ABN Amro Bank N.V. v Royal & Sun Alliance Insurance Plc & Ors.

In these proceedings, a claimant bank (ABN Amro) sued its insurers (the 1 ^st to 14 ^th defendants – altogether “the Underwriters”) and its insurance broker (the 15 ^th defendant – “Edge”). The claim against the Underwriters was for some £31.3 million allegedly due under a policy of marine cargo and storage insurance (the Policy). The claim against the broker, Edge, was brought on the basis of breach of contract and negligence by Edge in placing the Policy, in the alternative, in case all or part of the claim against the Underwriters failed (as in the event it did against the 8 ^th defendant (Ark) and the 10 ^th defendant (Advent)).

Mr Justice Jacobs (the judge) decided many issues in his 1,036 paragraph judgment after a 20-day trial. Significantly for these appeals, he decided (i) that the Appellant Underwriters (all excepting Ark and Advent) were liable to indemnify ABN Amro for the financial loss that it had sustained under the Transaction Premium Clause (the TPC) contained in the Policy, and (ii) that Ark and Advent were not liable under the Policy, because ABN Amro was estopped by convention from relying on the TPC as against them, and Edge was, in consequence liable to ABN Amro for the shares of the indemnity of Ark and Advent, which amounted to some £3.3 million.

The factual context was that ABN Amro had advanced working capital by way of structured commodities financing to Transmar and Euromar through a special purpose vehicle (Icestar) to finance their trade in cocoa beans and cocoa products (the Goods). In general terms, the “repo” transactions comprised Transmar and Euromar selling the Goods to Icestar/ABN Amro. Transmar and Euromar then had an obligation to repurchase the Goods from Icestar/ABN Amro at a specified time and (higher) price. Transmar and Euromar defaulted in doing so in 2016. The Goods were of poor quality and Icestar/ABN Amro sustained consequent financial losses when the Goods were sold. ABN Amro sought indemnity against those losses under the TPC which provided as follows:

The Underwriters had contended before the judge that, on its true interpretation and in the light of the admissible factual matrix, the TPC did not cover financial losses of the kind that ABN Amro had sustained. It is fair to say that the TPC is a non-standard clause in the marine market. The TPC was found within a marine cargo policy that insured the Goods that Icestar/ABN Amro bought as part of the repo transactions. The Policy covered mostly physical loss and damage, but also (as the judge found at [234] to be common ground) wider cover in the absence of physical loss and damage, namely Business Contingency Cover, CEND cover (Footnote: ¹) and Fraudulent Documentation cover. ABN Amro had suffered only financial, not physical, loss.

The factual background to Edge’s appeal is more complex. In brief, the story starts with the policy year preceding the Policy. In July 2015, during that previous policy year, Edge (through a Mr Mullen) broked a policy variation to the lead underwriter, Royal and Sun Alliance Insurance plc (RSA) (through a Mr Beattie). Mr Mullen of Edge told Mr Beattie of RSA that the amendments, which included the bespoke TPC and a non-avoidance clause (the NAC), had been drafted by ABN Amro’s lawyers. Mr Beattie of RSA scratched the amendment (the July Endorsement), but did not instruct Edge to circulate it to the following market, which included Ark and Advent. Accordingly, Ark and Advent did not know about the TPC and NAC. Mr Mullen and his colleague Mr Lockyer of Edge honestly believed that Mr Beattie of RSA had agreed the TPC on behalf of the entire following market, including Ark and Advent, under a Policy provision allowing the lead underwriter to agree contract changes.

The judge held, however, that only RSA, and not the following market, was bound by the TPC and the NAC contained in the July Endorsement.

When the prior year’s policy was renewed to become the Policy, each of the Underwriters, including Ark and Advent, scratched or signed, and was given a copy of, the Policy wording including the TPC and the NAC. The judge found that when Mr Mullen or Mr Lockyer of Edge broked the renewal to Mr Blewett of Ark, and when Mr Lockyer of Edge broked the renewal to Mr Cooke of Advent, those Respondent Underwriters were told by Edge that the renewal Policy was “as expiry”. Edge, of course, believed that the Policy was the same as the expiring one. Neither Mr Blewett of Ark nor Mr Cooke of Advent read the wording of the Policy. As a result, the judge found that Edge on the one hand, and Ark and Advent on the other hand, were at cross purposes about the meaning of “as expiry”. Edge understood “as expiry” to include the TPC and the NAC. Ark and Advent understood “as expiry” as not including the TPC and NAC.

Ark and Advent (amongst others of the Underwriters) contended at trial that they could avoid the Policy on the basis of the “as expiry” misrepresentation. The judge, however, held that the NAC prevented avoidance. Underwriters had affirmed the Policy by their conduct in dealing with ABN Amro’s claim without reservation of rights between notification in October 2016 and the first suggestion of avoidance in April/May 2020 after proceedings had started. Ark and Advent, however, established before the judge that the “as expiry” representation had induced them to write the Policy. And as I have said, the judge upheld Ark’s and Advent’s defence that the “as expiry” representation constituted an estoppel by convention preventing ABN Amro from relying on the TPC to make its claim for an indemnity against them.

There were, therefore, when the matter was argued over two days before the court, two distinct live appeals. 8 days after the argument concluded, the parties informed the court that the first appeal had been compromised on the terms of a Tomlin order that they invited the court to make, requesting that no judgment was issued in respect of the first appeal. The court responded by saying that a full draft judgment had already been prepared in respect of both appeals, and that it was concerned about a number of matters including the form of the agreed Tomlin order and whether it was appropriate, despite the compromise, to deliver the judgment that had been prepared. The court referred the parties to the White Book at 40.2.6, Brooke LJ in Prudential Assurance Co Ltd v. McBains Cooper [2000] 1 WLR 2000 , and R(Mohamed) v. Foreign and Commonwealth [2010] EWCA Civ 65 per Lord Neuberger MR, sought their submissions and directed a short hearing. On 23 November 2021, the parties responded with an amended agreed draft order (not this time in Tomlin form) dismissing the first appeal and making certain provision for costs. The parties submitted that they each recognised that whether or not to deliver judgment in the first appeal was a matter for the Court. They each nonetheless submitted detailed submissions on the point. The Underwriters asked us not to deliver our judgment, and ABN Amro and Edge were neutral, but pointed out the benefits to the market of so doing, despite the fact that the first appeal concerned a bespoke clause. In the result, we have decided to deliver judgment on both appeals. They were intimately connected. It would have taken significant time to disentangle the points. The parties were not unanimous in asking us not to deliver the judgment, and we have been told that the case has attracted interest in the insurance press. In delivering this full judgment, however, we do not wish to discourage parties generally from compromising any case at any stage. If, however, there are serious negotiations going on after the argument has concluded, it would help the court if it were informed informally so that it can avoid wasting time writing judgments that are not necessary. We have also informed the parties that we intend to approve the draft agreed order submitted on 23 November 2021 disposing of the first appeal.

