Tokio Marine Europe Insurance Ltd v Novae Corporate Underwriting Ltd

In autumn 2011 Thailand suffered severe floods. Unusually heavy rains throughout the rainy season caused the Chao Phraya River and its tributaries in central Thailand to swell and burst their banks in October 2011. The floodwater swept through Thailand from north to south, flooding over 20,000km of farmland and causing an estimated US$50 billion of damage.

Among the properties damaged were shops and distribution centres owned by a subsidiary of Tesco plc, Ek Chai Distribution System Co Ltd (“Ek Chai”).

Claims of approximately £125 million for property damage and business interruption were made by Tesco/Ek Chai who were insured under a global Master Policy issued by ACE European Group Ltd (“Ace Europe”) and local policies issued by local ACE entities pursuant to the Master Policy. The claim was adjusted by VRS Vericlaim and on 20 February 2012 the claim was settled in the amount of £82.5 million, less a deductible of £100,000 and a self-insured retention of £2.4 million.

ACE Europe and the relevant local ACE companies were reinsured under a facultative reinsurance (“the Reinsurance”) under which various reinsurers agreed to reinsure a 55% share of losses. The Claimant, Tokio Marine Europe Insurance Limited (“TMEI”), subscribed a 12.5% share to the Reinsurance and, in turn, purchased from the Defendant, Novae Corporate Underwriting Limited (“Novae”), a facultative excess of loss reinsurance (“the Retrocession”).

TMEI has paid its share of the settlement sum and in these proceedings seeks to recover from Novae its 12.5% share of losses excess of £53 million (i.e. £3,125,000). Novae denies liability on various grounds.

The trial of five preliminary issues which it is hoped will resolve the parties’ disputes was ordered. The issues are essentially questions of construction including in particular issues as to the scope and effect of the follow settlements clause. The trial was conducted on documents and without the need for oral evidence. Both parties made extensive written submissions and I have drawn on those submissions, with adaptations and amendments, in preparing this judgment, particularly in relation to matters of common ground and in setting out the parties’ arguments.

The parties produced a Statement of Agreed Facts which largely sets out the factual background.

The properties affected by the flooding in Thailand were insured under a local policy PIP0020253 AR-40124 - issued by ACE INA Overseas Insurance Company Limited (‘ACE INA’) and a Master Policy reference UKFRIC38309.11(the ‘Master Policy’) - issued by ACE Europe in respect of physical damage to the property insured and business interruption losses.

The insurance of the properties in Thailand was part of a multinational insurance programme by which properties owned and operated by Tesco plc, its affiliates, subsidiaries and joint ventures (together ‘Tesco’) in various territories throughout the world were insured under the Master Policy and, with certain exceptions, also under local policies issued by companies in the ACE group. For properties in those territories in which a local policy was issued, the Master Policy responded solely on a Difference in Conditions (“DIC”)/Difference in Limits (“DIL”) basis.

The Master Policy covered Tesco’s properties in the United Kingdom, Republic of Ireland, France, Hungary, Poland, Czech Republic, Slovakia, Thailand, South Korea, Malaysia, Hong Kong, Japan, Turkey, USA, China and India and local policies, issued by the respective local companies, covered Tesco’s properties in the same territories, except the United Kingdom, China and India.

ACE Europe and the relevant local companies were reinsured by ELH Insurance Limited (‘ELH’) up to £2.5 million each and every occurrence excess of the deductibles under the Master Policy and local policies.

ACE Europe and the relevant local companies were further reinsured under the Reinsurance by which various reinsurers, including TMEI, agreed to reinsure the cedants for, in total, a 55% share of losses sustained by Tesco during the period of 1 October 2011 to 28 February 2013 excess of the deductibles and captive/self-insured retention up to a limit of £100 million each and every occurrence excess of original policy deductibles.

TMEI subscribed a 12.5% share to the Reinsurance and, in turn, purchased from Novae the Retrocession.

John McGowan of TMEI’s broker AON Benfield supplied Novae’s underwriter, Richard Ikin, with the information contained in his e-mail of 19 August 2011 and its attachments and further spreadsheets of property values revised on 4 October 2011. Mr Ikin bound the Retrocession on 4 October 2011.

Under the Retrocession, Novae agreed to reinsure TMEI for 12.5% (its written line under the Retrocession) of losses sustained by Tesco during the period of 1 October 2011 to 28 February 2013 in excess of £53 million each and every “Loss Occurrence” subject to a limit of £25 million each and every “Loss Occurrence”, and subject to the terms and conditions of the Retrocession.

The “Reinsurance Conditions” of the Retrocession contain the term: “Following Original Policy Wording Reference Number: UKFRIC38309.10”. This is an error and should read “Following Original Policy Wording Reference Number: UKFRIC38309.11”

The Retrocession contains a follow the settlements clause in the following terms:

“This Contract is subject in all respects (excluding the rate and/or premium hereon and subject always to the Limits Reinsured hereon and except as otherwise provided herein) to the same terms, clauses and conditions as original and without prejudice to the generality of the foregoing, Reinsurers agree to follow all settlements (excluding without prejudice and ex-gratia payments) made by original Insurers arising out of and in connection with the original insurance and to bear their proportion of any expenses incurred whether legal or otherwise in the investigation and defence of any claim hereunder in addition to limits hereunder.”

Between 6 October 2011 and 6 November 2011 a number of Tesco stores and distribution centres were affected by the widespread flooding in Thailand as a result of which Tesco suffered losses.

ACE Europe and ACE INA appointed loss adjusters VRS Vericlaim to adjust Tesco’s claim and RGL Forensics to assist in the adjustment of Tesco’s business interruption claim.

VRS Vericlaim produced a number of reports. The final report dated 18 May 2012 records that the total claim made by Tesco and/or Ek-Chai under the Master Policy and the Local Policy for physical damage to property insured under those policies and business interruption was £125.32 million.

In 2011 Tesco occupied nearly 900 properties throughout Thailand and the floods affected a total of 165 of those properties.

The geographical distribution of the affected properties listed by province and approximate distance from Central Bangkok was as follows:

On 20 February 2012, a settlement agreement was entered into between Tesco plc, Ek-Chai, ACE Europe, ACE INA and ELH to settle Tesco’s claim under the Master Policy and the local policy for £82.4 million after application of one deductible of £100,000.

