WRBC Corporate Member Limited v AXA XL Syndicate Limited & Ors

This is a claim made by the reinsured, WRBC Corporate Member Limited (as the sole underwriting member of W.R. Berkley Syndicate 1967 at Lloyd’s for the Lloyd’s 2019 year of account) (“WRBC”) against the First to Fifth, Eighth, Tenth and Twelfth Defendant reinsurers (“the Reinsurers”) for indemnities and payment pursuant to two multi-line excess of loss reinsurance treaties (collectively “the Treaties”). The same claims made against the other Defendants have been settled on confidential terms.

The first treaty, dated 24 April 2019 and placed for WRBC by brokers, Aon UK Limited (“Aon”), is with (among others) five of the Reinsurers (“the Aon Treaty”). Those Reinsurers are the First, Fourth, Fifth, Eighth and Tenth Defendants (“AXA”, “XL Bermuda”, “Liberty”, “Chord” and “HCC” respectively, and collectively “the Aon Reinsurers”). The second treaty, dated 31 May 2019 and placed for WRBC by brokers, JLT Brokers Ltd (“JLT”), is with (among others) three of the Reinsurers (“the Treaty”). Those Reinsurers are the Second, Third and Twelfth Defendants (“Brit”, “Blenheim” and “TransRe” respectively, and collectively, “the JLT Reinsurers”).

Both treaties have materially the same terms and at the hearing, for convenience, reference was made to the terms of the Treaty. The period for both of them was 1 April 2019 to 31 March 2020.

Six classes of business were reinsured: (A) Fine Arts, Jewellers Block and General Specie; (B) Contingency; (C) Political Risks; (D) Accident; (E) Construction and Engineering; and (F) Political Violence.

This case concerns claims made for losses incurred by WRBC in its Contingency business. All such losses are said to arise from WRBC’s settlement of claims made against it as insurer, by its various insureds. The claims were in respect of conferences and similar events which were cancelled by the insureds as a result of measures imposed by the authorities in various locations which restricted or prevented large public gatherings, due to the Covid pandemic.

There are seven separate jurisdictions involved, each of which had governmental powers to impose those measures. They are the States of California, Colorado, Florida, Illinois, Nevada and New York in the USA, and England.

There were 174 individual losses in total. For the purposes of the claim made against the Reinsurers, WRBC has aggregated all of the losses in each jurisdiction together, so there are seven groups of losses. This is on the basis that all of the losses in each jurisdiction can be said to have been losses “arising out of one event” as set out in the relevant limits provision in the Treaties (“the Limits Clause”) which reads, in part:

WRBC says that the “event” concerned here is the relevant measure, or group of measures, taken in each jurisdiction that caused the underlying conferences etc. to be cancelled. On the basis of such aggregation, each group of losses comfortably exceeded the relevant deductible contained in the Treaties, being $1m.

The Reinsurers have resisted the claim on the basis that the word “event” in the phrase “arising out of one event” is not a causative event such as a Covid-related measure which restricts or prohibits large gatherings, including conferences. Instead, it imports the definition of “Any One Event” contained in the Class Definitions section of the Treaties. This reads as follows:

On that basis, the relevant loss or series of losses are such losses which arise from the relevant individual conference (which was cancelled). If so, there can be no aggregation and the $1m deductible must be applied to each of the 174 losses claimed. Since most of those losses were for sums under $1m, it means that the Reinsurers have no liability for them. For that reason, the amount to which the Reinsurers say WRBC is entitled is very much less than that claimed.

The financial consequences of this dispute are as follows, in broad terms:

I refer to this first dispute between the parties as “the Construction Issue”. If the Reinsurers succeed on this issue there are no further issues which arise, save in respect of the details of the quantum. There is a dispute about the rate of interest payable but it has been agreed that this can be dealt with, if it arises, after judgment.

If, on the other hand, WRBC succeeds on the Construction Issue, there is a further dispute. This has been referred to as “the Aggregation Issue” and it arises in this way: WRBC says that, within each of the seven jurisdictions, there is the same measure (or group of measures) which caused each of the individual cancellations (and therefore losses settled by WRBC). The Reinsurers do not accept this. In some cases, they contend that the reason for the cancellation was not a measure at all, or it was a different measure from that which caused other cancellations within the same jurisdiction. If the Reinsurers are correct in some or all respects here, this will diminish the sums they have to pay to WRBC. This means that for each of the 174 individual losses, I have to decide what the causative event was. However, I am not asked to calculate the quantum implications of such findings. This can be done by the parties, following my judgment on such matters (if it gets to that stage). I was not addressed orally on each of the 174 losses in the course of the trial, because it would have been too time-consuming to do so. Instead, I heard detailed oral submissions from the parties on each of the losses claimed for California (and England to some extent) which proved to be a useful example of the sort of points being made by the Reinsurers. In addition, I have written submissions from both parties which do address each of the 174 losses.

WRBC has adduced three witness statements (“WSs”) dated 4, 10 and 11 April respectively from the following:

None of these witnesses were required for cross-examination by the Reinsurers. Instead, the Reinsurers made a number of submissions about the relevance and/or admissibility of their WS’s. Nor did the Reinsurers adduce any witness evidence of their own. Insofar as I consider that the WS’s filed on behalf of WRBC are relevant and admissible, this means that, assuming there is no reason to doubt the credibility of the relevant witnesses, I can and should accept their evidence as being unchallenged.

In addition to that evidence, there is, of course, a significant body of contemporaneous documents to which both sides referred.

The following abbreviations are used below:

There was not, in my view, any real dispute between the parties as to the relevant principles, so I can deal with them relatively briefly.

The starting point, in the context of insurance contracts, is what Lords Hamblen and Leggatt stated at paragraph 47 of their judgment in Financial Conduct Authority v Arch Insurance (UK) Ltd [2021] UKSC 1 (“FCA Supreme Court”):

These principles, in the context of reinsurance, were referred to by the Court of Appeal in Unipolsai Assicurazioni SpA v Covéa Insurance [2024] EWCA Civ 1110. At paragraph 13 of the leading judgment, Sir Julian Flaux C referred to the statement of the law made by Foxton J (as he then was), the judge at first instance, without qualification; this referred to the passage quoted above from FCA Supreme Court, but also the extended statements of principle given by the Divisional Court in Financial Conduct Authority v Arch Insurance (UK) Ltd [2020] EWHC 2448 (Comm) (“FCA Divisional Court”) at paragraphs 62-70 of that Court’s judgment. Paragraph 63 itself quotes from the well-known judgment of Lord Hodge of the Supreme Court in Wood v Capita Insurance Services Ltd [2018] Lloyd’s Rep Plus 13:

The Divisional Court went on to add as follows:

The expression “Pick and Mix” describes a situation where the contract has used a selection of different clauses from other contracts without any attempt to ensure that the language is used consistently. In such a case, which is common in insurance contracts, the inference that terminology has been used consistently throughout “has little force”. See International Entertainment Holdings v Allianz Insurance Plc [2024] EWCA Civ 1281 at paragraph 18.

A good example of this in the insurance context is Sky UK Limited v Riverstone Managing Agency Limited [2024] EWCA Civ 1567 which has a certain resonance with the case before me. Here, the Court of Appeal considered the meaning of the phrase “any one event” in the relevant clause in a construction all risks policy (the “Policy”) which provided a deductible (“Retained Liability”) of “£150,000 any one event”. The insurers argued that the word “event” in the Retained Liability clause governed the damage suffered and not the cause of the damage. They pointed to a different clause (“the 72 hour clause”) which deemed certain losses to constitute a single event which such clause was stated to be for “the purpose of the Insured’s Retained Liability”, and said it followed that this definition should be applied also to the deductibles clause in issue.

However, the Court of Appeal rejected this argument. In particular, Popplewell LJ stated the following:

Then, when dealing with the insurers’ reliance on the 72 hours clause, he observed that:

So, even though the word “event” in the expression “any one event” was not defined in the clause in question, it did not follow that where “event” was defined elsewhere, that definition must apply to the relevant clause.

The key point to note is that an argument based on surplusage or redundancy is not infallible or conclusive. It is one factor in the unitary and iterative process of construing a contract in its entirety. Indeed, it may be unreliable.

Thus, in the House of Lords case of Beaufort Developments v Gilbert Ash [1999] 1 AC 266, Lord Hoffmann observed as follows at page274C:

Further, in the Supreme Court case of Triple Point Technology v PTT [2021] UKSC 29, Lord Leggatt stated as follows:

In Unipolsai Assicurazioni SpA v Covéa Insurance [2025] Lloyd’s IR 254, being the decision at first instance of Foxton J, he observed that:

He then referred to cases adopting this approach with regard to the ISDA Master Agreement and the ISDA Guide, explaining the purpose of changes made to it, a clause in a charterparty drafted by the Documentary Committee of The Baltic and International Maritime Council, where a circular issued by the Committee explained the thinking behind the clause, and a detailed review of the market history of the UNL clause in issue, when arriving at his conclusion as to the meaning of the words “actually paid”.

The Treaty here is, of course, not a market form of contract in the sense referred to by Foxton J but, provided that the caveats stated in paragraph 44 of his judgment are observed, I see no reason why the history of a contract like the Treaty cannot form part of the relevant factual matrix. On any view, the Treaty cannot be described as a “bespoke contract”.

At paragraph 65 of the judgment, FCA Divisional Court dealt with the decision of the House of Lords in Chartbrook Ltd v Persimmon Homes Ltd [2009] UKHL 38 as follows:

Paragraphs 24 and 25 of Chartbrook are also worth citing:

Further, as the Court of Appeal stated in Monsolar IQ Ltd v Woden Park Ltd [2021] EWCA Civ 961 at paragraph 40:

In Monsolar itself, the Court, in applying the Chartbrook principle stated at paragraph 41 that the cause of the error was simple enough. There, the clear and unambiguous language of the annual rent review clause in issue stipulated that the rent would increase, not by the increase in RPI over the previous year but rather the total increase in RPI since the lease commenced, which obviously produced a very much higher increase, and would do so at even higher levels for each succeeding year. This entailed the consequence that something had clearly gone wrong with the drafting. The Court of Appeal held that the reason for the mistake was obvious; either the total increase in RPI from inception could be taken and from it, the total increase in RPI from inception for the previous year could be deducted, or the increase in RPI, not from inception but from the previous year, could be used. Either way the amount of the increase would be the same. What the draughtsman had clearly done was to mix up both approaches. So this was a case where, on an objective basis, there was a plausible explanation for the mistake. Accordingly, the rent review clause had to be effectively reworded so as to produce the objectively intended rent increase formula.

It should be noted that in Chartbrook itself, the necessary correction did not so much require rewording as a syntactical alteration. In that case, the apparent effect of a clause which provided a possible additional payment by the developer which had purchased the claimant’s land and had obtained planning permission for the building of flats which would then be sold, was that the developer had to pay 23.4% of the sale price for each flat (less expenses) insofar as that figure was in excess of the guaranteed retail value of the flat. The problem was that the guaranteed retail value was calculated by reference not to the value of the flat as built but only the underlying land value of the flat. While, in round terms, each flat was expected to raise around £200,000, its retail land value (which was itself achieved by mathematical formula) was only £50,000. The latter was therefore clearly going to be exceeded in every case by the sale price. The House of Lords held that something had gone wrong with the drafting in order to produce such a commercially absurd or arbitrary result.

Applying the principles referred to above, it was held that what the relevant clause was meant to have said was that the additional payment was the difference between (a) 23.4% of the price achieved and (b) the minimum guaranteed value. This would only arise if the price achieved exceeded the “grossed up” minimum guaranteed value from its notional 23.4% to 100% which was taken to be the putative sale price of the flat.

Finally, the Court of Appeal in Scottish Widows Fund and Life Assurance Society v BCG International [2012] EWCA Civ 607 identified, at paragraph 23 of its judgment, two examples of where “something has gone wrong with the language”, namely where “the effect of the literal construction of the mistaken language [i]s to produce interpretations that made the instruments ineffective in law, or irrational or arbitrary” or where “the provision in question is inconsistent with some other provision which is clear must have precedence”.

The following propositions are clear from those cases:

The following is a summary of the relevance of the layers of protection to be found in reinsurance contracts such as the Treaty, and is taken from the judgment of Phillips J in Deeny v Gooda Walker Ltd [1996] IRLR 183 at p189:

I did not understand the Reinsurers to disagree with this summary, nor with the proposition that references to “catastrophe” are usually in a context where a single “catastrophic event” will give rise to numerous losses which collectively could be very substantial.

WRBC submits first, that on a straightforward textual analysis, it is clear that the expression “arising from one event” in the Limits Provision refers to the consequences of a causative event as opposed to the putative event which is cancelled as is contemplated in the Class Definition.

However, if that conclusion is not reached by a textual analysis alone, then it is justified when one takes into account the wider context or factual matrix. Here, WRBC relies on three matters:

As a final point, if necessary, WRBC contends that this is a case where a corrective construction on the basis of the Chartbrook principle should be adopted.

For their part, the Reinsurers contend first that on a textual analysis, their interpretation is the correct one.

Second, on the wider commercial context:

Finally, the Reinsurers deny that this is a case where a corrective construction can be applied.

In the light of the above, I shall deal first with the textual arguments and then the contextual arguments.

In setting out the material parts of the Treaty I shall, where appropriate, adopt the summaries of particular provisions which appear in the parties’ Skeleton Arguments where they are uncontroversial.

The Treaty was a Multi-Line Excess of Loss treaty. In other words, it provided reinsurance cover for different classes of insurance business effected by WRBC, here, six of them. Those classes of business were:

Class B is defined as follows:

The losses covered were those occurring in the period 1 April 2019 to 31 March 2020.

