Stevens (Respondent) v Hotel Portfolio II UK Ltd (In Liquidation) and another (Appellants)

Case summary

Case ID

UKSC/2023/0142

Parties

Appellant(s)

Hotel Portfolio II UK Ltd (In Liquidation),

Elizabeth Alexandra Aird-Brown (as Liquidator of Hotel Portfolio II UK Ltd (In Liquidation))

Respondent(s)

Anthony Edward Stevens

Judgment appealed

Judgment details

Judgment date

23 July 2025

Neutral citation

[2025] UKSC 28

Hearing dates

Full hearing

Start date

25 February 2025

End date

26 February 2025

Justices

Lord Reed

Lord Briggs

Lord Hamblen

Lord Burrows

Lord Richards

Judgment details

Trinity Term 2025

LORD BRIGGS (with whom Lord Reed, Lord Hamblen and Lord Richards agree):

1. The issue of law which arises on this appeal is whether a person who dishonestly assists a constructive trustee in dissipating the trust fund is liable to compensate the beneficiary for the consequential loss of its proprietary interest in the fund in circumstances where:

1. The fund held on constructive trust consisted of an unauthorised profit previously made by the trustee in breach of fiduciary duty to the same beneficiary, the making of which caused the beneficiary no loss, because it was a profit that could not have been made by the trust, and
2. The dishonest assistant also dishonestly assisted the trustee in making the profit in the first place.

The trial judge held the dishonest assistant liable, but the Court of Appeal allowed his appeal.

2. The answer to this question requires an analysis of the nature, terms and consequences of a constructive trust of unauthorised profits, a further examination of the basis for assessing compensatory liability for breach of trust, restated in Target Holdings Ltd v Redferns [1996] AC 421, and an in-depth examination, for the first time, of the principle that a trustee may not set off gains and losses incurred in the course of multiple breaches of trust, together with a possible exception to it where the breaches occur in the same or connected transactions.

3. On about 25 March 2008 there was credited to a bank account in the name of a nominee of Mr Ruhan a dividend in the sum of around £95m. Mr Ruhan was a constructive trustee of that dividend for the appellant Hotel Portfolio II UK Ltd (“HPII”). So, HPII was its beneficial owner. In ordinary parlance the dividend belonged to HPII. But HPII knew nothing about it at the time. The dividend consisted of a secret and unauthorised profit made by Mr Ruhan in breach of his fiduciary duty to HPII as its director.

4. Starting about a week later, and in clear breach of his duties as trustee under that constructive trust, Mr Ruhan spent the whole of the dividend upon speculative projects of his own in Qatar, and lost all of it, so that no part of it can now be traced or recovered by HPII. The respondent Mr Stevens dishonestly assisted Mr Ruhan in the dissipation of the dividend.

5. Those who dishonestly assist a trustee in a breach of trust are jointly liable with the trustee to the beneficiary for any loss caused by the breach. When in 2018, by then in liquidation, HPII found out about the dividend and its dissipation, it sued both Mr Ruhan and Mr Stevens, on the basis that Mr Ruhan had received the dividend as an unauthorised profit and had caused HPII to lose the whole of the dividend, and that Mr Stevens had dishonestly assisted Mr Ruhan in his breach of trust. The trial judge found this to have been proved and ordered Mr Ruhan to account for the dividend, and Mr Stevens to pay compensation which included £95m for HPII’s loss of the dividend. Nothing has been recovered from Mr Ruhan.

6. Mr Stevens appealed. His main argument was that HPII had suffered no loss by reason of the dissipation of the dividend, so there was no loss for which he was liable to pay compensation. This was, he said, because HPII only became the temporary beneficial owner of the dividend as the result of the earlier breach of fiduciary duty by Mr Ruhan in making the unauthorised profit represented by the dividend, and that both the gain and the loss were parts of a single pre-arranged fraudulent scheme, neither of which would have occurred if Mr Ruhan had performed his duties as fiduciary and trustee. Alternatively he argued that, if the dissipation of the dividend caused a loss to HPII, its earlier receipt was an equivalent gain also caused by a related breach of fiduciary duty by Mr Ruhan, such that both Mr Ruhan and therefore Mr Stevens could set off the gain against the loss, producing a nil return for HPII.

7. The Court of Appeal accepted these arguments and allowed Mr Stevens’ appeal. The question for this court, raised by HPII’s appeal, is whether it was correct to do so.

8. Standing right back, as Mr James Pickering KC, counsel for HPII, asked us to do, the argument that the complete dissipation by a trustee of £95m of his beneficiary’s property can have caused the beneficiary no loss appears to be contrary to any ordinary notion of justice, equity or common sense. A non-lawyer might well think that something had gone seriously wrong with the law. If someone wrongly takes my car, my boat or my dog, and I cannot get it back, why have I not suffered a loss equivalent to the value of that which has been taken? And why does it matter how I came to be its owner? Even if the taker had originally given it to me, why does his later decision to take it back from me cause me no loss? The result of the gift was that it became my property. I was the owner of it when it was taken. And why, if that is the correct approach to property gifted to me by the taker, should it make any difference that the property became mine as the result of an earlier breach of fiduciary duty owed to me? How can that put the taker in a better position? And why should it matter whether I was the legal owner of the property, or its beneficial owner under a bare trust? Either way, I was its real owner. It was mine.

9. In well researched and attractively presented arguments Mr Sebastian Kokelaar for Mr Stevens submitted that the analysis of the Court of Appeal was both correct in law, and produced a common sense and entirely just and equitable result. His submissions may be summarised under the following three main heads:

1. The constructive trust of the dividend which Mr Ruhan breached by his dissipation of it was in substance a remedy imposed by equity for his earlier breach of fiduciary duty in making the secret profit constituted by the dividend in the first place. Accordingly the subsequent breach constituted by its dissipation could not be looked at in isolation from the breach which gave rise to the dividend.
2. The modern approach of equity to identifying and quantifying a liability to pay compensation for loss arising from a breach of trust or fiduciary duty requires the court to compare the position of the beneficiary as the result of the breach with the position in which it would have been if the trustee or fiduciary had performed his duty. In the present case, the appropriate comparison is between HPII’s position following both breaches, and the position in which it would have been if neither had occurred, which the judge had found to be the same. HPII would neither have gained the dividend nor lost it.
3. There is a well-settled principle, which may be classified as an aspect of equitable set-off, that trustees and fiduciaries may not set off breaches which make gains for the beneficiaries against breaches which cause losses to them, unless the two (or more) breaches are connected. In the present case the breach constituted by the making of the profit could not be more closely connected with the breach constituted by its dissipation.

10. Analysis of those submissions requires a significant exploration of the relevant law. But it is first necessary to set out in more detail the relevant facts. The summary which follows is an abbreviated version of that provided by Newey LJ in the Court of Appeal, none of which is now in issue.

The Facts

11. HPII was in 2004 the owner of a portfolio of hotels in central London which included the Kensington Palace, Kensington Park and Lancaster Gate Hotels (“the Hyde Park Hotels”), each of which was recognised as having, at least in principle, potential for development and conversion into premium residential accommodation. When companies controlled by Mr Ruhan bought HPII for between £42m and £47m in mid-2003 it was already heavily indebted, including to Morgan Stanley Bank International Ltd (“Morgan Stanley”) and Thistle Hotels plc (“Thistle”). In a restructuring in late 2004 the ownership of HPII was equally divided between Mr Ruhan’s companies, Morgan Stanley and Thistle. Mr Ruhan was a director of HPII.

12. Morgan Stanley and Thistle wished to realise their investment (or that of note holders whom they represented) in HPII and authorised Mr Ruhan to accept bids for the Hyde Park Hotels in excess of £125m. In March 2005 a company incorporated in Madeira, Cambulo Comercio e Serviços Sociedade Unipessol LDA (“Cambulo Madeira”) agreed to buy the Hyde Park Hotels from HPII for £125m. The purchase price was attributed as to £56m to the Lancaster Gate Hotel and £69m to the two Kensington hotels, which the trial judge, Foxton J, found was their aggregate market value at the time. Cambulo Madeira was ostensibly beneficially owned by Mr Stevens, but the judge found that he and Cambulo Madeira acted in relation to the acquisition and on-sale of the hotels as Mr Ruhan’s nominee. This was dishonestly concealed throughout from HPII’s other directors and shareholders. Each of the Hyde Park Hotels was, on completion, vested in a different wholly owned subsidiary of Cambulo Madeira.

13. In April 2006 Cambulo Madeira and its relevant subsidiaries entered into a 50/50 joint venture with an unconnected company, CPC Group Ltd (“CPC”), to develop the two Kensington hotels, for which purpose they were transferred to a jointly owned company Cambulo Property Holdings Ltd (“CPHL”). Meanwhile, the Lancaster Gate Hotel was sold to an unconnected third party in August 2006 for £67.5m realising a net profit of £7.76m over the part of the March 2005 purchase price attributed to it. The net profit was received by Mr Ruhan and thereafter dissipated, so that no part of it is now traceable or therefore amenable to the pursuit of a proprietary claim.

14. In late March 2008 the Kensington hotels were sold to an unconnected third party for £320m, realising a gross profit over the part of the 2005 purchase price attributed to them of £251m. After payments to another stakeholder, Cambulo Madeira eventually distributed dividends of £100.2m to CPC and £95.2m to the Ruhan interests. The £95.2m dividend is that to which reference is made at the beginning of this judgment.

15. Mr Ruhan immediately used the £95.2m dividend to pay down loans taken out by him in connection with development projects in Qatar. He may have earmarked the anticipated profits for that purpose before they were realised in the form of the dividend, but not before he embarked on the scheme to acquire the hotels from HPII so as to make a large profit from their development value. Taken together Mr Ruhan made profits from his participation in the acquisition of the Hyde Park Hotels of £103.76m, less £1.5m which the judge found represented his payment to Mr Stevens for his services.

This litigation

16. HPII went into creditors’ voluntary liquidation in April 2008. Mr Ruhan’s part in the acquisition and profitable on-sale of the Hyde Park Hotels was thereafter discovered by its liquidators. Although Mr Ruhan was ordered to account for the whole of the £103m profit he made, nothing has been recovered from him. It is common ground that nothing will be recovered from the pursuit of proprietary claims in relation to those profits.

17. The judge found (and the contrary was not contended for by HPII’s liquidators) that HPII received full market value for the sale of the Hyde Park Hotels to Cambulo Madeira in 2005, that it would not have sold them for more if that sale had not happened, and that it would not itself have exploited the large development opportunity which led to the hugely profitable sale of the two Kensington hotels in 2008. It was never part of HPII’s case that the covert purchase of the three hotels by Mr Ruhan, or their subsequent profitable on-sale, caused HPII any loss.

18. HPII claimed against Mr Stevens that he account for the £1.5m profit which he personally made by dishonestly assisting Mr Ruhan. This liability has not been the subject of this appeal. HPII also claimed against both Mr Ruhan and Mr Stevens compensation for the dissipation of the £102.26m which Mr Ruhan had briefly held upon constructive trust for HPII. As against Mr Ruhan, it made no difference whether he was liable to HPII to account for the profit, or to compensate for the loss caused by its dissipation. Both were personal claims, and on the hypothesis that the dissipation of the profit caused a loss equivalent to it in value, the alternative personal claims would be broadly for an equivalent amount, provided that HPII did not recover twice over.