In the first appeal, the Appellant Underwriters were only given permission to pursue one ground of appeal concerning the proper interpretation of the TPC.

In these circumstances, the Appellant Underwriters contended that the judge ought to have interpreted the TPC as applying only to calculate, where applicable, the measure of indemnity in the event of physical loss or damage to the Goods insured under the Policy. The TPC did not, the Appellant Underwriters argued, either on the basis of its express terms or by necessary implication, extend the subject matter insured to include lost profits. The judge paid insufficient attention to (i) the factual matrix and the insurance market in which the Policy was placed, (ii) the nature of a marine cargo insurance policy, (iii) the correct identification of the subject matter of insurance and its significance, (iv) the language and purpose of the TPC, and (v) the entirety and structure of the Policy.

The substantive appellant in the second appeal is, of course, Edge, since it was Edge that, on the basis of the judge’s decision, picked up the shares of the indemnity for which Ark and Advent escaped liability as a result of the estoppel by convention that ABN Amro relied upon.

In the second appeal, therefore, Edge appeals the judgment against it on three main grounds: (i) estoppel by convention cannot arise when, as the judge found here, Edge on the one hand and each of Ark and Advent on the other hand, were at cross purposes, so that the broker (acting for ABN Amro) and these two underwriters did not share a common assumption (the cross purposes ground), (ii) the estoppel by convention defence was precluded by the NAC, which said that the Underwriters would not “seek to reject a claim for loss on the grounds of … misrepresentation other than … fraudulent misrepresentation”. The estoppel by convention defence had been based on a non-fraudulent representation made by Edge to each of Ark and Advent that the Policy was “as expiry”. The judge had held that the expiring policy was in their cases a policy that did not include the TPC on the basis of which ABN Amro sought its indemnity (the NAC ground), and (iii) the judge had been wrong to find that it would be unjust or unconscionable for ABN Amro to resile from the assumption made by Ark and Advent that the Policy they were writing did not include the TPC. The judge held that ABN Amro could not resile from that assumption and rely on the TPC against Ark and Advent, because, had they known of the inclusion of the TPC, Ark and Advent would not, as the judge found, have written the Policy (the injustice ground).

Ark and Advent argued by their Respondents’ Notice in the second appeal that the judge ought anyway to have allowed their claim to rely on an estoppel by representation. ABN Amro was estopped by representation from denying the truth of the “as expiry” representation of fact. The issues raised on estoppel by representation are (i) whether an estoppel by representation was precluded by the NAC, (ii) whether Ark and Advent could show detrimental reliance on the misrepresentations by relying simply on the judge’s findings that they were induced by them to write the Policy, (iii) whether the representations were sufficiently clear and unambiguous to found an estoppel by representation, and (iv) whether it would be unconscionable to hold Ark and Advent to the TPC.

With this introduction, I intend to deal with the two appeals by setting out the remaining salient terms of the Policy, summarising the judge’s reasoning on the critical points, and then dealing with the grounds and issues I have identified in each appeal.

The NAC provided as follows:

The Basis of Valuation clause provided as follows:

The General Conditions clause included:

I have included a copy of the main sections of the version of the Policy that Mr Peter MacDonald Eggers QC, leading counsel for the Underwriters, referred to during the hearing as Annex 1 to this judgment.

The judge’s judgment is too long even to attempt to summarise its entire contents here. Reference should be made to that judgment as necessary. I have limited myself to the most critical parts.

The judge set out the factual background in meticulous detail at [29]-[167]. He then set out the Policy in its various iterations.

At [175]-[188], the judge set out the principles of contractual interpretation applicable to the exercise he was about to undertake in terms that the parties have not criticised. I do not intend to repeat that treatment here.

After dealing in detail with the parties’ arguments, the judge dealt with interpretation at [206]-[314], under headings as to factual matrix, endorsements 3 and 4, Policy context, the terms of the TPC, the TPC as a Basis of Valuation provision, the balancing exercise, and the positioning [of the TPC] in the Policy. Any summary of the judge’s detailed iterative process of interpretation would be inadequate, but I shall seek to highlight the main points.

First at [209]-[210] he said that the most significant aspect of the factual matrix concerned the nature of the market, namely the London marine market where there were underwriters who specialised in writing insurance on cargo risks, whose core business was risks concerning physical loss and damage to cargo whilst in transit or in store. But, the evidence that the judge heard from underwriting experts agreed that coverage which did not require physical loss or damage was generally available in the London marine cargo market in 2015/16 as an add-on to marine cargo and storage policies. The parties did not dispute that the Policy included three add-ons which did not require physical loss or damage: business contingency cover in clause 20 of section 2, cover for fraudulent documentation in clause 17 of section 2, and the CEND cover in section 3.

Secondly, the judge found at [213] that the evidence showed that, at the relevant time, there was an incredibly difficult market-place: there was “clearly, overcapacity, with underwriters straining to attract and retain business”. Even so, at [214], the judge noted that there was no precedent in the marine cargo market for an add-on protecting “the insured in respect of the contractual default of a counterparty leading to a non-physical loss on cargoes which the insured had purchased”.

Thirdly, the judge noted at [217] that some London underwriters specialised in writing trade credit insurance, but underwriters and brokers in the marine cargo market only knew at a very general level of the existence of these specialist underwriters. The TPC was bespoke wording drafted by ABN Amro’s lawyers. In these circumstances, the judge thought that the factual matrix served to underline the need, discussed by Sir Ross Cranston in Engelhart CIP (US) LLC v. Lloyd’s Syndicate 1221 [2018] EWHC 900 (Comm) at [39]-[41] (Engelhart) for there to be clear words before the cover could be held to extend beyond the normal cover in the marine cargo market.