The three layers of cover are set out in the section headed “Limits”. In this regard, that section was amended by the third of three amending endorsements dated 1 July 2019 but effective from 1 April 2019 (“Endorsement 3”). As amended, Layer 1 varied as between different classes of business. For Classes B, D and F, it was US$7.5m in excess of US$1m

The latter wording (bold applied for emphasis) applied to every “in excess of” figure across all Layers and Classes. It is the critical clause for present purposes, because the issue is what “event” here means. I refer to it hereafter as “the Limits Clause”.

For Class A, Layer 1 was US$7.5m in excess of US$2.5m. Thereafter, Layers 2 and 3 had the same deductibles and limits for every class. Layer 2 was US$20m in excess of US$8.5m, and Layer 3 was US$45m in excess of US$8.5m. The overall limit was therefore US$72.5m. It will be noted at once that both Layers 2 and 3 were in excess of the same amount i.e. US$8.5m. What in practice this meant was that once the reinstatements for Level 2 were exhausted, Level 3 would engage for losses above (again) US$8.5m.

In addition, the following provision was added by Endorsement 3 (“the Cross-Class Deductible Provision”)

The Limits Section ends with the following:

The Reinstatement Provision allowed for automatic reinstatement of the various limits once they had become exhausted subject to payment of additional premiums. Four reinstatements were allowed for Layer 1, two for Layer 2 and one for Layer 3.

A later section headed “Conditions” contained a further section headed “DEFINITIONS”. Under “Class Specific” different provisions applied to different classes. Those affecting Class B were as follows (cited in full):

The section of text in bold is highlighted for emphasis. It is the provision relied upon by the Reinsurers as to the meaning to be given to “event” in the Limits Clause.

The DEFINITIONS section deals with the other classes of business (in summary) as follows:

Later on in the DEFINITIONS section, ultimate net loss is defined as follows:

Other parts of the CONDITIONS section contain a Notice of Loss and Loss Settlements Clause, a Reinsuring Clause, an Extra Contractual Obligations and Excess Limits Liability Clause, an Errors and Omissions Clause, a Special Termination Clause, an Insolvency Clause and various other detailed provisions governing the operation of the Treaties.

I shall refer to some other provisions of the Treaty in context, below.

It is here worth pointing out the brief background to the Treaty, which is not in dispute. First, for the year 1 April 2018 to 31 March 2019, there were similar multi-line XL treaties entered into between WRBC and groups of reinsurers, the majority of whom were also the Reinsurers.

In relation to previous years, the first scheme for reinsurance on a multi-line basis, in other words catering for a group of different classes of insurance business written by WRBC, was put into effect in 2015. This consisted of two separate treaties. The first, dealing with various classes of business, provided cover on a “per risk” basis of US$4.25m in excess of US$750,000 with a maximum aggregate limit for the total of all losses claimed of US$20m. I refer to this as “the Aggregate Treaty”. The second provided layers of cover starting from US$5m upwards. I refer to this as “the XL Treaty”. This structure was followed until 2018 when, as already noted, the structure changed to a single multi-line reinsurance contract of a form and content which was essentially repeated in 2019 with the Treaty.

In respect of each of those reinsurance years, the brokers for WRBC made submissions to the relevant groups of reinsurers for the purpose of the next year’s placement of reinsurance, typically a few months before the new reinsurance would take effect.

Prior to the treaties made in 2015, WRBC placed its reinsurance on an individual line of business basis. In other words, political risk, contingency, fine art etc were all separately reinsured.

A number of points can be made here and I have identified each one separately.

I agree that a useful starting point is to consider the meaning of “arising out of one event”, standing on its own, as it were. This is a very familiar expression in contracts of reinsurance (and insurance). Indeed, even without the words “arising out of” Popplewell LJ considered that the phrase “any one event” was “a classic term for aggregation of losses by reference to the cause of the losses.” See paragraph 118 of his judgment in Sky, quoted at paragraph 24 above. The phrase here clearly imports that the event is something which has causative effect. It is of course the same phrase which is used for all of the classes of business reinsured, so far as limits and deductibles are concerned. And as Sky once more demonstrates, the fact that there is elsewhere an expression using the word “event” does not negate the “potency” of the expression in its usual causal form; see paragraph 121 of the judgment of Popplewell LJ, quoted at paragraph 25 above. See also the causal interpretation given to a loss clause referring to “arising out of the same event” by Evans LJ in Caudle v Sharp [1995] C.L.C. 642 (1995) at pp7-8.

Next, the phrase “Any One Event” (“AOE”) is not actually present as such in the Limits Clause, nor are the words “one event” capitalised. I accept that the absence of capitals is not determinative here. After all, there is the expression “Any One Risk” (“AOR”) also used in the Limits Clause which is then referred back to the definition of Any One Risk (see paragraph 58 above), and it is hard to see there that the absence of capitals means that this definition does not apply.

However, the absence of the precise phrase “Any One Event” in the Limits Clause is important, in my view. The Reinsurers say that the word “any”, part of the definition but not present in the Limits Clause, is insignificant since the key word is “event”, or “one event”. I see that, but on the other hand, one would expect a phrase that is contained in a definitions section to be used precisely elsewhere if that definition is to be attracted. After all, that was the case (save for the capitals) for AOR. It is also the case for another defined term in the definitions section in respect of various Classes, namely “each and every loss”.

In this context it is also noteworthy that, while the Limits Clause makes a specific express reference back to the definitions section for “risk”, it does not do so for “event”. See paragraph 58 above. As to this, the Reinsurers say that this does not matter because the words of the Definitions Section apply the definitions to any part of the Treaty and thus to the Limits Clause. Certainly they do in some cases, for example where it is said that “each and every loss” shall have a particular meaning “For the purpose of this agreement”. In the context of AOE what is said is that “It is agreed that “Any One Event” is to mean…” which is not quite the same. In any event, the absence of the reference in the Limits Clause to the Definitions Section can at least add weight to the other factors suggesting that objectively it cannot have been intended to apply AOE to the Limits Clause.

Further, if the definition of AOE were to be deployed in the Limits Clause, it needs to be borne in mind that there are different elements of contingency which are covered by Class B. In respect of some of these, the application of AOE would make no sense at all. Such elements would include Confidential Life (Including the ASU Critical Asset Protection Cover), Trade Name Restoration, Failure to Survive and Mathematical Prize Indemnity. That would mean that some parts of Class B would be subject to aggregation because “event” would have its normal meaning (since it could not sensibly be governed by AOE), while other elements would not. That would be absurd in my view.

The point is not answered by saying that this does not matter because most of the business here related to Cancellation and Non-Appearance (79% as at 31 December 2019). The fact remains that for the other elements, the language of AOE would still not work. Moreover, some of the Non-Appearance cover written may not relate to a defined event like a conference anyway.

The Reinsurers here say that such other elements which could not be governed by AOE would still be governed by AOR. However, that then raises whether AOE was necessary at all (for any element), given the work to be done, on the Reinsurers’ case, by AOR. I deal with this point at paragraphs 89 - 90 below.

In addition, of course, even on the Reinsurers’ case since AOE can only apply to Class B, it follows that the word “event” is to be used in its normal causal sense in respect of claims made under every other Class which would therefore have the benefit of aggregation so far as deductibles are concerned, whereas Class B would not. In the latter case, “event” has a completely different function since on its face, it is defining that which is covered by the insurance as opposed to the limits of the reinsurance. Given the well-known use of aggregation clauses, that would be a very surprising result.

As to that, the Reinsurers argue that as this is a multi-line policy dealing with a variety of different types of risk insured by WRBC, with their own individual risk profiles, it would be rational and natural for the parties to provide differently for each Class. In theory, of course, that is true, but in this context, it would in my view be most unusual for parties to intend that a substantial part of the total business being reinsured (i.e. Contingency) should not benefit from aggregation on deductibles at all.

This last point is emphasised by the fact that, absent aggregation, there would be little commercial need for Layers 2 and 3 in relation to Class B, especially when the available reinstatements are taken into account. That is because many losses would not reach them. This is in a context where we know that on the Reinsurers’ case, and in the absence of aggregation for deductibles, out of 174 losses claimed by WRBC, only 40 would get past the Layer 1 deductible at all, let alone engage Layers 2 and 3. Indeed, of the 174 losses claimed, only two exceeded US$8.5m.

In this context, WRBC relies upon its stated maximum line exposure here of US$11.25m (“the Line Guide”). This was not a term of the Treaty but it was referred to by WRBC in the 2019 Submission. It was regarded as important, because when WRBC wrote business with an exposure of more than this, as in the case of the Tokyo Olympics, it sought a “special acceptance” from the Reinsurers to cover an increased amount of US$15m. Generally however, most of the lines written were very much less than this.

Since the Line Guide is referred to in the 2019 Submission, I shall deal with it in that context, below, noting that of course, if WRBC is entitled to rely upon it, it goes to the commercial utility of Layers 2 and 3 of the Treaty which can be regarded as a textual rather than a factual matrix question.

However, it is here important to note that part of the premium payable for the Class B business reinsured was for access to Layers 2 and 3. This is because the premiums payable under the Treaty were calculated by reference to a percentage of the underlying premium income estimated to be earned by WRBC for 2019 across each Class of business written. This can be seen from an email from Ms Robertson at Aon to Mr Jenkins at AXA, the lead reinsurer dealing with Aon, dated 11 June 2019 which attaches a schematic summarising AXA’s then offer. As the schematic shows, and as Ms Robertson explains at paragraphs 11-16 of SR1, the reinsurance premiums for Layers 2 and 3 were calculated by reference to the putative insurance premium income to WRBC for all Classes, which was a total of US$76.4m of which US$5.1m came from Contingency. It would make little sense for reinsurance premiums based on those figures to be payable if there was no real likelihood of much of Layer 2 and all of Layer 3 ever being engaged in respect of Contingency losses. Here, I should add that where there were specific qualifications to the availability of a layer, the schematic stated them; see the position in relation to Layer 1 and Political Risk and Construction/Engineering coverage, and the adjusted retention where a causative event involved losses in both Crisis Management and Asset Protection lines of business. Those qualifications were further explained at pages 4 and 22 of the 2019 Submission.

The Reinsurers respond that there is nothing surprising about the Layers operating differentially depending upon the particular Class of business reinsured and its own risk profile, where the reinsurance is to cover all Classes. However, that is not a sufficient answer, in my view, for the following reason: in the event of a pandemic (which must have been one of the sorts of “catastrophe” that the parties would reasonably contemplate), the Class most likely to be affected would be Class B. So it would be especially odd if this Class alone would have no real access to Layers 2 and 3.

In the light of that, there is force in WRBC’s point that if, in truth, there was to be no Layer 2 or 3 access, there could have been an express exception made for Class B to that effect. After all, Layer 1 was expressly excluded from Classes C and E.

A further point concerns the Cross-Class Deductible Provision referred to at paragraph 57 above. It is clear that the reason for the introduction of this provision in Endorsement 3 is because there were now two different deductibles under Layer 1: US$1m for Classes B, D and F, and US$2.5m for A. This provision apportions the relevant deductibles where there are claims for losses arising under Class A and another Class (a so-called “clash” of claims). The introductory wording “in respect of a loss event involving” clearly connotes some event which causes losses in more than one Class. Here, the phrase “loss event” which is clearly causal, cannot possibly import the AOE definition.

However, if AOE does not apply here, it suggests that there can be sets of losses across more than one Class which, if they involve Class B, can be aggregated. If so, it is very hard to see why there can be inter-Class aggregation involving Class B, as it were, but not aggregation within that Class. The Reinsurers’ answer to this puzzle was to say that if the Cross-Class Deductible Provision applies, there can indeed be aggregation across the Classes, but only one loss from Class B can be “brought over”, as it were, to be aggregated with, for example, Class A losses, where all the losses are caused by the same (loss) event. But that makes no sense at all and would be completely arbitrary. It can be asked, rhetorically, if 10 conferences were cancelled because of a single causative event (like a storm), why should only one of them be permitted to be aggregated with a group of losses from another Class, rather than all 10? In any event, even with the single loss which (on this hypothesis) is to be brought over, for that single loss a joint causative event has to be used (so as to bring it together with the similarly-caused losses from another Class) which means that for this particular loss, the operation of AOE is actually disapplied.

Finally, in this regard, the very use of the phrase “loss event” in the Cross-Class Deductible Provision, which is equivalent to “event” in the causal sense, again goes to emphasise the normal use of the word “event”.

A separate point arises as to the utility of the AOE definition in the Limits Clause which itself refers expressly to the “Any One Risk” definition. That definition seeks to include the interests of all (separate) insureds in a particular show etc. as constituting one risk. That would then suggest aggregation beyond a single deductible. However, on that basis, it is hard to see what the application of AOE, also tied to a particular show, would add.

The position is otherwise if “event” is to be interpreted in its causal sense. Of course, this is a “surplusage” argument, which must be treated with caution (see paragraphs 27- 29 above) but here it would be a redundancy in the very same phrase (in the Limits Clause) as the expression AOE was deployed. As to that, the Reinsurers accept that AOR is indeed wider than AOE because it could encompass a series of cancelled concerts. Nonetheless, they contend that if AOE was not stated to operate here, the word “event” would be left at large (and thus connote a causal event) which is not what the parties intended. Something needed to be added to ensure that “event” was not to be read causally. Put another way, the very fact that AOE was to apply here was a manifestation of the parties’ intention that the causal sense of “event” was not intended here, and that remains the case even if, on analysis and having regard to AOR, the actual definition of AOE was redundant. However, that seems to be using a circular logic. In truth, if the parties really wanted to say that “event” was here not to be used in a causal sense, surely they would just have said so.

