contract of indemnity, and that they indemnified the society if they paid it what it had to pay: to pay it more would be to enable the society, or its general creditors, to make a profit. For the society it was said that the whole object of reinsurance might be to save oneself from bankruptcy in case of claim, but that the contention of the insurance company involved that a person with no assets other than full reinsurance against his insurance liabilities would yet be driven into bankruptcy, and only recover from his reinsurers the nominal dividend his general assets could pay. Premiums on such reinsurances would have to be regulated not only by the solvency of the original principal debtor, but also by the solvency of the person whose risk was reinsured; the more insolvent the latter was the less the reinsurer would have to pay. And, it was said, how is what the reinsurer is to pay, that is the dividend,  [*332]  to be ascertained until you know his contribution? His contribution when made will increase the general dividend, and this increase will again increase his contribution. To this it was answered that the rule would be: find what dividend the general assets apart from claims on reinsurers would pay the general creditors apart from those whose claims were reinsured; collect this dividend from such reinsurer. It is perhaps sufficient to say there is no trace of this complicated rule in the policy, and it would not apply to cases where there were no general creditors whose claims were not reinsured. For instance, for simplicity take a man who has only insured one risk of £1,000 and, having no assets, has reinsured his whole risk. He becomes insolvent, and a claim for a total loss accrues on his policy. Left to himself his dividend to his creditor will be nil, and on the argument of the insurance company the amount payable by his reinsurer will be nil. Yet in fact he should remain solvent, his reinsurer providing the funds to satisfy his only debt. The same result would follow if he had 10 fully reinsured policies and 10 losses. If it be said that the society’s argument enables the general creditors to make a profit, the answer appears to be that it is not the society’s argument, but the operation of law which, though a creditor has a claim which the debtor has reinsured, takes the contribution of the reinsurer and divides it between the general body of creditors, because the creditor has no privity with the reinsurer, and no charge, lien, or equitable security, on his payment to the debtor.”

Later, after citing from the Eddystone case, he said, at pp. 649-650:

“The judgment may appear to be ambiguous, for the phrase ‘the payment to be made’ might mean liability or might mean actual payment. It has, however, been held that actual payment without legal liability will not do: Chippendale v. Holt (1895) 1 Com.Cas. 197; and Stirling J. says that the contention of the Western Company must prevail. The amount paid before the original insurer has paid can only be the whole liability, and the case is treated by the editors of Arnould on Marine Insurance, 8th ed., p. 422, para. 324, as deciding that the whole amount of liability and not the dividend only is payable. This view of the case has to my knowledge been taken in practice and business. The law in the United States seems to be the same. In Herckenrath v. American Mutual Insurance Co., 3 Barb. Ch. N.Y. 63 in 1848, the Chancellor held that the original assured had no claim for the whole sum received by his insurer from a reinsurer, but only to a dividend. The argument deals with the suggestion that the insurer makes a profit, pointing out that he makes no profit, but the state requires his assets to be distributed in a particular way; he parts with them all. ‘It is none the less a loss because they are unable to pay it in full.’ The Chancellor cited and apparently approved a decision of the Commercial Court of Marseilles in 1748, cited by Émérigon [Traité des Assurances, vol. i., cap. 8, s. 14, p. 253; translation by Meredith, 1850, p. 202], where the insolvent insurer paid 60 per cent. of the loss, but recovered 100 per cent. of the loss from his reinsurer, the court holding that the reinsurer was bound to  [*333]  pay what the original assurer became liable to pay, not what he had actually paid . . . It was suggested in argument that marine policies stood in a peculiar position. Having some acquaintance with them, I do not see why. The insurer reinsures a subject matter, the ship, in which he has an insurable interest, in that if she is lost he will have to pay under his policy. The measure of his loss is not what he pays but what he is liable to pay: Chippendale v. Holt.”

Scrutton J.’s judgment directs attention to another line of authority which is of unquestionable significance to certain parts of the present reinsurances. He had been unsuccessful counsel in Chippendale v. Holt (1895) 1 Com.Cas. 197 and was at pains to point out that it established that a simple provision for reinsurers “to pay as may be paid” does not affect the prima facie rule at common law that recovery under a reinsurance depends upon proof of legal liability (as well as the ascertainment of the reinsured’s financial commitment to his original insured by agreement, judgment or award: see e.g. Versicherungs und Transport A.G. Daugava v. Henderson (1934) 49 Ll.L.R. 252, 253, 254, per Scrutton and Maugham L.JJ.). The subsequent development of ever more comprehensive “follow the settlements” clauses shows however that insurers generally required, and were able to obtain, cover on a broader basis. This has been analysed in Insurance Co. of Africa v. Scor (U.K.) Reinsurance Co. Ltd. [1985] 1 Lloyd’s Rep. 312, (where Scrutton L.J.’s continuing involvement in this area is outlined, at pp. 319-321) and subsequent cases, and was also considered in Toomey v. Eagle Star Insurance Co. Ltd. [1994] 1 Lloyd’s Rep. 516. Its general purpose and effect is that the reinsured recovers for no more and no less than that for which he settles with his original insured. He neither profits, nor (provided he acts in a business-like manner) can he face arguments that he has settled with his original insured in circumstances where he had in law or in fact no actual liability. Scrutton J.’s remarks about “actual payment” in the Law Guarantee case [1914] 2 Ch. 617, 649 were thus designed to emphasise that, at that date, insurers could not look to their reinsurers (using modern terminology) “to follow their settlements,” but had to establish legal liability; he was not concerned with the distinction between settlement and satisfaction which is material in the present case.

Detailed analysis of the present reinsurance contracts

(a) The subject matter of reinsurance

I turn to the wording of the reinsurances. Each starts with a clause identifying the general nature of the cover afforded, described as a reinsuring clause in the first two contracts and an interest clause in the third. Charter submits that in each case reinsurance is granted of Charter’s interest, described in the first two contracts by the words “to pay all losses . . . on all interest under policies and/or contracts or insurance and/or reinsurance underwritten by the reinsured,” and in the third by the simple words “to cover the liability of the reinsured arising under policies and/or contracts of reinsurance.” The third contract is therefore expressly a cover on liability. In the case of the first two, it is possible for the syndicates to submit that “all losses howsoever and wheresoever arising during the  [*334]  period of this reinsurance on any interest” refers to original losses in the same sense as in the later period of reinsurance clause, rather than the reinsured’s loss in the same sense as in the ultimate net loss clauses. I doubt whether this argument is correct, since the reinsuring clause refers to “losses . . . on all interest,” which must, it appears to me, refer to Charter’s insurable interest under the relevant policies which it underwrites, as it also does in the liability clause. The reinsuring clause in the first two contracts makes the cover expressly “subject to the following terms and conditions.” Article 1 in the third contract has no such express qualification. I do not think that this makes any real difference. The point which Charter can properly make on all three contracts is that the subject matter of reinsurance is not directly or expressly defined so as to be related, or limited by reference, to the disbursement of moneys. The primary expression of cover under these reinsurances is against liability or loss affecting the reinsured’s interest under original contracts, not against outlay.

(b) The timing of reinsurance recoveries

The liability clause in the first two contracts contains phrases with temporal connotations—“shall only be liable if and when” and “shall thereupon become liable.” However, the focus of the clause is not on timing but on the scope and limits of the financial responsibility accepted on an excess of loss basis under the contracts. The equivalent clause in the third contract (article 7—reinsuring clause) has no phrases with any temporal connotations at all. The phrases quoted must anyway be approached with caution. Both the liability clause in the first two contracts and article 7 in the third employ the concept of “ultimate net loss” sustained by the reinsured, and both contain ultimate net loss clauses. In the case of the first two contracts, the ultimate net loss clause defines “net loss,” whereas in the third it defines “ultimate net loss.” Taking the liability clause literally, it provides that reinsurers are only to be liable “if and when the ultimate net loss sustained” exceeds the specified limit. But the ultimate net loss clause itself provides always “that nothing in this clause shall be construed to mean that losses under this reinsurance are not recoverable until the reinsured’s ultimate net loss has been ascertained.” The “industrious and anxious” draftsman who Mr. Hildyard suggested as the author of the first two contracts cannot have meant exactly what he was apparently saying in the liability clause in the first contracts. The draftsmanship of the third contract is different but also open to comment. Article 7 provides for indemnification by reference to ultimate net loss and article 10, the ultimate net loss clause, defines this to mean the sum actually paid by the reinsured in settlement of loss or liability. But article 10 then provides that nothing in it shall mean that losses are not recoverable from reinsurers until the ultimate net loss has been determined. Even reading this contract literally (as the syndicates invite) recovery cannot therefore depend on ultimate net loss as defined. Under all three contracts, the cover should no doubt be read as being (a) ultimately, against ultimate net loss as then “ascertained” or “determined,” to use the words of the provisos, and (b) at any intermediate stage, against such “net loss” as may at that stage have been ascertained or determined. That is  [*320]  part of a process of sensible rather than uncompromisingly literal interpretation.

(c) The object(s) of the ultimate net loss clause

In their skeleton argument, counsel for the syndicates suggested that the ultimate net loss clause is not a “measure of loss clause,” because of the proviso in its third sentence. There is no doubt that the clause is a measure of loss clause. It measures the ultimate liability of the reinsurers. It provides for provisional measurement of loss at earlier stages. The most obvious purpose of its extensive wording is to provide clear criteria for such measurement. Many of the criteria may now appear commonplace but insurance and reinsurance have over the years frequently generated disputes in this area about apparently fundamental matters. I need only instance disputes as to whether and when a reinsurer is bound to follow settlements without proof of liability; whether an insurer, if he settles for less than his actual liability, can recover from his reinsurer the full amount of his liability; and whether it makes any difference if a recovery is effected before or after indemnification by an insurer or reinsurer. In the first of these areas a real degree of certainty has only quite recently been provided through the Court of Appeal’s decision in the Scor case [1985] 1 Lloyd’s Rep. 312 and subsequent authorities. The second of these matters was the subject of two conflicting Court of Appeal decisions: Uzielli & Co. v. Boston Marine Insurance Co. (1884) 15 Q.B.D. 11 and British Dominions General Insurance Co. Ltd. v. Duder [1915] 2 K.B. 394. In the Uzielli case the Court of Appeal held that the underwriter could recover up to £1,000 under a reinsurance in that sum against total loss although he had himself settled with the original assured for 88 per cent. of the hull value, whilst also incurring sue and labour expenses to which the reinsurance did not apply, reliance being placed by the court on the “pay as may be paid” clause. In the Duder case the court reached an opposite conclusion, expressing its difficulty in understanding the Uzielli case. The third matter was the subject of litigation to the House of Lords as recently as 1992 in the context of stop loss insurance: Lord Napier and Ettrick v. Hunter [1993] A.C. 713. The caution exhibited by reinsurance draftsmen in the ultimate net loss clause in the 1930s and subsequently is understandable. In his oral submissions, Mr. Ruttle was ready to accept that the clause is substantially a “how much” clause, concerned with the measurement of reinsurance recoveries. The question is whether it also has what would be a further, and most significant, role in requiring disbursement or other satisfaction of amounts ascertained by insurers’ settlements as a precondition to the availability of any reinsurance recovery at all.

(d) “Actually paid . . . in settlement of losses or liability”

The phrase “actually paid by the reinsured in settlement of losses or liability” is at the core of the syndicates’ submissions, although they submit that the wording of the liability clause and claims clause in the first two contracts would alone establish their construction. The phrase looks necessarily, they say, to some disbursement of moneys. The syndicates’ submission is reinforced by the presence of the word “actually.”  [*336]

Had the clause spoken simply of “the sum paid by the reinsured in settlement,” Charter could have founded itself more directly on authority like the Eddystone case [1892] 2 Ch. 423 in its submission that the intention was simply to regulate the payment to be made by the reinsurer by that to be made on the original policy. The syndicates have other reasons why in their submission such authority affords no assistance, but the word “actually” is by itself an obvious difference. They say that this is particularly so when the summary of the argument of counsel in the Eddystone case contains a submission that the wording there did not make “actual payment” to the original assured a condition precedent to reinsurer’s liability. On the syndicates’ submission the draftsman of the ultimate net loss clause should be taken to have had this very submission in mind, although I think that references to the submissions of counsel go beyond anything which Hobhouse L.J. had in mind in his observations in the Toomey case [1994] 1 Lloyd’s Rep. 516 or which would in this context usually be appropriate.

The word “actually” is emphatic. Taking a definition from the Oxford English Dictionary (Compact Edition) cited by the syndicates, it means “in act or fact; as opposed to possibly, potentially, theoretically, ideally; really, in reality.” But it remains to identify what it is that is intended to be emphasised and why. If references to sums paid are otherwise made in the reinsurances to ensure that the payment to be made by the reinsurer is regulated by that to be made on the original policy, the word “actually” may simply be employed to make this more emphatically clear. This would be consistent with the main thrust of the ultimate net loss clause, that is the measurement (interim and final) of the recovery available under the reinsurance. It can be objected that the word “actually” was on that basis unnecessary, but this is an argument with limited force in the context of a reinsurance clause which shows anxiety to deal with every aspect of the measurement of recovery in detail. The fact that timing is dealt with in this way by a proviso or qualification in the third sentence of the clause appears to me to be at least consistent with the possibility that the earlier part of the clause may not have been directed to matters of timing at all.

(e) Condition precedent?

A similar comment applies to the syndicates’ repeated submission that the contracts contain the clear condition precedent that was lacking in authorities such as the Eddystone case [1892] 2 Ch. 423 and the Law Guarantee case [1914] 2 Ch. 617. The ultimate net loss clause on any view expresses some form of condition: recovery depends upon there being a “sum actually paid by the reinsured in settlement of losses or liability,” whatever that may in context involve. The issue is whether the phrase was intended not merely to establish a condition in the sense of a threshold and quantum of ascertained exposure, but also to introduce (i) considerations of disbursement or other satisfaction and (ii) in conjunction therewith, a temporal condition precedent to recovery consisting of prepayment by insurers. It is in my view legitimate to consider this issue with the expectation that, if (i) and (ii) had been intended, they would have been made clear. The authorities cited in Black King Shipping Corporation v. Massie [1985] 1 Lloyd’s Rep. 437 do not  [*337]  specifically relate to circumstances where the clause in question is on any view a condition of recovery of one sort or another, but they provide general support for Charter’s case, in so far as the syndicates’ interpretation involves reading the ultimate net loss clause as introducing a condition of a more expansive and onerous nature than Charter’s construction.

(f) Payment and settlement

When the ultimate net loss clause speaks of “the sum actually paid by the reinsured in settlement,” the syndicates submit that it contemplates two distinct stages in time: settlement first and payment second. In my view there is no reason to read the phrase as concerned with timing at all. However, even if it is, a two-stage analysis does not appear to me to be intended by the wording. The phrase is “actually paid . . . in settlement.” This on its face, either treats the two as concomitant or treats payment as an aspect of settlement. The second paragraph of the clause refers to salvages, recoveries or payments recovered or received subsequent to a “loss settlement.” The word “settlement” in the reinsurance context has a possible spread of meaning, but I accept the syndicates’ submission that it here embraces the establishment or ascertainment of the original assured’s claim (including its quantum) in particular by agreement or admission, or, it seems likely, by judgment or award. This is the same sense in which it is generally understood in the context of a “follow the settlements” provision: see the Scor case [1985] 1 Lloyd’s Rep. 312. It also reappears in this sense in the claims clause in the first two contracts, in article 14 (notification of loss and settlement of claims) in the third contract, and in the aviation settlements clause in the first two contracts. This last clause is a special provision for aviation business which contains no reference to amounts “paid” at all but caters specifically for situations in which a reinsured has to establish a fund to meet a prospective claim (see Boden v. Hussey [1988] 1 Lloyd’s Rep. 423 where the court did not have to focus on the question whether the reinsured would have in the first instance to establish the fund out of his own funds before looking to reinsurers).

Looking at other clauses in the contracts, payment may again either be equated with or treated as an aspect of settlement in the currency clause and the claims clause in the first two contracts. Indeed, there might be some anomalies if the currency clause were read as distinguishing between settlement and payment as different temporal stages: an insurer reaching an agreement with a Frenchman to settle a sterling claim at date A could, if the French franc declined, transfer at date B an equivalent number of francs at date B, which converted into sterling at the rate at date A would give a larger number of pounds than the original settlement. Whilst this sort of remote possibility may be of little assistance in the interpretation of the contracts, it also seems unlikely that the currency clause was concerned to draw a distinction between settlement and payment which would itself appear to be of relevance only on remote occasions.

The claims clause in the first two contracts makes provision for notice of losses which may cause a claim under the reinsurance, and goes on to provide for all loss settlements to be binding upon the reinsurers, provided that they are within the terms and conditions of the original insurance  [*338]  and the reinsurance. It then says that “the reinsurers shall pay the amounts due from them upon reasonable evidence of the amounts paid being given by the reinsured.” The syndicates rely on this as confirming an intention to require satisfaction of the original claim before any obligation arises on the part of reinsurers; and they point out that this particular clause does introduce some form of trigger or condition precedent to reinsurers’ obligation to indemnify, consisting of a requirement to submit certain evidence. Here, it is to be noted, the word “actually” is not used. The wording of the claims clause “shall pay . . . upon reasonable evidence of the amounts paid” is capable in my view of referring to payment in a sense and manner analogous to that in which it appears in the phrase “pay as may be paid thereon” (used in the Eddystone case [1892] 2 Ch. 423 and in article 18 in the third contract) whilst introducing an additional requirement of “reasonable evidence” of such payment. Since both the claims clause and the third contract also include “follow the settlements” provisions, there is the further difference from the position under a simple “pay as may be paid” provision that the syndicates are bound by the reinsured’s settlements; such settlements determine “the amounts due from them [the syndicates]” irrespective of the reinsured’s liability apart from any settlement. The phrase “amounts paid” on this basis connotes the amounts ascertained for payment on the reinsured’s settlements with its original insured. There is a further reason for construing it in this sense. Since it is the settlement, not the disbursement or satisfaction by which the reinsured is bound, one would expect the reinsured to have to make available reasonable evidence of the commitment constituted by the settlement, not—or certainly not only—reasonable evidence of disbursement. Yet on the syndicates’ two-stage analysis, the parties must have provided for the latter alone, and failed to provide for the former at all.

