Credit Suisse Life (Bermuda) Ltd v Bidzina Ivanishvili and 6 others (Bermuda) [2025] UKPC 53

[2025] UKPC 53

JUDGMENT

Credit Suisse Life (Bermuda) Ltd (Appellant) v Bidzina Ivanishvili and 6 others (Respondents) (Bermuda); Credit Suisse Life (Bermuda) Ltd (Respondent) v Bidzina Ivanishvili and 2 others (Appellants) No 2 (Bermuda)

From the Court of Appeal for Bermuda

before Lord Hodge Lord Briggs Lord Leggatt Lord Richards Lady Simler

Hearing JUDGMENT GIVEN ON 24 November 2025 Heard on 16, 17, 18 and 19 June 2025

Appellant/Respondent to Cross-Appeal Lord Falconer Sebastian Isaac KC Ben Waistell Grant Kynaston (Instructed by Gibson, Dunn & Crutcher UK LLP (London) and Walkers (Bermuda) Limited (Bermuda))

Respondents/Cross-Appellants Richard Morgan KC Louise Hutton KC Felix Wardle Akash Sonecha (Instructed by Blake Morgan LLP (London) and ASW Law Limited (Bermuda))

Between

Credit Suisse Life (Bermuda) LtdAppellant

Bidzina Ivanishvili and 6 othersRespondent

Date 2025-11-24

Introduction

[1]The former Prime Minister of Georgia, Mr Bidzina Ivanishvili, is a very wealthy businessman. In 2005 he formed a private banking relationship with Credit Suisse AG (“the Bank”). In 2011 and 2012, on the Bank’s advice, Mr Ivanishvili transferred over US$750 million, held on trusts for the benefit of himself, his wife and children, to Credit Suisse Life (Bermuda) Ltd (“CS Life”) as premiums under two life insurance policies. CS Life is a Bermuda insurance company which was a wholly owned subsidiary of the Bank. The policies were structured so that the premium was held in a segregated account of CS Life with the Bank and was to be invested on either a discretionary or nondiscretionary basis, according to the policyholder’s choice.

[2]In September 2015 Mr Ivanishvili discovered that his relationship manager at the Bank, Mr Patrice Lescaudron, had been dealing fraudulently with the policy assets. As established in these proceedings, Mr Lescaudron’s fraudulent conduct included misappropriating assets, transferring assets from the policy accounts to those of unrelated clients, transferring assets into the policy accounts at an overvalue to hide losses of those unrelated clients, and enriching himself by making investments of policy assets on which he received secret commissions. Following criminal complaints by Mr Ivanishvili, the Bank and others, Mr Lescaudron was prosecuted in Switzerland. In February 2018 he was convicted of offences of fraud, mismanagement, aggravated mismanagement and forgery and was sentenced to five years’ imprisonment. He later committed suicide.

[3]Mr Ivanishvili, members of his family and two companies which are the named policyholders began these proceedings against CS Life in August 2017, claiming damages for breach of contractual and fiduciary duties. In October 2020 they also claimed damages for fraudulent misrepresentation. The trial took place in November and December 2021 before Chief Justice Hargun, whose clear and comprehensive judgment was delivered on 29 March 2022: [2022] SC (Bda) 19 Civ.

[4]The Chief Justice decided almost every contested point in favour of the plaintiffs. CS Life’s case was not assisted by its conduct of the litigation. The company adopted what might be described as a “flat earth” defence of refusing even to admit Mr Lescaudron’s fraudulent conduct – even though by the time of the trial Mr Lescaudron had himself admitted it and been convicted of criminal offences on the basis of it. CS Life also failed to disclose – and was found to have deliberately withheld – many highly relevant documents. Vast numbers of documents were disclosed only on the eve of and during the trial, so that the plaintiffs had no reasonable opportunity to consider their relevance and impact on the case. These included documents relating to Mr Lescaudron’s fraud and documents which showed that CS Life had an asset monitoring department – a fact which CS Life had previously concealed. CS Life chose not to call witnesses who were involved in managing CS Life at the material time and/or who had knowledge of Mr Lescaudron’s fraud. The Chief Justice quite properly drew adverse inferences from this conduct.

[5]The judgment contains 31 specific factual findings. Based on these findings, the Chief Justice held that CS Life was in breach of contractual and fiduciary duties owed to the plaintiffs. He also concluded that the plaintiffs had been induced to enter into the policies by fraudulent misrepresentations. The Chief Justice awarded damages aimed at putting the plaintiffs in the same position financially as if the policy assets had not been fraudulently mismanaged by Mr Lescaudron and had instead been professionally managed.

[6]CS Life appealed to the Court of Appeal. The Court of Appeal (Sir Christopher Clarke P, Bell JA and Smellie JA) dismissed the appeal in relation to the claims for breach of contract and fiduciary duty but allowed the appeal in relation to the misrepresentation claim: [2023] CA (Bda) 13 Civ.

[7]From that decision CS Life brings this further appeal to the Board as of right, arguing that the Court of Appeal was wrong to uphold the award of damages for breach of contract and fiduciary duty. The plaintiffs have cross-appealed against the dismissal of their misrepresentation claim.

[8]All the relevant findings of fact made by the Chief Justice were affirmed by the Court of Appeal. It is the long-standing practice of the Board not to review concurrent findings of fact made by two courts below, save in special circumstances: see Devi v Roy

[1946] AC 508

; Sancus Financial Holdings Ltd v Holm (Practice Note)

[2022] UKPC 41

;

[2022] 1 WLR 5181

. CS Life has not argued that there are any special circumstances which would justify departure from this practice here and accepts that it is bound by all concurrent factual findings made by the courts below. II. THE APPEAL: BREACH OF CONTRACT AND FIDUCIARY DUTY

[9]To frame the issues in CS Life’s appeal, it is necessary to give more details of the two life insurance policies to which the plaintiffs’ claims relate.

The formation of the policies

[10]In 2005, on the advice and with the help of the Bank, Mr Ivanishvili founded the Mandalay Trust. Its beneficiaries are Mr Ivanishvili and his wife and children. Mr Ivanishvili transferred approximately $1.1 billion to the Mandalay Trust, which was managed for him by the Bank. From 2006, his relationship manager was Mr Lescaudron.

[11]At a meeting in Tbilisi, Georgia in March 2011, Mr Lescaudron proposed to Mr Ivanishvili that assets of the Mandalay Trust should be invested in a type of life insurance policy provided by CS Life called a “life portfolio international” (or “LPI”) policy. This financial product was marketed on the basis that it allowed clients to take advantage of the benefits of an offshore life insurance policy, in terms of wealth and inheritance planning, without changing the way their assets were managed.

[12]Mr Ivanishvili agreed to the proposal. He applied for a LPI policy on behalf of Meadowsweet Assets Ltd (“Meadowsweet”), a company owned by the trustee of the Mandalay Trust. CS Life issued the “Meadowsweet policy” on 7 November 2011. Meadowsweet was the policyholder. The insured person was Mr Ivanishvili. Most of the single premium of some US$480 million consisted of assets already held by Meadowsweet in accounts with the Bank which were transferred to an account of CS Life; the balance came from Mr Ivanishvili’s personal accounts with the Bank. The terms of the Meadowsweet policy are contained in the application form, the policy schedule and CS Life’s General Policy Conditions (the “General Conditions”). The General Conditions provide that the laws of Bermuda govern the policy and that the exclusive jurisdiction for all disputes between the parties in connection with the policy shall be Bermuda.

[13]In 2011 Mr Ivanishvili decided to sell his business interests in Russia; and at a meeting in Tbilisi in June 2012 Mr Lescaudron proposed that Mr Ivanishvili should invest some of the proceeds of sale in a second LPI policy with CS Life. Mr Ivanishvili agreed to do so. The policy was issued on 7 December 2012. The policyholder was Sandcay Investments Ltd (“Sandcay”), a company owned by the trustee of another family trust named the Green Vals Trust set up on the advice and with the help of the Bank. Almost all the single premium of some US$275 million was funded by a cash transfer. The terms of the “Sandcay policy” are materially similar to those of the Meadowsweet policy and are again contained in the application form, the policy schedule and the General Conditions.

The structure of the policies

[14]The LPI policies issued by CS Life are not conventional life insurance policies. They are in essence investment funds or asset portfolios managed by the Bank for the benefit of the policyholder within the framework of a life policy. The premium is a single premium which can be paid in cash or by transfer of assets to CS Life. All such cash and assets are allocated to a separate account of CS Life, which is a “segregated account” for the purpose of the Bermuda Segregated Accounts Companies Act 2000 (the “SAC Act”). The SAC Act requires the assets linked to a segregated account to be held in a separate fund which is only available to meet the claims of the policyholder and creditors of the segregated account and is not available to meet the obligations of the company to its shareholders or to those creditors whose claims are not linked to the segregated account. In the LPI policies this fund is referred to as “the internal fund”.

[15]The policyholder has a choice between two investment alternatives. If the investment alternative “with discretionary mandate” is chosen, the choice of investments is delegated to the Bank. At all relevant times this function was performed by a team within the asset management division of the Bank known as the MACS (Multi Asset Class Solutions) team. If the investment alternative “without discretionary mandate” is chosen, the policyholder chooses the investments, within certain limits set out in the General Conditions.

[16]The sum payable to the policyholder on the death of the insured person is the value of the internal fund at the time of death. The policyholder may also withdraw sums by a partial or total surrender of the policy at any time.

Policy terms

[17]The structure outlined above is reflected in the following relevant policy terms.

[18]The application form states in section B:

“The Life Portfolio International policy to be issued by [CS Life] will be linked to its own segregated account. This segregated account is a separate and distinct account of [CS Life] pertaining to an identified or identifiable pool of assets and liabilities which are separated, segregated or distinguished from other assets and liabilities of [CS Life].”

[19]Section C of application form, labelled “Investment alternative”, states:

“The single premium will be invested in an internal fund as stated below (which is invested separately to the insurance company’s other assets). More information about the alternative profile can be found in the ‘Description of the internal fund & asset management’ which is an integral part of the application.”

The “General Description of the Internal Fund & Asset Management” here referred to (which was part of the application form) includes the following:

“The Internal Fund The internal fund is invested separately from the insurance company’s other assets and managed according to the investment alternative chosen in the application. The internal fund is managed by [the Bank]. Asset Management with or without Discretionary Mandates [CS Life] will invest the insurance premium according to the investment alternative agreed with the policyholder. The policyholder may request at any time that [CS Life] change his/her investment alternative.”

Asset Management with or without Discretionary Mandates

[20]The preamble to the General Conditions states:

“Life Portfolio International is a life insurance policy (‘Policy’) that combines life insurance coverage with an investment in an internal fund of [CS Life]. … Each Policy to be issued by [CS Life] will be linked to its own segregated account. The Policy is based on a contract between the policyholder and [CS Life] as the insurance company.”

[21]Clause 1 of the General Conditions includes the following definitions:

“The internal fund Consists of the integrated assets invested separately from the other assets of [CS Life], in accordance with the investment alternative chosen by the policyholder and is linked to the segregated account in respect of the Policy. … The investment alternative The policyholder chooses an investment alternative to match his/her investment goals and risk tolerance. For discretionary mandates, the portfolio is managed according to the current investment policy of the custodian bank and in line with the guidelines relating to discretionary mandates issued by the Swiss Bankers Association. The insurance premium Compensation to [CS Life] for assuming the insurance coverage and acquiring the investments. … The custodian bank The bank managing the assets in the internal fund.”

The custodian bank

[22]Clauses 5 and 6 of the General Conditions provide:

“5) The insurance premium Life Portfolio International is a single premium Policy. … The payment is invested on the commencement date of the Policy according to the investment alternative referred to under paragraph 7 below. … The insurance premium can be paid in cash or by transfer of assets. If the premium is paid by transfer of assets, [CS Life] has an absolute discretion in deciding whether or not to accept the transfer of assets as a premium payment. … 6) Use of the premium The invested capital consists of the premium after deduction of any upfront-insurance fees or deductions. The net single premium and the net additional premium (if any) are invested in the internal fund in accordance with the investment alternative indicated in the application form and as set out in the Policy and any Policy addendum. …”

[23]Also relevant are the opening words of clause 7:

“7) The investment alternative and content of the internal fund The policyholder may choose an investment alternative with or without discretionary mandates. The investment alternative without discretionary mandate may comprise of investment funds, structured investments, direct investments and fiduciary deposits.”

The clause then specifies investments which the policyholder is not permitted to make when the investment alternative without discretionary mandate is chosen.

The findings of breach of contract and fiduciary duty

[24]On the plaintiffs’ claim for damages for breach of contract, the Chief Justice made the following key findings:(1) Under the terms of the policies, CS Life promised to invest the policy assets in accordance with the “investment alternative” chosen by the policyholder.

(2) For both policies, the investment alternative chosen was discretionary management by the Bank.

(3) The policy assets never were managed by the Bank but instead were invested or otherwise dealt with by Mr Lescaudron.

(4) Mr Lescaudron committed a long-running fraud involving the policy accounts, which included: (i) making investment decisions without proper authority; (ii) forging documents; (iii) executing investments for the purpose of obtaining unlawful commissions; (iv) directing the sale of assets at an undervalue; (v) directing the purchase of securities at an overvalue; and (vi) transferring assets to his other clients.

[25]As well as an obligation to invest the policy assets in accordance with the investment alternative chosen by the policyholder, the Chief Justice held that CS Life owed, and was in breach of, further contractual obligations. These were:(1) an implied duty to carry out the services provided under the policies with reasonable care and skill;

(2) an implied duty to monitor the investment of the policy assets for fraud and check that the chosen investment alternative was being complied with; and

(3) an implied duty to maintain accurate policy records.

[26]The Chief Justice also held that CS Life owed fiduciary duties to the policyholders to act in their best interests, safeguard the policy assets and account accurately to them, and that CS Life was in breach of these duties.

[27]The Chief Justice held that damages should be calculated as the difference between the actual value of the policy assets and what that value would have been if the assets had been managed by a reputable European bank on a discretionary basis (this being the best approximation to what the value would have been if the policy assets had been managed by the Bank’s MACS team).

[28]All these findings and conclusions were affirmed by the Court of Appeal.

CS Life’s grounds of appeal

[29]CS Life advances six grounds of appeal. Three contest the conclusions of the courts below as to the contractual and fiduciary duties owed by CS Life and three relate to the quantification of damages for losses caused by breach of those duties. The principal ground of appeal is the first, which concerns the scope of CS Life’s contractual duties.

Ground 1: CS Life’s contractual duties

[30]CS Life asks the Board to find:(1) that on the correct interpretation of the policies CS Life did not owe any of the contractual obligations to the policyholder found by the courts below; and

(2) that the courts below also erred in law in finding that the investment alternative chosen by the plaintiffs was a discretionary mandate. A.

The nature of the investment duty

[31]In addressing the scope of CS Life’s contractual obligations, most of the argument before the Board has focused on the key finding that CS Life undertook to invest the policy assets in accordance with the investment alternative chosen by the policyholder. The Chief Justice held (paras 447-452) that this obligation is imposed by clause 6 of the General Conditions (quoted at para 22 above) and the terms of the application form quoted at para 19 above. The Court of Appeal agreed (paras 219 and 308-309).

[32]On this appeal CS Life has not maintained an argument made below that the language of those provisions which states that premiums “are invested” or “will be invested” is merely descriptive. Leading counsel for CS Life, Lord Falconer, accepted that the language used is properly read as promissory and as expressing an obligation undertaken by CS Life. But what CS Life undertook to do, he submitted, was only(1) to pass the premium received from the policyholder to the Bank by placing it in the internal fund and

(2) to communicate to the Bank the policyholder’s choice of investment alternative. Managing the assets in the internal fund in accordance with the chosen investment alternative was solely the Bank’s responsibility. CS Life had no obligation to ensure that the assets were invested by the Bank in accordance with that choice.

[33]The Board is not persuaded that this is the correct interpretation of the policy wording. Lord Falconer focused on the words of clause 6 of the General Conditions which state that premiums “are invested in the internal fund in accordance with the investment alternative …”. (See also section C of the application form quoted at para 19 above: “The single premium will be invested in an internal fund …”). He submitted that these words show that the premium has to be put in the internal fund by CS Life; only after this has been done is the internal fund itself invested.

[34]In the Board’s opinion, this argument puts a weight on the phrase “invested in the internal fund” which it cannot bear. It does not make sense to take these words to mean that premium paid in cash or by transfer of assets to CS Life is then the subject of a further transfer by CS Life to the internal fund. Such an interpretation is inconsistent with the scheme of the policies.

[35]It is clear from the definition of “the internal fund” and other policy terms quoted above that the phrase “the internal fund” is simply a label for the portfolio or pool of assets which are segregated from other assets of CS Life and linked to the policy for the purpose of the SAC Act. This fund initially consists of the premium paid in cash or by transferring assets to CS Life. Nothing further needs to be done by CS Life after receiving the premium to constitute the internal fund. In particular, no process of transferring cash or other assets to the Bank to place them in the internal fund is contemplated or makes sense because the cash and other assets acquired by CS Life and held in a segregated account with the Bank already comprise the internal fund.

[36]It is also clear - and the Board did not understand CS Life to dispute - that the assets linked to the segregated account are owned (legally and beneficially) by CS Life. Not only is this implicit in the terms of the policy (eg the reference in the definition of “the internal fund” to “the other assets of [CS Life]”), but it is necessary if the policy is to operate as a contract of insurance. It would be inconsistent with the contract being one of insurance (and no doubt also with the intended benefit of avoiding liability for inheritance tax) if the sum payable on the death of the insured person represented assets already beneficially owned by the policyholder.

