CP Holdings Limited & Ors v Assicurazioni Generali SpA

The claimant policyholders seek indemnity from the defendant insurers for what the claimants say were substantial business interruption losses sustained during the Covid-19 pandemic. The claimants estimate their collective loss of revenue from the pandemic to be about €400m and claim to be entitled to recover about €160m under the insurance policies considered in this judgment. The exact figures, whatever they may be, do not matter for present purposes, where I am determining certain important preliminary issues only, as I explain below.

There are now 15 claimants, as follows, the 7^th and 13^th claimants having ceased to be parties under an order dated 25 September 2025:

When the Covid-19 pandemic hit in early 2020, the claimants between them operated a large number of business premises, including hotels, spas, restaurants and offices, located variously in the UK, Czechia, Romania, Slovakia and Hungary. The defendants are companies in the large and well-known Generali insurance and asset management group.

There are four insurance policies to consider:

A trial of preliminary issues was ordered at a first case management conference on 25 July 2024. Many of the individual questions included in the preliminary issues are or turn on questions of the proper interpretation of policy terms. There is also, however, a prior question of whether the 2019 Global Policy falls to be rectified, in relation to which the claimants allege, stating it in very general terms at this stage, that:

I shall use ‘conformity term’ as a generic label for a term of an insurance contract that provides, to such extent and manner as may have been agreed in the particular case, that the terms of a preceding year’s insurance are to prevail over those otherwise agreed for a current year. There was no insurance market expert evidence, so this is not an occasion for a general review of such terms. However, the factual evidence indicated that conformity terms, as I have thus defined them, are a known phenomenon, occasionally agreed, and that one situation in which parties may consider using a conformity term is where, as in this case, the insurer on a long-standing account seeks to introduce changes, for example a new policy wording, at short notice in a renewal process. Completing the renewal on time upon the insurer’s new terms, but with an additional term that the insured is, in effect, held covered on (not worse than) the old terms, to such extent, in such manner, and for such period, as may be agreed in the particular instance, may be mutually preferable to not renewing, or extending the prior year’s insurance giving rise to a deferred renewal.

Where a conformity term is agreed as part of, or alongside, a renewal, then an appropriately worded clause ought to appear in, or as an endorsement to, any insurance policy issued to document the terms of the insurance for the new policy year. As with any other insurance term, it is not a pre-requisite in law for the existence of a conformity term that will bind the parties that a wording for it has been provided and agreed. I am using the label ‘conformity term’ in part because ‘conformity clause’ might be taken to refer to a specific wording for a formal policy document, or to pre-suppose a need for such a wording. Of course, the question may arise in any given case, and does arise here, whether in a discussion of whether there is to be or might be a conformity term the parties intend there to be one without a specific wording for it having been settled between them.

The preliminary issues ordered are set out in the Appendix to this judgment. It was agreed at trial that they did not create an examination paper every question in which I had to answer, or set a fixed order in which I had to tackle the issues. For example:

Finally, by way of introductory comments, I have articulated the rectification claim as I have in paragraph 5 above because, although this was more obscured by than apparent from the pleadings, it was common ground before me that:

The 2019 Global Policy as issued by the first defendant in September 2019 contained no conformity term at all. The defendants placed some reliance on how the claimants and Aon UK reacted, or it may be did not react, to the policy as issued, but that was only for its value, if any, as evidence possibly casting light on the likely tenor or substance of what may have been discussed between Messrs Litchfield and Humphries in that last working week of June when the renewal was placed. It was not contended that what happened after the policy was issued gave rise to any variation of the contract concluded then, or any estoppel that might affect the rectification claim. Similarly, the claimants placed some reliance on how a claim relating to a fire in Czechia was dealt with under the 2019 Global Policy, but did not contend that that varied the contract as originally concluded in late June 2019 or created an estoppel.

The claimants called Mr Litchfield of Aon UK, already introduced, and Paul Filer, the first claimant’s long-standing Finance Director, now retired, who dealt personally with the 2019 renewal for the claimants. They also relied on the unchallenged witness statement evidence (so that the witnesses did not need to be called) of Dr Helga Sztanó, now and at the material time Head of Legal at the fourteenth claimant (Danubius Hotels ZRT) in Hungary, and Andrei Rusu, now General Manager, at the material time Operations Manager, at the sixteenth claimant (Balneoclimaterica Srl) in Romania.

I accept the evidence of Dr Sztanó and Mr Rusu, unchallenged as it was. My assessment of Messrs Litchfield and Filer was that they were both honest witnesses giving the court, as best they could, their genuine recollection of events and their contemporaneous thinking. More so for Mr Litchfield than for Mr Filer, the testing of the reality, clarity and plausibility of his testimony in cross-examination showed that some of what he said in his witness statement, so that it became his evidence in chief, amounted to over-statement or error as regards what truly he could say he remembered. But that did not give me doubt as to the honesty with which he had approached the initial exercise of giving his account in writing or engaged in the exercise of answering questions on that account from the witness box.

The defendants’ only factual witness was Stewart Stokes, Head of Property (Underwriting) at Generali (UK Branch). He succeeded Richard Brown in that post at the beginning of 2021. At the material time, Mr Stokes was UK Portfolio Manager within the Property Underwriting division of Generali. I shall hardly mention Mr Stokes again in this judgment. Save to confirm, which he did, that Mr Humphries and Matthew Cullen, the other Generali Birmingham office property underwriter, were in Mr Stokes’ experience of them good, experienced property underwriters who peer reviewed each other’s underwriting decisions, Mr Stokes had no relevant evidence to give, having had no involvement in or contemporaneous knowledge of the material events or underwriting decisions. Messrs Humphries and Cullen reported directly to Mr Brown.

Messrs Humphries, Cullen and Brown had all left Generali before this litigation commenced. If there was to be any factual witness evidence from what had been the defendants’ side (at the time) of the 2019 renewal, it would have needed to be evidence from one or more of them, or (perhaps) from Mr Kemp or Ms Austin, mentioned above. I was left with the bare fact that neither side called any of them.

To assist the court in relation to issues arising under the Hungarian and Romanian Policies, there was expert evidence on Hungarian and Romanian law.

On Hungarian law:

On Romanian law:

All of the expert witnesses on foreign law, in their written reports, went beyond informing the court as to, and explaining, the relevant substantive rules of the respective systems of law. They were led by their instructions to consider, and express opinions on, the application of those rules to the facts, which is a matter for the court and not for them as expert witnesses on foreign law. On the relevant rules of Hungarian and Romanian law, there was much common ground in the expert reports. Counsel provided, before the experts’ oral evidence, agreed documents setting out the principles of Hungarian and Romanian law on which the parties were agreed, in the light of the expert reports, and the questions concerning those principles on which the parties were not agreed and on which, therefore, the experts would be cross-examined. Those summaries were very helpful.

There was also expert evidence at trial on virology and epidemiology from Professor Sir John Bell (until recently Regius Professor of Medicine at the University of Oxford), Professor Keith Neal (Emeritus Professor in the epidemiology of infectious diseases at the University of Nottingham) and Professor Julian Hiscox (Chair in Infection and Global Health at the University of Liverpool). Professor Sir John and Professor Neal were called by the claimants; Professor Hiscox was called by the defendants. In the event, it will not be necessary to say very much about their evidence, learned and fascinating though it was.

The claimants pleaded the following as the foundation for their rectification claim (at paragraphs 15 to 18 of their Re-Amended Particulars of Claim):

That is an odd case, and distinctly unpersuasive, for a number of reasons.

Firstly, the following is an accurate statement of part of the law of rectification as it applies to insurance policies issued unilaterally by an insurer purporting to set out the terms of the insurance contract:

“If a binding agreement to insure was concluded prior to the issue of a policy, the purpose of which was merely to record the terms of that agreement, the insured has the option of suing upon the earlier contract in lieu of seeking rectification of the policy” (MacGillivray on Insurance Law, 16^th Ed. (2025), at 12-013).

Since it was common ground, albeit that perhaps emerged clearly only at trial, that a binding contract of insurance was concluded by the exchange of emails referred to in paragraph 9 above, and plainly on their own terms in any event, the emails relied on by the claimants (paragraph 20 above) were part of a negotiation of terms for renewal. Nothing in the content of those negotiation emails was apt to convey an intention, on either side, to be bound by anything other than the terms of renewal finally agreed, whatever they might be, if renewal did occur at or following the completion of that negotiation. The real question to which that exchange of emails gives rise, therefore, is whether the renewal terms finally agreed, on 28 June 2019, included a conformity term, and, if so, what was its effect.