(g) Specific clauses in the third contract

It is convenient to take first article 8 (original aviation loss warranty clause). This is a provision which limits the scope of the reinsurance to circumstances where the “total original incurred aviation loss” is equal to or exceeds $50m. It is not an excess, but a requirement that the total hull and liabilities claims (inclusive of adjustment, defence and representation costs and expenses) under direct insurances at the original level should be of that size. For this purpose the “total original incurred aviation loss” is defined in very expansive terms, providing not merely that it need not have been finally determined and paid, but also that its computation should include amounts “whether . . . paid or promised to be paid or reserved whilst under negotiation or dispute with the original claimant(s)” and that “in the event of no London aviation market reserve, the reserve for the purposes of this clause shall be that amount as advised to the original underwriters as an estimate of their liability by their adjusters, attorneys or solicitors.” This clause does identify payment as a separate matter to a promise to pay, though only for the purpose of making clear that they are both in this context to be equated with a reserve or even an adjusters,’ attorneys’ or solicitors’ estimate. This is clearly a special clause, whether it is tailor made or in a wording in more general use in the  [*339]  aviation market. It assists the syndicates’ case that it speaks of payment in the sense of disbursement and separates this from a promise to pay. This is however in the context of the original insurance level and of claims by original insureds against original insurers, who may not include Charter at all. Further, the language and subject matter of the clause in no way cross-relate to the standard language of the ultimate net loss clause, and it is notable that it does not mention the concept of loss settlement at all. I do not therefore regard the expanded and explicit wording of article 8 as carrying the syndicates’ case on construction of the reinsuring and ultimate net loss clauses very far.

I come now to a most striking feature of the third contract, which has no parallel in the first two contracts. Article 22 (the special cancellation clause) gives either party the right to terminate if the other has become insolvent or unable to pay its debts. It then continues:

“The Reinsurer shall remain liable for losses occurring up to and including the date of termination. Thereafter the liability of the Reinsurers shall cease outright other than so far as outstanding claims are concerned.”

This must, on the syndicates’ case, be a false friend. The message conveyed is that a reinsurer’s liability for outstanding losses remains unchanged despite the reinsured’s insolvency or inability to pay, even if reinsurers terminate cover in respect of future losses and a fortiori if reinsurers do not. It is irrelevant that there was in the present case no termination and that no issue directly arises under the special cancellation clause. What matters is the message conveyed by the clause about the position in case of the reinsured’s insolvency or inability to pay. The syndicates say that the message is true enough—their liability remains unchanged (in the amount presumably of nil in the event of inability to pay moneys reaching the excess point), because it has always been a term that they would not have to pay unless and until the reinsured disbursed or otherwise satisfied its obligations arising from any loss settlement. The syndicates can of course point to situations where insolvency or inability to pay debts in full does not prevent some recovery being made under an excess of loss reinsurance; as Scrutton J. pointed out in the Law Guarantee case [1914] 2 Ch. 617 there are situations where a partial dividend can be paid to all creditors, which is sufficient to enable a partial excess of loss reinsurance recovery which will enable a further partial payment to all creditors, and so on: there will then be an indefinite series of ever diminishing tranches until payment in full or until the amounts involved fall below the limits of legal tender. But I do not think for a moment that this can be accepted as the true significance of the words of article 22. The message implicitly conveyed is in my view straightforward. It is specifically directed to the events of insolvency and inability to pay, and it also suggests that the position in insolvency is no different from the position as it was before. The syndicates’ case under the third contract rests essentially on what they submit to be the general intention of the ultimate net loss clause, not specifically directed to situations of insolvency or inability to pay. The third contract lacks those clauses in the first two on which the syndicates rely to draw a distinction between settlement and payment. Whatever may  [*340]  be the purpose or effect in other situations of the ultimate net loss clause, the third contract in my judgment contains a clear indication that a reinsured who is insolvent or unable to pay his debts is not required to disburse or satisfy the amount of any loss settlement before acquiring a right to reimbursement under this reinsurance. If and in so far as this conflicts with the ultimate net loss clause, this indication is entitled to priority over that clause. It may however be that the situation is not one of conflict but of consistency, if the ultimate net loss clause is to be interpreted in the way in which Charter submits.

The history of the ultimate net loss clause

As a standard form the ultimate net loss clause derives, as the syndicates showed, from the 1930s and it has, with some expansion and amendment, been used in substantially the same form ever since. On the syndicates’ case, it can hardly have been intended as anything other than a deliberate departure from the approach identified with the Eddystone case [1892] 2 Ch. 423 and the Law Guarantee case [1914] 2 Ch. 617. If the clause was there to fulfil not merely its basic role of establishing the measure of indemnity, but also that of a condition precedent requiring satisfaction by the reinsured of the amount involved in any loss settlement, then it might, in this light, have been expected that this would have been stated in very positive terms.

If I look at the origin and history of the drafting of the ultimate net loss clause, as the syndicates invited me to do, then the departure also seems to have been unheralded. I must assume that the extracts put before me by the syndicates from contemporaneous publications of these clauses are complete as far as relevant. So far as appears, nothing whatever was said about any departure from past authority or about the purpose which the syndicates now ascribe to the clause. On the contrary, the revision of the original clause in 1937 to add its second sentence was accompanied by a letter dated 9 December 1937 from Lloyd’s Underwriters Fire and Non-Marine Association referring to three newly worded clauses, the salvage and recovery clause, the ultimate net loss (excess reinsurance) clause—as it was then known—and the costs clause. The titles and wording of the second and third of these clauses with their definitions of “net loss” indicate that they were designed for use in excess of loss reinsurance. The letter simply stated that the existing clauses were to be withdrawn and that:

“Your committee have adopted these new clauses in order to make the position of underwriters more secure in the matter of the proper apportionment of salvages and other recoveries received after the payment of a claim. Owing to the fact that without one of these alternative clauses underwriters might find salvages and recoveries apportionable in a manner they did not intend when writing a direct risk or a reinsurance containing an excess condition, your committee strongly recommend that members should protect themselves by specifying the appropriate one of these clauses in all new risks and in all renewals which contain an excess, particularly in those cases where ‘wording as before’ appears in the slip.”  [*341]

There is no hint in this letter that the purpose of the ultimate net loss clause or of its variant the costs clause was also to ensure (contrary to what must otherwise have been viewed as the position on authority) that excess of loss reinsurers were not liable unless and until their reinsureds had not only crystallised their position in settlement, but also satisfied it by disbursement or otherwise. It is, I think, a not unreasonable inference, if one looks at this material, that the ultimate net loss clause was conceived as a tool for accurate measurement and as a protection for reinsurers of a different nature to that now suggested by the syndicates.

Two further points appear to be indicated by the material and submissions before me. First, the precondition for which the syndicates contend would appear, on their case, to have been introduced and to apply generally in the field of excess of loss reinsurance. A measurement clause is an important feature of an excess of loss reinsurance. But the history of the present ultimate net loss clause and the standard clauses which the syndicates have put before me contain no hint of any alternative form of ultimate net loss clause designed to meet the requirements of insurers arranging excess of loss reinsurance who would need a detailed measurement clause to define “(ultimate) net loss” but who could view the approach of the Eddystone case [1892] 2 Ch. 423 and the Law Guarantee case [1914] 2 Ch. 617 as the only safe or appropriate basis of reinsurance. Secondly, any such precondition would appear to have been confined to excess of loss reinsurance. There is no suggestion of any equivalent clause being introduced or in use in the context of quota share reinsurance to alter the position resulting from the Eddystone and Law Guarantee cases. The syndicates’ submissions may therefore appear to suggest an oddly unbalanced view of acceptable reinsurance arrangements. These two points bring me naturally to the significance which the syndicates attach to the distinction between quota share and excess of loss reinsurance.

The suggested difference between the contexts of proportional and excess of loss reinsurance

The syndicates submit that it is important to remember that the ultimate net loss clause is a clause designed for use in the context of excess of loss reinsurance. Proportional and non-proportional reinsurances are different forms of reinsurance. Under the former the reinsurer shares pro rata in both premiums and losses on risks ceded whether from ground up or on a surplus basis. Under the latter, the premium may be fixed but is commonly (as under the present reinsurances) a minimum and deposit premium adjustable by reference to a percentage of the reinsured’s net premium income: claims on the risk(s) within the scope of the reinsurance are payable in excess of a specified figure, whether the cover relates to individual losses or to an accumulation of losses on the whole or a particular part of the reinsured’s account. The nature of excess of loss reinsurance is that there is a trigger to liability in the form of an excess point. However, this does not mean that disbursement or satisfaction of original claims is either more likely or more appropriate as a condition precedent to liability. Under a proportional agreement in the Eddystone case form reinsurers’ liability arises in respect of ascertained liability on the part of the reinsured in respect of business ceded (see Versicherungs [*342] und Transport A.G. Daugava v. Henderson (1934) 49 Lloyd’s Rep. 252) under a proportional agreement in modern form, it arises in respect of the reinsured’s commitment to pay ascertained by his settlement with the original insured. Under excess of loss reinsurances such as the present, reinsurers’ liability again arises in respect of the commitment to pay ascertained by settlement, but it is necessary to apply the excess point. None of this makes it more likely in commercial terms that the parties would wish, or that it would be acceptable, to introduce a further precondition of pre-payment in excess of loss reinsurance.

The syndicates submit that there is another fundamental difference, in that proportional reinsurance is designed to protect the solvency of reinsureds and to enable them to “borrow” solvency from their reinsurers and so to write more business, whereas (they submit) excess of loss reinsurance has no such purposes. No evidence was put before me to substantiate this surprising final proposition. The different forms of reinsurance represent different tools which an insurer may deploy, frequently in conjunction with another to pass on or protect exposure either on particular risks or on the whole or part of its insurance account. Much modern reinsurance is done on an excess of loss basis. The purpose is self-evidently to protect the insurer from exposures of the type reinsured which could otherwise either individually or collectively imperil the insurer’s solvency or profitability. To suggest that it is natural under an excess of loss reinsurance (as distinct from a proportional agreement) that the insurer should have to fund the entirety of any claim as a precondition to looking to his reinsurers is without warrant in any material before me, or I believe in common sense or experience. If one takes the present reinsurances alone, it is apparent that they cover in layers up to a high level, being for £2á5m. excess of £3m. and for £2á5m. excess of £7á5m. Probably there was an intermediate cover £2á5m. excess of £5á5m. and there may have been covers above £10m. Whether this is so or not, the purpose of the reinsurance must have been to ensure that the reinsured’s exposure in respect of a loss of up to £10m. would be protected to a very substantial extent by its reinsurers. To suggest that it is integral to excess of loss reinsurance that an insurer should operate with a capital base or with banking facilities to enable it to be sure of being able to fund the whole of every loss or accumulation of loss before making any recovery from its reinsurers appears to undermine, rather than represent the purpose of, such reinsurance. Nor do I accept the suggestion that pre-payment would be viewed as essential to exclude the possibility that the reinsured might otherwise seize a cash flow advantage, by recovering from its reinsurers and then holding on to the moneys for a period before disbursing them to the original assured. In the ordinary course original assureds would expect their money as soon as any settlement was reached, and one would expect accounting to take place in a way which would ensure as far as possible that payments and recoveries occurred at more or less the same time.

Commercial considerations

On this basis considerations similar to those which influenced the courts in the Eddystone case [1892] 2 Ch. 423 and the Law Guarantee case  [*343]  [1914] 2 Ch. 617 appear to me to have a part to play in the construction of the present contracts. The suggested precondition of prepayment would be detrimental to the aims and expectations which the parties as reasonable commercial people may be taken to have had in mind when effecting the reinsurances, as well as being capricious and potentially unfair in its operation. Stirling J.’s words in the Eddystone case also reflect the result of giving effect to the syndicates contention. If I may re-express them colloquially, excess of loss reinsurance incorporating the ultimate net loss clause would not serve as a direct mechanism for spreading the load; a reinsured would have first to show that he was by himself capable of lifting its full weight before claiming against his reinsurers, even although collapse might be the result. The fuller reasoning in the Law Guarantee case reinforces the strength of the commercial considerations militating against any such conclusion.

The same commercial considerations also provide a forceful answer to the syndicates’ contention that Charter’s case focuses inappropriately upon Charter’s insolvency and upon supposed hardship arising in that situation. Parties to a reinsurance may have in mind precisely the possibility of financial difficulties, and indeed provide for insolvency as the third contract does. In insurance, as much as (if not more than) in any other business, the matching of exposure and protection to assure both solvency and profitability is absolutely fundamental. Reinsurance—of whatever type—is a principal means to this end. It is unnecessary even to recall some of the recent Lloyd’s litigation in this court to see the problems which may emerge in this regard where exposure and reinsurance protection are not or not fully matched.

I would add that it cannot realistically be suggested, and was not here any more than in the Law Guarantee case, that reinsurance premiums under excess of loss reinsurances containing ultimate net loss clauses are weighted or reduced by reference to the strength or weakness of the reinsured’s cash flow position to take account of the possibility that reinsurers might not have to pay claims because the reinsured was unable to fund them in advance.

Regulatory and accounting considerations

Mr. Kentridge for Charter referred to the solvency and insurance company accounting requirements prepared under the regulations in force when the three reinsurances were written, and, though this cannot be of direct relevance, today. The legislature has repeatedly intervened with tighter and more comprehensive regulation in this important field. At the time of these contracts Charter was by section 32 of the Insurance Companies Act 1982 required to maintain a margin of solvency, defined as “the excess of the value of its assets over the amount of its liabilities, that value being determined in accordance with any applicable valuation regulations.” The regulations in force in 1988-1990 were the Insurance Companies Regulations 1981 (S.I. 1981 No. 1654). Regulation 52(1) required the amount of the liabilities of an insurance company to be determined in accordance with generally accepted accounting concepts, bases and policies or other generally accepted methods appropriate for insurance companies. Generally accepted accounting practice in the United [*344]  Kingdom contemplated (and contemplates) that incoming claims will be shown gross in the company’s revenue account and that reinsurance will be shown as a deduction giving a net cost of claims (cf. the Association of British Insurers’ Statement of Recommended Practice on Accounting for Insurance Business, reprinted January 1987). Regulation 41 provided that the value of any debt due to, or other rights in respect of, an insurance company under any contract of reinsurance should be the amount which could reasonably be expected to be recovered in respect of that debt or right.

The form and content of the annual accounts and returns required under section 17 of the Act of 1982 and submitted to the Department of Trade and Industry were governed by the Insurance Companies (Accounts and Statements) Regulations 1983 (S.I. 1983 No. 1811). The forms specified under Regulations 6 and 9 contemplated the entry of amounts outstanding in respect of incoming claims and amounts recoverable in respect thereof from reinsurers.

If the syndicates’ interpretation of the standard ultimate net loss clause in use in the market is correct, the solvency and accounting position of insurers would appear significantly complicated. Whilst it can in normal circumstances be presumed that a company is a going concern, generally accepted accounting concepts, bases and policies involve considering, where there is room for doubt, whether this basis remains applicable, and, if it is not or may not be, what effect that has on the company’s financial position and how this should be presented or noted. On the syndicates’ case, this must give rise to a potential problem of considerable significance—at least for so long as it continued to be the practice or permissible to use the ultimate net loss clause or to take into account for solvency purposes any potential recoveries under any excess of loss reinsurance including that clause. The syndicates submit that insurance company accountants calculating solvency margins and preparing accounts need simply to “factor into” their consideration, when considering whether the going-concern basis remains appropriate and for what reinsurance recoveries the company can take credit, an assessment of its prospects of being able to fund its incoming claims before looking to its reinsurers. I suspect that would add a hitherto unsuspected peril to the conduct of insurance business and to its regulation and accounting. The moment when an insurance company might fall from a plateau of solvency, when full credit could be taken for potential reinsurance recoveries, into an abyss of insolvency as a result of the operation of the precondition of pre-payment would be unpredictable. Any assessment of the prospects of its occurrence would fluctuate according to the size and volume of new claims and of potential reinsurance recoveries in respect of them. The assessment would have to take into account the prospects of well-disposed reinsurers being prepared to support the company by funding claims (on the syndicates’ case ex gratia) and to continue such support in circumstances where there could be very large financial benefit in an opposite attitude, which might (particularly if reinsurers themselves had any financial problems or were in liquidation or receivership themselves) be hard to resist. The syndicates submit that all this if true could mean at most that there had been a failure by those involved in insurance business,  [*345]  accounting and regulation to appreciate the true nature and effect of the ultimate net loss clause, and to introduce equivalent solvency provisions to those apparently introduced in the United States after the Supreme Court in one case in 1937 accepted the existence of a precondition such as that for which the syndicates contend. It seems to me however that, if it is material to look at the accounting and regulatory background about which insurers and reinsurers in this country must be taken to have general awareness, it reinforces Charter’s submission that the syndicates’ construction is unlikely to have been intended. At the end of the day, however, this is not a decisive factor in the conclusion which I reach.

The suggested aim of securing proper administration of the reinsured’s affairs in circumstances of insolvency

The syndicates submitted that, if protection in insolvency is to be assumed as an aim of these reinsurances, there could be valid commercial reasons why reinsurers should wish to limit their exposure to any amounts paid out by way of dividend to the original assured in the liquidation. Quoting counsels’ skeleton argument, “the interests of the creditors (for whom in effect the liquidator acts) will be in allowing debatable claims so as to maximise reinsurance recoveries,” whereas, “the reinsured’s interest prior to liquidation will have been to scrutinise claims rigorously.” This represents a comforting view of the efficiency in claims handling of companies destined for insolvency, and an apparent slur on the objectivity or competence of the professional liquidators and scheme administrators charged with the handling of insurance company insolvencies. I say apparent; counsel assure me that there was and is no such intention. If the distinction made were accepted, it would not explain why the ultimate net loss clause applies in the same form in all circumstances, before and after any problems of solvency arise. Reinsurers have the same legal protection in both situations, consisting in their right to refuse to follow any settlement not arrived at in a businesslike way: see the Scor case [1985] 1 Lloyd’s Rep. 312. With most reinsureds, the point is in any event likely to have a considerable element of artificiality because they act as following companies, on risks where the settlement decision is in effect taken not by individual companies, but by a leading underwriter or company.