[37]The terms of the policy make it equally plain that, from the commencement date of the policy, the investor of the assets in the internal fund is CS Life. This is implicit in the arrangement that the assets are owned by CS Life and in the description of the Bank’s role as “managing” (rather than “investing”) the assets. It is explicit in the description of “Asset Management with or without Discretionary Mandates” quoted at para 19 above.

[38]The role of the Bank in managing the assets in the internal fund is either to make investment decisions on behalf of CS Life if a discretionary mandate is chosen by the policyholder or to execute investment instructions given on behalf of CS Life if a nondiscretionary mandate is selected. The way in which the non-discretionary mandate operates is by CS Life granting a power of attorney by which CS Life as principal authorises the policyholder to give instructions to the Bank with regard to the purchase and sale of investments. Where the policyholder is a company (as in the case of Meadowsweet and Sandcay), there is then a sub-delegation of this authority to a named individual (here, Mr Ivanishvili). It is inherent in these arrangements that the investor on whose behalf investment instructions are given is CS Life.

[39]Clause 6 of the General Conditions cannot reasonably be understood to mean that the premium paid to CS Life must then be “invested” by making a further transfer of the premium to the internal fund because, as discussed above, “the internal fund” is simply a description of the pool of assets acquired by CS Life through payment of the premium. No further step of passing the assets to the Bank is contemplated because the assets comprising the internal fund are already held in an account with the Bank and it is CS Life, and not the Bank, which is responsible for investing the assets. Nor, if this was what the clause required, would it make sense for the investment to be “in accordance with the investment alternative”. The choice of investment alternative determines how the assets which comprise the internal fund are invested after they are acquired by CS Life, not how the internal fund is initially constituted. The words of clause 6 cannot therefore be read literally. Premium cannot be invested in the internal fund in accordance with the investment alternative, as it is only the internal fund itself that can be so invested.

[40]It is apparent that, for these reasons, clause 6 is loosely worded. In the context of the policy as a whole, a reasonable person would understand it to mean:

“… The net single premium and the net additional premium (if any) are invested in credited to the internal fund which will be invested in accordance with the investment alternative indicated in the application form and as set out in the Policy and any Policy addendum. …”

The obligation is more precisely expressed in the part of the application form quoted at para 19 above (under “Asset Management”) which states that “[CS Life] will invest the insurance premium according to the investment alternative agreed with the policyholder”.

[41]The Board is therefore satisfied that the Chief Justice and Court of Appeal correctly construed the policy documents as imposing on CS Life an obligation to invest the policy assets in accordance with the investment alternative chosen by the policyholder. Where the investment alternative chosen by the policyholder is a discretionary mandate, this requires CS Life to invest the assets in accordance with investment decisions taken by the Bank’s MACS team. B.

The choice of mandate

[42]The Chief Justice found that, for both policies, the investment alternative chosen by the plaintiffs was a discretionary mandate. For the Meadowsweet policy, this was agreed orally at the meeting in Tbilisi in March 2011 at which Mr Lescaudron proposed to Mr Ivanishvili investment in a LPI policy (see para 11 above). For the Sandcay policy, it was agreed at the meeting in Tbilisi in June 2012 at which the Sandcay policy was sold to Mr Ivanishvili (see para 13 above). The Chief Justice found that at these meetings the Bank (and therefore Mr Lescaudron as its representative) was acting on behalf of CS Life, which had delegated to the Bank responsibility for the sale of the LPI policies, including agreeing the investment alternative which would apply to the assets transferred to CS Life (paras 188-211). The Court of Appeal concurred in these findings (paras 213-215).

[43]On this appeal CS Life does not and cannot challenge the factual findings that at the meetings mentioned a choice of the investment alternative with discretionary mandate was agreed orally by Mr Ivanishvili (on behalf of the policyholder) with Mr Lescaudron. But CS Life submits, first, that the courts below were wrong in law to find that, in these communications, Mr Lescaudron was acting as agent of CS Life; they should have concluded that Mr Lescaudron was acting solely as agent of the Bank. Second, CS Life submits that the choice communicated orally at the meetings was in any event overridden by the policy documents which recorded a choice of the non-discretionary investment alternative.

Agency of Mr Lescaudron

[44]The first of these submissions, in the opinion of the Board, raises no arguable point of law and, in so far as it takes issue with findings of fact, is not open to CS Life.

[45]It is clear – and the Board does not understand CS Life to dispute – that the Bank (and thus its employee, Mr Lescaudron) had authority to represent CS Life in connection with the sale of the LPI policies. That authority was conferred by a collaboration agreement between CS Life and the Bank which provided for the Bank through its employees (including therefore Mr Lescaudron) to promote CS Life’s products as an intermediary. The collaboration agreement also stated that the Bank “shall not have the right to represent” CS Life. The agreement was governed by Swiss law and, on the basis of expert evidence of Swiss law, the Chief Justice found that its effect was to give the Bank authority to negotiate contracts on behalf of CS Life but not to conclude such contracts.

[46]Had the choice of investment alternative required the conclusion of a contract between the policyholder and CS Life, this would have meant that the Bank (and thus Mr Lescaudron) lacked the necessary authority to agree the investment alternative on behalf of CS Life. As it is, however, a contract was not required for that purpose. Under the terms of the policies the choice of investment alternative is a unilateral choice of the policyholder. Nor need the choice be communicated in writing, although if the policyholder later wishes to change the investment alternative, clause 12 of the General Conditions requires a written and signed notification. Thus, neither the collaboration agreement nor the policy terms precluded Mr Lescaudron from negotiating the sale of the policies on behalf of CS Life and, in these discussions, receiving on behalf of CS Life oral notification from Mr Ivanishvili (on behalf of the prospective policyholder) of the choice of investment alternative. Subject to CS Life’s second submission, addressed below, that choice became effective when the policy commenced. The Chief Justice found as a matter of fact that, in accepting Mr Ivanishvili’s choice that the policy assets would be managed by the Bank on a discretionary basis, Mr Lescaudron was acting on behalf of CS Life (para 237); and the Court of Appeal concurred in that finding (paras 219 and 310(i)).

[47]CS Life relies on the principle that knowledge of an agent will be attributed to the principal only if the agent acquired the knowledge while acting for the principal: see eg Bowstead & Reynolds on Agency, 23rd ed (2023), para 8-211. It follows that if an agent comes by knowledge while acting solely for principal A, this knowledge will not be attributed to principal B. Counsel for CS Life submit that it is therefore essential to analyse the relevant relationships to determine whether Mr Lescaudron was acting for CS Life when he was notified by Mr Ivanishvili of the choice of investment alternative for the proposed policy.

[48]The Board agrees that this was essential. It was an exercise which the Chief Justice carried out. He found on the facts that Mr Lescaudron was acting on behalf of CS Life (as well as the Bank) at the relevant time. Counsel for CS Life have not identified any error of law in the analysis. They say it is clear that in the relevant discussions Mr Lescaudron was acting on behalf of the Bank. That is no doubt correct. But it does not contradict the finding that he was also acting on behalf of CS Life. It is common for an agent to act on behalf of more than one principal in a transaction. Evidence accepted by the Chief Justice showed that CS Life had outsourced to the Bank (and specifically to the Bank’s relationship managers) functions which included selling LPI policies and, as part of the sales process, agreeing with the client matters which included the choice of investment alternative (see paras 194-198). In performing those functions Mr Lescaudron was therefore acting for the Bank which was itself acting on behalf of CS Life.

[49]Even if it were open to CS Life to raise them, none of the points made in its submissions casts any doubt on the correctness of this analysis. Thus, counsel for CS Life emphasised that Mr Lescaudron was an employee of the Bank and well known to Mr Ivanishvili as a representative of the Bank; and that, at the time when the policies were discussed, there was no contractual relationship between the policyholders and CS Life; nor could Mr Lescaudron have been the relationship manager for CS Life because he was only appointed to that role once the policy came into force. Yet these facts are perfectly consistent with the findings that the Bank, and Mr Lescaudron as its employee, were acting on behalf of CS Life in connection with the sale of the LPI policies and thus acquired knowledge of the choice of investment alternative made for each policy while acting for CS Life. CS Life must therefore be taken to have known that, for each policy, the option of discretionary management by the Bank had been chosen.

Was the choice of a discretionary mandate overridden?

[50]As mentioned, CS Life’s second argument is that this choice was overridden by the selection of a non-discretionary mandate contained in the policy documents. Clause 6 of the General Conditions refers to “the investment alternative indicated in the application form” and the application form similarly refers to “the investment alternative chosen in the application”. The application form does not contain any box which the policyholder is invited to tick or other means by which the policyholder is invited to indicate its choice of investment alternative. But CS Life submits that such a choice was conveyed by appending to the application in each case a further form signed by Mr Ivanishvili on behalf of the policyholder. The form appended to the application for the Meadowsweet policy was headed “Investment Alternative without Discretionary Mandate”; and in the case of the Sandcay policy it was headed “Investments without Discretionary Management Mandate”.

[51]In each case this form contained the following statement:

“Selection of the Investment Instruments The Policyholder/s will choose the investments, which should be part of the internal fund. …”

By signing the form, the policyholder confirmed that: “I have read and understood the risks associated with the investment alternatives mentioned in section J ‘Risk Consideration’ as well as the General Description of the Internal Fund & Asset Management. I furthermore confirm that if and when I decide to change the composition of the integrated assets, I am responsible for assessing all possible risks related to the chosen assets. The insurance company bears no responsibility for the investment decisions of the policyholder. I bear the capital investment risk. In particular, attention should be paid to price and currency risks. The value of the internal fund may be lower than the invested amounts. Under unfavorable conditions, a total loss may even be incurred.”

[52]There is a different form used by CS Life for “Investments with Discretionary Management Mandate” which Mr Ivanishvili was not asked to sign (save in relation to one sub-account for the Sandcay policy, after that policy had been issued).

[53]CS Life contends that by signing the form headed “Investment Alternative without Discretionary Mandate” Mr Ivanishvili chose the non-discretionary investment alternative, and that this choice superseded the earlier oral notification that the assets were to be managed by the Bank on a discretionary basis.

[54]The courts below did not accept this contention. The Chief Justice found, and the Court of Appeal agreed, that the form signed by Mr Ivanishvili was facilitative only. Its purpose was to obtain the policyholder’s consent to bear the risks associated with the underlying investments including any investments chosen by the policyholder and to absolve CS Life of any responsibility for investment decisions of the policyholder, if and when the policyholder decided to change the composition of the internal fund. Giving this consent and waiver of responsibility was necessary to enable the policyholder to make its own investment decisions if it wished to do so. Signing this form therefore made such self-management of the portfolio possible. But it did not signify a choice to forgo discretionary management by the Bank and therefore did not override the prior oral agreement between Mr Ivanishvili and Mr Lescaudron that the policy assets would be managed by the Bank on a discretionary basis.

[55]In the opinion of the Board, the courts below were correct to construe the form signed by Mr Ivanishvili in this way. How the document would reasonably be understood depends on the background knowledge of the reader. The Board would accept that a reader presented with the policy documents who knew nothing about the prior choice of investment alternative, and who knew that it was the practice when a discretionary management mandate was chosen to require the policyholder to sign another form not used in this case, would reasonably infer that the non-discretionary investment alternative had been selected. There is no finding, however, that Mr Ivanishvili was aware of that practice. Furthermore, the background facts known to him and, through Mr Lescaudron, to CS Life when the contract was made, as found by the Chief Justice, included the facts: that Mr Ivanishvili’s assets had previously been managed by the Bank on a discretionary basis; that no suggestion or request had been made that this arrangement should or would change if assets were transferred to (or more money was invested in) LPI policies; and that, to the contrary, Mr Ivanishvili had expressly chosen the investment alternative of discretionary management.

[56]It is not suggested that before or when Mr Ivanishvili signed the application forms he expressed any intention or wish to alter that choice. A reasonable person with knowledge of this background would not interpret his signature of the form headed “Investment Alternative without Discretionary Mandate” as a request or agreement to change the investment alternative from that already chosen. Rather, such a person would reasonably understand the purpose and effect of the form to be facilitative only, as it has been construed by the courts below.

[57]The Chief Justice found that Mr Lescaudron had an ulterior motive for asking Mr Ivanishvili to sign that form (and not the form for “Investments with Discretionary Management Mandate”). This was to engineer a situation in which:(i) Mr Ivanishvili believed the Bank was professionally managing the policy assets on a discretionary basis;

(ii) the Bank’s investment management team believed that Mr Ivanishvili had chosen to make and was making his own investment decisions; and

(iii) Mr Lescaudron had a free rein to do as he pleased with the policy assets, by the simple expedient of falsely recording that the policyholder had given instructions which he was relaying for the trades that he was telling the operational team to make (paras 244-245). These intentions, however, were obviously not apparent to Mr Ivanishvili at the time when he entered into the LPI policies on behalf of Meadowsweet and Sandcay and it was essential to Mr Lescaudron’s fraudulent enterprise that they were concealed.

[58]The upshot is that the choice of the discretionary mandate notified by Mr Ivanishvili on behalf of the policyholders was not displaced by the policy documents, when executed. Those documents, construed in light of the background facts known to both contracting parties when the contracts were made, did not have the effect of revoking the choice of discretionary management which in each case had already been communicated to CS Life. The courts below were therefore right to hold that CS Life had a contractual obligation to invest the policy assets in accordance with investment decisions made by the Bank’s MACS team. The Board will refer to this obligation for short as “the investment duty”. In breach of the investment duty, no such discretionary management took place and investment decisions were instead made by Mr Lescaudron acting dishonestly and without authority from CS Life. C.

Other contractual duties

[59]As it is enough for the plaintiffs to establish that CS Life owed, and was in breach of, the investment duty, it is unnecessary to decide whether the Chief Justice and the Court of Appeal were also right to hold that CS Life owed any of the further contractual duties mentioned at para 25 above. It has not been suggested that the plaintiffs might be able to recover any damages for breach of any of those alleged duties which they would not be entitled to recover as compensation for breach of the investment duty.

Ground 2: Abuse of process

[60]Ground 2 of CS Life’s grounds of appeal has arisen in this way. On its appeal to the Court of Appeal CS Life argued, as it has before the Board, that it did not owe any of the contractual duties to the policyholder which the Chief Justice found were owed. CS Life evidently recognised that this stance has the unreasonable implication that the policyholder would be left with no remedy where, as happened, the Bank (through its employee) fraudulently mismanaged the policy assets, causing losses which the Bank’s wholly owned subsidiary, CS Life, made no attempt to recover from the Bank. To try to avoid this unreasonable consequence, CS Life came up with an imaginative argument in the Court of Appeal that the value of the assets in the internal fund had not actually been diminished by Mr Lescaudron’s fraud because CS Life had a claim against the Bank for the sums lost which is itself a policy asset.

[61]It might be thought that this argument, if successful, would prove too much because it would merely provide the plaintiffs with an alternative basis for recovering from CS Life the damages awarded by the Chief Justice, on the footing that CS Life failed to make any such claim. But CS Life’s (partial) solution to this difficulty was to assert that its claims against the Bank were governed by Swiss law and that the measure of damages would not be the same under Swiss law as under Bermudian law. CS Life submitted that a fresh quantification was needed for which expert evidence on at least Swiss law and valuation would be required. To carry out this exercise, the case would need to be remitted to the Supreme Court.

[62]The Court of Appeal refused to allow CS Life to advance its new argument, describing the attempt to do so as “comprehensively abusive” (para 136). Circumstances which made it an abuse of the court’s process were that the new case:(1) was unpleaded;

(2) had not been raised at first instance and indeed was inconsistent with CS Life’s case at trial;

(3) would, if it had been raised at first instance, have materially affected the scope of the trial, including by requiring further expert evidence (of Swiss law and valuation);

(4) was not included in CS Life’s grounds of appeal to the Court of Appeal;

(5) was not even referred to in CS Life’s skeleton argument for the appeal hearing; and

(6) was raised Page 18 for the first time in oral argument in the Court of Appeal without any advance notice to the plaintiffs.

[63]A more realistic litigant than CS Life would have recognised these reasons as overwhelming and would not have continued this abuse. Realism, however, has not been a characteristic of CS Life’s approach to this litigation. Ground 2 of CS Life’s grounds of appeal is that the Court of Appeal was wrong to find its new argument to be an abuse of process. The justification put forward for pursuing this ground of appeal is that, in making the new argument, CS Life is doing no more than referring to and relying on a finding made by the Chief Justice that CS Life was in breach of an obligation to pursue a claim against the Bank to recover the losses caused to the policy assets (see para 536 of the judgment). CS Life maintains that there is nothing unusual or abusive about a party relying on appeal on a finding of the Chief Justice on a point on which it lost at trial.

[64]The Board regards this justification as specious. In making the new argument CS Life is not merely seeking to rely on a finding of the Chief Justice. Indeed, it is unclear how CS Life can rely on the finding to which it refers. The allegation made by the plaintiffs and accepted by the Chief Justice was that CS Life was in breach of a duty to take steps to recover losses caused to the policy assets by the Bank’s mismanagement. The source of this duty was the obligation which the plaintiffs alleged, and the Chief Justice held, that CS Life owed as an implied term of the policies to monitor the investment of the policy assets for fraud and check that the chosen investment alternative was being complied with. On this appeal CS Life contends under Ground 1 that no such contractual obligation existed. It is contradictory – and in the Board’s view abusive – for CS Life to seek to rely on a finding that it was in breach of this contractual obligation in support of an argument that it had no such contractual obligation.

[65]It is in any case plain that the new argument is not in truth merely an adoption of the finding made by the Chief Justice. It differs from that finding in two significant ways. The first is in alleging that CS Life’s claim against the Bank is governed by Swiss law. The second is in alleging that CS Life’s claim against the Bank forms part of the internal fund. The former allegation is needed to produce a lower measure of damages. The latter allegation is needed to give the plaintiffs the benefit of a claim against the Bank without admitting that CS Life owed any of the contractual duties alleged by the plaintiffs.