On 17 February 2012, Philip Dick, TMEI’s Chief Operating Officer, sent Andy McAvan of ACE an email in, inter alia the following terms:

“Given the urgency involved in providing you with a position from TMEI, I confirm that TMEI is prepared to support the proposed settlement with Tesco for its share of GBP80m net of one GBP2.5m deductible subject to the following points:

1.

We assume this is the intent of the proposed settlement with Tesco but, for the avoidance of doubt, our agreement to the settlement is on the basis that this is a ‘full and final’ settlement in respect of the Thailand flood event and that TMEI is fully released from all further liability to ACE in respect of damage to Tesco properties caused by the recent floods.

2.

Our agreement is on the strict understanding that it shall not, in any respect, be treated by ACE as a precedent to support any future claims it may make under its reinsurance contract with TMEI.

3.

TMEI’s support is also on the basis that the premium due will first be paid – a point we understand has already been agreed with QBE.”

The settlement amount was split £57,950,152 to the local policy and £24,449,848 to the Master Policy. Of the net settlement sum of £82.4m, ELH’s liability to ACE Europe and ACE INA (together referred to as “ACE”) was £2.4 million. The net settlement for the purposes of the Reinsurance was therefore £80 million. TMEI says it has paid in full its share of the settlement due under the Reinsurance.

The amount of the settlement concluded on 20 February 2012 that was apportioned to the Master Policy does not exceed the retention in the Retrocession.

The five preliminary issues are as follows:

Reinsurance contracts are to be construed by adopting the same principles applicable to the construction of commercial contracts - see, for example, Vesta v Butcher[1989] 1 Lloyd’s Rep 331 at p345-346.

The general approach is to consider what a reasonable person having all the background knowledge which would have been available to the parties at the time of contracting would have understood themselves to be agreeing in using the language in the contract: see Lord Steyn’s summary of the approach to construction of a commercial instrument in Sirius Insurance Co v FAI General Insurance Ltd [2005] Lloyd’s Rep IR 294 at [18]; and the Supreme Court’s recent statement on the correct approach to contractual construction in Rainy Sky SA v Kookmin Bank[2011] 1 WLR 2900at [14], [21].

If there are two possible constructions of a document or term, a court is entitled to prefer the construction which is more consistent with “business” or “commercial” common sense: see theRainy Sky case at [21], [30]. By contrast, where the parties have used unambiguous language, the court must apply it: theRainy Sky case at [23].

The relevant principles are set out by Aikens LJ in the recent case of BMA Special Opportunity Hub Fund Ltd & Ors v African Minerals Finance Ltd.[2013] EWCA Civ 416 at [24], which was accepted by the parties as a fair summary:

“There has been considerable judicial exposition of these principles by the House of Lords and the Supreme Court in recent years.⁷ [Footnote 7 refers to Chartbrook Ltd v Persimmon Homes Ltd [2009] 1 AC 1101; Re Sigma Finance Corp [2010] 1 All ER 571 and Rainy Sky SA v Kookmin Bank [2011] 1 WLR 2900] There is no point in my going over the same ground again at any length. The court’s job is to discern the intention of the parties, objectively speaking, from the words used in the commercial document, in the relevant context and against the factual background in which the document was created. The starting point is the wording of the document itself and the principle that the commercial parties who agreed the wording intended the words used to mean what they say in setting out the parties’ respective rights and obligations. If there are two possible constructions of the document a court is entitled to prefer the construction which is more consistent with “business common sense,” if that can be ascertained. However, I would agree with the statements of Briggs J, in Jackson v Dear, [2012] EWHC 2060 at [40] first, that “commercial common sense” is not to be elevated to an overriding criterion of construction and, secondly, that the parties should not be subjected to “…the individual judge’s own notions of what might have been the sensible solution to the parties’ conundrum”. I would add, still less should the issue of construction be determined by what seems like “commercial common sense” from the point of view of one of the parties to the contract.”

A particular issue which arises in relation to the construction of reinsurance contracts is the extent to which it can be presumed that the reinsurance or retrocession contract is intended to be back-to-back with the underlying insurance or reinsurance.

The courts have frequently emphasised that insurance and reinsurance contracts are separate and independent bargains entered into between different parties, and potentially also covering different risks: see, for example, Wasa International Ltd v Lexington Insurance Co [2010] 1 AC at [32]-[33] (per Lord Mance); Equitas Limited v R&Q Reinsurance Company (UK) Limited [2009] EWHC 2787 (Comm) at [46] (per Gross J).

In the context of proportional reinsurance in which the reinsured cedes a share of the risk to the reinsurer, it will frequently be inferred that it is the intent of the parties that the insurance and reinsurance are to be back-to-back. As stated by Lord Collins in Wasa v Lexington at [58]:

“..(3) In principle the relevant terms in a proportional facultative reinsurance – and in particular those relating to the risk – should be construed so as to be consistent with the terms of the insurance contract on the basis that the normal commercial intention is that they should be back to back”..

In the case of non-proportional reinsurance contracts, such as the Retrocession, there is, however, no presumption that the reinsurance contract is intended to be back-to-back with the underlying insurance contract. This was explained by Lord Mustill in Axa Reinsurance (UK) Ltd v Field[1996] 1 WLR 1026 at 1034A-H when addressing the first of three “themes” of the Court of Appeal judgment with which he disagreed:

“The first is an assumption that where a direct insurer takes out reinsurance, and where both policies contain provisions enabling the amount of losses to be added together, the parties are likely to have intended their effect to be much the same. This assumption may very well be correct where the reinsurance is of the proportionate kind, under which the reinsurer is sharing the risk assumed by the direct insurer. In such an event it is indeed likely that the treatment of multiple losses, and hence the outcome of the parallel contracts, was meant to be the same. But where a reinsurer writes an excess of loss treaty for a layer of the whole account (or the whole of a stipulated account) of the reinsured I see no reason to assume that aggregation clauses in one are intended to have the same effect as aggregation clauses in the other. The insurances are not in any real sense back-to-back. Thus, for example, a direct insurer may issue many policies on terms as to deductible and limit of liability which he can fix according to his knowledge of the policyholders and of the likely size and incidence of the kind of casualties which are insured. The financial outcome of these policies will depend on other factors besides the total monetary amount of the valid claims made. Thus, if many of the policyholders make a large number of small claims, comparatively few of them will exceed the deductible, and the underwriter's gross exposure will be small. At the other extreme, if the claims are large but few, most of them will be cut off by the upper limit, and again the exposure may be quite small. But if there are many claims of medium size the underwriter may find himself carrying them all in full. If when writing his policies he foresees that this could happen, he will consider limiting his liability under an individual policy by reference to aggregate claims made during the policy year, as well as by the size of each individual claim. Or, again, if the likelihood is that even when there are numerous losses a group or groups of them will share a more or less distant common origin, it may be prudent to impose not only a limit per claim but also a limit per group. These matters form an element in determining not only the premium charged, but also the amount and the nature of the reinsurance which it is prudent for the direct insurer to carry.