19. The judge found that Mr Stevens had dishonestly assisted Mr Ruhan in both the acquisition of the profits and in their dissipation. But any claim against Mr Stevens in relation to the £102.26m profit and its dissipation critically depended upon whether his dishonest assistance to Mr Ruhan caused HPII any loss. This is because, whereas a dishonest assistant is jointly liable with the defaulting trustee or fiduciary to compensate the beneficiaries or principal for any loss caused by the breach, the dishonest assistant is only liable to account for any profit received by him, not for the profit received by the trustee or fiduciary. That much is common ground. As already noted, the judge found that the dissipation of the profits did cause HPII an equivalent loss, and ordered Mr Stevens to compensate HPII accordingly. As against Mr Ruhan, HPII’s claim for (at its election) an account of profits or equitable compensation for breach of fiduciary duty succeeded and HPII elected for an account of profits.

Election

20. It is often said that a beneficiary must (at some stage) elect between an account of profits and compensation for loss, in a claim against a defaulting trustee or fiduciary. See generally Personal Representatives of Tang Man Sit v Capacious Investments Ltd [1996] 1 AC 514. In the present case HPII elected, as against Mr Ruhan, for an account of the profits derived from his breach of fiduciary duty involved in the making of them. But no separate election has been made, against either Mr Ruhan or Mr Stevens, in relation to the breach of trust constituted by the dissipation of the profits briefly held by Mr Ruhan for HPII on constructive trust. It is not submitted on Mr Stevens’ behalf that HPII has disabled itself by election from pursuing any available claim for compensation against Mr Stevens, either by its election to pursue Mr Ruhan for an account of profits or otherwise. Accordingly no issue about election arises in this appeal.

Legal analysis

21. I have already described some of the legal common ground between the parties. It is convenient to complete that description before directly addressing the legal questions in issue. First, there is no fundamental difference between the relationship between trustee and beneficiary and the analogous relationship between fiduciary and principal (such as director and company) in the present context. Most of the basic principles were originally fashioned to regulate the former and later applied analogically to the latter, once it was clearly established, over a century ago, that a company is both legal and beneficial owner of its property: see Rukhadze v Recovery Partners GP Ltd [2025] UKSC 10; [2025] 2 WLR 529, paras 3, 16, 24-25. In what follows I will refer generally to trustee and beneficiary, save where it is necessary to speak distinctly of fiduciary and principal.

22. Secondly, there is not suggested to be any difference in outcome as between the £95m dividend which represented Mr Ruhan’s illicit profit from the on-sale of the Kensington hotels and the smaller profit of £7.76m which he received a year and a half earlier from the on-sale of the Lancaster Gate Hotel. For convenience (and because the facts about their dissipation are clearer) I shall apply the law to the facts about the dividend of £95m received and dissipated in and after March 2008.

23. Thirdly and importantly, it is common ground that Mr Ruhan became a constructive trustee of the dividend immediately upon its receipt, under an institutional (rather than purely remedial) constructive trust. Furthermore, although there may be debate in particular cases about the precise nature and extent of the duties of the trustee under such a constructive trust, it is common ground that Mr Ruhan’s dissipation of the dividend was a breach of them. This is because at the very least the constructive trustee’s duty is to conserve the trust property for the benefit of the beneficiary, rather than to deploy it in a way which destroys the beneficiary’s proprietary interest in it, as Mr Ruhan did, dishonestly assisted in that regard by Mr Stevens. And it is inherent in that common ground that, whereas Mr Ruhan had been a fiduciary for HPII rather than a trustee stricto sensu, the relationship between them in relation to the dividend once received by Mr Ruhan was that of trustee and sole beneficiary, in which capacity HPII had a right to call on Mr Ruhan for the transfer of the property on demand, albeit in fact in ignorance of that right, or indeed of the existence of the dividend itself or of the constructive trust of it affecting Mr Ruhan. Its ignorance was of course the consequence of Mr Ruhan’s fraudulent concealment of his status as the real purchaser of the hotels in March 2005, aided and abetted in that regard by Mr Stevens, and not therefore something which either of them can pray in aid.

24. It is also I think inherent in that common ground that from the moment of Mr Ruhan’s receipt of it in March 2008, the dividend belonged not to Mr Ruhan but to HPII as its beneficial owner. It was property about which, had it known about it, HPII could have said: “it’s mine”. This is in my judgment a matter of cardinal importance to the outcome of this appeal. There used to be doubt whether this was true of the relationship of fiduciary and principal in relation to a bribe received by the fiduciary, until resolved by this court in FHR European Ventures LLP v Mankarious [2014] UKSC 45; [2015] AC 250. But it has been the law (or more precisely a principle of equity) since at least as long ago as the early 18th century in relation to unauthorised profits received in breach of trust, whether secret or not: see Keech v Sandford (1726) Sel Cas Ch 61. In the FHR case, Lord Neuberger of Abbotsbury said, at para 7:

“where an agent acquires a benefit which came to his notice as a result of his fiduciary position, or pursuant to an opportunity which results from his fiduciary position, the equitable rule (‘the rule’) is that he is to be treated as having acquired the benefit on behalf of his principal, so that it is beneficially owned by the principal.” (my emphasis).

He continued, at para 8:

“Where the facts of a particular case are within the ambit of the rule, it is strictly applied. The strict application of the rule can be traced back to the well known decision in Keech v Sandford (1726) Sel Cas Ch 61, where a trustee held a lease of a market on trust for an infant, and, having failed to negotiate a new lease on behalf of the infant because the landlord was dissatisfied with the proposed security for the rent, the trustee negotiated a

new lease for himself. Lord King LC concluded, at p 62, that, ‘though I do not say there is a fraud in this case’ and though it ‘may seem hard’, the infant was entitled to an assignment of the new lease and an account of the profits made in the meantime - a conclusion which could only be justified on the basis that the new lease had been beneficially acquired for the infant beneficiary.”

25. The present case is not of course about bribes, but it is an example of a profit made by a fiduciary “as a result of his fiduciary position”, squarely within the settled equitable principle which Lord Neuberger derived from Keech v Sandford and recently examined by this court in Aquila Advisory Ltd v Faichney [2021] UKSC 49; [2021] 1 WLR 5666 and Rukhadze. Applied to this case, it means that Mr Ruhan is to be taken as having made the profit constituted by the dividend on behalf of HPII, so that from the moment of its receipt it was beneficially owned by HPII. Furthermore, to the extent that there is any discernible distinction between Keech v Sandford and this appeal, it is that this is a plain case of fraud, whereas the older case was not. But the principle that a trustee or fiduciary holds such profits upon an immediate institutional constructive trust for the beneficiary does not depend at all upon the fiduciary having acted dishonestly. As Lord Russell of Killowen put it in relation to the parallel liability to account in Regal (Hastings) Ltd v Gulliver [1967] 2 AC 134 at 144:

“The rule of equity which insists on those, who by use of a fiduciary position make a profit, being liable to account for that profit, in no way depends on fraud, or absence of bona fides”.

26. Finally, subject to the question (left unresolved by the Court of Appeal) whether there can be compensation ordered for loss caused by a breach of a constructive trust of profits, it is not I think otherwise in contention that if the dissipation of the dividend caused HPII a loss, and if no set-off is available, then that is a loss for which the judge was right to order Mr Stevens to make compensation. This follows from the undoubted liability of a trustee to compensate the beneficiary for loss caused by a breach of trust, and the joint liability of a dishonest assistant in that breach to make compensation. It follows that the issues before this court are only really three:

1. Can the court order compensation for loss caused by breach of a constructive trust of unauthorised profits?
2. Did the dissipation of the dividend cause HPII a loss?
3. If so, can Mr Stevens (standing for this purpose in Mr Ruhan’s shoes) pray in aid by way of equitable set-off the gain made by HPII in having the dividend treated as having been made for its benefit?

Constructive trust really just a remedy?

27. The first submission (or group of submissions) made by Mr Stevens takes as its starting point the argument that the constructive trust of the dividend is in substance just equity’s remedy for the breach constituted by the making of the profit in the first place. Two outcomes are said to follow. The first is that this strongly reinforces the need to appraise loss by a but-for analysis which accommodates both the original breach and the breach of the constructive trust which operates as a remedy for the first breach. The second, more tentative, argument is that if the constructive trust is just a remedy for the original breach, then there ought not to be a compensatory remedy for the breach of the constructive trust. Its remedial function should be regarded as fully implemented by the pursuit of any available proprietary remedies. This is the point referred to in para 26 above as having been left unresolved by the Court of Appeal.

28. I would accept that, like the parallel liability to account (for which see Rukhadze), there are certain respects in which the constructive trust can usefully be seen loosely as equity’s remedy (or one of equity’s remedies) for a breach of trust. But that should not be allowed to detract in any way from the reality of the trust, or from equity’s intent that the beneficiary should be fully recognised as the beneficial owner of the profit, with all the rights which beneficial ownership ordinarily confers upon a sole beneficiary with an immediate right to demand the delivery up of the property or its proceeds. Just as in Rukhadze the majority concluded that the obligation to account for unauthorised profits was a separate free-standing duty, and not just an after the event remedy for breach of trust, so I would conclude that, once established, the constructive trust of the profits is a free-standing “real” trust in its own right, and not merely a way of conferring additional proprietary remedies upon the beneficiary, beyond the personal remedies of account or compensation for loss. Of course the availability of those proprietary remedies may be of critical importance, but they are not the only consequences of the existence of an institutional constructive trust. Other consequences include the usual personal remedies triggered by a breach of it.

29. The notion that something is a remedy for a wrong implies that it is something which is awarded by a court, just as damages are a court-provided remedy for the loss caused by a breach of contract or a tort. This is not how the constructive trust arises. It is equity’s automatic and immediate response to a set of facts, just as is the common intention trust which ordinarily comes into existence when two people together buy a home which is conveyed into the name of one of them, with the mutual intent that they should be co-owners of it.

30. The “reality” (i.e. institutional rather than purely remedial nature) of the constructive trust is not affected by the question, briefly debated on this appeal, whether it is to be regarded as a “class 1” or “class 2” trust, a classification originally made by Millett LJ in Paragon Finance plc v DB Thakerar & Co [1999] 1 All ER 400 and developed in subsequent authorities. This is a distinction which has consequences for the running of time under the Limitation Acts, but no other consequences of which I am aware, and the contrary was not suggested by counsel. Limitation is not an issue in these proceedings.

31. I would therefore reject the premise upon which the first group of submissions is based. The constructive trust of the dividend which bound Mr Ruhan upon his receipt of it may be equity’s response to the circumstances in which the profit was made, namely that it was made from the exploitation of former property of HPII which he had obtained in breach of fiduciary duty. But it was not just a remedy for that breach.

32. The first suggested outcome of the premise is that this reinforces the need to conduct the “but-for” analysis of the causation of alleged loss on the basis that the original breach (or breaches) of fiduciary duty are rolled up with the breach of the constructive trust of the dividend. Even if I had not rejected the premise, I would not have accepted that outcome. But since it is so closely tied up with Mr Stevens’ second, “no loss” submission, I will defer dealing with it until that stage.

33. The second asserted outcome is that there is no compensatory remedy for breach of this type of constructive trust. The argument proceeds by way of analogy with a remedy in damages for loss. It is said, I would think rightly, that the failure to comply with an order for damages does not amount to a new cause of action for compensation for any loss caused by that failure. Since the constructive trust is an equitable remedy for an earlier breach of trust or fiduciary duty, then it is argued by parity of reasoning that the breach of the constructive trust cannot sound in a remedy for compensation for loss either.

34. This was a matter briefly touched upon by Newey LJ in the Court of Appeal at paras 70-72, where he doubted whether a constructive trustee in Mr Ruhan’s position, or therefore Mr Stevens as his dishonest assistant, would incur a compensatory liability to the beneficiary for loss caused by a breach of that constructive trust. He pointed to what appeared to be the absence of any reported case in which such a claim had succeeded.