[66]It is a basic rule of the conflict of laws that, if a party wishes to rely on foreign law, this must be pleaded, otherwise the court will apply its own national law by default: see eg Dicey, Morris & Collins on the Conflict of Laws, 16th ed (2022), rule 2 (Vol 1, Part 1, Chapter 3); Brownlie v FS Cairo (Nile Plaza) LLC

[2021] UKSC 45

;

[2022] AC 995

. Here CS Life has never pleaded a case that it has a claim against the Bank governed by Swiss law (nor that Swiss law differs from Bermudian law in any relevant respect). Nor did CS Life seek to argue such a case or to adduce evidence of Swiss law to support it at the trial. It is therefore not open to CS Life to contend that the losses which it is entitled Page 19 to recover from the Bank should be quantified in accordance with Swiss law or that expert evidence of Swiss law on this question should now be admitted.

[67]The contention that any claims that CS Life has against the Bank form part of the internal fund is also one not raised at the trial. At the hearing in the Court of Appeal counsel for CS Life produced a note relying on clause 11 of the General Conditions. This requires CS Life, when a request is made for surrender of the policy, to “realise the integrated investments” and “pay the value of the internal fund” to the policyholder. The contention is that one of the assets in the internal fund is a right to recover losses from the Bank caused by the Bank’s mismanagement of other policy assets; and that, when a request is made for surrender of the policy, CS Life has a duty to exercise this right so as to realise its value and pay the proceeds to the policyholder.

[68]Even had the new argument been valid and raised at the proper time, it would not have affected the conclusion that CS Life had a contractual obligation to invest the policy assets in accordance with the discretionary mandate chosen by the plaintiffs. The Board has reached that conclusion for reasons given at paras 31-41 above which do not depend on the unreasonable consequences that would flow from CS Life’s interpretation of the policy wording (though those unreasonable consequences certainly reinforce the correctness of the conclusion reached). As it is, it was an abuse of the court’s process to advance the new argument in the Court of Appeal for the reasons which that court gave, and a further abuse to seek to renew it before the Board.

Ground 3: Breach of fiduciary duty

[69]By Ground 3, CS Life contends that the courts below were wrong to hold that it owed fiduciary duties to the plaintiffs after the policies commenced. CS Life accepts that, before each policy commenced, the premium paid to CS Life (including by transfer of assets) was held on trust for the plaintiffs and that CS Life owed a duty as trustee to hold and preserve these assets until the policy became effective. It is common ground that, once the policy became effective, CS Life no longer held the assets on trust for the plaintiffs. CS Life submits that, from that point on, its only obligations to the plaintiffs were contractual, deriving from the terms of the policies, and that it did not owe them fiduciary duties. In particular, it did not owe extracontractual duties in equity to safeguard the policy assets and act in the best interests of the policyholders, as alleged by the plaintiffs.

[70]This ground of appeal raises questions – including as to the effect of section 18 of the SAC Act – which are not easy to resolve. Their resolution will make no difference to the outcome of the case, as the plaintiffs do not suggest that they are entitled to any further or better remedy for breach of fiduciary duty than for breach of the contractual investment duty which the Board has held that they were owed by CS Life. In the circumstances, the Board prefers to reserve its opinion for a case in which these questions need to be decided and to abstain from obiter dicta which might, as Bowen LJ once cautioned, “like the proverbial chickens of destiny, come home to roost sooner or later” and prove a “source of embarrassment in future cases”: Cooke v New River Co

(1888) 38 Ch D 56

, 71.

Ground 4: The measure of damages

[71]Grounds 4 to 6 of CS Life’s grounds of appeal relate to the assessment of damages and renew three criticisms of the Chief Justice’s approach which were rejected by the Court of Appeal. For these purposes it is only necessary to consider the damages recoverable for breach of the contractual investment duty. The plaintiffs do not suggest that the assessment of damages would be more favourable and result in a larger award if CS Life were held to owe any of the other duties alleged.

[72]Grounds 5 and 6, addressed below, concern the start and end dates chosen for the assessment. Ground 4, which the Board will consider first, takes issue with the measure adopted by the Chief Justice as the best measure of the investment performance that would have been achieved if CS Life had complied with its duty to invest the policy assets in accordance with the plaintiffs’ chosen investment alternative.

[73]The basic measure of the damages to which the plaintiffs are in principle entitled is the amount of money which will put them as nearly as possible in the same position financially as if the breach of contract had not occurred. In that scenario the policy assets would have been managed on a discretionary basis by the Bank’s MACS team. There was no direct evidence tendered to show what investment decisions the MACS team would have made. But a forensic accountancy expert instructed by the plaintiffs, Mr Davies, prepared calculations estimating what returns would have been achieved if the policy assets had been invested in a medium risk portfolio managed by a reputable European bank. This model of investment returns was referred to as “the whole portfolio model”. Mr Davies also prepared calculations using alternative models. One of these, referred to as “the objectionable transactions model”, focused on specific transactions which were identified by an investment management expert instructed by the plaintiffs, Mr Morrey, as unauthorised or imprudent. This model assumed that the funds invested in those objectionable transactions were invested in the hypothetical medium risk portfolio but that the other assets in the internal fund were invested as they in fact were.

[74]The Chief Justice considered that the appropriate model to adopt was the whole portfolio model. CS Life submits that he was wrong to do so and that he should have adopted the objectionable transactions model. The financial difference that this would make is substantial. The losses calculated by Mr Davies using the objectionable transactions model were US$416 million, around US$140 million less than the comparable figure calculated using the whole portfolio model.

[75]CS Life points out that, of the many investments bought and sold or held in the policy accounts by CS Life, Mr Morrey identified only 31 objectionable assets and 13 “overconcentrated positions” (eight of which were profitable). The policy accounts contained large quantities of blue chip securities, treasury bonds and other investments to which no objection was taken by Mr Morrey and which were not held in overconcentrated positions. No basis has been identified for suggesting that these assets were less likely to have been held if CS Life had complied with its contractual investment duty than the assets assumed by the whole portfolio model. Indeed, the bulk of the assets invested under the Meadowsweet policy were already owned by Meadowsweet before they were transferred to CS Life and included many unobjectionable assets. It is said that there is no justification for assuming that, if the management of the internal fund had been entrusted to the MACS team as should have occurred, they would have undertaken a wholesale replacement of these assets with other assets that would have performed better.

[76]The Board rejects this argument and agrees with the Court of Appeal that the approach of the Chief Justice was a wholly legitimate one to take. The plaintiffs were entitled to the benefit of discretionary management services from the Bank’s MACS team which, as a result of CS Life’s breach of contract, they never received. As Bell JA observed: “A properly managed portfolio would have adopted an entirely different approach to one managed by a fraudster without any authority at all” (para 300). No doubt the MACS team would be unlikely to have undertaken a wholesale replacement of the existing assets. But there is also no warrant for assuming that, just because an investment was not objectionable, it would have been left in place. While any model necessarily involves approximation, one that treats the whole portfolio as if it had been professionally managed throughout the relevant period is, in the Board’s view, clearly more appropriate than one which assumes that the only decisions made by the professional manager would have been to replace those assets identified as positively objectionable.

[77]That conclusion is reinforced by the likelihood noted by the Court of Appeal (see footnote 11 to para 299 of the judgment) that the objectionable transactions model was incomplete. In his third report Mr Davies pointed out that the objectionable transactions which had been identified as such by Mr Morrey did not include investments in respect of which Mr Lescaudron had received secret commissions and certain heavily lossmaking transactions where, for example, Mr Lescaudron had taken contrarian positions. Mr Davies thought it likely that not all objectionable transactions had been identified and suggested that this would explain why the loss estimated by the objectionable transactions model was substantially less than the loss calculated using the whole portfolio model. It is reasonable to assume that the Chief Justice took account of these points. They make it all the more inappropriate to assume – as the objectionable transactions model does – that if CS Life had complied with its investment duty every asset not specifically identified by Mr Morrey as objectionable would have been invested as it in fact was.

Ground 5: The start date for the assessment of damages

[78]For each policy, the Chief Justice took as the start date for the assessment of damages a date before the policy commenced. The transfers of assets and payments of cash to CS Life which constituted the premium for the Meadowsweet policy were made on various dates between 13 September and 25 October 2011. The policy was issued on 7 November 2011 with a commencement date of 25 October 2011. The Chief Justice took as the start date for the assessment of damages the end of the month in which the payment of premium was made, ie 30 September 2011. (The experts had agreed that calculations should run from the last day of a calendar month for practical reasons.) The same approach was adopted for the Sandcay policy. The first payment of premium was made in September 2012, and so 30 September 2012 was taken as the start date for the assessment. But the policy was not issued until 7 December 2012 and had a commencement date of 27 November 2012.

[79]Clause 3 of the General Conditions states:

“Commencement of the insurance coverage The Policy becomes effective on the date stated in the Policy. The date is determined after [CS Life] has verified the application for Life Portfolio International and [CS Life] has received the premium payment. By issuing the Policy, [CS Life] confirms to the policyholder its acceptance of the application for Life Portfolio International and it is the date that the segregated account linked to the Policy is created.”

[80]CS Life’s argument is simple. No contractual obligation to invest the premium could arise before the commencement date stated in the policy on which, in accordance with the above term, the policy became effective. As noted above, the commencement date stated in the Meadowsweet policy was 25 October 2011 and in the Sandcay policy it was 27 November 2012. Maintaining the approach of running the calculations from the last day of the month, the start date for the assessment of damages should therefore be 31 October 2011 (instead of 30 September 2011) for the Meadowsweet policy and 30 November 2012 (instead of 30 September 2012) for the Sandcay policy. Although the difference is only of one or two months, CS Life claims (whether correctly or not the Board cannot say) that some US$58 million turns on this point.

[81]The Chief Justice gave no reason for awarding damages for periods before the policies commenced. But his assessment of damages did not distinguish between the claims for breach of contract and for fraudulent misrepresentation. The Chief Justice had found that, if Mr Ivanishvili had not been deceived, he would not have transferred assets Page 23 to CS Life but would instead have moved them to another European bank to be invested in a medium risk investment portfolio. In awarding damages for misrepresentation, it was reasonable, on this basis, to take as the start dates for the calculation the dates when assets were transferred to CS Life. The Chief Justice may have overlooked the fact that the same analysis did not apply to the claim for breach of contract.

[82]The Court of Appeal upheld the start dates adopted by the Chief Justice on the basis that, as soon as the assets were transferred to CS Life, CS Life owed fiduciary duties:

“(i) to act in the best interests of the policyholders; (ii) to hold the funds advanced for the policy premiums, (now in the legal ownership of CS Life), strictly in accordance with the purpose for which they had been advanced, namely to be invested by CS Life (via the Bank) in accordance with the agreed investment alternative, and (iii) thereby to safeguard the assets”

. The Court of Appeal held that “either mismanaging the assets or not managing them at all” before the policies commenced were clear breaches of these fiduciary duties: see para 266 of the judgment.

[83]It is not in dispute that, when CS Life received assets or cash intended for use as premium, it held the funds on trust for the prospective policyholder until the policy incepted. The Board considers that the trust created was a bare trust under which the trustee was obliged to hold the assets to the order of their beneficial owner. CS Life did not yet owe a duty to invest the funds in accordance with the chosen investment alternative. That duty could not arise unless and until the prospective policyholder and CS Life agreed that it should. It is not suggested that any request was made on behalf of Meadowsweet or Sandcay to CS Life or that any agreement was reached authorising CS Life to invest any funds before either policy was issued. Far from being obliged to do so, CS Life therefore had no right to invest those funds in a medium risk portfolio managed by the Bank’s MACS team. Its obligation was solely to hold the assets. When a contract was concluded by issuing the policy, it was open to the parties to agree that the policy should be treated as having taken effect from an earlier date, and they did so (see para 78 above). But it was only from that specified commencement date that CS Life was obliged (or entitled) to invest the premium in accordance with the investment alternative chosen by the policyholder.

[84]The Board therefore considers that the courts below were wrong to hold that damages should be awarded from dates before the Meadowsweet and Sandcay policies commenced. The correct approach is to take as the start date the end of the month in which the policy became effective. If in the period between receipt of funds by CS Life and that date any unauthorised transaction took place in relation to the funds received which caused loss, the policyholder is entitled to compensation for that loss. This will be achieved if, in assessing damages, the amount of any such loss is added to the value of the assets on the start date (which will also include any income earned or profits made since receipt) and is thus treated as having been invested in the model portfolio.

Ground 6: The end date for the assessment of damages

[85]The Chief Justice awarded damages for losses occurring up to the date of judgment (29 March 2022). In the Court of Appeal, CS Life argued that 31 August 2017 should have been taken as the end date for the calculation, as any loss suffered after that date was a result of Mr Ivanishvili’s free choice to leave assets with CS Life despite knowing of Mr Lescaudron’s fraud. The Court of Appeal held that this argument was not open to CS Life and affirmed the award of damages for the period up to the date of judgment. By Ground 6, CS Life contends that the Court of Appeal erred and should have held that the plaintiffs cannot recover damages (but only pre-judgment interest) for the period after 31 August 2017.

Relevant facts

[86]The following facts, taken from the judgment of the Court of Appeal, are not disputed or open to dispute on this appeal.

[87]The plaintiffs became aware of Mr Lescaudron’s fraud in September 2015. Mr Ivanishvili gave evidence that, having lost all confidence that the policy assets were being properly managed, he decided in early 2016 to transfer his assets to another bank as soon as possible and in summer 2016 approached Julius Baer & Co Ltd. After some initial meetings, he agreed to engage them to manage the policy assets.

[88]On 31 March 2017 Mr Ivanishvili signed letters of wishes asking the trustees of the Mandalay and Green Vals Trusts to procure a partial surrender of “the maximum permitted amount (94.9%)” of the value of the assets in the policy accounts and the transfer of those assets to Julius Baer. The letters also stated:

“To the extent that the assets of the Trust continue to be held in the custody or control of the Bank, including the remainder of the assets not surrendered under the [CS Life] policy, I request all such assets to be managed in accordance with existing management powers and agreements in place in respect of those assets, subject to any further letter of wishes being given by me.”

[89]The basis for the belief that any partial surrender had to be for less than 95% of the value of the policy assets is unclear. Clause 11 of the General Conditions permits a partial or total surrender of the policy at any time but provides that, upon a total surrender, the policy “will expire and all liabilities of [CS Life] will immediately and irrevocably cease from the date of the payment”. To preserve their claims against CS Life, the plaintiffs had therefore to keep the policies in effect and leave some assets in the internal fund. But there was no minimum required amount.

[90]In April 2018, and again in July 2018, the trustees requested confirmation from CS Life that no liabilities of CS Life would cease if a surrender of an amount greater than 94.9% of the value of the assets but less than 100% were to be made. This confirmation was duly given. (Indeed, in July 2018 CS Life replied – inconsistently with the policy terms – that even a total surrender of the policy would not affect CS Life’s liability.)

[91]When the fraud was discovered in September 2015, there was some US$117 million in the Meadowsweet policy account and US$141 million in the Sandcay account, making US$258 million in all. Before the action against CS Life was begun on 17 August 2017, only about US$8.3 million had been withdrawn. Even at the time of trial in late 2021 nearly US$60 million remained in the Sandcay policy account. The amount left in the Meadowsweet account at that time is unclear, but figures produced in closing submissions at the trial showed that nearly US$125 million remained in that account on 15 June 2020.

[92]The Court of Appeal noted, at paras 277 and 290, that, when asked at the trial to explain why he did not transfer the policy assets to another bank in 2015 or 2016, Mr Ivanishvili seemed to have no recollection and to be unable to give an explanation. In his first witness statement dated 14 May 2020, Mr Ivanishvili had said, without giving details, that “the nature of various of the assets held in the [LPI policies] has meant that transfer of the assets has not always been straightforward” and that “not all assets have been transferred … because CS Life has refused to confirm whether it would claim that full surrender of the policies somehow had the effect of defeating the claims made in these proceedings”. In a later witness statement, Mr Ivanishvili claimed (again without giving details) that CS Life had not allowed surrenders to be made of the amount remaining in the policy accounts.

[93]Although the Chief Justice accepted the explanation given in Mr Ivanishvili’s first witness statement (see para 438 of the judgment), it was rightly rejected by the Court of Appeal. While clause 11 of the General Conditions did indeed provide that all liabilities of CS Life would cease upon a total surrender of the policy, there was no obstacle to a partial surrender of any amount short of 100% - as CS Life confirmed when asked.

The plaintiffs’ loss

[94]At the trial CS Life did not contend that the investment duty, if it existed at all, had ceased. Clearly, both the Meadowsweet policy and the Sandcay policy were still in effect. Neither side alleged that either policy had been terminated, and it is common ground that neither policy had been totally surrendered. CS Life denied that it ever had an obligation Page 26 to invest the policy assets pursuant to discretionary mandates. But the courts below decided that issue against CS Life and the Board has upheld that decision. In particular, the Board has affirmed the finding that the investment alternative chosen for both policies was discretionary management by the Bank.

[95]In argument before the Board, counsel for CS Life pointed to passages in the plaintiffs’ statement of claim which allege that the Bank was not authorised to manage the policy assets because no written discretionary mandates had been executed. It was said to be “grievously unfair” for the plaintiffs to maintain that CS Life still had a duty to invest the policy assets pursuant to discretionary mandates after the statement of claim was served in August 2017, when the plaintiffs were positively asserting in their pleading that CS Life had no authority, let alone duty, to do this.

[96]There might have been a legal basis for this complaint if CS Life had shown that it was deterred from performing its investment duty because of what was pleaded in the statement of claim. But this has never been suggested. CS Life has never alleged that it relied in any way on the plaintiffs’ pleaded allegation that there was no valid discretionary mandate. In the absence of such reliance, the Board does not consider that there is any legal basis for contending that the pleading relieved CS Life from continuing to comply with the investment duty or prevented the plaintiffs from claiming damages calculated on the footing that CS Life was in continuing breach of that duty. In any event, no such case was made below, and it is not open to CS Life to raise it for the first time before the Board.

