This is an appeal from a decision of Mr Jonathan Hirst QC, sitting as a Deputy High Court Judge in the Commercial Court, who in a reserved judgment given on 29 March 2006 ([2006] EWHC Comm 607), dismissed a claim by the insured claimant, who was seeking to set aside, or otherwise to re-open, on the grounds of mistake and misrepresentation, a compromise of an insurance claim it had made against its defendant underwriters.

In short, the claimant’s contention was that the policy (“the Policy”) it had taken out with the defendant was “declaration-linked”, and was not therefore subject to average, whereas the settlement (“the Settlement”) of the business interruption component of its insurance claim following a fire was based on the common mistaken assumption and/or a misrepresentation by the defendant. That assumption or misrepresentation was to the effect that the Policy was not declaration-linked, but was on the “gross profits basis” and was accordingly subject to average. As a result, the claimant agreed with, and received from, the defendant the Settlement figure of some £205,500, whereas, on the basis that the Policy was declaration-linked, the claimant should have received some £100,000 more.

The most common form of business interruption cover is written on the gross profits basis. Such a policy states a figure for gross annual profit. If that figure is less than the actual level of profit which has been lost, average is applied and the sum paid out is appropriately reduced; further, the sum paid out will never be based on a level of profit which exceeds the stated figure. In this form of cover, the annual premium is not subject to subsequent adjustment. Where a policy is declaration-linked, the assured gives an estimate of gross annual profit (in this case the estimate was £350,000), and, in the event of a loss, the sum paid out is based on the actual profit subject to a maximum percentage uplift (or “escalator”), typically 133.3%, on the estimate, and average is not applied. The premium for declaration-linked cover is adjusted once the level of profit has been established.

In the present case, the difference between the two types of policy is important because the actual level of annual gross profit being (or which would have been) earned by the claimant during the period of the interruption was substantially more than the estimated £350,000. If the Policy was on the gross profits basis, the sum payable under the policy would have been based on the estimated profit figure. On the other hand, if the Policy was declaration-linked, the sum payable would have been based on the actual profit figure, subject to a maximum of the product of the estimated figure and the escalator recorded in the policy, 1.333` (albeit that the premium would have been retrospectively subject to increase).

While the Judge accepted much of the claimant’s case in so far as it is summarised in paragraphs [2] to [4] above, he rejected its contention that it should not be bound by the Settlement, or that it should be allowed to reopen the Settlement. That contention was based on the argument that the Settlement agreement was liable to be set aside on the ground that both the claimant and the defendant were negotiating under a common mistake as to a fact fundamental to the Settlement, or on the ground that the defendant had negligently misrepresented to the claimant a fact fundamental to the Settlement, namely that the Policy was not declaration-linked, whereas, in fact, it was.

The relevant facts are as follows. The claimant operates a night club. On 10 September 2001, its broker, a Mr Thomas of IRCS, approached the defendant’s agent, SK Underwriting Ltd (“SK”), with a request for cover on a non-average declaration-linked basis. According to the findings of the Judge (on fact, construction and rectification, which are, in my view quite rightly, not appealed), Mr Ing of SK did not intend the Policy to be written on such a basis, but it was in fact so written.

On 14 November 2001, a serious fire occurred at the club, and the claimant raised a claim for, inter alia, business interruption. Carr Greenwood Smith (“CGS”) acted for the defendant as their loss adjusters, and the claimant appointed Harris Claims Group plc (“Harris”) to act as its loss assessors.

It appears that Harris did not have a copy of the Policy, and asked both CGS and IRCS for a copy, which was not supplied by either entity. However, in early March, Harris were supplied with a poor, but legible, faxed copy of the schedule to the Policy (“the schedule”), which in terms recorded the sum insured as “declaration linked”. On 12 March 2002, Mr Lawrence of Harris wrote to the claimant, saying that, “if this was the case, it would be extremely advantageous to your Business Interruption Claim”. He also told the claimant that Mr Stafford of CGS was “surprised at this wording”, and that he had asked Mr Stafford to discuss it with the defendant underwriters.

On 20 March 2002, Mr Lawrence wrote to Mr Stafford pointing out what was recorded in the schedule, and inviting him again to “confirm and clarify this with your principals”. On 4 April 2002, Mr Stafford replied stating that, although the words relied on “would ordinarily” indicate that the policy was declaration-linked, in this Policy, this was not so in this case as “the item insured is Loss of Profit rather than Estimated Gross Profit”. This letter was passed on to Mr Dymant of Harris, who did not understand why, and pressed Mr Stafford for a further explanation.