Subrogation in insolvency

Another scenario in liquidation which the syndicates raised concerned the application of principles of subrogation. If incoming claims on which an insolvent insurer pays only a dividend of 10 per cent. entitle the insurer to a full reinsurance recovery, then, the syndicates submit, the reinsurers could lose the benefit of potentially valuable subrogated recoveries. The reinsurance recovery may have been spread (by the mechanism of dividends) across the general body of creditors. If the original insured makes a third party recovery for the whole of his loss, he cannot be called upon to repay the insurer more than the 10 per cent. dividend which he received. The syndicates question whether there would be any mechanism by which other creditors could be required to repay parts of the dividends which they had received, or by which the reinsurers could recover their  [*346]  100 per cent. payment, if the insolvent insurer had not retained sufficient other assets in the liquidation. In principle it seems clear that, if the original assured has (viewing the matter retrospectively) suffered no loss, because he has now made a complete third party recovery, he has no insurance claim on the insolvent insurer, who in turn has no reinsurance claim on the reinsurers, who should be repaid. On this basis, the problem suggested only arises if the original insured is paid a claim by the insurer and later makes a third party recovery, and, at the time of the third party recovery, the insurer has no further or future assets whatever out of which the reinsurers may be repaid. During the course of insolvency, one would expect liquidators and scheme administrators to reserve funds to cater for contingencies and not to distribute all available cash by way of dividends.

I must also say that remote circumstances of this kind postulated by the syndicates do not anyway appear to me a helpful guide to the construction of these reinsurances. They do not explain why it should be thought fair that reinsurers’ liability should, for practical purposes, be restricted to the amount of any dividend, which is likely in normal circumstances to afford them a windfall benefit for which they cannot have covenanted; and they do not of course provide a rationale for the ultimate net loss clause in situations of solvency. In the last analysis, there may be mechanisms for recovery from other creditors of part of the dividends paid, if it ever became necessary, on grounds of unjust enrichment, although I recognise that their pursuit could face practical difficulties. A final consideration is that it does not appear to me, even on the syndicates’ construction, that the ultimate net loss clause could give them the absolute protection which they suggest was intended as regards subrogation recoveries. Suppose that it had not been Charter but an insurer reinsured by Charter (on terms not including any ultimate net loss clause) that had gone into liquidation, and suppose that insurer were to distribute a 10 per cent. dividend to its creditors, and were to make a 100 per cent. recovery from Charter, leading to a full recovery by Charter from the syndicates. If the original assured thereafter made a full third party recovery, precisely the same exposure to possible difficulties in working out the operation of subrogation would apply between the insolvent insurer and Charter and would, if the problems raised by the syndicates have any validity, affect the syndicates notwithstanding the ultimate net loss clause.

Recent English authorities concerning ultimate net loss clauses

I turn to recent authorities in which a similar point to that now before me has been touched on, although in none was it necessary for the court to reach any concluded view. The first is Pine Top Insurance Co. Ltd. v. Unione Italiana Anglo Saxon Reinsurance Co. Ltd. [1987] 1 Lloyd’s Rep. 476, where Gatehouse J. had to consider a clause in an excess of loss reinsurance in respect of medical and other expenses incurred under travel insurances providing: “All terms, clauses and conditions as original. To pay as paid thereon, but subject nevertheless to the terms, clauses and conditions of this reinsurance.” He treated the words “to pay as paid thereon” as by themselves varying “the general rule that the reinsurer is obliged to pay even before his reinsured makes payment, i.e. as soon as  [*347]  the latter’s liability is assessed or capable or being assessed.” I think that this was in error; it does not appear that he was referred to the Eddystone case [1892] 2 Ch. 423 or the Law Guarantee case [1914] 2 Ch. 617. His judgment goes on to say that this variation was “carried through” to a limit of liability clause providing that “reinsurers shall not be liable for any loss under this reinsurance agreement unless the reinsured have actually paid in respect of each and every loss each and every person, an amount of £5,000.” Since there was nothing to “carry through” and since the policy wording is only briefly extracted and is in quite different terms to the present, I find the dicta in this case of no assistance.

Home and Overseas Insurance Co. Ltd. v. Mentor Insurance Co. (U.K.) Ltd. [1990] 1 W.L.R. 153 simply decided that the dispute under the particular reinsurance contract before the court should be referred to arbitration, where it would be determined by a tribunal expressly enjoined to interpret the agreement as an honourable engagement and to effect its general purpose in a reasonable manner rather than in accordance with a literal interpretation of the language. There are dicta of Hirst J. at first instance generally unfavourable to reinsurers’ present submissions, culminating with the statement that he would only have been prepared to uphold their construction if the words of the agreement as a whole plainly excluded any other meaning, which in his judgment was not by any means clearly the case. Mr. Kentridge asked me to note these words, and to follow a similar approach. The judgments in the Court of Appeal are in more guarded terms, and do no more than conclude that the question of construction was a serious, difficult and important one, which deserved mature consideration before the appropriate (arbitration) tribunal. Mr. Kentridge says that, if the point was difficult, that alone shows that there is a choice of meanings and that Charter’s construction should be preferred on grounds of policy and reasonableness. But this appears to me to be reading too much into the Court of Appeal’s decision which was clearly designed to leave the whole matter open for the arbitrators. Finally, there is In re A Company No. 0013734 of 1991 [1992] 2 Lloyd’s Rep. 415 the statements in which in my view also carry matters no further.

“Pay to be paid” clauses in Protection and Indemnity Association Rules

At the time of the Court of Appeal’s decision in the Home and Overseas Insurance case, the cases of Firma C-Trade S.A. v. Newcastle Protection and Indemnity Association [1989] 1 Lloyd’s Rep. 239 and Socony Mobil Oil Co. Inc. v. West of England Ship Owners Mutual Insurance Association (London) Ltd. (No. 2) [1990] 2 Lloyd’s Rep. 191 had been decided in the Court of Appeal, but not yet in the House of Lords. They were referred to by Parker L.J. in the Home and Overseas Insurance case as “considerable reinforcement” for reinsurers’ arguments there. The syndicates rely on the House of Lords decision as still stronger reinforcement. The Firma C-Trade and Socony cases concern a very different branch of insurance law and practice to the present. Under the mutual scheme embodied in the club rules which were there in question, the member was entitled to be indemnified against or in respect of claims, in the one case, “which he shall have become liable to pay and shall in fact have paid” for loss or damages to property carried on board an  [*348]  entered ship and, in the other case “which they as owners of the entered vessel shall have become liable to pay and shall have in fact paid.” The requirement of payment “in fact” was introduced as an essential element of entitlement to indemnity. It appears to have been accepted that the ordinary and natural interpretation of each rule was to make payment to the third party claimant a condition precedent to payment by the clubs to their members; the only contrary argument (raised in the Court of Appeal and developed in the House of Lords) was that it should be read subject to equitable principles or to an equitable jurisdiction to require the club to pay the third party direct or to put the member in funds to do so. This argument foundered on the clear language of the rule: see per Lord Brandon, at p. 197, and per Lord Goff, at pp. 200-202. In the present case, the issue concerns the proper construction of the ultimate net loss clause and the other clauses in the relevant contracts; the wording (as indeed the context) are different, and the provisions on which the syndicates place greatest reliance do not appear in the basic indemnity clauses, but in other clauses, the most obvious purpose of which is to measure the quantum of reinsurance recovery, rather than to introduce a condition precedent to such recovery. As to the context, in Ventouris v. Mountain [1992] 2 Lloyd’s Rep. 281, 298 Hirst J. referred to a “pay to be paid” rule of the nature identified in the Firma C-Trade case [1991] 2 A.C. 1 and the Socony case [1990] 2 Lloyd’s Rep. 191 as “entirely inappropriate in the non-club environment of a commercial insurance contract” such as that before him. However, Mr. Ruttle drew to my attention to collision liability clauses in various of the current institute clauses, which appear to be “pay to be paid” clauses, although the implications of that view do not seem to have been considered in any recent authority or in the light of The Fanti and the Socony case. Collision liability clauses in marine insurance themselves occupy a fairly special area, and the wording relied on was, as in The Fanti and the Socony case, part of the main indemnity provision. I do not gain any real assistance from the appearance of these clauses in the standard forms designed for use in direct contracts of marine insurance of hulls, freight and other interest.

American authorities

I was also referred by Mr. Ruttle to three American authorities on reinsurance of insolvent insurers. In the first, Allemannia Fire Insurance Co. of Pittsburgh v. Firemen’s Insurance Co. of Baltimore (1908) 209 U.S. 326, 328 a proportional contract “to reinsure” the reinsured provided by clause 11 that each risk insured should be subject to the same conditions as the original contract reinsured, that losses should be payable pro rata with, in the same manner, and upon the same terms and conditions as paid by the reinsured:

“and in no event shall this company be liable for an amount in excess of a ratable proportion of the sum actually paid to the assured or reinsured by the said reinsured company under its contracts hereunder insured, after deducting therefrom any and all liability of other reinsurers of said contracts or any part thereof.”  [*349]

There was also a provision for the reinsured to forward a proof of loss and claim “after said reinsured company shall have adjusted, accepted proofs of, or paid such loss or damage.” The Supreme Court rejected the submission that clause 11 was intended to introduce a requirement of prepayment before recovery as “to utterly subvert the original meaning of the term reinsurance and to deprive the contract of its chief value” (p. 336). The reinsurers had failed to show anything in the special provisions of the particular reinsurance to change the ordinary rule that a reinsured could recover without showing that he had paid the original loss. The purpose of the clause cited was, in effect, to quantify the amount payable by reinsurers.

The second case, Fidelity & Deposit Co. v. Pink (1937) 302 U.S. 224, involved a proportional contract providing that “the reinsurer does hereby reinsure against loss” and subject to conditions and provisions, including in section 4 (see p. 228):

“The reinsurer’s proportionate share of a loss under the bond, of costs and expenses as hereinafter defined, and of interest, shall be paid to the reinsured upon proof of the payment of such items by the reinsured, and upon delivery to the reinsured of copies of all essential documents concerned with such loss and costs and the payment thereof. The reinsured may, however, give the reinsurer written notice of its intention to pay the loss on a certain date, and may require the reinsurer to have its share of such loss in the hands of the reinsured by such date: provided, however, that the reinsurer in any event shall have a period of 48 hours, after the receipt of such written notice from the reinsured, to mail or otherwise despatch its payment; and provided further that in any such case the reinsurer, if it desires to do so, may pay its share of the loss by means of a check drawn in favour of the obligee of the bond.”

The court said, at p. 229:

“We do not question the general rules concerning liability of reinsurers announced in the Allemannia case; but the liability under any written contract must be determined upon consideration of the words employed, read in the light of attending circumstances. Here the two insurance companies stood upon an equal footing; both were experts in the field. The language used differs materially from that found in the policy of the Allemannia Company. There is no ambiguity and no circumstance requires disregard of the ordinary meaning of the language.”

Further, at p. 230:

“As the standard form of 1930 that was adopted twenty years after the Allemannia case it fairly may be assumed that the dissimilar language employed was intended to impose liability different from the one there found to exist.”

The Supreme Court’s decision in the Pink case must be viewed as depending on the very specific language of section 4. This was clearly designed to introduce a condition precedent of disbursement of money by  [*350]  the reinsured, and it went on expressly to cater for the problems of cash flow which this could involve by providing as an alternative a mechanism by which the reinsured might give notice of its intention to pay the loss on a certain date and to require the reinsurer to have its share of the loss in the reinsured’s hands before that date. What it did not do was cater for the situation of insolvency.

Finally, in Stickel v. Excess Insurance Co. of America (1939) 23 N.E. 2d 839 the Supreme Court of Ohio had to consider an excess of loss reinsurance covering “any amounts of ultimate net loss which the company may pay in excess of the first $5,000 on account if any one person injured or killed . . .” By section II (see p. 841):

“The term ‘ultimate net loss’ shall be understood to mean and shall mean the sum actually paid in cash in settlement of losses for which the company is liable, after making proper deductions.”

The court regarded the Pink case as more applicable than the Allemannia case to the particular wording before it, emphasising that “[i]t would seem superfluous to cite additional authorities for, after all, the decisions turn largely on the language of the particular contract under examination in the particular case.” The introduction of the phrase “in cash” in the ultimate net clause in the Stickel case showed clearly that the clause was concerned to introduce a requirement of disbursement of money.

The Allemannia case adopts a view of the ordinary objects and consequences of reinsurance not dissimilar to, though even more robustly expressed than, the judgments in the Eddystone case [1892] 2 Ch. 423 and the Law Guarantee case [1914] 2 Ch. 617. If I were to look to these three cases for guidance in the resolution of the present dispute, I would view the Pink and Stickel cases as special cases which turn on their particular wordings, which are distinguishable from the present. Every wording must, as these cases emphasised, turn on its own language.

Conclusion

Adopting an approach to construction which is neither uncompromisingly literal nor unswervingly purposive, I consider that the words of the ultimate net loss clause in all three contracts were intended to ensure that reinsurance recoveries were at each stage and ultimately measured in all conceivable circumstances by reference to the precise net amount of the commitment involved in the settlements achieved between the reinsured and the original assured; and that they did not have an additional purpose of introducing a temporal precondition to recovery in the form of disbursement or other satisfaction of any such amount. The precondition which the syndicates advocate could be achieved by clear words clearly designed for such a purpose, if that were acceptable to the parties. Putting the matter at its very lowest, the words of the ultimate net loss clause were not and are not in my view clearly designed for this purpose, Nor are the words of the contracts taken as a whole. On the contrary, in my judgment, they were designed, and are to be read as designed, for the different purpose of measurement. Indeed, the third contract contains a clause, article 22, which I regard as very specifically confirming that no such precondition can there have been envisaged, and as clear enough by  [*351]  itself to prevail there over any precondition in the ultimate net loss clause, had I thought that any was thereby intended. As it is, I regard article 22 in the third contract as confirmation, in the context of that particular contract, that the ultimate net loss clause is not to be read as involving any precondition in the form of disbursement or other satisfaction of amounts ascertained as due under loss settlements. I reach these conclusions upon consideration of each of the three reinsurance wordings before me and having regard to the general aims and consequences which these parties may be taken to have had in mind. In the latter connection, I have made clear my view that the commercial considerations identified in the Eddystone and Law Guarantee cases have considerable relevance. Other factors which support the same construction are (1) the history of the ultimate net loss clause, put before me by the syndicates, and (2) regulatory and accounting considerations. Each appears to me to represent part of the commercial background of which account may also legitimately be taken, but they do no more than confirm a conclusion at which I would anyway have arrived.

Construction of the contracts in situations of insolvency

Charter submitted that, if (contrary to the view just expressed) the contracts were to be read as involving a precondition of disbursement or satisfaction in situations of solvency, they should nonetheless be differently construed in circumstances of insolvency. The justifications submitted for this were (a) the unjust, uncommercial and arbitrary result of such a precondition in insolvency, and (b) the consideration that such a precondition would be contrary to public policy and void under the principle in British Eagle International Airlines Ltd. v. Compagnie Nationale Air France [1975] 1 W.L.R. 758. As to the first, if the precondition exists in solvent situations despite the evident problems which it could pose for solvency, it seems to me difficult to see how it is to be necessarily implied, in the absence of any positive provision, that the precondition is not to apply in insolvency. That is so at all events under the first two contracts. There could be an argument for such an implication under the third contract, in view of the provisions of article 22. The second justification appears to me circular. The principle in the British Eagle case applies when and if a contract on its true construction is repugnant to principles of insolvency law, because for example it purports to secure a party a greater advantage in bankruptcy or liquidation than previously or provides for the distribution of what are, at least in substance, assets otherwise than pari passu. It is not itself a canon of construction or means of altering the true construction of a contract in insolvency.

Alternatively, Mr. Kentridge submitted that the principle in the British Eagle case was directly applicable in view of the appointment of provisional liquidators in respect of Charter by order of Blackburn J. dated 23 June 1994. The effect of the syndicates’ construction of the contracts was in his submission to require Charter to meet its liabilities in a manner other than pari passu if it was to get in its assets, contrary to the compulsory scheme applicable in liquidation and contrary, he submitted, to a strong public interest (cf. the dictum of Kennedy L.J. in the Law Guarantee case [1914] 2 Ch. 617, 639). Mr. Hildyard’s response is  [*352]  twofold: that Charter is not in winding up and the principle can have no application, and that, even if Charter were in winding up, the principle would not apply; any asset was and is, in his submission, “flawed” in that its realisation is subject to a precondition which cannot be fulfilled in insolvency; he says that its diminished economic worth in circumstances where the reinsured cannot pay its claims in full is irrelevant.

The order gives the provisional liquidators powers, inter alia, to get in all the company’s property and assets and to administer and settle such claims either by or against Charter as advised in the interests of creditors, as well as to prepare and consider a scheme of arrangement. I understand Mr. Kentridge to accept that the power to administer and settle claims would not embrace the pari passu distribution of assets in relation to claims, which would be a matter for the liquidators after a winding up order. This insolvency may well follow the pattern of other recent insurance company insolvencies and lead not to winding up but to a scheme of arrangement. I see the force of Mr. Hildyard’s submission that the British Eagle case can have no relevance unless and until there is a winding up order. However, in the view which I take on construction, it is unnecessary for me to go further into this or any issue concerning the scope and application of the British Eagle case and I shall not do so.

The effect of insolvency as payment

Finally, Mr. Kentridge raised very shortly the possibility that the making of a winding up order or the admission of a proof thereunder operates as a notional discharge of liability, which would meet any precondition of disbursement or satisfaction to be found in the contracts. Mr. Kentridge referred to the reasoning of Neville J. in In re Law Guarantee Trust and Accident Society; Godson’s claim [1915] 1 Ch. 340, but found difficulty in supporting this on the authorities, including Scrutton J.’s judgment in the Law Guarantee case, cited by Neville J. The judgment of Oliver J. in In re Dynamics Corporation of America [1976] 1 W.L.R. 757 is on this point also of potential interest. Oliver J. was dealing with the question at what date a foreign currency claim should be converted into sterling when the defendant is in liquidation, having regard to the rule established by Miliangos v. George Frank (Textiles) Ltd. [1976] A.C. 443, that judgment may be entered for a sum in foreign currency or the sterling equivalent at the date of payment, instead of (as previously) in sterling converted at the date of accrual of the cause of action. His judgment includes this passage, at pp. 774-775:

“What the court is seeking to do in a winding up is to ascertain the liabilities of the company at a particular date and to distribute the available assets as at that date pro rata according to the amounts of those liabilities. In practice the process cannot be immediate, but notionally I think it is, and, as it seems to me, it has to be treated as if it were, although subsequent events can be taken into account in quantifying what the liabilities were at the relevant date. In the context of liquidation, therefore, the relevant date for the ascertainment of the amount of liability is the notional date of discharge of that liability, and . . . that date must, in my judgment, be the same for all creditors and it must be ‘the date of payment’ for  [*353] the purposes of any judgment which has been entered for the sterling equivalent at the date of payment of a sum expressed in foreign currency.”

