Cavell USA, Inc & Anor v Seaton Insurance Company & Anor

The appellants (“Seaton” and “Stonewall”) were at all material times insurance companies which in 1999 and 2000 were not accepting new business and were in run-off. The run-off was conducted by the first respondents (“Cavell”) who were employed by Seaton and Stonewall for this purpose at a remuneration of $4-5 million per annum. As between themselves they were, of course, principal parties to the contractual arrangements made in respect of the run-off but in the course of conducting the run-off Cavell acted on behalf of Seaton and Stonewall particularly in adjusting and paying claims on the one hand and in making reinsurance recoveries from reinsurers (“NICO”) on the other. Mr Ken Randall was the moving spirit of Cavell. The fact that the judgment of Gross J has been reported [2009] Lloyd’s Rep. IR 616 relieves me from the necessity of setting out the facts and contentions of the parties in more than the barest outline.

Seaton and Stonewall allege (inter alia) that in breach of contractual or fiduciary duty Cavell and/or Mr Randall sub-contracted the conduct of the run-off to NICO and that they were therefore unable to look after the proper interests of Seaton and Stonewall; that is because Seaton and Stonewall wanted a speedy run-off and, if possible, a sensible and reasonably priced commutation from reinsurers, whereas NICO’s interest as reinsurers was to benefit from the substantial premiums which they had received and defer the payment of claims for as long as possible. It is said that NICO provided substantial separate business to Cavell as an inducement to Cavell to sub-contract the run-offs and Cavell dishonestly benefited from that arrangement and concealed the existence of the sub-contracting agreement (called “the Collaboration Agreement”) from their principals.

The owners of Seaton and Stonewall felt they had no alternative but to terminate the relationship and this was done by what was called a Term Sheet of February 2006. By this time Seaton and Stonewall’s interests were substantially owned by an entity called Dukes Place Holdings LP (“Dukes Place”) while the Randall interests were substantially owned by an entity called Randall & Quilter Investment Holdings Ltd (“RQIH”) and both these companies were parties to the Term Sheet on behalf of their respective affiliates and subsidiaries. The preamble stated as follows:-

“This Term Sheet documents the agreement between the parties identified as Parties below with respect to the orderly termination of the contractual and other commercial relationships amongst them and the orderly handover by Cavell Management Services Limited (“Cavell UK”) and Cavell USA Inc, (“Cavell USA”) of run-off management and other services in connection with Seaton Insurance Company (“Seaton”), Stonewall Insurance Company (“Stonewall”), Unione Italiana (UK) Reinsurance Company Ltd (“Unione”) and Cavell Insurance Company Limited (“CIC”) having regard to the regulatory responsibilities of Dukes Place (as defined below) and Randall (as defined below) and the interests of the policyholders of Seaton, Stonewall, Unione and CIC.”

Clause 12 provided that the existing run-off management agreements between Cavell and Seaton and Stonewall were to terminate on 31^st March 2006. Clause 13 provided:-

“Dukes Place hereby releases and forever discharges Randall of and from all actions, causes of action, suits, claims and demands whatsoever, whether at law or equity, whether known or unknown, suspected or unsuspected, disclosed or undisclosed, fixed or contingent, accrued or unaccrued, asserted or unasserted, which Dukes Place ever had, now has or hereafter can, shall or may have against Randall for, upon, or by reason of any matter, cause or thing whatsoever arising out of or in connection with any business, commercial, contractual or other arrangements between or involving either of them as at the date of this Term Sheet, save (i) in respect of any obligations expressly set out in this Term Sheet, (ii) in respect of any actions, causes of action, suits, claims, and demands arising from any breach by Randall of any provision of this Term Sheet, and (iii) in the case of fraud on the part of Randall. This release will not inure to the benefit of any third party, and Dukes Place expressly reserves the right to pursue, against any such third party, any of the claims released herein.

Clause 29 provided:-

“This Term Sheet shall be governed by and construed in accordance with English law and the parties submit to the exclusive jurisdiction of the English Courts.”