In dealing with the Policy context, the judge said that the Underwriters had referred to many provisions within the Policy which indicated that the Policy was intended to cover the risk of physical loss and damage to cargo. That was hardly surprising. Clear words were needed if wider cover was to be provided. But the three add-on clauses in the Policy also answered the Underwriters’ reliance on the General Conditions, which were applicable to all sections of the Policy, that it was agreed in general terms that the Policy was to protect “the interest of” ABN Amro at all times and in all circumstances, and that it was, to quote the wording of the Policy, “ to protect against all risks of physical loss of or damage to the Subject Matter Insured from whatsoever cause arising”.

At [239], the judge concluded that whilst that part of the clause did refer to all risks of physical loss and damage, the immediately following wording indicated that “it was necessary to have regard to all the contractual terms in order to ascertain the scope of the cover”. The three add-ons showed that the cover was not concerned exclusively with physical loss and damage. The General Conditions could not be construed as an express limitation denying cover for risks of non-physical loss or damage.

At [245], the judge emphasised that the language was the most important consideration in the overall unitary exercise of construction. If the language were clear, it was unlikely that arguments based on factual matrix would result in a different conclusion being reached.

In dealing with the words of the TPC itself, the judge interpreted them first in the light of the Policy as a whole, doing so without reference to the fact that the TPC appeared in two places in the Policy – both within section 1 and in a free standing section 4.

The judge dealt at [252]-[264] with essentially the following points on interpretation:

It was indisputable that the TPC confirmed in its first paragraph the existence of coverage for the matters set out in it.

It was clear from the language that the TPC went beyond cover that was either for, or dependent upon, physical loss and damage to goods.

There was nothing in the express language of the TPC which provided a link either to the coverage for physical loss and damage to cargo or to the other non-physical loss and damage perils covered by the policy, for example the CEND and Business Contingency Coverage.

The TPC used a series of key expressions (Transaction, Transaction Premium and Default) which were not used elsewhere in the policy; each of those terms was defined in terms which were unrelated to physical loss and damage to the cargo.

The start of the TPC and the definition of “Transaction” introduced the concept, which was critical to its operation, of “Default”, which was ordinarily alien to a policy or clause simply covering the risk of physical loss or damage.

The Underwriters’ argument was primarily based not on the language but upon factual matrix, the contract as a whole, and commercial consequences leading to the ordinary and natural meaning being circumscribed. That was not appropriate in the context of this carefully drafted clause.

The Underwriters’ argument had the effect of negating express words which were found in the definition of “Default”, which referred to the client’s failure or refusal “(for whatever reason) to purchase or repurchase”.

The Underwriters’ argument produced the result that the TPC had no application in a situation in which the cargo was destroyed (for example in a warehouse fire) or so damaged that it could not in practice be sold and delivered to an exchange or sold on the open market. It was common ground that the TPC would have no application in such circumstances; because the coverage was referrable to the “Actual Sale Price”, which was defined as the sum received upon the sale of the Subject Matter Insured to the applicable Exchange or to a third party on the open market. As Ms. Sabben-Clare had said in opening, it would be a “curious beast” for a clause which provided a measure of loss or basis of valuation to be inapplicable in a situation of loss or damage (i.e. destruction or extensive damage to goods rendering them unsaleable) which could foreseeably occur and for which the parties would have expected the insurance to provide coverage.

The Underwriters’ case that the TPC was no more than a “Basis of Valuation” clause was neither sound nor persuasive (explained in detail at [265]-[279]).

The wording of the provision concerning “Sellers Insurance Interest” marginally reinforced the case that there was no relevant link in the TPC to physical loss or damage.

At [280]-[295], the judge conducted the balancing exercise tying his analysis back to the approach identified in Rainy Sky v. Kookmin Bank [2011] UKSC 50; [2011] 1 W.L.R 2900 (Rainy Sky) and Wood v. Capita Insurance Services Ltd [2017] UKSC 24; [2017] AC 1173 (Wood v. Capita), concluding that commercial considerations did not tell against ABN Amro’s interpretation. That would have ignored the fact that the Policy was placed and then renewed in soft market conditions. No point had been taken that ABN Amro’s claim was at odds with commercial common sense.

Finally on interpretation, the judge dealt with the position of the TPC in the Policy at [296]-[314]. He said that the TPC should mean the same wherever it appeared. The fact that it appeared in a separate section 4 was, in the judge’s view, “the clearest indication” that the coverage provided by the TPC was standalone coverage, not linked to the need for physical loss or damage or the operation of the perils covered by the other non-physical loss and damage add-ons.

The judge dealt with the law on rectification and estoppel at [315]-[331], on the basis of the Underwriter’s case that three conversations between Mr Beattie of RSA and Mr Mullen of Edge prevented ABN Amro from relying on the terms of the Policy as he had interpreted them, and led to a case for either rectification of the TPC to make it applicable only where there was physical loss, or ABN Amro were precluded by estoppel from asserting otherwise.

The judge said that there was no significant dispute between the parties as to the applicable legal principles, even though such a dispute has now emerged before this court.

The judge set out the test for rectification from the Court of Appeal’s decision in FSHC Holdings v. GLAS Trust [2020] Ch 365 (FSHC) at [176] to the effect that: “it is necessary to show either (1) that the document fails to give effect to a prior concluded contract or (2) that, when they executed the document, the parties had a common intention in respect of a particular matter which, by mistake, the document did not accurately record. In the latter case it is necessary to show not only that each party to the contract had the same actual intention with regard to the relevant matter, but also that there was an outward expression of accord meaning that, as a result of communication between them, the parties understood each other to share that intention”. He referred to [80]-[87] of Leggatt LJ’s judgment under the heading “Tacit Agreement” making it clear that the concept of an “outward expression of accord” did not require that the parties’ common intention should be declared in express terms. The shared understanding could be tacit. It might therefore include understandings that were so obvious as to go without saying, or that were reached without being spelled out in so many words. It was “fundamental that contractual rights and obligations should be based on mutual assent which the parties have manifested to each other”. The judge then cited MacGillivray on Insurance Law 14th edition at [12-002] for the proposition that “[t]here is a presumption that a policy which is issued by the insurer and accepted by the insured contains the complete and final contract between the parties. Consequently, the court’s equitable jurisdiction to rectify insurance policies is exercised with restraint inside certain well-established limitations, or else it would tend to destroy certainty in insurance business”. The judge thought that it might be “that there is nothing particularly special about insurance contracts, because there is in any event a “natural presumption” that a written contract is an accurate record of what the parties agreed”. FSHC at [46] had made clear that it was necessary to show that “at the time of executing the written contract the parties had a common intention (even if not amounting to a binding agreement) which, as a result of mistake on the part of both parties, the document failed accurately to record. This requires convincing proof to displace the natural presumption that the written contract is an accurate record of what the parties agreed” (see also the Contract Certainty Code of Practice published by Lloyd’s, the Association of British Insurers and other industry bodies in October 2012, which reflected best practice).