Causation of loss

[97]The argument made by CS Life in the Court of Appeal and renewed before the Board does not dispute that at the time of the trial CS Life remained in breach of the investment duty and that the plaintiffs were continuing to suffer loss. The argument is that in law such loss was not caused by CS Life’s breach of contract. Rather, for the period after 31 August 2017 the sole cause in law of the plaintiffs’ further losses was their own voluntary action in choosing to leave most of the policy assets with CS Life.

[98]CS Life did not plead such a case or raise it in opening submissions at the trial. It was raised for the first time in written closing submissions served at the conclusion of CS Life’s closing oral argument. In those submissions counsel for CS Life asserted that the plaintiffs’ losses had “crystallised” at the date when Mr Ivanishvili ought to have realised that he was free to move his assets, which was said to be no later than August 2017. Any losses suffered after that date were caused by the plaintiffs’ own independent decisions and not by the wrongdoing of CS Life and thus are not recoverable.

[99]The Chief Justice did not directly address this argument in his judgment. Given the very late stage and informal way in which the point was raised, he was in the Board’s view entitled to disregard it. The argument was squarely raised in the Court of Appeal. But the Court of Appeal held that the argument was not open to CS Life because it amounted to a defence that the plaintiffs had failed to mitigate their loss. The burden of pleading and proving such a defence lies with the defendant. No such defence was pleaded and, because it was raised for the first time by CS Life only in closing submissions at the trial, relevant facts were not investigated. In particular, the Court of Appeal could not be sure that it had all the relevant evidence, including evidence about the circumstances in which the letters of wishes issued on 31 March 2017 were not in fact put fully into effect, the time taken by CS Life to action surrender requests and complications which were said to have arisen in respect of the transfer of assets from the Bank (para 286). The Court of Appeal also declined to make a finding that Mr Ivanishvili realised that the policy assets were not being managed by a discretionary team and concluded that the extent of Mr Ivanishvili’s awareness was not clear (para 289).

Arguments on this appeal

[100]On this appeal CS Life submitted, in its written case, that the Court of Appeal was wrong to treat its argument as being that the plaintiffs failed to mitigate their loss. Rather, the issue is one of causation, which is for the plaintiff to plead and prove. In support of this submission, counsel for CS Life relied before the Board, as they did below, on the seminal judgment of Robert Goff J in Koch Marine Inc v d’Amica Societa di Navigazione ARL (The Elena d’Amico)

[1980] 1 Lloyd’s Rep 75

, 88-89.

[101]In the Board’s opinion, this submission misses the fundamental point made by Robert Goff J in The Elena d’Amico that mitigation and causation are not two separate principles. Having referred to the discussion in McGregor on Damages, 13th ed (1972), para 205, which distinguished three different aspects of the principle of mitigation, Robert Goff J said, at p 88:

“Now, in my judgment, these three aspects of mitigation are all really aspects of a wider principle which is that, subject to the rules of remoteness, the plaintiff can recover, but can only recover, in respect of damage suffered by him which has been caused by the defendant's legal wrong. In other words, they are aspects of the principle of causation. It follows that what is alleged to constitute mitigation in law can only have that effect if there is a causative link between the wrong in respect of which damages are claimed and the action or inaction of the plaintiff.”

[102]Robert Goff J went on to explain the market mitigation rule that, where there is an available market in which the plaintiff can obtain an adequate substitute for goods or services of which it has been deprived by the defendant’s breach, damages will be assessed as if the plaintiff had entered the market at the earliest reasonable opportunity and obtained such a substitute, whether or not the plaintiff in fact did so. That is because (p 89):

“If … [the plaintiff] decides not to take advantage of that market then, generally speaking, that will be his own business decision independent of the wrong; and the consequences of that decision are his.”

In other words, the plaintiff’s voluntary choice breaks the chain of causation between the defendant’s breach of duty and any subsequent loss (or gain).

[103]This legal analysis has been endorsed, including at the highest level, in many later cases: see eg Bunge SA v Nidera BV

[2015] UKSC 43

; [2015] Bus LR 987, paras 80-81 (Lord Toulson); Sharp Corpn Ltd v Viterra BV

[2024] UKSC 14

; [2024] Bus LR 871, paras 85-98 (Lord Hamblen).

[104]The same underlying reasoning is therefore involved when the mitigation principle is relied on as when the plaintiff’s conduct is alleged to constitute a new intervening act which breaks the chain of causation between the defendant’s breach of duty and damage suffered by the plaintiff. As noted by Andrew Burrows, Remedies for Torts, Breach of Contract, and Equitable Wrongs, 4th ed (2019), pp 84-85, the courts sometimes use the two principles interchangeably. In each case the question is whether the plaintiff is to be regarded as having made a voluntary choice to act in a particular way following the defendant’s breach of duty, such that the plaintiff alone should be held responsible for the consequences of that choice. The plaintiff’s conduct is usually viewed through the lens of mitigation, rather than intervening act, when the action which the plaintiff takes or fails to take is one calculated to avoid damage that the plaintiff might otherwise have suffered from the defendant’s breach.

[105]CS Life was also wrong to suggest that how the issue is characterised affects the burden of proof. Just as the burden of pleading and proving a failure to mitigate loss lies with the defendant, so does the burden of pleading and proving that an event subsequent to the defendant’s breach of duty operated as a new intervening cause: see Armstead v Royal & Sun Alliance Insurance Co Ltd

[2024] UKSC 6

;

[2025] AC 406

, paras 60-61.

[106]Here the argument made by CS Life about the appropriate end date for the assessment of damages was, in the Board’s view, rightly characterised by the Court of Appeal as an argument that the plaintiffs failed to take reasonable steps to mitigate their loss. The thrust of the argument is that, after Mr Lescaudron’s fraud was discovered and the plaintiffs knew that they were free to withdraw the policy assets, the course of action which a reasonable and prudent person in the plaintiffs’ position could be expected to take was to transfer all but a token amount of the policy assets to another fund manager. But it makes no relevant difference if the argument is presented, as CS Life prefers to present it, as alleging a break in the chain of causation. Either way, the burden of pleading and proving this case was on CS Life.

[107]Once this is recognised, the conclusion reached by the Court of Appeal is unassailable. When Mr Lescaudron’s wrongdoing was discovered, Mr Ivanishvili was entitled to assume that management of the policy assets would now be put in the hands of the Bank’s MACS team in accordance with the investment alternative that had been agreed. He was also entitled thereafter to assume continuing performance of this contractual obligation, unless and until it was made clear to him that no such discretionary management was being or would be provided. Then but only then could it be said that a reasonable and prudent person in the plaintiffs’ position who wanted the policy assets to be professionally managed would not have left them with CS Life but would have moved them to another fund manager. Put another way, only then could it be said that the lack of professional management of the assets and any losses to which it gave rise were the result of a voluntary choice made by Mr Ivanishvili.

[108]As discussed, the burden was on CS Life to prove that, by the end of August 2017, Mr Ivanishvili knew that the policy assets were not being professionally managed by the Bank. CS Life made no attempt to show this. Mr Ivanishvili gave evidence – and the Chief Justice found as a fact (para 216) – that he understood “throughout” that the Bank was managing the investments in the policy accounts. While Mr Ivanishvili may have been referring, at least primarily, to the period before Mr Lescaudron’s fraud was discovered, CS Life did not assert that this understanding later changed.

[109]Thus, the Chief Justice was not asked to find, and lacked an evidential basis for finding, that, before the end of August 2017, Mr Ivanishvili had become aware that the policy assets were not being managed on a discretionary basis by the Bank. The Court of Appeal was invited to make such a finding but declined to do so, on the grounds both that CS Life had not properly raised such a case and that the extent of Mr Ivanishvili’s awareness was unclear. Those grounds are unimpeachable and there is no basis on which the Board could properly interfere with that decision.

[110]The Board accordingly rejects the argument that the Court of Appeal erred in not imposing an end date for the assessment of damages of 31 August 2017.

Conclusion on the appeal

[111]The effect of the Board’s conclusions is that CS Life’s appeal should be dismissed except in one respect. This concerns the start dates for the assessment of damages (Ground 5). The damages awarded to the plaintiffs should be re-calculated, starting from 31 October 2011 (instead of 30 September 2011) for the Meadowsweet policy and 30 November 2012 (instead of 30 September 2012) for the Sandcay policy. In performing the calculations, there should be added to the actual value of the assets on these start dates any losses resulting from any unauthorised transactions during the period when the premium was held by CS Life before the policy commenced (see para 84 above). III. THE CROSS-APPEAL: MISREPRESENTATION

[112]By the cross-appeal, the plaintiffs are asking the Board to restore the decision of the Chief Justice that they are entitled to damages for fraudulent misrepresentation. They contend that the Court of Appeal was wrong to reverse that decision.

[113]If CS Life’s appeal against the award of damages for breach of contract had failed entirely, the outcome of the cross-appeal would have had no practical significance because the Chief Justice awarded the same damages for fraudulent misrepresentation as for breach of contract. But the Board has upheld Ground 5 of CS Life’s appeal and concluded that the start dates for the assessment of damages for breach of contract should be later than those adopted by the Chief Justice. There is no challenge to his assessment (including the start dates) of the damages payable for fraudulent misrepresentation, if that claim succeeds. It follows that the cross-appeal is not academic because, if the crossappeal and hence the misrepresentation claim succeeds, the plaintiffs will obtain a larger damages award (although the additional amount may be relatively small compared with the award for breach of contract).

The misrepresentation claim

[114]The misrepresentations on which this further claim is based were made by Mr Lescaudron (on behalf of CS Life) to Mr Ivanishvili (on behalf of the prospective policyholders) at the meetings in Tbilisi, Georgia, at which the LPI policies were proposed (see paras 11 and 13 above).

[115]The Chief Justice found that, by recommending investment in the LPI policies, Mr Lescaudron impliedly represented that the Bank (including Mr Lescaudron himself) was not managing the plaintiffs’ accounts fraudulently and did not intend to manage the policy assets fraudulently. Those representations were false and known to be so by Mr Lescaudron. They were intended to, and did, induce the plaintiffs to enter into the LPI policies. (In this regard, the Chief Justice found that Mr Lescaudron could not have induced the plaintiffs to enter into the LPI policies without making the implied representations: para 699.) These findings were affirmed by the Court of Appeal and are not challenged before the Board.

[116]The issues on the cross-appeal are whether the Court of Appeal was right to hold that the misrepresentation claim failed because:(1) the plaintiffs had not pleaded and proved that Mr Ivanishvili had any conscious awareness or understanding of the representations made to him; and/or

(2) the claim was subject to the law of Georgia relating to limitation and was brought after the expiry of the three-year limitation period prescribed by Georgian law.

What law applies?

[117]A preliminary question is: what system of law governs the misrepresentation claim? What law applies to claims in tort has proved a vexed question for courts in common law jurisdictions. In Boys v Chaplin

[1971] AC 356

a majority of the House of Lords held that, as a general rule, a claim in tort arising from an act of the defendant done in a foreign country is “actionable” in England and Wales only if the act is actionable both(1) as a tort under English law and

(2) under the law of the foreign country where it was done. This is known as the “double actionability” rule.

[118]In Boys v Chaplin Lords Hodson and Wilberforce thought that the rule should be given a flexible interpretation such that it need not always be applied. In Red Sea Insurance Co Ltd v Bouygues SA

[1995] 1 AC 190

the Privy Council, on an appeal from Hong Kong, recognised the existence of such an exception. This allowed the court, in a clear case, to apply (only) the law of the country which, with respect to a particular issue between the parties or the case as a whole, “has the most significant relationship with the occurrence and the parties” (pp 206-207). In Imanagement Services Ltd v Cukurova Holdings AS [2008] ECarSC 119, para 56, the Eastern Caribbean Court of Appeal took this to require that “all or almost all the significant connecting factors point in the direction of [the jurisdiction in question]”.

[119]In Sophocleous v Secretary of State for Foreign and Commonwealth Affairs

[2018] EWCA Civ 2167

;

[2019] QB 949

the Court of Appeal of England and Wales had to apply these rules to decide what law governed historic claims for assaults allegedly committed by British soldiers and seconded police officers in Cyprus in the 1950s. Longmore LJ (who gave the court’s reasons) observed that “[t]he width of the flexible exception has never been defined” and that “there is no mechanical rule determining when to apply it” (paras 11 and 30). But he also noted that there is “high authority that the flexible exception should not be too readily available” and cited the opinion of Lord Wilberforce in Boys v Chaplin, at p 391, that the general rule should apply “unless clear and satisfying grounds are shown why it should be departed from” (para 33).

[120]In the United Kingdom the double actionability rule was strongly criticised by the Law Commissions, leading to its statutory abolition: see Private International Law: Choice of Law in Tort and Delict, Law Com No 193 and Scot Law Com No 129 (1990). Four defects identified by the Law Commissions are usefully summarised in Dicey, Morris & Collins on The Conflict of Laws, 16th ed (2022), para 35-013. First, it is anomalous to require actionability by two systems of law, such an approach being unknown in any other area of UK private international law. Second, as well as being almost unknown in the private international law of any other country, it is “parochial in appearance” and unjustifiable in principle to apply the substantive law of the forum in cases involving a tort regardless of the foreign complexion of the factual situation. Third, the general rule gives an unfair advantage to the defendant because the claimant cannot succeed unless liability is established under two systems of law, whereas the defendant can escape liability by taking advantage of any defence available under either the law of the place where the tort was committed or the law of the forum. Fourth, the formulation and application of the “flexible exception” to the general rule is speculative and uncertain.

[121]The double actionability rule and the exception to it were abolished in the United Kingdom (except in relation to defamation claims) by Part III of the Private International Law (Miscellaneous Provisions) Act 1995. They were replaced by a general choice of law rule that “the applicable law is the law of the country in which the events constituting the tort or delict in question occur”: section 11(1). That general rule is displaced if the comparative significance of factors connecting the tort with another country make it “substantially more appropriate” to apply the law of that other country: section 12. The 1995 Act has itself been largely superseded by the Rome II Regulation (EC) No 864/2007 of 11 July 2007 on the law applying to non-contractual obligations, which has been retained in UK law after the withdrawal of the United Kingdom from the European Union.

[122]The double actionability rule has also been rejected in several other common law jurisdictions, either by judicial interpretation or legislation. In 1994 the Supreme Court of Canada replaced it with a rule requiring application of the law of the place where the tort was committed (subject in international cases to a narrow exception to avoid injustice): see Tolofson v Jensen

[1994] 3 SCR 1022

. A similar approach has been taken by the High Court of Australia: see John Pfeiffer Pty Ltd v Rogerson [2000] HCA 36;

(2000) 203 CLR 503

; and Régie Nationale des Usines Renault SA v Zhang [2002] HCA 10;

(2003) 210 CLR 491

. In New Zealand, the double actionability rule was replaced in 2017 by legislation modelled in part on the 1995 Act in the United Kingdom.

[123]Against this background, it would have been open to either side on this appeal to invite the Board not to follow Boys v Chaplin and to hold that the general rule in Bermuda, as in Canada and Australia, is that the court should apply the law of the place where the tort was committed without any further requirement that the tort is also actionable under the law of the forum. But neither party has chosen to advance such a case. The parties are agreed that the law summarised at paras 117-119 above should be applied and the Board will do so.

The place where the tort was committed

[124]To apply the double actionability rule, it is first necessary to identify the place where the act constituting the alleged tort was committed. Here it is common ground that this place was Georgia. That was where the misrepresentations were made and communicated to Mr Ivanishvili. It is also where the misrepresentations were acted on by Mr Ivanishvili by signing letters of instruction to the trustees of the Mandalay and Green Vals Trusts to enter into the LPI policies and by signing the application forms for the policies on behalf of Meadowsweet and Sandcay.

The issue of domestic tort law

[125]Before considering the question of actionability under the law of Georgia, the Board will first decide whether the misrepresentation claim is actionable under the law of Bermuda. Here the only issue now in dispute is whether the Court of Appeal was right to hold that the claim failed because the plaintiffs had not pleaded and proved that Mr Ivanishvili understood Mr Lescaudron to be making the implied representations on which the claim is based. Although the plaintiffs have submitted that on the findings made by the Chief Justice any such requirement is satisfied, the Board agrees with CS Life that this submission is untenable. The plaintiffs did not allege, and Mr Ivanishvili gave no evidence at the trial, that he was aware of or understood Mr Lescaudron to represent that he was not managing the plaintiffs’ accounts fraudulently and did not intend to manage the policy assets fraudulently. (Nor did Mr Ivanishvili even give evidence that he assumed this to be the case.) Instead, the plaintiffs elected to advance their case solely on the footing that awareness and understanding of the representation is not a legal requirement. The Chief Justice accepted that contention (see paras 682, 689 and 691 of his judgment). But the Court of Appeal rejected it and allowed CS Life’s appeal on this ground (paras 258-263).

[126]The issue for the Board is therefore a pure question of law: is it a legal requirement of a claim in deceit that the plaintiff was aware of the representation on which the claim is based? The issue has been argued on the basis that the relevant Bermudian law is the same as the law of England and Wales.

The tort of deceit

[127]The plaintiffs have not sought to avoid (ie set aside) the policies on the ground that they were induced to enter into them by fraudulent misrepresentation. Their misrepresentation claim is solely a claim for damages for what is traditionally called the tort of deceit. This tort involves a simple and perfectly general principle: a person who causes another person to suffer loss by deceiving that other person is liable to compensate the other person for such loss.