It will be noted that Randall were discharged from all actions claims and demands “whether at law or equity” save (relevantly) “in the case of fraud” on Randall’s part. Seaton and Stonewall have in the light of the provision instituted what they call “fraud proceedings” in the Southern District of New York against both Cavell and Mr Randall who have resisted the jurisdiction of the New York court on the basis that such claims (which, it is accepted, can only be such claims as would be recognised as being in respect of fraud in English law) should have been brought in England. Cavell and Mr Randall have accordingly issued their own proceedings for declaratory relief in England. In the course of these proceedings Flaux J on 23^rd May 2008 ordered that there be a trial of preliminary issues to determine 2 questions.

“A.

Whether, on a proper construction of the Term Sheet documenting the agreement between Dukes Place Holdings LP and Randall & Quilter Investment Holdings Limited and signed in February 2006 (“the Term Sheet”) and in particular clause 29 thereof, the parties have agreed to submit all disputes (including claims in fraud against Randall (as the phrase is used in the Term Sheet)) to the exclusive jurisdiction of the English Courts.

B.

Whether, on a proper construction of clause 13 of the Term Sheet, the claims advanced by the defendants in the First Amended Complaint in the US District Court for the Southern District of New York (“the First Amended Complaint”) are claims in fraud within the meaning of that clause.”

Those issues came on for trial before Gross J who answered question A in the affirmative. With the consent (if not at the suggestion) of counsel he divided question B into two parts as set out in para 16 of his judgment:-

Mr Swainston QC for Seaton and Stonewall submitted that his permissible fraud claim (whatever that precisely was) was a claim which already existed before the Term Sheet had been concluded and that there was nothing in the Term Sheet which stopped him from bringing that claim in any appropriate forum and particularly in New York which was the jurisdiction specified in the two original management agreements. The fact that the Seaton agreement provided for jurisdiction in the courts of New York while the Stonewall agreement provided for disputes to be resolved by arbitration in New York did not detract from that submission. If Cavell wanted to apply for a stay in relation to the disputes under the Stonewall agreement, Cavell could always do so. The claim against Mr Randall was not caught by the arbitration agreement anyway since he was not a party to the Stonewall agreement.

Mr Hofmeyr QC for Cavell and Mr Randall submitted that clause 29 of the Term Sheet applied to all the claims which were not released but were saved by clause 13. The parties could never have intended that any claim for fraud should be brought in New York if it was only permissible to bring a claim which English law would regard as a fraud claim. A New York court would have to be educated about the English law of fraud in order to see whether whatever claim Seaton and Stonewall decided to bring constituted fraud according to English law. That would be an absurdly difficult exercise. He further relied on the fact that the whole point of the Term Sheet was to disentangle and put an end to a whole series of relationships (not just those with Seaton and Stonewall). That had to mean that a clean break was intended and that the Term Sheet was to be the source of all obligations save where old obligations were expressly preserved, as they were in relation to an Advisory Agreement with Renaissance Capital Partners in clause 23 of the Term Sheet and Cavell UK’s lease of premises in Norwich in clause 24. Otherwise all disputes had to be resolved in accordance with clause 29.

To this Mr Swainston riposted that both original management agreements had specifically provided that the jurisdiction/arbitration clauses in those agreements were to survive termination of the management agreements.

These arguments were finely balanced but I agree with the judge. What the parties wanted was identified in the preamble as being

“the orderly termination of the contractual and other commercial relationships amongst them and the orderly handover … of run-off management and other services”

in connection with four separate insurance companies. There was to be a general release of all claims and demands save for 3 specific cases. The first two exceptions related to obligations set out in “this Term Sheet” and claims for breach of any provision of “this Term Sheet” and would have to be dealt with in England pursuant to clause 29. The same must, in my view, apply to the third exception.

“In the case of fraud on the part of Randall.”

The complaints against Randall related to breach of their contractual and fiduciary obligations but those claims have now been released “save in the case of fraud”. That is a new concept provided for in the Term Sheet and it is a concept which must, in my view, fall within the provision of clause 29 just like the other two exceptions.