On estoppel by convention, the judge referred to Blindley Heath Investments Ltd. v. Bass [2015] EWCA Civ 1023 (“Blindley Heath”) and Aras v. National Bank of Greece SA [2018] EWHC 1389 (Comm) for the proposition that “[a]n estoppel by convention arises if (i) there is a relevant assumption of fact or law, either shared by both parties, or made by party B and acquiesced in by party A, and (ii) it would be unjust to allow party A to go back on that assumption” (see Chitty on Contracts 33rd edition at [4-108], which reflected the summary by Lord Steyn in Republic of India v. India Steamship Co Ltd (No 2) [1998] AC 878 (“Republic of India”) at page 913E-G:

Lord Steyn’s principles were elaborated upon by Briggs J in HM Revenue v. Benchdollar [2009] EWHC 1310 (Ch) (“Benchdollar”), in the context of an estoppel by convention arising out of non-contractual dealings:

The parties referred before us to the UK Supreme Court’s decision in Tinkler v. HMRC [2021] UKSC 39 (“Tinkler”), which was delivered after the judge’s judgment, and approved Blindley Heath and Benchdollar.

The judge referred to Mitchell v. Watkinson [2014] EWCA Civ 1472 at [52], where it was suggested that the Benchdollar principles applied to contractual cases. Burrows JSC said they did at [78] in Tinkler.

The judge then said at [331] that “[u]ltimately, the present case turns upon the facts”, rather than on fine differences in the precise formulation of the principle.

After a lengthy treatment of the evidence, the judge concluded at [426] that the evidential foundation for the Underwriters’ case on rectification, collateral contract and estoppel did not exist. Mr Beattie for RSA and Mr Mullen for ABN Amro intended to contract on the terms set out in the various documents which Mr Beattie signed. At [438], the judge made clear that the estoppel argument in relation to Ark and Advent (and two others) was quite separate: “[t]hese [“as expiry”] representations were primarily relied upon as misrepresentations which entitled those underwriters to avoid” the Policy. The Underwriters had argued in closing that “these representations could found an estoppel which prevented [ABN Amro] from relying upon the TPC and also the [NAC]”.

At [439]-[447], the judge decided that the following market was not bound by the July 2015 endorsement.

At [448]-[484] the judge dealt with non-disclosure and misrepresentation, saying first that the Underwriters’ case of non-disclosure and misrepresentation was introduced belatedly in April 2020. The arguments focused on the contentions by ABN Amro and Edge that (i) the NAC provided a complete answer to the suggested avoidance, since fraud was not alleged, (ii) affirmation provided a second complete answer, and (iii) even if fraud were not required, there were no material non-disclosures or misrepresentations. There were two main non-disclosures relied upon: (i) the purpose or intention behind the TPC – such purpose or intention being a matter of fact – was a material circumstance which should have been disclosed, and (ii) the presence of the NAC. Specific underwriters then relied on other specific non-disclosures, and Ark and Advent (and two others) relied on the “as expiring” misrepresentation. The judge noted at [462] that it was a feature of the alleged non-disclosures that each arose from, or was directly related to, the wording that was actually presented to the underwriters and which each of them scratched. Edge’s core argument was that it was the underwriter’s responsibility to read and understand the slip with which he is presented for agreement.

As regards the non-disclosure of the NAC, the judge accepted ABN Amro’s and Edge’s submission that the principle was that a person who signs a document knowing that it is intended to have legal effect is generally bound by its terms, whether he has actually read them or not (see Saini J in Higgins & Co Lawyers Ltd. v. Evans [2019] EWHC 2809 (QB) at [75]). Clauses such as the NAC were effective in accordance with their terms (see Toomey v. Eagle Star (No. 2) [1995] 2 Lloyd’s Rep 88, HIH Casualty and General Insurance Ltd. v. New Hampshire Insurance Co and others [2001] EWCA Civ 735, HIH Casualty and General Insurance Ltd. v. Chase Manhattan Bank [2003] UKHL 6 and Mutual Energy Ltd. v. Starr Underwriting Agents Ltd. [2016] EWHC 590 (TCC)). The judge held that the NAC was comprehensive, and its effect was that any non-disclosure or misrepresentation relied upon in support of avoidance had to be alleged and shown to be fraudulent, including an allegation of non-disclosure relating to the NAC itself, whether taken on its own or in combination with the TPC.

As regards affirmation, the judge said at [486] that the legal principles were not in dispute. At [502], the judge said that the key document relied upon as constituting the affirmatory act was the Underwriters’ defence. After a detailed analysis of the pre-action correspondence and the pleadings, the judge said at [529] that there was no suggestion in the lengthy defence that any point on avoidance for misrepresentation or non-disclosure was either taken or reserved. The judge concluded at [541] that the service of the defence, when viewed in the context of the surrounding circumstances, demonstrated objectively and unequivocally that the insurers were making an informed choice to affirm the policy which had been concluded. This was demonstrated by the background, and by the fact that the defence asserted that other matters had not been disclosed, and that the underwriters retained the premium with no offer (until 2020) to return it. The Underwriters had knowledge of all material facts at the relevant time. At [566], the judge concluded that avoidance was barred on the facts of the present case by affirmation.

At [568]-[582], the judge dealt with some basic and uncontentious legal principles concerning non-disclosure, misrepresentation, materiality and inducement.

At [595], the judge concluded that he did not consider that any of the Underwriters could properly allege that the TPC was not disclosed to them. At [610]-[611], the judge said that he did not “accept the proposition that the insured has a duty to tell the insurer of unusual policy terms, or to explain their purpose or effect”, nor did he accept the Underwriters’ supposed reason for that proposition which was that, in the marine market, “an insurer cannot reasonably be expected to read the terms of the policy that he is subscribing”. The judge pointed out that “the proposition that underwriters should not read the slips which they sign would in [his] view come as a surprise to generations of insurance lawyers”.

At [634]-[640], the judge dealt with the non-disclosure of the NAC, holding that the claim failed for the same reasons as the alleged non-disclosure of the TPC had failed. At [638], the judge noted that the other underwriters (apart from RSA) were in no better position to allege non-disclosure of the NAC. The NAC was in the slip which all of them signed. They are therefore presumed to know about it, even if (as some of them alleged) they did not.