Mr Stafford then telephoned Mr Ing on 24 April, who told him that the Policy was not declaration-linked. Mr Stafford’s note of the conversation records Mr Ing also saying that “no escalator was asked for or agreed”. (In fact, the standard form policy referred to a maximum escalator of 133.33%. This could have been a reference to the maximum escalator which had been agreed as claimable on the particular declaration-linked policy, which is how the Judge construed it or it could have been a reference to the maximum level of escalator which the underwriters would be prepared to contemplate on any declaration-linked policy, on the basis that the level remained to be agreed for any such particular policy, and had not been agreed here because the policy was not declaration-linked, which is how Mr Ing construed it.)

Mr Ing repeated his view that the Policy was not declaration-linked, to Mr Stafford in a further telephone conversation on 3 May. This view was then relayed by Mr Stafford to Mr Dymant first by telephone and then in a letter of 9 May 2002. In that letter, Mr Stafford said that the defendant underwriters “confirm that this is not a declaration linked sum insured. The sum insured of £350,000 is subject to the average clause detailed in the policy and any payment made will be proportionately reduced”.

This was apparently accepted, at least for the purpose of the negotiations, by Mr Dymant, who wrote to Mr Williams of his client, the claimant, on 16 May 2002, saying that “it would appear that the insurers are saying that this is…subject to average which, of course, greatly reduces our claim”. However, the letter immediately went on to explain that “[t]here was some uncertainty about this” in view of what was stated in the schedule, and to raise questions as to the intentions when the policy was taken out. The letter invited Mr Williams to telephone Mr Dymant “to discuss this point”.

Negotiations then proceeded on the basis that the Policy was not declaration-linked, and the final Settlement of the claimant’s business interruption claim was negotiated between Harris and CGS on 11 March 2003, in the sum of £205,511.76, which was duly paid to the claimant.

Thereafter, pursuant to the efforts of Mr Williams, the claimant obtained sight of IRCS’s file, and appreciated that declaration-linked cover had been expressly sought and given. Accordingly, after some further discussion, the claimant began the instant proceedings on 24 January 2005, alleging that the Settlement agreement was vitiated by common mistake, or negligent misrepresentation on the part of the defendant (through CGS), as to the effect of the policy. It was agreed in the statements of case that, on the basis that the Policy was declaration-linked, the claimant should have received another £108,319.56 (less the additional premium payable on the basis that this was a declaration-linked policy and the actual profit exceeded the estimate).

The issues before the Judge were more extensive than those before us. He had to consider whether the Policy was indeed declaration-linked as a matter of construction, which he decided that it was, and, if so, whether it should be rectified, which he decided that it should not. On the issue of mistake, he accepted that the Settlement had been entered into on the basis of a common mistake, but held that the nature of the mistake was such that the claimant was not entitled to contend that the Settlement agreement was thereby vitiated. As to misrepresentation, he held that the claimant’s case failed on the grounds that (a) there was no representation by Mr Stafford on behalf of the defendant as to the Policy not being declaration-linked, on the basis that he had merely raised a contention or argument to that effect, and (b) in any event, the alleged misrepresentation did not induce the claimant (through Mr Dymant) to enter into the Settlement.

So far as the evidence was concerned, the Judge said that the witnesses who gave evidence to him, namely Mr Williams of the claimant, Mr Ing of SK, Mr Dymant of Harris, and Mr Stafford of CGS, were “all honest and doing their best to assist the Court”. He also had a witness statement from Mr Thomas of IRCS, whom he found to be unreliable in one respect, namely on the issue of whether he had ever received a copy of the schedule. On the Judge’s finding, he had apparently received it around 22 November 2001.

The Judge also found that, although they were plainly wrong in their view, Mr Ing and Mr Stafford were both telling the truth as they saw it when they stated in April and May 2002 that the Policy was not declaration-linked. Mr Ing had not intended to write a declaration-linked policy when he agreed the proposal put to him by Mr Thomas in September 2001, and believed that he had given effect to his intention. The Judge also concluded that Mr Stafford believed Mr Ing.

The Judge further found that Mr Stafford had not concealed the Policy from Mr Dymant; and that he had “proceed[ed] on the fair assumption that Mr Dymant was fully briefed, and that he had a copy of the standard policy terms”. Indeed, the Judge thought that Mr Stafford “would have been incredulous if he had been told that Mr Dymant, a professional loss assessor working for a well known firm, was proceeding without knowledge of the full policy terms”. The Judge concluded that, although he had written to Mr Thomas and Mr Stafford seeking a copy of the Policy and had only received a copy of the schedule, Mr Dymant had “made no real effort to obtain a full copy of the Policy or to investigate with the brokers whether they considered the underwriters to be correct, or to examine the broker’s placing file”.