The strategy of the underwriter who takes a line on a layer of an excess of loss treaty is not necessarily the same. He cannot rate the individual policyholders and individual risks directly, and must take a much broader view. For him, the relationship between the inward and outward policies is essential to profitability. Not only the limits for each loss, but the aggregation of losses, both causally and in other ways, and the numbers and circumstances of permitted reinstatements, make all the difference. It is, I believe, plain that the elements of the prudent underwriter’s judgment when writing policies of this kind need not be at all the same as if he were writing the underlying business direct. In particular, according to circumstances it may suit him, but not the direct underwriter, to have an aggregation clause which sweeps up many losses into one aggregate; or the opposite may be the case. It is simply impossible to generalise, and I have no predisposition to start the inquiry by assuming that parties intended the provisions for aggregation in the direct policy and in the reinsurance treaty to be the same. The natural way to achieve that result is to make sure that the aggregation clauses are the same.”

The Retrocession provided as follows:

Novae stressed the references made in the terms delineating the scope of the cover to the “Original Policy Wording” – i.e. the Master Policy – and the fact that no reference is made to local policies. It submitted that this made it clear that the Retrocession only provided cover for losses falling within the Master Policy, that this was the only available meaning of the words used and that there was therefore no room for a consideration of business or commercial common sense.

Focusing first on the wording of the Retrocession, I do not agree with Novae that it is to be construed in the limited manner for which it contends. The coverage provided is expressed in general and unqualified terms. In essence it covers TMEI in respect of its exposure for “all risks of Direct Physical Loss, Destruction or Damage” suffered by Tesco in respect of “all real and personal property of every kind and description” belonging to Tesco including business interruption. TMEI’s exposure to such losses is not limited according to whether it arises under the Master Policy or local policies. It is correct that the property covered, the insured perils and the insured are as defined in the Master Policy, but this is for definitional purposes and does not detract from the general coverage provided. It is also correct that the applicable terms and conditions are those of the Master Policy but those terms do not determine coverage. In this case there was loss and damage suffered by Tesco to insured property belonging to it and resultant business interruption from an insured peril within the terms of the Master Policy. TMEI’s exposure to those losses arose both under the Master Policy and the local policies, but Novae’s liability under the Retrocession is neither expressly nor impliedly limited to the former.

Novae stressed the contrast between the way in which the local policies are dealt with in the Reinsurance and the Retrocession. In particular, the “Interest” clause in the Reinsurance states that the Reinsurance was provided in respect of “Real and Personal Property and Business Interruption as defined in Original Policy(ies) wording” and the Reinsurance expressly defined “Original Policy(ies)” to include the Master Policy and “all locally issued policies reinsured by or issued as part of the Master Policy” and there were other references to the local policies in the wording agreed. In the Retrocession there are no such references. Since this is a document which did not cross the line to Novae it is of limited assistance. At most it shows that there are other and perhaps clearer ways in which the matter could be addressed. It does not alter the natural meaning of the words actually used in the Retrocession.

The correctness of TMEI’s construction of the Retrocession is borne out when regard is had to the factual matrix and the presentation materials.

The Retrocession refers to “Information” provided by TMEI. This is a one-page sheet of property damage and business interruption values at risk produced by Tesco. These figures show the full ‘ground-up’ values at risk in each location around the world, including in Thailand. There would be little reason for this information to be provided or noted if the Retrocession were not providing cover in respect of the local policies, because the majority of the ‘value at risk’ would be protected under the local policies and not the Master Policy.

Further, this information was expressly provided in order to help with “the rating of this risk” and the “assessment of this risk” respectively. This information would have been of little assistance if the intention had been that cover under the Retrocession should be restricted to exposure arising under the Master Policy. In such a case the relevant information would have been figures on a DIC/DIL basis (although the Court was informed by TMEI that there were in fact no difference in limits under the local policies).

Novae’s answer to this was that this information would have been difficult to provide and it would still have been helpful for underwriters to know what the total values locally were, even if their exposure was only on a DIC/DIL basis. However, that does not explain how underwriters could meaningfully assess and rate the risk without some DIC/DIL information.

It is also striking that there is nothing in the covering email from brokers or the presentation itself that suggests any such restriction to Master Policy coverage. I agree with TMEI that if such a restriction had been intended it would surely have been mentioned – apart from anything else it would undoubtedly have had a significant effect on the pricing if the risks covered were restricted as Novae now claims.

The correctness of TMEI’s construction of the Retrocession is further borne out when regard is had to commercial or business sense considerations.

If Novae’s construction were correct it would lead to a radical mismatch between the cover provided under the Reinsurance and the cover provided under the Retrocession. It would not simply be a question of the cover not being back to back; the coverage provided would be fundamentally different. There is no obvious reason why the parties should have intended that the cover provided by the Retrocession should be so very different from the cover provided by the Reinsurance. It makes little commercial sense for TMEI to have purchased reinsurance only in respect of exposure arising to it under the Master Policy and not the same exposure arising under local policies: whether the claims are treated as arising under the local policies or Master Policy makes no difference to TMEI.

Further, Novae’s construction would render the cover provided by the Retrocession unlikely to be called upon in respect of losses outside the UK. It was a term of the Master Policy that a local policy must be issued in every jurisdiction (except the UK) covering the property and interests located in that jurisdiction if such cover is obtainable in that jurisdiction (General Condition 6). The Master Policy acts as excess/difference in conditions cover on top of the local policies (General Condition 7). The limit of cover under the original insurance (i.e. the Master Policy and local policies combined) and the Reinsurance is £100 million any one Occurrence. The Retrocession provides cover in respect of £25m xs £53m any one Occurrence. On Novae’s case, the Retrocession would only respond if there were a loss to the Reinsurance that fell in the range £53m - £78m but which was made up purely of losses under the Master Policy. Given that the maximum amount of any loss to the Reinsurance is £100m, it is difficult to conceive of any claim under the Master Policy from a jurisdiction outside the UK that will not have generated a significant loss to a local policy. This means that there is little prospect of any loss to the Reinsurance on property or interests outside the UK that exceeds £53m composed purely of losses under the Master Policy. The Retrocession would therefore effectively be restricted in its application to losses on property or interests in the UK, a commercially unlikely result for a Retrocession purportedly providing worldwide coverage.