That gave rise to some discussion at trial of the oft-stated principle that very clear evidence of the true agreement is required before rectification can or should be granted (see, e.g., Thomas Bates v Wyndham’s (Lingerie) Ltd [1981] 1 WLR 505 at 521F-G, per Brightman LJ; The Olympic Pride [1980] 2 Lloyd’s Rep 67, per Mustill J (as he was then), proposition 5 at 73 lhc; FSHC Group Holdings Ltd v GLAS Trust Corp Ltd [2019] EWCA Civ 1361, [2020] Ch 365, at [174], per Leggatt LJ (as he was then), giving the judgment of the court).

I respectfully wonder, as a matter of first principle, why, in an insurance case, if (a) the contract of insurance is concluded prior to and otherwise than by the policy being issued, and (b) the conclusion of a normal exercise of ascertaining the terms of that contract from the contemporaneous materials is that the contract included a particular term, anything more should be required to justify a (strictly unnecessary) claim to rectify the policy to include that term so that it conforms to the true bargain thus ascertained. In FSHC Group, supra, for example, having referred to Lord Hoffmann’s 2015 Combar Lecture, “Rectification and other Mistakes”, Leggatt LJ said this at [141]:

“… it can be seen that the judges who at one time espoused the view that it was necessary to find a prior concluded contract before an order for rectification could be made were treating the only permissible form of rectification for common mistake as the first form of rectification described by Lord Hoffmann, based on the principle that the court should give effect to what the parties have contractually agreed to record in their document. … if parties make a binding agreement to execute a document containing particular terms but instead execute a document containing different terms, the court may specifically enforce the agreement by rectifying the document; … in such a case, the terms of the contract to which the subsequent document is made to conform must be objectively determined in the same way as any other contract.”

(See also per Hobhouse LJ, as he was then, cited by Leggatt LJ, in Britoil plc v Hunt Overseas Oil Inc [1994] CLC 561, who (at 572) said of the case of prior concluded contract with no variation, “the court will have to construe the earlier agreement as a matter of contract and as a matter of law. Having decided as a matter of law what its effect is, the court will give effect to the legal rights of the parties. This is a different situation from that which exists where there was no prior contract.”)

Mr Masefield KC submitted that the authorities bound me against the approach posited by the first sentence of paragraph 25 above. It can be said that, for example, Hobhouse LJ in Britoil immediately continued by saying that it was “fundamental that the intention of the parties as stated in [a final written contract which they have made] is only to be displaced upon the most cogent evidence”. On the other hand, of course, an insurance policy issued unilaterally by the insurer is not, or may not be, such a final written contract as Hobhouse LJ described. If it mattered, Mr Masefield’s submission would have required detailed consideration. However, it does not matter. On any view, if on an ordinary analysis of the contract formation exchanges, a term contended for in subsequent litigation was proposed but not agreed, or never really proposed at all, there can be no claim to rectify the policy to include it. In such a case, the claim that could properly be brought without reference to the policy, suing upon the contract as in fact concluded, would fail if it depended on proof of the contentious term. It would be bizarre then to allow a claim to rectify the policy with a view to allowing that bad claim to be made upon it, as rectified.

Secondly, it is peculiar, and a distraction, to find it pleaded, seemingly as the claimants’ primary case, that there was an ‘outward expression of accord’, continuing until the policy was issued in September 2019, which is the language of rectification by reference to a non-binding prior agreement relating to a contract entered into when the policy was issued. On any view, this case is one of a binding agreement, indeed strictly the binding agreement, viz. the contract concluded between the parties on 28 June 2019. If that contract included a conformity term, then rectification is both strictly unnecessary and straightforwardly available (subject, perhaps, to the point identified in the preceding paragraph), since the 2019 Global Policy failed to include that ex hypothesi agreed term and (again) no variation or estoppel is alleged.

In relation to both of those difficulties with the rectification claim, the claimants doubled down on their erroneous analysis in reply. The defendants pleaded (at paragraphs 12-13 of their Re-Amended Defence) that the claimants were relying on only a partial summary of emails exchanged between the parties prior to renewal, and referred to additional exchanges, including emails sent on 27 June 2019. In reply (at paragraphs 12-13 of the Re-Amended Reply) the claimants admitted the further emails cited by the defendants, but insisted that the exchange on 27 June 2019 was irrelevant “in circumstances where, on the Claimants’ case, a concluded agreement and/or outward expression of accord had already been reached on 26^th June 2019 and the email [sic.] dated 27^th June 2019 did not purport to affect that agreement or outward expression of accord.”

Thirdly, on any view the tenor of the exchange of emails relied on by the claimants is not to the effect of the ‘Conformity Agreement’ they allege was concluded on 26 June 2019. As alleged, that is an agreement as to the meaning and effect of any policy that might be issued for the 2019-2020 year, whereas the whole point of the exchange was a recognition, slightly in the abstract (Mr Litchfield having not yet gone through the proposed new policy wording), that the new policy for which the first defendant had issued its renewal quote might well not have that meaning and effect. The subject matter of the exchange was whether there would or might be, in addition, a conformity term.

That last concern about the pleaded case rests upon a somewhat fine distinction, however, and it would not be right to allow it to determine the result of the rectification claim. The other two difficulties, however, are in my view basic and fatal. The pleaded rectification claim is, I consider, obviously unsustainable. That is sufficient basis to dismiss that claim, and if it were dismissed on that basis I would question the utility of working through the list of specific questions set out as part of preliminary issue 2 (see the Appendix).

Taking that approach, however, might not do justice to the case developed, and responded to, at trial. I have therefore considered whether, upon the exchanges between the parties culminating in the renewal confirmation on 28 June 2019, the correct view is that the contract then concluded did contain a conformity term and, if so, what such term. I have done so de bene esse as to whether a claim on the basis of a decision as to that, if favourable to the claimants, should be treated as open to them given their pleadings.

The general background to the 2019 renewal negotiation was that the claimants had a long-standing relationship with Aon UK. Throughout his time as the first claimant’s Finance Director (2000 to 2022), Mr Filer was responsible for arranging its insurance programme through Aon UK. From about 2005 or 2006, the principal broker who dealt with the claimants’ account at Aon UK was Mr Litchfield. When he took on the account, the claimants’ overseas companies were not included within a single insurance policy, and the idea of a global programme was picked up between him and Mr Filer, following some initial discussions between Mr Filer and Mr Litchfield’s predecessor on the account at Aon UK, David Hutton. A first global policy in line with that idea was placed for 2007.

The scope of the global cover expanded over time until all of the jurisdictions and properties within the claimants’ group were covered. At all stages, this global insurance programme involved at least some local policies issued by Generali group companies, each covering a particular jurisdiction and the properties operated there, and also the annual predecessors to the 2018 and 2019 Global Policies. Since the broad intention was for the annual global policy to provide the cover ultimately desired by the claimants (assuming Generali was happy to underwrite it), whether or not available in any given local jurisdiction on the terms of cover the local Generali insurers would underwrite:

A policy wording for the global policy, which was bespoke to the claimants’ account, not a standard Generali wording or a collection of standard wordings, was developed and used consistently from 2007 until, and including, the renewal for the 2018 Global Policy. It included property damage (‘PD’) cover and business interruption (‘BI’) cover; and the latter provided for certain cases of ‘pure’ BI cover, i.e. cover for business interruption loss not resulting from damage to any property insured by the PD cover, as well as for BI losses consequent upon accidental loss of or damage to an insured’s property.

The Policy Loss Limit (a defined term in the global policy wording) was calculated from the costliest combined PD and BI loss, based on a spreadsheet provided annually that set out building reinstatement values and annual rent and revenue exposures. That costliest combined PD and BI loss was an imagined total loss of the Regents Park Hotel in London, operated by the fourth claimant.

Although it was not the start of the renewal process, the start of the sequence of events that now gives rise to the rectification claim was a telephone call on Monday 24 June 2019. The renewal date was 30 June 2019, i.e. Saturday 29 June 2019 was the last date of the insurance period under the 2018 Global Policy (and its local policy counterparts), so that the renewal discussions had as their target a commencement date of Sunday 30 June 2019.

On that Monday afternoon, Mr Humphries called Mr Litchfield. The gist of his call was that he had some bad news, namely that he had to move the account onto a new policy wording with effect from the imminent 2019 renewal. There was some discussion of the possibility of including a conformity term so that the insured would not be disadvantaged by moving to the new policy wording.