The Reinsurers also contend that if AOE is not used to define “event”, then the meaning of “event” is “at large” in the sense that there is no particular definition to apply to it. I do not think that this point takes the Reinsurers anywhere. First, given its extensive use in a causal sense especially with the additional words “arising out of” there does not need to be a specific definition anywhere else. Second, it is already “at large” in the Limits Clause so far as it applies to the other five Classes.

As against all of the above, the Reinsurers point out that if AOE does not govern the Limits Clause, it effectively has no work to do and is therefore redundant. That would be especially surprising when it was contained in an important and detailed Conditions section which WRBC does not suggest was inoperative with regard to any other of its elements. See, for example the geographical and time limits applied to Classes D, E and F, contained in the Definitions section, referred to in paragraph 62(2) - 62(4) above. The Reinsurers say that on that basis, WRBC is simply “cherry-picking” the AOE definition for Class B for inapplicability. I do not accept that where, for this particular definition, there are other linguistic and commercial considerations which militate against its applicability to the Limits Clause.

I should add, although it was not its main point, that WRBC also says that at least in one respect AOE was not redundant. This is by reference to Annex A to the Treaty which deals with Exclusions, at p45-46. In respect of the exclusion for Terrorism under Class B, there are various exceptions to the exclusion, as it were, and here there is this passage:

In fact, even if this did not rule out entirely the surplusage point made by the Reinsurers, WRBC says that notwithstanding the placement of the AOE definition in an important Conditions section, one can and should treat the surplusage argument with real caution here. See the authorities referred to at paragraphs 27-29 above. Not only is this an area where contracts may have been cut and pasted over the years without any true regard for consistency (and in this regard see, again, Sky at paragraph 121 quoted at paragraph 25 above), but there are powerful counter-arguments based on the language and commercial consequences of the use of this phrase which is otherwise “surplus”. Those counter-arguments can and in my view do trump any surplusage argument which can be maintained here.

Taking all the above matters into account, in my judgment, and even without recourse to contextual matters, WRBC’s interpretation of the word “event” in the limits clause is clearly to be preferred. In other words, “event” refers to a causal event leading to the losses claimed, as opposed to the AOE definition. That said, I still consider the contextual analysis below.

WRBC relies upon three distinct forms of factual or contextual matrix, all of which it says support its interpretation of the Limits Clause. They are:

As already noted, the Reinsurers’ position on all of this is as follows:

The 2019 Submission is expressed as “2019 Multi-Class Submission” and sets out “Placing Information dated 28 February 2019”. The reason why it cannot be said that this is inadmissible as factual matrix is not simply that it was a document passing between the parties for the purpose of agreeing the terms of the Treaty; it is because the Treaty itself makes express reference to it at page 41 as follows:

WRBC submits as follows in respect of the context of the 2019 Submission:

The simple point made here is that it would have made no sense to make all of these references to “catastrophe” events in the specific context of Contingency business if such events were unlikely to lead to substantial claims potentially utilising all 3 layers of reinsurance, because each loss arising from the same causal event would not be aggregated and would therefore be subject to a deductible which might mean there was no claimable loss at all. The observations of Phillips J in Gooda Walker quoted at paragraph 41 above are relevant here.

As to this, the Reinsurers made an initial point that all of this information was about WRBC’s own book, not, for the most part, how the Treaty was to operate. That is strictly correct, but goes nowhere, because there would be no point in WRBC giving this placing information if it was not relevant to the nature and extent of claims that would be contemplated under the proposed Treaty.

This point is given further force by the fact that in relation to Class E, Engineering and Construction, it is made clear at p4 that “due to a specific quota share this class of business is only covered for individual risks / event losses greater than USD10m FGU”. In other words, when there was a departure from the usual case of layer protection, this was noted. So, it can be reasonably inferred that if it was common ground that Class B did not benefit from aggregation, in contrast to the other Classes, this would have been mentioned.

In other words, the 2019 Submission contemplated that, as with the other classes of business reinsured, Contingency Business would have the benefit of all three layers and aggregation, so far as losses from a single major causal event were concerned.

There were then references in the 2019 Submission to Line Guide figures. In particular, the “Maximum Line Guide” for Contingency ranged from a high of US$11.25m each and every risk for Cancellation and Miscellaneous to a low of US$5m each and every risk for Non Appearance (see p. 27). At p12 it stated that for Contingency, “The largest event is around USD 24m with our line at 39.5% however the average size of events is USD 4m with our line therefore being around USD1.5m. On the schedule, only 18 events would give us an exposure of greater than USD 2.5m”. Finally, in the diagram at p29, it was shown that the majority of Contingency risks fell within the US$0-2m limit bands. Accordingly, losses associated with these risks would not be recoverable unless aggregated, because were below the (then contemplated) deductible of US$2.5m under Layer 1.

Further, if one looks at the actual losses for the prior year 2018-2019, as recorded in the 2019 Submission at page 29, one can see that, of the 145 risks written in respect of Contingencies, only 10 were for more than US$7.5m (and up to US$10m) and only 8 were above US$10m. If one assumes that each of those risks would have a maximum loss of US$11.25m (the line guide), then each of them would produce a maximum loss of US$2.75m above Layer 1, being the difference between US$11.25m and US$8.5m, the latter being the limit of Layer 1. On the basis of the 18 losses in excess of Layer 1, that would produce a total figure of US$49.5m. However, as Layer 1 has 4 available reinstatements, its true effective capacity is US$37.5m, being 5 x US$7.5m (the latter figure being the excess over the US$1m). So Layer 1 would respond to most of the US$49.5m anyway, although of course, reinstatement premiums would be payable. On that basis, only US$12m would engage Layer 2 which, with its two reinstatements, has a maximum capacity of US$34.5m. Layer 3 would not be engaged at all.

As to all of this, the Reinsurers say that it is misleading to focus on a single loss figure of US$11.25m. This is because, if one takes a large event (like the Tokyo Olympics, but disregard the higher sum of US$15m for these purposes), WRBC might very well write a number of different lines, for different insureds, all of whom have an interest to protect if the event was cancelled. If the event was cancelled, then there would be a total claim of US$56.25m (5 x US$11.25m) which would certainly reach Layer 2 on the basis that there would be one deductible only, because there was only one event which was cancelled (ie according to AOE). That is technically correct (leaving aside arguments about AOR etc made above). However, it does not affect the general point that in practice, there are likely to be very many claims made in the light of something like a storm where each relates to a different “Event” in the sense of “conference” etc. Again all of this is demonstrated by each of the 174 different losses clamed here.

In this context, the Reinsurers then argue that there could be substantial losses from a single event like a storm, not by way of aggregation of losses which are all due to that causal event, but because, using AOE, there could be, say, 5 large conferences or other events all cancelled as a result of a storm. If each of them generated losses of US$8.5m each, that would be (after deductibles) US$7.5m each, making a total of US$37.5m. That is a large amount and the Reinsurers would be interested in that anyway because it would trigger all 4 reinstatements under Layer 1. I follow that but typically, the losses from one risk were much smaller (see above) and the point about reinstatement is that it was still unlikely to trigger Layer 2 or much of it. See paragraph 106 above. In fact, the way that the material was presented in the 2019 Submission looks very much as if the context is aggregation anyway, not reinstatement, which is not really discussed.

The Reinsurers also make a related point that concentrating on the Line Guide of US$11.25m is unrealistic anyway, because this is on a “per line” basis. There could be 4 or 5 such lines written in respect of one event like Glastonbury, and using the application of AOE, there would just be one deductible, as they are all losses arising out of that festival. If the total loss was, say, US$28m, based on 4 lines at US$7m, Layer 2 would certainly be engaged. This is in a context where we know that for 2018, there were 8 risks written for more than US$10m. I see that, but again, I do not think that it removes the notion that the Line Guide of US$11.25m was, as against Layers 2 and 3, relatively low and that most risks written were no more than US$2m.

As for Layer 3, the Reinsurers point out that this is a “top and drop” layer because it is engaged once losses exceed US$8.5m, as with Layer 2, albeit that it has a higher limit and is only engaged once Layer 2 is fully exhausted (unless the underlying loss is more than US$20m). So, a loss of, say, US$12m would engage Layer 3, provided that all the reinstatement under Layer 2 had been used up. I see that, and would agree that this would give some commercial utility to Layer 3 even in the absence of a single very large loss (which might be an aggregated one, on WRBC’s case). Nonetheless, there is still force in the suggestion that on the Reinsurers’ case it would not be likely that Layer 3 would be engaged in its “upper limit” sense. Moreover, to some extent, the proof is in the pudding, because we know how matters panned out, in terms of size of individual losses, in the real word of when a very substantial catastrophe did happen, namely Covid.

The Reinsurers also rely upon RDS figures shown at p31 of the 2019 Submission. These are modelled scenarios which are mandatory for Lloyds Syndicates. The Reinsurers point out that the maximum single gross loss across the whole reinsurance book, as it were, is stated as US$33.9m (US Terrorism One World Trade), each of the other losses being less than the upper limit of Layer 2 which is US$28.5m. Further the gross loss (for Contingency only - see PH1 paragraph 66) for communicable disease is only US$13.2m. So this suggests that Layer 3 was not thought to be needed much anyway. Perhaps, but the real point about the RDS modelling results is that, as Mr Harris said at paragraph 50 of PH 1,

In other words, it is all about aggregation across the whole book. There is no reason for me not to accept that evidence.

A further point made by the Reinsurers is that there are many references to “event” in the 2019 Submission which are clearly talking about events in the AOE sense, i.e. events which have been cancelled. This, it is said, makes it reasonable to think that the AOE definition was apt for the Limits Clause. I disagree. Of course there are references to “events” in that sense in the 2019 Submission, but that is inevitable when one considers what the subject-matter of Contingency insurance was. However, none of this explains why AOE was apt in the Limits Clause, nor can one infer from the references to “events” in this way that both parties must objectively have been proceeding on the basis that AOE was to apply.

Notwithstanding the points made by the Reinsurers, I agree that overall, the clear impression conveyed by the 2019 Submission, is that both parties were proceeding on the basis that no distinction was being drawn between Class B Contingency and the other Classes for how catastrophe losses would be calculated and that this was on the basis of aggregation. It is therefore supportive of WRBC’s case on the Construction Issue.

Pausing there, if this was the extent of the contextual evidence submitted by WRBC, I would agree that, taken with the textual analysis set out at paragraphs 71 - 96 above, it is plain that the Construction issue should be resolved in favour of WRBC. Indeed, as already noted, I reach that view, even without recourse to the 2019 Submission.

However, I proceed to deal in any event with the other contextual matters relied upon by WRBC. It is to be noted that while the Reinsurers deny that they have any relevant part to play, the Reinsurers do not themselves advance any argument (at least not of any real substance) that such other materials, positively support its construction of the limits clause.

I propose to deal with them as follows:

WRBC contends that the formation, content and history of the Aggregate Treaty and the XL Treaty shows, first, how the AOE definition in the Treaty was indeed an inapt “relic” from its original use in the Aggregate Treaty, and second, how, objectively, the parties knew that the XL Treaty (and later the Treaty) was intended to cover “catastrophe” losses through aggregation. In both respects, therefore, these matters support WRBC on the Construction Issue. The Reinsurers deny this.

As a preliminary point, the following table shows which of the Reinsurers had participated in which earlier treaties as from 2015 to 2018:

I did not understand the contents of this table to be disputed by the Reinsurers. It can be seen therefore that only two of the Reinsurers (Blenheim and Chord Re) have no prior participation at all. Even then, they both received the 2018 Submission in late March 2019 from Mr Harris at JLT although in the event they did not participate in the 2018 Treaty. AXA (taking 15%, and which became the lead reinsurer dealing with Aon) participated from 2016, including in both the Aggregate Treaty and the XL Treaty in 2017. Liberty, the next largest of the Aon Reinsurers, participated for two years but not in the Aggregate Treaty, while Bermuda participated in two of these years including in the Aggregate Treaty. Of the JLT Reinsurers, Brit (taking 17.5% and the main contact with JLT) participated fully in all four years, while Trans Re did for three years, but not in the Aggregate Treaty.

As already noted above, WRBC first embarked upon a structured reinsurance for all of its classes of business in 2015. This was by means of the Aggregate Treaty and the XL Treaty. There is no material dispute between the parties as to how they were to operate.

I deal with the Aggregate Treaty first, and the evidence shows that it was drafted before the XL Treaty, although both are dated 1 April 2015. At that stage, there were just the two treaties, placed for WRBC by JLT. There was no other broker involved.

The Aggregate Treaty worked on the basis of deemed reinsurance programmes for each class of business called “Qualifying Reinsurances”. The limits clause for Class B, Contingency, read as follows:

One notes immediately the words “any one Risk, or any one Event”, and the absence of the words “arising out of one event” or a similar expression. As with the Treaty, there is a Conditions section which then contains a DEFINITIONS section which is in precisely the same terms as the DEFINITIONS section in the Treaty (i.e. defining AOR and AOE). The deductible of US$750,000 applies to AOR and AOE, as well as losses. The applicable limit is US$4.25m and there are no upper layers.

The Aggregate Treaty then provides as follows:

The first part of this provision therefore imposes a total maximum payable of US$20m. The second part imposes a further US$2.5m limit explicitly in cases where losses across more than one class are caused by a single “occurrence” which here plainly means a causative event. To this extent there is aggregation, but in a way which is adverse to the reinsured because it imposes an upper limit on losses across classes all deriving from the same event. Subject to that, each loss or series of losses has an upper limit of US$4.25 million.