Like the Judge, I propose first to consider the claimant’s case on mistake, and then to turn to its case on misrepresentation.

On this issue, the facts are simple and were not in dispute before the Judge. The settlement for the claimant’s business interruption claim against the defendant underwriters was settled at about £205,000 on the common assumption that the Policy was not declaration-linked, whereas it was so linked, and, had the parties been aware of this, they would (I assume for present purposes) have settled at a figure about 50% higher.

As the Judge said, the leading modern case in which the circumstances in which a common mistake can vitiate a contract were considered was the decision of this court in Great Peace Shipping Ltd –v-Ttsavliris Salvage (International) Ltd [2002] EWCA Civ 1407, [2003] QB 679. Relying on what Lord Phillips of Worth Matravers MR (giving the judgment of the court) said at paragraphs [73] to [76], the Judge held that the proper test to apply in this case was whether the mistake in question rendered the contract in issue “impossible of performance” (the expression also used by Lord Phillips when ultimately formulating the critical question in the Great Peace case itself at paragraph [162]). At least on the face of it, it seems difficult to quarrel with the Judge’s view that, if that is the right test, it was self-evidently not satisfied here.

Mr Butler, who appears for the claimant, runs as his main argument the contention that this test was inappropriate in a case such as this; his alternative argument is that the test, if properly applied, was in any event satisfied here. I should in this context refer to the decision of this court in Brennan –v- Bolt Burdon [2004] EWCA Civ 1017, [2005] QB 303. In that case, a personal injuries action was settled on the common assumption that the claim form had been served out of time, and a subsequent decision of this court showed that that assumption was wrong. The claimant unsuccessfully sought to impeach the settlement.

In paragraph [22] of his judgment, Maurice Kay LJ gave three reasons why the Great Peace decision gave rise to difficulties for the claimant, the first of which was that it was “quite simply not a case of impossibility of performance. The compromise has at all times remained performable…”. Sedley LJ, however, was more concerned about the application of the “impossibility of performance” test in cases of common mistake of law – see at paragraphs [60] to [61]. At the end of paragraph [59], he had identified the “difficulty…in seeing how the effect of [a common mistake of law…on an agreement by which litigation is compromised] can be equiparated with the impossibility of a contractual venture”. The third member of the court, Bodey J, did not discuss this aspect.

In my opinion, it is unnecessary for us on this appeal to decide which view is preferable. Indeed, I suspect that ultimately, the two approaches may essentially amount to the same thing. If the doubts of Sedley LJ are justified, then, as Mr Butler argues, the right test is that propounded by Steyn J in Associated Japanese Bank (International) Ltd –v- Credit du Nord SA [1989] 1 WLR 255. In a passage at p. 268F, cited and expressly approved in the Great Peace case at paragraphs [90] and [91], Steyn J said that, in order to vitiate a contract, “the mistake must render the subject matter of the contract essentially and radically different from the subject matter which the parties believed to exist”. He justified this at p. 268E on the basis that “the law ought to uphold rather than destroy apparent contracts”.

It appears to me that, by approving Steyn J’s observations and by applying the “impossible to perform” test, this court in the Great Peace case must have considered that the two approaches amounted to much the same thing. In practice, the concept of impossibility of performance, at least in a case such as this, can be said to raise an issue of definition: if one defines the contract as the assessment of compensation under a declaration-linked policy, then it is, at least in a sense, impossible to perform if both parties negotiate on the basis that the policy is not declaration-linked. It seems to me, therefore, that there is much to be said for applying, as Mr Butler argues we should, Steyn J’s test in the Associated Japanese Bank case in the instant case.

In my judgment, applying that approach, the mistake in this case did not render what the parties believed to be the “subject matter of the [Settlement agreement] essentially and radically different” from what it actually was. The parties correctly believed that they were settling a business interruption claim resulting from a fire at certain premises at which the claimant ran a night club; they made no mistake as to the period of interruption or the estimated level of gross profit, or indeed any other mistake about the claim or the nature of the cover, save that they assumed that it was on the gross profits basis, rather than on the declaration-linked basis. The difference between the actual and assumed subject matter of the settlement can in my view certainly be characterised as significant, but it is not an “essential…and radical…” difference.