Mr Litchfield gave evidence that he thought at the time that he had a binding commitment from Mr Humphries in that telephone call. However:

Mr Litchfield also said in his evidence that what he understood had been discussed was a conformity term such that moving to Generali’s new policy wording, as Mr Humphries was saying had to happen straight away, “wouldn’t matter for this year [i.e. the 2019-2020 policy year], because CPH would not be disadvantaged by the terms of the 2019 policy as compared to the 2018 policy.” That obviously concerns the nature or scope of what it was discussed might be agreed; it does not improve the quality of the evidence as to whether what was said on the telephone on 24 June 2019 might have conveyed a firm and binding commitment to that or any other effect.

So far as is material, next came the exchange of emails at 15:27 hrs / 15:32 hrs on the Wednesday, 26 June 2019 (paragraph 20(1) above).

By his email at 15:27 hrs, Mr Humphries wrote as follows:

“Hello Bob,

Further to our discussions regarding renewal I now have pleasure in enclosing our CCQ [Contract Certainty Quote] for renewal, along with a copy of the new Generali wording that this is based on.

Please note that the quote system isn’t able to accommodate mixed currencies in the Premium Section so the correct total is €1,057,807. The rate of exchange utilised for the conversion from GBP to EUR is 1.13827, as expiring.

As we’ve discussed the differential pricing for the individual territories has had the effect of eroding the mean rate for the programme (ex. [viz. excluding] Buxton) and I have therefore adjusted rates slightly to reflect this. Our proposed allocation is shown below but I am happy to modify this in line with your discussions with the Insured, subject to the overall premium remaining at the level quoted.

[ A table followed breaking the c.€1m total premium down into various categories (mostly, but not exclusively, geographical). ]

With regard to LCR [Low Claims Rebate] and LTA [Long Term Agreement] I can advise as follows:

LCR

1.

LCR to be 10% of the programme premiums excluding Buxton, Buxton Re and Risk Engineering subject to Loss Ratio (as currently calculated) being below 25%;

2.

Property and Casualty LCRs to operate independently, rather than cross-class;

3.

LCR to be subject to renewal of the programme at the end of the expiring period and calculated in accordance with current timescales.

LTA

Unfortunately at this stage we are unable to confirm if we can offer an LTA but are continuing to work on this aspect.

Thank you for bearing with us Bob whilst we finalised our offer, it is appreciated.”

I should explain the reference to a long term agreement. On the claimants’ behalf, Mr Litchfield had been looking to secure a commitment to premium levels for multiple policy years (whether or not, strictly, that would have tied Generali’s hands as a matter of contract for future renewals). Mr Humphries’ inability to help, and the fact that strictly there was only ever a contract in place for one year at a time, does not make a nonsense of the idea of agreeing a conformity term for the (entire) 2019-2020 policy year, as some of Mr Masefield KC’s submissions sought to suggest. Such a term might mean, depending on the detail, that for that year, the only year for which any contract was being concluded, the 2018 Global Policy might as well have been just re-issued for a further year, in return for whatever 2019 premium was finally agreed (the low claims rebate arrangement being one aspect of that). But the business concept here had been and plainly remained that this was a long-standing and annually renewed insurance programme that the parties expected to continue for years to come.

Mr Humphries’ email attached a 13-page quote with “Quotation ID 19DM04134000”, and a 109-page Generali policy wording. As his email said, that was a CCQ, i.e. a Contract Certainty Quote, for renewal, a reference to the UK insurance industry’s Contract Certainty Code of Practice issued in October 2012. The evident purpose of the CCQ, therefore, and the explanation for its form and content, was to satisfy Principle A of the Code, requiring that “The insurer and broker (where applicable) must ensure that all terms are clear and unambiguous by the time the offer is made to enter into the contract or the offer is accepted. All terms must be clearly expressed, including any conditions or subjectivities.” The first of the specified ‘Insurer actions’ to ensure compliance with that Principle is that “The insurer should check that the proposed contract [i.e. the document which contains the offer and which can take many forms] clearly identifies all of the terms by the time it formally commits to the contract.”

By his email at 15:32 hrs, Mr Litchfield replied in these terms:

“Thanks for this Derek, I will study it and revert to you with any questions.

In terms of the Risk Engineering fee, to which country would you suggest I allocate this?

I assume that Liability rates are on their way?”

The second question, as I understand it, concerned occupiers’ liability insurance cover also provided by Generali with which this litigation is not concerned.

Mr Humphries responded almost instantly, at 15:34 hrs, on the Risk Engineering fee, saying it “would normally be collected in the UK and would be subject to VAT rather than IPT – presumably CP are registered and would be able to reclaim?”

Mr Litchfield’s email about ten minutes later (paragraph 20(2) above) replied on that, as well as raising the conformity question. Thus:

“Yes they can re-claim VAT but I guess what I am really asking is to which territories does it relate? CP would want to allocate that internally.

One of our requirements is that we have to show expiry premiums as well as the invited premium and I’m not sure I have that? Are you able to tell me please what the expiry premium would be for the first 5 entires [sic., entries]?

We spoke the other day about including a clause to cover anything that would have been covered under the existing wording but isn’t catered for in the new wording. Is that agreed?”

Mr Litchfield’s reference to requirements upon Aon UK concerned what he was required, by the claimants, to set out in the renewal report he was expected to prepare, setting out and explaining to the claimants the insurance offer he had obtained for them upon which they could then give their renewal instructions.

That generated the response from Mr Humphries upon which the rectification claim now principally hangs (paragraph 20(3) above). Mr Humphries clarified Generali’s intentions on Risk Engineering visits for the 2019-2020 insurance period, and provided a table of expiring premiums for the core exposures by territory, i.e. Czechia, Slovakia, Romania, Hungary and the UK. Between the two he responded on conformity in terms which it is convenient to repeat here, namely:

“I can confirm that we are able to include a Conformity Clause (wording to be agreed) in the new Policy to ensure that the Insured is not disadvantaged by the move to the new wording.”

No revision to the CCQ, or revised CCQ, came with that response, or was ever issued by Mr Humphries or requested by Mr Litchfield.

The next day, Thursday 27 June 2019, at 12:28 hrs, sent as a reply to that email, Mr Litchfield emailed Mr Humphries, stating:

“I’ve just about gone through everything you’ve sent me.

The one remaining query is re the conditions precedent. There are far more in the new wording than the old one and I’m not overly comfortable with that.

So, a couple of queries:

1.

Do we have to use the new wording?

2.

If we do, will all of the conditions precedent therein apply or just the ones under the current programme?

Await your early advices as I’m looking to finalise our report.”

Mr Humphries replied on that a couple of hours later, by email at 14:38 hrs:

“As discussed I am happy to Endorse the new policy to reflect the expiring position with regards to Conditions (Hot Work and the like) with a view to moving towards the new versions over time.”

In the context of the new Generali policy wording under review, the references to “conditions precedent therein” (Mr Litchfield) and “Conditions (Hot Work and the like)” (Mr Humphries) were plainly enough, I think, references to the section of that wording entitled “Special Conditions to Section A” (vs. “special conditions applying to section a” in the 2018 Global Policy); possibly also to “Special Conditions Applicable to Section B” (which had no equivalent in the 2018 Global Policy), although that might be more debatable and the parties did not address any argument to that point.

There is no hint in the documents, and it was not Mr Litchfield’s evidence, that those special conditions of cover (e.g. as regards hot works at any of the insured properties) had been a specific topic discussed on the Monday telephone call. Mr Humphries’ email suggests that there had been a further call; and indeed Mr Litchfield’s evidence, having seen the sequence of emails, was that there was a call, after his email raising the query and before Mr Humphries’ response two hours later, which seems plausible.

As to the key exchange on 26 June 2019, Mr Litchfield’s evidence was that the sense he intended by his question to Mr Humphries was, “Is that agreed?”, i.e. (as I would paraphrase his evidence) ‘I thought we had agreed that, have I got that right?’. It is plausible that if Mr Litchfield thought something had been agreed firmly on the Monday telephone call, and he then expressed himself as he did in his Wednesday email, that might have been what he intended by it. The more natural reading of the email, though, is that he was there asking for a first indication from Mr Humphries of whether the inclusion of a conformity term in the new policy, and one of very general scope, was agreed.