35. I do not share Newey LJ’s doubts. It would seem to me extraordinary and contrary to basic equitable principle for the dissipation of a fund held on an institutional constructive trust to give rise to no remedy by way of equitable compensation for any consequential loss. It is suggested that the real remedies arising from the constructive trust are proprietary. But a dissipation of the fund held upon constructive trust strikes at the heart of any proprietary remedy. The very concept of dissipation means that there is neither the fund nor any traceable proceeds of it to which a proprietary remedy can usefully attach. And the dissipation thereby causes a loss to the beneficiary which will generally be at least equivalent in value to the property dissipated, since the beneficiary was its beneficial owner, and has thereby been deprived of any proprietary claim to the property. It might be said that, strictly, the beneficiary’s loss was the value of his beneficial interest in the property dissipated rather than the value of the property itself. But since it is accepted that the minimum terms of this constructive trust required Mr Ruhan to deliver or transfer the property on demand to HPII, that strikes me as a distinction without a difference, at least in the mouth of Mr Ruhan or Mr Stevens. They cannot be heard to say that, for example because they dishonestly kept secret the existence of the dividend until they had dissipated it, HPII’s beneficial interest in it was of no significant value.

36. Generally speaking, it will make no difference to the outcome of a personal claim against the constructive trustee who has dissipated the fund, whether the claim is pursued by way of account or for compensation for the loss thereby caused. Both are personal claims, and the quantum of each is usually likely to be the same, namely the value of the property dissipated. But that difference is said to be critical to the claim against the dishonest assistant in the dissipation. If he is not liable in respect of the trustee’s profits, it is submitted that he cannot be liable, jointly with the trustee, to compensate the beneficiary for loss caused by the dissipation of those profits.

37. I consider that argument to be wrong. The rule that a dishonest assistant is not liable to make good the trustee’s or fiduciary’s account for an unauthorised profit is based on the perception that each contributory to the making of an unauthorised profit is only liable to disgorge that which he has received: See Ultraframe (UK) Ltd v Fielding [2005] EWHC 1638 (Ch); [2007] WTLR 835, paras 1595–1601 in particular, especially para 1600 where Lewison J said:

“I can see that it makes sense for a dishonest assistant to be jointly and severally liable for any loss which the beneficiary suffers as a result of a breach of trust. I can see also that it makes sense for a dishonest assistant to be liable to disgorge any profit which he himself has made as a result of assisting in the breach. However, I cannot take the next step to the conclusion that a dishonest assistant is also liable to pay to the beneficiary an amount equal to a profit which he did not make and which has produced no corresponding loss to the beneficiary.” (Emphasis in original.)

See also Novoship (UK) Ltd v Mikhaylyuk [2014] EWCA Civ 908; [2015] QB 499, at para 77 where Longmore LJ (giving the judgment of the court) said:

“even in Australian law a knowing participant is not generally required to account for profits that he did not make”.

I will call that the Novoship principle. But it needs to be borne constantly in mind (and is common ground) that the breach consisting of the making of the profit is separate and distinct from the breach consisting of the dissipation of the property subject to the constructive trust. That latter breach is not about making unauthorised profits. Rather it is about dissipating or destroying trust property. And there is no rule or principle that a person who dishonestly assists the constructive trustee in the dissipation is only liable for that which he has received, merely because the property in question came to be held upon constructive trust because it represented an unauthorised profit.

38. The point may be tested by three examples, each based upon the fiduciary making the same unauthorised profit, and then his subsequent dissipation of it in breach of constructive trust. First, take the case in which A dishonestly assists in the making of the profit, but not in its dissipation. A can rely upon the Novoship principle. Then take the case in which B dishonestly assists only in the dissipation. B cannot invoke the Novoship principle, because it has nothing to do with aiding and abetting a dissipation of trust property. Lastly take this case, where C dishonestly assists in both the making of the profit and in its dissipation. There is no reason why the fact that the Novoship principle may insulate him from a liability to make good the trustee’s account arising from the making of the profit should also insulate him from liability for the loss caused by the dissipation. And it makes no sense, nor is it equitable, to allow the dishonest assistant in the position of C a defence which pleads dishonest involvement in an earlier breach as the only reason why he should not be liable for dishonest assistance in the dissipation of the trust property. The trial judge was of exactly the same view, at the end of para 287 of his judgment.

39. It was put to Mr Kokelaar that if his submission to the contrary were correct, it would mean that a dishonest assistant who assisted for no reward in the dissipation of profits held upon constructive trust would get away scot-free. He was constrained to accept that this would be so, but made no attempt to explain why that should be just or equitable.

40. There is, it has to be admitted, little useful authority on the question whether a constructive trust of this kind (i.e. one which arises from the making of an unauthorised profit) gives rise to any liability to compensate a beneficiary for loss caused by its breach. The trial judge, Foxton J, found some authority, in text books and reported cases, for the proposition that there can be compensatory liability for dishonest assistance in a breach of constructive trust. At paras 276-277 of his judgment he refers to Paul McGrath QC’s Commercial Fraud in Civil Practice, 2nd ed (2014), paras 9.34-9.35, Goff & Jones: The Law of Unjust Enrichment, 9th ed (2016), para 38-15, Bank Tejarat v Hong Kong and Shanghai Banking Corpn (CI) Ltd [1995] 1 Lloyds’s Rep 239, 247, Fitzalan-Howard v Hibbert [2009] EWHC 2855 (QB) and Heinl v Jyske Bank (Gibraltar) Ltd [1999] Lloyd’s Rep Bank 511. He concluded from those citations that it was settled that there could be compensatory liability for dishonest assistance in relation to breach of a type 2 constructive trust, even a bare trust not imposing fiduciary obligations, and that the arguments in favour of recognising such a claim for dishonest assistance in relation to a type 1 constructive trust seemed even stronger. I agree. In none of those cases was that proposition challenged, and nor was it before Foxton J. Nor indeed did the claims for dishonest assistance succeed on the facts of those cases. But they did succeed in Papamichael v National Westminster Bank plc [2003] EWHC 164 (Comm); [2003] 1 Lloyd’s Rep 341. The claimant’s husband paid her lottery winnings into his bank account with a correspondent bank of the defendant bank. The defendant converted the drachmas into US Dollars and credited them to an account of the husband. The bank was held liable for dishonestly assisting the husband’s breach of constructive trust. The amount of the loss was the value of the claimant wife’s winnings. There may be an issue whether that was a constructive or resulting trust but, as Etherton J said in London Allied Holdings Ltd v Lee [2007] EWHC 2061 (Ch), para 276, and I agree, that makes no difference to the principle.

41. If the dishonest assistant can be liable for breach of a constructive trust then a fortiori the trustee must be liable, and the liability of both of them must include liability to compensate for loss caused to the beneficiary by the breach. Unfortunately none of those cases concerned constructive trusts of unauthorised profits made in breach of trust or (as here) of fiduciary duty, and none have been found. But the absence of specific authority cannot on its own justify treating such constructive trusts as exceptions from the generally accepted principle. There may be a number of explanations for the silence. Alleging and proving dishonesty against the assistant is itself a serious hurdle, not surmounted in any of the above cases other than Papamichael, and it may be that the trustee defendants in the secret profits cases were of sufficient substance to repay or that they had not dissipated the profits before being held to account. In many of the secret profits cases there was no dishonesty involved from start to finish, on the part of anyone.

42. I would therefore reject both the premise for, and the second supposed consequence of, the proposition that this type of constructive trust is, in substance, just a remedy. It is entirely inconsistent with recent judicial thinking about this constructive trust, in the bribery cases, which are just about a particular species of secret or unauthorised profit. It runs counter to the very recent decision of this court, in Rukhadze, which was about unauthorised profits. The constructive trust of profits imposes the usual obligation on the constructive trustee not to dissipate the trust property, and the usual obligation on both him, and upon any dishonest assistant in the dissipation, to compensate the beneficiary for any loss caused thereby.

No loss caused by the dissipation?

43. The basis upon which loss qualifying for equitable compensation for breach of trust is to be assessed was so fundamentally re-stated in Target Holdings Ltd v Redferns [1996] AC 421 that (although affirmed in AIB Group (UK) plc v Mark Redler & Co [2014] UKSC 58; [2015] AC 1503) it stands as an almost complete statement of the relevant principle. It requires the assessment of loss to be conducted through a “but-for” counterfactual: ie what would have been the beneficiary’s position but for the breach of trust? The loss is assessed by comparing that but-for position with the position in which the beneficiary finds himself as the result of the breach.

44. This new principle has generally been regarded as a healthy implant of common sense, and from a common law origin: eg see Rukhadze, paras 6, 59. But it does not bring with it the rest of the scope of duty, causation, remoteness and loss analysis usually applied by the common law to a tort or a breach of contract. The facts of the Target case are a stark illustration of the need to apply the new but-for principle in order to avoid manifest injustice and inequity. Solicitors holding £1.525m of their client Target’s money on bare trust, with instructions to exchange it (by way of loan) for valid security by way of mortgage of specified property, paid almost all of it to the borrower, in breach of trust, about a month before receipt of valid mortgage security. That security later proved inadequate to repay the loan, selling for only £0.5m, leaving the client with a shortfall of approximately £1m, which it sought to recover from the solicitors. By way of summary judgment the Court of Appeal held the solicitors liable to restore the whole of the money paid away, less the realised value of the security property.

45. The House of Lords held that Target had suffered no loss from the solicitors’ breach of trust because, but for that breach, there was a triable issue that the money would still have been duly paid to the borrower in exchange for the security over the property a month later, and the same shortfall would still have occurred. There was of course a risk of a much larger loss during the month between payment out and provision of the security but, by carrying out the loss assessment at the trial date rather than the date of the breach, it could be seen that the short-term risk of losing the whole advance did not in the event materialise.

46. The sensible decision of the House of Lords put right a manifest injustice to the solicitors. The real cause of the loss was not that they had paid out the advance early, but that Target had been provided with a gross overvaluation of the security property (at £2m) and had been lied to by the borrower about the purchase price (£775,000) and that there was then a fall in the market value of the property before the security was realised. None of that was the result of the early payment of the trust fund by the solicitors. There was a separate negligence claim against the solicitors for failing to inform Target of aspects about the transaction of which Target had been kept in the dark, but that was not the basis of the Court of Appeal’s order, which was made while the negligence claim was awaiting trial. Although Lord Browne-Wilkinson described the transaction as “redolent of fraud and negligence” (at p 432B), he tested the underlying principle on the hypothetical basis of an entirely innocent (rather than fraudulent or even negligent) payment by solicitors one day ahead of receipt of the security documents, followed by a loss to the lender client entirely attributable to a fall in the property market.

47. The House of Lords’ analysis broke new ground in two respects. First, it moved away from the notion (accepted by the Court of Appeal) that the liability of a trustee for a breach of trust consisting of an unauthorised payment away of the trust fund was always to reconstitute the fund. At least their Lordships did so in relation to a commercial or business trust under which, by the time of trial, there was a single beneficiary absolutely entitled. Secondly (and in reliance in particular upon In re Dawson decd [1966] 2 NSWR 211 and Canson Enterprises Ltd v Boughton & Co (1991) 85 DLR (4th) 129), their Lordships decisively preferred the trial date rather than the breach date as the time for the assessment of loss.

48. In that latter context Lord Browne-Wilkinson said at p 437D-E that there was no justification for “stopping the clock” immediately upon the breach. To do so, he said:

“may… lead to compensating the trust estate or the beneficiary for a loss which, on the facts known at trial, it has never suffered.”

His conclusion, at p 439A-B, was that:

“Equitable compensation for breach of trust is designed to achieve exactly what the word compensation suggests: to make good a loss in fact suffered by the beneficiaries and which, using hindsight and common sense, can be seen to have been caused by the breach.”