Finally, if it really had been intended to restrict the Retrocession coverage in this way so that it was fundamentally different to that provided under the Reinsurance one would expect that to be clearly spelt out, both in the Retrocession and in the presentation materials – it is not.

For all these reasons, I conclude that the answer to the first preliminary issue 1 is “Yes”.

Coverage under the original insurance is provided for up to £100 million any one “Occurrence”.

“Occurrence” is then defined in the Master Policy wording to mean “any one Occurrence or any series of Occurrences consequent upon or attributable to one source or original cause.” The same wording applies to the Reinsurance.

The Retrocession provides that the sum reinsured is “GBP 25,000,000 each and every Loss Occurrence in excess of GBP 53,000,000 each and every Loss Occurrence which in turn is excess of Original Policy Deductibles.”

The Retrocession is expressly subject to “the same terms, clauses and conditions as original” and follows “Original Policy Wording”, which is defined by reference to the Master Policy wording, and thereby incorporates the Master Policy definition of “Occurrence”.

The short issue of construction which arises is whether “Loss Occurrence” in the Retrocession means an “Occurrence” as defined in the incorporated Master Policy wording, or whether it has some different meaning.

It was TMEI’s case that there was not intended to be any difference between the meaning of the word “Occurrence” used in the original insurance and incorporated into the Retrocession, and the phrase “Loss Occurrence” used in the Retrocession. In other words, the addition of the word “Loss” was surplusage not intended to have any effect on the definition of the term. On any view, the addition of the word was not intended completely to alter the definition of the insured “Occurrence” and nullify the incorporation of the defined meaning of “Occurrence” from the original insurance.

It was Novae’s case that the fact that the Retrocession uses different aggregating words when defining the limits and retentions in that policy (i.e. “Loss Occurrence”, as opposed to “Occurrence”, the word used in the Master Policy and the Reinsurance) clearly demonstrates that TMEI and Novae are to be taken as having intended, viewing things objectively, something different to what the parties to the Master Policy and the Reinsurance had agreed.

As to what the parties did intend thereby, Novae submitted that the phrase “Loss Occurrence” is to be given the meaning ordinarily attributed to “occurrence” or “event” in the insurance context, namely it is “something which happens at a particular time, at a particular place, in a particular way” – see Countrywide Assured Group Plc v Marshall [2003] 1 Lloyd’s Rep IR 195 at 201, col. 1; Kuwait Airways Corporation v Kuwait Insurance Co. [1996] Lloyd’s Rep 664 at 685-686. On their case also the word “Loss” is therefore ultimately surplusage.

This is a potentially significant point. First, if “Loss Occurrence” is not to be construed in the same manner as the definition of “Occurrence”, then the Reinsurance Settlement will not bind Novae in relation to the application of that aggregation provision, as TMEI accepted. Secondly, if “Loss Occurrence” is taken to mean something which happens at a particular time, at a particular place, in a particular way, the likely consequence is that the Tesco losses were made up of a number of distinct “Loss Occurrences”, each of which does not exceed the £53 million retention in the Retrocession. Indeed, as this illustrates, if Novae’s construction is correct it is likely to deprive the follow settlements clause of much of its practical effect.

As a matter of language “Loss Occurrence” is clearly capable of having the same meaning as “Occurrence” and in my judgment that was what was intended here.

The context is a Retrocession to cover TMEI’s exposure to losses arising from Occurrences which have a defined meaning, which defined meaning is incorporated into the Retrocession.

Against that background if the parties had intended for a different type of Occurrence to be covered by the Retrocession they would surely have clearly spelt out i) that that was the intention and ii) what the different meaning was to be.

This is all the more so when one considers the fundamental change in coverage which Novae’s construction supposes to have been intended. The Reinsurance was to cover same source or cause Occurrences. The Retrocession was to be limited to same time/place/way Occurrences. No reason was advanced as to why the parties should wish to create such a major mismatch. I agree with TMEI that it makes far more commercial sense for the parties to adopt a consistent approach. The considerations set out in paragraph 46 above equally apply to this further suggested radical mismatch.

Further, there is nothing in the presentation materials to suggest that such a fundamental difference in coverage was being suggested, considered or contemplated.

On Novae’s case this fundamental change was intended to be carried out by the addition of the single word “Loss”, without any explanation of what that was to mean and even though, on its own case, the word ultimately adds nothing. Novae’s case is that “Loss Occurrence” means “Occurrence” but not in accordance with the defined meaning given to the word in the incorporated Master Policy wording, but some other unspecified and undefined meaning. I do not accept that this was likely to be or was the parties’ intention.

For all these reasons I conclude that the answer to Preliminary Issue 2 is that “Loss Occurrence” in the Retrocession is be construed in the same manner as the word “Occurrence” in the Master Policy.

The relevant clause provides:

TMEI submitted that it is plain that the relevant settlement is that made by ACE as the “original Insurers” under the Master Policy and local policies thereunder as the “original insurance”.

I agree that this is the natural meaning of the words, particularly when considered in their textual context. That includes the admitted fact that other references to “original” elsewhere in the slip, and indeed in the same clause, refer to ACE and the policies it issued. Further, later on in the same clause the term “Reinsured” is used, in contrast to “original Insurers”.

Novae submitted that this was sloppy drafting and that the only sensible meaning is that it is referring to TMEI, its contractual counterparty. I accept that follow settlement clauses often refer to an obligation on a reinsurer to follow the settlement of his immediate reinsured, of which the cases relied upon by Novae provide examples - Hill and others v Mercantile and General Reinsurance Co. Plc [1996] 1 WLR 1239 and Equitas v R&Q[2009] EWHC 2787 (“All loss settlements by the Reassured…). I also accept that there may be good reasons for preferring to follow the settlement of your counterparty rather than that of a third party.