Either way – and this is what matters – what Mr Litchfield got back from Mr Humphries was not agreement to include, or confirmation that there was already agreement to include, a conformity term. He received instead confirmation that Generali was (i) “able to include” (ii) “a Conformity Clause (wording to be agreed)”, to ensure that the new wording did not disadvantage the insured. That was no more than a confirmed willingness in principle to agree a conformity term, if asked to do so and if a specific wording, with the stated aim, was agreed.

The immediate context was that Mr Litchfield had only just received the new Generali wording, and had promised to review it and revert. In that context, whether a conformity term, and if so what term, would ultimately be requested would surely depend, and at any rate might well depend, on the outcome of that review. I was not impressed by the argument that the full terms of the quote, and the new policy wording offered, could not realistically be reviewed for acceptability in the time available, given the experience and sophistication of Aon UK and the claimants, and the resources available to them. There is no reason to put upon Mr Humphries’ confirmation a meaning beyond its plain and ordinary meaning of willingness in principle and subject to wording.

The exchange, and related telephone call, the next day, then makes entire sense, again just reading it naturally:

Strikingly absent from Mr Litchfield’s post-review response to the 26 June quote and the new Generali wording was any request, thus fully informed as to the terms offered for what would become the 2019 Global Policy, for a general conformity term such as Mr Humphries had indicated a willingness in principle to agree. The only push back against the simple adoption of the new wording, and the full detail of the 26 June quote, was the proposal to retain the outgoing conditions precedent instead of those in the new wording.

That exact proposal was not accepted. Mr Humphries’ response made clear that if the claimants confirmed renewal, the 2019 policy would be issued on the new wording. He counter-proposed, however, a limited conformity term by way of policy endorsement that would “reflect the expiring position with regards to Conditions (Hot Work and the like) with a view to moving towards the new versions over time”. The question arises, but it is not necessary to decide it, whether that counter-proposal was accepted and, if so, whether the new contract of insurance concluded on the Friday therefore did include a limited conformity term concerning conditions precedent or how, if it did, that term should now be articulated most accurately given that no specific wording for an endorsement was ever proposed or agreed.

It is not necessary to decide that question because the only relevant claim made is for rectification to include a general conformity term to the effect that the claimants were in all respects entitled to the better, for them, of that which they had in the 2019 Global Policy and that which they would have had under the terms of the 2018 Global Policy had they been carried over. That claim would not be made good by a finding that there was a conformity term that had only the effect, in one form or other, that compliance with the conditions precedent in the 2018 Global Policy would be sufficient for the new policy year, as regards that section of the policy. Furthermore, the claimants’ ultimate claims to be indemnified in respect of their Covid-19 business interruption losses, as alleged, are not resisted by reference to any of the conditions precedent in the 2019 Global Policy that might be impacted by such a limited conformity term.

Mr Litchfield’s evidence was that when, on Friday 28 June 2019 and following all those exchanges, he confirmed the renewal, he understood his instructions from the claimants to be “to renew on the basis of the new policy, but with the benefit of the conformity clause – so that the new policy was the starting point and CPH took the benefit of any enhancements in cover. However, if the prior year’s policy was more favourable to CPH, then that would apply. Therefore, CPH would not be disadvantaged.” If that was his understanding, it will have been based on the fact that he received instructions to renew on the basis of his final renewal report provided to Mr Filer for, and discussed at, a final renewal meeting involving just the two of them (Messrs Litchfield and Filer) at the first claimant’s office at 10.30 am that day, 28 June 2019.

The minutes of that meeting, prepared by Mr Litchfield, note various matters, including Mr Filer’s disappointment at the unavailability of a new “3 year deal” (a reference to the absence of any quote for an LTA) and at the LCR terms offered for renewal, which were less favourable for the claimants than the outgoing terms. The minutes make no mention of any conformity term, nor do they in any other way indicate directly any instruction or understanding that renewal was to be, so far as material, on anything other than the terms of Mr Humphries’ 26 June quote. (A specific instruction was noted that Aon UK was to reject the Risk Engineering fee, Mr Filer’s view having been that it was an operational cost for insurers to bear as they had done previously, but that is not material to the rectification claim.)

Those minutes do however state that Mr Litchfield provided his renewal report, which “formed the basis for the review of individual policies as noted below”, and that as regards PD and BI, “Generali’s renewal terms were discussed”. In that renewal report, the claimants were informed that:

That evidences that Mr Litchfield indeed seems to have thought at the time that he was in possession of (i) a substantive offer of insurance on the outgoing terms of the 2018 Global Policy, at the premium (with revised LCR term) quoted for the new policy year, (ii) assurance that although the policy when it was issued would be on Generali’s new wording, an endorsement would be agreed such that the claimants would have the benefit of at least everything they had under the outgoing wording, and (iii) a specific agreement that the outgoing conditions precedent, not those of the new wording, would apply for the whole policy year (the new set of conditions precedent to be introduced only at a subsequent renewal, so at the earliest from 30 June 2020, and perhaps not even then).

For the reasons I have given, that was in my judgment a long way from what, through Mr Humphries, Generali had actually offered.

The evidence showed that the claimants were, by annual premium, the biggest client of Generali’s Birmingham property underwriters, Mr Humphries and Mr Cullen; and that they did in fact agree a general conformity term, of the type for which the claimants contend, with one of their other biggest clients, Dechra Pharmaceuticals, for its 2019-2020 policy year incepting on 1 November 2019. The underwriters were supposed to engage in peer review of each other’s work, Mr Stokes confirmed that Messrs Humphries and Cullen did so, and I draw the inference that indeed each will have peer reviewed the other’s work as respective primary underwriter for the claimants (primary underwriter, Mr Humphries) and Dechra (primary underwriter, Mr Cullen).

Dechra’s 2019 policy was generally on the new Generali policy wording, like the 2019 Global Policy issued to the claimants. The technique used for Dechra was a policy endorsement, one of three at the end of the policy, as follows:

“Endorsements

…

3.

Conformity Endorsement

Definition

Previous Policy

[ Dechra’s 2018 policy was identified ]

The following Claims condition is added to this Policy.

If any claim is made by the Insured under this Policy, but the Insured would be entitled to greater recovery for any portion of that claim according to the terms and conditions of the Previous Policy, the terms and conditions of this Policy will be deemed amended to provide cover according to the terms and conditions of that Previous Policy, with regard to the relevant portion of the claim, but only to the extent necessary to facilitate that greater recovery.”

In those circumstances, I find that Generali, through Mr Humphries (under the approval, if required, of Mr Kemp or Mr Brown, and there is no reason to suppose it would not have been given as it must have been given for Dechra) would have been content, if Mr Litchfield had pressed the request, to agree a general conformity term to similar effect. Whether the same wording would have been used is conjecture; the effect of the Dechra endorsement could be achieved, I think, more simply and concisely.

The insuperable problem for the claimants, however, is that Mr Litchfield on their behalf did not press any request for a general conformity term. Indeed, to the contrary, on an objective reading of the relevant pre-contract exchanges in late June 2019, he dropped the idea and pursued instead only a specific request for conformity as regards conditions precedent. Rectification cannot be founded upon what might have been, but was not in fact, agreed; it always requires a clear accord that has not been given effect by the document to be rectified.

The claimants relied on an exchange of emails in September 2019 primarily between Mr Litchfield and a loss adjuster at Sedgwick concerning a fire at one of the insured premises, a sauna in Czechia. When the loss adjuster asked for a copy of the applicable policy, by email at 11:43 hrs on 23 September 2019, Mr Litchfield replied, at 11:49 hrs, attaching the draft policy that he had received with the 26 June quote and stating: “Attached is the latest policy wording although there is an agreement that if anything was included under the previous version but not under this one, the cover will still be operative at least for the first year.” That supports my finding that Mr Litchfield thought at the time that he had an agreement to that effect. It does not affect my conclusion that he was wrong about that.

That email also went to Mr Humphries, because the loss adjuster’s request for the up to date wording had also gone to him. Mr Humphries had himself replied to the loss adjuster (and Mr Litchfield) a couple of minutes before, at 11:47 hrs, with the actual 2019 Global Policy, saying he had just finalised it and so was attaching it for reference. Even if I were entitled to draw adverse inferences from the defendants’ failure to support their case with any evidence from Mr Humphries, which was contentious given that Mr Humphries left Generali in 2020, in my judgment it would be a stretch too far to infer that he noticed Mr Litchfield’s comment and said nothing about it because he agreed with it. It is more likely, inherently, that he paid Mr Litchfield’s email no attention, knowing he had just sent the signed policy as issued in reply to the loss adjuster’s request.