It is clear that a different statutory scheme takes over on winding up (cf. also In re Lines Bros. Ltd. [1983] Ch. 1), but, if the present contracts are otherwise to be construed as including a precondition of prior disbursement, a “notional” discharge may not suffice. That no winding up order has been made anyway poses an immediate obstacle to Charter’s submissions. Once again it is unnecessary for me to go further into, or to express any concluded views on, these matters.

Declarations

For the reasons given I shall declare that the contracts do not contain the precondition of prepayment for which the syndicates contend, and grant Charter such further relief as may be appropriate after hearing counsel.

Declaration accordingly.

The reinsurers appealed.

Court of Appeal

25 October. The following judgments were handed down.

STAUGHTON L.J.  The appeal of Mr. Fagan and his Names (“the reinsurers”) raises issues of general interest for the insurance market and in the law of contract generally. To what extent should one depart from the plain meaning of the words in a contract in order to avoid a result which is unreasonable, or even absurd? Who is to be the judge of reasonableness? And on what material is it to be decided? These issues are also of some difficulty. One of them involves a choice between conflicting decisions of the United States Supreme Court.

Charter Reinsurance Co. Ltd. (“the insurers”) are now in provisional liquidation. They formerly carried on the business of insurance and reinsurance. In the course of it they concluded three contracts of reinsurance with the reinsurers: (1) on their 1989 whole account, for £2m. in excess of £3m. each and every loss and/or catastrophe and/or calamity and/or occurrence and/or series of occurrences arising out of one event; (2) on the same risk, for £2á5m. in excess of £7á5m.; (3) on aviation risks,  [*354]  for £2á5m. in excess of £28m. each and every loss (as defined). There were also limits in U.S. and Canadian dollars, which I need not mention. In each of the contracts there was a clause saying that it covered “losses occurring” during a period, that is to say the calendar year 1989 for the two whole account contracts and the calendar year 1990 for the aviation contract.

Losses have arisen on business underwritten by the insurers. In particular, the losses with the names of Exxon Valdez, Phillips Petroleum and Australian earthquakes attach to the insurers’ whole account; and India Airlines and C.A.A.C. Airways attach to their aviation account. We are asked to assume that claims have been presented to the insurers in respect of those losses, that they have agreed their liability in respect of those claims, but by reason of inability to pay debts as and when they fall due they are at present not able to pay those claims.

In case it be thought that misfortune has fallen only upon the insurers, I would add that Mr. Fagan’s two syndicates ceased underwriting in 1990 and are now under run-off management.

The question is whether the reinsurers are liable to indemnify the insurers in respect of sums which the insurers have not yet paid, either by a transfer of funds or by a settlement in account, although the amount of the loss has been agreed. In an agreed statement of facts there is this cri de coeur:

“Both [the insurers and the reinsurers] recognise that the ‘actually paid’ issue is a London market-wide problem. The three contracts at issue in these proceedings are but examples of one type of excess of loss reinsurance contract wording in wide usage in the reinsurance market. Both parties wish this issue to be determined for the benefit of the reinsurance market as a whole.”

Mance J. decided the question in favour of the insurers; and the reinsurers now appeal. I should say at once that the argument on the appeal has evidently been more closely confined than it was before the judge.

The terms of the contracts

The two whole-account contracts contained the same clauses that are material. [His Lordship then set out the reinsuring clause, liability clause, ultimate net loss clause, currency clause and claims clause, see ante, pp. 319B-320C, 321B-D, and continued:] The third contract, relating to aviation risks, was in somewhat different terms, so far as material. [His Lordship then set out articles 7 to 10, 15, 18 and 22, see ante, pp. 322A-323D, 324A, D-325B and continued:]

Approach to interpretation of the contracts

Like the judge I take this topic first. It is now well settled that no contract is to be construed in a vacuum. Although we may start with the literal meaning of the words, they should not be considered in isolation, but rather in the light of the surrounding circumstances (or, if one prefers it, the background, context or matrix) when the contract was made. That process may even, on occasion, lead one to reject the literal meaning of the words and to adopt some other interpretation.  [*355]

So too it may be appropriate to consider whether the literal meaning of the words leads to a result that is unreasonable or even absurd. The classic exposition of that doctrine is, in my opinion, to be found in the speech of Lord Reid in Wickman Machine Tool Sales Ltd. v. L. Schuler A.G. [1974] A.C. 235, 251:

“The fact that a particular construction leads to a very unreasonable result must be a relevant consideration. The more unreasonable the result the more unlikely it is that the parties can have intended it, and if they do intend it the more necessary it is that they shall make that intention abundantly clear.”

That short passage comprehends almost everything that needs to be said on this topic. It shows (1) that there is in general, and apart from consumer protection legislation, no law against people making unreasonable contracts if they wish; (2) that whether they have done so is to be decided by ascertaining their intention (which of course has to be found in the language they used read in the light of the surrounding circumstances); and (3) that it is a matter of degree in two respects—the more unreasonable the result, the clearer the language needed.

We are also to have regard to the object, purpose, genesis or aim of the transaction, viewed objectively and ascertained from the factual background: Utica City National Bank v. Gunn (1918) 222 N.Y. 204, 208 (Cardozo J.), Prenn v. Simmonds [1971] 1 W.L.R. 1381, 1385 (Lord Wilberforce) and Reardon Smith Line Ltd. v. Yngvar Hansen-Tangen (trading as H.E. Hansen-Tangen) [1976] 1 W.L.R. 989, 995 (Lord Wilberforce). Lord Diplock deprecated the use of the word “purposive” in connection with the interpretation of private contracts as opposed to statutes: Antaios Compania Naviera S.A. v. Salen Rederierna A.B. (The Antaios) [1985] A.C. 191, 200. Parker L.J. in Home and Overseas Insurance Co. Ltd. v. Mentor Insurance Co. (U.K.) Ltd. [1990] 1 W.L.R. 153, 162 questioned what Lord Diplock’s objection was. For my part, however, I can see that interpreting the language of Parliament in order to achieve its manifest objective is a somewhat different process from the interpretation of a private contract. That view is, I think, supported by a passage from Palm Shipping Inc. v. Kuwait Petroleum Corporation [1988] 1 Lloyd’s Rep. 500 quoted below.

It is not difficult to find instances where the literal meaning of words has been modified or even rejected, by reason of the surrounding circumstances, or the absurdity of the results which would otherwise ensue, or both considerations. Thus in the Wickman Machine Tool Sales case [1974] A.C. 235 itself it was held that the words “it shall be a condition of this agreement” did not mean what lawyers would ordinarily understand by a condition. By contrast in The Antaios [1985] A.C. 191 it was held that the words “on any breach of this charterparty” did not mean any breach, but any repudiatory breach. A good example is to be found in Segovia Compagnia Naviera S.A. v. R. Pagnan & Fratelli [1977] 1 Lloyd’s Rep. 343. There a charterparty provided that the charterers might order the vessel “to any United States port east of Panama Canal.” It so happens that all ports in the U.S. Gulf and most or all of those in  [*356]  Florida are west of the Panama Canal. Nevertheless they were held to be within the charterers’ range of choice.

A recent example quoted by Mance J. is Arbuthnott v. Fagan (unreported), 30 July 1993; Court of Appeal (Civil Division) Transcript No. 1024 of 1993. A claim against agents for negligent underwriting was held not to arise out of or be in any way connected with the agents’ requirement to be put in funds to meet insurance claims.

Mance J. thought that considerations of reasonableness were of particular relevance in a commercial context; but I am not sure that I would agree with him on that. A contract between a house-owner and his builder or decorator is just as much to be construed in accordance with Lord Reid’s formula as one between mercantile concerns operating on a shipping or commodity exchange. One often reads in the cases of commercial common sense, or business common sense; but that is not some arcane substance of a special and unusual nature. It is common sense or reasonableness in the context of the business in question.

So in this case we must first consider what the contracts say on the literal meaning of their wording, and how clearly they say it; and then whether that meaning would be unreasonable, and if so how unreasonable. (I do not of course exclude surrounding circumstances.) For the proposition that it is legitimate to start with the literal meaning I can, as the judge did, refer to what Saville J. said, quoting Lord Goff of Chieveley, in Palm Shipping Inc. v. Kuwait Petroleum Corporation [1988] 1 Lloyd’s Rep. 500, 502:

“The starting point must be the words and phrases the parties have chosen to use. It is not a permissible method of construction to propound a general or generally accepted principle for sharing the risk of delay between owners and charterers or seeking in the abstract to determine a reasonable allocation of risk of delay and then (to use the words of Lord Goff in Société Anonyme Marocaine de l’Industrie du Raffinage v. Notos Maritime Corporation (The Notos) [1987] 1 Lloyd’s Rep. 503, 506) to seek to force the provisions of the charter into the straitjacket of that principle or into that concept of reasonableness. To do so is to rewrite the bargain that the parties must be taken to have made by the words that they have chosen to use.”

The language used in the contracts

The whole account contracts are in somewhat different terms from the aviation contract. Yet I doubt if the insurance industry will be altogether pleased if we hold that this leads to different results. So are we to reach an interpretation on one of them first, and then hold that it must apply to the others? And if so, which should we begin with? This illustrates a perennial problem in the interpretations of contracts commonly used in a particular branch of commerce: one is referred to decisions on other wording which is different but not very different, and then asked to proceed from that to the contract in question. Yet, if there is a difference, it may be critical. I cannot escape the conclusion that in this case we must give independent consideration to the different forms of contract.  [*357]

(A) The whole account contracts

The liability clause seems to me as clear as can be. It says that the reinsurers shall only be liable if and when the ultimate net loss exceeds £3m. in one case, and £7á5m. in the other. It follows as the night the day that until that event happens the reinsurers are not liable.

One then turns to the ultimate net loss clause. At first there is a slight problem—it refers to “The term ‘net loss.’ “ Those words are not used elsewhere on their own. However, I can see why it was thought necessary to define them. The net loss is to be the amount paid by the insurers plus expenses but less recoveries. Then it is said in the second sentence that recoveries made (sc. by the insurers) after the reinsurers have paid shall, in effect, go to the credit of the reinsurers. That necessarily entails that the reinsurers may have been called upon to pay before recoveries are received. That implication is confirmed in the third sentence or proviso. So the insurers must be entitled to payment as soon as the net loss exceeds the lower limit in the contract, even if the final net loss remains to be ascertained.

The remaining problem is that this interpretation gives a different meaning to ultimate net loss in the liability clause and ultimate net loss in the proviso or third sentence of the ultimate net loss clause. In the earlier place it means the net loss as existing at any given moment; in the later, the net loss when all has been resolved. I would be shocked by such inconsistency in a document drafted in Chancery; but I am not sure that we are entitled to expect such a high standard in an insurance contract. And as it happens we do know that the proviso was drafted by a different hand, or at any rate at a different time, from the rest of the clause (cf. the “collage of clauses” in Bremer Handelsgesellschaft m.b.H. v. Vanden Avenne-Izegem P.V.B.A. [1978] 2 Lloyd’s Rep. 109, 113).

Once that point is resolved, the combined effect of the liability clause and the ultimate net loss clause is to my mind perfectly clear: the reinsurers are to be liable only for sums actually paid by the insurers. The word “actually” means, in that context, in fact or in reality. It must have been added so that there could be no argument about sums which were payable, or deemed to be paid, or liable to be paid.

Mr. Kentridge for the insurers argues that the ultimate net loss clause is concerned with amount, and with ensuring that the reinsurers do not, for example, pay 100 per cent. of a claim when the insurers have settled it for 60 per cent. I would agree that one of the matters that it deals with is amount. But the second sentence and the proviso are clearly concerned with timing; and so, beyond all doubt, is the Liability clause.

In those circumstances it is unnecessary to consider whether the currency clause and the claims clause add further support to the conclusion that the reinsurers are only liable for sums that have been paid. But in my opinion the claims clause does to a significant extent, since it requires “reasonable evidence of the amounts paid.”

(B) The aviation contract

There are a number of points to note. (i) The reinsurance is to “indemnify the reinsured . . . for all losses sustained.” (ii) But they must be “in excess of an ultimate net loss of £28,000,000.” (iii) The original  [*358]  aviation loss warranty clause is somewhat obscure to an untutored eye. But it does not seem to me to bear on the present problem. (iv) The ultimate net loss clause is for practical purposes the same as that in the two whole account contracts. It presents the same problem that the words “ultimate net loss” refer in one place to the final figure but must in the reinsuring clause include an interim figure. Apart from that the wording is plain; once more there is express reference to “the sum actually paid.” (v) The aviation contract does not have the requirement in the whole account contracts that “reasonable evidence of the amounts paid” should be given by the insurers. Instead the reinsurance clause says that the reinsurance “shall pay as may be paid” by the insurers. That phrase has been held not to impose a requirement that the insurers shall have paid before the reinsurers are liable. (vi) The special cancellation clauses give either party the right to terminate upon a number of events, of which one is the insolvency of the other party or inability to pay its debts. In the event of termination the reinsurers “shall remain liable for losses occurring up to and including the date of termination.” Mance J. attached importance to this clause; and he described it as “a false friend” in so far as it was relied on by the reinsurers. However, it is important to identify the “losses,” which are referred to in the clause as occurring “up to and including the date of termination.” These must in my view be the losses of the original insured persons, which have to occur during the calendar year of the relevant reinsurance contract. (In the whole account contracts there is express reference in the reinsuring clause to “losses howsoever and whensoever arising during the period of this reinsurance.” As Mr. Pollock pointed out, if the “losses” there referred to were those of the insurers when they agreed a claim and became liable to pay it, none of that happened during 1989, and for that reason alone the reinsurers would be under no liability. The judge was not inclined to accept this argument.) I do not see that the special cancellation clauses are in any way inconsistent with the reinsurers’ interpretation of the aviation contract: they are not to be liable if the original insured’s loss, albeit occurring during the calendar year 1990, was after the reinsurance contract had been terminated for any of the reasons set out; but they remain liable in other cases, provided of course that the net loss reaches the minimum payment level to engage the reinsurers’ liability; and they cease to be liable when it exceeds the maximum. Whether the loss has reached the minimum payment level depends upon the words “actually paid.” It must in my opinion have been contemplated that termination under article 22 would take place during the period of reinsurance, and not afterwards: the provision for adjustment of premium shows that. So “losses occurring before termination” will very likely have to be the losses of the original insured rather than of the insurers.

There was little detailed argument before us by either party aimed at distinguishing the aviation contract from the two whole account contracts. That is understandable. Each sought to succeed on all three contracts. In my opinion the aviation contract, on its literal wording, does provide that the reinsurers’ liability only arises when the insurers have paid, just as the whole account contracts so provide. But the aviation contract is less clear on that point. If I am asked how much less clear, I can only answer:  [*359]  somewhat less clear. That, as it seems to me, does not necessarily amount to an ambiguity; but I do not think that the presence or absence of an ambiguity is the determinative factor in Lord Reid’s test.

Reasonableness or absurdity

Judges in the past have on occasion thought it unreasonable that a reinsurer should not be liable if the primary insurer does not and cannot pay the claim. In re Eddystone Marine Insurance Co; Ex parte Western Insurance Co. [1892] 2 Ch. 423 was concerned with a reinsurance contract which was “to pay as may be paid” by the primary insurer. Both companies went into liquidation. Stirling J. said, at p. 427:

“Now, a main object of reinsurance is to relieve the reinsured from a portion of the risk previously undertaken by him; and the result of giving effect to the liquidator’s contention would be that, before the reinsured obtains the benefit of his reinsurance, he must himself have paid on the original insurance, even though bankruptcy might be the result. I think that this could not be intended, and that such a construction ought not to be put on the language of the policy unless it is clearly called for. In my opinion the words do not clearly require to be so construed.”

In re Law Guarantee Trust and Accident Society Ltd.; Liverpool Mortgage Insurance Co.’s Case [1914] 2 Ch. 617 was concerned with a contract between a guarantee society (which later went into liquidation) and a mortgage insurance company. The contract in question provided, p. 620:

“The Liverpool Mortgage Insurance Co. Ltd. . . . hereby guarantee the within named Law Guarantee and Trust Society Ltd. . . . to the extent of £5,000 being two-elevenths of the risk assured by, and subject to, the conditions of the within policy of debenture insurance, and also the like proportion of all costs and expenses incurred by the society. . .”

Buckley and Kennedy L.JJ. and Scrutton J. all referred to the Eddystone case [1892] 2 Ch. 423 with approval. As to reasonableness or the lack of it, Buckley L.J. said [1914] 2 Ch. 617, 634:

“I am anxious to point out some consequences which would ensue if the view I have expressed were not the right one. The contrary view is one which makes it to the interest of the company that the society should be insolvent. For to such extent as the society cannot pay in full the argument is that the company are not liable to make payment. It is obvious that such a consideration cannot have entered into the contemplation of the parties in fixing the premium. The company has received a certain premium upon the terms that in an event it shall pay a certain sum. If its liability is reduced to 10s. in the pound, it has received payment of premium as the price of an obligation to pay 20s., but is, by reason of circumstances not material at all so far as the company is concerned, relieved of one half of the liability. If this were true a society whose credit was bad ought to pay a less premium than a society whose credit was good, because the obligation would  [*360]  in the former case result in a smaller liability. The fact is, I suppose, that as matter of business if the credit of the society were bad a larger, not a less, premium would be demanded as the price of the guarantee.”

It is to be observed that both the Eddystone case [1892] 2 Ch. 423 and the Law Guarantee case [1914] 2 Ch. 617 were concerned with reinsurance against a share of the original loss, whereas here we are concerned with reinsurance of the excess above a given figure. That might well make a difference on a question of interpretation of the contracts. But it is not a relevant distinction on the question presently under consideration, unreasonableness or absurdity.