I suspect that it is normally not easy to say precisely why a difference such as there is in this case is not, or indeed is, radical and essential. If that is right, this case is no exception to the norm. However, it seems to me that the following factors strongly drive one to the conclusion that the difference in this case was not radical or essential. In conceptual terms, once one appreciates what was correctly assumed or agreed (as discussed in the preceding paragraph), it is hard to say that if one corrects the one aspect which was wrongly assumed, it would radically and essentially alter the nature of the contract. What was wrongly assumed was a detail, albeit a significant detail, of the basis on which the Policy was written: it did not go to the validity of the Policy, the parties, the property or nature of the business, or even the nature of the risks covered. In addition, if it is appropriate to look at the matter in commercial terms (as I believe it is in this case at any rate), although the claimant received some 33% less than it should have done, which is a significant, even a substantial, reduction on its entitlement, I do not think it can fairly be characterised as an “essentially or radically” different sum from its entitlement.

Accordingly, I think the Judge was right to dismiss the claim in so far as it was based on common mistake. I would leave open the question whether it would, in any event, be right to hold the settlement vitiated by mistake, given that the question of whether the Policy was declaration-linked had been raised by the claimant and considered by the parties in the very negotiations which resulted in the Settlement which is now sought to be impeached on the ground of a common mistake that it was not declaration-linked.

Before the Judge, the misrepresentation issue was more contentious factually. However, in the light of his primary findings of fact (especially Mr Ing’s and Mr Stafford’s honest belief that the Policy was not declaration-linked , and Mr Stafford’s reasonable belief that Mr Dymant had a copy of the Policy), he concluded that the alleged misrepresentation was really a contention or statement of position on behalf of the defendant underwriters, which gave the claimant no cause of action in misrepresentation against the defendant.

Mr Butler attacks this conclusion on a number of grounds. First, he says that the Judge was wrong to consider the statement by reference to a test postulated by Mance LJ in paragraph [30] in Primus Telecommunications PLC –v- MCI Worldcom International Inc. [2004] EWCA Civ 957, where this court reversed the grant of summary judgment. In that connection, the Judge said this:

The approach as so described by the Judge appears to me to be right in principle. It is true that, in paragraph [30], Mance LJ expressed himself somewhat diffidently in the Primus case, but that was because it was unnecessary to state a firm conclusion (as the matter was to go to trial); however, in my judgment his approach was correct in principle. The fact that one is here concerned with a unilateral statement made by one party to another, rather than a statement in a bilateral document does not justify a different approach to interpretation. If there was any doubt about this, it would be put to rest by the reasoning in Mannai Investment Co Ltd –v- Eagle Star Life Assurance Ltd [1997] AC 749, where the established approach to interpretation of contracts was plainly treated as equally applicable to the interpretation of unilateral notices.

Mr Butler’s next point is that the consequences of a false statement which is acted on should lie with the person who makes the statement, at least to the extent of the onus being on him to show why it should not be an actionable misrepresentation. That appears to me to be an assertion which does not reflect the law. The question whether a statement can amount to a representation capable of founding a claim in misrepresentation must depend on the words involved in the context in which they were so used. Until one has decided that, in the circumstances of the particular case, the statement should be treated as a representation rather than, say, a contention or argument, any question of liability for misrepresentation cannot arise.

Mr Butler’s final point on this aspect was that, even if the Judge was right in the approach he adopted, the letter of 9 May 2002 could not have properly been interpreted as anything other than a representation in all the circumstances. I appreciate that, as a matter of linguistics, the letter of 9 May 2002 can be said to be purporting to state facts (in particular in the light of the verbs used – “confirm”, “is”, and “will be”). However, read in the context which the Judge considered (after applying the correct principles) to be appropriate, I think that, to put it at its lowest, the Judge was entitled to conclude that it should be read as a contention, not as a representation. As has been said (albeit perhaps a little extravagantly) in relation to issues of interpretation, context is all. Thus, barristers, when making submissions in court, frequently express themselves in the form of unqualified positive averments: that does not change the characterisation of how they will and should be understood by a judge, namely as making submissions.

In this case, the defendant insurer’s position as to the meaning of the relevant insurance policy was relayed to Mr Dymant, an experienced and professional loss assessor, acting for the insured claimant. He had the schedule, which itself should have made the matter pretty clear, and Mr Stafford was entitled to assume that he had a copy of the Policy, which would have rendered it even clearer (especially to an experienced loss assessor). Mr Dymant could, indeed would, reasonably have been expected to form his own informed view and to advise his client accordingly. Mr Dymant’s position was, as I see it, little different from that of a solicitor acting for an insured, to whom an interpretation of the relevant policy was forcefully put by or on behalf of the insurer.