49. Much reliance upon the disapprobation of “stopping the clock” was placed by Mr Kokelaar. He argued by parity of reasoning that there should be no starting of the clock only at the time of the breach consisting of the dissipation of the dividend, and thereby to ignore the earlier connected breach consisting of the making of the unauthorised profit. There are I think three answers to this superficially simple and attractive submission, at least on the question whether it can be derived from anything said in Target v Redferns. The first is that Target was about a single breach of trust, consisting of the unauthorised payment out of a fund, which did not represent unauthorised profits or the proceeds of any earlier breach. It is not to be supposed that their Lordships had the present question remotely in mind.

50. The second is that the reason for the disapprobation of stopping the clock in Target was to cure the court of the occasional myopia which arises from a slavish observance of the supposed common law breach-date rule for the assessment of loss. It had nothing to do with the question whether the assessment of loss occasioned by the dissipation of a trust fund is to be affected by the fact that, looking back before the breach, the fund itself represents unauthorised profits derived from an earlier breach of trust.

51. The third answer is that Target is essentially about whether an undoubted loss (there the shortfall in recovery of the advance) was itself caused by the defendant’s breach of trust, or by something else. Here the quite different question is whether there was a loss at all. No-one would doubt that the dissipation of the dividend caused HPII to be deprived of the value of its beneficial interest in it under the constructive trust. To see that does not require the benefit of hindsight. And but for the breach constituted by the dissipation HPII would still enjoy the beneficial interest in the dividend. But the question is whether the fact that the dividend represented the fruits of an earlier breach of trust means that its deprivation is not to be regarded as a loss to HPII. Nothing in Target sheds any light on the answer to that very different question.

52. The same is to be said of AIB v Redler. Solicitors were instructed to complete a loan transaction for their client lender by obtaining a first charge over the security property. Negligently and in breach of trust they paid out the whole of the loan while omitting to secure the discharge of an existing charge. So the lender only obtained a second charge. When the borrower defaulted the property had fallen greatly in value and was insufficient to discharge the first charge and repay the loan. It would have been insufficient to repay the loan even if there had been no first charge. The question was whether the solicitors were liable to their client for the whole of the shortfall or only for the amount outstanding under the first charge. Affirming the Court of Appeal (which had applied the principle laid down in Target) this court held the latter.

53. The decision is notable mainly for its careful rejection of a barrage of mainly academic criticism of Target, and also for Lord Toulson’s careful explanation that (contrary to what might have been deduced from Lord Browne-Wilkinson’s reasoning) the same but-for principle applied to the assessment of loss for breach of a traditional trust as applied to a commercial trust created for the purposes of a finance transaction. At para 67 he said:

“A traditional trust will typically govern the ownership-management of property for a group of potential beneficiaries over a lengthy number of years. If the trustee makes an unauthorised disposal of the trust property, the obvious remedy is to require him to restore the assets or their monetary value. It is likely to be the only way to put the beneficiaries in the same position as if the breach had not occurred. It is a real loss which is being made good.” (my emphasis).

54. Neither Target nor AIB v Redler were about an institutional constructive trust, but I can conceive no reason why the same principles should not be applied to the assessment of loss caused by breach of it, as were applied in those two famous cases to an express trust. The only complication may be that, whereas Lord Toulson pointed to the contractual terms which governed the trust of the loan money in the solicitors’ hands (i.e. the terms of their retainer) for the purpose of establishing what would have happened but for the breach of trust (i.e. the but-for counterfactual), in the case of a constructive trust there are by definition no such express terms.

55. But that is no real obstacle. In the present case the breach consists of the unauthorised dissipation of the dividend, and it is not in dispute that, at the very least, the duty imposed on Mr Ruhan as constructive trustee was to inform HPII that he was holding the dividend on their behalf, to seek HPII’s directions as to what to do with it, and in the meantime to keep it safe and whole. The loss caused by the breach is therefore the same as that in Lord Toulson’s example in para 67, namely the value of the assets paid away.

56. Like Target however, AIB v Redler was not about a trust fund representing the profits derived from an earlier breach of trust. It was just another single breach case. Although therefore it may be said that Lord Toulson’s analysis in para 67 fits the present case like a glove (thereby supporting the appellant’s submissions on this appeal), we may be sure that he did not have in mind the present question, whether the beneficiary has to aggregate all prior breaches of trust with the breach actually complained of, to ascertain whether there has been any net loss. In short, the but-for technique laid down in Target and affirmed in AIB v Redler, although undoubtedly applicable to the present case (and the contrary is not seriously suggested) begs rather than answers the present question.

57. I return to the issue whether the supposed fact that the constructive trust is a remedy for an earlier breach of trust, in making the profits in the first place, supports the aggregation of the two breaches (making the profit and then dissipating it) for the purposes of applying the Target but-for test. I have of course rejected the premise for this supposed outcome, but I will address the outcome nonetheless. In my judgment it does not support the aggregation of the two breaches. My main reason for that conclusion is that, if it did, then it would render the intended real effect of the constructive trust illusory, at least in depriving the beneficiary of any remedy for the loss of his beneficial interest in the trust property, against the dishonest assistant in the dissipation, and for no reason having anything to do with fairness, equity or justice.

58. I have already noted that the question whether the breaches are to be aggregated for the purpose of assessing the beneficiary’s loss makes no practical difference to the personal claim of the beneficiary against the profiteer constructive trustee. He is already personally liable to account for the profit. If he does not dissipate it, he can perform that account by delivering the trust property representing the profit in specie to the beneficiary. If he does dissipate it, he is still personally liable to account for it, and must find the means to do so from his own resources. He does not have to pay twice. But he may, like Mr Ruhan, now be a man of straw, and/or bankrupt. That is why, by means of the constructive trust, equity gives a proprietary remedy against the property and its proceeds, and a personal remedy against the dishonest assistant in any dissipation of the fund.

59. It must be supposed that the constructive trust of an unauthorised profit is intended by equity to have some useful effect. And the function of the court is to validate rather than inhibit that effect. The intent of equity is that the beneficiary should be the owner of the unauthorised profit, because the constructive trust is a real (ie institutional) trust. The potentially valuable proprietary remedies may be rendered nugatory by dissipation of the trust property. But the constructive trust also provides a vital means of recourse for compensation for loss caused by the dissipation, against both the trustee and the dishonest assistant (i.e. one who assists in the dissipation), but only if the breach consisting of the making of the profit is not airbrushed out of the counterfactual by being aggregated with the breach consisting of its dissipation.

60. I have recorded Mr Kokelaar’s inevitable concession that (apart from any fee for his assistance) the dishonest assistant would get away scot-free from his dishonest assistance in the dissipation if that were to be the outcome of any case where the subject matter of the constructive trust was an unauthorised profit. I can envisage no reason why that should be a just or equitable outcome. Nor does it follow from the application of the Novoship principle, which does not protect the dishonest assistant from joint liability with the trustee to compensate for loss. There is nothing in Target or AIB v Redler to suggest that this is how the but-for test should be applied in such a case.

61. My second, perhaps more technical, reason is that if the application of the Target but-for test requires it to be assumed that there was no breach consisting of the making of the profit in the first place, then the counterfactual would not merely operate “but-for the breach” but also but for the constructive trust which creates the duty not to dissipate. If the trustee is assumed not to have made the profit, there can be no profit, but also no constructive trust of it. That seems to me to be a case of using a technique to assess a loss caused by a breach in a way that throws out the baby with the bath water. The but-for test is a familiar tool for use in assessing loss caused by a breach of contract. It requires the court to assume that, but for its breach, the contract was performed rather than broken. But it never, as far as I know, requires you to assume that there was no contract in the first place. On the contrary, as Lord Toulson explains in AIB v Redler, the terms of the contract or express trust (or, I would add, the constructive trust) which has been broken are a vital tool in answering the question where the claimant would be but for the breach.

62. I would therefore conclude that, applying the but-for analysis of the loss caused by the dissipation of the whole of the dividend, which Mr Stevens dishonestly assisted, it caused HPII to lose the whole of the value of its real beneficial interest in that dividend, regardless of the fact that the dividend was itself the fruit of an earlier breach of trust in making the profit in the first place. The counterfactual is that Mr Ruhan performed his duty under the constructive trust and preserved HPII’s beneficial interest in it, not that there was no profit and no constructive trust, and therefore nothing for HPII to lose.

Equitable set-off

63. The answer to this further issue lies in my judgment in a close analysis of the supposed sub-rule, first enunciated in Bartlett v Barclays Bank Trust Co Ltd (Nos 1 and 2) [1980] Ch 515, that the general rule that a trustee is not permitted to set off gains caused by one breach of trust against losses incurred by another breach is subject to the exception that he may do so when the two breaches arise in the same transaction. I will call the general rule the no set-off principle. If the exception to it were as simple and mechanical as depending upon whether the multiple breaches formed part of a single transaction, that would not assist Mr Stevens, because the transactions by which the profits were made were in no recognisable sense the same transactions as those by which the profits were then dissipated. But might it be possible to build on that same transaction exception by treating the making of the profit and its dissipation as parts of one dishonest scheme, in the sense that neither was probably going to happen without the other?

64. There is, at the outset, something rather fantastical (rather than equitable) about the notion that a liability which would ordinarily be incurred from participation in a series of events can be expunged by alleging they were all part of the same fraudulent scheme. It sounds like using fraud to ravel rather than unravel all. But that is a consideration better addressed at the end, rather than the beginning, of the analysis.

65. There is surprisingly little authority, other than Bartlett itself, for this exception to the no set-off principle. Prior to 1980, when Bartlett was reported, all the main equity and trust text books stated the no set-off principle in fairly rigid terms. That it might be subject to an exception was referred to mainly in footnotes. Thus, in Snell’s Principles of Equity, 27th ed in 1973 it is said (at p 276) in apparently unqualified terms that:

“…if in fact a profit accrues from the breach of trust, the beneficiaries may claim it; and a profit made in one unauthorised transaction cannot even be set off against a loss incurred in another unauthorised transaction.”

Avid readers of the footnotes will have their attention drawn to Dimes v Scott (1827) 4 Russ 195; 38 ER 778 and Wiles v Gresham (1854) 2 Drew 258; 61 ER 718 as authority for the no set-off principle. Fletcher v Green (1864) 33 Beav 426; 55 ER 433 is briefly mentioned as explicable on the ground that the gain and the loss were part of a single transaction. They are analysed below.

66. In Lewin on Trusts, 16th ed in 1964, under the heading “Profit not set off against loss” there is a full description of Dimes v Scott, followed by the observation that, at the (then) present day, set-off would be allowed if there was any connection between the loss-making and profitable transactions, “and perhaps even if they were merely part of the same policy of investing in breach of trust”. As will appear, the quoted passage proved to be prophetic. Fletcher v Green is cited in the footnotes as authority for the exception based upon connection. But that case is a very doubtful basis for an exception couched in such very wide terms.

67. Halsbury’s Laws of England, 3rd ed at vol 38 (1962) states the no set-off principle in unqualified terms, provided that the breaches of trust are distinct, under the heading at para 1804: Absence of right of set-off in case of breaches of trust. Adye v Feuilleteau (1783) All ER Rep 632 is added to the authorities in support of the no set-off principle. Fletcher v Green is referred to in a footnote as a case of a single transaction, and in para 1810 as allowing a trustee who has invested money on insufficient security to take the benefit of a gain realised on the investment of the proceeds of its realisation.

68. Finally, Underhill’s Law of Trusts and Trustees, 12th ed (1970) at p 634 states the no set-off principle as its Article 90: “Set-off of Gain on one Breach against Loss on another not allowed”. The editors then mention Vyse v Foster (1872) LR 8 Ch App 309 and summarise Wiles v Gresham and Dimes v Scott before finishing with Fletcher v Green and concluding that its outcome seemed reasonable and in accordance with common sense and justice.