Here, the application of AOE is appropriate, and that precise expression is used in the limits clause, albeit not wholly capitalised. What are covered are single losses or a series of them, single risks or single events in the AOE sense. As Mr Harris stated in paragraphs 30 and 31 of PH1, and as shown in the diagram contained in the 2015 Submission for this treaty, the Aggregate Treaty provided cover where WRBC experienced multiple individual losses over a year of up to US$5m (subject to the deductible) per loss with an aggregate total of US$20m. It therefore protected against what he described as “an accumulation of attritional losses, that is providing horizontal cover in respect of multiple unrelated individual losses occurring over the course of the year.”

The XL Treaty covered the same classes of business as the Aggregate Treaty. There were three layers. Layer 1 was US$20m in excess of US$5m, and this reflected the individual loss or risk limit in the Aggregate Treaty of US$5m. Layer 2 was US$25m in excess of US$25m, and Layer 3 was US$25m over US$50m, so the upper limit was US$75m.

The language of the limits clauses for each layer was the same and is the same as that contained in the Limits Clause in the Treaty, namely:

Equally, the DEFINITIONS section of the Conditions section for Class B is identical to that in the Treaty, and in the Aggregate Treaty.

The above language was repeated in the XL Treaty for the later years.

If one had to interpret the limits clause in the XL Treaty, then, for all the reasons given in the textual analysis for the Treaty, the result would be the same, namely, WRBC’s interpretation to be preferred.

I should add that in reaching that conclusion I have considered the Reinsurers’ point that the Aggregate Treaty and the XL Treaty treaties were complimentary and so the same definition of “event” should logically apply to both. It is, of course, correct that both were part of an overall reinsurance strategy for all of WRBC’s classes of business. It is also correct that they contained deductibles. However that hardly means that if the same definition was used in both, it must be operative in both if there are clear reasons to suggest that it is not. In addition, each treaty had a different overall structure and set of provisions, including, critically, differently-worded limits clauses. Further, while there was a simple “top-up” treaty for the Contingency business, this was not the (general) XL Treaty, but rather a separate treaty giving cover excess of US$5m up to US$6.25m – see paragraph 144 below.

In addition, the Reinsurers point out that it is not the case that the AOE definition was first created for the Aggregate Treaty in 2015. It had appeared in, for example, the 2014 class-specific per-risk reinsurance (“Risk Excess of Loss”) treaties. However, this is irrelevant for present purposes. WRBC is not saying that the AOE definition first appeared in 2015. Rather it is saying that having been used first in the Aggregate Treaty, aptly, it was then copied over to the XL Treaty, inaptly. Further, the “per risk” nature of the 2014 Treaty is shown by the fact that the limits clause was in the same terms as the Aggregate Treaty, in other words the specific expressions “Any One Risk” and “Any One Event” were used.

In that regard, the Reinsurers also point out that, WRBC had a separate “Clash Excess of Loss Reinsurance” (“the Clash Reinsurance”) for its contingency business, made at the same time as the Risk Excess of Loss treaty referred to at paragraph 134 above, which also uses the same AOE definition. I do not see that this assists, because the preamble to the Limits section of the Clash Reinsurance refers specifically to the Risk Excess of Loss treaty and the limits clause thereunder, and then to the fact that Reinsured “may from any one cause or related, similar or associated causes have to bear more than a single Risk retention (“clash loss”). In such a case the Clash Reinsurance provides for up to US$2.5m in respect of “each and every Loss Occurrence or series thereof, Any One Proximate Cause” in excess of US$1.5m, in respect of the same. This is therefore dealing with a situation where there were multiple losses on risks all arising from the same cause, hence the “Clash”. One understands why the AOE is there, because the Clash Reinsurance operates in the context of the Risk Excess of Loss treaty. However, the critical point is that AOE plainly does not apply to the limits clause itself, which makes no reference to an event at all. Indeed, if anything, the Clash Reinsurance is plainly providing cover in respect of one causative event which is what the limits clause is all about.

I now refer to the 2015 Submission. This first summarises the Aggregate Treaty at p5 with a diagram. This includes references to the different per-class qualifying losses and “RAD/LOD, Any One Risk/Any One Event as per class”. At p6 it then summarises the XL Treaty with a further diagram. In the diagram, the heading for Layer 1 reads: “Multi Class Large Risk / Catastrophe Excess of Loss”, the heading for Layer 2 was the same, under the heading for Layer 3 was simply “Multi Class Catastrophe Excess of Loss”. The rubric to the right of the diagram reads: “Protecting both large risk and event losses emanating from any one and/or combination of the above accounts greater than USD5m”.

What this shows, according to Mr Harris at paragraph 36 of PH1, is that the XL Treaty is providing protection for large (single) risks (in Layers 1 and 2), with all 3 layers covering multiple losses arising from the same causative event such as a “catastrophe”. I agree that this is the correct objective reading. It is clear from this and many other documents that when reference is made in communications between the parties to “catastrophe” what is being contemplated is some major causative event which gives rise to a whole series of different losses.

On that basis, the obvious inference is that cover for catastrophe would enable such losses to be aggregated with the consequence that in the absence of deductibles being applied to each and every loss, there is the potential for very large claims. In particular, there is no other sensible meaning to be attributed to “large [risk and] event losses”. It is impossible to read the word “event” here as meaning anything other than a causative event. I do not accept, as the Reinsurers suggest, that it is unclear what “event” means here. Of course, when it comes to describing the Contingency cover (at p10), reference is made to the provision of an indemnity “to Insureds should their event, through an insured peril and from no fault on their part, be cancelled abandoned postponed interrupted or relocated”; but this hardly means that the word “event” in relation to the diagram at page 5 means the same thing. Indeed it could not, because it is not suggested by the Reinsurers that the XL Treaty does not provide cover on an aggregated basis in respect of losses caused by one event for all the other classes. So there would have to be a distinction made so far as Class B is concerned, yet there is no such distinction in the 2015 Submission.

At p10, there is also a reference to the fact that “The Cancellation account results have been affected in recent years by adverse weather in the UK, Superstorm Sandy and the North East US polar vortex. As a result the underwriters have increased their Winter Loading in these areas”. I agree that this reference provides some additional support for the notion that what would be covered in the XL Treaty in this context would be catastrophe-induced multiple losses which would be aggregated. Of course, in theory, and as the Reinsurers suggest, all of that information could be relevant, even if the cover was on a simple “per risk” basis where the putative catastrophe caused losses in respect of a number of different, say sporting, events which were all cancelled as a result of that catastrophe. Each separate cancelled event would then attract its own deductible but if there were enough insured events with a very high line (up to, again, US$11.25 - see p25) they could in total lead to a very large claim. Therefore, references to storms etc. here might be of interest to reinsurers in any event. I follow all of that, but one has to remember that here, there was a deductible of US$5m which would lessen the overall claim. Further, it is worth noting that there are also references to Superstorm Sandy in the context of Class A, Fine Art where there was an overall loss of US$82m, which must have been on an aggregate basis. It is hard to see why the same would not apply for Class B.

It is true that the diagram at p6 for the XL Treaty does also state at the bottom of each of the three layers, “LOD, Any One Event”, “LOD, and Any One Risk / Any One Event” and “LOD Any One Risk/ Any One Event RAD Any One Risk as per class” respectively, and “Any One Event” is the phrase contained in the Definitions section of the XL Treaty. However, I do not consider that this removes the significance of the reference to “catastrophe” and “large… event losses emanating from any one” described above.

Accordingly, if one had to interpret the limits clause in the XL Treaty, the 2015 Submission would assist the proposition that the AOE definition here did not govern the limits clause.

I now refer to some correspondence from 2015. The first is an email from Mr Bird to Mr Harris and others, commenting on the draft XL Treaty. Among other things it says:

This clearly assumes that presenting a “cat” cumulative loss of US$12.5m would have only one deductible hence the net claim of US$7.5m, which means that there has to be aggregation. The ultimate figures here are the same, whether the loss was presented to the Aggregate Treaty or the XL Treaty, but that does not affect the point, nor do I accept that this email is “neutral” as the Reinsurers suggest. The words “event” in the context of “cat” here plainly means a causative catastrophic event.

In fact, in addition to the Aggregate Treaty and the XL Treaty, WRBC purchased some additional separate class-specific risk treaties including a separate contingency risk treaty, placed by JLT. This was because WRBC could not place all of its exposure under the new programme in 2015, although it was able to from 2016. See paragraphs 11-12 of PH1. In the additional contingency treaty the Class B definitions were again used. The Reinsurers say that this is of some significance but I do not agree. It was simply a specific top-up providing US$6.25m in excess of US$5m and contained the same limits language that the Aggregate Treaty did, as opposed to the XL Treaty. That is consistent with WRBC’s case.

If one takes stock, as it were, as at 2015, for those who were party to the relevant treaties, Submissions and correspondence (ie Liberty, Brit and TransRe), it can be said that they clearly understood what was intended for the structure and operation of the Aggregate Treaty and the XL Treaty. The use of expressions like “cat” and “large event”, the extent of cover under the XL Treaty, the reference to catastrophic events in the context of the latter and the absence anywhere of any suggestion that Class B was to be treated differently from the other classes, all suggest, singly or cumulatively, that losses caused by one event or “catastrophe” would be aggregated under the XL Treaty. That mutual understanding “rolls forwards”, as it were, to later years and it adds cumulatively to those parties’ respective knowledge as at 2018 and 2019. They knew that there had been a continuing structure of reinsurance which differentiated between the Aggregate Treaty and the XL Treaty where there was a fundamental distinction between what each of them was designed to cover, and where there was no basis for thinking that Class B was to be treated differently from any other class for the purpose of aggregation.

Of course, one needs to feed in to this equation the fact that in 2018, the Aggregate Treaty and the XL Treaty were replaced by the first of the single composite multi-line treaties. Whether, and if so how, that affected the position I shall consider below.

Before doing so, however, I should return to the question of the extent to which what happened before and when the Aggregate Treaty and the XL Treaty were entered into can assist on the invocation by WRBC, if it were necessary, of a Chartbrook corrective construction.

As to this, Mr Harris explains how the wording in the Aggregate Treaty and the XL Treaty emerged, as follows in PH1:

WRBC relies on that evidence, in part, to show that the AOE definition was lifted from the Aggregate Treaty which was drafted first and that this was a mistake. As to this, while I have no reason to doubt Mr Harris’s evidence as to the sequence of drafting, he does not say in PH1 that this was communicated to the relevant reinsurers at the time. WRBC says that, at least as far as JLT and Brit were concerned, this was known by reason of the email of 13 March 2015 from Mr Harris to Mr Bird of Brit. It said:

I agree that the email does show that what became the Aggregate Treaty was drafted first, and Brit can be taken to have known that. However, there is no evidence that this was communicated to any of the other putative reinsurers, and in fact, none of the Reinsurers entered into the Aggregate Treaty in 2015. On that basis, I do not think that this email is sufficient to show either that the Reinsurers (other than Brit) did know, or could have known that the Aggregate Treaty was drafted first. Nor can I safely conclude that, through Brit, all of those who entered the Aggregate Treaty in 2015 knew it or could reasonably have known it. The upshot is that I cannot take into account the fact that the Aggregate Treaty was entered into first.

As to whether the copying over to the XL Treaty of the AOE was a mistake, I agree with the Reinsurers that it is not clear that one can get this from what Mr Harris said in PH1. He just refers to copying over everything because both treaties were part of the same overall scheme. I would be prepared nonetheless to infer that what he meant (especially bearing in mind that he denies that anyone made the (contrary) suggestion that the XL Treaty did not provide for aggregation for Class B (as opposed to other classes)) was that the terms were all transferred over without giving much or any thought as to whether they in fact worked or not. Indeed, as a matter of common sense and given how contracts like this might develop, this thesis sounds plausible enough. However, the point is that, in my view, there is insufficient evidence to say that both sides must have appreciated that there was a clear mistake. Probably, no-one gave this potential anomaly any real thought, because it was never put to the test, as it were.

In this context, the Reinsurers have referred to the fact that the definition applicable to Political Risk business was changed in 2017, in respect of the JLT but not the Aon placement, in that the Definitions section which previously governed the definition of “one risk/contract” was added to by saying at the beginning that that coverage was to apply on an “each and every loss” basis which latter term was then defined. The rest of the Definitions section including the actual guidelines (i) to (iv) were exactly the same. Nonetheless, the Reinsurers argue that this particular alteration shows that there was at least some consideration of the Definitions section in that respect in that year, so it cannot be said that no attention was paid to them at all. I follow that, but whatever the reason for the change in the JLT-placed XL treaty, it does not mean that the other Definitions were considered or reviewed. In any event, as I explained below, the evidence as to how and why a mistake was made is of limited effect.

Recognising that the application of the corrective construction principle requires knowledge by the relevant parties or the reasonable observer, WRBC contends that irrespective of knowledge as to which treaty was drafted first, what could be said, objectively, is that if the XL Treaty was understood to provide for aggregation for all classes, it must follow that the AOE was inapt for the purposes of that treaty, which had a differently worded limits clause,although the AOE definition worked perfectly well for the Aggregate Treaty. That being so, and regardless of how or why it happened, but being conversant with the fact that AOE was in both treaties, the only logical view would be that it had appeared in the XL Treaty in error. On that basis, the fact of the mistake (if not the precise way in which it had been made) was clear and the required correction would have been obvious, namely not to apply it to the underlying limits clause.

In that way, it can be said, at least so far as the 2015 XL Treaty was concerned, that had its interpretation been the subject of dispute, the corrective principle would apply. In order then to get to the contract in question, namely the Treaty, one could then say that the presence of this clear mistake and its solution would apply to the Treaty itself as well as its forerunners.