[128]What it means to deceive someone can be unpacked into a number of separate elements. It involves(1) making a representation of fact (or law) which

(2) is false,

(3) the maker does not believe to be true,

(4) is intended to be believed by the representee, and

(5) causes the representee to believe that the representation is true.

[129]The scope of what counts as a representation for this purpose is very broad. The concept is not limited to statements which expressly assert the truth of a proposition. Indeed, it is not limited to statements: it includes actions as well as words. For the purpose of the law of deceit, the term “representation” encompasses any words or act calculated to cause another person to believe a proposition.

[130]The breadth of the concept can be illustrated by some examples, most of them taken from a longer list given in Clerk & Lindsell on Torts, 24th ed (2023), para 17-09. The following have been held to constitute deceit: pledging goods as security for a loan knowing that one has no title to the goods or authority from the owner to pledge them (Advanced Industrial Technology Corpn Ltd v Bond Street Jewellers Ltd

[2006] EWCA Civ 923

); ordering goods on credit on behalf of a company known to be insolvent (Contex Drouzhba Ltd v Wiseman

[2007] EWCA Civ 1201

;

[2008] BCC 301

); presenting company accounts to a buyer knowing that they had been dishonestly prepared (MAN Nutzfahrzeuge AG v Freightliner Ltd

[2005] EWHC 2347 (Comm)

, para 78); and inviting someone to invest in a company known to be insolvent (Sinha v Taylor

[2022] EWHC 1096 (Comm)

, para 57).

[131]Two cases particularly relied on by counsel for the plaintiffs are Gordon v Selico Ltd

(1986) 18 HLR 219

and Spice Girls Ltd v Aprilia World Service BV

[2002] EWCA Civ 15

; [2002] EMLR 27. In Gordon v Selico the seller of a flat who deliberately covered up dry rot so that the prospective buyer would not see it was held liable in deceit. In Spice Girls the participation in photoshoots of all five members of the Spice Girls group and their approval of promotional material depicting all five of them for use in advertising motor scooters manufactured by their sponsor (Aprilia) were held to represent that they did not know that any member had declared an intention to leave the group (as Geri Halliwell had in fact done). A claim for damages for misrepresentation succeeded although, as the Court of Appeal in that case noted at para 67, “no one at [Aprilia] gave any consideration at the time to what representations were to be implied into the statements and conduct of the Spice Girls”.

[132]There is nothing recent or novel in the notion that deceit can be perpetrated by entirely non-verbal conduct, including conduct of which the claimant is unaware. An old example is Schneider v Heath (1813) 3 Camp 506, where the seller of a ship, to hide the fact that the hull was worm-eaten and the keel broken rendering the ship unseaworthy, had the ship removed from the ways where she lay dry and floated in a dock so that the defects would not be seen when the buyer came to bid for her. Sir James Mansfield CJ had no hesitation in holding that on these facts the buyer was entitled to succeed in a claim to recover back his deposit on the ground that he was induced to pay it by deceit.

[133]Two further examples discussed in oral argument were, first, the case of a person who orders food in a restaurant. By doing so the person (absent special circumstances) represents that he or she has the means and intention to pay for the meal. The second, similar, example discussed was that of a person who hails a taxi available for hire by waving her arm. Counsel for CS Life accepted that by this action the prospective passenger ordinarily represents that she has the means and intends to pay the fare.

[134]A variant of the first of these situations occurred in Director of Public Prosecutions v Ray

[1974] AC 370

. The defendant went with friends to a Chinese restaurant intending to have a meal there and pay for it. After eating the main course, they decided not to pay after all, waited until the waiter went out of the room and then did what is popularly known as a “runner”. The defendant was convicted of dishonestly obtaining a pecuniary advantage by deception. On an appeal to the House of Lords, the question was whether, on these facts, the defendant was guilty of deception. This turned on whether, after forming the dishonest intention not to pay for the meal, the defendant had represented and induced the waiter to believe that he did intend to pay and to act on that belief. It was held (by a majority of three to two) that the defendant’s conduct, viewed as a whole, was a continuing representation of his present intention to pay and that his change of mind caused the waiter not to take steps that he would otherwise have taken to ensure that the group did not leave without paying. The conviction was accordingly upheld. The reasoning would apply equally to a civil claim for deceit.

[135]What is significant about the case of DPP v Ray is not so much the conclusion reached on the particular facts but the starting-point accepted by all the law lords (and described by Lord Hodson at p 389D as “trite law and common sense”) that, by ordering food in a restaurant, a person impliedly represents that he or she intends to pay for it before leaving. This and other cases of ordering goods or services without an intention to pay for them are straightforward instances of deceit.

[136]Another type of case, which must be increasingly common, is where a fraudulent misrepresentation is made not to any human being, but to a machine. In Renault UK Ltd v Fleetpro Technical Services Ltd

[2007] EWHC 2541 (QB)

, para 122, the judge held that a fraudulent misrepresentation giving rise to liability in deceit can be made to a machine, rather than to an individual, if the machine is set up to process certain information in a particular way in which it would not process information about the material transaction if the correct information were given; see also Skatteforvaltningen v Solo Capital Partners LLP

[2025] EWHC 2364 (Comm)

, paras 531-532.

Authority for the alleged requirement of awareness

[137]Statements of the essential elements of a claim in deceit do not traditionally include a requirement that the claimant was aware of the representation relied on and understood it to have been made. For example, in a summary that has often been cited, Viscount Maugham in Bradford Third Equitable Benefit Building Society v Borders

[1941] 2 All ER 205

, 211, said that an action for deceit requires four things to be established: “First, there must be a representation of fact made by words, or, it may be, by conduct. … Secondly, the representation must be made with a knowledge that it is false. It must be wilfully false, or at least made in the absence of any genuine belief that it is true … Thirdly, it must be made with the intention that it should be acted upon by the plaintiff, or by a class of persons which will include the plaintiff, in the manner which resulted in damage to him … Fourthly, it must be proved that the plaintiff has acted upon the false statement and has sustained damage by so doing ...” (Citations omitted.) Many similar statements of the law can be found: see eg Derry v Peek

(1889) 14 App Cas 337

, 360-361, 374; Clerk & Lindsell on Torts, 24th ed (2023), para 17-01.

[138]The earliest case cited to the Board in which a requirement of awareness or understanding was said to be an essential element of the cause of action is Raiffeisen Zentralbank Österreich AG v Royal Bank of Scotland plc

[2010] EWHC 1392 (Comm)

;

[2011] 1 Lloyd’s Rep 123

. There Christopher Clarke J, at para 80, identified the matters which the claimant had to show to succeed as including:

“(a) that [the defendant] made representations to it; (b) that it understood that those representations were being made; (emphasis added) (c) that such representations were false; (d) that it was induced by those representations … [to enter into the contract] …”

The second element, which the Board has emphasised, was expanded at para 87:

“the claimant must show that he in fact understood the statement in the sense (so far as material) which the court ascribes to it: Arkwright v Newbold (1881) 17 Ch D 301; Smith v Chadwick (1884) 9 App Cas 187; and that, having that understanding, he relied on it. This may be of particular significance in the case of implied statements.”

It will be necessary to return to the two nineteenth century cases cited in this passage. But the Board notes now that in neither of them was it said to be a necessary element of a deceit claim that the claimant was aware of the representation and understood it to have the meaning alleged.

[139]In Raiffeisen Christopher Clarke J found that none of the representations alleged by the claimant was in fact made. But he also considered whether the claimant had understood the defendant to be making any of the alleged representations and concluded that it had not.

[140]Counsel for CS Life cited no fewer than 11 later cases in which Raiffeisen has been cited as authority for a requirement of awareness and understanding. In most of these cases the point was mentioned only in passing. But in three of them (all decisions at first instance) the matter has been considered in depth. They are: Marme Inversiones 2007 SL v Natwest Markets plc

[2019] EWHC 366 (Comm)

, paras 278-286; Leeds City Council v Barclays Bank plc

[2021] EWHC 363 (Comm)

;

[2021] QB 1027

, paras 34-153; and Loreley Financing (Jersey) No 30 Ltd v Credit Suisse Securities (Europe) Ltd

[2023] EWHC 2759 (Comm)

, paras 374-425.

[141]In Marme the relevant discussion was obiter because Picken J had already concluded that the implied representations contended for by the claimant had not been made and the need to show contemporaneous awareness of a representation was conceded by the defendant. But there was argument about whether conscious thought is necessary. The judge accepted that the authorities cited to him, starting with Raiffeisen, supported a requirement that a claimant “should have given some contemporaneous conscious thought to the fact that some representations were being impliedly made, even if the precise formulation of those representations may not correspond with what the court subsequently decides that those representations comprised” (para 286). Picken J also accepted a submission that to hold otherwise would substantially “water down” the requirement of showing reliance in implied representation cases and would remove the distinction between actionable non-disclosure and misrepresentation. Only one of the cases to which the Board has referred at paras 130-135 above (MAN Nutzfahrzeuge AG v Freightliner Ltd) appears to have been brought to the court’s attention.

[142]In Leeds the claimants sought rescission of loan agreements on grounds of fraudulent misrepresentation. They alleged that, when entering into the loan agreements, they had relied on representations impliedly made by the defendant bank to the effect that the LIBOR rates used as benchmarks for the interest payable on the loans were being set honestly and properly. The bank applied to strike out the claims and, for the purposes of the application, it was assumed that the alleged representations(a) had been made,

(b) were false and

(c) were made fraudulently. The bank argued that, even on these assumptions, the claims could not succeed because the claimants did not assert that, when entering into the agreements, they were aware (in the sense of giving contemporaneous and conscious thought to the matter) that any of the alleged representations had been made. The bank submitted that as a matter of law proof of such awareness is essential to establish reliance, which itself is an essential element of a claim for misrepresentation.

[143]Cockerill J discussed this question at length and was ultimately persuaded by authority that “there is some requirement of awareness” (para 146). As to the precise nature of the requirement, she said that “in some cases the question will be what the claimant consciously thought, but in other cases it may be better expressed by a focus on active presence” (ie on whether the representation was “actively present” to the claimant’s mind). The judge did not identify a criterion for determining which formulation is more suitable in a given case. She also had difficulty accommodating in her analysis cases such as DPP v Ray and a hypothetical case of a bidder at an auction who raises a paddle and thereby impliedly represents their willingness and ability to pay a certain sum. She recognised that in such cases it would be artificial to attribute conscious awareness of the representation to the waiter or auctioneer and considered that a “quasi-automatic understanding” would suffice (para 148). But such an understanding was still, in her view, different in principle from an assumption on the part of the claimant, which would not suffice, even though in some cases the dividing line might be “thin to non-existent” (para 147).

[144]As these observations indicate, Cockerill J may not have been entirely convinced that her conclusions were correct in principle. Indeed, she said that, had it not been for cases involving similar representations where “the judges involved have said in one form or another that awareness is required”, she “would certainly be tempted to say” that it depended on the precise facts (paras 149-150). Cockerill J nevertheless held that, because the claimants had not pleaded any facts capable of showing that they were consciously aware of the alleged representations or understood them to have been made, the claims had no real prospect of success and should therefore be struck out. The judge gave permission to appeal against her decision, but the case was settled before the appeal was heard.

[145]The judgment in Leeds was the subject of incisive criticism in articles by William Day, “Recent travails of fraudulent misrepresentation” [2021] LMCLQ 636 and Peter de Verneuil Smith QC and William Day, “Reliance: a comparison between the common law and s 90A FSMA” (2021) 6 JIBFL 389. It has been doubted by Clerk & Lindsell on Torts, 24th ed (2023), para 17-37, on the ground that:

“[The] holding seems inconsistent with the jurisprudence on half-truths and misrepresentation by deliberate concealment; furthermore, there seems nothing incoherent in the idea of a party holding, and acting on, an implicit if subconscious belief that there is nothing unusual or untoward about a given transaction.”

[146]In Crossley v Volkswagen AG

[2021] EWHC 3444 (QB)

; [2023] 1 All ER (Comm) 107 Waksman J declined to follow the approach taken in Leeds. In group litigation brought by car purchasers against manufacturers found to have used a “defeat device” during emissions testing for their cars, the manufacturers applied to strike out claims in deceit on the ground that the purchasers could not show that they had been consciously aware of any representation that the cars complied with emissions standards. The judge refused to do so. After a detailed examination of the authorities, he concluded (at para 97) that there are issues raised where implied representations by conduct are alleged which have yet to be fully worked out; and that, given the decisions in Spice Girls, Gordon v Selico and DPP v Ray, the deceit claim had a real prospect of success.

[147]In Loreley Cockerill J was asked to revisit her conclusions in Leeds in the light of these developments. She did so, although on her factual findings the issue did not arise as she found that the representations alleged by the claimant were not made (para 374). After again considering the topic in detail – including now the judgment of Waksman J in Crossley and also Gordon v Selico (the case of covering up dry rot) which had not been cited in Leeds – the judge adhered to her earlier view that, to satisfy the requirement of reliance, the claimant “must be aware of [the representation]/have it actively present to their mind when they act on it” (para 421). She again grappled with “the distinction between representation and assumption” which she thought “is also the division between representation and non-disclosure” and which she saw as particularly acute when dealing with implied representations of honesty (para 422).

[148]In Loreley Cockerill J accepted that there are cases – of which she took Gordon v Selico and Spice Girls to be typical –where reliance is found “without any distinct evidence of understanding or awareness being identified” (para 423). She suggested that there are two hallmarks of such cases which distinguish them from those where the issue of understanding/awareness is “really live”. The first is that “in the Gordon v Selico/Spice Girls type case the representation is simple and cannot well be missed by the representee” (para 423). The second is that, in cases of this type, the representation is “at the heart of the transaction” (para 424). These factors mean that “the question of awareness is one to which the answer is obvious”. Such cases are to be contrasted with complex transactions where “awareness is far from obvious” and “there are real questions as to whether a particular implicit message is received and understood” (para 424).

The dispute in this case

[149]Except for Loreley which had not yet been decided, the cases just mentioned were cited in these proceedings to the courts below. The Chief Justice characterised the implied representations made on behalf of CS Life as “simple but fundamental” and not derived from a complex web of communications or ambiguous in any way (paras 682 and 696). He regarded this case as in the same general category as Gordon v Selico and Spice Girls (paras 683 and 686) such that evidence of contemporaneous awareness and understanding of the representation is unnecessary.

[150]In overruling the Chief Justice on this issue, the Court of Appeal considered that he had misunderstood the significance of the distinction drawn by Cockerill J. She was not suggesting – and it is not the case – that there is a category of case in which awareness and understanding are not required; the distinction between complex and simple transactions goes only to how the requirement is proved (para 259). As put by Bell JA, at para 262:

“It is no doubt true that the more obvious the implication of a representation the more likely it is that the representee would have understood it to be made; but that is not to say that the evidence of the understanding is unnecessary.”

[151]It is worth quoting how the Court of Appeal attempted (at para 261 of Bell JA’s judgment) both to square the Gordon v Selico and Spice Girls type of case with the proposition that awareness and understanding of the representation is an essential element of a claim in deceit and at the same time to distinguish the present case from cases of the Gordon v Selico/Spice Girls type:

“I do not think that this case falls into the [Gordon v Selico] category, where there was an act on the part of the landlord designed to deceive the tenant, and the covering up of the dry rot was a plain indication that there was none; nor does it fall into the Spice Girls category where the promotional material implied that there was no reason to believe that the band was about to lose one of its members. Neither is it in the auction paddle category, where the knowledge and awareness operate in a split second. It seems to me to be much more in the category of assumption, where if Mr Ivanishvili had been asked the question whether he assumed that the relevant Portfolio had not been and was not in the future to be fraudulently managed, he would have said ‘of course’. He had no reason to think otherwise … But that is not to say that he applied his mind to whether CS Life was making any representation to him.”

[152]The plaintiffs submit that the Court of Appeal was wrong to hold that, to establish liability for fraudulent misrepresentation, it is necessary as a matter of law to show that the claimant was consciously aware of the representation and understood it to have the meaning alleged. CS Life argues the contrary and invites the Board to affirm the reasoning and conclusion of the Court of Appeal on this issue.

The suggested distinctions

[153]The Board has given at paras 130-136 above a series of examples of cases of deceit in which it is either unrealistic to suppose that the claimant was consciously aware of the representation made by the defendant or plain that the claimant was unaware of it. The waiter who serves a customer in a restaurant does not stop to think that, by placing an order, the customer is representing that he has the means and intention to pay for the meal before leaving. The waiter just assumes this to be the case. It is an assumption that arises naturally through established social norms and expectations. The same applies to the cab driver who stops to pick up a person who waves her arm to hail the cab or the auctioneer who treats a raised hand as a bid to pay the price asked. Describing the waiter, cab driver or auctioneer in such cases as having a “quasi-automatic” (Leeds, para 148) or “quasireflexive” (Loreley, para 388) understanding that the conduct in question represents an intention and ability to pay should not be allowed to obscure the reality that no conscious thought of that kind usually takes place in such situations. It will not do to say that in these cases there is always a process of awareness and understanding which “is always distinct from assumption”, even though “[it] may look like assumption” (Loreley, para 388), or that “the knowledge and awareness operate in a split second” (see the above quotation from the Court of Appeal judgment). Of course, in these examples the claimant is aware of the defendant’s conduct. But there is no separate awareness of a representation that the defendant intends and has the means to pay for the service or goods which he or she is impliedly offering to purchase. The claimant’s belief that this is so does not just look like an assumption. That is exactly what it is.

[154]The attempt to reconcile cases such as Gordon v Selico with a requirement to show awareness or understanding of the representation is even more forlorn. In cases of this type not only is there no conscious awareness or understanding of the representation being made; the claimant is not even aware of the conduct which gives rise to the representation. Thus, in Gordon v Selico itself the buyer did not know that the seller had covered up dry rot. The buyer’s ignorance of that conduct was essential to the deceit.