69. The authorities referred to in the textbooks current in 1980 provide little beyond example in the way of an explanation for the existence of, and apparent exception to, the no set-off principle. An early trenchant statement of the principle is to be found in Adye v Feuilleteau, as follows, at p 633 (per Lord Loughborough):

“In some of the reports a confused notion prevails that an executor or trustee is not answerable for the loss, where he would be answerable for the profits, but I take that to be quite erroneous, and that it has been long established in this court, that in these cases everything shall be taken against the executor: if any profits are made, he must account for them; if any loss happens, he must bear it. It does not alter the case that the executor has improved the estate by lending money on personal security, for the court will not consider the whole account of his dealings together, but must consider every single transaction by itself.”

70. Decisions which simply support the no set-off principle include Dimes v Scott and Wiles v Gresham. In the latter case Kindersley V-C said, at p 271:

“When there are two separate funds, subject to trusts, and the trustees commit a breach of trust as to one, by which it is lost, I think it impossible to permit the trustees to say, ‘We have improved the other fund, and that fund is bound to make up the loss on the other.’ That I cannot hold. If the trustees have lost one part of the settled funds, they must answer for it, whatever may be the improvement of the other part.”

71. The only authorities before Bartlett which might be said to support the exception are Earl of Winchelsea v Norcliffe (1686) 1 Vern 435; 23 ER 569, Fletcher v Green and Vyse v Foster. In the Earl of Winchelsea trustees used trust money in breach of trust to buy land to be held as an accretion to the trust, if the infant beneficiary approved the transaction upon attaining his majority, but he died. The trustees were held liable for interest on the sum so used, but were allowed to set off gains made on the purchased property. The reasoning is obscure (largely due to the passage of time) but it may perhaps be regarded as an example of gains and losses made and incurred on a single transaction. On any view it was a fair and equitable outcome.

72. In Fletcher v Green trustees lent trust money upon inadequate security, and in breach of trust. A mortgage action was started to realise the security, during which the property was sold and the proceeds paid into court and invested in Consols. They appreciated in value before the fund in court was released and returned to the trust fund. The trustees were allowed to set the gain in the Consols against the shortfall in the security. This was not really a case of multiple breaches of trust at all. The only breach was the original loan. The gain arose not from breach of trust, but from a competent salvage operation to realise the security. The gain was made while the proceeds were held in court in the mortgage action, rather than part of the trust fund at the free disposal of the trustees. Again, the outcome made obvious equitable good sense, because it really did reduce the shortfall on the realisation of the security.

73. Finally, in Vyse v Foster executors used income from the personal estate of the deceased to improve property forming part of the real estate, thereby improving its value. The estate was a mixed fund of realty and personalty held for the same beneficiaries. The executors were given credit for the appreciation in value of the realty against the amount of the personalty deployed in its development. It is hard to see how there was more than one breach of trust. On appeal to the Court of Appeal (later approved in the House of Lords) James LJ said (at pp 336-337):

“As the real and personal estate constituted one fund, we think it neither reasonable nor just to fix the trustees with a sum, part of the estate, bona fide laid out on other part of the estate in the exercise of their judgment as the best means of increasing the value of the whole. If they were mistaken in this, which does by no means appear, the utmost they could be fairly chargeable with would be the loss, if any, occasioned by the mistake in judgment.”

74. This was the rather unpromising material facing Brightman J (LJ by the time he gave judgment) in Bartlett. The defendant trust corporation (a department of Barclays Bank) held all the shares in a property company upon the trusts of a family settlement. The directors of the company announced that they intended to embark upon a policy of speculative investment in property development projects, designed to raise the value of the company for the purpose of paying anticipated future death duties. The bank did nothing as controlling shareholder to stop them. Two projects were undertaken, one in Guildford and the other near the Old Bailey in London. The Guildford project made a modest profit, part of which was ploughed into the London project, which made a disastrous loss. The bank was held liable for breach of trust for failing to stop the directors pursuing the speculative development policy, but allowed to set off the profit on the Guildford project in diminution of their liability for the loss on the London project. Brightman LJ described the bank’s liability as one of non-feasance rather than misfeasance (p 531A). It was not conduct amounting to breach of the fiduciary aspects of its duty as trustee, but rather breach of a professional duty of care as trustee.

75. The set-off issue was dealt with summarily at the end of a long judgment following a 40 day trial. Brightman LJ said this, at p 538:

“The general rule as stated in all the textbooks, with some reservations, is that where a trustee is liable in respect of distinct breaches of trust, one of which has resulted in a loss and the other in a gain, he is not entitled to set off the gain against the loss, unless they arise in the same transaction”.

Brightman LJ then referred to the textbooks reviewed above, and continued:

“The relevant cases are, however, not altogether easy to reconcile. All are centenarians and none is quite like the present. The Guildford development stemmed from exactly the same policy and (to a lesser degree because it proceeded less far) exemplified the same folly as the Old Bailey project. Part of the profit was in fact used to finance the Old Bailey disaster. By sheer luck the gamble paid off handsomely, on capital account. I think it would be unjust to deprive the bank of this element of salvage in the course of assessing the cost of the shipwreck. My order will therefore reflect the bank’s right to an appropriate set-off.”

76. The first point to note about that passage is that Brightman LJ was plainly not applying any “same transaction” exception to the no set-off principle. The Guildford and London projects were clearly separate transactions, albeit animated by the same foolhardy policy. Nor was Brightman LJ expanding the “same transaction” exception into a general “connected transaction” exception as suggested by the editors of the then edition of Lewin. His observation that the authorities were not easy to reconcile with the rule which he extracted from the textbooks suggests that he was well aware that Lewin’s very much wider statement of the exception was not at all supported by Fletcher v Green. Nor was he simply treating the case as an example of equitable set-off.

77. Rather, if Brightman LJ’s reasoning is viewed in the round, he was concluding that the particular facts about the case before him simply made it unjust to apply the no set-off principle in the circumstances. The relevant facts were that the breach of trust consisted of a negligent failure, by the bank acting in good faith, to use its powers as shareholder to stop the directors embarking upon and then implementing the inappropriate and risky property development policy that led to both the relevant transactions. It was not just that there was a connection between them, but because the connection consisted precisely of the same unwise policy. The bank’s failure to prevent the adoption of this policy by the directors was the bank’s single but continuing breach of trust. Furthermore, his reference to salvage in a shipwreck rather suggests that he had Fletcher v Green primarily in mind, which was, as I have already described, a sort of salvage case.

78. Bartlett has received no in-depth analysis or criticism in the subsequent editions of the text books relied upon by Brightman LJ. In Snell’s Equity, 35th ed (2025) at para 30-019, Fletcher v Green continues to be cited as the sole authority for a same transaction exception, while Bartlett is added as illustrative of a “same wrongful policy” exception. In Lewin on Trusts, 20th ed (2020) at para 41-025, the exception is widened further to accommodate connected breaches, on the supposed authority of Fletcher v Green, Vyse v Foster and (surprisingly) AIB v Redler, at para 57, where Lord Toulson simply said (in a single breach case) that “in determining the value of what has been lost, the court must take into account any offsetting benefits received”. Lewin then states that Bartlett justified a “same improper policy” exception.

79. The current edition of Halsbury’s Laws of England states at vol 98, para 681 that the no set-off principle is subject to same transaction and same policy exceptions, relying upon Bartlett for the latter. Finally, the current edition of Underhill cites Bartlett in support of the no set-off principle, and at para 93.7 treats it as in substance a single breach case.

80. Both Males LJ in the Court of Appeal in the present case and Mr Kokelaar in response to a question from the court treated the exception to the no set-off principle as really an aspect of the principles applied in equitable set-off generally. I do not think that this can be so, other than at a very high level of generality. The basic modern requirement for equitable set-off is that there should be both a close connection between claim and cross-claim and a finding that it would be unjust to enforce the claim without taking the cross-claim into account: see Halsbury’s Laws of England, 5th ed (2020), vol 11, para 404 and Geldof Metaalconstructie NV v Simon Carves Ltd [2010] EWCA Civ 667; [2011] 1 Lloyd’s Rep 517. The close connection is described as a formal requirement and the injustice as a functional requirement.

81. The first difficulty with a direct application of this principle is that the multiple breach of trust situation is not one which gives rise to a claim and a cross-claim, at least in any conventional sense. A trustee who like Mr Ruhan makes an unauthorised profit from his fiduciary position does not thereby have a cross-claim against his beneficiary. On the contrary, equity treats the profit as belonging to the beneficiary. The second more fundamental problem is that whereas equitable set-off is equity’s positive response to the occasionally harsh strictures of the common law, the starting position of equity in the case of multiple breaches of trust is that there should be no set-off. The third more general problem is that whereas equitable set-off ordinarily operates between persons (such as contracting counterparties) with no special or fiduciary relationship between them, the no set-off principle is specifically fashioned to regulate the quantification of loss in a fiduciary relationship of trustee and beneficiary, and as part of a generally rigorous regime designed to enforce fiduciary obligations: see again Rukhadze.

82. In my judgment, for the reasons which follow, the true principle (a label which I prefer to a rule) is better expressed as follows. The general principle applied by equity where gains and losses are made and incurred for the trust estate by a series of breaches of trust is that one breach may not be set off against the other. Its effect is that the beneficiary is entitled to the gains, but the trustee must bear the losses. This may seem tough, but equity is habitually strict in compelling the discharge of fiduciary obligation. But the court may recognise an exception where the application of the no set-off rule would, usually because of a particular type of connection between the breaches concerned, produce a clearly inequitable result. The recognition of this exception traditionally formed part of the court of equity’s discretion to ameliorate the strict consequences of the traditional rules as to accounting by trustees, but I can see no reason why it is not equally applicable to claims for equitable compensation. Indeed, the language of some of the old cases, in particular Vyse v Foster at p 333, suggests that the discretion was so applied. I have elsewhere described another aspect of this discretion as tempering the wind to the shorn lamb.

83. By way of amplification, it is not the mere existence per se of a connection, or even a close connection, between the relevant breaches of trust which may make the denial of a set-off inequitable. The connection merely sets the scene for asking the often fact-intensive question whether the ordinary application or the disapplication of the no set-off principle better serves the objectives of equity in the particular circumstances. Nor is that question always to be answered by asking the often rather metaphysical question whether the relevant breaches all form part of the same transaction. For some purposes they may have, for other purposes they may not. Finally, the burden of demonstrating a qualifying inequity lies on the person who seeks to disapply the no set-off principle, here Mr Stevens. That is because the no set-off principle is itself there to serve a purpose of equity, so that a person seeking to demonstrate an inequitable result must demonstrate that the underlying purpose of the principle is outweighed by other weightier considerations of an equitable nature.

84. There is little learning about equity’s underlying purpose for the no set-off rule. That may be because nineteenth century judges thought it too well settled to need to be justified, and perhaps because equitable principles had temporarily degenerated into something of a rule book. But the purpose is not difficult to divine. It is because gains made as the result of a breach of trust are, generally, for the account of the beneficiary, not the trustee. This is plainly the case where the gain consists of an unauthorised profit, as the cases from Keech v Sandford, including the recent bribery cases, make crystal clear.

85. The various cases on the no set-off principle reviewed above fairly fall within the principle and the exception to it which I have tried to explain. In Fletcher v Green the gain was made by the trustees in the course of a perfectly proper attempt by them to mitigate the loss already caused to the beneficiaries by their unauthorised and under-secured loan of part of the trust fund. It was not really a separate breach of trust at all, but a partially successful salvage exercise following a breach of trust. There was an obvious connection between the original loss and the later gain, such that it would have been manifestly unfair and therefore inequitable for the trustees not to have been allowed to pray it in aid in diminution of the shortfall, and therefore of the beneficiaries’ loss.