In my judgment, however, those considerations cannot override what I consider to be the clear and unambiguous meaning of the words used in this Retrocession contract. Further, there are good commercial reasons why in this case the ‘settlement’ that Novae must follow, and in relation to which it is entitled to information, is the original settlement between the ACE insurers and Tesco. That is the “coalface” and the level at which the substance of the claim will have been thrashed out and coverage issues determined. It is much less likely that the discussions between TMEI and ACE, already at one remove from the loss and under a Reinsurance with its own follow settlements clause, will be of as much relevance.

For all these reasons I conclude that the answer to Preliminary Issue 3 is that Novae agreed to follow the settlements of ACE.

The determination of this issue involves a consideration of what was decided by the Court of Appeal in Assicurazioni Generali SpA v CGU International Insurance plc (“Generali”) [2004] 2 All ER (Comm) 114 and in particular what Tuckey LJ said at paragraph 18 of his judgment.

In order to resolve that issue it is necessary to consider in some detail the background to that decision.

As stated by Lord Mustill in Hill v Mercantile[1996] 1 WLR 1239at 1251F, there are only two rules in considering the meaning and effect of follow settlement clauses:

“First, that the reinsurer cannot be held liable unless the loss falls within the cover of the policy reinsured and within the cover created by the reinsurance. Second that the parties are free to agree on ways of proving whether these requirements are satisfied.”

The history of reinsurance is replete with examples of how the parties have agreed ways of so proving. The present case is concerned with an unqualified follow settlements clause such as that authoritatively considered by the Court of Appeal in Insurance Company of Africa v Scor Reinsurance [1985] 1 Lloyd’s Rep 312. In that case Robert Goff LJ described the effect of such a clause in the following terms (at p330):

“In my judgment, the effect of a clause binding reinsurers to follow settlements of the insurers, is that the reinsurers agree to indemnify insurers in the event that they settle a claim by their assured, i.e., when they dispose, or bind themselves to dispose, of a claim, whether by reason of admission or compromise, provided that the claim so recognized by them falls within the risks covered by the policy of reinsurance as a matter of law, and provided also that in settling the claim the insurers have acted honestly and have taken all proper and businesslike steps in making the settlement.”

The two underlined sections in the above passage are provisos to the application of unqualified follow the settlements clauses of the kind found in Scorand the Retrocession, and are frequently referred to as the ‘first and second provisos’ respectively.

There are other types of follow the settlement clauses which are qualified, such as the “double proviso” clause considered in Hill v Mercantile, which included the proviso: “providing such settlements are within the terms of the conditions of the original policies and/or contracts…and within the terms of this reinsurance”. These are clauses which, while they start out by suggesting that the reinsurer is to follow the settlements of the reinsured, then go on to upend the position by specifying that the reinsured must nevertheless prove that the claim is in fact covered by the insurance and/or reinsurance.

In relation to the Scor type of unqualified follow settlements clause and its first proviso, an important question arises as to what is meant by “the claim so recognized by them falls within the risks covered by the policy of reinsurance as a matter of law” and how that works in practice. This was considered by Evans J in Hiscox v Outhwaite (No.3) [1991] 2 Lloyd’s Rep 524 and by the judge at first instance and by the Court of Appeal in Generali [2003] 2 All ER (Comm) 425 (Mr Kealey QC) [2004] All ER (Comm) 114 (Court of Appeal).

In Hiscox, the reinsured sought to recover from reinsurers in respect of sums paid under the Wellington Settlement in relation to asbestos liabilities. It was common ground that the sums paid under the Wellington Settlement were not calculated by reference to the reinsured’s actual liabilities for specific claims but rather by reference to its market share, and that as a result it was inevitable that the sums paid by the reinsured included sums paid in respect of claims that were not covered by the insurance or the reinsurance. The reinsured sought to contend that the reinsurer was nevertheless bound to follow its settlement. Evans J held that it was not, because the claims were not even arguably covered by the insurance or reinsurance as a matter of law. He stated (at p530):

“The principle stated in Scor’s case is unexceptional, but its application gives rise to difficulty where the terms of the reinsurance contract are the same as those of the underlying insurance contract, and the reinsurer has agreed to follow the settlements of the original insurer/reinsured. This difficulty was recognized by the Court of Appeal of Hong Kong in Insurance Co. of the State of Pennsylvania v. Grand Union Insurance Co., [1990] 1 Lloyd’s Rep. 208. In the leading judgment, Hunter J.A. said this:

‘Two points, I think, have to be noticed about these two provisos. The first, I have no doubt, was very carefully worded and deliberately limited to the policy of reinsurance. Mr. Collins argues that where, as is usual, the policy of reinsurance refers to the terms of the original insurance, the reinsurer can look through to those terms and complain, as was sought to be done here, of breaches of condition in the underlying policy. I reject that. If Lord Justice Goff meant that, he would in Mr. Justice Mortimer’s words “be nullifying the conclusion that he had already reached”. I am satisfied that he meant no such thing. He was well aware that many settlements include compromises on liability and quantum and that to permit reinsurers to go back to an alleged strict construction of the policy would destroy the value of the clause. If there is any question as to the sufficiency or propriety of the settlement it arises under the second proviso.’

In my judgment, the reinsurer is always entitled to raise issues as to the scope of the reinsurance contract, and where the risks are co-extensive with those of the underlying insurance he is not precluded from raising such issues, even when there is a “follow the settlement” term of the reinsurance contract. Ultimately, this is the only sure protection which the reinsurer has against being called upon to indemnify the reinsured against payments which were not legally due from him to the original insured, however reasonable and businesslike the payments may have been. But this is subject to one proviso which I have already assumed in the Syndicate’s favour, and which is supported by the judgment of Hunter J.A. in the Grand Union case, quoted above. The reinsurer may well be bound to follow the insurer’s settlement of a claim which arguably, as a matter of law, is within the scope of the original insurance, regardless of whether the Court might hold, if the issue was fully argued before it, that as a matter of law the claim would have failed.” (emphasis added)

This judgment, and how the first Scor proviso works in practice, was then addressed in some considerable detail by Mr. Gavin Kealey QC, sitting as a Deputy High Court Judge, in Generali at first instance.

Generali involved a claim by Pirelli under a policy of works insurance which was (re)insured by Generali. The claim was settled and Generali sought to recover from its reinsurers. An issue arose between the parties as to the scope of the follow settlements clause. Generali applied for summary judgment on that issue. The question, as formulated by the Court of Appeal, was “to what extent does the first [Scor] proviso preclude reinsurers from raising coverage issues relating to terms which are the same in both the insurance and the reinsurance contracts?”