The judge dealt with the “as expiry” misrepresentation and the question of estoppel at [647]-[705], with Ark at [672]-[676], and with Advent at [677]-[682].

As regards Ark, the judge accepted that the underwriter, Mr Blewett, was a straightforward witness. He was an ardent note taker, who had recorded after the broke that “[a]ll rates terms and conditions [were] as expiry”, which the judge found was based on what he was told either by Mr Mullen or Mr Lockyer of Edge. Mr Blewett believed that the only material changes to the prior year’s policy were those which had been specifically discussed. That was a misrepresentation because “Ark had not been told of or given the July 2015 endorsement”. Mr Blewett did not read through the policy, for which the judge did not criticise him as a following underwriter. The judge found that there was a satisfactory evidential foundation for Ark’s case that, if the representation had not been made, Ark would not have written the policy on the terms that they did. On that basis, and on the basis that Ark had a system of peer review which would otherwise have picked up the point about the TPC covering non-physical loss, inducement was established.

As regards Advent, the judge found that Mr Cooke, the underwriter, was a good and fair-minded witness. He had noted “[a]s expiry but brokerage up from 22.5%” after the broke, and the judge found that meant that “the renewal terms were broked as being on the same terms as the expiring policy year”. The judge did not accept Mr Lockyer’s late evidence that it had been his standard practice to say that “the slip incorporated endorsements agreed during the expiring year”. There was, therefore, a misrepresentation that the Policy was “as expiry”, and the judge accepted that, if the TPC had been shown to Mr Cooke and the reasons for its inclusion explained, he would not have agreed it. Although Advent’s evidence of inducement was not as strong as Ark’s, the judge decided on a balance of probabilities that, had he known about the changes, internal discussions would have led to Mr Cooke declining to write the risk.

The judge dealt with “[t]he misrepresentations as the foundation for an estoppel” at [692]-[703]. He first recorded the Underwriters’ defence pleading that ABN Amro was prevented by estoppel and/or estoppel by convention from asserting that the Policy was on terms including the TPC. He dealt with the fact that the focus was on avoidance, the exiguous nature of submissions on the point, and Edge’s submissions that the necessary elements were not made out. But the judge said that he had nonetheless to resolve the issue “as best [he could] on the basis of the evidence adduced at trial”. He said that “[t]he potential advantage of the argument from the perspective of [Ark and Advent] is that it potentially circumvents the difficulties in their avoidance case, and in particular the effect of the NAC and affirmation”.

The judge noted at [696] that the “only ‘species’ of estoppel discussed in [the Underwriters’] closing submissions” was estoppel by convention. An estoppel by convention could arise “if (i) there is a relevant assumption of fact or law, either shared by both parties, or made by party B and acquiesced in by party A, and (ii) it would be unjust to allow party A to go back on that assumption”.

The judge then said at [697] that “[h]ere, the evidence shows that … Mr Blewett of Ark and Mr Cooke of Advent … made the assumption that the terms which they were considering were “as expiry” … That assumption was acquiesced in by Edge, and indeed was the result of positive statements made by Edge during the 2016 renewal broke”. On that basis, the question was whether it would be unjust to allow ABN Amro “whose broker acquiesced in the assumption and was responsible for making the representation which induced it, to go back on it”.

At [700], the judge said that “the question of injustice should also be considered in the light of [his] conclusion on inducement. [He accepted] that the issues on inducement consequent upon a misrepresentation and estoppel by convention [were] not quite the same. The former [depended] upon whether the policy would still have been written, on the same terms, if the representation had not been made. The latter [depended] upon whether the requirements set out in [696] [were] satisfied. However, where an underwriter has sought to rely upon avoidance based on the same representation that is alleged to give rise to an estoppel, and in that context has failed to establish inducement, it would be surprising to reach the conclusion that there was sufficient injustice to give rise to an estoppel by convention”.

Nonetheless, the judge reached different conclusions in relation to both Ark and Advent, to the effect that it was not just to allow ABN Amro to go back on the assumptions on which Ark and Advent proceeded. He concluded that ABN Amro was estopped from asserting that the Policy included cover in respect of credit risks and/or financial defaults.

In sections G and H of the judgment, the judge concluded that ABN Amro had not been reckless in its approach to taking collateral (see [750]ff), and had not been in breach of the “sue and labour” clause [852]. At [854], the judge concluded that “the claim against underwriters succeeds in full, except in relation to Ark and Advent where it fails because of the estoppel on which those underwriters can rely”.

At [861]-[940], the judge concluded that Edge was in breach of duty to ABN Amro, whether or not the TPC in fact provided the coverage which ABN Amro was seeking. The judge dealt with a number of other issues concerning Edge towards the end of his judgment. For present purposes, it is enough to record that he held that ABN Amro succeeded against Edge in respect of 100% of the liability that Ark and Advent would otherwise have borne.

In their written argument, the Appellant Underwriters submitted that certain specific elements of the factual matrix found by the judge ought to have been sufficient to displace the natural meaning of the language:

There was no precedent for marine cargo underwriters adding to a marine cargo policy cover protecting for the contractual default of a counterparty leading to a non-physical loss on cargoes which the Bank had purchased ([214] and [286]).

Underwriting and broking participants in the marine cargo market would not know the detail of the risks written by and the approach of and rating applied by specialist underwriters for trade credit risks ([217]).

The marine cargo market had no mechanism for pricing credit default or financial guarantee risks, and none of the materials by which a trade credit underwriter would assess the risk of counterparty default were provided by ABN Amro to Underwriters ([217]).

Due diligence information had been provided for the likelihood of physical loss or damage, but not for the likelihood of ABN Amro’s counterparty not delivering product of the quality it was paying for ([761]).

The TPC was carefully drafted by ABN Amro’s lawyers but the wording had not been carefully drafted or considered by Mr Beattie of RSA ([290]-[291]).

Underwriters enter their lines at the box under pressure of time and peer review does not take place until post-placement ([595] and [607]).

No additional premium was paid to the Insurers in respect of the TPC ([243]).

The Contract Certainty Code of Practice (Oct 2012) reflected best practice and required the insurer and broker to ensure that all terms of the policy were clear and unambiguous ([321]).