The Eddystone case was again cited in British Dominions General Insurance Co. Ltd. v. Duder [1915] 2 K.B. 394 and Versicherungs und Transport A.G. Daugava v. Henderson (1934) 49 Ll.L.R. 252. But I can find no assistance in those cases as to what is or is not unreasonable. In the latter case Scrutton L.J. began his judgment, at p. 253: “This appeal raises a very short point of construction on very short agreed facts, and has been dealt with by Roche J. in a very short judgment.”

Firma C-Trade S.A. v. Newcastle Protection and Indemnity Association (The Fanti); Socony Mobil Oil Co. Inc. v. West of England Shipowners Mutual Insurance Association (London) Ltd. (No. 2) (The Padre Island) [1991] 2 A.C. 1 sheds a different light on reasonableness. It was concerned with the liability of mutual protection and indemnity associations, which afford cover to their members against liabilities and expenses under various heads in the business of shipowning. The relevant contractual provision for the Fanti was:

“4. The member shall be protected and indemnified against all or any of the following claims and expenses which he shall have become liable to pay and shall in fact have paid . . . (q) For loss or damage caused to . . . property . . . carried on board a ship entered in this class . . .”

The wording of another protection and indemnity association, in which the Padre Island was entered, was not materially different.

There is little discussion in the speeches of the House of Lords as to whether the rules in question were or were not unreasonable, since the wording was clear. Lord Goff of Chieveley said, at p. 36:

“if the condition is in sufficiently clear terms, it would be contrary to principle that equity should grant specific performance of the contract inconsistently with the terms of the contract. In the present cases, the condition of prior payment is perfectly clear. The clubs are only bound to indemnify a member against claims or expenses which he shall become liable to pay and shall in fact have paid. The reason why this provision was included in the rules seems to me to be immaterial. I should record that Mr. Boyd, for the West of England Club, put forward a rather different reason for its inclusion from that proposed by Mr. Sumption. He suggested that, in a mutual insurance association such as a P. & I. Club, it is essential that members should be able to assume the financial probity of other members, because all  [*350]  of them are insurers as well as insured. To that end, it is customary to require each member to discharge his own liability before he can be indemnified against it by the club. Each member is, after all, running his own business: it is up to him to make sure that a claim against him is well founded, and the best way of ensuring that is to require him first to pay the claim before seeking indemnity from the club. I must confess that I was much attracted by this submission. Your Lordships do not however have to choose between Mr. Sumption’s and Mr. Boyd’s submissions on this point, especially as it is not inconceivable that they are both correct. For the fact remains that the rules provide unambiguously that there is no present obligation on the club to indemnify the member unless the condition of prior payment has been fulfilled; and for equity to grant specific performance of the contract, inconsistently with that condition, would, as I have said, be contrary to principle.”

The reason proposed by Mr. Sumption was, I think, “to prevent a member from making a profit from his insurance cover by receiving payment from the club but failing to pay the third party:” see pp. 33-34.

What I find of some significance for present purposes is that shipowners had for a great many years been content to contract with protection and indemnity associations on terms that they could only recover what they “shall in fact have paid.” Can one nevertheless say that such a contract is unreasonable, uncommercial or absurd? The argument gains further modest support if one turns to the rules of the Sunderland Steamship Protection and Indemnity Association, cited in my judgment at first instance in The Fanti [1987] 2 Lloyd’s Rep. 299, 303:

“Provided always that in the case of a liability actual payment (which shall be made out of moneys belonging to him absolutely and not by way of loan or otherwise) by the member or other insured persons of the full amount of such liability shall, unless the directors otherwise decide, be a condition precedent to the right of the member or other insured person to recover and the obligation of the association to satisfy and make good.”

Presumably the members of the Sunderland Association were content with that rule.

I now turn to the substantive arguments on unreasonableness as they were put before us.

There was no direct evidence that on the reinsurers’ construction the contracts were unreasonable. Whether such evidence would be admissible or not—we must bear well in mind that the undue proliferation of expert evidence is a major cause of the present high cost of litigation—I very much doubt whether it would be helpful. Witnesses can explain the background, context, surrounding circumstances or matrix. As we have no such evidence, or very little, this is not the case to embark on consideration of what would qualify under that head. Surrounding circumstances must at least be known, or capable of being known, to both parties at the time when the contract is made Reardon Smith Line Ltd. v. Yngvar Hansen-Tangen (trading as H.E. Hansen-Tangen) [1976] 1 W.L.R. 989, 996 by Lord Wilberforce. Perhaps they might include the general [*362] nature of the risks which the insurers would insure, the circumstances and manner in which claims were liable to arise on an excess of loss reinsurance, the proportions in which different kinds of claim occur, the circumstances which reinsurers take into account in fixing the premiums, and the ostensible solvency of the insurers. That information might be helpful. But, once it is available, two conflicting opinions on whether the contract is unreasonable on one construction are of little value. Even though it be a question of fact, the judge can as well make up his own mind.

The argument for the insurers emphasised the case where one single claim is made upon them for what is unquestionably one loss, and the claim exceeds the lower limit of the reinsurers’ liability. An example would be a claim for £5m. where the reinsurers (as here) are liable for £2m. in excess of £3m. It is said to be unreasonable that the insurers must first pay £5m. to their insured before they can recover from the reinsurers.

One answer to that argument is that the insurers can borrow from their bank. Or, if they cannot do that, they need only pay their own share of £3m. and a little bit more, before they start recovering from the reinsurers; and each recovery can then fund the next payment. This scheme might run into difficulty if the insurers’ financial position was so precarious that a payment by them would be a fraudulent preference. But the main answer to the argument is that the single large claim is not the typical situation which excess of loss reinsurance has to guard against. In support of that we were shown a sheet of figures dated January 1995 stating the ultimate net loss to date for Exxon Valdez as 24 March 1989, for Phillips Petroleum as 23 October 1989 and for the Australian earthquake as 28 December 1989. That, it was suggested, illustrates how claims arising from a particular casualty accumulate as time goes by, and how from time to time the insurers would claim upon the excess of loss reinsurers. If a time comes when the insurers can no longer fund payments to their insured owing to insolvency, it is then that the reinsurance ceases to pay on the reinsurers’ construction. As the parties and the judge recognised, excess of loss reinsurance is often arranged in layers or slices. Those layers will continue to pay in sequence, and so provide funds, for as long as the insurers can and do pay the claims by their insured. It is only when the insurers nevertheless become insolvent that the flow ceases.

Buckley L.J. in the Law Guarantee case [1914] 2 Ch. 617, 634 considered it important that the solvency of the primary insurer cannot have been a circumstance that was considered in fixing the premiums. Factually that must be right. But it leads me to ask what circumstances are taken into account in fixing the premium for this type of insurance. We did ask; and despite counsel taking instructions we received no answer. In the first of the whole account contracts, for example, there is a minimum and deposit premium of U.S. $600,000; and it is to be adjusted so as to equal 1á09 per cent. of insurers’ net premium income for the year in question. How was that figure reached, for a reinsurance of £2m. in excess of £3m.? I can only guess that it was based on the underwriter’s experience of past claims with these and other primary insurers, and on his judgment of what his rivals would be asking. I do not find much  [*363]  weight in that as an argument that the reinsurers are—or are not—to pay in full when the insurers become insolvent.

Mance J. also attached importance to what he described as regulatory and accounting considerations, that is to say (i) the margin of solvency required by section 32 of the Insurance Companies Act 1982 and the Insurance Companies Regulations 1981 (S.I. 1981 No. 1654), (ii) section 17 of the Act and the Insurance Companies (Accounts and Statements) Regulations 1983 (S.I. 1983 No. 1811). For my part, I cannot see that these requirements are of any real significance as a guide to the intention of the parties to the contracts which they made in this case. It is no doubt of the highest importance in the national interest that insurance companies should remain solvent; it is no doubt important, to that end, that they have adequate reinsurance on suitable terms. But it is not the task of the courts to interpret private contracts in such a way as to ensure that the national interest is well served. It is for the regulators to carry out their task of regulating, and not for the judges to do so indirectly by the way that they construe contracts. We have some material on the history of the ultimate net loss clause, and this was considered in the judgment of Mance J. However, neither Mr. Kentridge nor Mr. Pollock wished to rely on that material in this court; and I too find it of no assistance. In the course of the argument Simon Brown L.J. asked if there was a well recognised clause different from that which we are considering. The answer, according to my note, is that Mr. Ruttle, junior counsel for the reinsurers, had seen a contract which said “payable” instead of “paid.” That, I am afraid, does not take us very far.

Mr. Pollock produced extracts from a booklet published in 1986 by the Reinsurance Offices Association. This was said to be relevant to the question whether the contracts were, on the insurers’ construction, unreasonable or absurd, because it was material which the parties would have had in mind. The commentary on the ultimate net loss clause reads in part:

“The clause is intended to indicate that the reinsured must indemnify the original insured before it is entitled to claim against the reinsurer. . . . However . . . the better opinion is that this does not affect the liability of the reinsurers towards the reinsured in the event of a liquidation. . . . If it is thought necessary a clause to this effect may be added.”

The October 1990 version of the booklet, which did not exist when these contracts were made, has this passage:

“The clause also provides as do most, although not all such reinsurance covers, that the ultimate net loss consists of losses ‘actually paid’ by the reinsured which is intended to make it clear that the reinsurance is not meant to provide some form of contingency cover but a contract of indemnity against loss.”

We were also shown extracts from a booklet of the London Insurance and Reinsurance Market Association, published in June 1992. This  [*364]  reproduces the first of the passages quoted above. It also has this comment upon a liquidation clause contained in the booklet:

“Although traditional London market ‘ultimate net loss’ clauses oblige the reinsurer to indemnify the reinsured only in respect of losses which have been ‘actually paid’ by the reinsured, there was always a problem should the reinsured be in liquidation. Under those circumstances the liquidator quite simply could not ‘actually’ pay losses under the reinsured’s original policies because he did not have the funds available to enable him to do so. In the United States this problem has in recent years been surmounted by the inclusion in wordings of a liquidation clause which obliges the reinsurer, in the event of the insolvency of the reinsured, to settle losses in full to the liquidator under the original policy. It is felt that, to make the position absolutely clear in U.K. wordings, it is advisable likewise to include a liquidation clause. This clause, whilst conceding that a contract of reinsurance remains essentially a contract of indemnity, nevertheless obliges the reinsurer in the event of the reinsured’s liquidation to pay the liquidator in full even though the liquidator may not already have ‘actually paid’ the loss concerned.”

I have grave doubt as to whether the material in any of those booklets is admissible, quite apart from the fact that it was not produced before the judge. But what it does show, to my mind, is that judges should hesitate before coming to an a priori conclusion as to what the object of a particular clause in a contract was, or that its literal meaning would lead to absurdity. The clause in its present form has existed since 1937; the market has apparently realised that there could be some doubt as to its effect when the primary insurer goes into liquidation (and see the American cases cited below); but the parties to these contracts, like many others, made no effort to clarify the position. If one draws any conclusion, it is that they were not much concerned as to whether the reinsurers would have to pay if the insurers were in liquidation. If the point ever arose, which they may have thought very unlikely, then it would be up to the courts to provide a solution.

My conclusions on unreasonableness or absurdity are first, that there is no evidence to support it. Secondly, if the point had expressly been drawn to the parties’ attention when the contracts were first made, I am by no means sure that they would have said, “Of course we do not mean that.” Rather the insurers might well acknowledge that if, despite their margin of solvency, there came a time when they were unable to fund payment of claims, then their business would dissolve in ruins whatever these particular reinsurers did. And the reinsurers would say that they felt no obligation to contribute to the dividend received by other creditors of the insurers. True this would leave many unhappy creditors, including those who were insured under the primary cover. But I repeat that it is not for us to regulate for the public good. It is plain, and Mr. Kentridge accepts, that the parties are lawfully entitled to make contracts such as Mr. Pollock says they did make. They are not obliged to consider the interests of possible future creditors in a possible future liquidation, provided that their present trading does not infringe the insolvency law.  [*365]

Mr. Kentridge suggests that there can be no motive which would influence the reinsurers to stipulate for payment to cease upon liquidation of the insurers. I am not sure that he is right; but, even if he is, he is answering the wrong question. We should ask whether there is a motive for stipulating that the insurers must pay before the reinsurers reimburse them. To that question there are a number of possible answers: (i) to ensure that the insurers have satisfied themselves that the claim is well founded (see Mr. Boyd’s argument in The Fanti [1991] 2 A.C. 1, 36), (ii) to prevent the insurers receiving the money themselves and then failing to pay their insured (Mr. Sumption, at pp. 33-34), or (iii) to ensure that if there is delay in the money reaching its ultimate destination, the interest on it accrues to the reinsurers rather than the insurers. This last is Mr. Pollock’s suggestion; most people are familiar with the habit of the insurance industry, or some parts of it, to ask for money promptly when it is due as premium and to pay rather more slowly in response to claims. None of those three reasons is appropriate when the insurers are in liquidation; but each is a sufficient explanation of why the contract was meant to mean what it says.

Cases directly on the point

Allemannia Fire Insurance Co. of Pittsburgh v. Firemen’s Insurance Co. of Baltimore (1908) 209 U.S. 326 arose out of the great fire in Baltimore in 1904. The clause in a reinsurance contract read:

“Losses, if any, shall be payable pro rata with, in the same manner, and upon the same terms and conditions as paid by the said reinsured company under its contracts hereunder reinsured, and in no event shall this company be liable for an amount in excess of a ratable proportion of the sum actually paid to the assured. . . .”

The United States Supreme Court (including Harlan and Holmes JJ.) held that the primary insurers were not obliged to pay the loss before enforcing their claim against the reinsurers.

The opposite conclusion was reached by the same court in Fidelity & Deposit Co. v. Pink (1937) 302 U.S. 224. There the contract provided that the reinsurer’s share of the loss “shall be paid to the reinsured upon proof of the payment of such items by the reinsured, and upon delivery to the reinsurer of all essential documents . . .” Again there were distinguished names on the bench. The court said, at pp. 228-229:

“Both courts below thought that Allemannia Fire Insurance Co. of Pittsburgh v. Firemen’s Insurance Co. of Baltimore, 209 U.S. 326 required approval of respondent’s contention. This was error. The defense was well taken and should have been sustained. We do not question the general rules concerning liability of reinsurers announced in the Allemannia case; but the liability under any written contract must be determined upon consideration of the words employed, read in the light of attending circumstances. Here the two insurance companies stood upon an equal footing; both were experts in the field. The language used differs materially from that found in the policy of the Allemannia Co. There is no ambiguity and no  [*366]  circumstances requires disregard of the ordinary meaning of the language.”

Of English authority the first case that we were referred to was Gether v. Capper (1855) 15 C.B. 696. There the contract was that the owner was to receive “The highest freight which he could prove to have been paid on the same voyage.” It would seem that the owner could prove what had been contracted to be paid, but (not surprisingly) was unable to prove that it had been actually paid. The interlocutory observation of Jervis C.J., at p. 701, shows that some members of the court thought proof of the contract sufficient. This was in a wholly different context and is no help at all in the present case.

In the Home and Overseas Insurance Co. Ltd. v. Mentor Insurance Co. (U.K.) Ltd. [1990] 1 W.L.R. 153 the defendants were insurers who had gone into liquidation and the plaintiffs were reinsurers. The ultimate net loss clause was in the same terms as that which I have quoted (in part) from the two whole account contracts in this case, save that it began “The term ‘ultimate net loss’ “ instead of “The term ‘net loss.’ “ The reinsurers sought a declaration under R.S.C., Ord. 14 that they were only liable to pay in respect of sums which had been paid by the insurers. The insurers applied for a stay of the action in favour of arbitration.

Hirst J. [1989] 1 Lloyd’s Rep. 473 at first instance dismissed the Order 14 summons and granted a stay. He considered the solution proposed by the reinsurers, at p. 480, “both unjust and discordant with commercial good sense.” However, he did quote from Butler & Merkin, Reinsurance Law, pp. D.2.2-09-D.2.2-10 including this passage:

“it is of interest to note that, following the Pink decision, 302 U.S. 224, legislation was enacted in New York—subsequently to be adopted in other states—to ensure that where an original loss has occurred, the reinsurer is liable to pay the full reinsured sum attributable to that loss to the liquidator irrespective of whether or not the ceding company had paid its policyholders. It may well be the case that if reinsurance agreements with English ceding companies were to begin to incorporate clauses which make it clear that the reinsurer’s responsibility to pay was limited to sums actually paid, legislation might intervene in a similar way to negative their effect.” (Emphasis added.)

“Actually paid” were of course the words used in that case, and in this case, and presumably in many other contracts at that time.

The Court of Appeal (Parker, Lloyd and Balcombe L.JJ.) agreed that the appeal should be dismissed, but not because they agreed with the view of Hirst J. as to the meaning of the contract. Parker L.J. said [1990] 1 W.L.R. 153, 160:

“I come to the construction issue. I fully appreciate the force of Mr. Clarke’s arguments which received considerable reinforcement from the decisions in another division of this court Firma C-Trade S.A. v. Newcastle Protection and Indemnity Association [1987] 2 Lloyd’s Rep. 299 and Socony Mobil Oil Co. Inc. v. West of England Shipowners Mutual Insurance Association (London) Ltd. (No. 2) [*367]  [1987] 2 Lloyd’s Rep. 529 which were handed down whilst argument on this appeal was in progress. Despite the force of those arguments, however, I have no doubt at all that the question raised is a serious, difficult and important one which deserves mature consideration. Indeed, if it be different, I am satisfied that the plaintiffs are on the material before us not entitled to judgment here and now.”

So no conclusion was reached, and the dispute went to arbitration. Mr. Kentridge relies on the case as showing that there was ambiguity in the ultimate net loss clause; and he submits that if there is ambiguity the insurers should succeed. But the same argument would have occurred to the Court of Appeal, and they might in consequence have adopted the view of Hirst J. They did not do so.