So, as a pure matter of construction of the Term Sheet, any claims for fraud must be brought in England. This conclusion is supported to my mind by the genuine difficulty that a New York Court would have in deciding whether a particular claim would be regarded as a case of fraud as a matter of English law. As the argument before the judge and before us showed, there is scope for considerable debate about what constitutes “fraud” as a matter of English law. It took the judge 57 paragraphs of factual and legal analysis to conclude that only claims which would sound in the tort of deceit could be maintained pursuant to clause 13 of the Term Sheet. Mr Swainston (not wholly implausibly) maintains that that is too narrow an approach and English law must (and does) recognise that the concept of “fraud” is wider than the tort of deceit and devotes 21 paragraphs of what is supposed to be a “skeleton argument” in support of that submission. Mr Hofmeyr then needs 35 paragraphs of his skeleton argument to show that the judge was right. The parties cannot have intended that a concept of such legal uncertainty should be subject of New York litigation or New York arbitration. Both Mr Swainston and Mr Hofmeyr maintained that once it was clear that the definition of “fraud” that they espoused was established, New York courts or arbitrators would have no difficulty in applying that definition but would have great difficulty if the other side’s definition was found to be the correct one. But for my part I cannot believe that the parties to the Term Sheet can have contemplated that New York courts or arbitrators would have to become involved in debate about the scope of the English law of fraud at all. The judge, who presumably knew that he was shortly going on to decide that “save in the case of fraud” meant “save in a case where the facts disclosed the tort of deceit”, said that it would be a judicial nightmare for the New York court to have to apply the English concept of fraud. I agree. I would agree just as much, if the judge had adopted Mr Swainston’s submissions as the meaning of fraud.

Seaton and Stonewall relied on the case of Satyam Computer Services v Upaid Systems [2008] EWCA Civ 487 [2008] 2 All E.R. (Comm) 465 to show that the terms of an original agreement can survive a comprehensive settlement. But since that settlement expressly preserved the original (assignment) agreement whereas the Term Sheet which is the subject of this appeal does not expressly preserve the original management run-off agreements, and indeed expressly terminates them, Satyam is readily distinguishable. The fact that the jurisdiction provisions of the original agreement were originally intended to survive termination does not, to my mind, mean that they survive through to the wholly new legal landscape contemplated by the Term Sheet.

Having answered the first question, ordered by Flaux J to be tried as a preliminary issue, by saying that the parties had agreed to submit their disputes (including any fraud claims) to the exclusive jurisdiction of the English courts, the judge could have been forgiven for saying that the second question, relating to the claims advanced in the First Amended Complaint, did not arise and did not need an answer. Instead of doing that, however, he broke down the second question into the two parts set out in para 6 above. It is clear that in so doing he was acting with the consent and, most probably, the active encouragement of counsel because he wished to help the parties by saying what sort of claim could legitimately be brought in the English court, even though no such claim had yet been pleaded in England.

This laudable attempt was not, in my view, sufficiently thought through. In the first place the phrase “claims in fraud” does not appear in clause 13 at all. Claims were what were released “save …. (iii) in the case of fraud”. To kaleidoscope the release and the exception into “claims in fraud” already starts to narrow down the phraseology of the release as a whole. Secondly, as Toulson LJ pointed out towards the end of the argument, the subject – matters of the release are:-

“all actions, causes of action, suits, claims and demands whatsoever, whether at law or equity … save … (iii) in the case of fraud.”

Since both claims at law or equity are envisaged as being released except in the case of fraud, it might be thought that both legal and equitable claims were to be preserved in the case of fraud. This should, of course, hardly matter since the fusion of law and equity in 1873 but the use of the phrase whether at law or equity by those, whom Mr Hofmeyr submitted had been “careful lawyers” in the drafting of the agreement, might be an indication that something wider than the common law action of deceit was envisaged to be preserved by the term “in case of fraud”.