86. In Vyse v Foster the breach of trust consisted in the use of one part of the fund in gainfully developing property consisting of another part of the fund. The gain was closely connected with the “loss” (if that is really what it was) in the sense that the money was only paid away to enable the gain to be made, which in the event it was. The disallowance of the gain made by the use of the money would have been wholly unjust because the same persons beneficially owned both parts of the fund (personalty and realty). And the only reason why the development of the realty could even be described as a breach of trust is because an unauthorised use of trust money was made to bring it about.

87. In Bartlett there were undoubtedly two transactions, both of which the bank trustee failed negligently to prevent. But they were closely connected because the trustee’s involvement in them consisted in the same single continuing breach by omission, namely not preventing or stopping the directors’ adoption of a foolishly risky policy. In all three cases, unlike the present, the inequity arose as between the trustees and the beneficiaries.

88. In the present case there were three (or strictly four) distinct transactions. That is why Mr Stevens can derive no assistance from the supposed single transaction exception. The first was the dishonest purchase by Mr Ruhan (through and concealed by a nominee) of the three hotels from HPII in March 2005, in breach of the self-dealing rule. The second was the dishonest but highly profitable sales of the hotels in two lots in August 2006 and March 2008, in breach of the profit rule. They were dishonest because Mr Ruhan’s participation in the on-sales continued to be concealed from HPII, so that it was unable to lay its hands on, or freeze, the profit held for it upon constructive trust. I shall treat them, for convenience but contrary to the strict fact, as a single transaction. The third was the subsequent dissipation of the profits received in the form of the dividends on entirely unconnected loss-making projects of Mr Ruhan overseas, in breach of the terms of the constructive trust under which he held the profits. That was again dishonest because it was concealed from HPII.

89. There is an irregular pattern of connection between the three transactions. The first and second were quite widely separated in time, but they did form two parts of a dishonest scheme by Mr Ruhan to sell what started out as HPII’s property for a big hoped-for profit which HPII did not have the ability to make for itself. Although the profitable on-sales were not themselves actually pre-ordained (since the prospect of profit was too speculative to have affected the market value of the hotels when Mr Ruhan bought them in 2005), there can be no doubt that the 2005 purchase was motivated by a desire by Mr Ruhan to make a profit on re-sale or development.

90. In contrast the connection between the second and third transactions, although much more closely connected in time, is more diffuse. The loss caused to HPII by the third transaction was not because Mr Ruhan used the money for his own purposes, which of course he always intended to do. It was caused by his use of it in a way that destroyed HPII’s proprietary interest in it, mainly on paying down loans on overseas property schemes which later failed. There is no suggestion that his destruction of HPII’s beneficial interest by dissipation of the dividend was deliberate (in the sense that a dissipation outcome was either inevitable or part of any pre-ordained plan). Mr Ruhan might have spent what he probably (but wrongly) thought were his ill-gotten profits on the purchase of a villa in the South of France, or upon a large house in Belgravia. In either case HPII would have retained a valuable beneficial proprietary interest in either property, as the traceable proceeds of its own property held on constructive trust. And he probably hoped when he used the dividend to support his property development schemes in the Middle East that he would end up richer at their conclusion.

91. The only connecting factors which may be said to bind together all three transactions can best be described as dishonesty and greed. All three were dishonestly concealed under the same cloak of secrecy, to keep them from HPII’s attention. And Mr Ruhan’s desire, in breach of fiduciary duty, to use HPII’s property to enrich himself may fairly be categorised as greed. These do not look like connecting factors which ought to enable Mr Ruhan to complain that the application of the no set-off principle causes him an injustice. They are utterly removed from the connecting factors which existed in the four authorities reviewed above.

92. But the equity of the matter does not lie simply or even mainly between Mr Ruhan and HPII, but rather between Mr Stevens and HPII. As already explained, the question whether Mr Ruhan now faces a personal liability to compensate HPII for loss is of no concern to Mr Stevens. Mr Ruhan is liable to account for the whole of the profit, and the effect of his dissipation of it merely deprives him of a ready means of performing that account. But it critically affects Mr Stevens, since his liability is limited to compensating HPII for its loss. The unappealing nature of the connection between the three transactions, as between Mr Stevens and HPII, is even more stark than as between Mr Ruhan and HPII. Mr Stevens simply provided the same dishonest cloak of anonymity which hid Mr Ruhan’s involvement in all three of them from HPII. Even if (as is submitted in Mr Stevens’ written case) an assistant may incur liability to compensate from relatively modest acts of assistance, it is the dishonesty which makes him liable. If his dishonest assistance in the dissipation were otherwise to make him liable, it seems strange that he should be permitted to assert a set-off contrary to general equitable principle on the ground that his dishonesty extended to earlier transactions, and supplied the necessary connection.

93. I turn therefore to ask whether the purposes of equity would be better served by allowing a set-off, as between Mr Stevens and HPII, or refusing it, on the facts of this case. I derive an answer from an examination of the underlying purposes of equity in this context to avoid a concern that a recourse to a purely subjective perception of what is or is not unjust or inequitable looks too much like measuring it by reference to the length of the Lord Chancellor’s foot. An outcome is inequitable because it is contrary to equitable principles, not because it may give rise to a feeling of unease.

94. The starting point, as already explained, lies in the determination of equity to recognise unauthorised profits made by a fiduciary as belonging in equity to the beneficiary from the moment of their receipt. This is the purpose of the constructive trust, as already described and recently affirmed both in the bribery cases and in Rukhadze. Such profits belong to the beneficiary, and cannot be the basis for a set-off by the trustee or by any dishonest assistant in their dissipation. If set-off of the gain represented by the profit is allowed against the loss caused by its dissipation, then the purpose of the constructive trust will have been entirely defeated by the combination of the dissipation and the set-off, since the usual personal liability of the trustee and of any dishonest assistant for loss caused by the dissipation will be wiped out by the set-off.

95. Take a case where the settlor and the trustee are the same person. The settlor who dissipated the trust fund which he has set up cannot claim the setting up of the fund as a gain for the beneficiaries against the loss caused by the dissipation, any more than a thief can defend a civil claim in conversion upon the basis that he gave the stolen property to the victim in the first place. And a constructive trustee cannot be in a better position than such a settlor, merely because the trust property consists of a profit made by him in breach of fiduciary duty.

96. By contrast, if a set-off is refused, then those responsible for the dissipation of the dividend will be personally liable for the loss of the beneficiary’s proprietary interest in the trust property, and the purpose of the constructive trust of the profits will have been achieved, at least to the extent of the personal resources of those responsible for the dissipation, including Mr Stevens as dishonest assistant. This analysis provides a strong, even compelling, reason for not allowing the set-off or, which is the more precisely correct test, for permitting the no set-off principle to take its usual course because the outcome is not unjust or inequitable.

97. It has been suggested that refusing a set-off in this context would be to make the dishonest assistant personally liable for the trustee’s profit, by the back door, and in conflict with the Novoship principle. I have already explained at paras 36 to 39 above why I disagree. We are here concerned with the second of two separate fiduciary relationships, first that between Mr Ruhan and HPII which was broken by his making of a secret profit, and secondly that created by the constructive trust of those profits once received, which was broken by their dissipation. If the dishonest assistant has contributed to a loss suffered by the beneficiary, it is neither here nor there that the beneficiary’s loss is equivalent to the trustee’s profit, in particular if (as is conceded here) there has been no election as against Mr Ruhan to take an account rather than compensation for loss, in relation to the breach constituted by the dissipation. The Novoship principle is there to protect a dishonest assistant from liability to disgorge the trustee’s profit, where there has been no corresponding loss to the beneficiary. It cannot be used to wipe out a real loss, by allowing a set-off of the gain which constitutes, or is represented by, the trust property which has been dissipated.

98. Mr Stevens submits that it would be unjust to refuse the set-off because the profits represented a pure windfall for HPII, so that the dissipation of them was not a loss. For that purpose he relies upon the opinion of Professor Dame Sarah Worthington in “Four Questions on Fiduciaries” (2016) 2(2) CJCCL 723 that the purpose of the constructive trust of profits is that it should deprive the fiduciary of them, but not that it should give them to the beneficiary. I consider that this understates the purpose of the constructive trust. It is clearly designed to confer ownership of the profit upon the beneficiary, as already explained. It has been doing so for over 300 years, and surely not just by accident. In any event, if the trust property was a pure gift (and so a windfall in that sense), that is no reason why its dissipation does not cause the beneficiary a loss and thereby an injustice.

99. I can conceive of no other reason why the outcome of applying the no set-off principle to the facts of this case should be regarded as inequitable or unjust, and none has been put forward. It follows that Mr Stevens’ defence based upon the exception to the no set-off principle also fails.

Summary

100. It may assist in the digestion of this over-long judgment if I summarise my essential conclusions of law, as follows:

1. Like any other trust, a constructive trust of unauthorised profits gives rise to an immediate proprietary interest of the beneficiary in the fund representing those profits, from the moment of their receipt by the trustee.
2. A dissipation of the fund by the trustee is a breach of trust for which the trustee is liable to compensate the beneficiary for the loss of its proprietary interest. That loss is generally to be assessed by reference to the value of that proprietary interest, but for the dissipation of which would still belong to the beneficiary.
3. A person who dishonestly assists the trustee in the dissipation is jointly liable with the trustee for the loss caused by the dissipation.
4. Those general principles are unaffected by the facts that (a) the fund held on constructive trust is or represents unauthorised profits made in an earlier breach of fiduciary duty to the same beneficiary, (b) the making of the profits caused the beneficiary no loss and (c) the effect of the constructive trust of the profits was to confer a gain on the beneficiary.
5. Nothing in Target v Redferns or in AIB v Redler requires or suggests that, in asking what would be the beneficiary’s position but for the breach of trust complained of (here the dissipation of the trust fund), the earlier breach of trust or (here) fiduciary duty consisting of the making of the profits, is also to be assumed not to have happened. On the contrary, the counterfactual required by those cases requires attention to be given to the terms of the trust which has been breached, in order to ascertain what would have been the consequences of the observance of those terms but for the breach. If no profits were made, there would be no constructive trust.
6. Nothing in the Novoship principle (which insulates a dishonest assistant from liability to disgorge profits which he has not himself made) prevents a dishonest assistant from being jointly liable for the loss caused by a dissipation, merely because the fund which has been dissipated consisted of, or represented, unauthorised profits in the making of which the assistant also dishonestly assisted.
7. The general equitable principle that a trustee may not set off gains against losses made or incurred by successive breaches of trust is subject to a potential exception where the disallowance of a set-off would be inequitable, but no such inequity occurs in a case where to allow the set-off would undermine the integrity and effect of the constructive trust, in particular by enabling the dishonest assistant to escape scot-free from having to compensate the beneficiary for the loss caused by the dissipation of the fund.

Conclusion

101. I would therefore allow the appeal, and restore the order made by the judge.

LORD BURROWS (DISSENTING):

1. Introduction

102. This appeal raises some important issues about equitable compensation for the equitable wrong of dishonestly assisting a breach of fiduciary duty. The issues that arise follow on some modern developments in English law. Two are of central importance.

103. The first was the clear recognition by the highest court in Target Holdings Ltd v Redferns [1996] AC 421 and AIB Group (UK) plc v Mark Redler & Co [2014] UKSC 58; [2015] AC 1503 that there is an equitable compensation remedy for loss against a trustee (or other fiduciary) for breach of trust (or other breach of fiduciary duty). In the past, the compensatory nature of that remedy was often lost in the language of accounting.