In this connection, the decision of Romer J in Wauton –v- Coppard [1899] 1 Ch 92 was relied on by Mr Butler. I accept that that case assists the claimant to the extent that it shows that a statement as to the meaning or effect of a document can amount to an actionable misrepresentation. However, the decision, in my judgment, is readily distinguishable from this case. It involved a statement by a vendor’s agent as to the effect of a restrictive covenant to a lay person who, as a prospective purchaser, did not (to the knowledge of the vendor’s agent) have a copy of the covenant. Here, the statement was as to the meaning of a policy to an agent of the claimant who had relevant professional qualifications and experience, and who had, or at least would reasonably have been taken to have, a copy of the policy.

The fact that Mr Stafford, as a loss adjuster, may have owed a duty to both the claimant and the defendant does not alter my view. It was clear that he was relaying to Mr Dymant (and hence to the claimant) the view of the defendant. The fact that it may have been (and it must be emphasised that I am not suggesting that it was) surprising that Mr Stafford accepted what the defendant said is nothing to the point.

Mr Butler also contended that the Judge gave insufficient weight to the fact that Mr Dymant had requested Mr Stafford to provide him with a copy of the standard terms (i.e. of the full terms of the Policy) and had not been provided with it. As I understand it, this contention is intended to excuse Mr Dymant from not knowing the terms of the Policy and to show that Mr Stafford was not entitled to assume that Mr Dymant knew those terms. However, the contention does not meet the points, first, that Mr Dymant had a copy of the schedule, and, secondly, that the Judge found (and was plainly entitled to find) that Mr Dymant could, should, and would have been expected to, try much harder to get such a copy, and would have been able to do so. In connection with this second point, Mr Stafford explained, to the Judge’s (reasonable) satisfaction, that Mr Dymant’s initial request was a standard one, which he did not regard as significant and which was not, as far as he recalled, repeated during his “very many meetings with Harrises on this claim”; he also said that “Harrises could have accessed both the policy document and the schedule, and they should have”.

I turn to the second ground upon which the Judge rejected the case on misrepresentation, namely lack of reliance. On the facts stated so far, that may seem a surprising conclusion. However, it was primarily based on a passage in the cross-examination of Mr Dymant by Mr Reed, who appeared (and appears) for the defendant. Following Mr Dymant’s invitation in his letter of 16 May 2002 to Mr Williams to telephone him to discuss the basis of the cover, he said in his witness statement that there was such a conversation. In that conversation, Mr Dymant said that he advised his client to go along with the notion that the policy was not declaration-linked “for the present”.

When cross-examined about this, Mr Dymant explained that he would have liked to have had sight of the broker’s file “to see what the true position was”, and that what was in his mind at the time of the Settlement, was “that there was probably a case against the broker” as well as a “very very small chance of a claim against the underwriters”.

In a passage in the judgment which is rather compressed (perhaps because it was a secondary reason for rejecting the misrepresentation claim), the Judge held that this evidence showed that Mr Dymant “treated the Underwriters’ contentions as exactly that” (by which he clearly meant “as contentions, not as representations”). The Judge went on to say that Mr Dymant “decided for reasons that I have not managed to fathom that the claim should be settled on the basis proposed by the Underwriters rather than investigating the point further, as he had ample opportunity to do”.

At any rate on the unusual facts of this case, I consider that the Judge was entitled to reach the conclusion that there was no inducement in this case. I can see the force of the point that the statements as to the effect of the Policy, as relayed by Mr Stafford, did induce the Settlement, in the sense that they were a causative factor in the claimant and Mr Dymant agreeing on a settlement figure of around £200,000, rather than around £300,000.

However, I think the Judge effectively found that Mr Dymant merely treated the statements as contentions which he decide to assume were right for the purposes of the Settlement, and which he would thereafter investigate (as he did) with a view to obtaining compensation for the claimant. In effect, he chose to treat them as correct for the time being, without necessarily believing that they were accurate, and with the intention of investigating their accuracy and ramifications later. As already indicated, the facts of the present case appear to me to be very unusual, mainly, as the Judge said, because the approach of Mr Dymant appears mystifying. On the unusual facts, I think the Judge’s second reason was, at the very least, open to him.

On that basis, the second reason for dismissing the misrepresentation claim becomes congruent with the first. The first reason is that the statements, viewed objectively, were contentions and not representations; the second is that the statements were subjectively understood by the person to whom they were made as contentions and not representations.

The defendant lodged a respondent’s notice with a view to supporting the judgment on additional grounds. Having considered the transcripts of the evidence, Mr Reed, very sensibly in my opinion, decided to abandon those grounds. I say no more about them.

In all the circumstances, despite Mr Butler’s well and economically sustained arguments, I would, for the reasons given above, dismiss the appeal.

I agree.

I also agree.