[155]It is unclear how in the passage quoted above the Court of Appeal considered that it was distinguishing the facts of this case from cases of the Gordon v Selico and Spice Girls type. In Gordon v Selico the prospective buyer would naturally have assumed that there was no dry rot because he would have seen nothing to suggest that there was. But that is not to say that he applied his mind to whether the seller was making any representation to him about the absence of dry rot. He assuredly did not. Similarly, it would never have crossed the sponsor’s mind in Spice Girls that Geri Halliwell might be about to leave the group or that a contrary representation was being made. These are cases plainly in the category of assumption. They are no different in this respect from the present case where – as Bell JA rightly observed – Mr Ivanishvili would certainly have assumed that the relevant portfolio had not been and was not in the future to be fraudulently managed; but that is not to say that he applied his mind to whether CS Life was making any representation to him.

[156]What about the suggested distinction between cases where the representation is simple and at the heart of the transaction and cases where a representation is said to be implied in a complex transaction? Like the Court of Appeal, the Board does not understand Cockerill J’s suggestion to be that in the former type of case awareness of the representation is unnecessary; but rather that in such cases awareness and understanding of the representation may readily be inferred without the need for evidence specifically dealing with point. But this suggestion, in the Board’s view, conflates the question whether it is obvious that the representation was made with whether it is obvious that the representee would have understood it to be made. In some circumstances the two may go together. But cases of ordering a meal, raising an auction paddle or covering up dry rot are not such cases. They are cases in which it may be obvious that the representation contended for was made but not at all obvious that the representee would have understood it to be made. Indeed, it may be plain that the representee would have had no conscious awareness of the making of the representation.

[157]There is no escaping from the choice that must be made. Either the Board must conclude that all cases of the kind described at paras 130-136 above were wrongly decided; or it must reject the theory which has gained some currency following the dicta in Raiffeisen that contemporaneous awareness and understanding of the representation is in law an essential element of a deceit claim. The Board has no hesitation in adopting the latter course.

[158]What merits further consideration is why it should ever have been thought that there is a legal requirement to show awareness. This idea appears to stem from three misconceptions.

Reliance and awareness

[159]The first is that, unless the claimant can show that it was aware of the representation and understood what representation was being made, it cannot show reliance on the representation, which is an essential element of the cause of action. The point is put in a variety of ways. But they all amount to saying that a misrepresentation cannot cause the claimant to act in a way that results in loss unless it has an impact on the claimant’s mind; and that it cannot have such an impact unless the claimant is consciously aware of the representation.

[160]This is a critical premise underlying Cockerill J’s conclusions in Leeds and Loreley. In Leeds she described awareness of the representation as “the logical bridge” between the representation and reliance (para 70) and said (at para 67) that:

“if the representation was not understood to have been made, or was not understood in the sense relevant for the complaint (but rather in some other sense), then inducement logically cannot be made out.” (Emphasis added.)

In Loreley, para 385, Cockerill J confirmed that:

“At the heart of [the decision in Leeds] is what I saw as a necessary and logical bridge between the representation and inducement (Leeds, at paras 67, 70). The starting point - which is not controversial - is that representation has to cause inducement … and that causative link has to be at least capable of being discerned.”

[161]There is no doubt that reliance or inducement is an essential element of a claim for deceit (or other claim for damages for misrepresentation). There are two aspects to the requirement. The first is that the representation must have deceived the claimant (C) by causing C to hold a false belief (“reliance in belief”). The second is that C must because of holding that false belief have acted so as to suffer loss (“reliance in action”). Both aspects of reliance require the representation to operate on the mind of C. But neither logically requires C to be consciously aware of the representation at the time when C acts on it. Nor is there any good reason to insist on such an additional requirement.

[162]It is an everyday feature of human experience that people form and act on beliefs without any conscious awareness or thought. If someone takes advantage of such unconscious mental processes to deceive another person and cause her to act to her detriment, there is no reason why a claim for damages should not lie. The mischief is no less than in a case involving conscious awareness.

[163]Compare two cab drivers, each of whom is hailed by a person who has formed an intention to run off without paying the fare at the end of the ride. Watson is a normal cab driver who naturally assumes without giving the matter any thought at all that the passenger intends to pay, and who would be just as surprised to learn that he was relying on a representation as Monsieur Jourdain was to be told that for more than 40 years he had been speaking prose. The second cab driver, Holmes, gives evidence that, before stopping, he consciously reasoned that, by hailing a taxi with an illuminated sign saying “For Hire”, the passenger was impliedly representing that she had the means and intention to pay the fare at the end of the journey. Holmes testifies that this representation was “actively present to his mind” when he agreed to drive the passenger to her destination.

[164]It is difficult for a court to determine what conscious thought did or did not occur in the mind of the claimant. A rule of law which allowed Holmes to recover damages only if his evidence about his conscious mental processes is accepted would cause obvious practical and evidential difficulties. But the bigger question is why, even if judges had a window into the claimant’s mind, determining whether such conscious thought took place should matter. In the hypothetical example given, Watson is equally a victim of deceit. The passenger caused him to believe that she intended to pay the fare and to act on that belief by driving her to her requested destination. It would be unreasonable and unworldly – and a charter for fraudsters – if the law were to distinguish between the two cases.

[165]In the taxi driver and other similar examples, the claimant is aware of the conduct which gives rise to the representation but acts on it without conscious thought. Generally, no claim can arise if the claimant was not even aware of the conduct itself, particularly where it takes the form of words. In Leeds, para 70, Cockerill J gave the example of a very clear representation made just at the moment when a meeting participant’s wifi dropped out momentarily or when a pneumatic drill had started up outside her window so that she could not hear anything but that noise for the crucial 30 seconds. In such a case the representation clearly can have no impact, whether conscious or unconscious, on the mind of the claimant. And as Lord Toulson observed in Zurich Insurance Co plc v Hayward

[2017] AC 142

, para 62:

“A misrepresentation which has no impact on the mind of the representee is no more harmful than an arrow which misses the target.”

[166]Yet such examples do not justify the conclusion drawn in Leeds, para 70, that “[i]f there is no awareness of the making of the representation, logically it cannot operate”. The requirement is not one of logic but a feature of the factual situation. Cases such as Gordon v Selico in which defects are deliberately covered up show that it is possible to make a fraudulent misrepresentation that gives rise to liability for deceit without the claimant even being aware of the conduct which gives rise to the representation. Another illustration is the case mentioned earlier of a computer which acts automatically in response to the input of information known by the person entering it to be false. There is no reason in principle why in such a case that person should not be held liable in deceit, even though the claimant is unaware of the making of the representation. The requirement to prove reliance can be met by showing that the claimant assumed and acted on the assumption that the information received by the computer was true, when it was not.

[167]Suggestions that a representation must be “actively present to [the claimant’s] mind” have their source in a dictum of Bowen LJ in Edgington v Fitzmaurice

(1885) 29 Ch D 459

. The Reverend Edgington claimed that he had been induced to buy debentures in a loss-making company by false statements made in a prospectus. The Court of Appeal upheld the decision of the judge that, although the prospectus stated that the money raised by the issue of debentures was to be used to develop the company’s business, the actual intention, as the directors knew, was to use it to meet pressing liabilities. The plaintiff had given evidence, and the judge found, that he had relied on the misstatement. But he had also admitted that he would not have taken the debentures but for a mistaken belief (not induced by the prospectus) that they were secured by a charge over property of the company.

[168]The defendants argued on the appeal that this evidence showed that what caused the plaintiff to advance money was his own mistake and not the misstatement in the prospectus. All the members of the Court of Appeal rejected this argument, holding that, on the facts, there were two operative causes of the plaintiff acting as he did. It was sufficient that he had been induced to take the debentures by the misstatement in the prospectus and did not matter that he had also been induced to do so by his own mistake. In addressing this point, Bowen LJ said, at p 483:

“But such misstatement was material if it was actively present to his mind when he decided to advance his money. The real question is, what was the state of the plaintiff’s mind, and if his mind was disturbed by the misstatement of the defendants, and such disturbance was in part the cause of what he did, the mere fact of his also making a mistake himself could make no difference. It resolves itself into a mere question of fact.”

[169]This passage has been widely cited. But, in the Board’s view, it is wrong to treat the language used as authority for a requirement that, to prove causation, a plaintiff must always show that he gave conscious thought to the representation made. As with any judgment, the language used by Bowen LJ must be read in context, by reference to what was in issue in the case. In Edgington the court was not concerned with whether, to show that he was influenced by a misrepresentation, it is always necessary for the plaintiff to prove that he gave conscious thought to it. The issue was whether, in a case where the plaintiff would not have entered into the relevant transaction had he not made a mistake for which the defendant was not responsible, the defendant may nevertheless be held liable in an action for deceit. The answer given by the Court of Appeal was “yes” and that the only relevant question was whether the misrepresentation had in fact caused the plaintiff to act as he did; it did not have to be the sole operative cause. This was the only point that Bowen LJ was addressing in the passage quoted. It is not an authority on the question whether conscious awareness is logically or legally necessary for liability in deceit.

Ambiguous representations

[170]There are certainly cases in which, to establish that the defendant’s words and/or conduct caused the claimant to hold a false belief, it will in practice be necessary to show that the claimant understood them to convey a particular meaning. This is so whenever the meaning is unclear or ambiguous and the representation is false only if it bears one particular meaning. The two nineteenth century cases cited by Christopher Clarke J in Raiffeisen are examples. In Smith v Chadwick

(1884) 9 App Cas 187

a company prospectus contained a statement which, in the view of a majority of the House of Lords, was capable of two meanings; on one meaning the statement was true and on the other meaning it was false. In an action for deceit alleging that he had been induced by fraudulent representation to purchase shares in the company, the plaintiff was not asked and gave no evidence at the trial about what interpretation he had put on the statement. The majority of the House of Lords held that in these circumstances the claim must fail because, to establish that he had been deceived into taking the shares, the plaintiff needed to prove that he had understood the words in the prospectus in the sense in which they were false, which he had not done.

[171]The same point arose in Arkwright v Newbold

(1881) 17 Ch D 301

, 324-325, where Cotton LJ said:

“In my opinion it would not be right in an action of deceit to give a plaintiff relief on the ground that a particular statement, according to the construction put on it by the court, is false, when the plaintiff does not venture to swear that he understood the statement in the sense which the court puts on it. If he did not, then, even if that construction may have been falsified by the facts, he was not deceived.”

[172]This statement, in the Board’s view, is not quite an accurate statement of the law because what matters in a claim for deceit is not whether the plaintiff understood the defendant’s statement according to the construction put on it by the court, but whether the plaintiff understood the statement in the sense that the defendant intended the plaintiff to understand it (and knew to be false or at least did not believe to be true): see eg Akerhielm v de Mare

[1959] AC 789

, 805. How the court interprets the statement may well be relevant to the court’s assessment of how either or both parties probably understood it. But it is the parties’ subjective beliefs that are critical. That said, in a case such as Arkwright v Newbold where a particular statement made is false only if interpreted in a particular non-obvious sense, it is clearly correct that to prove “reliance in belief” the plaintiff must show that it understood the statement in that sense.

[173]The mistake made in Raiffeisen and cases which have followed it has been to treat what in practice needs to be shown to establish deceit in particular factual circumstances as if it were a necessary element of the cause of action.

Representations and assumptions

[174]As well as the notion that reliance on a representation is never possible without conscious awareness, another misconception apparent in the judgments in Marme, Leeds and Loreley is that a distinction needs to be drawn between cases where the claimant has relied on a representation and cases where the claimant has acted on an assumption. The idea is that, if the claimant has acted on an assumption, it cannot be said that the defendant’s representation caused the claimant to hold a false belief. The source of the error must lie with the claimant.

[175]Yet this is a false dichotomy. The categories of representation and assumption are not mutually exclusive. It is possible, and indeed common, for a person to act on the basis of an unconscious assumption and in reliance on a representation. The waiter who brings food to the table assumes without thinking that the diners who ordered the food intend to pay for it. But that belief has been caused by the conduct of the diners when they sat down at the table and placed their order. One way of describing such situations is to say that they involve exploiting an unconscious assumption of the claimant. A defendant who speaks or acts with the intention of causing the claimant to make and act on an erroneous assumption so as to suffer loss is just as much liable for deceit as a defendant who brings about that result by operating on the conscious mind of the claimant.

[176]What matters is whether, in a case where the claimant has acted on an assumption, the assumption was one which the claimant would naturally be expected to make in response to the defendant’s words or actions or whether it was one made independently by the claimant. If the claimant has acted as a result of an erroneous belief not caused by the defendant, the defendant will not be liable.

Misrepresentation and non-disclosure

[177]A related misconception is that requiring awareness of representation is necessary to preserve the distinction between misrepresentation and non-disclosure. This idea appears to have originated in Marme (para 286) and to have been accepted in Leeds (para 65) and Loreley (para 422). The distinction between misrepresentation and non-disclosure – which reflects that between acts and omissions – is an important one because, except in those cases (such as the formation of contracts of insurance) where there is a duty to disclose material facts, non-disclosure does not give rise to liability.

[178]The distinction turns on whether the defendant(1) has done something to cause the claimant to hold a false belief on which the claimant has acted to its detriment or

(2) has merely failed to inform the claimant of a material fact or to correct a false belief which the claimant independently holds. A case may fall in the first category without the claimant being aware of what the defendant has done, as Gordon v Selico and other examples mentioned earlier show. Such ignorance does not turn the case into one of nondisclosure. The seller who takes active steps to conceal a defect in order that a buyer should not discover it stands in a different position from the seller who is aware of a defect not apparent to the buyer but does nothing actively to hide it. The line is not always easy to draw. But it depends entirely on what the defendant has or has not done and not at all on the claimant’s awareness or understanding of acts done by the defendant.

Conclusion on the alleged need to show awareness

[179]The Board concludes that under the law of England and Wales and of Bermuda it is not a legal requirement of a claim for deceit that the claimant was aware of the representation or understood it to have been made. The Court of Appeal was wrong so to hold. The Chief Justice was entitled to find that, by proposing the LPI policies, Mr Lescaudron acting on behalf of CS Life induced Mr Ivanishvili to believe that Mr Lescaudron did not intend to manage the policy assets fraudulently and, on the strength of that belief, to arrange for the plaintiffs to enter into the policies. It was unnecessary for the plaintiffs to plead or prove that Mr Ivanishvili had a conscious awareness of the representations impliedly made to him.

[180]CS Life has not sought to argue before the Board, and did not argue in the Court of Appeal, that the Chief Justice erred for any reason other than this alleged error of law in finding that the misrepresentation claim is actionable under Bermudian law. It follows that the Court of Appeal should have affirmed that finding and held that the first limb of the double actionability rule is satisfied.

Liability under Georgian law

[181]The Board therefore turns to the second limb, which requires the act constituting the alleged tort to be actionable under the law of the place where it occurred. As noted earlier (see para 124 above), it is agreed that this place was Georgia. The Chief Justice found that – subject to the question of limitation – the plaintiffs have a good claim under Georgian law. That finding has not been appealed. What remains in issue is whether the Georgian law of limitation applies to the claim. It is agreed that, if it does, the claim was brought outside the limitation period of three years prescribed by Georgian law.

[182]Under the common law, foreign rules of limitation apply only if classified as substantive rather than procedural. But in Bermuda, as in England and Wales, the position is now governed by legislation. Section 34A of the Bermudian Limitation Act 1984, which is equivalent to section 1 of the English Foreign Limitation Periods Act 1984, states:

“Application of foreign limitation law 34A (1) Subject to the following provisions of this Part, where in any action or proceedings in a court in Bermuda the law of any other country falls (in accordance with rules of private international law applicable by any such court) to be taken into account in the determination of any matter – (a) the law of that other country relating to limitation shall apply in respect of that matter for the purposes of the action or proceedings; and (b) except where that matter falls within subsection (2), the law of Bermuda relating to limitation shall not so apply. (2) A matter falls within this subsection if it is a matter in the determination of which both the law of Bermuda and the law of some other country fall to be taken into account. … (5) In this section ‘law’, in relation to any country, shall not include rules of private international law applicable by the courts of that country or, in the case of Bermuda, this Part.”

[183]As can be seen, section 34A(1) establishes a general rule that, where foreign law applies to a claim, that country’s law of limitation also applies. Section 34D makes it expressly clear that this is so whether the foreign rule of limitation is classified as procedural or substantive. Section 34A(2) is clearly directed to the double actionability rule in tort, which (so far as the Board is aware) is the only context in which “both the law of Bermuda and the law of some other country fall to be taken into account”. The result is that, where the double actionability rule applies, the claim is barred if it is brought after the expiry of either the limitation period prescribed by Bermudian law or the limitation period prescribed by the law of the country where the act constituting the tort occurred (whichever is the shorter). Thus, in Sophocleous (referred to at para 119 above), pursuant to the equivalent English statutory provision both the Cypriot and English rules of limitation applied, and the claims failed because they were brought after the expiry of the relevant limitation period under Cypriot law.

[184]On its face, therefore, section 34A of the Limitation Act leads to the conclusion that the misrepresentation claim made in this case is time-barred because it was brought after the Georgian limitation period had expired.

[185]To seek to avoid that conclusion, the plaintiffs make three, alternative arguments. The Chief Justice accepted all of them, but they were all rejected by the Court of Appeal. All three alternative arguments are renewed before the Board and give rise to the three remaining issues on the cross-appeal. In the Board’s opinion, all three arguments lack merit. (1) The exception to the double actionability rule

[186]First, the plaintiffs contend that the “flexible exception” to the double actionability rule (described at para 118 above) should be applied in this case such that only the law of Bermuda and not that of Georgia applies to the misrepresentation claim.