104. The second was the clarification in Royal Brunei Airlines Sdn Bhd v Tan [1995] 2 AC 378, Twinsectra Ltd v Yardley [2002] UKHL 12; [2002] 2 AC 164, and Barlow Clowes International Ltd v Eurotrust International Ltd [2005] UKPC 37; [2006] 1 WLR 1476, that there is an equitable accessory wrong of dishonestly procuring or assisting a breach of trust or other fiduciary duty. This “dishonest assistance” wrong used to be referred to as “knowing assistance of a dishonest and fraudulent design by [a trustee]” (see, eg, Barnes v Addy (1874) LR 9 Ch App 244, 251-252). But the need for the design to be dishonest or fraudulent is no longer necessary provided the design is a breach of fiduciary duty. When this equitable wrong is described as an accessory wrong what is meant is that, as with accessory liability in tort, it is dependent on a primary wrong and that the dishonest assister will be jointly and severally liable with the primary wrongdoer for loss caused (see Byers v Saudi National Bank [2023] UKSC 51; [2024] AC 1191, at paras 41 and 147).

2. The facts in outline

105. In this case Hotel Portfolio II UK Ltd (“HPII”), which is the claimant and the appellant, is alleging that it has an equitable compensation remedy against Anthony Stevens, who is the defendant and the respondent, for dishonestly assisting a breach of fiduciary duty by Andrew Ruhan. Mr Ruhan was a director of HPII and, as such, owed fiduciary duties to HPII. HPII sold various London hotels (referred to as the “Hyde Park Hotels”), at a fair market value, to a company, “Cambulo Madeira”. Mr Stevens controlled Cambulo Madeira but did so as nominee for, and on behalf of, Mr Ruhan. Mr Ruhan did not inform HPII of his interest in the sale of the hotels to Cambulo Madeira.

106. Mr Ruhan made very significant profits (£102.26m) when Cambulo Madeira resold the hotels, as premium residential property developments, to third parties. It is not in dispute that, in buying the hotels and making those unauthorised profits by self-dealing and allowing his fiduciary duty and self-interest to conflict, Mr Ruhan was acting in breach of fiduciary duty to HPII. It is also not in dispute that Mr Ruhan has since dissipated those profits. Further there is no dispute that Mr Stevens dishonestly assisted Mr Ruhan with both the making and dissipation of those profits. There is also no dispute that Mr Stevens is liable to account for the profits that he has personally made through his dishonest assistance. These were assessed at first instance (on the evidence before the court) at £1.5m, a relatively small sum in the context of the very significant profits made by Mr Ruhan.

107. The essential question is whether HPII has a claim for equitable compensation for dishonest assistance against Mr Stevens for the loss (by dissipation) of the profits made by Mr Ruhan in a situation where Mr Ruhan’s dishonest design, taken as a whole, has caused HPII no loss.

3. The decision at first instance

108. Foxton J in a very impressive judgment, [2022] EWHC 383 (Comm), dealt with various causes of action brought against Mr Ruhan and Mr Stevens, including the tort of unlawful means conspiracy as well as the equitable wrongs with which we are directly concerned. As regards the facts, it is important for our purposes that Foxton J accepted that HPII had not suffered any overall loss. In particular, he made clear that HPII did not seek to advance any case: (i) that it was paid less than the market value of the hotels; or (ii) that if it had not sold the hotels to Cambulo Madeira it would have sold them on more favourable terms to someone else; or (iii) that it would have exploited the development opportunity itself. It followed (at least in the way that the claim was pleaded) that the tort of conspiracy could not be made out because there was no loss caused to HPII and one could not treat the profits made by Mr Ruhan as a loss to HPII.

109. But in relation to Mr Ruhan’s breach of fiduciary duty, Foxton J reasoned that Mr Ruhan held the profits on constructive trust for HPII. The dissipation of those profits constituted a dissipation of trust assets by Mr Ruhan and HPII was entitled to equitable compensation against him for that loss. Following on that, Foxton J held that HPII also had an equitable compensation claim against Mr Stevens for his dishonest assistance of Mr Ruhan’s breach of fiduciary duty (ie breach of a constructive trust) by dissipating the profits. In this respect, it was central to Foxton J’s reasoning that there were separate breaches of fiduciary duty by Mr Ruhan first in making the profits by acquiring and reselling the hotels and then in dissipating the profits. Mr Stevens had assisted each of those breaches of fiduciary duty.

110. Foxton J’s decision, therefore, was that HPII’s claim against Mr Ruhan for, at its election, an account of profits or equitable compensation for breach of fiduciary duty succeeded; and HPII’s claim against Mr Stevens for, at its election, an account of profits or equitable compensation for dishonest assistance in Mr Ruhan’s breach of fiduciary duty also succeeded.

111. Nevertheless, Foxton J indicated that he had some misgivings about his decision. He said at para 296:

“I have not found the answer entirely satisfactory or wholly intuitive:

i) It might be said that the success of the argument elides many of the distinctions between claims for an account of profits and claims for equitable compensation, despite the very different nature of those two remedies and the legal regimes which govern them.

ii) In substance, HPII’s complaint here is that Mr Ruhan abused his position as a fiduciary to make a profit which HPII would not have made for itself, and that Mr Stevens dishonestly assisted him in that. It might be said that, as a matter of substance, that is a claim for an account, and it should carry whatever legal consequences follow from that categorisation.

iii) In certain factual scenarios, including this one, the argument might be said to come close to rendering the dishonest assistant liable for the profits made by the fiduciary even though English law has not chosen to render dishonest assistants directly so liable, and to permit such a claim ‘as of right’, notwithstanding the ‘strong’ discretion which exists in determining whether to order the dishonest assistant to account for their profits and (perhaps) without the benefit of the more exacting causation test which would have applied to such a claim.

iv) The result might be thought particularly strict, because of the consequences which follow from applying the causation test [applicable] to claims for dishonest assistance in the breach of purely custodial duties (as opposed to a test considering the effect on the beneficiary of the acts of dishonest assistance).”

4. The decision of the Court of Appeal

112. The Court of Appeal (Newey, Males and Birss LJJ) [2023] EWCA Civ 1120; [2024] Bus LR 160, allowed the appeal of Mr Stevens against the decision of Foxton J. The leading judgment was that of Newey LJ, with whom Males and Birss LJJ agreed. Males LJ added some further reasoning in a concurring judgment.

113. Newey LJ’s primary reasoning was that the equitable compensation claim against Mr Stevens for dishonest assistance failed because Mr Ruhan had only one scheme or course of conduct or plan and that had caused no overall loss to HPII. Newey LJ expressed this as follows at para 67:

“Standing back from the detail, there was a single and uninterrupted course of conduct which, taken as a whole, caused HPII no loss. That being so, it strikes me as just that Mr Stevens’ liability should be limited to his personal profit…. Whether or not HPII has suffered a loss should be determined by reference to the total effect of Mr Ruhan’s scheme. To put things differently, the ‘loss’ stemming from Mr Ruhan’s treatment of the profits must be balanced against the claim to recover those very profits which arose from the same plan.”

114. He also relied on what one might describe as an “election” point. He said at paras 66 and 72:

“Another way in which the sale of the Hyde Park Hotels and the compensation claim are tied together is to be found in the basis on which the profits to which the compensation claim relates are said to have been subject to a trust in favour of HPII. The trust reflects, and is a product of, the liability to account arising from the original sale. Had HPII opted against any account of profits as against Mr Ruhan, it could not have maintained the claim that the profits were held on trust for it and the foundation for the order requiring Mr Stevens to pay compensation equal to the amount of the profits would have fallen away.

I find it hard to see that HPII could both have made the election in favour of an account of profits without which there would have been no trust and have had a claim … for compensation for breach of that trust.”

115. Newey LJ also pointed to Ultraframe (UK) Ltd v Fielding [2005] EWHC 1638 (Ch); [2007] WTLR 835, para 1600, (“Ultraframe”) and Novoship (UK) Ltd v Mikhaylyuk [2014] EWCA Civ 908; [2015] QB 499 (“Novoship”) as making clear that, in English law, the assister should be liable only for the profits that the assister has made personally. That is, the liability for an account of profits is several not joint. Yet, as Newey LJ made clear, the effect of Foxton J’s decision was that, through the remedy of equitable compensation, the assister, Mr Stevens, was made liable for the profits made by Mr Ruhan rather than the profits which Mr Stevens had himself made.

116. It is to be noted that Newey LJ preferred to leave open, at paras 71-72, some of the deeper questions as to the precise nature of a constructive trust. In particular, Newey LJ preferred not to commit to whether or not there could be an equitable compensation claim against Mr Ruhan for breach of this type of constructive trust. He said:

“It is … open to serious question whether a fiduciary can incur liability to pay compensation for breaching a trust of this type…”

But, even assuming that equitable compensation could be awarded for breach of such a trust, Newey LJ’s other reasons meant that, in any event, the claim for equitable compensation against Mr Stevens for dishonest assistance should fail.

117. For completeness, I should add that Newey LJ made clear (and HPII accepts that Newey LJ was correct on this) that, in so far as Foxton J indicated that the hotels themselves were held on constructive trust for HPII, that was incorrect because the sales had not been rescinded by HPII.

118. Males LJ in his concurring judgment went on to explore the nature of a linked transaction and looked at some cases (including Barlett v Barclays Bank Trust Co Ltd (Nos 1 and 2) [1980] Ch 515) on the extent to which gains on one transaction can offset losses on another transaction.

5. The central submissions of the appellant

119. HPII has now appealed to this court against the decision of the Court of Appeal in favour of Mr Stevens. James Pickering KC, counsel for HPII, makes the central submissions that, in this case, the profits made by Mr Ruhan in breach of fiduciary duty owed to HPII were held by him on constructive trust. In dissipating those trust assets, Mr Ruhan was acting in breach of his duty as a constructive trustee and HPII was entitled to equitable compensation from him for the loss of those trust assets. It therefore follows, so it is argued, that HPII also has an equitable compensation claim for that loss against Mr Stevens who dishonestly assisted Mr Ruhan’s breach of the constructive trust.

120. In deciding whether that argument is correct, or deceptive in its simplicity, I shall first consider whether HPII was entitled to equitable compensation from Mr Ruhan for breach of this type of constructive trust by dissipation of the trust assets. This requires plunging into some deep waters in exploring the nature of this type of constructive trust.

6. The nature of this type of constructive trust

121. The precise nature of a constructive trust imposed by law (for example, as in this case, a constructive trust of profits made in breach of fiduciary duty) and not based on intention, express or implied, has long perplexed commentators. In particular, the question has been asked whether such a constructive trust is in reality a form of remedy. It is commonly said that, in English law (in contrast to, for example, Canada and Australia), a constructive trust is institutional and not remedial: see, eg, Westdeutsche Landesbank Girozentrale v Islington London Borough Council [1996] AC 669, 714-715; FHR European Ventures LLP v Cedar Capital Partners LLC [2014] UKSC 45; [2015] AC 250, at para 47; Angove’s Pty Ltd v Bailey [2016] UKSC 47; [2016] 1 WLR 3179, at paras 27-28. But it is not entirely clear what that terminology means.

122. There is a huge body of academic writing discussing these issues, an early example being Roscoe Pound, “The Progress of the Law” (1920) 33 Harv LR 420. I have found particularly helpful: Gbolahan Elias, Explaining Constructive Trusts (1990), pp 159–163; Peter Birks, “Proprietary Rights as Remedies” in The Frontiers of Liability (ed Birks, 1994), pp 214–223; Craig Rotherham, Proprietary Remedies in Context (2002), ch 1 and pp 57–63; Birke Hacker, “Proprietary Restitution after Impaired Consent Transfers: a Generalised Power Model” (2009) 68 CLJ 324; William Swadling, “The Fiction of the Constructive Trust” (2011) 64 CLP 399; Goff & Jones on Unjust Enrichment, 10th ed (2022), paras 38-16 – 38-23; Hanbury & Martin, Modern Equity, 23rd ed (2024), paras 12-029 – 12-032.