Finally there is the decision of the House of Lords in the Firma C-Trade case [1991] 2 A.C. 1. It had not been argued at first instance, before either Saville J. or myself, that the words “and shall in fact have paid” meant anything other than they said. But there was an argument in the Court of Appeal and the House of Lords that equity would treat them as satisfied if the association’s members were merely liable to pay. This was rejected. We have the authority of the House of Lords as to the effect of “those express provisions” (Lord Brandon of Oakbrook, at p. 28) which were “perfectly clear” (Lord Goff of Chieveley, at p. 36). I can see no relevant distinction between “sum actually paid” and “shall in fact have paid.”

It is said that members of a mutual insurance association, sometimes called a “club,” are different from those who carry on insurance business and are reinsured. I cannot for my part see that either is likely to be less in business to make profits and avoid losses than the other. But like Lord Goff (if I may say so) I can see some force in Mr. Boyd’s submission that a P. & I. association is concerned that its members should remain solvent so as to be able to pay supplementary calls if they occur. So too in the present case the deposit premium is provisional, to be adjusted later; and there may be recoveries which the insurers are obliged to pay over to the reinsurers. Consequently the reinsurers have an interest that the insurers remain solvent. I can see no distinction of any great significance on that ground.

The respondent’s notice

The first point argued was that it was contrary to public policy for a contract to provide for a result which was repugnant to the insolvency laws; and contracts should be construed so as to avoid that result.

The cases cited in support of that submission were In re Johns; Worrell v. Johns [1928] Ch. 737 and British Eagle International Airlines Ltd. v. Compagnie Nationale Air France [1975] 1 W.L.R. 758. They are, as it seems to me, concerned with a situation where the parties by contract seek to provide that one creditor shall, in the event of insolvency, receive more than the appropriate dividend on his debt. They do not touch upon this case, where the contract is concerned with sums payable to the company. A contract may lawfully provide that they shall not be payable in the event of insolvency.  [*368]

Mr. Kentridge mentioned a second argument, based on what was said by Neville J. in In re Law Guarantee Trust and Accident Society; Godson’s Claim [1915] 1 Ch. 340, 345, that “in law bankruptcy in an action on an indemnity was equivalent to payment.” However, this principle could not be found in the sources to which it was attributed, and Mr. Kentridge did not feel able to pursue it.

Conclusion

This dispute is about the meaning of two words, “actually paid.” There must come a time when efforts to bend meaning (or, as I would say, reverse it) have to stop. The literal meaning of the words in the contracts requires that the insurers shall have paid before the reinsurers are liable. To the extent, if at all, that this produces a result which is unreasonable, it is not so unreasonable that it requires us to depart from the plain meaning of the words. Indeed I doubt whether it is unreasonable at all. Mance J. held that the object of reinsurance is to spread the load. So indeed it is. But it does not follow that any term in a reinsurance contract which results in anything other than an equal division of the loss is to be disregarded. The load is to be spread upon and subject to the terms of the contract. I would allow this appeal. The judge had been asked by both parties to make a decision under Order 14A; and he granted a declaration that

“it is not a condition precedent to the liability of [the reinsurers] to indemnify [the insurers]. . . that [the insurers] should first have made payment in respect of the ascertained amount of the relevant claim by its insured by way of transfer of funds to its insured or otherwise have satisfied the ascertained amount of such claim.”

I would delete the word “not,” and otherwise grant a declaration in the same terms.

SIMON BROWN L.J.  Staughton L.J.’s judgment helpfully sets out the material relevant to this appeal so that I can express my own contrary opinion altogether more briefly.

The appeal was argued on the basis that it raises a short point of construction: does the phrase in these reinsurance policies “the sum actually paid” mean what it appears to mean or does it mean “the sum actually payable?” In other words, for the reinsurers to be liable, must the reinsured actually have disbursed (or otherwise satisfied) the claim against which the reinsurance is to indemnify them, or it is sufficient that such claim has been established to be immediately payable?

It would not be helpful for me either to cite or seek to rephrase the principles emerging from the many authorities governing the correct approach to construing a commercial contract. Plain it is that the contractual words used may, whatever their context, be so abundantly clear that however unreasonable the result, they must be given that clear meaning. Failing that, however, then as I understand what Lord Reid said in Wickman Machine Tool Sales Ltd. v. L. Schuler A.G. [1974] A.C. 235, 251, the more unreasonable the result of a given construction, the readier should the court be to adopt some less obvious construction of the words:  [*369]

“detailed semantic and syntactical analysis . . . must be made to yield to business commonsense,” as Lord Diplock put it in Antaios Compania Naviera S.A. v. Salen Rederierna A.B. (The Antaios) [1985] A.C. 191, 201.

Is the apparent meaning of the words used here—“the sum actually paid”—so abundantly clear in context, does it so obviously involve actual disbursement, that, however unreasonable this result, one must so construe them?

Powerful though the arguments are to that effect, and hesitant though I am to disagree with Staughton L.J.’s conclusions, I have finally reached the view that the words here are not so clear. I readily acknowledge that the most obvious meaning of the words is that for which the reinsurers contend. But if it be asked how such a meaning could have been made any clearer, a simple answer is to be found in the House of Lords decision in Firma C-Trade S.A. v. Newcastle Protection and Indemnity Association (The Fanti) [1991] 2 A.C. 1—namely by adopting the words used in the club rules there under consideration, whereby the members were indemnified against claims which they “shall have become liable to pay and shall in fact have paid.” I recognise, of course, the close similarity between “the sum actually paid” and claims . . . which he . . . shall in fact have paid.” But the inclusion in The Fanti of the additional words “shall have become liable to pay” made it impossible to argue there, as was successfully argued before Mance J. here, that the critical words meant just that—in that case the words so construed would have been simply tautologous; here such a construction provides an arguable alternative.

To say that words of yet clearer meaning could have been used is not, however, I accept, a sufficient answer to the submission that the words used here are themselves unambiguously clear—“clear beyond rational argument” as Mr. Pollock boldly contended. Additional responses need to be given. First is that, as always, the words have to be construed in their context. That context here has two aspects. One, the wider aspect, is that these words appear in an excess of loss reinsurance contract, a context in which it is unlikely that the reinsurers’ duty to pay will be made conditional upon the reinsureds’ prior disbursement of incoming insurance claims—as to this improbability see particularly In re Eddystone Marine Insurance Co; Ex parte Western Insurance Co. [1892] 2 Ch. 423; In re Law Guarantee Trust and Accident Society Ltd.; Liverpool Mortgage Insurance Co.’s Case [1914] 2 Ch. 617 and Home and Overseas Insurance Co. Ltd. v. Mentor Insurance Co. (U.K.) Ltd. [1990] 1 W.L.R. 153.

The second and narrower aspect is that the words are used in an “ultimate net loss clause,” a clause which on any view is concerned essentially with the measurement of reinsurance recoveries and where it would be surprising to find imposed a condition requiring prior disbursement.

In both these respects, be it noted, the present context differs from that existing The Fanti [1991] 2 A.C. 1. There the contract was not one of reinsurance but of mutual insurance between club members. Perhaps more importantly, however, the critical words appeared there in the indemnity rule itself—as an essential element of entitlement to indemnity—rather than in what is essentially a loss measurement clause. True, in the first two of the reinsurance contracts here under consideration, the liability  [*370]  Clause itself provides that “the reinsurers shall only be liable if and when the ultimate net loss sustained by the reinsured [reaches a certain figure] and the reinsurers shall thereupon become liable. . . .” But, as Mance J. pointed out, “the focus of the clause is not on timing but on the scope and limits of the financial responsibility accepted on an excess of loss basis under the contracts.” There is, moreover, no equivalent provision in the main indemnity clause of the third contract and, although clearly the appeal could be decided differently regarding that contract, no one has contended for such a result.

Mance J. furthermore pointed out that, read literally, the liability clause in the first two contracts is inconsistent with the ultimate net loss clause, there being no liability under the former to pay on an interim basis despite the clear implication to this effect in the proviso to the later clause. This to my mind is another factor justifying a less literal approach to the words in question.

The importance of context is neatly illustrated by Mr. Kentridge’s example of the situation between exchange and completion of contracts for a property purchase, where the buyer could well say that he had bargained down the asking price and “actually paid” something less. In that admittedly very different context the words would not connote disbursement. In the present context too, I conclude that it is possible to construe the words as not requiring disbursement but rather as emphasising that the recoverable loss must be immediately due and net of all deductions. That I readily accept is a difficult construction, but not, I think, any more difficult than to construe “any breach” of contract as meaning “any repudiatory breach,” as in The Antaios [1985] A.C. 191.

This difficult construction, however, I would only finally adopt were the more obvious construction to cause plainly unreasonable results. I turn next, therefore, to consider this aspect of the case.

How (if at all) unreasonable is it to make prior payment by the reinsured a condition of indemnity under these policies? A convenient starting point is again The Fanti [1991] 2 A.C. 1. There, of course, because the rules were found—indeed accepted, subject to as unsuccessful argument in equity—to provide unambiguously for prior payment, the reasonableness of that condition was ultimately immaterial. Nevertheless it was considered and two justifications were suggested. First, that the purpose of the rule was to prevent a member from making a profit from his insurance cover by receiving payment from the club but failing to pay the third party. Second, a suggestion to which Lord Goff of Chieveley expressed himself much attracted, at p. 36:

“that, in a mutual insurance association such as a P. & I. Club, it is essential that members should be able to assume the financial probity of other members, because all of them are insurers as well as insured. To that end, it is customary to require each member to discharge his own liability before he can be indemnified against it by the club. Each member is, after all, running his own business; it is up to him to make sure that a claim against him is well founded, and the best way of ensuring that is to require him first to pay the claim before seeking indemnity from the club.”  [*371]

No such justification, submits Mr. Kentridge, can exist in the context of reinsurance. Take excess of loss reinsurance against catastrophe loss. Sometimes, inevitably, the reinsured will be unable to meet the entire loss from his own pocket—that, indeed, will often be why he obtained reinsurance in the first place. It would be wholly unreasonable on that account to deny him indemnity under the reinsurance, to drive him into insolvency, and to give the reinsurer the windfall benefit of not paying without even the liability to return premiums.

That, replies, Mr. Pollock, is just rhetoric. There is no known case where the reinsured has in fact been forced into insolvency by the requirement to make prior payment of the entire loss. Provided only that he can pay his share of the loss, he will invariably be able to find (if necessary with the aid of a bank loan) sufficient in addition to pay at least part of the reinsured excess and thereby trigger at worst a process of staged recovery under the indemnity clause.

I find it helpful to consider the rival constructions in three different situations: first, where the reinsured is in any event insolvent and unable to pay even his own share of the loss; second, at the opposite end of the spectrum, where without difficulty he can pay the entire loss; and third, where he can pay his own share but not easily more.

(1) Insolvency

I can readily see that reinsurers might wish to provide that in the event of the reinsured becoming insolvent, indemnity—which in those circumstances would go to the general body of creditors—should not be payable, and clearly—subject only to an argument on the application of the principle in British Eagle International Airlines Ltd. v. Compagnie Nationale Air France [1975] 1 W.L.R. 758—that could be stipulated. But there was no such stipulation here, and in the one policy where the reinsured’s insolvency was expressly contemplated (by article 22 of the third contract), as Mance J. points out:

“The message conveyed is that a reinsurers’ liability for outstanding losses remains unchanged despite the reinsured’s insolvency or inability to pay, even if reinsurers terminate cover in respect of future losses and a fortiori if reinsurers do not.”

(2) Reinsured able to pay entire loss

In this situation it really makes little difference to the parties whether prior payment is required or not. Even if it is required and made, the reinsurer comes under an immediate liability to reimburse his reinsured and accordingly gains nothing from the provision. The very fact that he must immediately reimburse the loss denies the provision any value as a guarantee that the claim is well founded.

(3) Reinsured unable to pay entire loss

This surely is the paradigm situation and I for my part find Mr. Kentridge’s argument upon it compelling, Mr. Pollock’s unpersuasive. Theoretically, no doubt, the reinsured, if solvent, could seek a bank loan or otherwise pay part at least of the excess loss so as to trigger some  [*372]  reimbursement under the indemnity policy. But to what end? Assuming full recovery was ultimately achieved as intended under the policy, the reinsurers would have gained nothing (save perhaps very marginally postponed liability), whereas the reinsured would probably have been put to very considerable inconvenience and expense. The condition could not rationally have been imposed to achieve those consequences. It serves, in short, no legitimate (i.e. reasonable interest) of the reinsurers. In reality, the only effect of introducing a condition of prior payment would be that sometimes it could not be satisfied in which event the entire object of the reinsurance contract would be thwarted and the reinsurers would receive a pure windfall gain. That cannot have been the mutual intention of the parties. I conclude therefore that such a construction of the clause is wholly unreasonable and that it must yield to business common sense.

For these reasons, which I believe do no more than reproduce in shorter form the essential reasoning to be found in the very full and helpful judgment below, I for my part would dismiss this appeal.

NOURSE L.J.  The contracts out of which this appeal arises are excess of loss reinsurances. Liability under such an insurance, while it may be provisionally assessed, cannot be conclusively established until there has been a final settlement between reinsurer and reinsured, at which point the reinsured must produce an account showing that the aggregate sum paid under the insurances covered is within the excess. In practice no doubt, even at the stage of final settlement, a reinsurer will often act on the basis of sums payable by the reinsured but not yet paid. But his strict right must be to insist on actual payment, because until then his own liability is not conclusively established.

That being the nature of the insurance, how should the contracts in this case be construed? Their material provisions are fully set out in the judgment of Staughton L.J. Like him, I will start with the two whole-account contracts, the first three and most important clauses of which are headed “Reinsuring clause,” “Liability clause” and “Ultimate net loss clause” respectively. The reinsuring clause provides that the reinsurance is to pay “all losses howsoever and wheresoever arising” subject, however, to the terms and conditions following. So the outcome primarily depends on the interaction between the liability and ultimate net loss clauses.

If the definition of “net loss” in the ultimate net loss clause is incorporated into the liability clause, the words “sustained by the reinsured” being omitted as tautologous, the first part of the latter clause reads as follows:

“The reinsurer shall only be liable if and when the ultimate sum actually paid by the reinsured in settlement of losses or liability [after making deductions etc.] in respect of interest coming within the scope of the reinsuring clause exceeds £3,000,000. . . . “

Bearing in mind the nature of the insurance as I have stated it, I am unable to see why a provision headed “Liability clause” and expressed to deal with liability should not look forward to the final settlement between the reinsurers and the reinsured, at which point, as I have said, the reinsurers’ liability will be established by reference to the aggregate sum  [*373]  that the reinsured has actually paid under the insurances covered. On this view of the matter the words “the sum actually paid” mean what they say, neither more nor less. An objection to this construction might possibly be based on the words “and when” in the first line of the liability clause. Certainly the effect might be even clearer if it read “the reinsurers shall only be liable if the ultimate sum actually paid etc.” But I do not think that the objection could be a good one. The words “if and when” are merely words of contingency equivalent to “in the event that.” Accordingly, the inclusion of the words “and when” does not alter the effect of the liability clause.

It is then necessary to consider the second and third sentences of the ultimate net loss clause. The second contemplates that there may be “a loss settlement,” necessarily an interim settlement, and it provides for all necessary adjustments to be made thereto in respect of salvages, recoveries or payments recovered or received subsequently. The third sentence contains a proviso that nothing in the ultimate net loss clause shall be construed to mean that losses under the reinsurance are not recoverable until the reinsured’s ultimate net loss has been ascertained. Thus the second and third sentences of the ultimate net loss clause provide for what is to happen during the period before final settlement and are consistent with the view that, in contrast, the liability clause looks forward to that point. In my view none of the other provisions of the whole-account contracts has any significant role in the decision of the question that confronts us.

I turn to the third contract, which relates to aviation risks. Here the decisive provisions are contained in articles 7 and 10 headed “Reinsuring clause” and “Ultimate net loss clause” respectively, with article 7 doubling up for the reinsuring and liability clauses in the whole-account contracts. If the definition of “ultimate net loss” in article 10 is incorporated into article 7, the first part of the latter clause reads as follows:

“This reinsurance is to indemnify the reinsured, subject to its provisions, for all losses which may be sustained by the reinsured in excess of a sum actually paid by the reinsured in settlement of losses or liability [after making deductions etc.] of £28,000,000. . . .”

Again, I am unable to see why that provision should not look forward to the final settlement between the reinsurers and the reinsured. Again, the words “the sum actually paid” mean what they say, neither more nor less. Again, the second and third sentences of article 10 perform the same function as the second and third sentences of the ultimate net loss clause in the whole-account contracts and are consistent with the view that article 7 looks forward to the point of final settlement. Again, none of the other provisions of the aviation contract has any significant role in the decision of the question that confronts us.

By this route I have come to the conclusion that under none of the three contracts are the reinsurers entitled, before the point of final settlement, to make payment to the reinsured conditional on prior payment under the insurances covered. Not until that point do the liability clauses come into play. Until then the relationship between the parties is essentially governed, first, by the reinsurers’ general obligation to  [*374]  indemnify the reinsured against the losses specified and, secondly, by the second and third sentences of the ultimate net loss clauses.

I am conscious that this construction of the contracts, although put briefly to Mr. Pollock, for the reinsurers, at the end of his reply, was not canvassed in argument either in this court or before Mance J. Yet a question of construction of a document, like that of a statute, is invariably at large; cf. Bahamas International Trust Co. Ltd. v. Threadgold [1974] 1 W.L.R. 1514, 1525, per Lord Diplock. The first task of the court is always to construe the particular words of the particular contract against the factual background known to the parties at or before the date it is entered into. Here it is only the nature and objective aim of an excess of loss reinsurance of which account can properly be taken. Having carried out that process of construction, I am of the opinion that the meaning and effect of the material provisions of the contracts in this case are plain. Since they result in nothing unreasonable, far less absurd, no further inquiry is necessary. For these reasons, I agree with Simon Brown L.J. that the appeal should be dismissed.

Appeal dismissed.

No order as to costs.

Leave to appeal granted.

The reinsurers appealed.

[*381]

22 May.

LORD GOFF OF CHIEVELEY.  My Lords, I have had the advantage of reading in draft the speech of my noble and learned friend, Lord Mustill, and for the reasons he gives I, too, would dismiss this appeal.

LORD GRIFFITHS.  My Lords, I have had the advantage of reading in draft the speech of my noble and learned friend, Lord Mustill, and for the reasons he gives I, too, would dismiss this appeal.

LORD BROWNE-WILKINSON. My Lords, for the reasons given in the speech by my noble and learned friend, Lord Mustill, I, too, would dismiss this appeal.