The rectification claim therefore must fail. The answers I would give to the specific preliminary issues that were ordered in relation to that claim, in the light of my analysis and conclusions above, are as follows (see the Appendix for the questions):

The particular questions raised as part of preliminary issue 1 are all questions of the proper construction of either the 2018 Global Policy (questions 1.1(1) to 1.1(7)) or the 2019 Global Policy (questions 1.2(1) to 1.2(3)), or possibly (in whole or part) both (questions 1.3 and 1.4).

Question 1.2(1) asked what is the true construction under the 2019 Global Policy of the exclusion of infectious diseases “designated as ‘Phase 5’ or ‘Phase 6’ pandemic alert”. In the normal way, the preliminary issues will have been formulated to reflect matters that were contested on the pleadings. Also in the normal way, not everything contested on the pleadings remained in dispute by the time of trial, in this instance this preliminary issues trial. At trial, there was in fact no issue between the parties about the meaning or effect of the pandemic alert infectious diseases exclusion in the 2019 Global Policy. For the avoidance of any possible doubt, the claimants opened the trial expressly on the basis that, they accepted, cover for losses caused by Covid-19 were excluded by the 2019 Global Policy if it is not rectified.

Question 1.1(5) asked what is the true construction under the 2018 Global Policy of the exclusion relating to “Severe Acute Respiratory Syndrome (SARS), Avian Flu, and/or atypical pneumonia, or fear of threat”. That did remain in issue between the parties, the focus being on the meaning of “atypical pneumonia”, and the effect of including it in that exclusion. The ultimate concern there would be whether as the defendants claim (the burden being on them since this is an exclusion clause) the claimants’ Covid-19 BI losses were excluded even if the 2018 Global Policy wording had possible relevance after a rectification of the 2019 Global Policy. That becomes academic following the rejection of the rectification claim.

I shall not lengthen this judgment by setting out the equivalent analysis for all of the other parts of preliminary issue 1. The conclusion overall is that to the extent they ask questions as to the proper construction of the 2019 Global Policy, they do not now affect the claims made in these proceedings because the claimants accept that all their claims are excluded by the infectious diseases exclusion in that Policy; and to the extent they ask questions as to the proper construction of the 2018 Global Policy, they do not now affect the claims made in these proceedings because the rejection of the rectification claim removes any possible foundation for a Covid-19 BI loss claim based upon the terms of that Policy.

That would suggest a simple answer to preliminary issue 1, viz. that it no longer arises, but I shall consider any submission, if made when this judgment is handed down, that I can and should say more about any of the questions of construction raised. Come what may, I say now rather more about question 1.1(5), because it is the reason there was expert evidence at trial on virology and epidemiology, from the eminent experts I mentioned in paragraph 19 above. I wish to pay tribute to them for the quality of their evidence and the impeccable manner in which they discharged their duty to the court as expert witnesses. It was a privilege to have the assistance through their evidence of such distinguished individuals. As should be the case, but in litigation all too often is not, the truly independent and expert nature of their work resulted in a very large measure of agreement, clearly and concisely expressed in a joint memorandum, with thoughtful, constructive and balanced answers to questions asked at trial when they gave their oral evidence. I mean therefore no disrespect whatever to those eminent expert witnesses by setting out in this judgment only minimally (in paragraphs 76 to 80 below) how I would have decided the ‘atypical pneumonia’ point if it had continued to matter.

In paragraphs 76 to 80 below, I should add, I have assumed in the defendants’ favour, without deciding, that the SARS etc. exclusion clause would be relevant. Strictly, that skips over a question of construction arising from the fact that the SARS etc. exclusion clause was the final paragraph of Clause 3 (loss of attraction) of the Supplementary Clauses Applicable to Section B of the 2018 Global Policy, whereas under the 2018 Global Policy wording claims for Covid-19 BI losses would be made against Clause 8 (contagious diseases, food poisoning and other health risks) of those Supplementary Clauses. The SARS etc. exclusion stated that “Notwithstanding the foregoing, this Policy excludes” SARS etc. BI losses. The claimants argued that ‘this Policy’ should not be read literally there but referred, in context, either only to the cover under Clause 3, or at all events not to cover under the subsequent Clause 8.

It was common ground between the experts (as regards scientific usage), and between the parties (as regards how the SARS etc. exclusion clause in the 2018 Global Policy should be read), that Covid-19 BI losses are not losses resulting from SARS or avian flu. Covid-19 is a disease or syndrome caused by the SARS-CoV-2 virus, indeed strictly it is the condition of being infected with that virus, even asymptomatically and whether or not, if symptomatic, it results in pneumonia. All of that distinguishes Covid-19 from SARS (the disease caused by the virus now called SARS-CoV-1 (previously SARS-CoV)) and from avian flu, an entirely different disease with little or no close genetic homology to a coronavirus, caused by an orthomyxovirus and associated with a different epidemiology and clinical picture. That is why, as I have indicated already, to be excluded by the SARS etc. clause, if applicable, the claimants’ Covid-19 BI losses would have to be regarded as BI losses caused by “atypical pneumonia”.

That term, it was common ground between the experts, has been used variously in clinical practice and epidemiological contexts, without unanimity as to its meaning. In other words, it has been used over the years, and still finds itself used, amongst medics or medical scientists, in different ways. Its original use, the better scientific use, and the use to which it usefully should be confined, is as a description of pneumonia caused by a group of bacterial pathogens without conventional cell walls that grow intra-cellularly. Those pathogens include Mycoplasma, Chlamydia, Coxiella and Legionella bacteria; and those bacteria are universally associated with the term ‘atypical pneumonia’.

The term has also been used for, or associated with, pneumonias of, at the time of the writing, unidentified cause. In that loose usage, the pneumonia in question might be viral. The experts agreed that in that meaning the term is not useful in current epidemiological or clinical practice, since the molecular diagnosis of a pneumonia-causing pathogen will nowadays be readily achieved. That did not stop the term appearing in certain literature in the early days of what became the Covid-19 pandemic.

I agree with the claimants’ submission that the exclusion is not rendered futile by reading ‘atypical pneumonia’ as having its longer-standing, better defined and more rigorous scientific meaning. Business interruption causing loss as a result of Legionella causing an outbreak of disease, for example, might well have been in mind sufficiently to explain the exclusion. Given the availability of that better meaning of the term, and the fact that with that meaning the exclusion has narrower scope, the proper construction of the SARS etc. clause is that it uses the term in that sense. If, which was the main theme of the defendants’ contrary submission, the clause had in mind pandemics (actual or threatened), that language could and would be expected to be used, rather than a seemingly specific medical term for which, it ought reasonably to have been appreciated, an insured curious to understand its scope would readily encounter the specific, long-known meaning.

In developing that main theme, Mr Masefield KC’s major line of argument was that this is a case for the application of the maxim noscitur a sociis, that is to say ‘atypical pneumonia’ takes its meaning from its associates in the SARS etc. clause, SARS and avian flu. The notorious background, it was suggested, was a modern history of SARS, avian flu, swine flu and MERS as identified potential pandemics of major concern. The proper conclusion, it was contended, was that a reasonable policyholder would see in the language of the SARS etc. clause an insurer seeking to exclude “pandemics of concern, SARS, avian flu, swine flu, MERS and other similar respiratory viruses”. By a clear margin I preferred Mr Blackwood KC’s submission in reply that (in summary):

The 2019 Global Policy included a ‘Difference in Conditions/Limits (DIC/DIL)’ provision in the following terms:

“This Policy includes difference in conditions and difference in limits where any Specified Local Policy does not cover the Insured in whole or in part when:

a)

the cover provided by this Policy is broader in meaning or in scope than the cover provided under the relevant Specific Local Primary Policy; and/or

b)

any Sum Insured or Limit of Liability under this Policy is higher than any corresponding limit under the relevant Specific Local Primary Policy; and

to the extent such coverage is not provided but would have been provided had such Specified Local Primary Policy followed the scope of coverage, Sum Insured and Limit of Liability of this Policy.

…”

A set of five provisos (“Provided always that: …”) and a set of six effective exclusions (“No amount will be recoverable under this Policy: …”) followed.

The DIC/DIL provision in the 2018 Global Policy was in different terms, and the failure of the rectification claim means that it does not now matter whether they were more favourable to the claimants than the provision in the 2019 Global Policy. I note only for completeness therefore that, in particular, the claimants relied, if rectification enabled them to do so and then if necessary, on the following paragraph in the 2018 Global Policy which has no counterpart in the 2019 Global Policy:

“Without waiving any of the foregoing [viz., the main DIC/DIL wording], if any Insurers affording other insurance to the Insured denies primary liability under its policy, the Insurers hereunder will respond under the Policy as though such other insurance were not available provided that they shall be subrogated to all rights of the Insured to such other insurance and the Insured shall do all things necessary to enforce such rights.”