123. If the question is whether the constructive trust confers a wide judicial discretion to order or grant an appropriate remedy for a cause of action (for example, for an equitable wrong or for unjust enrichment) then it is clear that that is not how the constructive trust operates in English law. In this jurisdiction, there is no such remedial constructive trust.

124. However, the clear rejection of that analysis of a constructive trust leaves open difficult and disputed questions as to the extent to which a constructive trust, of the type with which we are here concerned, is dependent on the choice of a claimant and/or the involvement of a court (whether by the making of an order or a declaration). But whatever view is taken on those questions, it is hard to deny that, in many situations, a constructive trust is a legal response to an equitable wrong or unjust enrichment. As counsel for both parties accepted in their oral submissions, in the context of unauthorised profits made in breach of fiduciary duty, with which we are here dealing, the constructive trust of the profits can be viewed, along with an account of profits, as a legal response to the breach of fiduciary duty. Moreover, I would add, it is an initial response that then creates the possibility of the claimant being awarded, or choosing, proprietary remedies in respect of identifiable (ie traceable) assets that are effective, and in that sense, confer priority on the defendant’s insolvency.

125. If this type of constructive trust is operating as a response to a breach of fiduciary duty, along with an account of profits, it may be thought unnecessary to recognise that there is also an equitable compensation remedy for dissipation of the profits because that would merely replicate what would be achieved by requiring the constructive trustee to account for the profits made. That may perhaps explain why there appears to be no previous example in the case law of equitable compensation having been awarded in such a situation. And if there is no equitable compensation remedy against the constructive trustee in that situation, it follows, from the requirement of joint and several liability, that there can also be no equitable compensation remedy against the dishonest assister.

126. However, the important point is that, even if one accepts Mr Pickering’s submission that the imposition of a constructive trust of the profits does carry with it the possibility of an equitable compensatory remedy against the constructive trustee for dissipation of those profits, it does not necessarily follow that Mr Pickering can build on it to establish that HPII has a dishonest assistance claim against Mr Stevens for equitable compensation for the loss caused by dissipation of those profits. On the contrary, there are several significant objections to accepting Mr Pickering’s submissions on dishonest assistance.

127. Before looking at those objections, I should make clear that, with one exception (Papamichael v National Westminster Bank plc [2003] EWHC 164 (Comm); [2003] 1 Lloyd's Rep 341, a decision of HHJ Chambers QC), it cannot be said that dishonest assistance of the breach of a constructive trust has been the reasoned basis of a previous decision or fully discussed in any past case. Nevertheless, there have been several cases where it has been assumed generally that there can be dishonest assistance of a breach of a constructive trust. These include one Court of Appeal case: Heinl v Jyske Bank (Gibraltar) Ltd [1999] Lloyd’s Rep Bank 511. See also Goff & Jones on Unjust Enrichment, 10th ed (2022), para 38-15; and Paul McGrath QC, Commercial Fraud in Civil Practice, 2nd ed (2014), paras 9.34-9.35.

7. Objections to accepting the appellant’s submissions on dishonest assistance

128. A major objection to accepting Mr Pickering’s submissions is that, on the facts, there was one dishonest scheme (or design or plan) which involved the fiduciary and the assister from the start. It is wholly artificial to divide it up into the acquisition of the profits and their dissipation. The scheme comprised both. Its whole purpose was for Mr Ruhan to acquire unauthorised profits that would be laundered away out of reach. The purpose of the scheme would have been completely undermined if, having made the profits, Mr Ruhan handed them across to HPII. Mr Ruhan was assisted throughout by Mr Stevens.

129. Applying a compensatory analysis, the correct question as regards Mr Stevens is, what position would HPII have been in if the scheme had not been carried out, that is if there had been no breach of fiduciary duty by Mr Ruhan from the start? The answer is that HPII would have been in the same position as it now is. It has suffered no overall loss. In particular, it would never have made the profits that Mr Ruhan himself made (see para 108 above) so that his profits do not represent a loss to HPII. HPII no longer has the hotels but it sold them at, what the judge held, was a fair market value. Its overall financial position has been made no worse by Mr Ruhan’s and Mr Stevens’ wrongdoing.

130. When one is considering equitable compensation for the wrong of dishonest assistance, it is therefore important that one considers the breaches of fiduciary duty that have been assisted as a realistic whole. Where the principal wrongdoer has committed a breach of fiduciary duty both by making the profits and dissipating them, an assister who assists either from the start (as on the facts of this case) or only with the dissipation cannot be held liable for loss attributable just to the dissipation because the principal wrongdoer’s single wrongful scheme has produced no overall loss to the victim of the wrongdoing. It follows that I agree with the primary reasoning of Newey LJ set out at para 113 above.

131. It is noteworthy that, although now regarded as incorrect to have focused on dishonesty and fraud by the trustee, the old terminology of “knowing assistance of a dishonest and fraudulent design” by the trustee may have had the merit, in this type of case, of looking at the “design” of the fiduciary. The modern formulation of dishonestly assisting a breach of fiduciary duty should not lose sight of the “design” of which the breach of fiduciary duty forms part.

132. A second objection to awarding equitable compensation for dishonest assistance in disposing of profits obtained in breach of fiduciary duty is that it is well-established in English law that the wrong of dishonest assistance does not lead to the dishonest assister being liable to account for the profits made by the principal wrongdoer. Rather the dishonest assister is severally liable to account only for the profits that it has personally made: see the Ultraframe and Novoship cases; and analogously, in the context of being an accessory to an intellectual property tort, Lifestyle Equities CV v Ahmed [2024] UKSC 17; [2025] AC 1. To make the dishonest assister liable to compensate the beneficiary for a loss, where that loss is nothing more than the principal wrongdoer’s failure to account for the profits (or to transfer their traceable substitute), would be inconsistent with the principle that a dishonest assister is not liable to account for the profits made by the principal wrongdoer. One would be, inconsistently, and by a backdoor “compensation” route, making the assister liable to account for the profits made by the principal wrongdoer. I therefore agree with the reasoning of Newey LJ summarised at para 115 above.

133. A third objection is that, along with an account of profits, a constructive trust of the profits can be viewed as a disgorgement response imposed by the law for equitable wrongdoing by the principal wrongdoer. The constructive trust is not a compensatory response and, on the facts of this case, the principal wrongdoer has, overall, caused no loss to the beneficiary. It would be inconsistent with the law’s imposition of disgorgement, as against compensation, if a dishonest assister were held liable for compensation where the only possible way of formulating a loss is to view it as a failure, or frustration, of the disgorgement response.

134. Fourthly, Mr Pickering’s submissions in respect of the dishonest assister would produce an odd consequence. Leaving aside any claim for an account of the personal profits made by the assister, it would mean that a dishonest assister, who assisted in the actual making of the profits could not be held liable through an equitable compensation claim if, as on these facts, there was no loss to the principal. Yet a dishonest assister who merely assisted in the dissipation of those profits could be held liable, through an equitable compensation claim, for the profits dissipated. That is a distinction that appears to have no sound basis in principle or policy.

135. A fifth objection concerns what I have described at para 114 above as an “election” point. HPII, when faced with Mr Ruhan’s breach of fiduciary duty and following Foxton J’s judgment, can be said to have elected for an account of profits against Mr Ruhan rather than equitable compensation. For the need to elect between inconsistent remedies, see generally United Australia Ltd v Barclays Bank Ltd [1941] AC 1; Personal Representatives of Tang Min Sit v Capacious Investments Ltd [1996] AC 514 (although for criticism, see Stephen Watterson, “Alternative and Cumulative Remedies: What is the Difference?” [2003] Restitution Law Review 7). HPII did so for the obvious reason that it had not suffered any overall loss. It was clear therefore that an account of profits would yield a higher sum than equitable compensation. In relation to Mr Ruhan, therefore, HPII can only switch back to claiming equitable compensation for loss if it gives up its claim for an account of profits which, in principle, it could do on the ground that judgment for an account of profits against Mr Ruhan has not been satisfied: see Island Records Ltd v Tring International plc [1996] 1 WLR 1256. But if HPII were to give up its claim to an account of profits, then it could not at the same time mount an argument that Mr Ruhan’s failure to account for profits (or to transfer their traceable substitute) constitutes a loss. That is, that “constructive loss” (as I shall label it) is dependent on having elected for an account of profits. There is therefore no possible claim for compensation, including a constructive loss, against Mr Ruhan. And without a possible claim for compensation against the principal wrongdoer, Mr Ruhan, the dishonest assister, Mr Stevens, cannot be held jointly and severally liable with Mr Ruhan for equitable compensation. It again follows that, in my view, Newey LJ was correct in what he said in the passages set out at para 114 above.

136. A final objection (and very closely related to what has been said in the previous paragraph) is that Mr Pickering’s submissions lead to the potential for double recovery. Indeed, Matthew Frey and Zihang Liu in their case-note on the Court of Appeal’s decision “Equitable set off, election and constructive trusts” (2024) 140 LQR 331 argue that the key to understanding why the decision was correct is that the central problem of principle faced was the need to avoid double recovery. The problem of double recovery can be starkly shown if we hypothetically amend the facts so that, instead of HPII suffering no loss from Mr Ruhan’s breach of fiduciary duty, it is assumed that HPII suffered a loss of, say, £75m because, had there been no breach, it would itself have exploited the opportunity and would have made profits of £75m. If the appellant’s arguments were correct, it would appear that HPII would have an equitable compensation remedy for dishonest assistance against Mr Stevens for £75m plus £102.26m. The former was the loss caused by assisting Mr Ruhan’s exploitation of the opportunity and the latter was the loss caused by assisting the dissipation of the profits made by Mr Ruhan of £102.26m. If Mr Pickering’s submissions were correct, it would appear that both compensatory measures could be accumulated by HPII as against Mr Stevens because they are separate and consistent in covering different losses. Yet it surely cannot be correct – and produces double recovery – if HPII were able to accumulate both measures of compensation against Mr Stevens in that way. The conventional (albeit blunt) means of avoiding any such double recovery is by recognising that a claimant cannot combine, and must elect between, equitable compensation for compensating loss and an account of profits for disgorgement of profits. Once the claimant has elected for an account of profits against the principal wrongdoer, and not equitable compensation for loss, the possibility of double recovery would be avoided if it were to be accepted that there could be no equitable compensation against either the principal wrongdoer or the dishonest assister. That would mean that, on these facts, where HPII has elected for an account of profits against Mr Ruhan, there can be no equitable compensation remedy against Mr Stevens.

137. In my view, these objections undermine and contradict key components of Mr Pickering’s submissions on dishonest assistance, which should therefore be rejected. I essentially agree with Newey LJ’s judgment in the Court of Appeal and, in my view, Foxton J was correct to have had the misgivings about his decision set out at para 111 above.

138. For completeness I should add that I am not convinced there is much help to be gained by the thrust of the additional reasoning of Males LJ in the Court of Appeal. In his concurring judgment he focused on exploring the nature of a linked transaction and relied on some cases exploring whether gains on one transaction can offset losses on another transaction. I am unconvinced that that analogy takes matters any further. On this issue, I think it is better to focus on whether there was a single scheme, whether involving one or more breaches of fiduciary duty. Here, there clearly was such a single scheme.

139. The law abhors the dishonest wrongful conduct of Mr Stevens. But the temptation to distort equitable principles so as to award a substantial compensatory remedy against him, where remedies against the principal wrongdoer are thwarted (eg by insolvency) must be resisted. Moreover, it would be inaccurate to suggest that Mr Stevens escapes scot-free. Leaving aside any possible criminal sanctions, Mr Stevens is required to account for all the profits he has made from his dishonest assistance and, because of the dishonesty involved, there is no question of a dishonest assister being entitled to any equitable allowance for work and skill. That is the established and appropriate remedy against him given that HPII has suffered no overall loss.

8. Conclusion

140. For all the above reasons, I would therefore dismiss the appeal.