Thirdly, as I have already observed, the concept of fraud is notoriously difficult to define. In his letter to Lord Kames of 30^th June 1759 (which the editors, if not the original author, of Snell on Equity, have done well to preserve, see 25^th edition 1960 ed. Megarry and Baker p. 496 and 31^st edition ed. McGhee para 8 – 01) Lord Hardwicke said

“Fraud is infinite, and were a Court of Equity once to lay down rules, how far they would go, and no further, in extending their relief against it, or to define strictly the species or evidence of it, the jurisdiction would be cramped and perpetually eluded by new schemes which the fertility of man’s invention would continue.”

Twenty three years after the fusion of law and equity Lord Macnaghten agreed:-

“Fraud is infinite in variety: sometimes it is audacious and unblushing; sometimes it pays a sort of homage to virtue, and then it is modest and retiring; it would be honesty itself if it could only afford it, see Reddaway v Banham [1886] AC 199, 221.”

One can only reflect how brave it was of Gross J to volunteer to define the phrase “claims in fraud” and track a path which Lord Hardwicke and Lord Macnaghten feared to tread. It is perhaps not surprising that he felt that the common law of the tort of deceit was the only rock to which he could safely cling.

Other great judges have had similar difficulties in defining fraud. In Welham v DPP [1961] AC 103, 133 Lord Denning drew a distinction between “intent to deceive” and “intent to defraud” by saying that the former conveyed the element of deceit which induced a state of mind “without the element of fraud which induces a course of action or inaction”. When he came to consider a person’s entitlement to trial by jury on the ground that “a charge of fraud against that party was in issue” pursuant to section 6(1)(a) of the Administrative of Justice (Miscellaneous Provisions) Act 1933, he said

“’Fraud’ in ordinary speech means the using of false representations to obtain an unjust advantage … Likewise in law ‘fraud’ is proved when it is shown that a false representation has been made knowingly or without belief in its truth, or recklessly, careless whether it be true or false, see Derry v Peek per Lord Herschell.”

He concluded in Barclays Bank v Cole [1967] 2 QB 738 that a claimant alleging robbery did not make a “charge of fraud” because robbery did not “involve a false representation”. This equation of fraud and the tort of deceit was much relied on by Mr Hofmeyr in his submissions in support of the judge.

Diplock LJ thought that the phrase “a charge of fraud” as used in the 1933 Act had had a special meaning as a term of art for over a hundred years. He cited the 3^rd edition of Bullen & Leake and said

“’fraud’ in civil actions at common law, whether as a cause of action or as a defence, has meant an intentional representation (or, in some cases, concealment) of fact made by one party with the intention of inducing another party to act on it which does induce the other party to act on it to his detriment.”

This is, of course, the language of the common law tort of deceit, although the reference to concealment shows that silence can sometimes be enough. When, however, the House of Lords had to consider whether deceit was a necessary ingredient of the criminal offence of conspiracy to defraud, it was held in Scott v Commissioner of Police for the Metropolis [1975] A.C. 819 that the criminal offence did not necessarily involve deceit. Lord Diplock (as he had by then become) said at page 841:-

“The intended means by which the purpose [of the conspiracy to defraud] is to be achieved must be dishonest. They need not constitute the civil tort of deceit. Dishonesty of any kind is enough.”

This enabled Mr Swainston to submit to the judge and to us that “claims in fraud” meant “claims that any dishonest act had been committed which caused loss to the claimant”. Not surprisingly the judge recoiled from endorsing such a wide definition and decided that the only definition he could give was that “claims in fraud” meant “claims in deceit”.

Faced with the prospect that the same result might follow in this court and faced with the same difficulty, about defining fraud in any wider sense than deceit, as that faced by Lord Hardwicke and Lord Macnaghten, Mr Swainston in his reply said that what he had meant to encourage the judge to do was to frame issue B(1) in this form.

“Did the word ‘fraud’ in clause 13 of the Term Sheet include the concept of “dishonest abuse of fiduciary position?”

Not unnaturally Mr Hofmeyr objected to this re-reformulation of the preliminary issue at such a late stage and pointed out that even now Mr Swainston had made no attempt to define “dishonest” any more than he had made any real attempt to define fraud.