Preliminary issue 5 asked whether, if a claim were time barred under (as might be relevant) the Hungarian Policy or the Romanian Policy, that claim would be recoverable under the 2019 Global Policy in any event. One main strand of argument in relation to that concerned the paragraph quoted immediately above from the 2018 Global Policy, if the rectification claim succeeded.

If the posited claim would not have been a valid claim under the relevant local policy, any time bar would be irrelevant to the DIC/DIL cover provided by the 2019 Global Policy. There is no reading of that policy under which an insured could be deprived of rights under it by a failure to bring a bad claim under a local policy before it became time barred. Such an insured’s rights under the 2019 Global Policy would simply turn on an application of its terms (other than the DIC/DIL provision itself) to that claim.

The contentious aspect of preliminary issue 5, therefore, concerned a valid claim under a local policy that cannot be pursued because of time bar. If through rectification the claimants effectively had the benefit for the 2019-2020 policy year of the paragraph quoted in paragraph 82 above, did it mean that an otherwise valid but time barred claim under the Hungarian or Romanian Policy could be brought against the first defendant under the 2019 Global Policy, if and to the extent covered by its terms? That would depend on whether such a claim would fall within the ‘Insurers will respond’ language (the claimants said it would, the defendants said not) and whether it would fall foul of the proviso about rights of subrogation (the defendants said it would, the claimants said not).

Without the potentially complicating factor of time bar, the DIC/DIL cover of the 2019 Global Policy, with or without rectification, might in principle be invoked if the Hungarian or Romanian Policy, respectively, did not cover a Hungarian or Romanian Covid-19 BI claim, but the 2019 Global Policy did, or (and then to the extent that) both local and global policies covered the claim but the global policy afforded a more generous recovery. However, that is not relevant since the 2019 Global Policy is not to be rectified and it is now common ground that it does not respond to any of the claimants’ Covid-19 BI claims due to the infectious diseases exclusion. Preliminary issue 5 therefore also now does not matter.

In the result, the Covid-19 BI claims left to the claimants are only such claims, if any, as are available to the relevant claimants, if they are not time barred, under either the Hungarian Policy or the Romanian Policy. Preliminary issue 4 addresses that, asking:

For the reasons set out in the remainder of this section of this judgment, my answer to all four questions is no:

The Hungarian Policy was a seven-page document issued on 4 September 2019, by which the second defendant undertook “to indemnify the Assured for losses and the related expenses according to the respective Insurance Conditions and in the manner hereinafter provided”. That short policy document stated all of its provisions in both Hungarian and English, but with a note that the Hungarian text was decisive. It identified the “Insured Period” as 30 June 2019 to 29 June 2020, and listed the “Insured Locations”, “Insured Risks” and “Insurance Conditions”.

The “Insured Risks” and “Insurance Conditions” were listed in bullet points, the former a list of types of cover, the latter a list of sets of detailed terms. The former were:

“• All Risks Vagyonbiztositás (All Risks Property Damage)

• All Risks Üzemszünetbiztositás (All Risks Property Damage [sic.])

• Géptörés Biztositás (Machinery Breakdown)

• Géptörés-Üzemszünetbiztositás (Machinery Breakdown Business Interruption)

• Elektronikus Berendezések Biztositása (Insurance for Electronic Appliances)”

It was common ground that the English in the second bullet point was an error and that, since the Hungarian text prevails, it should be read as “All Risks Business Interruption”.

The “Insurance Conditions” were compendiously described at the head of the list (in Hungarian) as the second defendant’s applicable insurance conditions (or, as the document stated it, slightly less well, in English, “As per local policy wording of [the second defendant]”). The list included the second defendant’s (i) General Property Insurance Conditions, Reg No 20819 (Effective from 28.07.2018) (‘the Hungarian General Conditions’), (ii) Special Provisions of All Risks Business Interruption Coverage (ARBISP) (Effective from 15.3.2014) (‘the ARBISP Terms’).

The final page of the policy included the following declaration:

“The Policyholder having signed the above document acknowledges the conditions that are listed in the policy and has full awareness of them, furthermore, expressly accepts the terms stated in “Provisions substantially deviating from the Civil Code” of the insurance conditions pertaining to this insurance contract.”

The ARBISP Terms provided, at Clause I.1. that the insurance covered business interruption caused by “material damage to the insured assets in the course of insured events…”; and by Clause I.2(c) that, “The insurance does not cover that part of the business interruption loss caused by any of the following … c) authority restrictions on reinstatement and operation, or any other dilatory measures”.

Clause XIII.1 of the Hungarian General Conditions provided that the applicable limitation period for a claim under the Hungarian Policy was one year.

The claimants accepted in closing that the ARBISP Terms were incorporated into the Hungarian Policy. It was common ground, in the light of the expert evidence of Hungarian law, that under the principle of contractual freedom in Section 6:59 of the Hungarian Civil Code, parties are free to determine the content of their agreement and may depart from default rules unless deviation is expressly prohibited. Dr Molnár agreed in cross-examination that Hungarian law imposes no doctrinal requirement that business interruption insurance cover be contingent upon property damage. Whether such cover has been granted is therefore a matter on which the parties have freedom of contract.

That means the question becomes one of interpreting the Hungarian Policy. In particular, as regards Clause I.1 the question is whether “material damage to the insured assets” refers to property damage, i.e. physical damage to property insured at one or more of the insured locations. As regards Clause I.2(c), the particular question is whether the relevant claimants’ Covid-19 business interruption loss was loss caused by “authority restrictions on … operation”.

There was no material dispute between the Hungarian law experts or, therefore, between the parties, on the applicable principles. Thus:

It is appropriate to set out Section I of the ARBISP Terms in full. The key language, as it appears in that context, is clear and unambiguous. The full wording is as follows:

“I. Insured Event and Exclusions

I.1 Pursuant to the following provisions, the insurance covers loss and damage caused by a complete or partial interruption of the plant insured, provided that the operation of the plant insured is interrupted for the following reason:

material damage to the insured assets in the course of insured events caused by risk circumstances defined in the insurance policy … and covered by the insurance company.

I.2 The insurance does not cover that part of the business interruption loss caused by any of the following (exclusions).

a)

any extraordinary event or permanent condition caused by other than an insured event during business interruption,

b)

plant equipment extensions or plant renovation works completed after the insured event, in the course of reconstruction,

c)

authority restrictions on reinstatement and operation, or any other dilatory measures,

d)

extraordinary delays in the course of plant equipment reconstruction, such as clarification of ownership, possession, or lease conditions, settlement of probate proceedings, litigations, etc.,

e)

any delays of the policyholder (insured) in commencing the repeated procurement of the assets damaged, destroyed or lost, including financial difficulties,

f)

if any piece of a collection of assets is destroyed and the piece remaining intact cannot be used for operational purposes any longer.”

The plain sense of the language of Section I.1 is that the second defendant covered loss caused by an interruption of the physical operations of an insured business at an insured location resulting from physical damage to property at that location caused by an insured peril. Furthermore, the exclusions only really make sense on that basis, considering especially the language of Sections I.2(b) and (d) to (f), providing or assuming as they do that for there to be a BI loss that might becovered, if not excluded, there must have been physical damage to property requiring repair or replacement.

The claimants’ contrary argument was that business revenues exposed to the risk of loss if an insured peril operated were themselves insured assets, so that there was “material damage to the insured assets” whenever there was a loss of business revenue caused by an insured peril. That was said to follow because a schedule to the Hungarian Policy identified “two insured assets for each insured location, the physical property identified for each row in the “Total Sum Property Damage” column and the business revenue identified in the column “Annual Rent/Revenue Receivable”.” However:

The Covid-19 business interruption losses claimed by the relevant claimants in these proceedings are, as alleged, losses of revenue unrelated to property damage but resulting from the Hungarian government’s emergency measures taken in response to Covid-19, in the nature of travel prohibitions, curfews, and restrictions on the operation of inter alia hotels and restaurants. Such losses were not covered by Section I.1 of the Hungarian Policy.