Mr. Kealey QC explained the words “claim so recognised” (used by Robert Goff LJ in Scor) as follows at [38]-[40]:

“38 The distinction between having to prove that an original loss falls within the cover provided by a contract of insurance and also by a contract of reinsurance, and having to prove that a claim that has been recognised by the insurers as falling within the cover provided by a contract of insurance also falls within the cover provided by a contract of reinsurance, is significant. In the former, one is examining what in fact happened and whether, on the basis of what actually happened, the insurers are liable to indemnify the assured under the contract of insurance and the reinsurers are liable to indemnify the insurers under the contract of reinsurance, according to their respective terms. In the latter, one is examining the claim recognised by the insurers by their settlement of it by admission or compromise and whether on that basis the claim falls within the reinsurance cover as a matter of law.

39...When one is examining the claim so recognised by the insurers when they settle it by admission or compromise, one is examining the real basis on which the claim has been settled...

40 In examining the real basis on which a claim has been settled, one is looking to identify the factual and legal ingredients of the claim embodied and thus recognised in the settlement...”

He then considered how one identifies the “real basis” of a settlement and the factual and legal ingredients thereof and the circumstances in which the reinsurer may be bound thereby. This led him to consider Evans J’s judgment in Hiscox.

In relation to Hiscox he stated at [49]-[50]:

“49 It is apparent that Evans J was of the view that, in a case where the risks reinsured are co-extensive with those originally insured, the effect of the reinsurers’ agreement to follow the settlements of the insurers may be to bind the reinsurers by a compromise of a dispute between the insurers and their assureds as to liability, including as to whether the claim is covered by the risks insured under the contract of insurance as a matter of law, provided that the insurers have acted honestly and have taken all proper and businesslike steps in making the settlement: with the consequence that the reinsurers cannot reopen the precise same question for the purposes of disputing liability under the terms of the contract of reinsurance or contesting that the claim does not fall within the risks covered by the contract of reinsurance as a matter of law.

50 ….He concluded …. that the reinsurers ‘may well be bound’ to follow the insurers' settlements and, implicitly therefore, may well be precluded from raising under the contract of reinsurance the same issues in relation to liability and scope of cover as had already been disputed and compromised, or admitted, by settlement under the contract of insurance..”

Mr Kealey QC expressed his agreement with Evans J’s provisional view and identified three reasons for doing so at [51]:

“….First, if the reinsurers were to have the ability to reopen the same issues of coverage as a matter of law under the contract of reinsurance as those which the insurers had admitted or compromised under the contract of insurance in circumstances where the parties had deliberately ensured that the reinsurance should be back to back with the insurance, the result would in practical and legal terms be similar to that which would obtain if the parties had agreed that the promise by the reinsurers to follow the settlements of the insurers should be subject to the settlements coming within the terms and conditions of both the contract of insurance and the contract of reinsurance. That, however, was not agreed either expressly or impliedly. Secondly, reinsurers having entrusted the insurers with their confidence in determining and compromising any disputes in relation to whether any claims made under the contract of insurance fell within the risks covered by the contract of insurance as a matter of law in circumstances where the risks covered by the contract of reinsurance were identical, it would be inconsistent for the reinsurers to be able thereafter to dispute the self same questions of law arising in connection with the identical terms and the equivalent scope of cover under the contract of reinsurance. It is implicit in the combination of the fact that the contracts were deliberately back to back with the promise by the reinsurers to follow the settlements of the insurers that, if the insurers upon the taking of all proper and businesslike steps honestly compromise a dispute concerning the scope and application of the contract of insurance, the reinsurers should be bound to follow the insurers in respect of that settlement. Thirdly, and following from the above, the follow the settlements promise binds the reinsurers to a proper and businesslike compromise by the insurers of the question of liability or a proper and businesslike admission of liability under the contract of insurance whether the insurers were legally liable to the assured or not. For the insurers then to have to prove to the reinsurers that, on the settled facts, they were legally liable to the assured, which would be the inevitable consequence (if indirect) if the reinsurers were entitled to require the insurers to establish that on the settled facts the claim fell within the risks reinsured, would replace the unqualified follow the settlements promise with something significantly different.”

His conclusion was:

“61….The defendants are bound by all settlements made by Generali (except ex gratia and without prejudice settlements) (a) provided that the claims so recognised by Generali fall within the risks covered by the contract of reinsurance as a matter of law: by which is meant in this context as explained above, provided that the claims were settled on a basis which, if and assuming it to be valid, falls within the risks covered by Generali's outward contract of reinsurance as a matter of law; and (b) provided that Generali acted honestly and took all proper and businesslike steps in making such settlements.”

Mr Kealey QC’s decision was upheld by the Court of Appeal in somewhat shorter terms. The judgment of the Court was given by Tuckey LJ who, having considered the proviso suggested by Evans J in Hiscox, stated as follows at [17]-[18]:

“17 The proviso to which Evans J refers is that reinsurers are bound by reasonable compromises on liability and quantum between the insurers and their assured under the terms of the original policy. That, as I have already said, is well established: the insurer does not have to prove that if the original claim was fully argued it would in fact have succeeded. No investigation as to whether it was arguably within the terms of the original policy is required. But what Evans J says about the reinsurance is clear. Like the judge, I agree with what he says.

18 What I have said so far disposes of Mr Hofmeyr's submissions on this part of the case which I do not accept. But none of the earlier cases have considered how the first proviso works in practice in a case such as this. Here the judge has done so. By reference to the words ‘the claim so recognised’ he has concluded that the insurers do not have to show that the claim they have settled in fact fell within the risks covered by the reinsurance, but that the claim which they recognised did or arguably did. I think this gives effect to what Robert Goff LJ said and gives some sensible added meaning to the clause. It gives substance to the fact that the reinsurer cannot require the insurer to prove that the assured's claim was in fact covered by the original policy, but requires him to show that the basis on which he settled it was one which fell within the terms of the reinsurance as a matter of law or arguably did so. This and the need for the insurer to have acted honestly and taken all reasonable and proper steps in settling the claim provide adequate protection for the reinsurer.” (emphasis added)

The point made by TMEI is that Tuckey LJ is here stating that the test is whether the basis on which the claim is settled arguably falls within the terms of the reinsurance. It is on this basis that TMEI argue that the applicable standard of proof is arguability rather than balance of probabilities.