Mr Swainston responded that to send him away completely empty handed, particularly if the court felt that clause 13 was not intended to be confined to claims in the tort of deceit, would do a positive disservice to the parties, since in relation to what was truly a claim of dishonest abuse of fiduciary position (whether well-founded or not) the parties would have to go through the whole process of ascertaining whether such a claim was permissible or not all over again with further possibilities of appeals. He recognised that there was no pleading before the court but that was only because the application had started as a jurisdiction application in relation to which it is not customary to go to the expense of drafting a comprehensive pleading since the case might never be proceeding in England at all. The claim he wished to bring was, he said, sufficiently adumbrated in the First Amended Complaint and in the Statement of Case which he had been required to submit for the purpose of the trial of the preliminary issues.

For my part, I am persuaded that when the parties agreed that ‘all claims … whether at law or equity … save in the case of fraud’ were to be released, they did not envisage that the only claims which Seaton and Stonewall could thereafter bring were claims which sounded in the English tort of deceit. I have already said that the phrases “or equity” and “in case of fraud” do not sit well with confining fraud to deceit. Nor do I consider that a decision on an English statute entitling someone charged with fraud to a jury trial in a civil action can be a decision of the meaning of “fraud” in a commercial document which has a decided international flavour. (There are similar decisions about the right of someone accused of fraud to be released from an arbitration agreement under the Arbitration Act 1950 but I do not think they can be decisive either). The run-offs were basically run-offs of Seaton and Stonewall’s American business and were originally subject to New York jurisdiction and arbitration clauses. Despite the fact that the Term Sheet was drawn up by English lawyers after substantive agreement between two principals had been reached, I do not think the word “fraud” was ever intended to mean only deceit in the sense of being only liability that follows from a fraudulent representation.

Mr Hofmeyr submitted that, once one abandoned the “rock” of deceit, one was only left with the “shifting sands” of an elusive and amorphous concept of fraud which could not be what the parties could have intended. One sees the force of that argument and it would be particularly attractive if one was forced to give a comprehensive definition of ‘fraud’ for all purposes under the agreement. But I am unwilling to accept the straightjacket which the judge adopted when he agreed to define the expression “claims in fraud” and there is no need to do so.

The question that remains, therefore, is whether to allow Mr Swainston at this later stage to re-reformulate issue B as set out in para 19 above. In spite of all the dangers inherent in trying to assist commercial parties to resolve their differences when it is not strictly necessary to do so, there is nevertheless a long tradition of judges of the Commercial Court so doing and I think this court should do the same, at any rate in a commercial case. Since Mr Swainston always had a secondary argument, both in this court and below, to the effect that “fraud” at least included dishonest abuse of fiduciary position, Mr Hofmeyr has been able to deal with the argument and I would therefore permit issue B(1) to be re-formulated as asked. I would, however, emphasise that the court has seen no pleading and that the somewhat theoretical exercise on which I am about to embark is not intended, in advance, to authorise or to deny the validity of any pleading which may be forthcoming, let alone any particular paragraphs of it.

The primary reason given by the judge (para 90(i)) for declining to include the concept of dishonest abuse of fiduciary position in the concept of fraud was that it was not a cause of action known to the law. Dishonesty is not a necessary element in the cause of action of abuse of fiduciary position. There is moreover authority for the proposition that dishonesty is not, as in the common law crime of conspiracy to defraud considered in Scott’s case, the touchstone of fraud but rather that deception is, see Kensington v Republic of Congo [2008] 1 Lloyds Rep 161, 174 per Moore-Bick LJ. Moore-Bick LJ also considered that conduct involving dishonest abuse of a position (in which a person is expected to safeguard the financial interests of another person) with a view to gain for himself or another, or causing loss or risk of loss to another, could be described as deception of a kind, "since the wrongdoer deliberately deceives the person whose interests he is bound to safeguard by allowing him to believe in his trustworthiness while actively falsifying that belief”.