[187]The Chief Justice accepted this argument despite having found that “the substance of the tort did not take place in Bermuda” and that “it is not obvious that any of the constituent elements of the tort of misrepresentation took place in Bermuda” (para 588). Adopting the test used in Imanagement, he nonetheless decided that “all or almost all of the significant connecting factors point to the direction of Bermuda” (para 589), so that there is no need to show that the misrepresentation claim is actionable under the law of Georgia.

[188]The Board agrees with the Court of Appeal that this ruling cannot be supported. The assertion that “all or almost all of the significant connecting factors point to the direction of Bermuda” is inconsistent with the Chief Justice’s own, clearly correct finding that “the substance of the tort did not take place in Bermuda” (para 588). In fact, the only significant connection with Bermuda is that CS Life is incorporated there. All other significant connecting factors point to other countries, particularly Georgia.

[189]For a start, none of the plaintiffs is or was situated in Bermuda: Mr Ivanishvili is domiciled and resident in Georgia, Meadowsweet is incorporated in the British Virgin Islands and Sandcay is incorporated in the Bahamas. The Chief Justice said that Meadowsweet and Sandcay entered into the LPI policies with CS Life in Bermuda (para 587). But that statement is contrary to his own findings that Mr Ivanishvili signed the application forms for the policies in Georgia (para 575) and that the policies were signed on behalf of CS Life and issued in Switzerland (para 588). Although the Chief Justice attached importance to the fact that the policies are governed (through the parties’ choice) by Bermudian law, it is hard to see how this is relevant to the deceit claim. Far more significant is the fact that the fraudulent misrepresentations were made by Mr Lescaudron (on behalf of CS Life) to Mr Ivanishvili (on behalf of the prospective policyholders) in Georgia. Nor can it be said that any of the fraudulent conduct which caused the plaintiffs to suffer loss occurred in Bermuda or that the loss was suffered in Bermuda. Mr Lescaudron was based in Switzerland, as were all the Bank employees who operated the policy accounts; and the accounts on which the losses were suffered were held with the Bank in Switzerland. It was not even suggested at the trial that Mr Lescaudron or Mr Ivanishvili had ever visited Bermuda (see para 575 of the judgment).

[190]In these circumstances there is no justification for applying the exception to the double actionability rule and declining to apply the law of Georgia as the law of the place where the tort occurred. (2) Renvoi

[191]The next argument advanced by the plaintiffs to avoid the limitation period under Georgian law relies on the doctrine of private international law known as renvoi (meaning “re-sending” or “referral”). The core idea is that, when an issue is governed by the law of a foreign country, the court should seek to decide the issue in the same way as a court of that country would decide it. To achieve this result, the court should look first, not to the internal law of the foreign country, but to its rules of private international law. If those rules would require a court of the foreign country to apply a different system of law from its own, then that law should be applied. Thus, as Lord Collins of Mapesbury explained in Dallah Real Estate and Tourism Holding Co v Ministry of Religious Affairs of the Government of Pakistan

[2010] UKSC 46

;

[2011] 1 AC 763

, para 124: “Renvoi is concerned with what happens when the … court refers an issue to a foreign system of law … and where under that country’s conflict of laws rules the issue is referred to another country’s law.”

[192]At the trial both experts on Georgian law agreed that, if the misrepresentation claim were brought in Georgia, then under that country’s conflict of laws rules (ie its rules of private international law) the claim, including the issue of limitation, would be regarded as governed by the law of Bermuda (paras 591-593). The plaintiffs argue that in these circumstances the Bermuda courts, applying the doctrine of renvoi, should likewise treat the misrepresentation claim, including the question of limitation, as governed solely by Bermuda law.

Renvoi does not apply in tort cases

[193]A fatal objection to this argument is that the law of England and Wales, from which Bermuda law is not said to differ, has never applied the doctrine of renvoi in the field of tort (or any other branch of the law of obligations). Briggs on Private International Law in English Courts, 2nd ed (2023), para 3.77, observes that:

“Renvoi was generally assumed to play no part in the private international law rules for tort, where the common law rules of private international law identified the lex loci delicti commissi, the law of the place of the tort, with the domestic law of the place of the tort.”

To similar effect, Dicey, Morris & Collins on The Conflict of Laws, 16th ed (2022), para 35-127, states:

“It has never been suggested in the English case law that the lex loci delicti means anything other than the domestic rules of that law. In other words, the doctrine of renvoi does not apply …”

See also Cheshire, North & Fawcett on Private International Law, 15th ed (2017), pp 69- 70:

“In countless cases dealing with such matters as torts … the English courts, when referred to ‘the law’ of a foreign country, have never had the slightest hesitation in applying the internal law of that country.”

[194]The only case cited to the Board in which the notion of applying the private international law rules, rather than the domestic rules, of the place of the tort appears even to have been mentioned is a decision of the Scottish Court of Session Inner House, M’Elroy v M’Allister, 1949 SC 110, 126, where Lord Russell in an obiter dictum thought it desirable to note that:

“in referring to the lex loci delicti to ascertain by what rules the rights and liabilities of the parties to this action are there regulated this court refers to the internal domestic law of that locus and not to its private international law.”

This has generally been taken to be an accurate statement of English law.

[195]The UK Parliament has also rejected any suggestion that renvoi should operate in the field of tort. The Law Commission report (see para 120 above) which led to Part III of the Private International Law (Miscellaneous Provisions) Act 1995, at para 3.56, recommended that renvoi should be excluded, as it would create uncertainty and would not accord with the reasonable expectations of the parties. That recommendation was implemented by section 9(5) of the 1995 Act. Renvoi is also excluded by article 24 of the Rome II Regulation.

[196]It is not only in the field of tort that the doctrine of renvoi has failed to gain any foothold in the law of England and Wales. According to Cheshire, North & Fawcett on Private International Law, 15th ed (2017), p 69:

“Its scope appears to be limited to certain matters concerning either status or the disposition of property on death”

. Similarly, Dicey, Morris & Collins on The Conflict of Laws, 16th ed (2022), para 2-116, records that, outside the field of succession, the doctrine seems to have been applied in England only to legitimation by subsequent marriage. In Macmillan Inc v Bishopsgate Investment Trust plc (No 3)

[1995] 1 WLR 978

, 1008, Millett J observed that the doctrine of renvoi “has not been applied in contract or other commercial situations. It has often been criticised, and it is probably right to describe it as largely discredited”.

Reasons for rejecting renvoi

[197]Some of the reasons for rejecting the doctrine are practical. Renvoi makes it far harder to determine what country’s internal law applies to a claim. Instead of selecting the law applicable to the substantive issue, the forum’s choice of law rule is treated merely as a starting point which identifies a country whose choice of law rules are to be applied. Whenever that country is a foreign one, evidence of its rules of private international law will therefore be required. If the aim is to do exactly what a court of that country would do, this inquiry will need to include its rules as to renvoi. (This doctrine of “double” or “total” renvoi – as opposed to “single” or “partial” renvoi, in which the foreign rules of renvoi are left out of account – is indeed the version that English courts have applied in those few cases where renvoi has been applied at all: see Dicey, Morris & Collins, para 2-114.) This increases the complexity and uncertainty of the inquiry, as renvoi is one of the more esoteric areas of private international law and whether, when or how it is recognised by a foreign legal system may be obscure.

[198]If the foreign country’s rules of private international law dictate that its own internal law applies to the claim, then the law ultimately applied will be the same as if the forum court had not applied the doctrine of renvoi. The only effect of considering the foreign rules of private international law in such a case is thus to make the inquiry more complicated and opaque than it would otherwise have been. Renvoi may affect the outcome only if the foreign court would apply a law other than its own. If that law is the law of a third country and the foreign court would itself apply renvoi, then that third country’s rules of private international law must in turn be investigated (including potentially its approach to renvoi). If the law which the foreign court would select is the law of the forum court and the foreign court would apply total renvoi, a stalemate is created (or, as academic writers have called it, a “circulus inextricabilis”: see Dicey, Morris & Collins, para 2-131). In this situation it is impossible to decide the case just as the foreign court would decide it and the only solution is to abandon renvoi – though whether at this point the forum court applies its own internal law or that of the foreign country appears arbitrary.

[199]The possibility of such a stalemate shows that the ideal of applying the same law as the foreign court would apply is impossible to realise in all cases if the foreign court would adopt the same approach. Total renvoi is in this sense self-defeating because it cannot work if everyone adopts it. If both the forum court and the foreign court adhere to the doctrine, then deciding how the foreign court would decide the case collides with the foreign court’s determination to decide the case according to how the forum court would decide it. No decision can be reached if each court defers to the other.

[200]More fundamentally, it may be wondered why it should ever be thought desirable to subordinate the choice of applicable law to another country’s choice of law rules. The basic reason for applying foreign law, it might be thought, is that the case is judged to be more closely connected with the foreign jurisdiction than with the forum. A primary purpose of private international law rules is to specify the criteria by which that judgment is to be made. It is hard to see why, rather than applying whatever rules and policies the legislature or courts of the forum have decided are appropriate for this purpose, the courts of the forum should defer to the rules chosen by another country’s legal institutions, whatever those rules and the policies underlying them might be. This appears all the more objectionable when, as discussed, the strategy only works if it is not reciprocated.

[201]In light of these practical and theoretical objections, the Board has no difficulty in agreeing with the view of Beatson J in Blue Sky One Ltd v Mahan Air

[2010] EWHC 631 (Comm)

, paras 174-185, that renvoi should not be adopted in English (or Bermudian) law when its application is not required by binding precedent. To do so would introduce considerable complexity and uncertainty into the law for no good reason.

The Australian approach in Neilson

[202]Although it is not an authority on which the plaintiffs rely, and Beatson J in Blue Sky One declined to follow it, it is right to note that renvoi was applied to a claim in tort by the High Court of Australia in Neilson v Overseas Projects Corpn of Victoria Ltd [2005] HCA 54;

(2005) 223 CLR 331

. That decision has proved controversial in Australia. In Neilson both parties were Australian residents, but the tort occurred in the People’s Republic of China under whose law the relevant limitation period was one year. If that law had been applied, the claim (for compensation for personal injury) would have been time-barred. Commentators have suggested that the High Court resorted to renvoi as a way of avoiding what it regarded as an undesirable result, when a better way of doing so would have been to allow the court exceptionally to apply a law other than the law of the place where the tort occurred where there are sufficient connections with another country. The High Court did not feel able to adopt that solution, as in earlier cases it had steadfastly refused to admit a flexible exception to the general rule. See Reid Mortensen, “‘Troublesome and Obscure’: The Renewal of Renvoi in Australia” (2006) 2 J Priv Int L 1; Elsabe Schoeman, “Renvoi: Throwing (and Catching) the Boomerang” (2006) 25(1) UQLJ 203; Anthony Gray, “The Rise of Renvoi in Australia: Creating the Theoretical Framework” (2007) 30(1) UNSWLJ 103.

[203]Without expressing any opinion on the merits of the decision in the context of Australian law, the Board does not regard Neilson as persuasive authority in this jurisdiction.

The plaintiffs’ argument

[204]Counsel for the plaintiffs have not tried to advocate the acceptance of renvoi as a general principle. They recognise that, under the common law, renvoi has been applied only in exceptional cases. They have advanced a narrower argument, tied to the operation of the double actionability rule and the object of that rule.

[205]The argument seeks to build on an academic article by Professor Adrian Briggs entitled “In Praise and Defence of Renvoi” (1998) 47 ICLQ 877. In this article Professor Briggs contended that the purpose of the second limb of the double actionability rule, as explained in Boys v Chaplin by Lord Wilberforce, was to prevent “forum shopping”, and that this object would be most fully achieved by applying the choice of law rules, and not merely the internal law, of the place where the tort occurred: in other words, by applying renvoi.

[206]In Boys v Chaplin each of the five law lords stated the law differently; but it is the speech of Lord Wilberforce which has generally been treated as authoritative. On his formulation, the general rule with regard to foreign torts required actionability as a tort according to English law, subject to the condition that civil liability also exists under the law of the foreign country where the act was done (see p 389F). In justifying this condition, Lord Wilberforce referred to two examples of foreign tort laws which were more restrictive than English law: an Ontario statute which precluded passengers in a car (treated as guests) from obtaining a remedy against the driver (host); and the provision of the Maltese Civil Code relied on by the defendant in Boys v Chaplin itself, which permitted only financial losses and not compensation for pain and suffering to be recovered as damages for personal injury. Lord Wilberforce said, at p 389E-F:

“I can see no case for allowing one resident of Ontario to sue another in the English courts for damages sustained in Ontario as a passenger in the other’s car, or one Maltese resident to sue another in the English courts for damages in respect of pain and suffering caused by an accident in Malta.”

[207]In his article, at p 879, Professor Briggs argued that this logic pointed to applying renvoi. The examples given by Lord Wilberforce made it clear that the purpose of requiring actionability under the law of the foreign country where the act was done was to prevent forum shopping. That purpose would only be completely achieved if the law of the foreign country where the act was done meant the law which a court of the country would itself have applied, as distinct from its domestic law. Professor Briggs wrote:

“If this were so, renvoi and the new choice of law rules would provide a good and sturdy defence against forum shopping. For if an English court would give a plaintiff only what a Maltese judge would give, there would be no incentive to forum shop; if an English court would do something different from what a Maltese judge would do, the incentive might still remain.”

[208]Professor Briggs did not go so far as to claim that Boys v Chaplin is direct authority for applying renvoi. The broader theme of his article was that choice of law rules ought not to be viewed separately from rules governing territorial jurisdiction: their interrelationship should be considered. In the absence of a more direct means of preventing claims from being litigated in England when another forum would be more appropriate, renvoi may achieve a similar result. It does so by ensuring that the English court applies the law which a court of the natural or appropriate forum would apply.

[209]Professor Briggs saw this, however, as a path not taken by English law. His suggestion was only that Boys v Chaplin contained the seeds from which such an approach might have developed. But, he said, “before anyone had time to think this through”, that possibility was pre-empted by the decision of the House of Lords in Spiliada Maritime Corpn v Cansulex Ltd (The Spiliada)

[1987] AC 460

, which provided a direct means of controlling forum shopping through the doctrine of forum non conveniens. This meant that:

“control of forum shopping did not depend on choice of law. Forum shopping plaintiffs in tort cases could be dealt with directly: they could, on the defendant’s application, be sent to the foreign court from which they had sought to escape.”

Professor Briggs acknowledged that, in those circumstances, “it may have been acceptable” to interpret the law of the foreign country where the tort occurred as referring to the domestic law of that country (p 879).

[210]The plaintiffs’ argument in this case is more ambitious. They urge the Board to follow the path not taken by English law and to interpret the law referred to in the second limb of the double actionability rule as the private international law of the foreign country where the tort occurred, and not the domestic law of that country. This is said to be necessary to give effect to the object of ensuring that there is nothing to be gained by forum shopping.

[211]In the Board’s view, this would involve an illegitimate development of the common law. It is beyond question that, when members of the House of Lords in Boys v Chaplin referred to “the lex loci delicti”, or the law of the foreign country where the act was done, they meant the domestic law of that country, and not its rules of private international law. In Boys v Chaplin the plaintiff claimed damages in England for serious injuries sustained in a road traffic accident in Malta, caused by the admitted negligence of the defendant. The premise on which the entire argument took place was that, if the court were to apply the law of Malta where the tort occurred, the plaintiff could not recover damages for pain and suffering. This assumed that the relevant Maltese law was its domestic law contained in the Maltese Civil Code. Both parties were normally resident in England but were stationed in Malta at the relevant time as members of the British armed forces. There were therefore significant connections with England and Wales. Yet there was no consideration at all of what country’s law a Maltese court would have applied if the action had been brought in Malta. Renvoi played no part in the reasoning or decision.

[212]The Board is in any case not persuaded that the speech of Lord Wilberforce in Boys v Chaplin provides any support for developing the law to incorporate renvoi. When Lord Wilberforce said that one Maltese resident should not be allowed to sue another in the English courts for damages for pain and suffering caused by an accident in Malta and gave a similar example involving residents of Ontario where one suffered injury as a passenger in the other’s car, he was making a simple point. He was saying that a claim brought in England should not succeed if liability did not exist under the domestic law of the place where the tort occurred. Lord Wilberforce did not contemplate the possibility that the rules of private international law applicable in Malta or Ontario might be relevant.

[213]Nor, in the Board’s view, is it necessary to make the law any more complicated to remove the incentive to forum shop. That incentive is removed, so far as choice of law rules can do so, by requiring the plaintiff to establish civil liability under the domestic law of the foreign country where the tort occurred. If that condition is applied, the foreign resident in Lord Wilberforce’s examples cannot achieve a better outcome by suing in England. It is unnecessary for the English court to apply the foreign country’s rules of private international law to eliminate any incentive to mount such a forum shopping expedition, as Professor Briggs suggested in his article.

[214]The only situation in which applying Maltese rules of private international law might make a difference is one where a Maltese judge would apply a law which is more generous than Maltese domestic law (eg by allowing recovery of damages for pain and suffering). But in such a case there would be no incentive to forum shop by suing in England regardless of whether renvoi is applied. If renvoi were applied, the English court would award the same damages as a Maltese court would award. If renvoi is not applied, the plaintiff would actually be worse off suing in England rather than in Malta.

[215]It therefore appears to the Board that the condition formulated by Lord Wilberforce which requires civil liability under the domestic law of the place where the tort occurred is sufficient to avert the prospect that concerned him: the arrival of the Maltese forum shopper coming to England to sue, hoping thereby to avoid the exclusion of damages for pain and suffering, or the Ontarian hoping to avoid the exclusion in Ontario law of claims by passengers against drivers.