Having said all of that, I do not in fact consider that the application of the corrective principle in this way is either required or appropriate. That is because, in my view, and unlike cases such as Chartbrook and Monsolar, it is not here clear and unambiguous that the AOE definition would otherwise govern the Limits Clause. Indeed, for all the reasons given in connection with my textual analysis above, that is far from the case. At best, the position is ambiguous and can be resolved in favour of WRBC, without recourse to the corrective principle. After all, this is not a case where any wording or syntax needs to be changed. Rather, it is a case where the only implication from the construction favourable to WRBC is that the AOE definition is redundant, at least for almost all purposes, but I have already concluded that any such redundancy argument makes no difference.

All of this is demonstrated by the fact that the basis on which it is said that the mistake here is clear is not because of some inherent peculiarity in the actual language or syntax but rather because the putative reasonable observer has already concluded that the contract must be interpreted in a way which excludes the operation of the AOE Definition. That then becomes circular, because if that is the position, there is no need for a corrective principle to operate. Overall, then the corrective principle cannot be invoked here.

What all of that means is that there is value in considering the position as at 2015, not in order to provide the basis for the application of the corrective construction principle, but as further support for a factual matrix argument pertaining as at 2019 to the extent indicated in paragraph 145 above.

In 2016 and 2017, the position was much the same. There were further Aggregate and XL treaties in materially the same form as before, although more of the Reinsurers joined. A second XL treaty was placed now by Aon (with AXA) as well as that placed by JLT. In 2017, Aon placed a second Aggregate treaty.

Submissions in materially the same form were made by JLT and Aon for 2016. They contained diagrams for the Aggregate Treaty similar to those in the 2015 and also described the XL Treaty as "protecting both large risk and event losses emanating from any one and/or combination of the above accounts greater than USD5m.” To that extent, the points made about the use of such language in the 2015 Submission apply equally here. Liberty was provided with the 2016 Submission, although it did not enter the XL Treaty for that year.

The schematic for the XL treaty is slightly different. As Mr Wakefield explains in TW1:

That analysis seems to me to be clearly correct and phrases such as “any one event” clearly connote one causative event leading to multiple losses. As with 2015, there is no suggestion that losses from one event under Class B were not to be aggregated whereas the other classes were.

Paragraphs 33 and 34 of TW1 refer to emails dated 21 and 22 March 2016 between Mr Jenkins at AXA and himself making reference to “per event potential” and “event cover” in respect of the intended XL treaty. These references to “event” are plainly in the sense of one causative event. As with the 2015 email referred to at paragraphs 142- 143 above, I do not accept that these emails were neutral. Rather, they provide further support for WRBC’s position.

The position for 2017, and the Submission made then, is essentially the same.

I note that in this context, the Reinsurers rely on an email from Mr Aikman at Aon to Mr Briggs at China Re dated 5 May 2017, which replies to a number of comments made by Mr Briggs in relation to the wording of the proposed 2017 XL treaty. Point 6 of those comments questions whether the definition of “Loss Occurrence” for the Construction and Engineering class of business was circular. Mr Aikman’s response here was:

The Reinsurers say that it is noteworthy that Mr Aikman did not at this time suggest that the Definitions section dealing with Contingency was in any way inapt, or included by mistake. I fail to see how this goes anywhere, since he was dealing specifically with one particular provision relating to C & E business. He was not asked to conduct a review of all the Definitions or considering particular the AOE definition.

Accordingly, the mutual understanding of the parties for 2015, is now added to, with the same effect, through 2016 and 2017. By 2017, AXA, XL Bermuda and Brit were parties to the Aggregate and the XL treaties, while Liberty, HCC and TransRe were parties to the XL Treaties. That is now a sufficient proportion of the Reinsurers (being all of them, save Chord and Blenheim, each with 2%) to say that the Reinsurers had knowledge of the XL Treaties in particular, as set out above, and it was in any event information reasonably available to them as at 2019. That information yielded the understanding of the structure behind in particular the XL Treaty and how it was designed to provide for aggregation in respect of the same causal event across all classes of business. That is not some arcane point of detail which may have been lost on the Reinsurers as at 2019. It was a simple structural understanding of how the cover they had been providing was to work and its commercial purpose.

In all material respects, the 2018 Treaty was in the same terms as the Treaty. It now of course replaced the previous Aggregate and XL Treaties. All but four of the Reinsurers subscribed to the 2018 Treaty and the three which did not, Chord Re, HCC, Blenheim and TransRe (all at 2%) received the 2018 Submission anyway.

At page 4 of the 2018 Submission, it refers to the underlying performance of the portfolio from a “risk and catastrophe” perspective. There is again reference to storms and I make the same point here as in other Submissions discussed above. The diagram at p29 shows the low level exposure of the individual losses for Class B, as with the 2019 Submission. Again, there is the specific qualification for C & E cover, but no suggestion that Class B would be treated differently from the other classes.

Because the prior year losses engaged the previous Aggregate and XL treaties, there is specific reference to them both at p30 and 31. On the former treaty there was a claim for just over US$2.5m for the cancellation of the SurfShow due to Hurricane Irma and then for the XL Treaty there was a contingency loss of over US$3.6m related to Hurricane Irma. This included the SurfShow losses of US$2.5m but clearly exceeded them. In other words, it was an aggregation of different losses arising from Hurricane Irma. In addition, as Mr Harris points out at paragraph 53 of PH1, the loss data included the "Cat Codes" which Lloyd's assigns to each catastrophe for the purpose of tracking multiple individual losses arising from the same event; so, for example, the losses in relation to Hurricane Irma are provided against the Cat Code 17G. It is therefore beyond doubt that this record of losses under the XL treaty was on an aggregated basis for Class B.

Indeed, paragraph 92.3 of the Reinsurers’ Skeleton Argument accepts as much in terms. It is correct that this particular information was not in the 2019 Submission, but that does not remove its cumulative effect, as it were, in showing the parties’ overarching understanding of how the XL Treaty was intended to work. The Reinsurers further say that because of the relatively modest size of these particular Contingency losses, the Reinsurers reading about them would not have paid much attention to them. I do not accept this because, yet again, it just demonstrates their general understanding. I take the same view of the argument that because there was now a new structure, the Reinsurers were likely to have focussed on losses in excess of US$2.5m going forwards. All that had really happened was that the Aggregate Treaty (which is what had provided the cover up US$2.5m) had now been dropped. See paragraph 23 of TW1 and paragraph 29 of PH1. Anyway, the 2018 Submission saw fit to include these losses and of course the US$3.6m losses in question did exceed US$2.5m.

It is then said that the reader of information contained in the 2018 Submission should not be required do “detective work” so as to recall this particular point in the 2018 Submission, which was not repeated in the 2019 Submission “in order to draw the inference that WRB did not consider the Treaty wordings to work in accordance with the terms.” As to that, first, I do not see why the 2018 Submission should not be regarded as part of the materials reasonably available to the Reinsurers for the purpose of construing the Treaty made a year later. Second, even if it was not strictly reasonably available to the Reinsurers, as at 2019, it can still be relevant so as to support the overarching understanding, first of the separate XL Treaty and then the 2018 Treaty and then the Treaty itself. It needs to be recalled that four of the Reinsurers (including Brit and AXA) entered into the 2018 Treaty and the others were provided with the 2018 Submission. Third, the point made by the Reinsurers which I have quoted assumes that the starting point for the Reinsurers was that the AOE definition was understood to apply and then they had to do their detective work to disapply it. That, however, assumes that the Reinsurers’ interpretation is correct to begin with. Finally, if the Reinsurers really wanted to make points of this kind about pre-contractual materials, and what they considered or could reasonably have considered about them, there was nothing to stop the Reinsurers putting in evidence to this effect, in particular from those who were party to the various emails referred to above and who read the 2018 Submission. Not only is there no such evidence, but there was no challenge to WRBC’s evidence in this regard.

I now refer to some correspondence from 2018. First, there is the email from Mr Harris to Ms McKenzie of TransRe dated 5 March 2018. The two last bullet points read as follows:

One sees here the clear and familiar distinction made between risk and catastrophe losses in these bullet points, and the reference to exposure “on [the on] an event basis” is to a single causative event. This email in fact ends by saying that there would be a restructure in 2018 with the XL treaty now being $72.5m excess of US$2.5m and the Aggregate treaty no longer being purchased. Subject to figures, therefore, the essence was simply that the latter treaty would now fall away.

Then, on 9 March, Mr Harris emailed Mr Bird (at Brit), making a similar distinction between catastrophe and per risk exposure:

One of the attachments to this email was a spreadsheet showing the RDS modelling results. One can see the figure of US$13.2m for communicable disease in respect of Contingency. While the US$13.2m figure may not seem very high, it is obviously higher than the US$11.25m Line Guide, and see also paragraph 111 above. It is not to the point that the modelled RDS loss was less than WRBC’s exposure on one particular risk, that is, the Tokyo Olympics, which was up to US$15m agreed by the “special acceptance”. That was a special case, and the real point is that communicable disease is a catastrophe scenario because it can cause losses on multiple risks.

Mr Harris’ later email to Ms McKenzie dated 20 March 2018 also included the spreadsheet, which he said was giving a breakdown of RDS by class.

Information about historical losses was also supplied to Mr Berry of AXA by Mr Aikman of Aon on 8 March. This in included a reference to just under US$3.5m for Hurricane Irma losses, as to which see paragraph 169 above.

By reason of the matters referred to above, I consider that the understanding to be gained from the 2018 Submission and associated correspondence was, again, that the structure of the XL Treaty was to provide against catastrophe losses across all classes of business, which supports the notion that there was to be event-based aggregation for all of them in the manner contended for by WRBC.

It follows from everything which has been referred to in paragraphs 118 to 178 above, that if (contrary to my findings at paragraphs at paragraphs 96, 114 and 115 above) further contextual support for WRBC’s case on the proper meaning of “event” for the purpose of the Limits Clause is required, it is to be found in the materials cited from 2015 onwards.

For all of the above reasons, it is clear, in my judgment, that WRBC must succeed on the Construction Issue. “Event” is to be construed as meaning the event which causes the relevant losses, rather than in the terms of the AOE definition. Accordingly, and as is common ground, it is therefore necessary for me now to determine the Aggregation Issue.

The Aggregation Issue is concerned with the amount of monies due from the Reinsurers to WRBC in respect of the 174 individual losses claimed. Each of those losses is in respect of a separate cancelled conference, assembly or the like.

This gives rise to two overarching questions in respect of each such loss:

In this judgment I will need to address each of those questions. The answer to the second will depend on my answer to the first, once all the claimable losses for each of the 7 jurisdictions have been determined.

However, it is not then necessary for me to calculate the particular amounts of claimable losses in their entirety so as to reach a total figure. This can be worked out by the parties in the context of the consequential matters to be decided, once I have given judgment.

To that end, the Reinsurers have produced (at A/19) a spreadsheet which shows what their liability would be on the assumption that they are correct as to what losses have been caused by what event (if any). However, the key elements of the spreadsheet can be changed to reflect my determination of the losses. This is likely to be a useful starting point for any consequentials hearing, but it was only produced on 4 February 2026 and WRBC says that it has some errors in it. At any rate, I am not going to use it at this stage to arrive at particular figures.

Rather, and for present purposes, there are three important working documents produced by the parties:

The breakdown of the claimed losses by jurisdiction is as follows:

Each party has separately labelled the measures in each jurisdiction upon which it relies or refers to. In some cases, WRBC has pointed to more measures than the Reinsurers and in other cases it is the reverse. Since it is not the case that, for example, the Reinsurers always refer to measures relied upon by WRBC (or vice versa) I will have to make reference to both parties’ designations where necessary, when considering the individual jurisdictions. In any event, I will make clear which particular measure I am referring to.

Given the number of items of loss in issue, I do not intend to produce a detailed narrative for each and every one of them in this judgment. Instead, I will deal in this judgment with the losses claimed to have arisen in California and England and that should exemplify my approach for the other 5 jurisdictions. For losses in the latter jurisdictions, I have set out my determination of them in a separate Appendix to this judgment.

Before doing all of that, however, I need to set out the parties’ position on the losses and then deal with certain points of law.

It is common ground that in principle, a measure which has an impact on the viability or otherwise of a conference or other event insured by WRBC within its Contingency business, can constitute an “event” for the purposes of the Limits Clause in the context of the Aggregation Issue.

For each jurisdiction, WRBC has set out a number of alternative causative events. These are summarised at paragraph 116 of its Skeleton Argument as follows:

Those sets of alternative events are then particularised for each jurisdiction. Thus, to take California as an example, paragraph 31 of the Re-Amended Particulars of Claim states as follows:

As can be seen from those passages, WRBC’s case is that whichever is the appropriate event, it caused all of the losses claimed for that particular jurisdiction. In other words, there would only be one deductible to be applied, because all the losses would arise from the same event. So, there would be the maximum aggregation possible.

That said, it is, of course, possible that within a particular jurisdiction one loss may have been caused by one event and another caused by a different event. If so, more than one deductible would be applied since those particular losses would not be aggregated together. Plainly, it is most advantageous for WRBC’s case if the first iteration of its event alternatives, namely a combination of measures, was correct because there would then be the least difficulty in attributing all the losses (whenever they occurred) to it.