Novae submitted, however, that Evans J in Hiscox, with whose proviso the Court of Appeal seemingly agreed, was there using the word “arguably” to refer to the settlement of the claim under “the original insurance”, which he had earlier assumed could “cover disputed issues of law as well as fact” (at 529, col. 1). He was saying that where a claim arguably falls within the inward (re)insurance contract, and is compromised on that basis, that settlement might bind the reinsurer where the terms of the insurance and reinsurance are identical.

Novae further submitted that neither Evans J, nor Mr Kealey QC in following Evans J’s approach, was saying that the standard of proof to be applied in showing that the settlement (or to use the words in Scor, “claim so recognised”) falls within the outward contract of reinsurance is merely “arguability”, as opposed to balance of probabilities. Indeed, such a lower standard would appear to be contrary to the right of the reinsurer, expressly recognised by Evans J, to say that the settlement falls outwith the reinsurance.

Novae accordingly submitted that Tuckey LJ’s judgment is to be understood as a short hand reference to the test as applied by Evans J and Mr Kealey QC, with whose judgments he was agreeing. The threshold is arguability, but arguability under the inward rather than the outward contract.

Novae submitted that this is supported by the decision of Burton J in IRB Brasil v CX Re [2010] EWHC 974. In that case, one of the key issues that Burton J had to determine was whether the arbitrators whose award was being appealed had wrongly applied the test of “arguability” rather than “balance of probability” in deciding whether the claim fell within the scope of the underlying and reinsurance cover in the context of a reinsurance policy containing a qualified “double proviso” follow the settlements clause (see [8]).

Having referred to the use of “arguably” by Tuckey LJ in Generali, Burton J went on, at [31], to explain that the arbitrators had “[i]nfelicitously... use[d] the word arguably in relation, not only to the underlying cover, but also to the reinsurance cover...” The same point was made at [37] where Burton J stated that “the question of arguability is inapt to the terms of the reinsurance”.

Novae submitted that its case is further supported by Gross J’s decision in Equitas v R&Q. That case also concerned qualified “double proviso” clauses of the kind found in Hill v Mercantile rather than the unqualified follow settlements clause.

In such a case Gross J held that the burden of proof is on the reinsured to prove that the settlement falls within the Retrocession on a balance of probabilities: see [65], [68]. Novae submitted that the same approach should be followed in respect of the first Scor proviso.

In this context, Gross J emphasised, at [66], that, although the clause he was considering, “unlike the variant of the follow the settlements clause found in Scor... and in the Assicurazioni litigation”, required “proof that the settlements are within the cover of the policy reinsured as a matter of law”, it remained “a follow the settlements clause”.

There is considerable force in Novae’s submissions. I agree in particular with their analysis in this respect of the judgments of Evans J in Hiscox and Mr Kealey QC in Generali. I also agree that in principle it is difficult to see why a lesser standard of proof than the usual civil standard should govern the application of the first Scor proviso to the reinsurance. However, I am bound by the decision of the Court of Appeal in Generali and I consider that paragraph 18 of the judgment is the ratio of that case. That paragraph states in terms (twice) that the test of arguability relates to the reinsurance. Moreover, that this was deliberate is borne out by paragraph 17. In that paragraph Tuckey LJ states that “no investigation as to whether it was arguably within the terms of the original policy is required” – yet that is what Novae submits is the effect of the judgment. He then states that “what Evans J states about the reinsurance is clear” thereby emphasising that the issue of whether the claim is arguably within the policy relates to the reinsurance rather than the original insurance. Whilst I do not agree with that analysis of what Evans J was saying, that is the Court of Appeal’s analysis of it and I am bound by the decision to which that leads, as reflected in paragraph 18 of the Court of Appeal judgment.

For all these reasons the answer to preliminary issue 4 is that the standard of proof to which TMEI has to show that the claim so recognised by ACE falls within the Retrocession as a matter of law is arguability.

The determination of this issue requires a consideration of when the Scor first proviso is to be applied in the manner set out in Generali, as set out above.

It was TMEI’s case that the settlement by ACE involved the compromise of the underlying construction argument which Novae wishes to raise concerning the relationship between the definition of “Occurrence” in the original policy and the 72 Hours clause in the same policy. Whether the settlement did involve this compromise is not an issue which I am being invited to determine, but for the purpose of Preliminary Issue 5 I am to assume that that it did so.

As set out above, the definition of “Occurrence” provides that “Occurrence” is to mean “any one Occurrence or any series of Occurrences consequent upon or attributable to one source or original cause.” TMEI contended that the argument that Tesco and its lawyers pressed was that the Tesco flood loss was attributable to one source or original cause and so constituted a single Occurrence giving rise to a single deductible. ACE agreed to settle the claim on this basis.

The contrary position, which TMEI contended that ACE had argued, was that the definition of Occurrence was displaced in situations in which the 72 Hours clause applied. That clause provided in relevant part as follows: “All loss, destruction or damage… caused by inundation from the sea or the rising, overflowing or breaking of boundaries of any lake, pond, reservoir, river, stream or other body of water, all whether or not driven by wind, and occurring during a period of seventy two consecutive hours… shall be deemed to have been caused by a single Occurrence.”

Novae’s argument is that where claims fall within the 72 Hours clause, there is a separate Occurrence for each group of losses occurring within periods of 72 hours, and that it is impermissible to have regard to the definition of Occurrence then to treat these separate Occurrences as a single Occurrence on the grounds that they arise out of a single source or original cause. Novae also says that on the proper construction of the definition of “Occurrence” and in particular the phrase “single source of original cause”, the Tesco claims are consequent upon or attributable to more than one source or original cause.”

The position which TMEI contended was advanced by Tesco and accepted in the settlement by ACE as the basis upon which the claim was settled is, so TMEI submitted, at the very least arguable. It is at least arguable that the 72 Hours clause does not displace the definition of Occurrence, and that the correct construction of the aggregation provisions is that (a) where losses fall within the 72 Hours clause then that clause operates so as to bring all losses within a consecutive 72-hour period into a single Occurrence but (b) where losses occurring over a period greater than 72 hours all arise out of a single source or original cause then the effect of the definition of Occurrence is to aggregate them all into a single Occurrence. If, as held in answering Preliminary Issue 4, TMEI is required to prove that the claim as recognised by ACE arguably falls within the coverage provided by the Retrocession as a matter of law then TMEI submitted that it is clear that it can do so and Novae is not entitled to re-open the issue as to construction of the aggregation provisions.