LORD MUSTILL.  My Lords, this appeal turns on the meaning of the words “actually paid” in three contracts of reinsurance. The question is whether the words prescribe that no sum will be paid by reinsurer to reinsured in respect of a loss, or more accurately that no sum will be brought into the balance of account between the two parties, until the reinsured has paid out a sum of money to the person whose claim against him has brought the reinsurance into play. At first sight this seems the shortest of questions, requiring a very short answer; and so in the end it proves to be. But the instinctive response must be verified by studying the other terms of the contract, placed in the context of the factual and commercial background of the transaction. I will therefore go straight to the nature of the business and to the terms of the contract in which it was embodied, concentrating for the moment on only one of the three policies, namely policy no. X 20693/5386.

By this contract two syndicates, represented in these proceedings by Mr. P. F. Fagan (“the syndicates”) reinsured for small percentages of a total line Charter Reinsurance Co. Ltd. (“Charter”) in respect of Charter’s whole account for losses occurring during the calendar year 1989. The contract formed part of a programme which also comprised “specific reinsurances” taken out with others on four of Charter’s accounts viz., non-marine LMX; non-marine international; marine; and aviation. These accounts were reinsured in a series of tranches to limits of, respectively, £23m., £11m., £32á25m. and £31á5m. Above these reinsurances of separate accounts were the levels of whole account reinsurance with which two of the three contracts in suit were concerned. Above a retention of £100,000, there were successive layers of £2á9m., £2m., £2á5m. and £2á5m. Policy no. 5386 insured the second of these layers, for £2m. excess of £3m. and one of the other policies sued upon covered the fourth layer up to £7á5m. For the purposes of the present litigation it is assumed that a series of major casualties arising from perils insured under the policy have caused valid claims to be made against Charter under policies issued by it to other reinsured or insured companies or syndicates (“the inward policies”). These claims are so large as to exhaust all the reinsurances comprising the specific accounts of the programme, and to encroach upon the relevant layers of whole account reinsurance. The problem arises from the fact that Charter is in provisional liquidation, being unable to pay its debts as they fall due, and these debts include claims under the inward policies. For  [*382]  their part, the syndicates do not for present purposes dispute that all the requirements of a valid claim against them by Charter are present, save only one: that Charter have not paid, and cannot pay, the inward claims which they have reinsured. Thus, say the syndicates, Charter have no cause of action under the reinsurance.

The practical importance of this defence, if sound, is obvious; and its implications have been multiplied by the levels of financial frailty experienced in the London insurance market in recent years. Across the market as a whole very large sums depend upon it, and the litigation from which this appeal stems has been brought in practice, if not in form, as a test case. The proceedings take the shape of an action by Charter for a summary declaration that payment by way of transfer of funds or other means of satisfaction by Charter under the inward policies was not a condition precedent to the liability of the syndicates. Within a very few months it proved possible to obtain the opinion of the Commercial Court in the shape of a meticulous and thoughtful judgment of Mance J., granting a declaration in those terms. Upon recourse to the Court of Appeal this decision was upheld by a majority, Staughton L.J. dissenting. The syndicates now appeal to this House.

This being, I believe, a sufficient summary of the dispute I turn to policy no. X 20693/5386. It is important to quote its terms at some length.

For ease of reference I have added numbers and letters, and have placed in italics the words around which the controversy revolves.

“1. Reinsuring clause

“This Reinsurance is to pay all losses howsoever and wheresoever arising during the period of this Reinsurance on any Interest under Policies and/or Contracts of Insurance and/or Reinsurance underwritten by the Reinsured in their Whole Account. Subject however to the following terms and conditions.

“2.(a) Liability clause

“The Reinsurers shall only be liable if and when the Ultimate Net Loss sustained by the Reinsured in respect of interest coming within the scope of the Reinsuring Clause exceeds £3,000,000 or U.S. or Can.$6,000,000 each and every loss and/or Catastrophe and/or Calamity and/or Occurrence and/or Series of Occurrences arising out of one event and the Reinsurers shall thereupon become liable for the amount in excess thereof in each and every loss, but their liability hereunder is limited to £2,000,000 or U.S. or Can.$4,000,000 each and every loss and/or Catastrophe and/or Calamity and/or Occurrence and/or Series of Occurrences arising out of one event.

“(b) Warranted Reinsurers hereon to have benefit of Specific Reinsurances as per Schedule attached.

“Ultimate net loss clause

“(c) The term ‘Net Loss’ shall mean the sum actually paid by the Reinsured in settlement of losses or liability after making deductions for all recoveries, all salvages and all claims upon other Reinsurances whether collected or not and shall include all adjustment expenses arising from the settlement of claims other than the salaries of employees and the office expenses of the Reinsured.  [*383]

“(d) All Salvages, Recoveries or Payments recovered or received subsequent to a loss settlement under this Reinsurance shall be applied as if recovered or received prior to the aforesaid settlement and all necessary adjustments shall be made by the parties hereto. Provided always that nothing in this clause shall be construed to mean that losses under this Reinsurance are not recoverable until the Reinsured’s Ultimate Net Loss has been ascertained.

“(e) Notwithstanding anything contained herein to the contrary, it is understood and agreed that recoveries under all Underlying Excess Reinsurance Treaties and/or Contracts (as far as applicable) are for the sole benefit of the Reinsured and shall not be taken into account in computing the Ultimate Net Loss or Losses in excess of which this Reinsurance attaches nor in any way prejudice the Reinsured’s right of recovery hereunder.

“3. Period of reinsurance clause

“This Reinsurance covers Losses Occurring during the period commencing with 1 January 1989 and ending with 31 December 1989 both days inclusive, Local Standard time at the place where the loss occurs. . . .

“4. Premium clause

“The Minimum and Deposit Premium for this Reinsurance shall be U.S.$600,000: 10 per cent. Payable in Sterling, namely £37,500; 891Ú2 per cent. Payable in U.S. Dollars, namely $537,000; 1Ú2 per cent. Payable in Can. Dollars, namely $3,000 . . .

“5. Currency clause

“Losses (if any) paid by the Reinsured in currencies other than Sterling, shall be converted into Sterling at the rate of exchange ruling at the date of the settlement of loss or losses by the Reinsured other than losses paid in U.S. or Can. Dollars which will be paid in those currencies.

“6. Reinstatement clause

“In the event of loss or losses occurring under this Reinsurance, it is hereby mutually agreed to reinstate this Reinsurance to its full amount of £2,000,000 or U.S. or Can.$4,000,000 from the time of the occurrence of such loss or losses to expiry of this Reinsurance and that an additional premium shall be paid by the Reinsured upon the amount of such loss or losses when they are settled in the first instance calculated at 100 per cent. of the Minimum and Deposit Premium hereunder subject to a further payment hereunder (if any) when the Final Earned Premium is known. Reinstatement premiums to be paid in the currency of loss settlement hereunder for which purpose U.S. or Can.$1á60 = £1. Nevertheless the Reinsurers shall never be liable for more than £2,000,000 or U.S. or Can.$4,000,000 in respect of any one loss and/or series of losses arising out of one event, nor for more than £6,000,000 or U.S. or Can.$12,000,000 in all.”

The case for the appellants concentrates almost exclusively on the words in italics. It is very simple. These words plainly create a condition  [*384]  precedent to any liability of the syndicates. The condition is that Charter shall have “actually paid” under the original policies. If this expression has a natural and ordinary meaning, effect should be given to it. The expression and the words which comprise it do have such a meaning. By no stretch of language can it be extended to cover a situation in which Charter has not made any disbursement, actual or even notional, and will never do so.

My Lords, to a substantial degree I accept this argument. I believe that most expressions do have a natural meaning, in the sense of their primary meaning in ordinary speech. Certainly, there are occasions where direct recourse to such a meaning is inappropriate. Thus, the word may come from a specialist vocabulary and have no significance in ordinary speech. Or it may have one meaning in common speech and another in a specialist vocabulary; and the context may show that the author of the document in which it appears intended it to be understood in the latter sense. Subject to this, however, the inquiry will start, and usually finish, by asking what is the ordinary meaning of the words used. I begin, therefore with “actually.” In my opinion this word is used by way of qualification or precaution, in the sense of “really,” “in truth,” “not notionally” or “not prospectively.” On this, I feel no doubts. The word “paid” is more slippery. Unquestionably, it is no longer confined to the delivery of cash or its equivalent. In ordinary speech it now embraces transactions which involve the crediting and debiting of accounts by electronic means, not only transfers between bank accounts by payment cards and direct debits, but also dealings with credit cards and similar instruments. Conditional payment by cheque would also be covered, at any rate outside a strictly legal context. Furthermore, I think it plain that in a document created to govern a transaction in the London insurance market payment would extend beyond remittances from debtor to creditor and would include the settlements in account with brokers which are a feature of that market. None the less, even giving “paid” an extended meaning the word would at first sight, and even without the qualifier “actually,” fall well short of encompassing a situation in which the debtor had suffered no immediate financial detriment through a transfer of funds in the direction of the creditor, and would never do so.

My Lords, I have used the expression “at first sight” because I had initially thought that the meaning of the words was quite clear, and that the complexities and mysteries of this specialist market had hidden the obvious solution, and had led the courts below to abjure the simple and right answer and to force on the words a meaning which they could not possibly bear. I was not deflected from this opinion by any of the cases cited, which with few exceptions (to which I must return) seemed too remote from the present to offer any useful guidance.

This is, however, an occasion when a first impression and a simple answer no longer seem the best, for I recognise now that the focus of the argument is too narrow. The words must be set in the landscape of the instrument as a whole. Once this is done the shape of the policy, and the purpose of the terms which I have grouped as clause 2 become quite clear. As one would expect, four essential features of the insurance are described: the perils insured against; the measure of indemnity; the  [*385]  duration of the cover; and the premium. Clause 1, read together with various later clauses of enlargement and restriction, which I have not quoted, describes the nature and geographical scope of the perils insured against. In principle, all events happening within the period laid down by clause 3 (construed in association with special provisions relating to liability insurance) which constitute losses by perils insured under the original policies are to be losses insured under this policy. This is not the place to discuss the question, perhaps not yet finally resolved, whether there can be cases where a contract of reinsurance is an insurance of the reinsurer’s liability under the inward policy or whether it is always an insurance on the original subject matter, the liability of the reinsured serving merely to give him an insurable interest. This may be important in the context of regulation, but it makes no difference here, for it is quite plain that payment by reinsurer is not the insured event. There has still been an insured loss, and even if the argument for the syndicates is right the consequence is only to reduce or eliminate the amount of Charter’s recovery under clause 2 in respect of a loss which has undoubtedly occurred. Clause 1 therefore has no bearing on the present dispute. Nor of course is the premium provision in clause 4 of any relevance.

What does matter is the group of provisions which establish the measure of indemnity, once a loss by an insured peril has taken place. I would break these down as follows.

(i) Clause 2(a) fixes the level at which financial prejudice suffered by Charter under the inward policies in consequence of a loss by a peril insured under this policy causes a liability to attach. This happens when the ultimate net loss in relation to each and every loss and/or catastrophe and/or calamity and/or occurrence (which I will call a set of linked losses) exceeds £3m. This sub-clause also fixes the upper limit of indemnity under the policy. An additional limit, this time fixed by reference not to each set of linked losses but to the cover for the entire policy year, is imposed by the last sentence of clause 6.

(ii) Clause 2(b) incorporates into the scheme of the policy the four sets of layered “specific” insurances (i.e. the “accounts”) identified in the schedule. When an event occurs which is a peril insured under one of those sets of insurances and also under this policy the limits of all the insurances comprising that account must be exceeded before any indemnity begins to fall due under this policy.

(iii) Clause 2(c) gives meaning to clause 2(a) by defining ultimate net loss. (In fact the sub-clause omits “ultimate.” This must be a mistake, for otherwise the entire group of provisions makes no sense. The word does appear in the clause as typed in the aviation policy). The purpose of clause 2(c) is to make clear that the syndicates are not to pay losses gross, but that there is to be a netting-down for recoveries, salvage and the like when ascertaining whether, and if so by how much, the relevant liabilities of Charter cross the boundary into the layer covered by this policy.

(iv) The first sentence of clause 2(d) elaborates clause 2(c) by making clear that the fixing of an ultimate net loss in respect of any set of linked losses is provisional, in the sense that the amount of it, and hence its impact if any on this layer of insurance, is to be open to recomputation if  [*386]  and when items of the identified description subsequently accrue to the benefit of Charter.

(v) The proviso in the second sentence of clause 2(d) emphasises that even though the computation of an ultimate net loss is provisional, if it yields a figure broaching the bottom of the layer insured under this policy it will then be “recoverable” even if a subsequent recalculation when all the figures are in may lead to an upward or downward adjustment, or even to the elimination of any recovery at all.

(vi) Clause 2(e) is puzzling at first sight, because the use of initial capitals may suggest that, like “Specific Insurances” in clause 2(b), the expression “Underlying Excess Reinsurance Treaties and/or Contracts” has a meaning specifically ascribed for the purpose of this policy. Yet one finds it nowhere defined. In fact, however, a reading of the document as a whole shows that capitals are used indiscriminately throughout, and that they have no special significance in clause 2(e). In the light of the explanations given in argument, I accept that the purpose of the sub-clause is simply to ensure that the calculation of the ultimate net loss under sub-clause (a) does not involve a deduction of the liabilities on the underlying layers, so as to diminish the possibility of a recovery on the layer covered by this policy.

Analysed in this way, the policy makes complete sense, and works perfectly well in practice when understood as requiring the satisfaction of only two conditions before an indemnity falls due. First, that an insured event shall have occurred within the period of the policy, and second that the event shall have produced a loss to Charter of a degree sufficient, when ultimately worked out, to bring the particular layer of reinsurance into play. This reading accommodates without strain the words “if and when,” in clause 2(a); for they are concerned only with the point, not of time but of arithmetic, at which the figures for the ultimate net loss reach the appropriate level. Equally, I am now satisfied that the purpose of “the sum actually paid” in clause 2(c) is not to impose an additional condition precedent in relation to the disbursement of funds, but to emphasise that it is the ultimate outcome of the net loss calculation which determines the final liability of the syndicates under the policy. In this context, “actually” means “in the event when finally ascertained,” and “paid” means “exposed to liability as a result of the loss insured under clause 1.” These are far from the ordinary meanings of the words, and they may be far from the meanings which they would have had in other policies, and particularly in first-tier policies of reinsurance. But we are called upon to interpret them in a very specialised form of reinsurance, and I am now satisfied that, as Mance J. expressed it in his judgment at first instance, the words in question did not have the purpose of introducing a temporal precondition to recovery in the form of disbursement or other satisfaction of the precise net commitment between Charter and its reinsured, but were there “for the purpose of measurement.”

Whilst I have come to this conclusion simply from a study of the document I ought to comment on a number of other matters which are said to bear upon it. In the first place, there is an argument ad absurdum to the effect that the parties cannot have intended Charter to retain such liquidity as would enable it to answer claims under the incoming policies  [*387]  without recourse to the reinsurance. At a time when the use of money was a vital element in the profitability of insurance business it is impossible to suppose (the argument runs) that Charter should have agreed to finance its own outlays, the more so since, if the syndicates’ interpretation of clause 2 is right, Charter would have to find, not only the funds required to disburse the sum due under this particular layer, but also the total of the underlying reinsurances. This would be a wholly impracticable arrangement, and would bear especially hard on Charter if it fell into financial trouble and lacked the means to make the payments necessary to unlock the reimbursements due under its contracts with the syndicates.

This argument draws strength from the shape of the policy. As I have already suggested, under this form of words, although perhaps not under all forms, the policy covers not, as might be thought, the suffering of loss by the reinsured in the shape of a claim against him under the inward policies, but the occurrence of a casualty suffered by the subject-matter insured through the operation of an insured peril. The inward policies and the reinsurance are wholly distinct. It follows that in principle the liability of the reinsurer is wholly unaffected by whether the reinsured has satisfied the claim under the inward insurance: see, amongst several authorities, In re Eddystone Marine Insurance Co.; Ex parte Western Insurance Co. [1892] Ch. 423. This result can undoubtedly be changed by express provision, but clear words would be required; and it would to my mind be strange if a term changing so fundamentally the financial structure of the relationship were to be buried in a provision such as clause 2, concerned essentially with the measure of indemnity, rather than being given a prominent position on its own.

Further arguments, to my mind some way short of conclusive, were advanced on each side. The syndicates pointed out a possible disconformity between the postponement of the reinsurers’ liability to pay with the statutory provisions governing margins of solvency. For Charter attention was drawn to long-established contractual provisions creating just such a condition precedent as is argued for here: for example, in the running down clause and in protection and indemnity club cover against third party liabilities, the effect of which was discussed in Firma C-Trade S.A. v. Newcastle Protection and Indemnity Association [1991] 2 A.C. 1. Each side suggested reasons why such a provision would or would not make commercial sense; and proposed ways in which the hardship to the reinsured might be ameliorated by devices such as the making of a series of small “pump priming” payments, which would produce a sufficient trickle of cash to satisfy ultimately the inward claim in full, hence unlocking a recovery under the reinsurance.

These arguments are fully explored in the judgments delivered below. Intending no disrespect I do not enter into them here, for in my opinion they cannot be decisive. If, as I believe, a proper reading of the policy discloses no condition precedent, there is little profit in considering whether it would have been absurd to include one. If, per contra, the words “actually paid” can only as a matter of language and context mean what the syndicates maintain, I would hesitate long before giving them any other meaning, just because the result would be extraordinary. The words of Lord Reid in Wickman Machine Tool Sales Ltd.v. Schuler A.G. [*388]  [1974] A.C. 235, 251 do, of course, reflect not only a method of constructing contracts but also the common experience of how language is understood:

“The fact that a particular construction leads to a very unreasonable result must be a relevant consideration. The more unreasonable the result the more unlikely it is that the parties can have intended it, and if they do intend it the more necessary it is that they shall make that intention abundantly clear.”