[216]But even if it were right that renvoi would provide a greater deterrent to forum shopping, there remains the point made by Professor Briggs that adopting the doctrine of forum non conveniens removed any need to resort to renvoi for this purpose. As The Spiliada is good law in Bermuda, forum shopping can be dealt with directly. The most that could be said is that, if the doctrine of forum non conveniens had not been developed to control forum shopping, an alternative way in which the common law might have sought to do so indirectly in the field of tort would have been to adopt renvoi. As it is, there is no need to have recourse to such an indirect approach. Concern to prevent forum shopping provides no possible reason to take the step, unprecedented in English and Bermudian law, of introducing renvoi to the choice of law rules for tort - a step that would also be inconsistent with the legislative policy in the United Kingdom (see para 195 above).

[217]The Board concludes that there is no scope for renvoi in this case and that the only rules of Georgian law which are to be taken into account in determining whether the misrepresentation claim is actionable under Georgian law are its internal rules of law and not its choice of law rules.

Legislative exclusion of renvoi

[218]Given this conclusion, it is unnecessary to decide whether two alternative arguments relied on by CS Life and accepted by the Court of Appeal are well founded. The first is that, even if the second limb of the double actionability rule required the application of renvoi, this would not actually help the plaintiffs. For, as the Court of Appeal held (paras 241-244), a Georgian court would not in fact apply the Bermudian limitation period but would apply Georgia’s own limitation period of three years. This is because the Georgian rules of private international law (as agreed by the expert witnesses) include the doctrine of renvoi; a Georgian court applying Bermudian law would therefore apply Bermudian choice of law rules; but the Georgian renvoi rules are to the effect that, if the foreign choice of law rules would refer the claim back to the law of Georgia (as the double actionability rule does), Georgian domestic law would be applied.

[219]The second alternative argument, again accepted by the Court of Appeal, is that, even if the second limb of the double actionability rule requires the application of renvoi, section 34A of the Bermudian Limitation Act 1984 excludes renvoi as regards the law of limitation. CS Life submits that this is the effect of section 34A(5), quoted at para 182 above and equivalent to section 1(5) of the English Foreign Limitation Periods Act 1984, which defines the term “law” in section 34A so that it does not include rules of private international law. This means that, where a foreign law applies in proceedings in Bermuda, section 34A requires only the internal law of that country relating to limitation and not its rules of private international law to be applied.

[220]The plaintiffs accept this so far as it goes. But they argue that the “law” of another country referred to in section 34A(1) is, on the express wording of that provision read with the definition in section 34A(5), the internal law of the country identified by applying the Bermudian rules of private international law. So in a field where those rules of private international law include the doctrine of renvoi, it is the internal law identified by applying that doctrine that is to be applied - both in relation to the substance of the dispute and, pursuant to section 34A(1)(a), in relation to limitation.

[221]The plaintiffs’ interpretation is consistent with the language of the legislation and with the explanation given by the Law Commission in its report on Classification of Limitation in Private International Law 1982 (Law Com No 114), paras 4.33-4.34, of the mischief to which the Foreign Limitation Periods Act 1984 was directed; see also PA Stone, “Time limitation in the English conflict of laws” [1985] LMCLQ 497, 506-507. It is, however, unnecessary to decide this point. For what is clear on either view is that, in any field where the Bermudian rules of private international law do not include renvoi, it is the internal law as to limitation of the country whose law governs the substance of the dispute that applies, and the choice of law rules of that country are to be ignored. Since, as discussed above, the second limb of the double actionability rule looks to the internal law of the country where the tort was committed (here Georgia) and not to its choice of law rules, it is the Georgian internal law relating to limitation that applies pursuant to section 34A(1).

Public policy

[222]A yet further argument advanced by the plaintiffs is that this conclusion is displaced by section 34B of the Limitation Act 1984 (equivalent to section 2 of the English Foreign Limitation Periods Act 1984). This provides:

“(1) In any case in which the application of section 34A would to any extent conflict (whether under subsection (2) or otherwise) with public policy, that section shall not apply to the extent that its application would so conflict.”

The plaintiffs argue that it would conflict with public policy to apply a limitation provision that neither the Bermudian court nor the Georgian court would apply in a domestic case.

[223]The short answer to this argument is that in a domestic case the Georgian court would apply a three-year limitation period and the fact that this period is shorter than the Bermudian limitation period does not make its application contrary to public policy. Periods of limitation vary significantly between countries and it is inherent in a rule that requires another country’s law relating to limitation to be applied that a claim may become time-barred much sooner than would be the case under the internal law of limitation of the forum.

[224]What the plaintiffs mean by a “domestic case” so far as a Georgian court is concerned is presumably a case which is not in fact purely domestic but has a foreign element such as would cause a Georgian court applying the Georgian rules of private international law to apply a foreign limitation provision. The suggestion then is that it is contrary to public policy to apply a limitation provision which(a) a Bermudian court would not apply in so far as its own internal law of limitation is applicable and

(b) a Georgian court would not apply if it applied the Georgian rules of private international law. But there is no public policy which this offends. Since the possibility of such a situation arising could be avoided only by applying renvoi, the suggestion amounts to an assertion that it is contrary to public policy not to operate the doctrine of renvoi. Even the most enthusiastic proponent of renvoi could not make such an unrealistic claim.

[225]For all these reasons, the attempt to avoid the effect of the Georgian time bar by invoking the doctrine of renvoi is flawed. (3) Permission to amend out of time

[226]The third route by which the plaintiffs seek to avoid the time bar relies on Bermudian rules of procedure. In Bermuda as in England and Wales, an amendment to a writ or statement of claim is treated as “relating back” to the date when the action was begun. For this reason, a new claim may not normally be added by amendment after the expiry of a limitation period current when the action was commenced. Allowing the amendment circumvents the limitation period because the new claim is then treated as if it had been made when the action was begun and therefore brought in time. Exceptionally, however, the court has power to allow such an amendment even though its effect is to add a new cause of action if “the new cause of action arises out of the same facts or substantially the same facts as a cause of action in respect of which relief has already been claimed in the action by the party applying for leave to make the amendment”: see the Rules of the Supreme Court 1985 (“RSC”) Order 20, rule 5(5), which corresponds to what is now CPR r 17.4 in England and Wales.

[227]When the plaintiffs applied to the court in October 2020 to amend their writ and statement of claim to add the misrepresentation claim, the Chief Justice did not at that stage decide whether to allow the application. Had he given permission to amend, the result – as just explained – would have been to treat the misrepresentation claim as if it had been made when the action was begun (17 August 2017), which is agreed to have been before the three-year Georgian limitation period had expired. Instead, the Chief Justice adjourned the amendment application to the trial and decided it at the same time as the substantive issues in the proceedings - including those issues which were contingent on permission to amend being given to plead the misrepresentation claim.

[228]When the Chief Justice addressed the question in his judgment, he held that the misrepresentation claim arose out of the same facts or substantially the same facts as the causes of action in respect of which relief had already been claimed in the original statement of claim (para 633). On this basis he decided that the court had power under RSC Order 20, rule 5(5) to grant permission to amend the statement of claim even if the Georgian law of limitation applied, and he granted permission to amend. The effect was to treat the claim as having been brought in time.

[229]The Court of Appeal reversed that decision, and in the Board’s opinion they were right to do so. The proposition that the misrepresentation claim arose out of the same facts or substantially the same facts as the claims for breach of contract or fiduciary duty is unsustainable.

Comparing the causes of action

[230]As the Court of Appeal noted, at para 251, the exercise of deciding whether a new cause of action arises out of the same facts or substantially the same facts is generally to be determined by examination of the pleadings alone. In analysing the pleadings, it is the material facts necessary to formulate the cause of action (or, where relevant, a defence) that matter. Background facts and matters of evidence – which are pleaded all too often although this is contrary to good practice – have no part to play in the analysis.

[231]Further:

“The selection of the material facts to define the cause of action must be made at the highest level of abstraction.”

Paragon Finance plc v D B Thakerar & Co

[1999] 1 All ER 400

, 405F (Millett LJ). Thus, as stated by Robert Walker LJ in Smith v Henniker-Major & Co

[2002] EWCA Civ 762

;

[2003] Ch 182

, para 96:

“in identifying a new cause of action the bare minimum of essential facts abstracted from the original pleading is to be compared with the minimum as it would be constituted under the amended pleading.”

Or as David Richards J put it in Revenue and Customs Comrs v Begum

[2010] EWHC 1799 (Ch)

, para 32:

“The exercise to be undertaken … is therefore to compare the essential factual elements in a cause of action already pleaded with the essential factual elements in the cause of action as proposed.”

[232]Where damages are claimed for breach of contract, the essential factual elements in the cause of action are:(1) the making of the contract between the parties (C and D);

(2) the term(s) of the contract which D has allegedly breached;

(3) facts constituting a breach of such term(s) by D; and

(4) the fact that the breach caused C to suffer loss. In this case the plaintiffs pleaded these essential facts in their original statement of claim seeking damages for losses caused by breaches of obligations owed to them by CS Life under the two LPI policies. (The further allegations of breach of fiduciary duty add nothing material to the analysis.)

[233]These essential factual elements have only to be compared with the essential factual elements of the cause of action in deceit to see that they are completely different. As discussed earlier, the essential factual elements which must be proved in a claim for deceit are:(1) that a representation was made to C by or on behalf of D;

(2) that the representation was false;

(3) that it was made fraudulently in that D did not believe it to be true;

(4) that D intended C to rely on the representation; and

(5) that C did rely on it (in the sense discussed above) so as to suffer loss. None of these elements is an essential factual element in a cause of action claiming damages for breach of contract.

[234]The Chief Justice said he was satisfied that “the facts underpinning the misrepresentation claim were pleaded in the original [statement of claim], including the sale of the policies, the Bank’s role as CS Life’s agent … and Mr Lescaudron’s fraudulent conduct and wrongful trading on the plaintiffs’ accounts …” (para 633). Yet conspicuously lacking from the original statement of claim was any allegation that Mr Lescaudron had made the representations on which the misrepresentation claim was founded, let alone that he did so fraudulently or that those representations were intended to or did induce the plaintiffs to enter into the two LPI policies.

[235]Counsel for the plaintiffs have pointed out that the representations were implied from the acts of proposing investment in the LPI policies and that the original statement of claim pleaded (at para 34) that, on various dates between 2011 and 2013, the Bank recommended investments in life insurance policies issued by CS Life. But in the context of the contractual claim for which alone relief was being claimed, this was no more than background to the making of the contracts. It was not a material fact, let alone an essential factual element in the cause of action. Furthermore, only when the plaintiffs applied to amend the statement of claim did they allege that, in recommending the investment in life insurance policies issued by CS Life, the Bank was acting on behalf of CS Life. In the original statement of claim the only reference to the Bank’s role as CS Life’s agent was an allegation that the Bank acted as agent in holding and investing the premium for the Meadowsweet and Sandcay policies. That fact had no relevance to the misrepresentation claim.

[236]The only significant area of overlap between the two claims consists in the allegations of wrongdoing by Mr Lescaudron which caused the plaintiffs to suffer loss. On the most generous view, this falls far short of showing that the respective causes of action arise out of the same or substantially the same (material) facts.

The purpose of the amendment power

[237]Ms Louise Hutton KC, who presented the oral argument on the cross-appeal, urged the Board to consider the purpose of asking whether the facts out of which the two causes of action arise are “substantially the same”. In Goode v Martin

[2001] 3 All ER 562

, 566, Colman J at first instance identified that purpose as being “to avoid placing a defendant in the position where if the amendment is allowed he will be obliged after expiration of the limitation period to investigate facts and obtain evidence of matters which are completely … unrelated to those facts which he could reasonably be assumed to have investigated for the purpose of defending the unamended claim”. In BP plc v AON Ltd

[2006] 1 Lloyd’s Rep 549

, para 54, Colman J repeated this statement and described the power to give permission to amend as “thus based on the assumption that the party against whom the proposed amendment is directed will not be prejudiced because that party will … already have had to investigate the same or substantially the same facts”. In Ballinger v Mercer Ltd

[2014] EWCA Civ 996

;

[2014] 1 WLR 3597

, para 34, the Court of Appeal of England and Wales in obiter dicta approved these statements as “helpful guidance”.

[238]Ms Hutton emphasised that the Chief Justice was here in the relatively unusual position of deciding whether to give permission to amend at the end of the trial. That meant that he could assess the extent to which CS Life had in fact been obliged to investigate facts and obtain evidence of matters unrelated to the claims originally pleaded. The Chief Justice observed that CS Life gave no additional discovery and called no additional witness in relation to the misrepresentation claim. He judged that the misrepresentation claim had not materially widened the factual inquiry at the trial. Ms Hutton submitted that, given the purpose of RSC Order 20, rule 5(5), the Chief Justice was in those circumstances entitled to conclude that the misrepresentation claim arose out of the same or substantially the same facts as the claim for breach of contract.

[239]In support of the argument that it is permissible to look further than the pleadings, Ms Hutton cited Akers v Samba Financial Group

[2019] EWCA Civ 416

;

[2019] 4 WLR 54

, para 52, where McCombe LJ said that:

“in the vast majority of cases what is ‘in issue’ in an existing claim will usually be determined by examination of the pleadings alone. It will be the primary, and probably the only, source of material for deciding the question. In some cases, however, such as those considered above where, for example, there has been an extensive evidential battle on a summary judgment application or on a jurisdictional question, it may be possible to discern that facts are already in issue in a case prior to being crystallised in formal pleadings. None the less, I consider that such cases will be rare.”

[240]This statement, with which the Board agrees, does not support the plaintiffs’ argument. The Board has no difficulty in accepting that there may be cases in which it is possible to identify and compare the essential factual elements in an existing and proposed new cause of action before the proposed new cause of action has been formally pleaded. But it does not follow that, when there are pleadings, a more wide-ranging inquiry is legitimate. It is, after all, the function of pleadings to define the issues in the litigation by identifying the material facts that give rise to a claim (or defence).

[241]Although the discussion of this point in Akers was obiter, Floyd LJ also addressed it. He said, at para 56:

“Whilst I can understand that there might be situations where it would strike one as fair to inquire more widely, there are obvious practical difficulties in defining the scope of the permissible inquiry if it is not limited by the pleadings. Does the pool of facts in which it is permissible to fish for the basis of the new cause of action include facts alleged in party and party correspondence, or in every witness statement which has been filed? Do transcripts of submissions and evidence count? … My provisional view is that neither the Act nor the rule contemplates such a broad-ranging inquiry to determine what facts are in issue.”

The Board agrees with that provisional view.

[242]Leaving the decision whether to allow an amendment that adds a new cause of action until the end of the trial, as the Chief Justice did here, is indeed unusual. There is good reason for that. If a new claim is advanced after the applicable limitation period has expired, then unless it arises out of the same facts or substantially the same facts as an existing claim, the defendant ought not to be vexed with it. The plaintiffs themselves make this point in submitting that the purpose of the “substantially the same facts” test is to avoid prejudicing a defendant by requiring the defendant, after the limitation period has expired, to investigate facts and obtain evidence of matters which it would not otherwise have had to investigate. Postponing the decision whether the test is satisfied until the end of the trial undermines that purpose. It means that, if the court ultimately decides that it lacks the power to, or should not, allow the amendment, the time and expense of investigating and litigating the new claim will have been unnecessarily incurred.

[243]Given these practical considerations as well as those pointed out by Floyd LJ, it is undesirable that the “substantially the same facts” test should depend on whether, in hindsight, it can be seen that litigating the new claim has required additional discovery to be given or additional witnesses to be called. Treating such matters as relevant would provide a reason to adjourn the application to amend until the end of the trial on the ground that the court will then be better placed to judge the question. This would encourage the very waste of time and expense which the procedure is designed to avoid.

[244]The objections to taking such matters into account are not only practical. In the Board’s view, the plaintiffs’ argument involves the error of treating a purpose of having a rule that allows a new claim to be added after the expiry of the limitation period only if a restrictive test is satisfied as if it were the test itself. Assessing whether the defendant would be prejudiced by having to investigate new facts is certainly relevant when the court is exercising its discretion whether to allow the amendment. But RSC Order 20, rule 5(5) requires a condition to be satisfied before the court can exercise discretion. This condition does not involve or permit an evaluation of potential prejudice. It is a relatively hard-edged test. Its wording requires the court to identify the proposed new “cause of action” and each existing “cause of action” and to compare the facts out of which each cause of action arose. In applying the test, the court has therefore to analyse the factual elements which constitute each cause of action. There is no licence for undertaking a more broad-ranging inquiry.

[245]Thus, the Board does not consider that there is any reason to relax or depart from the well-established approach which requires the question whether a new cause of action arises out of the same facts or substantially the same facts as a cause of action in respect of which relief has already been claimed to be determined by examination of the pleadings alone. There was, in the Board’s opinion, no good reason in this case to postpone the determination of that question until the end of the trial. It was plain when the application to amend was made that the condition in RSC Order 20, rule 5(5) was not satisfied.

[246]The Chief Justice might still properly have deferred the decision whether to allow the amendments until the end of the trial because it was unclear whether the applicable limitation period had in fact expired. That was going to depend on expert evidence of Georgian law and the resolution of the issues discussed above. But it was plain when the application to amend was made that the facts required to plead the misrepresentation claim were substantially different from those already raised by the contractual claim. Nothing which emerged afterwards could affect that assessment. The suggestion that the plaintiffs could and should be permitted to pursue the misrepresentation claim even if the applicable limitation period had expired was therefore misplaced when the claim was first advanced and remains so now.

Conclusion on the cross-appeal

[247]On the cross-appeal, the Board has concluded that the Court of Appeal was wrong to hold that the plaintiffs’ misrepresentation claim failed because conscious awareness or understanding of the representation made is a legal requirement of a claim in deceit. But the Court of Appeal was correct to decide that the misrepresentation claim was timebarred because it was brought after the expiry of the three-year period of limitation prescribed by Georgian law. IV. ADVICE

[248]The Board will humbly advise His Majesty to allow the appeal, but only to the extent stated at para 111 above, and to dismiss the cross-appeal.