As against that, the Reinsurers’ position is as follows:

In general, it is common ground between the parties that a particular Covid-related measure imposed in any of the 7 jurisdictions can amount to an “event” for present purposes. The one major exception is where there is declared a state-wide state of emergency. Such declarations are sometimes relied upon by WRBC, whereas the Reinsurers deny that they could constitute a relevant causal event. Save in a relatively small number of losses, and for the reasons set out below, this issue between the parties does not affect the outcome. The real debate between the parties is whether any particular measure caused the relevant loss, and if so, which one. As Mr Templeman KC put it in argument, the key question here is not whether there has been some “break in causation” in respect of the loss claimed, but whether any relevant measure has caused it in the first place.

I should add that while there was originally a dispute between the parties as to whether an event which was postponed rather than cancelled would be covered, it is now agreed that a postponement “counts” for these purposes.

For present purposes, it is common ground in the light of the case-law that:

Further, the various measures which were imposed in each of the 7 jurisdictions in question are set out in the Chronology, which lists all of the measures in chronological order and indicates for each measure which party places reliance upon it. As already noted, for the most part there is a considerable overlap, but the initial measures tend to be relied upon by WRBC (in particular the declaration of a state of emergency or equivalent), while there are a number of later measures relied upon only by the Reinsurers.

As for other points which are the subject of disagreement between the parties, both WRBC and the Reinsurers made extensive reference to the trilogy of cases involving insurance claims for business interruption losses arising out of Covid. These were all determined by Butcher J in Stonegate Pub Co Ltd v MS Amlin Corporate Member Ltd [2023] Lloyd's Rep IR 672, Greggs Plc v Zurich Insurance Plc [2022] EWHC 2545 (Comm), and Various Eateries Trading Ltd v Allianz Insurance plc [2022] EWHC 2549 (Comm).

In Stonegate, which was a trial of preliminary issues, the claimant made claims against insurers under business interruption policies covering 760 pubs which it owned and operated. Insurers argued that these losses should be aggregated and if so, an overall limit of £2.5m would apply. Absent aggregation, the claims were worth very much more. The relevant events concerned the various measures taken in the UK in relation to the Covid pandemic. The key measures are the same as those relied upon by WRBC here for the purpose of the claims made in respect of losses incurred in England. The £2.5m limit was said to apply to, among other things, one Single Business Interruption Loss (SBIL) which itself was defined as:

It is common ground that the language of the limits clause there was broader than the language of the Limits Clause here, as a result of which the causal link which had to be shown was somewhat weaker.

At paragraphs 79-90 Butcher J set out principles which were approved by the Court of Appeal in Various Eateries Trading Ltd v Allianz Insurance plc [2024] 1 Lloyd’s Rep 490. They are as follows:

In the event, Butcher J held that the decision which had been taken at the government’s COBRA meeting on 16 March 2020, that the public should be advised to avoid pubs, restaurants and clubs was a single occurrence. Further, from the perspective of the claimant there, which was the operator of the pubs and clubs across the country, this decision would be regarded as a unitary matter which would have effect on its business throughout the country rather than only at local level whereby the impact on each venue would have to be differentiated individually.

Butcher J then went on to consider a further argument raised by insurers which was to the effect that it could rely upon a number of measures and announcements which took place between 16 and 26 March 2020 so as to constitute a single occurrence. As to that, he stated as follows:

Finally, Butcher J accepted the third “single occurrence” proffered by insurers which was the instructions given to all pubs, bars and restaurants to close on 20 March and not to reopen the next day.

In respect of both the 16 and 20 March single occurrences, Butcher J could not determine at that stage which losses could be said to arise from, be attributable to or connected with either of them. However, he did add this:

That passage has a particular resonance for the case before me.

It also follows from Stonegate that it cannot be said that a number of measures can never collectively constitute a single occurrence or event. It all depends on the nature of those measures and their factual context. That is why Butcher J distinguished Silversea, rather than suggest that it was wrong in law. I will therefore have to consider WRBC’s “combination” argument (which the Reinsurers reject) separately for each jurisdiction, albeit recognising that the nature and sequence of the measures introduced have broad similarities. It is also necessary to pay particular regard to what Butcher J held in respect of the measures introduced in England, since that it is one of the 7 jurisdictions in issue here.

Greggs wasanother case involving aggregation where, again, Butcher J had to identify relevant occurrences, this time in respect of the closure of a number of Greggs shops due to Covid measures. At paragraph 86, Butcher J said as follows:

Again, this has a particular resonance for the case before me. As WRBC has pointed out, Butcher J’s reasoning here was described by Males LJ (giving the judgment of the Court of Appeal) in Various Eateries Trading v Allianz [2024] EWCA Civ 10 at paragraph 84 as “clearly right”. The same view was taken by Dias J in World Challenge Expeditions v Zurich [2024] 1 All ER (Comm) 786 at paragraph 322, where she noted that the continuation of restrictions (as opposed to their initial imposition) cannot be said to be an identifiable and separate occurrence for the purposes of aggregation.

What all of this means is that it is not possible, as a matter of law, simply to assert that the measure which immediately precedes the relevant cancellation must be the event which caused it, as opposed to some earlier event. Further, the fact that a later measure continued the earlier measure does not deprive the earlier measure of its causative force. Equally, the fact that a later measure reduced the impact of the earlier measure (but not so as to affect the cancellation in question) is irrelevant. This also follows from what Butcher J said in paragraph 192 of Stonegate (quoted above).

Another aspect of Greggs relates to a concern which had been expressed by insurers about any findings to the effect that there were multiple relevant occurrences, because this would lead to difficulties in assigning losses to them. Butcher J treated that argument with caution on the basis that calculations had already been made of Greggs’ losses on a regulation-by-regulation basis and there was no evidence that such an exercise of attribution would be particularly difficult or impossible. What he did say was this:

It is not entirely clear to me what is being postulated as a substitute if it could not be said that there were two “singular occurrences”. One possibility is that they would both be regarded as comprising a single occurrence. That would appear to reflect the observation Butcher J made at paragraph 187 of his judgment in Stonegate (quoted at paragraph 206 above). If so, then it would depend on the facts of each case whether there was any such real impossibility. Otherwise, he may have been suggesting that one would have to regard both announcements and regulations. If so, there would be no particular problem because, on the basis of what he said at paragraph 192 of Stonegate, one would simply take the relevant occurrence as being the first in time.

I should add that in any particular case, ascertaining whether a measure relied upon by WRBC was an effective cause requires consideration of the facts. Where the relevant measure is referred to by the relevant insured or loss adjuster, either specifically or by inference (for example state bans on public gatherings) that is clearly a strong pointer and indeed much of the time in argument was taken up by looking at such references (or the absence of them). However, that does not mean that a relevant measure could not be an effective cause, simply because it is not referred to, when it is obvious that the event could not take place without the promoter being in breach of it. In such a case, it should be regarded as an effective cause. Such a measure is only likely not to be regarded as an effective cause where it is plain from the documents that there was some other matter which was entirely separate from Covid-related restrictions that was in fact the cause of the cancellation. That said, and as will be apparent from my findings, in almost every case where I have upheld the loss claim because it is caused by a measure relied upon by WRBC, there will have been a specific or inferential reference to it in the documents.

A further issue which has arisen is this: as paragraph 149 of WRBC’s Skeleton Argument makes clear, there are some pleaded losses concerning events which were cancelled or postponed prior to the date of the final measure relied upon by WRBC. In one case, there was no preceding measure at all. In most cases, the preceding measure relied upon by WRBC was the declaration of a state of emergency or equivalent. The Reinsurers for their part do not accept that such a declaration amounted to a relevant measure at all.

WRBC’s argument is that the fact that the relevant cancellation occurred before a particular measure does not mean that such a measure cannot be relied upon as causing the cancellation. Partly that is said to be because the subsequent measure was “imminent” or could be seen as inevitable once there had been a declaration of a state of emergency. The Reinsurers deny this.

The starting point is, of course, the words of the relevant provision. Here, they are “arising out of” one event. That very much suggests that the latter must precede the cancellation. One case where this issue arose was the decision of the Court of Appeal in Scott v Copenhagen Reinsurance Co (UK) Ltd [2003] Lloyd's Rep IR 696. Here, following the Iraqi invasion of Kuwait in 1990, Iraqi troops seized Kuwait Airport and then removed 15 aircraft owned by Kuwait Airways (“KAC”) to Iraqi. Two of the KAC aircraft were subsequently destroyed at Mosul airfield in Iraq, following the attack upon it by coalition forces in January 1991. At the time when Kuwait Airport was seized, there was also present a BA Boeing 747, which was not removed by Iraqi forces and remained there. It was subsequently badly damaged while still at Kuwait Airport, apparently by coalition fire.

Claims were brought against reinsurers by the insurer of both the KAC and BA aircraft, which argued that the losses claimed should be aggregated because the loss of all the aircraft arose from one event, namely the invasion. The Court of Appeal held that, unlike the KAC aircraft, the BA aircraft was not lost as a result of the invasion and capture of the airport. The point was that this aircraft was not damaged or seized by Iraqi forces at the time. Its damage came later, in different circumstances.

In that context, at paragraph 61 of his judgment, Rix LJ approved the approach taken by the Court of Appeal in Caudle to the operation of an “arising out of one event” clause, as follows:

The Reinsurers here also point to the obiter observations of Males LJ in Various Eateries at paragraph 69 that government instructions or subsequent legislation could not amount to an aggregating occurrence in respect of Covid-related losses suffered here before 16 March 2020.

As to that, WRBC first relies upon FCA Supreme Court in the context of an issue there as to whether, for the purposes of business interruption claims due to Covid, the relevant restriction imposed which caused the loss had to have the force of law. The Court held that while one would generally expect “restrictions imposed” to have the force of law or carry the imminent threat of legal compulsion, such a phrase was not limited to an exercise or threatened exercise of legal powers. The example given was where the Prime Minister on 20 March 2020 instructed named business to close “tonight”, stating that this was an operative mandatory instruction which the public would understand had to be complied with. I follow all of that, but fail to see how it assists WRBC in its present argument. Of course, in that case, it meant that there was an operative “restriction” to cause the relevant loss, where the relevant loss was then followed by a particular order with the force of law, but all of that is about the nature of the restriction relied upon. It is not concerned with a situation where the measure relied upon actually post-dates the loss.

I can quite see that if the relevant measure was announced on a particular day to take place in, say, two days time, as a result of which a conference was cancelled because it was inevitable that such a measure would prohibit it, that announcement could be sufficient to operate as a relevant causative event. However, that is not how the case is put here. WRBC does rely upon a prior measure in most of these cases, which is the declaration of a state of emergency. That is not the same as announcing particular measures which are about to come into force, even if it might be possible to contemplate that with a declaration of a state of emergency in relation to a pandemic, one might expect further specific measures to follow. That is not sufficient, in my view, to say that the declaration of emergency is itself, in all cases where it is necessary to rely upon it, a sufficient causative event from which everything else follows.

A second argument deployed by WRBC relies upon the decision of the Privy Council in Canada Rice Mills Ltd v Union Marine & General Insurance Company Ltd [1941] AC 55. In this case, an insurance claim was made for damage to the ship’s cargo of rice which was found to have been heated up, on arrival at the port of discharge. The question was whether this was caused by an insured peril. The proximate cause of the heating was the closure of cowl ventilators and hatches during the voyage. That closure had occurred so as to avoid seawater entering the cargo holds through the ventilators and hatches during a storm. While the damage to the rice was not caused by the storm in the sense of creating the incursion of seawater, it was caused by steps taken to avoid such incursion and the damage which that would have caused. That being so, this was sufficient to say that the steps taken which resulted in the heating of the cargo with themselves the result of the storm. It is important to note, as Lord Wright said at page 71 of the judgment, that this was not a case of where a ship had deliberately changed course so as to avoid the peril, i.e. the storm, where the loss might not have been attributable to the peril so as to be recoverable. In this case, as in other cases, the subject of the insurance was in “the grip of the peril”. The analogy would be with insured goods damaged by water which was used in order to put out the relevant peril i.e. a fire.

However, I fail to see how that case assists WRBC either. Here, the actual loss is the same, whenever it would have happened, namely the cancellation of the conference. This did not occur because the insured was in “the grip of a peril”. The “peril” i.e. a measure which would prohibit the conference taking place, the putative “storm”, in other words, had not yet emerged.

In this regard, WRBC also referred to what Jacobs J said at paragraph 791 of his judgment in ABN Amro Bank NV v Royal & Sun Alliance Insurance Plc [2021] EWHC 442 (Comm), where he referred to the notion of mitigation of loss and the “grip of a peril” principle, but I do not see how this takes the matter any further for this case.

For those reasons, I do not accept that there will have been a measure constituting a relevant causative event in respect of those claims where there was no prior (or co-terminous) measure, or the only measure relied upon is a declaration of emergency or similar.

There was a further issue between the parties in this context in relation to cases where an event was initially postponed but then cancelled altogether. The Reinsurers say that in such cases postponement is the relevant factor because it came first and/or because some losses may have been incurred as a result of the postponement itself. I do not agree. The fact that there was a postponement first cannot affect the fact of the later cancellation which took place. Nor did the postponement cause the cancellation. Further, it is obvious that the real losses incurred by the insured will have resulted from the cancellation, not the postponement. Even if the latter caused some initial loss (which may be difficult to discern from the claim documents), it makes no difference in my view because postponement and cancellation are two different things. Thus, I agree with WRBC that the key factor is the cancellation, not the postponement regardless of the lead-up to the cancellation, as it were. That is how claims which involve both tended to be put. My determination of this issue is relevant for those cases where, for example, it is said that there was no relevant causative measure before a postponement, but there was by the time of the cancellation.