Section I.2(c) of the Hungarian Policy is relevant only if, to the contrary, Section I.1 were somehow read as applying to such losses. On that basis, the losses alleged in this case would seem to me to be squarely within the language of Section I.2(c), and therefore excluded, being losses caused by restrictions on the operation of the relevant claimants’ insured businesses imposed by the Hungarian authorities.

The claimants’ contrary argument was that an exclusion relating to restrictions on reinstatement and operation was “aimed at strictures placed on necessary re-commissioning or re-building work at an insured location preventing operation for health and safety reasons and the like” and that “other dilatory measures” had in mind “strictures imposing waiting periods on necessary re-commissioning or re-building work” for similar reasons. There is the obvious irony that if correct, that argument reinforces the primary conclusion that the business interruption cover under the Hungarian Policy was an extension of the property insurance, not freestanding cover. However, on its own terms there is no such restriction in the wording of Section I.2(c); and I do not agree that it can be read in as a matter of interpretation. The words are simple, plain words. They apply to the situation at hand.

If therefore, contrary to my conclusion, the losses claimed by the relevant claimants in this case fell within Section I.1 of the Hungarian Policy, they were excluded by Section I.2(c).

From the expert evidence on Hungarian law, it was common ground that the default limitation period under Section 6:22(1) of the Hungarian Civil Code is five years, but that the parties to a contract may stipulate in writing for a shorter limitation period and there is no general power in the court to review the fairness or suitability of such a contractual limitation period, if duly incorporated.

There was a substantial dispute between the parties as to whether the Hungarian General Conditions were duly incorporated into the Hungarian Policy. There was also an issue whether those Conditions, as purportedly incorporated into the Hungarian Policy, satisfied the requirement of 6:22(3) of the Hungarian Civil Code that any agreement altering the default limitation period must be in writing. The claimants argued that because of Section 6:7(2) of the Code, the Conditions themselves would need to have been signed by or on behalf of the insured (“Unless otherwise provided by this Act, a legal declaration is considered to be made in writing if it has been signed by the declarant”). I do not need to express any view on that argument, on which it seems there is no learning in the Hungarian case law, and I prefer not to do so.

As regards incorporation, the experts were agreed that for standard terms to be successfully incorporated into a contract, four conditions had to be met pursuant to Section 6:78 of the Hungarian Civil Code:

The admitted evidence of Dr Sztanó was that a set of the second defendant’s General Conditions was provided, prior to the Hungarian Policy, only on 7 April 2014. However, the Hungarian General Conditions purportedly incorporated were a set of terms having effect only from 28 July 2018. There was neither allegation nor evidence that that set of General Conditions was provided to or drawn to the attention of any of the claimants, or (if relevant) that of Aon (in the UK or in Hungary) prior to the conclusion of the Hungarian Policy.

Dr Molnár expressed the view in his supplementary report that the publication of general terms and conditions on a service provider’s website was sufficient for their incorporation, so that delivery of those terms to the other contracting party was not required. This was said to be founded on a Hungarian Curia decision, reference Gfv.VII.30.394/2018/8, that related to a bank’s standard terms. However, Dr Molnár accepted in cross-examination that: (i) the Curia in that case did not endorse the judgment of the court of second instance on incorporation, it conducted only a limited review into alleged procedural impropriety; (ii) the court of second instance did not rely only on the availability of the bank’s terms and conditions on its website, but cumulatively on that fact and the facts that the bank’s terms and conditions were posted in its branches and that the customer had acknowledged the debt in a notarial deed; and (iii) the process of acknowledging before a notary in Hungary involves the reading out of the relevant document by the notary, in full, to those attending. Dr Molnár therefore agreed, ultimately, that the decision on which he had relied was not a case in which the Hungarian Supreme Court had considered that publication of general terms and conditions on a service provider’s website to be sufficient for their incorporation.

I therefore do not find it established, and to the extent that Dr Molnár may have been putting forward any independent opinion on the point I do not accept his evidence, that mere website publication is enough for incorporation of terms in Hungarian law.

In any event, there was no evidence upon the basis of which I could find that the Hungarian General Conditions were published and available on a Generali website prior to the conclusion and inception of the Hungarian Policy in September 2019.

I agree with the claimants’ submission that the considerations above are fatal to the defendants’ case on incorporation of the one-year time bar in Clause XIII.1 of the Hungarian General Conditions. It is unnecessary to consider whether, if incorporated, it would have been unenforceable as an unfair standard term contrary to Section 6:102(1) of the Hungarian Civil Code, or a rather specific point taken on claim notification requirements that was said to affect matters under Clause XIII.2 of the Hungarian General Conditions.

The defendants’ time bar defence would have failed anyway, in my judgment, because these proceedings were brought within 12 months of the relevant date, for limitation purposes. Under the Hungarian Policy, if it covered freestanding business interruption losses, contrary to my conclusion on cover, the insured were entitled to claim only their relevant net loss at the end of the 36-month period of cover. The Hungarian law experts were agreed that Section 6:22(2) of the Hungarian Civil Code provides that a limitation period commences “when the claim becomes due”.

It was also common ground between the experts that there is no Hungarian case law considering commencement of a limitation period for non-damage business interruption losses, still less for such losses with a 36-month indemnity period. However, when Dr Molnár was invited to consider in cross-examination the possibility (which in my judgment matches this case) of a policyholder claiming non-damage BI losses who does not know whether it has sustained insured loss, or if it has in what amount, until the end of the indemnity period, because the insurance requires a calculation to be done at that point inter alia to take account of the insured’s experience over that full period, including any rebound in customer numbers after the business interruption has ceased, his opinion supported the claimants:

“Q. So if you’re looking for a logical starting point for the commencement of the limitation period in cases of non-physical damage business interruption loss, that couldn’t logically be before the end of the indemnity period, could it, if you have one insured event?

A.

That’s correct in your assumption, considering one insured event, yes.”

On the facts, therefore, a contractual limitation period of one year, if there had been one, could not have commenced before March 2023 and so could not have expired before March 2024. However, the Claim Form in these proceedings was issued on 22 September 2023. That means the claim pursued on the Hungarian Policy was not time barred and I do not need to lengthen this judgment, and will not do so, by considering the further question raised by the claimants of whether the running of the limitation period, if it commenced in 2020, was suspended by operation of Section 6:24(1) of the Hungarian Civil Code and, if so, whether it was suspended for long enough for the Claim Form to have been in time under Section 6:24(2) of the Code (by which, on the facts of this case, the lifting of any suspension would have had to be later than 22 June 2023 for the claim to have been in time).

The Romanian Policy was an eight-page document issued on 15 July 2019, entitled “Property Damage Policy”. Its contents are set out in English and Romanian, and it provides that the Romanian text is definitive. It states that the insured’s business description is “Hotels”, and lists six “Risk Location[s]”, each a Romanian hotel. It states the “Insured Risks” to be, ““All Risks” policy for property damages and business interruption, according to the standard conditions of Generali Romania”.

The policy document used the convention throughout of “Aplicabilitate:”, followed by initials, to indicate whether provisions that it included or identified applied to property damage (“PD”), business interruption resulting from property damage (“PD from BI”), or both (“PD, BI from PD”). It was not said for any provision that its “aplicabilitate” was “BI” simpliciter. That drafting convention was applied inter alia to a list, set out in the policy document, of all the general exclusions in the 2019 Global Policy. For those exclusions, that was done by an “Applicable to:” column (stated, as it happens, only in English), split into two sub-columns, “PD” and “BI from PD”, with an entry of either “YES” or “NO” in each sub-column for each exclusion.

As Mr Walsh KC submitted, there is room to wonder, if those general exclusions are scrutinised individually and closely, why in some cases they might ever be capable of being in play in a property damage claim (with or without consequential business interruption). That does not draw the sting of the drafting convention for present purposes. What matters is that it evidently displayed that ‘PD’ and ‘PD from BI’ were the only two categories of cover understood and intended to be provided by the Romanian Policy. The structure and content of that section of the policy document raises and answers a simple question, thus: Q: Where’s the sub-column for (pure) BI? A: There isn’t one because there is no such cover under this policy. In his closing argument, Mr Walsh fairly and candidly acknowledged that that was the argument against him and that if I were minded to read this aspect of the policy document in that way, then the claimants would not succeed on the Romanian Policy. I am so minded; in fact, I think it is not sensibly capable of any other reading.