This practical rule of thumb (if I may so describe it without disrespect) must however have its limits. There comes a point at which the court should remind itself that the task is to discover what the parties meant from what they have said, and that to force upon the words a meaning which they cannot fairly bear is to substitute for the bargain actually made one which the court believes could better have been made. This is an illegitimate role for a court. Particularly in the field of commerce, where the parties need to know what they must do and what they can insist on not doing, it is essential for them to be confident that they can rely on the court to enforce their contract according to its terms. Certainly, if in the present case the result of finding a condition precedent would be anomalous there would be good reason for the court to look twice, and more than twice, at the words used to see whether they might bear some other meaning. In the end, however, the parties must be held to their bargain. Thus, if I had adhered to my first impression that the expression “actually paid” could possess, even in the context of the policy, only the meaning which it has in ordinary speech I would have wished to consider very carefully whether the opinion expressed in the dissenting judgment of Staughton L.J., austere as it might seem, ought to be preferred. In the event however, for the reasons stated, this is not my present understanding of the words, and since the broader question does not on this view arise I prefer to say no more about it.

Next, I must notice three decisions from the United States. The first is Allemannia Insurance Co. of Pittsburgh v. Firemen’s Insurance Co. of Baltimore (1908) 209 U.S. 326. A proportionate policy of reinsurance stipulated (p. 328) that:

“11. Each entry under this compact . . . shall be subject to the same conditions, stipulations, risks and valuations as may be assumed by the said reinsured company under its original contracts hereunder reinsured, and losses, if any, shall be payable pro rata with, in the same manner, and upon the same terms and conditions as paid by the said reinsured company under its contracts hereunder reinsured, and in no event shall this company be liable for an amount in excess of a ratable proportion of the sum actually paid to the assured or reinsured . . .”

After the great fire in Baltimore of 1904 the direct insurer became insolvent and could not pay more than 55 cents in the dollar, and therefore was unable to satisfy claims under its policies unless it could first recover from the reinsurer. The Supreme Court of the United States held that payment by the reinsured was not a condition of recovery under  [*389]  the reinsurance. Delivering the opinion of the court Peckham J. stated, at p. 336:

“We agree with the court below, that the language of the eleventh subdivision, taken in connection with the fact that it is used in a contract designated by the parties as one of reinsurance, means that the reinsuring company shall not pay more than its ratable proportion of the actual liability payable on the part of the reinsured, after deducting all liability of other reinsurers. To hold otherwise is to utterly subvert the original meaning of the term reinsurance and to deprive the contract of its chief value. The losses are to be payable pro rata with, in the same manner and upon the same terms and conditions as paid by the reinsured company under its contracts. This means that such losses, payable pro rata, are to be paid upon the same condition as are the losses of the insurer under its contract. . . . [This] . . . does not mean there must be an actual payment of such liability by the insurer before it can have any benefit of the contract of reinsurance which is made with defendant.”

In the second case, Fidelity & Deposit Co. v. Pink (1937) 302 U.S. 224 the contract was in very different terms. It stipulated that the reinsurer’s proportionate share of the loss should be paid to the reinsured upon proof of payment by the reinsured, and on tender of documents in support. It was furthermore stipulated that the reinsured might give the reinsurer prospective notice of its intention to pay on a certain date, and might require the reinsurer to put its share of the loss in the hands of the reinsured by that date. Distinguishing the Allemannia case, without differing from the statement of general principle therein contained, the Supreme Court held that on this occasion the contract was effective to make prior payment a condition precedent to liability.

Finally, in Stickel v. Excess Insurance Co. of America (1939) 23 N.E.2d 839, an ultimate net loss clause defined that term (p. 841) as “the sum actually paid in cash in settlement of losses for which the company is liable, after making proper deductions . . .” Founding on the language of the particular policy in question, the Supreme Court of Ohio found the case closer to Pink than to Allemannia, and held that once again actual disbursement was a condition precedent to recovery.

There was some suggestion in argument that there is an inconsistency between these cases. On examination I can detect none. Even the brief account given above is sufficient to make the individual decisions perfectly understandable. Whether they were all right it is unnecessary and inappropriate to consider; and it is of course true that the Allemannia case, 209 U.S. 326 was concerned with proportionate insurance, whereas the present is not. What it is permissible to say however is that the brief statement of general principle in that case accords with the law as it has been understood for many years, in common law jurisdictions and elsewhere. I can see nothing in these cases to cast doubt on the opinion which I have expressed as to the effect of the present policy.

Finally, there are the inferences about the purpose of the words “actually paid” which may be drawn from the history of the “follow settlements” clause. The matter is fully developed in the speech of my  [*390]  noble and learned friend, Lord Hoffmann. If I own to hesitation in adopting this as a direct answer to the problem it is because the historical materials presented in argument are incomplete, and subsequent reading has not filled the gaps. It is however clear that in the long time-frame of the insurance industry excess of loss reinsurance is comparatively modern, probably dating from transactions arranged by C.E. Heath in the United States in the last two decades of the 19th century. It was not until after the Baltimore fire that the need for an excess of loss non-proportionate cover written on a treaty basis became obvious. Such cover would of course need to provide for a means of ascertaining the point at which the reinsurance (or its first layer) attached; equally important however, was that this determined the amount of the reinsured’s retention, always a matter of prime importance when writing reinsurance. I think it a reasonable surmise that this retention was expressed in terms of net rather than gross. It is likely therefore that there was from the start some form of ultimate net loss clause in American excess of loss policies. Given that the Allemannia proportional reinsurance, effected in 1903, already included these words, I think it as likely that they were simply copied into excess of loss policies, as that they were deliberately included to combat a puzzling English decision some 20 years old, not referred to at all in the report of the Allemannia case, and not yet the subject of acute controversy even in England. This is however surmise but it is possible to say with some confidence that there is nothing in the available history to suggest that the words “actually paid” were and are included in order to create a condition precedent.

There is one final point, directed to the wording of this particular policy. It will be recalled that clause 2(c) defined net loss as “the sum actually paid . . . after making deductions for all recoveries [etc] whether collected or not.” There is a discontinuity here, if the syndicates are right. There is good reason why the provisional ascertainment of the effect which the losses will have on the reinsured layer should be made in the light of forecasts about the funds which will be transferred out, and the funds which will be transferred in, on future occasions before the ultimate net loss is finally ascertained, but I can see no reason why uncollected funds should be used as a contra sum at a time when through the absence of payment under the inward policies there is nothing against which to set them. Here again, I do not regard the point as conclusive, but it does reinforce the solution at which I have independently arrived.

For these reasons therefore I consider that the interpretation given by Mance J., and Simon Brown L.J. to policy no. X 20693/5386 was correct. This makes it unnecessary to consider the alternative line of reasoning which led Nourse L.J. to conclude in favour of Charter. The position under the second policy is acknowledged to be the same.

There remains the aviation policy. There are differences between this and the first two policies which might for other purposes be important. Mance J. has drawn attention to some of them. But in my opinion none of them bear on the present dispute, and the reasoning which I have proposed applies equally to all three contracts.

In these circumstances I would dismiss the appeal.  [*391]

LORD HOFFMANN.  My Lords, this appeal turns upon the construction of a standard clause known as the ultimate net loss (“U.N.L.”) clause which is in common use in the London excess of loss reinsurance market. Although the action concerns three particular policies of reinsurance written on behalf of two Lloyd’s syndicates, it raises an issue which affects the whole reinsurance market.

The relevant provisions are set out in the speech of my noble and learned friend, Lord Mustill, and I need not repeat them. The question is whether the words “actually paid” mean that the liability of the reinsurers is limited to the sum in respect of which Charter Reinsurance has discharged its liabilities in respect of the risks which it insured. Mr. Sumption says that this is the natural meaning of the words. There is nothing in the context which requires them to be given a different meaning and that is the end of the matter.

I think that in some cases the notion of words having a natural meaning is not a very helpful one. Because the meaning of words is so sensitive to syntax and context, the natural meaning of words in one sentence may be quite unnatural in another. Thus a statement that words have a particular natural meaning may mean no more than that in many contexts they will have that meaning. In other contexts their meaning will be different but no less natural.

Take, for example, the word “pay.” In many contexts, it will mean that money has changed hands, usually in discharge of some liability. In other contexts, it will mean only that a liability was incurred, without necessarily having been discharged. A wife comes home with a new dress and her husband says “What did you pay for it?” She would not be understanding his question in its natural meaning if she answered “Nothing, because the shop gave me 30 days’ credit.” It is perfectly clear from the context that the husband wanted to know the amount of the liability which she incurred, whether or not that liability has been discharged.

What is true of ordinary speech is also true of reinsurance. In re Eddystone Marine Insurance Co.; Ex parte Western Insurance Co. [1892] 2 Ch. 423 the policy contained the form of reinsurance clause then in common use “and to pay as may be paid thereon.” As in this case, the reinsured company was in insolvent winding up and could not pay its debts. Stirling J. said, at p. 427, that the policy did not mean that the liability should have been discharged. They meant only that “the payment to be made on the reinsurance policy is to be regulated by that to be made on the original policy of insurance.” In other words, the clause is concerned with the amount of liability and is indifferent to whether or not it has been discharged.

But, said Mr. Sumption, there is the word “actually.” Stirling J. might have been willing to accept that paid could in some artificial or figurative sense mean “liable to be paid.” But the word “actually” was surely added to make it clear that money must have changed hands. “Actually paid” said Mr. Sumption, meant actually paid.

One speaks of something being “actually” the case to point a contrast; perhaps with what appears to be the case, or with what might be the case, or with what is deemed to be the case. The effect of the word therefore  [*392]  depends upon the nature of the distinction which the speaker is wanting to make. This can appear only from the context in which the phrase is used. It is artificial to start with an acontextual preconception about the meaning of the words and then see whether that meaning is somehow displaced. The context might indicate that the word was used to reverse the ruling in the Eddystone case and require the liability of the reinsured to have been discharged. On the other hand, it might suggest that a different contrast was intended.

To revert to my domestic example, if the wife had answered “Well, the dress was marked £300, but they were having a sale,” and the husband then asked “So what did you actually pay?” she would again be giving the question an unnatural meaning if she answered “I have not paid anything yet.” It is obvious that the contrast which the husband wishes to draw is between the price as marked and the lower price which was charged. He is still not concerned with whether the liability has been discharged. This is not a loose use of language. In the context of the rest of the conversation, it is the natural meaning.

What then is the context? Is the draftsman wanting to draw a contrast with the meaning given to “paid” in the Eddystone case [1892] 2 Ch. 423 or does he have some other contrast in mind? My noble and learned friend, Lord Mustill, has analysed the structure of the policies and for the reasons which he gives, I agree that the context points to a wish to emphasise the net character of the liability as opposed to what, under the terms of the policies, the liability might have been.

I think that these conclusions are reinforced by the history of reinsurance clauses. Contracts of reinsurance were unlawful until 1864. Such a contract is not an insurance of the primary insurer’s potential liability or disbursement. It is an independent contract between reinsured and reinsurer in which the subject matter of the insurance is the same as that of the primary insurance, that is to say, the risk to the ship or goods or whatever might be insured. The difference lies in the nature of the insurable interest, which in the case of the primary insurer, arises from his liability under the original policy: see Buckley L.J. in British Dominions General Insurance Co. Ltd. v. Duder [1915] 2 K.B. 394, 400.

The difference in the nature of the insurable interest does however mean that, insurance being a contract of indemnity, the amount recoverable will not necessarily be the same as under the primary insurance. For example, the liability of the primary insurer will not necessarily be for the whole loss suffered by the original insured but may be subject to exceptions and limitations. His net outlay can also be reduced by recoveries under his right of subrogation. It therefore became customary in the market to have a special clause or clauses which defined the extent of the reinsurer’s liability. It appears that the most commonly used form in the early years of reinsurance was to add the words “Being a reinsurance, subject to all clauses and conditions of the original policy or policies, and to pay as may be paid thereon:” see McArthur, The Contract of Marine Insurance, 2nd ed. (1890), p. 332 and the form of policy in Uzielli & Co. v. Boston Marine Insurance Co. (1884) 15 Q.B.D. 11, 12.  [*393]

As construed by the courts, however, the phrase “and to pay as may be paid thereon” disappointed the expectations of the market on both sides. The original insurers assumed that it meant that if they agreed in good faith to pay under the original policy, they would be able to recover without having to prove their own legal liability. Reinsurers assumed that whatever the loss of the original insured might be, their liability would not exceed the net outlay of the reinsured, after taking all recoveries into account. Both assumptions were to prove false.

The story of how the expectations of original insurers were disappointed by the decision of Mathew J. in Chippendale v. Holt (1895) 1 Com.Cas. 197 and the subsequent development of the “follow settlements” clause to restore what had been thought to be the effect of the old clause has been told more than once, including by Scrutton L.J., who was junior counsel in Chippendale v. Holt, in Gurney v. Grimmer (1932) 44 Ll.L.R. 189, 192-194. (For subsequent developments, see Robert Goff L.J. in Insurance Co. of Africa v. Scor (U.K.) Reinsurance Co. Ltd. [1985] 1 Lloyd’s Rep. 312.)

The second assumption, on the part of reinsurers, had however been shaken by an even earlier decision. In Uzielli & Co. v. Boston Marine Insurance Co., 15 Q.B.D. 11 the defendants were reinsurers of the reinsurers of the Rose Middleton, which had been insured in the sum of £1,000. The ship went aground and the owners gave notice of abandonment to the original underwriters. The underwriters disputed the validity of the notice but eventually settled the claim for 88 per cent. But they also spent more money in getting the ship off th  [*350] e rocks than they eventually realised in selling her. The result was that they incurred a total loss of 112 per cent. They recovered the additional sum from the plaintiffs, their reinsurers, under a “sue and labour” clause in the policy which entitled them to recover such expenditure reasonably incurred by the insurers or their “factors or servants or assigns.” The plaintiffs in turn claimed £1,120 from the defendants. Matthew J. held that there had been a constructive total loss, that the reinsurers were entitled to add the expenditure of the underwriters on salvage under the “sue and labour” clause and gave judgment for £1,120. The reinsurers appealed and the Court of Appeal held that, as against the defendants, the “sue and labour” clause did not cover expenditure by the original underwriters because they were not the “factors or servants or assigns” of the first reinsurers. One might have thought that the result would be that the plaintiffs could recover only the 88 per cent. of the £1,000 for which the claim of the shipowner had been settled. That was what had been paid on the original insurance policy. Instead, however, the court substituted a judgment in favour of the underwriters for £1,000.

The Uzielli case caused a good deal of puzzlement in the market and among marine insurance lawyers. Mr. McArthur (The Contract of Marine Insurance, p. 335) said that “as the facts in the case were peculiar, no general principle can be deduced from the decision.” In Western Assurance Co. of Toronto v. Poole [1903] 1 K.B. 376, 387-388, Mr. Hamilton K.C. and Mr. Scrutton K.C. offered Bigham J. different explanations of the case, neither of which he found satisfactory. In British Dominions General Insurance Co. Ltd. v. Duder [1915] 2 K.B. 394, Buckley L.J. said that he could not find any principle in the case: p. 402. Pickford L.J. likewise  [*394]  said, at p. 405, that it was very hard to understand and Bankes L.J. was similarly perplexed: see p. 413. Although the principle of indemnity is fully reaffirmed in Duder it would not be surprising if the market felt nervous that the House of Lords might one day see some light in Uzielli which had eluded other judges since the time it was decided.

Although the commercial history of the matter is not as well documented as that of the “follow settlements” clause, it is clear that the formula “pay as may be paid thereon” disappeared from standard forms of reinsurance. The objects which it had sought to achieve on behalf of the original insurers were taken over by the follow settlements clause. It does not seem unreasonable to infer that its function in delimiting the liability of the reinsurers was taken over by the ultimate net loss clause. The U.N.L. clause shows throughout a preoccupation with ensuring that the reinsurer cannot be called upon to pay more than the reinsured has been required to pay. In Uzielli the words “pay as may be paid” had proved ineffective to achieve this result, even though they had been thought apt to do so. In his argument in Duder’s case [1915] 2 K.B. 394, 398 Mr. Roche K.C., arguing for a similar result to that in Uzielli, said plaintively but truthfully, that the words “pay as may be paid thereon:” “weakened the case of the plaintiffs, and yet this court held that they could recover the full 100 per cent. and not merely the 88 per cent. for which they had settled the claim against them.” It would not therefore be surprising if underwriters thought that if “paid” was not good enough to satisfy the courts, “actually paid” might drive the point home.

The U.N.L. clause in the policies before the House has been traced back in unaltered form to the early 1930s and I would not be surprised if it went even further back than that. The words “actually paid” can be found in the policy considered in Allemannia Insurance Co. of Pittsburgh v. Firemen’s Insurance Co. of Baltimore, 209 U.S. 326, where they were given the construction which I suggest in this case.

I find further support for my view in the fact that the U.N.L. clause has been thought suitable for use in the London excess of loss reinsurance market. There are certainly forms of reinsurance in which it may be commercially appropriate to make discharge of his liability by the reinsured a condition of the liability of the reinsurer. It may be, as in cases of mutual insurance, that the reinsurer has an interest in making certain that the reinsured maintains sufficient liquid assets to meet his liabilities. Or it may be a protection against fraudulent claims. But the London excess of loss market operates on the assumption that a reinsurance programme will relieve the insurer of the burden of having to pay claims covered by the reinsured layers. The regulation of insurers in this country uses a test of solvency which treats reinsurance cover as a proper deduction from the insurer’s liabilities. None of this would make sense if the insurer had first to satisfy the claim out of his own resources before he could call upon his reinsurers to pay.

Mr. Sumption suggested a stratagem which insurers might use to avoid having to pay the whole claim themselves. They could pay a part, even a very small part, of the reinsured liability and then, having to this extent actually paid, they could call upon the reinsurer to reimburse them. Having thus primed the pump, they could by successive strokes draw up  [*395]  the full amount from the reinsurance well. I cannot imagine that the parties could ever have contemplated such a strange procedure and one is bound to ask what commercial purpose the reinsurer could have expected to achieve by being able to insist upon it.

Considerations of history, language and commercial background therefore lead me to the conclusion that the word “actually” in the U.N.L. clause is used to emphasise that the loss for which the reinsurer is to be liable is to be net and that the clause does not restrict liability to the amount by which the liability of the reinsured for the loss has been discharged. I think that this is the natural meaning of the clause.

In conclusion I would like to pay tribute to the judgment of Mance J. which deals comprehensively with the issues and all the relevant authorities and with which I am in full agreement. I would dismiss the appeal.