Finally, WRBC contended that when it comes to the assessment of the losses for the purpose of this reinsurance claim, a broader and more flexible approach should be taken to whether they qualify, than that taken by an insurer such as WRBC, when considering whether or not to admit the underlying claim as falling within an insured peril or not. That is not a matter which I need to decide (and in any event, it was put only on a high-level basis by WRBC), because I am satisfied that the way in which I have assessed the losses is correct having regard to the fundamental causal approach which is at the root of this exercise.

First, the essential material adduced by WRBC in respect of the losses claimed for each of the jurisdictions consists of letters of claim by the relevant insured and/or reports to WRBC by loss adjusters in respect of the individual insured’s claims. I have read and considered all such materials, even if I do not make specific reference to them in the context of any particular claim.

Second, I have had to treat such materials with a degree of caution because the loss adjusters’ in particular are often made considerably later than the cancellations to which they refer, by which time further measures may have been adopted by the relevant jurisdiction. Obviously, where those reports refer to such later measures, as they often do, the later measures must be disregarded.

Third, the relevant measures are often applied to gatherings of particular size, for example, 50 or 250 or 500 people. The claim documents and loss adjusters’ reports sometimes say what the expected number of attendees was, but sometimes they do not. In the latter case I have exercised my judgment to determine what the likely attendance size was which will then be relevant to the applicability or otherwise of a particular measure.

Fourth, in a number of cases, it will be apparent that there were two particular measures which were each candidates to be an effective cause of the cancellation. In each case, I will set out which measure should be regarded as an effective cause. However, where this situation arises, it often does so in relation to a large number of the claimed losses. This means that if Measure 1, was not an effective cause, it would then have to be Measure 2. An example would be Nevada where the choice would be between WRBC’s NV2 and NV4. In those cases, the ultimate outcome, which is concerned with the application of deductibles, is unlikely to be very different, if different at all.

Fifthly, often, the question arises as to whether one of the earlier measures which would have entailed the cancellation of the event is concerned, was still in force at the date of cancellation. I have, again, exercised my judgment on the facts to work out whether it was or not and where necessary, have consulted the terms of the actual measure.

Sixthly, and consistent with the approach of Butcher J in Stonegate, the fact that an event is cancelled at a time when it was not benefiting from a relaxation of an earlier restriction, does not mean that it was cancelled because of that relaxation. It could not be because the relaxation did not apply to it. Normally, what that means is that the prior restriction will remain an effective cause. I make this point because the Reinsurers sometimes rely upon the separate relaxation measure, simply because it is the measure most proximate to the cancellation but that is not the correct approach.

Further, the Reinsurers on occasion have contended that WRBC is, or must be relying upon a measure which it has not in fact pleaded. I deal with such a point on a case-by-case basis. I should, however, say that there are many instances where the Reinsurers positively assert that the likely cause is a measure which they have referred to but which is not relied upon by WRBC. In those cases, there can be no objection at all if I were to find that, for example, a large number of losses in a particular jurisdiction had as their effective cause one of the measures cited by the Reinsurers.

Also, in relation to a number of the claims, the reasons for the cancellation refer to a number of different measures, some of them federal, as well as the pandemic generally. The fact that all of these are mentioned does not mean that where a measure relied upon by WRBC is also mentioned, or is plainly a reason, that measure cannot be an effective cause of the cancellation. That is particularly so where it is plain that under such a measure that event could not take place. For that reason, when considering the claims in detail, as I do below and in the Appendix, I focus on where reference or reliance is made upon one of the measures invoked by WRBC; the absence of references to other measures or matters by me does not mean that I did not consider them.

Finally, and on any view, my determination of each of the 174 losses claimed should be sufficient to enable the parties to calculate, or agree, a final figure. Where I have not expressly disallowed a claim, I have allowed it, and this should be obvious from the context, anyway.

I have set out at paragraph 193 above the three particular measures in California relied upon by WRBC either as amounting collectively to a single event or individually so. To reiterate, they were:

The first question is whether all of these, or the second and third, amount to a single event. In my judgment, and following the reasoning of Butcher J at paragraphs 186 and 187 of his judgment in Stonegate (see paragraph 206 above), they clearly do not do so. Each of them is of a different character, albeit arising in quick succession. While the first is general in its terms, the second and third each have specific, but different consequences for gatherings. Nor can it seriously be suggested that in the case of any given loss, it is impossible to attribute it sensibly to one of them as opposed to another of them. In addition, of course, it would be entirely possible for at least one of these measures not to have any effect at all; for example if the gathering concerned was clearly less than 250 people, it would not be affected by the 11 March Announcement, but only by the 16 March Announcement subject to any argument that it was already affected by the 4 March Proclamation.

That being so, it is necessary to consider each of the losses claimed by reference to those three measures or the further five measures relied upon by the Reinsurers. The former three measures are WRBC’s CA1, CA2 and CA3. WRBC’s CA2 and CA3 correspond to the Reinsurers’ CA1 and CA2. Unless indicated otherwise below, the references here are to the Reinsurers’ labelled measures.

The loss adjuster claim letter dated 25 April 2020 states that the new claim was received on 17 April, which means that the event cannot have been cancelled later than 17 April. The letter refers to the Reinsurers’ CA1, CA3 and governance guidance on 15 March asking for closures of bars and nightclubs and restaurants to focus on take-out (which may or may not be their CA2). The letter also states, however, that the insured advised that after CA1 which limited gatherings to 250 or less and implemented social distancing, the company cancelled the event. It plainly follows that the Reinsurers’ CA1 caused this cancellation, with the intended number of delegates being 250 to 300. That is accepted by the Reinsurers. The fact that CA2 would also have applied is irrelevant because of the pre-existing effect of CA1.

The claim letter dated 1 May 2020 states that cancellation was on 16 March. While referring to a number of federal and other measures, the letter stated that on 15 March Governor Newsom, while not yet ordering bars and restaurants to close, said that bars should close in California and restaurants should focus on ‘take out’ for those isolating. It was this which prompted the insured to cease operations on 16 March. It also stated that “as of today” (i.e. 1 May) the public gathering bans remained in force. However, the cancellation could not have been caused by CA1, because that only applied to gatherings over 250 persons. CA2 would have done, and it was co-terminous with the cancellation, and would inevitably have prompted it at the same time. Although not relied upon in the claim letter I consider that it should be regarded as an effective cause here.

The claim letter of 23 March 2020 shows that cancellation was on 23 March 2020 which is common ground. It makes express reference to both CA2 and CA3, in addition to CA1 (as well as federal measures). The number of attendees was 1,800. The immediate catalyst for the cancellation at that time appears to be the fact that CA3 would prevent the holding of face-to-face planning meetings for the event. However, holding the event would have been impossible in any event because of CA1, which is referred to. I consider that this is at least one of the effective causes of the cancellation which is sufficient.

It is common ground that these were cancelled on 8 April; see the announcement of that date. This simply said that “Sadly, but obviously, the face-to-face trainings below have been postponed.” The Reinsurers say that this cancellation would appear to arise from CA3. While this is nearest in time to the cancellation, and may have been one of the causes, the intended training sessions would inevitably have been cancelled as a result of CA2. For that reason, it was an effective cause.

A claim letter dated 2 February 2022 states that camping events ceased “effective 11 March 2020.” This letter does not refer to any California measure causing that cancellation but rather some specific federal measures. It does go on to say that on 19 March 2020 there was CA3 with large gatherings ultimately being banned in California until well into mid-2021, and with California restricting large gatherings through the Summer and Fall of 2020. In fact, there was an earlier claim letter dated 29 September 2020 which stated much the same thing. Although the Reinsurers say that it appears that CA3 caused this cancellation of bookings, it could not have done so, since this was only ordered on 19 March. In this case, the only candidate for a causative event was the proclamation of a state of emergency on 4 March. However, for reasons given in paragraphs 217 - 228 above, this is insufficient. The fact that subsequent measures would have caused a cancellation of bookings, or prevented any resumption of bookings at a later stage (even if only 5 days later, as here) does not in my view make a difference. In fact, CA1 occurred on 11 March itself. I would have been prepared to accept this in principle as a cause, except that it is far from clear that any particular scout camp would have involved 250 or more attendees, so it cannot count. Accordingly, this particular claim should be disallowed.

It is common ground that this event, together with those which are the subject of Claims 22 and 27, was cancelled on 25 March 2020. This was stated in Gartner’s SEC filing of that date, which said that it had cancelled or postponed all conferences scheduled for April through August 2020 “in consideration of guidelines and travel restrictions implemented by governments around the world” due to Covid-19. An individual announcement said that this event was being postponed “Due to ongoing concerns about the coronavirus”. Attendance was to be 1,800. Although the cause was stated in general terms, and it appears that Gartner cancelled events worldwide at this stage, I consider that the reference to guidelines implemented by governments is sufficient to include CA1, and this was an effective cause for present purposes.

The attendance here was to be 2,000. Otherwise the position is the same as for Claim 16, and the answer is the same.

The position here is the same as for Claim 16, and the answer is the same.

This was cancelled on 15 June 2020. There were to be 1600 attendees. The announcement of the cancellation stated that recent developments had made it impossible effectively to execute collective shows. This was said to be due to a number of different developments including federal measures, stay-at-home orders and restrictions on congregating and large gatherings imposed by states. The Reinsurers say that this would appear to be a result of CA4, a Public Health Order dated 7 May 2020 which permitted the re-opening of lower risk workplaces and other spaces, but would not in fact affect this event. However, that is simply a modification of existing restrictions and cannot be seen as a cause of the cancellation here. The, or at least one effective cause was, and remained, CA2. There are other documents in connection with this claim referring to the status of Anaheim’s reopening plan and CA3, but in my judgment neither of these deprive CA2 of its causative effect. Indeed, the insurance claim document referred among other things to current restrictions on congregating and large gatherings.

It is common ground that this was cancelled on 1 July 2020. The claim letter states that reasons for this included government orders limiting the size of public gatherings. CA2 is also mentioned specifically. The expected attendance is not stated but I am prepared to infer that it was more than 250, in which case CA1 made cancellation inevitable. It may be that CA2 would have put the matter beyond doubt but, I consider that CA1 was still an effective cause.

It is common ground that cancellation was on or before 3 July. Attendance would have been 56,000. The letter from the insured to loss adjusters referred to reasons for the cancellation including “local orders regarding large public gatherings” which would prevent many 2020 ASW events from being held in July and then referring, by way of an example, to a California state order and a Los Angeles local order banning public gatherings of any size. While the California state order is likely to have been CA2, with the event being this size, CA1 would plainly have prevented it, and there was a reference to orders banning large public gatherings as well. Accordingly, CA1 was an effective cause.

It is common ground that the cancellation was made on 1 July 2020. The claim document made reference to CA3 and CA4. Reasons for cancellation referred among other things to California state restrictions on large group gatherings. It is not clear that the individual events which appear to have been arranged to take place at wineries in San Francisco and Monterey, would have been for more than 250 people, and in fact J8/60/24 suggests that the attendance was around 113. This event would inevitably have been prevented by CA2. In my judgment, this was an effective cause of the cancellation.

The venue, the Moscone Center, notified the ADA on 13 November 2020 that these events could not take place. The loss adjusters' letter dated 7 December 2020 stated that insured reported that cancellation was due to regulations and restrictions in California which included stay-at-home order and the venue advising it could not hold the event because state and local authorities had not provided it with guaranteed and approved protocols necessary for a meeting of the size and scope intended. The latter could only be referring to the failure to secure approval in advance for the events in March 2021 which would currently have been prohibited by CA1, where the attendance was going to be around 16,000 people. However, that failure cannot be viewed in isolation as the only cause of the cancellation, because what it meant was that for this event CA1 was still in force. That was an effective cause of the cancellation in my view. For the reasons given above, it cannot be correct that the cause was CA4.

It is common ground that this was cancelled on 2 November 2020. The venue, Anaheim Convention Center, was already closed due to restrictions, as at the date of cancellation and as at the date of the event. Reasons given for being unable to proceed with the event included current restrictions on congregating and large gatherings imposed by states which made it impossible to execute an event. The underwriters report also referred to CA3. Given that the event attracted activewear retailers from across the world, the anticipated number attending must have been more than 250. On that basis, CA1 was an effective cause.

The position here is identical to that under Claim 9 and the result is the same.

This was postponed on 6 November 2020 and cancelled on 26 February 2021. WRBC relies on the cancellation date here. The claim form refers, among other things, to stay-at-home orders and restrictions on congregation congregating on large gatherings imposed by states. There would have been 15,000 people in attendance and CA1 clearly applied and was an effective cause.

The letter from the Society dated 18 December 2020 refers to significant restrictions on non-essential commerce and activity with gatherings such as this event prohibited into the foreseeable future, at state and local level. This would seem to have been an event with more than 250 delegates and CA1 would clearly have prevented it, so it is an effective cause. The fact that the underwriters report also refers to the Diego Convention Center remaining closed “due to the California mandates” does not affect the position so far as CA1 is concerned. Indeed, CA1 would have forced its initial closure anyway.

This was cancelled on 17 May 2021. The claim form refers to restrictions on congregating in large gatherings imposed by states. It also refers to progressive lifting of restrictions, but these would not at the time have covered this event. The fact that there were measures relaxing some early restrictions is not relevant where the event concerned was still prohibited because of the original restriction. The number of attendees here was clearly more than 250 and therefore CA1 was an effective cause.