The Romanian Policy purportedly incorporated, by express reference, the “Generali “All Risks” Insurance Conditions for Property Damage (Mod. DMT-C / Version 05 - 1 July 2015)” (‘the Romanian General Conditions’). There were four different relevant references in the policy document:

The Romanian General Conditions defined the scope of cover, or insured event, for BI cover at Clause 1.1 of Section II – Business Interruption, in these terms, namely:

“loss of the insured interest, except for the excluded provisions, if at any time during the period of insurance the property insured under Section I of the policy - Material Damage [i.e. the Property Damage section] suffers a loss or damage which is compensable under Section I if the franchise [Deductible] would not have been applied, thereby causing an interruption or disruption of the insured activity.”

Unarguably, in the light of all of the foregoing, BI cover was provided by the Romanian Policy, if the Romanian General Conditions were incorporated, only as an extension to the PD cover, available if and only if there was loss or damage for which Section I (PD) provided an indemnity (or would have done so but for any applicable deductible).

The claimants advanced a submission that a list in the policy of the “Sub-limits of indemnity according to Master Policy [i.e. the 2019 Global Policy]” was (i) a list of insured perils potentially defining the scope of cover and (ii) somehow provided freestanding BI cover because it included sub-limits, taken from the 2019 Global Policy, for “Contagious disease, food poisoning and other health risks”, “Public Authority Measures”, and “Denial of Access”. Quite apart from the fact that reciting in that way the sub-limits of the 2019 Global Policy is plainly not a purported definition of the scope of cover, the policy applicability convention was followed, and those three sub-limits had “aplicabilitate” to, respectively, “BI from PD”, “PD” and “BI from PD”. The claimants’ argument of construction was, with respect, hopeless.

The claimants sought to resist the conclusion that therefore the Romanian Policy did not extend to the Covid-19 BI losses claimed by arguing that Section II, Clause 1.1 of the Romanian General Conditions is an ‘unusual clause’ that under Romanian law required to be, but (it was said) had not been, expressly accepted in writing. There were two insuperable problems with that argument:

As with the claim under the Hungarian Policy, that is an end of the claim put forward on the Romanian Policy. The claimed Covid-19 BI losses did not arise from property damage covered by Section I of the insurance. Therefore they fell outside the BI cover provided by Section II of the insurance.

Even if that might not be true, simply on the language of Clause 1.1 of Section II, it is undoubtedly true when that Clause is read with the exclusions set out in Clause 5 of Section II. They included an exclusion (by Clause 5.1.3.3) of loss occurring during periods of inactivity including but not limited to when operations cease for any reason other than property damage insured under the policy. That squarely catches the Covid-19 BI losses claimed in this case.

The defendants also relied on Clause 5.1.1 of Section II, which excluded loss due to interruption or disruption of insured activity caused by (in translation), “any restriction of access or otherwise to the location of the insured property imposed by governmental or public authority (central or local)”. It is not clear to me that that exclusion quite captures the losses claimed here. I find the meaning of the phrase “restriction of access or otherwise to [a] location” a touch elusive and I do not need to take a final view on whether it covers the Covid-19 measures in Romania that are said by the claimants, in the pleading of the claim on the Romanian Policy, to have generated Romanian Covid-19 BI losses.

The Romanian law experts did not agree whether an exclusion clause in an insurance contract will constitute a ‘limitation of liability’ within the scope of Article 1203 of the Romanian Civil Code, so as to require to be expressly accepted in writing. But that requirement was satisfied, if it applied, because of the signed declarations in the policy document itself. For completeness, I would have preferred Mrs Peli’s view that at all events exclusions such as those in Clause 5 of Section II of the Romanian General Conditions are essential terms of the policy defining what is and is not insured, not clauses caught by Article 1203. They are negatively expressed counterparts to the positively expressed insuring clause, to be read with that clause so as to define the scope of cover.

It was common ground that under Romanian law, there is a two-year limitation period from which parties may not derogate by contract and under which time begins to run, in a claim under an insurance contract, from the moment when the insurer has the obligation to pay the indemnity provided for by the contract. It was also common ground that there is no relevant decision of the Romanian courts as to the date on which, therefore, the limitation period will commence for a claim for non-damage BI loss.

What loss, when capable of being assessed and how to be measured, will be (as was the case for the Hungarian Policy) a question of, or function of, the proper construction of the policy, rather than something governed by its own rule of law. Here (again as was the case for the Hungarian Policy), the policy obliged the insurer to indemnify the insured in respect of its final net loss, if any, at the end of the 36-month BI indemnity period. Only then, that is to say in early 2023, could there have been any obligation on the third defendant to pay an otherwise valid non-damage BI claim resulting from the Covid-19 pandemic, if such a claim had existed. The Claim Form here having been issued in September 2023, no Romanian law time bar had come into existence.

I noted at the outset (paragraph 8 above) that the preliminary issues, as ordered, did not stand as an examination paper every question in which had to be tackled and answered. In the event, it has not been necessary to consider or decide quite a few of the individual questions posed.

The key conclusions I have reached, in each case for the reasons set out in the corresponding main section of this judgment, above, are that:

I have indicated in the course of the main sections of this judgment, above, how I would answer certain of the specific questions identified as preliminary issues. Counsel will no doubt assist as to the most appropriate form and content for an Order that can be drawn up to give effect to my conclusions.

Construction of the 2018 and 2019 Master Global Policies / Limits:

The true construction under the 2018 Master Global Policy of:

the “Policy Loss Limit”;

the Disease Clause and any sub-limit;

the maximum indemnity period;

General Clause 5 (regarding the composite nature of the Policy);

the exclusion relating to “Severe Acute Respiratory Syndrome (SARS), Avian Flu, and/or atypical pneumonia, or fear of threat”;

the Difference in Conditions / Difference in Limits cover; and

the Hours clause.

The true construction under the 2019 Master Global Policy of:

the exclusion of infectious diseases “designated as ‘Phase 5’ or ‘Phase 6’ pandemic alert”;

“Occurrence”; and

the Difference in Conditions / Difference in Limits cover.

How the Limits and maximum indemnity period apply to multiple insureds and/or multiple premises or locations and/or whether the 2018 Global Master Policy contained any annual aggregate limit for all claims for loss resulting from interruption of or interference with the Business under the Disease Clause.

Whether each of the Claimants’ hotels, workspaces and restaurants constituted a separate “Business” and/or was agreed to be a separate “Business” by the Defendants.

Rectification:

Whether there was an outward expression of accord and/or a concluded agreement between the Defendants and the Claimants (whether via Aon or otherwise) that, when cover renewed for the 2019 policy year, the cover would conform to the extent of (or be no less than) the cover in the 2018 Master Global Policy such that the Claimants would not be disadvantaged in particular when bringing a claim under the 2019 Policy. In particular:

The true content, meaning and effect of communications between the Defendants and Aon prior to the issue of the 2019 Master Global Policy.

If there was an outward expression of accord and/or concluded agreement between the Defendants and Aon regarding a conformity clause, whether it continued until 18^th September 2019.

If there was an outward expression of accord and/or concluded agreement between the Defendants and the Claimants (whether via Aon or otherwise) regarding a conformity clause, was it only limited to:

conditions precedent in the 2018 Master Global Policy; and/or

the wording of the 2018 Master Global Policy but not the Schedule.

If there was an outward expression of accord and/or concluded agreement between the Defendants and the Claimants (whether via Aon or otherwise) regarding a conformity clause

did the Claimants rely on the statements made by Aon;

if so, in what way; and

was it reasonable for the Claimants to rely on such statements that the Defendants should be taken as legally bound with wording yet to be agreed?

Whether there was an error or a common mistake between the parties in issuing the 2019 Master Global Policy without any wording to reflect the conformity agreement.

As to any differences between the policy terms in the 2018 Master Global Policy and the 2019 Master Global Policy:

was the cover provided under the 2018 Master Global Policy materially less extensive overall than the cover provided under the 2019 Master Global Policy?

did the 2019 Master Global Policy disadvantage the Claimants compared to the 2018 Master Global Policy?

If the 2019 Master Global Policy Wording and/or Schedule is to be rectified, whether, and if so to what extent, claims arising out of Covid-19 are covered under the 2019 Master Global Policy.

Cover under the Hungarian and Romanian Local Policies

Whether business interruption losses arising from the Covid-19 pandemic are covered under the Hungarian Local Policy.

Whether any claim under the Hungarian Local Policy is time-barred.

Whether business interruption losses arising from the Covid-19 pandemic are covered under the Romanian Local Policy.

Whether any claim under the Romanian Local Policy is time-barred.

If any claim is time-barred under the Hungarian Local Policy or the Romanian Local Policy, whether that claim is recoverable under the 2019 Master Global Policy in any event.