The fact that a fiduciary may be liable for abuse of position without being dishonest does not seem to me to be conclusive of the proposition that a person abusing his fiduciary position can never be regarded as fraudulent. Some abusers of position will be acting dishonestly; others (probably rather fewer) will not be acting dishonestly in the sense that they might consider that they were entitled to act as they did or they might think that, even if not entitled, they were acting for the benefit of their principal. Those, however, who were acting dishonestly may well be acting fraudulently or ‘in fraud of’ their principal. In 1998 Mr Jack Straw, as Home Secretary, asked the Law Commission

“As part of their programme of work on dishonesty, to examine the law of fraud.”

The result was their comprehensive 2002 report called “Fraud” (Cm.5560) and one of the matters they considered was whether abuse of fiduciary position should be a criminal offence.

They concluded that there should be a criminal offence of “fraud” where

“with intent to make a gain or to cause loss or to expose another to a risk of loss, a person dishonestly

(1)

makes a false representation;

(2)

wrongfully fails to disclose information; or

(3)

secretly abuses a position of trust.” Para 1.8

In para 4.37 the Law Commission gave various instances of frauds which did not (necessarily) involve deception. The first of these was ‘making a secret gain or causing loss by abusing a position of trust or fiduciary duty’. In para 7.20 they say

“If an employee embezzles her employer’s money, both lawyers and non-lawyers would agree that her conduct can properly be described as fraud even if she makes no representation (for example, by falsifying the accounts)”

I can only say that, as a lawyer, I agree with that statement and I would expect both the lawyers who drafted the Term Sheet and their respective lay clients to agree also.

The Law Commission then recommended that secret abuse of position was fraud and should be the subject-matter of potential criminal liability. They explained this in terms of what would generally be regarded as a fiduciary relationship in para 7.37

“The essence of the kind of relationship which in our view should be a prerequisite of this form of the offence is that the victim has voluntarily put the defendant in a privileged position, by virtue of which the defendant is expected to safeguard the victim’s financial interests or given power to damage those interests. Such an expectation to safeguard or power to damage may arise, for example, because the defendant is given authority to exercise a discretion on the victim’s behalf, or is given access to the victim’s assets, premises, equipment or customers.”

As a matter of history, the recommendation was accepted by the Government and is now contained in section 4 of the Fraud Act 2006.

These considerations gain considerable support from the obiter dicta of Millett LJ in Armitage v Nurse [1998] Ch. 241 in which a trustee was exempted from liability for loss or damage “unless such loss or damage shall be caused by his own actual fraud”. The court held that the trustee was under no liability in absence of any dishonest intention. Millett LJ went on to say (page 251) that care was needed when applying concepts relevant to the tort of deceit to a breach of trust because breaches of trust were of many different kinds. The phrase “actual fraud “was” not used to describe the common law tort of deceit”.

To my mind all this shows that in the commercial context of this case the concept of ‘fraud’ is wider than the concept of the tort of deceit where a fraudulent misrepresentation (or equivalent) is required. I do not consider, therefore, that the second paragraph of the judge’s order can stand and I would set it aside and substitute a declaration that the exception “in case of fraud” in clause 13 of the Term Sheet is not confined to claims in deceit but extends to at least some cases of dishonest abuse of fiduciary position. I would be prepared to add in this judgment (but not as part of the declaration) that the defendant is not to be held to have acted dishonestly unless he was doing something that ordinary people would regard as dishonest and he knew that ordinary people would so regard it. This, as I understand it, is the criminal law test of dishonesty and I see no reason why it should not apply to any dishonest abuse of fiduciary position contemplated by the Term Sheet.

Further than that I am not prepared to go. It will be for the trial judge to determine whether any further element of “deception” or “secrecy” is required in addition to the above and to determine all the facts of the matter. To this limited extent I would allow the appeal in relation to the answer B(1). As far as any personal claim against Mr Randall is concerned, I do not see that the position can be any different, save that it must be readily arguable that it was not he but his companies which occupied any relevant fiduciary position.

I agree.

I also agree.