The two documents relied upon by WRBC in its Response in fact relate to the cancellation by the AMA announced on 3 April 2020 that its Annual Meeting would be suspended. The announcement made reference to government mandates prohibited large-scale gatherings in Chicago, not California. The Reinsurers’ comment in fact refers correctly to the fact that this claim is about cancellation of a different event, namely the AMA’s Health care Executive Summit. They are correct to point out that no cancellation date has been given in the relevant documents (the reference to J15/8 is wrong because that refers to the Annual Meeting). However the relevant documents demonstrate that there clearly was a cancellation and it cannot have been earlier than 1 June 2020. J8/3/4 suggests that the claim was approved by the insurer on 16 April 2020, and if so, cancellation was obviously before then. Although the documents relied upon by WRBC do not refer to particular reasons for the cancellation, the expected attendance was 30,000 and on that basis, CA1 must have been an effective cause.

This was cancelled no later than 24 May 2021 when the insured first notified the claim. There were expected to be 5,100 attendees and over 1,100 exhibitors. It was expected that on 15 June 2021 indoor conferences and meetings could resume at full capacity with no limits provided that there would be no more than 5,000 people. In the latter case the event was not itself prohibited but the organisers had to verify that the attendees were either fully vaccinated or had tested negative within 72 hours of the event start time. In this particular case, I cannot see that CA1 was still effective because the regime was now changing and the question would be compliance with what was in fact the Public Health Order dated 11 June 2021 but effective as from 15 June 2021. The problem here is that the insured’s reasons did not focus on original restrictions, or indeed any restrictions on conferences, but rather the fact that attendees were themselves healthcare professionals who had an obligation to treat those with Covid, so there was a conflict. There was also still a risk of contracting Covid for those who had to travel and there were travel restrictions in some other countries. The insured cancelled the event but rebooked it swiftly at the same venue for 2027 (the date here may be an error) and thereby avoided cancellation fees.

I do not see that the cancellation of this particular event is covered at all, and accordingly I disallow this claim.

The cancellation date was no later than 11 May 2020 when the loss was first notified. The loss adjusters’ report referred to “gathering bans” and CA1, CA2, and CA3. The likelihood is that more than 250 people would have attended and accordingly CA1 was an effective cause.

This was cancelled on 6 May 2020. It was expected that there would be at least 18,500 attendees. The insurance claim questionnaire refers to state orders on prohibiting large gatherings, and the letter from the San Diego Convention Center dated 11 May 2020 referred to the current situation there and in California regarding when large gatherings can resume is presently still very unclear. On that basis, and given the number of attendees, CA1 was an effective cause.

The cancellation date was 15 June 2020. On the face of it, attendees would be more than 250. Insured’s letter of that date referred to the fact that the venue, the San Diego Convention Center, could not guarantee that it would be available in time for the event and additionally “California’s restrictions and prohibitions on gatherings make it impossible, if not illegal, for AAP to hold the Insured Event”. On that footing, CA1 was plainly an effective cause. Insured’s reasons for cancellation included state orders on prohibiting large gatherings. This means that CA1 was an effective cause.

This was cancelled on 17 March 2020 by an SPIE announcement. There would be around 4,500 attendees. The loss adjusters’ report refers to the original State of Emergency proclamation on 4 March 2020 and CA3, in terms, but also refers to California restrictions on gatherings and other operations as being some of the toughest in the US. That would encompass CA1 which I consider was an effective cause.

This was cancelled no later than 15 June 2020 as can be seen from the email exchange at J15/13. The loss adjusters’ report refers to the original State of Emergency proclamation on 4 March 2020 and CA3, in terms. The venue, the San Diego Conference Center announced that it would be closed through August and this seems to have been the immediate catalyst for the cancellation. However, in two of its communications it said that the exact timeline for reopening depending on state and local revisions to restrictions on public gatherings. That must have contemplated CA1 and CA2. Since these are why the Conference Centre had to close in August, I consider that there is a sufficient connection between CA1 and CA2 and the cancellation. Given the number of attendees (expressed to be “thousands” but in any event would clearly be more than 250), the causative measure was CA1 rather than CA2.

This was cancelled on 16 March 2020. WRBC say that the attendees must have been at least 300, going on what number attended in 2019, and I am prepared to accept that the number must be more than 250. The loss adjusters’ report refers to present mandates in California including social distancing and gathering limitations. CA1 was therefore an effective cause.

According to a letter from the venue, cancellation was no later than 10 June 2020. The loss adjusters’ report refers to reports from the insured that cancellation was due to the current state of the pandemic and associated government regulations restrictions and advisories throughout California. Such mandates included, but were not limited to, social distancing requirements, “stay-at-home orders” and non-essential travel conditions. That is sufficient to encompass CA1 and CA2. WRBC says that the number of attendees was unknown but this was to take place over a week at the San Francisco Marriot Marquis Hotel, and I would infer that there would be at least 250 attendees. Accordingly, CA1 was an effective cause.

The insured notified the cancellation on 19 February 2021, so cancellation was no later than then. The loss adjusters’ report refers to present mandates in California including social distancing and gathering limitations. The expected attendance was just over 14,000. On that basis, CA1 was an effective cause.

This event was cancelled on or around 23 October 2020 when the venue said it could not be made available. That letter gave as the reason, health risks associated with the pandemic. The loss adjusters’ report referred to the fact of local and state authorities issuing orders restricting the movements of people along with restrictions on mass gatherings. Further, it said that the event was scheduled to be held on a US Military Base with its own restrictions closely related to federal guidelines. As a result, the Reinsurers contend that the actual causes here were distinct regulations binding on the military. Because this was a specific point which only arose in the Reinsurers’ Skeleton Argument for trial, WRBC reserved the right to take time to investigate the question of military regulations, but in the event its response was to refer to what the loss adjusters had first said about restrictions on mass gatherings, and to say that this must have at least been a concurrent cause, which is enough. I agree, it may be that the military restrictions were more severe than the general state restrictions but in the overall context, that does not mean that CA1 could not be an effective cause, and I consider that it was.

Cancellation took place on 17 May 2021. There would have been around 1,600 attendees. The reasons for cancellation set out in the claim form included ongoing restrictions on congregating in large gatherings imposed by states. I consider that CA1 was therefore an effective cause.

There are four measures relied upon by WRBC. Because not all of them have an equivalent reference from the Reinsurers I will here use WRBC’s terminology. In relation to other measures referred to by the Reinsurers, I will use the latter’s terminology.

So far as the first four measures mentioned are concerned, they are:

WRBC’s primary case is that the collective impact of the combination of measures E1 to E4, was (among other things) to prevent or restrict large gatherings from taking place in England, and they constituted a single aggregating event, alternatively the relevant combination was E1, E3 and E4; in the further alternative, the aggregating event was E1, alternatively E3, alternatively E4, or such other aggravating event as the Court might determine.

I reject WRBC’s “combination” argument here, as to which there is already subsisting authority, by reason of the finding of Butcher J in Stonegate at paragraph 186 of his judgment (quoted at paragraph 206 above) that the combination of measures E1 to E4 constituted a single aggregating event; they were each to be treated as single “occurrences” albeit taking place in quick succession. There is no reason not to follow his judgment in this respect here. For the same reasons, I do not accept that there were the alternative combinations put forward by WRBC. Accordingly, and as with California, it is necessary to consider each loss claimed by reference to a particular measure.

As with California, I will deal with each of the 20 losses claimed here in the order in which they are dealt with in the Response.

The claim letter states specifically that all of these events were cancelled on 16 March 2020, and the cancellation date is accepted by the Reinsurers. This letter refers expressly to E1, and indeed, given the date, it could refer to no other. This claim is therefore made out.

An email dated 17 March 2020 said that the Point-to-Point authority had shut down such events as a result of which this event had to be cancelled. The loss adjusters’ report said that the event was cancelled on 1 April due to government advice concerning the pandemic and the associated withdrawal of medical provisions. Particular reference was made to E1. Given the content of the email I consider that E1 was an effective cause.

It is agreed that this event was cancelled on 20 March 2020. Again, the loss adjusters’ report said that the event was cancelled due to government advice to avoid mass gatherings (i.e. E1), and the associated withdrawal of medical provisions. Given the cancellation date, the cancellation was plainly caused by E1.

It is agreed that this event was cancelled on 24 March 2020, and the loss adjusters’ report referred specifically to government advice to avoid mass gatherings. The loss adjusters’ report made specific reference to E1 and the cancellation of British Point to Pointing for the 2019-2020 season. Accordingly, E1 was the, alternatively an, effective cause of the cancellation.

It is common ground that Wimbledon 2020 was cancelled on 1 April 2020. The press release refers to concerns about mass gatherings and the government’s measures. This could include E1 and E4, but plainly E1 would in any event have been an effective cause.

The loss adjusters’ report, dealing with claims in respect of 23 Emerald events around the world, makes reference to “gathering limitations” in England, among other places. Again, while both E1 and E4 qualify as causing this cancellation E1 is plainly an effective cause and is the earliest in time.

Emails dated 1 April 2020n refer to this claim, so this event was cancelled by that date at the latest. The loss adjusters’ report makes specific reference to E3 and E4, with E3 itself meaning that the 115,000 spectators expected over the course of the weekend could not attend. However, there is a further reference to the nationwide withdrawal of event support by emergency services, all of which came into effect with E1. Such services were said to be integral to the staging of this event. On that basis, I consider that E1 was also an effective cause and being the earliest it should be regarded as the relevant cause for present purposes.

Emails show that this was cancelled by no later than 25 June 2020. Indeed, it appears that the actual decision to cancel had been taken internally on 29 May and there may have been an announcement of the cancellation before 25 June. However, this makes no difference. According to loss adjusters, the current government social distancing restrictions would have led to running the event behind closed doors at massive financial loss. Also, the provision of emergency services was an issue, with ambulance services not being able to guarantee they could provide the required services for the event safely to take place. Given that such an event would obviously constitute a mass gathering where emergency services would need to be present, I consider that E1 was an effective cause.

A general point is made by the Reinsurers that it is for WRBC to prove that all or any of those events were cancelled by reason of relevant measures in England, as opposed to a general worldwide decision. For the reasons given in respect of each of the claims referred to below at paragraphs 288 - 299 below, I consider that WRBC has proved that cancellations were caused by one of the measures relied upon.

According to the content of the archived announcement, cancellation of this particular event took place no later than 4 April 2020. However Gartner’s press release dated 25 March 2020 indicated the cancellation or postponement of all conferences scheduled for April through August 2020 “in consideration of guidelines and travel restrictions implemented by governments around the world due to Covid.” As with California (see paragraph 248 above), I consider that this is sufficient to encompass the measures in England. Those measures would include E1. There were, of course, later measures but, consistent with the approach taken by Butcher J at paragraph 192 of his judgment in Stonegate (see paragraphs 208 - 210 above), the correct analysis is to say that E1 was an effective cause.

There is a press release referring to the decision by Gartner to cancel or postpone all conferences scheduled for April through August 2020, and then a specific press release cancelling this event. In my judgment, the position here is the same as for Claim 6.

There was a specific announcement about this which said that this event was cancelled due to ongoing concerns about Covid. However, I consider this was also covered by the general announcement made on 25 March 2020, and so the position here is as with Claim 6 above.

I consider that the position here is the same as with Claims 6 and 8.

Emails referring to the cancellation of this event refer to the decision to cancel all events through August 2020 which must have been a reference back to the US SEC announcement of 25 March 2020. Accordingly, the position is the same as Claims 6, 8 and 9.

The position here is the same as with Claim 8, and so again, an effective cause was E1.

There was a press release announcing this cancellation, making reference to continued uncertainty relating to travel restrictions, government directives and health organisation reports. The date of that document is not clear but its SEC filing for the quarter ended 30 September 2020 noted that on 2 July 2020, Gartner announced the cancellation of all in-person destination conferences for the remainder of the year. I have not seen that press release but there is no reason to think that it was not in similar terms to the 25 March announcement (see paragraph 288 above), in which case the position is as with Claim 6. I therefore consider that E1 was an effective cause, consistent with my treatment of Claims 6 to 11 above.

The press release said that this event was postponed to 17-18 November 2020. Its date is not clear but in any event, the position would be the same as for Claim 13. Accordingly, E1 was also an effective cause here.

Consistent with its SEC report dated 2 July 2020, the cancellation announcement of this event referred to the immediate re-evaluation of Gartner’s conference portfolio for the remainder of 2020. The announcement relied upon in fact goes on to say that the IT Infrastructure Conference scheduled to take place in Sydney 27-28 April 2020 would be cancelled. That particular event does not fall within the scheme of claims here, but it does not alter the references to the cancellation of the event which is the subject of Claim 15. On that basis, the position, again, is the same as for Claim 13, and E1 was an effective cause of this cancellation.

I have taken the date of the event from the Schedule under Claim 16. The reference to J14/49/1 cannot be correct because this is about a virtual conference on 7-10 December 2020. The reference by WRBC to the 25 March 2020 SEC statement cannot be correct, either, since this event was going to take place after August 2020. However the Schedule correctly identifies the relevant overall decision as being that contained in the 2 July 2020 SEC statement which referred to the cancellation of events for the rest of the year. On that basis, the position is the same as for Claim 13, and E1 was an effective cause of the cancellation of this particular event.

The document referred to by both parties here (J/14/51) is a press release which announced the rescheduling of this event from an earlier date to 3-4 December 2020. However, the event could plainly not have taken place then and indeed it did not. On that basis, as it seems to me, and with, again, 2 July 2020 being given as the date of cancellation, the position is the same as for Claim 13. E1 is, therefore, an effective cause.

There is a press release which announces this cancellation which may be from August 2020 but in any event, the cancellation followed from the announcement in Gartner’s 2 July 2020 SEC statement. Accordingly, the position is the same as in Claim 13.

For all the reasons given above and in the Appendix,

I am most grateful to Counsel for their very helpful written and oral submissions and for their assistance generally in the course of the trial.