These arguments were overshadowed in oral argument by contentions based on sections 1, 3, 5 and 26 of the Marine Insurance Act 1906 (“MIA 1906”). Section 1 defined a marine insurance policy as one where the insurer indemnified against marine losses “that is to say the losses incident to marine adventure”. Section 3 of the MIA 1906 defined “marine adventure” by saying that there was such an adventure in three situations. First, where insurable property was exposed to maritime perils. Secondly, where the earning of any freight or other profit was endangered by the exposure of insurable property to maritime perils. Thirdly, where there was a liability to a third party by the owner of insurable property by reason of maritime perils. So, submitted Appellant Underwriters, even the second financial loss provision in the definition of marine adventure was not at large, but only applied to financial loss where property was exposed to maritime perils. Section 5 of the MIA 1906 provided that a person had an insurable interest who was interested in a marine adventure, with section 5(2) defining when that person has such an interest. Finally, section 26 of the MIA 1906 provided that the subject matter insured by a marine insurance policy must be designated with reasonable certainty. Moreover, profits arising from a sale of goods must be specifically named in the policy (see Arnould: Law of Marine Insurance and Average, 20 ^th edition, at 10-25).

The Appellant Underwriters argued that the policy has to identify the subject matter of the insurance, and the Policy does so in the definition of “Interest” which is “all goods or merchandise appertaining to the Insured’s business for which the Insured is the legal owner …”, and in the “General Conditions” that specifically says it is “applicable to all sections of this contract”. The General Conditions make clear that “[n]otwithstanding anything contained herein to the contrary, it is agreed … in general terms that it is the intention of this contract to protect the interest of the Insured at all times and in all circumstances. This contract is to protect against all risks of physical loss or damage to the Subject Matter Insured …”. Then, in section 1 of the Policy, the Limit is identified by reference only to conveyances and storage facilities, not profits or financial losses. And one finds the TPC in section 1 sandwiched between the provisions on storage risks and the provisions on basis of valuation, both of which are clauses concerning physical loss. For those reasons, the TPC had, according to the Appellant Underwriters, to be interpreted as requiring physical loss and damage even if that was not mentioned within the TPC itself.

Mr MacDonald Eggers submitted that, where there was a physical loss or damage to goods insured, which was the subject matter of the Policy, the amount recoverable was defined either by the basis of valuation provision or by the TPC. The latter would apply, where its requirements were satisfied, and, for example, ABN Amro’s client failed to reacquire goods (which had been damaged) from ABN Amro, when it sold those goods to a third party. Even a total loss might be resold. The TPC gave the insured a higher measure of indemnity. He submitted that the TPC should be given the same interpretation in section 4 where it was repeated, and that it referred in terms to the Subject Matter Insured as defined earlier in the Policy. There was no separate definition of the Subject Matter Insured in the TPC itself, in section 4 or elsewhere.

I do not intend to set out the authorities on contractual interpretation at any length. The judge did so, and nobody has suggested that his treatment was inaccurate in any way. I do, however, cite two passages that the Appellant Underwriters rely on specifically as follows.

In Wood v. Capita, Lord Hodge said this at [13]:

MacGillivray on Insurance Law (14th ed., 2018) said this about reinsurance at [35-049]:

I can deal with this single ground of appeal shortly, because, despite having looked carefully at the oral and written arguments of the Appellant Underwriters, I cannot see how their appeal can get off the ground.

I regard the judge’s decision-making on the interpretation of the TPC, which I have summarised at [23]-[33] as sound and comprehensive. He dealt with each of the Appellant Underwriters’ points, as is acknowledged by the suggestion that specific elements of the factual matrix found by the judge ought to have been sufficient to displace the natural meaning of the language of the TPC. The judge dealt exhaustively with each of those elements of the factual matrix (summarised at [24]-[26] above). He dealt with the interpretation of the TPC by applying the unitary and iterative process recommended first by Lord Mance in In re Sigma Finance Corporation [2010] 1 All ER 571 at [12], and then endorsed by Lord Clarke in Rainy Sky at [21]-[28], and by Lord Neuberger in Arnold v. Britton [2015] AC 1619 at [76]-[77].

As the judge acknowledged, elements of the factual matrix pointed against a literal interpretation of the TPC. He said at [195] that ABN Amro accepted that marine cargo insurance was normally a different class of business from credit risk insurance, and that this was an important part of the factual matrix. The problem for the Appellant Underwriters was, as the judge also said, that add-ons to standard physical loss and damage cover were common in the market, and there was no reason why such an add-on could not give protection for financial default. In considering the policy context at [232], the judge said that the clauses in the Policy relating to physical loss and damage provided a firm foundation for the principle, illustrated by Engelhart, that the starting point was that the Policy covered physical loss and damage to the cargo, unless there were clear words which provided wider cover. At [221], the judge said in terms that clear words would be needed for there to be such extended cover in the light of the factual matrix.

The question before us is whether the words of the TPC were clear as the judge held. Mr MacDonald Eggers accepted that the words were unambiguous, but suggested they were unambiguously concerned only with physical loss and damage. I cannot agree. The TPC says expressly that “Underwriters … agree that, in respect of any Transaction, it is hereby confirmed that the Insured is covered under this contract for the Transaction Premium that the Insured would otherwise have received and/or earned in the absence of a Default on the part of the Insured’s client”. These are words of coverage, not simply of a basis of valuation or measure of indemnity. Moreover, the references in the definitions included in the TPC to the Subject Matter Insured are apt, since the losses envisaged by the TPC all arise from a failure by ABN Amro’s clients to repurchase the insured goods in accordance with the agreed transactions. The provisions of the MIA 1906 relied upon by the Appellant Underwriters cannot override the clear terms of what the parties agreed. Mr MacDonald Eggers did not submit that it was unlawful to combine two types of insurance in a single policy. That is what was done by these parties in the Policy that the judge was interpreting.

I would dismiss the first appeal for the reasons the judge gave in his judgment.

As already explained, Edge made the argument on the second appeal, even though the true issue was between ABN Amro on the one hand and Ark and Advent on the other. At first sight, it is strange that, in an insurance case, it should be suggested that underwriters who signed the slip, and are bound by that slip whether they read it or not, should be able to escape liability on the grounds of an estoppel by convention or an estoppel by representation. That oddity is compounded when one knows that the same underwriters failed in all the claims they made to avoid the Policy on the grounds of misrepresentation and non-disclosure on the basis of the validity of the NAC and the judge’s holding that it bound all the Underwriters including Ark and Advent, who sought to rely on the estoppels.

The judge concluded at [697] that Ark and Advent could rely on an estoppel by convention, reasoning as follows:-

The judge decided at [701]-[705] that it would be unjust to allow ABN Amro to go back on the assumption on which Ark and Advent had proceeded because they would not have written the risks had they known of the TPC’s inclusion in the Policy. I will consider in due course how this fits with the judge’s holdings that the Policy could not be avoided for non-disclosure or misrepresentation because of the NAC.

In reply to Edge’s appeal, the Respondent Underwriters considerably clarified their case. First, they made clear that, on appeal, they maintained two species of estoppel: first, an estoppel by convention based on acquiescence rather than any common assumption, and secondly an estoppel by representation. Both cases were based on the same substantive representations made by Edge to each of Ark and Advent that the rates and terms and conditions of the Policy were “as expiry” or simply that the Policy was “as expiry”.

This change of approach meant that the first ground of appeal raised by Edge was argued as being, first, that an estoppel by convention cannot exist where the parties do not share a common assumption and are at cross purposes, and secondly that an estoppel by convention based on acquiescence can only succeed where the party said to be estopped (here ABN Amro) knows what it is said to be acquiescing in (namely that Ark and Advent were not bound by the July endorsement, or to put it another way, that Ark and Advent thought that the expiring policy did not contain either the TPC or the NAC).

I do not intend to repeat the summary of the law that is taken from the judge’s judgment and the later case of Tinkler at [37]-[40] above. In the end, it was common ground that an estoppel by convention can arise from a representation and that, in this case, it was only a single representation that was relied upon by each of Ark and Advent as founding the estoppel. Those representations were, as I have said, held by the judge, in a finding that is not appealed, to be misrepresentations (see [49]-[50] above and [673] and [680] of the judgment).

As is clear from the seminal speech of Lord Steyn in Republic of India (see [37] above and [328] of the judgment) an estoppel by convention can arise either where parties to a transaction act on a shared assumed state of facts or law or where one party to a transaction has made an assumption as to the state of facts or law and the other party acquiesces in that assumption. We are concerned here, as I have said, only with the second situation.

As Lord Burrows said in Tinkler, different types of estoppel continue to be seen as having their own particular requirements and effects [28], it made no difference to the operation of an estoppel that a common assumption had been induced by a representation [32], and a common subjective assumption is not sufficient unless a statement or conduct crosses the line [34]-[37]. I interpose that the Respondent Underwriters submitted, and I accept, that a representation made by one party to the other must be regarded as an archetypal crossing of the line. Lord Burrows also confirmed that the so-called Benchdollar principles (set out at [45] in Tinkler, at [328] in the judgment, and at [38] above), as amended by the Court of Appeal in Blindley Heath to add the need for a statement or conduct to cross the line (see [49]-[50] and [78] of Tinkler), represented the current law in both contractual and non-contractual cases. Lord Burrows commented at [64] that, in many cases, the 5^th Benchdollar principle of unconscionability was unlikely to add much where there was detrimental reliance, but that one could envisage exceptional circumstances where it would have a useful additional role to play. Edge relied particularly on this passage, because here it was said that Ark and Advent were the authors of their own misfortune by failing to read the Policy.

The Respondent Underwriters submitted that the two limbs of estoppel by convention were separate doctrines and that the requirement for a shared assumption did not apply to estoppel by convention by acquiescence. In one sense that is correct, but the real question is the second one raised under this head, namely whether acquiescence involves the party said to be estopped (here ABN Amro and its broker, Edge) knowing what it is said to be acquiescing in (namely that Ark and Advent were not bound by the July endorsement). Edge submitted that knowledge is an inevitable part of acquiescence, because one cannot acquiesce in something that one does not know about, and Ark and Advent submitted that acquiescence is an objective concept. Here, therefore, Edge represented that the Policy was “as expiry”, and Ark and Advent acquiesced in that being the case even though the parties were at cross purposes.

Two first instance cases were referred to as throwing light on this point. First in Jones v. Lydon [2021] EWHC 2321 (“ Lydon ”) (decided after Jacobs J’s judgment) Mann J said at [58] that it was common ground between the parties in that case that “[i]n the light of the fact that the acquiescence case has come to the fore it is necessary to note that in order for there to be acquiescence in the assumption of another it is necessary for the first person to know of the assumption of the other” (see also [81] of Mann J’s judgment).

Secondly, in Starbev GP Limited v. Interbrew Central European Holding BV [2014] EWHC 1311 (Comm) (“ Starbev ”), Blair J recorded at [127] the common ground there that silence could only amount to acquiescence where there was a duty to speak. There cannot have been such a duty here on the part of Ark and Advent, as underwriters, even if Edge might have had such a duty (see sections 18 and 20 of the MIA 1906, which were in force at the relevant time). Blair J said this at [132]-[133] in relation to unconscionability:

In this context, Ark and Advent submitted that, if a representation is made, the person making that representation cannot contend that the person to whom it was made was negligent in relying on it (see Cartwright on Misrepresentation, Mistake and Non-disclosure, 5^th edition, at 10-26). Of course, this is true if the representation was intended to be relied upon, but it is not directly relevant to a claim for an estoppel, save perhaps in relation to unconscionability. Ark and Advent accepted that the two sides were at cross purposes, but said that estoppel by convention was concerned with the manifestation or objective meaning of the assumption or representation. This case was at odds with other estoppel by convention cases because Edge’s representation must be taken to have been intended to be relied upon, which is why Ark and Advent were to be taken to have acquiesced in its objective meaning. Other such cases were about the duty to speak; this case was not.

Finally, Ark and Advent submitted that there were two reasons why the estoppel case was not inconsistent with the judge’s dismissal of the claim to avoid based on misrepresentation and non-disclosure. First, none of the other findings of non-disclosure or misrepresentation was inconsistent with the finding that there was a misrepresentation to each of Ark and Advent that the policy was as expiry. Secondly, the judge dismissed the case based on a duty of utmost good faith by reference to the “as expiry” representation by the application of the NAC and the Underwriters’ affirmation of the Policy. Neither the NAC on its proper interpretation, nor the Policy affirmation detracted from the establishment of an estoppel.

I agree with Ark and Advent that the judge’s finding of affirmation of the Policy is not inconsistent with the estoppel, simply because paragraph 29 of the Underwriters’ defence (see [692] of the judgment), which raised a plea of estoppel and estoppel by convention, pre-dated the affirmation found by the judge.

I do not, however, agree that the finding that there was no right to avoid the Policy based on the same misrepresentations that are said to found the estoppel is irrelevant. In short, if the NAC were applicable to prevent the Respondent Underwriters avoiding the Policy on the grounds of misrepresentation, it would be strange at least if an estoppel based on those same representations could succeed. But I agree that that must ultimately depend on the proper interpretation of the NAC, which is an issue dealt with under the next ground of appeal.

Since the Respondent Underwriters now rely on an estoppel by representation, there is perhaps less need to resolve the question of whether an objective interpretation is sufficient for an acquiescence. Nonetheless, in my judgment, both Mann J in Lydon and Blair J were right to think that an estoppel by convention based on acquiescence can only exist where the parties are subjectively in agreement; i.e. where in this case, the party making the representation knows that the other party has a different understanding. It may be noted that this approach aligns the two species of estoppel by convention, as might be expected, with the approach adopted in rectification for common mistake (see FSHC at [36] above) and unilateral mistake (see, for example, A Roberts & Co Ltd v Leicestershire County Council [1961] Ch 555 cited in FSHC at [103] and [105]) . Moreover, the understanding of acquiescence in equity goes some way towards confirming the approach: it generally requires knowledge (see Spry on Equitable Remedies, 9 ^th edition, 2014, at pages 249-250 and Re Pauling’s Settlement Trusts [1962] 1 WLR 86 at page 108).

Since, as I have said, the judge held that there were misrepresentations, the question of whether there were also common assumptions or an assumption in which the Respondent Underwriters acquiesced is not significant to the outcome. Even if, as I have suggested, Edge is right about the subjective nature of a conventional understanding and acquiescence, there is clear authority for the proposition that the meaning of a representation depends upon how a reasonable representee would understand it (see Cartwright supra at [3-06]). Plainly, the reasonable representee in the position of Ark and Advent would understand the representations that the Policy was as expiry to mean what it said, namely that the Policy was indeed on the terms of the expiring policy, which did not, in their cases, include the TPC and the NAC.

I turn then to decide whether an estoppel by representation and/or an estoppel by convention claim is contrary to the NAC.

It is important first to understand that the judge held the NAC to be included in the Policy written by Ark and Advent as I have explained. That finding was not challenged.

The NAC has two important provisions, only one of which was relied upon by the judge. The first is that “[t]he Underwriters will not: seek to avoid or repudiate this contract for non-disclosure or misrepresentation other than fraudulent non-disclosure or fraudulent misrepresentation”. The second is that “[t]he underwriters will not: … seek damages for or seek to reject a claim for loss on the grounds of: Non-disclosure or misrepresentation other than fraudulent non-disclosure or fraudulent misrepresentation”.

The judge held that the first part of the NAC meant that none of the Underwriters could avoid the Policy for the non-fraudulent “as expiry” representations as they had claimed to do.

The question now is whether the second part of the NAC providing that “[t]he underwriters will not … seek to reject a claim for loss on the grounds of … misrepresentation other than … fraudulent misrepresentation” bites on the alleged estoppel by representation and/or convention. Since Ark and Advent accept, as I have said, that the estoppel is founded on the “as expiry” misrepresentations found by the judge, it looks at first sight as if the NAC would bite.

Ark and Advent made a number of points in response. First, the NAC does not expressly preclude reliance on an estoppel. It is a common exclusion clause in the marine market, intended to apply to a non-disclosure or misrepresentation made on behalf of the insured in breach of the duty to disclose under sections 18 and 20 of the MIA 1906. The NAC needed to refer expressly to an estoppel if it were to be effective, since, as Lord Bingham had said in HIH Casualty v. Chase Manhattan Bank [2003] 2 Lloyd’s Rep 61 at [11]: “the courts should not ordinarily infer that a contracting party has given up rights that the law has conferred upon him to an extent greater than the contractual terms indicate he has chosen to do”. Exclusion clauses were to be construed strictly against those they protected, and the clearest language was required to exclude or waive legal rights. Moreover, the representations embraced both the TPC and the NAC, though the judge did not deal with the point, so that, if the estoppel succeeds as regards the TPC, it should also succeed so as to exclude the NAC.

In my judgment, however, the estoppel defence, however it is framed, depends, as I have said on the “as expiry” misrepresentations. Put simply, the clear words of the NAC say that the “[t]he underwriters will not … seek to reject a claim for loss on the grounds of … [a non-fraudulent] misrepresentation”. Ark and Advent are founding their estoppel on non-fraudulent representations, and rely on that estoppel claim in order to reject ABN Amro’s claims for loss against them. That is precisely what the NAC says they cannot do. Accordingly, construing the NAC strictly as is required, it bites on the estoppel whether it is an estoppel by convention or representation – both being based on the misrepresentations the judge found to have been made by Edge.

For completeness, I should say that, in my judgment, the judge was wrong to say at [695] that “[t]he potential advantage of the [estoppel] argument from the perspective of [Ark] and [Advent] is that it potentially circumvents the difficulties in their avoidance case, and in particular the effect of the NAC and affirmation”. The judge assumed that the NAC only prohibited avoidance for non-fraudulent misrepresentation and overlooked the fact that it also prohibits rejection of a claim for non-fraudulent misrepresentation.

That, in my judgment, concludes Edge’s appeal in its favour. Since, however, other points have been argued, I will deal briefly with them.

The judge found that it would be unconscionable for ABN Amro to resile from the assumption that had been made to the effect that the Policy did not include the TPC. That was a finding of fact that might have been difficult by itself for Edge to overturn, bearing in mind the huge amount of material and evidence put before the judge. Thankfully, it is not necessary for me to review the finding. Had I needed to do so, however, I might have put more weight on the undoubted fact, as the judge found, that neither Ark nor Advent read the Policy before agreeing to it, despite the clear law he cited that underwriters are bound by the terms of the slip to which they subscribe, whether they read it or not.

I have dealt with this question above. In my judgment, an estoppel by representation is indeed precluded by the NAC.

This question does not in the circumstances arise.

It seems to me, through there is no need to decide it, that the representations made to Ark and Advent were misrepresentations as the judge found, and must therefore have been, in theory, sufficiently clear as to found an estoppel by representation.

I have answered this question above.

As I have said at [10] above, we have decided to give judgment on the first appeal, despite the fact that the parties agreed after the argument and after this draft had been prepared that the first appeal should be dismissed.

For the reasons, I have tried to give concisely, I would anyway have dismissed the Appellant Underwriters’ appeal, and would allow Edge’s appeal on the estoppel finding made in favour of Ark and Advent.

I agree.

I also agree.