Novae disputed that this was so on the basis that a number of facts distinguish the present case from Generali and that the clause in this case is not to be applied in the same manner. In particular, it was submitted that these facts meant that this is not the type of case, unlike Generali, in which a reinsurer would consciously entrust its liabilities under the Retrocession to TMEI, still less ACE. Further, Novae reserved the right to argue before a higher court that Generali was wrongly decided.

The main distinguishing facts relied upon were:

As to i), it was stressed that whereas in proportional reinsurance, as a result of the reinsurer sharing in the risk of the original insurance (in exchange for a proportion of the premium), the fortunes and interests of the reinsurer are likely to be similar, if not identical, to that of the reinsured, that is not the case in non-proportional excess of loss reinsurance. In this connection reliance was placed on the judgment of Lord Mustill in Axa v Field and the passage cited at paragraph 35 above. It was further pointed out that as an excess of loss retrocessionaire, Novae’s interests were unlikely to have been the same as TMEI or ACE, particularly in relation to aggregation issues.

I accept that the fact that this is non-proportional excess of loss Retrocession means that Novae’s interests will not always be the same as those of TMEI or ACE. In many cases, however, their interests are likely to be similar. In relation to the issue of aggregation in this case, for example, it was in the interests of the insurer/reinsurer under each contract under which it was insuring/reinsuring for the claim to be made up of a number of Occurrences, albeit that Novae’s interests are more affected by that issue.

As to ii), in the light of my findings on Preliminary Issues 1 and 2, no material difference in the terms and conditions governing coverage under the Reinsurance and the Retrocession have been identified, so in that sense they were back to back. It is true that while ACE, as a direct insurer of Tesco, and TMEI, as a proportional reinsurer of ACE, provided coverage (albeit for different percentage shares) for all covered losses exceeding the deductibles in the Master Policy and the local policy, Novae only covered losses which exceeded the £53 million per Occurrence deductible, but that is essentially a different way of expressing the point made under i).

As to iii), Novae submitted that it stands one removed from the direct insurance provided by ACE Europe under the Master Policy. It submitted that its position was analogous to that of the reinsurer in Hill v Mercantilewho Lord Mustill described (at 1252D-E) as “a remote reinsurer, who may know nothing beyond the identity of his reinsured, and the terms of his own cover” who would be likely to “hesitate to entrust his liabilities to a stranger, which is what will happen if all the reinsurances down the chain embody unqualified follow settlements clauses.”

Hill v Mercantile, however,was concerned with a reinsurer under the LMX spiral who was truly remote from the direct insurance and at the end of a long chain of reinsurances. In this case Novae was only one step removed from ACE, and indeed had itself reinsured ACE the previous year.

I do not therefore accept that the factual distinctions drawn are as significant as Novae contended. In any event, I would not regard them as sufficient in principle to displace the meaning given to the materially same unqualified follow settlements clause in Generali.

The ambit of the decision in Generali is illustrated by the question which was formulated by the Court of Appeal as arising on the appeal, namely: “To what extent does the first (Scor) proviso preclude reinsurers from raising coverage issues relating to terms which are the same on both the insurance and reinsurance?”

As Wasa v Lexington illustrated, the terms must be both the same and to the same effect (in that case different governing laws meant that they were to different effect). Lord Mance at [36] described Generali as being a case “where the insurance and reinsurance incorporated materially identical terms with materially identical effect (and the issue was whether and on what basis the facts fell within such terms)”.

Where the relevant terms are materially identical one can well understand why reinsurers should be bound by the legal basis of the settlement, provided it arguably falls within the reinsurance. If it were otherwise the reinsurer, having agreed an unqualified follow settlements clause, would be entitled to attempt to avoid paying its share of a settled claim by re-litigating an issue of construction arising on a materially identical clause, which has been the subject of dispute between the original insured and the original insurer, resulting in a compromise that accepted the insured’s argument as correct and which cannot be criticised as unreasonable. If, as is accepted, the unqualified follow settlements clause prevents this in relation to issues of fact, it is difficult to see why that should not equally apply where the issue that has been settled is one of construction of a materially identical clause.

In this connection the three reasons identified by Mr Kealey QC in Generali at [51] as to why the reinsurer should generally be bound by an unqualified follow settlement clause where the identical legal issue arises under the insurance/reinsurance are apposite. Although Novae submitted that this reasoning mainly applied where the contracts were deliberately back to back, in this case the terms governing coverage and the relevant terms in issue were, on my findings, deliberately back to back. The defined meaning of Occurrence/Loss Occurrence and the 72 hour clause are materially identical in the Reinsurance and the Retrocession.

Further, the effect of Novae’s argument is that the unqualified Scor type follow settlements clause in this case is to be construed in the same manner as if it contained the second proviso of the qualified Hill v Mercantile type follow settlements clause. However, it did not so provide. The general effect of an unqualified follow settlements has been clear from the time of the Generali decision (at the latest). If it is wished to make it possible for reinsurers to dispute that the settlement was not within the legal extent of the cover provided, there is a well established means of so doing – namely, the qualified Hill v Mercantile type follow settlements clause. That course was not taken. Instead an unqualified follow settlements clause was used.

As Lord Hobhouse noted in Lloyds TSB General Insurance Holdings & Ors v Lloyds Bank Group Insurance Co Ltd. [2003] Lloyd’s Rep IR 623 at [31]:

“Another preliminary observation which needs to be made, which is true of very many professionally drafted commercial and financial contracts, and is particularly true in the present case, is that there are often well established alternatives open to the parties in the drafting of their agreement. The choice made from among these alternatives represents part of the bargain struck by the parties and must be respected by anyone (judge or arbitrator) adjudicating upon a dispute arising under the document. As noted by Lord Mustill in AXA Re v Field [1996] CLC 1169 at pp. 1173–7; [1996] 1 WLR 1026 at pp. 1031–5, and, as I will explain below, aggregation clauses come in different well established forms. The clause in the present case is no exception.”

The same observation can be made in respect of follow settlements clauses and the choice made in this case.

For all these reasons I conclude that the answer to Preliminary Issue 5 is that Novae is bound by a determination by ACE (if any) as to the construction and application of the aggregation provisions in the Master Policy to the Tesco losses.

For the reasons outlined above I answer the five preliminary issues as follows: