Mark Byers v The Saudi National Bank

This is the judgment of the Court. All three of us have contributed to it.

The appeal is brought by the claimants against the dismissal by Fancourt J (“the Judge”), in a judgment dated 15 January 2021, of their claim for knowing receipt. The appeal raises issues as to, first, whether such a claim depends on the claimant having had a continuing proprietary interest in the property in question when in the hands of the defendant and, secondly, whether such an interest existed in the present case having regard to the relevant Saudi Arabian law.

The claimants are Saad Investments Company Limited (“SICL”), a company registered in the Cayman Islands, and its joint official liquidators, Mr Mark Byers and Mr Hugh Dickson. The Grand Court of the Cayman Islands made a winding-up order against SICL on 18 September 2009 pursuant to a petition presented on 30 July 2009. The Cayman Islands proceedings were recognised by the English Court as foreign main proceedings pursuant to the Cross Border Insolvency Regulations 2006 by orders made on 20 August and 25 September 2009.

On 14 August 2013, Mr Byers, Mr Dickson and Mr Stephen Akers, who at the time was also one of SICL’s liquidators, issued proceedings against Samba Financial Group (“Samba”), a Saudi Arabian bank, whose assets and liabilities were transferred on 1 April 2021 to the defendant, the Saudi National Bank (“SNB”, formerly the National Commercial Bank). The claim challenged the transfer to Samba in September 2009 by Mr Maan Al-Sanea of shares in five Saudi Arabian banks (“the September Transfer”). Mr Al-Sanea was registered as the owner of the shares (“the Disputed Securities”) either (in the case of four of the five banks) at the Saudi Arabian Securities Depository Centre or (in the fifth case) on the bank’s register of shareholders, but the liquidators alleged that he had come to hold the shares on trust for SICL as a result of various transactions in 2002-2008 (“the Six Transactions”). The shares were said to have been worth about US$318 million at the date of the September Transfer.

The liquidators’ proceedings were based on section 127 of the Insolvency Act 1986, rendering void unless the Court otherwise orders “any disposition of the company’s property” made after the commencement of its winding up. That case eventually foundered when the Supreme Court held in judgments given on 1 February 2017 (Akers v Samba Financial Group [2017] UKSC 6, [2017] AC 424) that “for the purposes of section 127 of the Insolvency Act 1986 there was no disposition of any rights of SICL in relation to the shares by virtue of their transfer to Samba” (to quote from paragraph 57 of Lord Mance’s judgment). However, the Supreme Court considered that the liquidators should be given the chance to apply to the High Court for permission to re-amend their particulars of claim. The liquidators then sought to re-formulate their existing claim as one for knowing receipt of trust property, but the attempt failed and the proceedings were dismissed on limitation grounds: see Akers v Samba Financial Group [2019] EWCA Civ 416, [2019] 4 WLR 54. In the meantime, though, on 31 May 2017, the claimants had issued new proceedings, again alleging knowing receipt, and it was that claim which came on for trial before the Judge in October 2020 and with which we are concerned now. The receipt relied on was the legal transfer of the Disputed Securities into Samba’s name. There was no pleaded claim of any antecedent agreement to transfer, or of any receipt of rights under such an agreement as being receipt of property founding a claim in knowing receipt.

The trial was of limited scope. Samba had failed to comply with an order for disclosure and, as a result, had been debarred from defending the claim otherwise than on specific grounds which the Judge had decided in a judgment dated 8 April 2020 could fairly be tried without the missing disclosure. As the Judge noted in paragraph 19 of the judgment now under appeal (“the Judgment”), the effect of his April 2020 decision was that “all factual questions other than the content of Saudi Arabian law and the valuation issues were deemed to have been resolved in accordance with the claimants’ pleaded case”. In the circumstances, the only substantive issues which fell to be determined at the trial were those identified in paragraph 4 of the Judgment, namely:

Having regard to the Judge’s April 2020 decision, the following allegations by the claimants were (and are) to be taken to be true:

Until the September Transfer, Mr Al-Sanea held the Disputed Securities on trust for SICL under Cayman Islands law as the law governing their relationship in respect of the Disputed Securities, Cayman Islands law being materially identical to English law;

The purpose of the September Transfer was to discharge indebtedness of Mr Al-Sanea to Samba;

Samba knew that Mr Al-Sanea was holding the Disputed Securities on trust for SICL; and

“a reasonable bank in [Samba’s] position would have appreciated that (alternatively would or ought to have made inquiries or sought advice which would have revealed the probability that) … the September Transfer was a breach of trust; and/or … Samba recklessly failed to make such inquiries about the September Transfer … and the Disputed Securities as an honest and reasonable bank would make”.

At trial, Samba did not dispute that the knowledge which it was alleged to have had was “sufficient in principle to establish the overarching requirement for knowing receipt liability, namely that ‘the recipient’s state of knowledge should be such as to make it unconscionable for him to retain the benefit of the receipt’” (see paragraph 32 of the Judgment). As, however, the Judge noted in paragraph 33 of the Judgment, “neither in the re-amended particulars of claim nor in the amended reply to Samba’s amended defence have the claimants alleged that Samba acted dishonestly” and so the claimants “in argument – though at times they came close to alleging that Samba was an accessory to theft of the Disputed Securities – did not (and could not) argue that their pleaded case could be taken as an allegation of dishonesty, such as would be required to establish liability as a constructive trustee for dishonest assistance in a breach of trust”.

The Judge determined both the Law of Knowing Receipt Issue and the Saudi Arabian Law Issue in favour of Samba and, accordingly, dismissed the claim. He concluded in paragraph 117 of the Judgment that, “absent a continuing proprietary interest in the Disputed Securities at the time of Samba’s registration, the claim in knowing receipt as pleaded will fail” and, in paragraph 206, that “SICL had no continuing proprietary interest in the Disputed Securities after the September Transfer capable of supporting a claim against Samba in knowing receipt”.

The Judge nevertheless went on to consider the Valuation Issue, viz., “whether, in valuing the Disputed Securities, there should be a discount (‘block discount’) from the quoted market price on the relevant dates for each of the substantial holdings in the five banks, and if so the size of each discount”. The Judge explained that, had he reached a different conclusion on liability, he would have agreed with Samba that a block discount was applicable in the case of each of the holdings comprised in the Disputed Securities.

The claimants now challenge the Judge’s decision in this Court. They contend that the Judge erred in relation to each of the three issues he identified in paragraph 4 of the Judgment. In short, it is their case that (a) the claimants did not need to have a continuing proprietary interest in the Disputed Securities to succeed in their knowing receipt claim, (b) in any event, SICL’s interest in the Disputed Securities was not in fact extinguished as a matter of Saudi Arabian law and (c) the Judge was also mistaken in thinking it appropriate to apply a “block discount”.

In Barnes v Addy (1874) LR 9 Ch App 244, Lord Selborne LC said at 251-252:

The final three lines of this dictum are reflected in the two species of ancillary liability (or liability as a “constructive trustee”) referred to as “knowing receipt” and “dishonest assistance”. Liability for “dishonest assistance” arises where a person dishonestly procures or assists in a breach of trust or fiduciary duty. While “the standard by which the law determines whether [a mental state] is dishonest is objective”, negligence does not suffice. For a defendant to be liable for dishonest assistance, his mental state has to have been such as “by ordinary standards … would be characterised as dishonest” (see Barlow Clowes International Ltd v Eurotrust International Ltd [2005] UKPC 37, [2006] 1 WLR 1476, at paragraph 10, and also Ivey v Genting Casinos (UK) Ltd [2017] UKSC 67, [2018] AC 391). However, there is no requirement that the defendant should have received property to which the trust or fiduciary obligation has ever attached.

Turning to “knowing receipt”, Hoffmann LJ said in El Ajou v Dollar Land Holdings plc [1994] 2 All ER 685, at 700, that, to establish such a claim, a claimant must show:

The last of these ingredients was the subject of consideration by the Court of Appeal in Bank of Credit and Commerce International (Overseas) Ltd v Akindele [2001] Ch 437 (“Akindele”). It was there held that “[t]he recipient’s state of knowledge must be such as to make it unconscionable for him to retain the benefit of the receipt” (per Nourse LJ, with whom Ward and Sedley LJJ agreed, at 455). Nourse LJ pointed out at 448 that, “[w]hile a knowing recipient will often be found to have acted dishonestly, it has never been a prerequisite of the liability that he should”.

Before Akindele was decided, it was sometimes suggested that liability for knowing receipt was restitutionary. Thus, in El Ajou v Dollar Land Holdings plc at first instance ([1993] 3 All ER 717) Millett J, at 736, described knowing receipt as “the counterpart in equity of the common law action for money had and received” and said that “[b]oth can be classified as receipt-based restitutionary claims”. Similarly, Lord Nicholls observed in Royal Brunei Airlines v Tan [1995] 2 AC 378, at 386, that liability for knowing receipt is “restitution-based” and, writing extra-judicially in Cornish, “Restitution Past, Present and Future” (1998), said this at 238-239:

That view has not prevailed, however. In Akindele, Nourse LJ noted at 456 that, at Court of Appeal level, it would have been a “fruitless exercise” to present an argument based on Lord Nicholls’ suggestions and added this:

Liability in knowing receipt thus derives from the combination of “the beneficial receipt … of assets which are traceable as representing the assets of the plaintiff” and “[t]he recipient’s state of knowledge” having been “such as to make it unconscionable for him to retain the benefit of the receipt”. Neither is enough on its own. While it is essential that the defendant should have “received the property of another”, liability is not considered to be “triggered by the mere fact of receipt”; there must also be unconscionability. On the other hand, dishonesty is not required: the fact that the defendant must have received relevant property makes a lesser test of fault appropriate.

A recipient need not necessarily have had any knowledge or even notice of any breach of duty at the point of receipt to be liable for knowing receipt. As Millett J said in Agip (Africa) Ltd v Jackson [1990] 1 Ch 265 (“Agip”), at 291, a person who has received trust property transferred to him in breach of trust can incur liability either if he received it with notice that it was trust property and that the transfer to him was a breach of trust or “if he received it without such notice but subsequently discovered the facts”. What matters is that the recipient’s state of knowledge should have become “such as to make it unconscionable for him to retain the benefit of the receipt” (to quote, with emphasis added, from Nourse LJ in Akindele, at 455).

A recipient will, however, escape any liability if he no longer has the property by the time he learns of the relevant breach of duty. In Agip, Millett J observed at 290 that “even a volunteer who has received trust property cannot be made subject to a personal liability to account for it as a constructive trustee if he has parted with it without having previously acquired some knowledge of the existence of the trust”. Likewise, Lloyd LJ explained in Independent Trustee Services Ltd v GP Noble Trustees Ltd [2012] EWCA Civ 195, [2013] Ch 91 (“Independent Trustee Services”), at paragraph 76, that “to the extent that, before she had notice of the claim to the funds, [a Mrs Morris] had disposed of any of the money without receiving traceable proceeds, she would not be liable to the trustee”. Knowledge and possession must thus coincide for liability to arise.

Mr Jeff Chapman QC, who appeared for the claimants with Mr David Murray and Mr Adam Cloherty, suggested that it can suffice for liability that the defendant benefited from trust property regardless of whether he actually received such property. A defendant could thus, Mr Chapman submitted, be liable for knowing receipt if, say, a trustee had wrongfully spent trust money on a holiday for the defendant albeit that none of the money had ever passed through the defendant’s hands.

In our view, however, that is not correct. Success in a knowing receipt claim depends, as it seems to us, on establishing, as Hoffmann LJ’s summary indicates, “beneficial receipt by the defendant of assets which are traceable as representing the assets of the [claimant]”. A defendant must have received trust assets, not just benefited from them. As the Court of Appeal said in Novoship (UK) Ltd v Mikhaylyuk [2014] EWCA Civ 908, [2015] QB 499, at paragraph 89, “receipt of trust property is the gist of the action”. See too e.g. Satnam Investments Ltd v Dunlop Heywood & Co Ltd [1999] 3 All ER 652, at 671, and Trustor AB v Smallbone (No 2) [2001] 1 WLR 1177, at paragraphs 18-19.

There is, however, no requirement that a defendant should still have the property in question by the time a knowing receipt claim is brought. In fact, such a claim is most obviously likely to be useful where the defendant no longer has the property. If the defendant yet retains the property, the claimant may be able to recover it by asserting proprietary rights rather than by making a personal claim for knowing receipt. In Re Montagu’s Settlement Trusts [1987] 1 Ch 264, Sir Robert Megarry V-C distinguished at 276 between “the equitable doctrine of tracing” and “liability as a constructive trustee”, pointing out that “if the recipient of trust property still has the property or its traceable proceeds in his possession, he is liable to restore it unless he is a purchaser without notice”. In Arthur, Sir Terence Etherton, giving the judgment of the Privy Council, said in paragraph 34:

The question raised by the Law of Knowing Receipt Issue is whether a knowing receipt claim is possible where, although the defendant has or had property which was formerly trust property, the claimant never had a proprietary claim in respect of it against the defendant. As the quotations from Sir Robert Megarry V-C and Sir Terence Etherton indicate, that may be so where the recipient of trust property is a bona fide purchaser for value.

A proprietary claim against a recipient of trust property may be impossible, too, as a result of overreaching, the process by which trusts are transferred to other property. Lewin on Trusts, 20^th ed. (2020), explains in paragraph 44-014:

Statutory intervention may also preclude the assertion of a proprietary claim against a recipient of trust property in other ways. The regime established by the Land Registration Act 2002 provides an obvious example. Where registered land is subject to a trust, a transferee for valuable consideration will take free of the beneficiaries’ interests even if he had notice that the transfer was in breach of trust unless the beneficiaries were in actual occupation at the time of the disposition. That follows from section 29 of the 2002 Act, subsection (1) of which provides:

The Judge concluded in paragraph 114 of the Judgment that “a claim in knowing receipt, where dishonest assistance is not alleged, will fail if, at the moment of receipt, the beneficiary’s equitable proprietary interest is destroyed or overridden so that the recipient holds the property as beneficial owner of it”. A knowing recipient, the Judge said in paragraph 116, “must have held trust property, not property to which from the moment of receipt he had good title”: “[i]t is of the essence of such a claim that the beneficiary asserts that the recipient has, or had, the beneficiary’s property”. The reason for a transferee to be liable for knowing receipt, the Judge observed in paragraph 110, “is that the transferee has knowingly dealt with (or retained) property that belongs to the trust inconsistently with his duty” and “[i]f the property is not trust property, there cannot be liability of that kind”. The Judge considered there to be “a consistent line of case law” where it has either been decided that “a claim in knowing receipt cannot succeed unless the claimant has a continuing proprietary interest following the impugned transfer” or that has been assumed to be correct: see paragraph 74. In paragraph 106, he concluded that “absent dishonesty or a pre-existing equity, a registered transferee of land, whose title has priority under the [Land Registration Act 2002] over the equitable interest of the beneficiary, is not liable in knowing receipt, even if he has knowledge that the transfer was made in breach of trust”.

Mr Chapman submitted that the Judge had wrongly introduced an “additional requirement, previously unknown to the law,” into a knowing receipt claim. The Judge’s approach, Mr Chapman maintained, was justified by neither authority nor principle. Knowing receipt is a species of equitable wrongdoing in which liability is based on the defendant’s fault, viz. receiving trust property in circumstances which make it unconscionable for the defendant to retain the benefit of the receipt. The defendant’s conscience, Mr Chapman said, is equally affected whether or not the effect of the relevant transfer happens to be to override the beneficiary’s equitable proprietary right in the assets. The policy considerations likely to underlie a registration scheme (such as the importance of third parties being able to rely on accuracy of the register) may well warrant the overriding of the beneficiary’s proprietary claim to the relevant assets, but this has no effect at all on the equity of the situation as between the beneficiary and the (ex hypothesi) guilty transferee, who has unconscionably received the property with knowledge of the beneficiary’s interest. Likewise, the fact that trust interests have been overreached need not preclude a knowing receipt claim. The right of a beneficiary to bring a personal claim for knowing receipt is not dependent or parasitic on any property right he might have. Further, the Judge was mistaken in his analysis of the scope for knowing receipt claims in the context of registered land.

In contrast, Mr Brian Green QC, who, appearing with Mr Andrew Onslow QC, Mr Alan Roxburgh, Mr Edward Harrison and Ms Sarah Tulip, argued this part of the appeal for SNB, supported the Judge’s decision. The liability of a knowing recipient, Mr Green said, is a personal liability for breach of a duty to deal with property that he has received as if he were a trustee of it. The imposition of such a duty can, Mr Green submitted, be justified only where the recipient has received property which remains trust property in his hands. A claim in knowing receipt involves the assertion of a trust interest against a third party, and the fact that a knowing receipt claim is one in personam does not change the fundamental point that a property right is required to maintain such a claim against a third party. The claim is founded on the defendant’s wrongful failure to give effect to the core custodial duty to restore the property when the unconscionability of its retention is apparent. If the position under the Saudi Arabian lex situs (recognised and given effect by English private international law) is that Samba obtained good title to the Disputed Securities, free of SICL’s equitable proprietary interest, notwithstanding knowledge of that interest and of Mr Al-Sanea’s breach of trust, it would be illogical for English law to treat Samba’s receipt as nevertheless giving rise to a personal obligation to return the shares or account for their value. Further, the Judge was correct both in his treatment of the authorities to which he referred and his analysis of the position as regards registered land.

In Macmillan Inc v Bishopsgate Investment Trust plc (No 3) [1995] 1 WLR 978, shares in Berlitz, a company incorporated in New York, which were held by Bishopsgate as nominee for Macmillan were wrongfully used to secure loans to the Maxwell group of companies. Macmillan sought relief against banks which claimed to have security interests having priority to its own interest. The relevant shares having been sold, Macmillan’s writ asked for “compensation and/or damages for breach of constructive trust” (see 981). Millett J, the trial judge, summarised the parties’ positions as follows at 983:

Millett J identified the “question for determination” at 983 as “whether Macmillan retained an interest in the shares superior to that of any of the defendants and is accordingly entitled to the corresponding part of the proceeds of sale”. He explained at 990:

Millett J concluded at 987 that “the question whether any of the defendants is a bona fide purchaser for value without notice must be determined in accordance with the law of New York” and so, having regard to expert evidence on New York law, that “the critical question is whether the defendant acted honestly and in good faith when it took delivery of the certificate or whether it did not act in good faith because it knew or suspected that Macmillan was the beneficial owner of the shares or that the transfer was otherwise improper”. On the facts, Millett J found that, approaching matters on that basis, each bank either was, or derived title from, a bona fide purchaser for value without relevant notice, with the result that the action was dismissed.

In the course of his judgment, Millett J said at 988 that it was “manifestly correct to characterise Macmillan’s claim as lying in restitution”, but distinguished in that context between “the unjust enrichment of the defendant which is occasioned by depriving the plaintiff of his property” and “enrichment which results from a wrong done to the plaintiff by the defendant”. Millett J said of the distinction at 988-989 that it was “that drawn by equity between the claim of an equitable owner to recover his property, or compensation for the failure to restore it, from a person into whose hands it has come and a claim by a plaintiff in respect of a breach of fiduciary obligation owed to him” and continued:

Macmillan’s claim, Millett J said at 989, “is of the former kind”. He went on:

At 990, Millett J said:

Millett J thus regarded Macmillan’s claim as depending on whether it had a “continuing proprietary interest” in the Berlitz shares and, hence, on whether each bank was, or derived title from, a bona fide purchaser without notice. In other words, a bank which was a bona fide purchaser without notice could defeat Macmillan’s claim against it because it would have taken free of Macmillan’s interest.

Mr Chapman dismissed Millett J’s decision as irrelevant to the Law of Knowing Receipt Issue on the basis that it concerned the bona fide purchaser defence to a proprietary claim. If the writ encompassed a personal claim in knowing receipt, Mr Chapman said, the edited version of Millett J’s judgment to be found in the report does not reveal what happened to it. Macmillan, Mr Chapman argued, was simply an action by which an equitable owner sought to vindicate its proprietary rights in misapplied trust assets.

In our view, however, Millett J’s analysis cannot be discounted in this way. The terms of the writ were apt to refer to a personal claim (“compensation and/or damages for breach of constructive trust”) and, as we have already mentioned, Millett J spoke at 983 of Macmillan seeking “both proprietary and personal relief”; at 988, of “[e]ven [Macmillan’s] claim to the return of the Berlitz shares or their proceeds in specie, … which can loosely be described as proprietary”, implying the existence of another claim which could not be so described; and, at 989, of liability “to restore the shares or their proceeds to Macmillan or to pay compensation for their failure to do so” (emphasis added). Further, the category of unjust enrichment claim to which Millett J thought Macmillan’s claim belonged (viz. “the unjust enrichment of the defendant which is occasioned by depriving the plaintiff of his property”) is very far from confined to proprietary claims.

Mr Chapman also submitted that the claim brought by Macmillan would no longer be characterised (if it ever would have been) as a claim based on unjust enrichment. There is force in that: the reference in the writ to compensation/damages for “breach of constructive trust” must, we think, have included a knowing receipt claim and, as we have said, such a claim is not now seen as one for unjust enrichment. What that means, however, is that Millett J considered the availability of knowing receipt claims against the banks to depend on defeating their bona fide purchaser for value without notice contentions.

When Macmillan Inc v Bishopsgate Investment Trust plc (No 3) reached the Court of Appeal, Millett J’s decision was affirmed on different grounds, the Court of Appeal considering that “an issue as to who has title to shares in a company should be decided by the law of the place where the shares are situated (lex situs)” (to quote from Staughton LJ at 405) rather than, as Millett had thought, the lex loci actus. Consistently with Millett J’s analysis, however, the issue in the case was taken to be “whether the defendants have a defence on the ground that they were purchasers for value in good faith without notice of Macmillan’s claim” (see 398-399). Moreover, it was recognised that Macmillan’s claims were not exclusively proprietary: Staughton LJ noted at 398 that Macmillan’s claims “are to some extent proprietary” (emphasis added) and that counsel for Macmillan “points out that Macmillan claim not only a declaration as to their proprietary rights, but also an order that the defendants restore the shares to Macmillan and compensation or damages”.

Millett J’s judgment was referred to approvingly by Peter Gibson LJ in Lightning v Lightning Electrical Contractors Ltd (2009) 1 TLI 35 (“Lightning”, a 1998 decision). In that case, the plaintiff, who was resident in England, brought proceedings against a company also resident in England, for a declaration that land in Scotland was held on trust for him. The company disputed the jurisdiction of the English Court, but the Court of Appeal allowed the claim to proceed. In his judgment, Peter Gibson LJ, with whom Henry LJ agreed, said at 38:

Mr Chapman rightly pointed out that Lightning did not involve any issue as to whether a personal claim for knowing receipt could run in the absence of a continuing proprietary interest. Even so, we agree with Mr Green that the decision lends support to his case. The Court of Appeal took it that the English Court could not grant relief against a third party who had notice of an equity when it was transferred to him if “by the lex situs governing the transfer, the transfer extinguished the plaintiff’s equity”.

Mr Chapman suggested that Millett LJ’s judgment was consistent with his case. However, Millett LJ was not retreating from what he had said in Macmillan Inc v Bishopsgate Investment Trust plc (No 3). He was commenting on the law governing the relationship between the parties to a “trust”, not suggesting that the lex situs could be ignored when considering the impact of a trust on a third party.

Both sides sought support for their cases in Arthur. The central issue in that case, as Sir Terence Etherton said in paragraph 2 of his judgment, was “whether the Torrens system of land registration, as enacted in the Turks and Caicos Islands’ Registered Land Ordinance (‘the RLO’), precludes a claim that the registered proprietor of land is a constructive trustee of the land by virtue of his knowledge that the transfer was in breach of trust or fiduciary duty”. The Privy Council gave an affirmative answer on the strength of a provision in the RLO to the effect that registration of a person as a proprietor did not exonerate the person from any duty or obligation to which he was subject “as a trustee”.

At paragraph 38 of his judgment, Sir Terence Etherton said:

The assumption by the parties to which Sir Terence Etherton referred clearly accords with Mr Green’s case, but Mr Chapman said that there was more than a hint that the Privy Council did not consider the assumption to be correct. In our view, this passage offers no assistance to either party in the present case. Sir Terence Etherton was simply recording the basis on which argument had proceeded, not expressing any view, or even giving a hint, as to whether the parties’ assumption was well-founded.

The previous paragraph of Sir Terence Etherton’s judgment is of more significance than paragraph 38. Sir Terence Etherton said in paragraph 37:

The book chapter by Charles Mitchell and Stephen Watterson which Sir Terence Etherton cited argues that “when the courts say that a knowing recipient is ‘personally liable to account as a constructive trustee’, they mean exactly what they say: because of the circumstances in which knowing recipients acquire title to the misapplied property, Equity fixes them with custodial duties which are the same as some of the duties which are voluntarily assumed by express trustees” (page 130). The recipient’s “primary duty, and generally his only duty, is to restore the property immediately, so that the trust can be reconstituted and duly administered” (page 149). With regard to the requirement that a defendant must have received trust property before he can be liable in knowing receipt, the authors say at pages 157-158:

As we see it, a key premise underlying Mitchell and Watterson’s account of knowing receipt is that the beneficiaries still had an equitable interest in the relevant property at a time when the defendant had knowledge of the breach of trust. It is the existence of that equitable interest which gives rise to the custodial duties to which Sir Terence Etherton referred. In particular, it is incumbent on a recipient with knowledge of a breach of trust “to restore the property immediately” because the beneficiaries have equitable rights to it. Conversely, it is inapt to talk of a custodial duty, or a duty to restore, if the recipient acquires full and unencumbered title as a result of the transaction by which he receives the property. The need, on this understanding of the law, for the beneficiaries to retain an equitable interest can also be seen in Goff & Jones, “The Law of Unjust Enrichment”, 9^th ed. (2016), of which Mitchell and Watterson are two of the three editors, at paragraph 8-196:

Lloyd LJ analysed the law in a similar way in Independent Trustee Services, citing among other things Mitchell and Watterson’s chapter in “Constructive and Resulting Trusts”. In Independent Trustee Services, the trustee of various pension schemes alleged that money paid to the respondent had been misappropriated from the schemes. Lloyd LJ said in paragraph 81:

The proposition that a person liable as a knowing recipient will have been subject to “custodial duties” also finds support in other recent cases. In Williams v Central Bank of Nigeria [2014] UKSC 10, [2014] AC 1189 (“Williams”), Lord Sumption (with whom Lord Hughes agreed) said in paragraph 31:

Mr Chapman argued that the passage in Williams quoted above supported his submission that the essence of liability in knowing receipt is simply receipt of assets transferred in breach of trust, with the requisite knowledge, that being what “therefore” made the recipient’s possession wrongful and adverse to the interests of the true trustees and beneficiaries. We do not consider that the passage carries the implication for which Mr Chapman contends. To the contrary, the emphasis on the sole obligation of practical significance being one of restoration assumes some continuing proprietary interest in the property to be restored.

Mr Chapman further submitted that the explanation of the law given in the Mitchell and Watterson book chapter was wrong. In that connection, he pointed out that the book pre-dated Williams, which, he said, shows liability for knowing receipt to be based on the commission by the defendant of an equitable wrong. He relied in this respect on passages in Williams in which Lord Sumption, at paragraph 9, described both dishonest assisters and knowing recipients as “persons who never assumed and never intended to assume the status of a trustee, whether formally or informally, but have exposed themselves to equitable remedies by virtue of their participation in the unlawful misapplication of trust assets” and said that “[t]he intervention of equity in such cases does not reflect any pre-existing obligation but comes about solely because of the misapplication of the assets” and “is purely remedial”.

However, views very similar to those voiced by Lord Sumption had been expressed before Mitchell and Watterson wrote their book chapter, notably in a case quoted in the chapter, Paragon Finance plc v D B Thakerar & Co [1999] 1 All ER 400, in which Millett LJ referred at 413 to “the distinction between an institutional trust and a remedial formula” and said at 409 of the class of “constructive trust” to which the liability of a knowing recipient belongs that such a person “never assumes the position of a trustee, and if he receives the trust property at all it is adversely to the plaintiff by an unlawful transaction which is impugned by the plaintiff”. Further, the Supreme Court’s decision in Williams has not caused Mitchell and Watterson to alter their analysis (see e.g. Goff & Jones, “The Law of Unjust Enrichment”, at paragraph 8-201, and Underhill and Hayton: Law of Trusts and Trustees, 19^th ed. (2018), at paragraphs 98.5 and 98.33), and similar views have been expressed by other commentators notwithstanding Williams (see e.g. Chambers, “The End of Knowing Receipt” (2016) 2(1) CJCCL 1, Virgo, “The Principles of Equity & Trusts”, 3rd ed. (2018), at 596, Hanbury and Martin: Modern Equity, 22^nd ed. (2021), at paragraph 25-008, and Lewin on Trusts, at paragraphs 42-044 and 42-083). In any event, we cannot see that Mitchell and Watterson’s analysis of the law conflicts with Williams. To say, as Lord Sumption did in Williams, that knowing recipients “never assumed and never intended to assume the status of a trustee” and that equity’s intervention “does not reflect any pre-existing obligation but comes about solely because of the misapplication of the assets” is not inconsistent with Mitchell and Watterson’s conclusion that “Equity fixes [knowing recipients] with custodial duties” or with it being essential to a knowing receipt claim that the defendant should not have taken free of the beneficiaries’ interests. Lord Sumption himself spoke of the possession of a defendant having been “wrongful and adverse to the rights of both the true trustees and the beneficiaries” and of a defendant having an obligation “to restore the assets immediately”.

Turning to the Supreme Court decision in Akers v Samba (“Akers v Samba (SC)”), the leading judgment was given by Lord Mance, with whom Lords Neuberger, Sumption, Toulson and Collins all agreed. In paragraphs 20-22, Lord Mance said this:

Lord Mance thus distinguished between the position of a trustee and that of a third party. A trustee such as Mr Al-Sanea cannot deny the validity or effect of a trust on the basis that the jurisdiction in which the assets are situated does not recognise trusts or equitable proprietary interests. On the other hand, “where under the lex situs of the relevant trust property the effect of a transfer of the property by the trustee to a third party is to override any equitable interest which would otherwise subsist, that effect should be recognised as giving the transferee a defence to any claim by the beneficiary, whether proprietary or simply restitutionary” (emphasis added in each case). In other words, the availability of even a personal claim against a third party transferee depends on the continued existence of an equitable interest.

For his part, Mr Chapman sought support for his case in passages from the judgments of Lord Neuberger and, especially, Lord Sumption. In paragraph 63, Lord Neuberger said:

In our view, there is no inconsistency between this passage and Lord Mance’s judgment or, indeed, SNB’s case. It does not follow from the fact that SICL retained personal rights against Mr Al-Sanea that it also has or had such rights against SNB.

Turning to Lord Sumption’s judgment, Mr Chapman stressed paragraphs 83 and 86, where Lord Sumption said:

Mr Chapman pointed to Lord Sumption’s reference to the “personal and proprietary rights of the beneficiary exist[ing] independently”, “neither [being] dependent on the continued existence of the other”, and to the alleged constructive trustee’s amenability to personal remedies in El Ajou v Dollar Land Holdings plc being “unaffected by any issue as to existence of rights in rem”. However, (a) it can be seen from Lord Sumption’s reference to “the beneficiary’s personal rights, if any, against the trustee or his amenability to personal remedies” that the “personal rights” on which he was focusing in paragraph 83 were those against a trustee, (b) the reference in paragraph 86 to the amenability of the “alleged constructive trustee” to personal remedies must be read in the context of the justification which Millett J himself gave in El Ajou for considering the “temporary repose” of assets in civil law jurisdictions to be immaterial, (c) Lord Sumption agreed with Lord Mance’s judgment, endorsed Lightning in paragraph 86 and, in paragraph 87, referred to the availability to Samba of “the usual equitable defences” and (d) Lord Sumption was anyway only one of the five members of the Court.

Courtwood Holdings SA v Woodley Properties Ltd [2018] EWHC 2163 (Ch) (“Courtwood”) points in the same direction as Akers v Samba (SC). In paragraph 201 of that judgment, Nugee J quoted from a judgment he had given earlier, on the first day of the trial. In paragraph 59 of that latter judgment, he had said:

Finally, it is relevant to refer to views which have been expressed about the availability of knowing receipt claims in the context of registered land. As already mentioned, section 29 of the Land Registration Act 2002 has the consequence that a transferee for valuable consideration of registered land will take free of prior beneficial interests even if he had notice that the transfer was in breach of trust unless the beneficiaries were in actual occupation at the time of the disposition. Moreover, section 26(1) provides for “a person’s right to exercise owner’s powers in relation to a registered estate or charge … to be taken to be free from any limitation affecting the validity of a disposition”. However, section 26(3) states that the section “has effect only for the purpose of preventing the title of a disponee being questioned (and so does not affect the lawfulness of a disposition)”.

The explanatory notes for the 2002 Act said this about section 26:

As Mr Chapman pointed out, the explanatory notes contemplated that a transferee with an unimpeachable title might nonetheless be vulnerable to a knowing receipt claim. However, Matthew Conaglen and Amy Goymour, in another chapter in “Constructive and Resulting Trusts” (ed. Mitchell, 2010), “Knowing Receipt and Registered Land”, concluded at page 181 that “in situations where trust property is registered land and the trustee transfers title to that land in breach of trust, if the disposition was made for valuable consideration so that the transferee can claim the benefit of section 29 of the Land Registration Act 2002 to avoid the beneficiaries’ pre-existing equitable interests in the land, the transferee ought also to be immune from a personal claim for knowing receipt”. At page 177, they said:

In comments on Arthur, Martin Dixon and Nicholas Hopkins both differed from Conaglen and Goymour. Dixon said in a note ([2012] Conv 439, at 445):

Further, the Law Commission has maintained the view it expressed in its 2001 report. In its 2018 report, “Updating the Land Registration Act 2002” (Law Com No 380), with Professor Hopkins as one of its Commissioners, the Commission said this:

The case cited in paragraph 5.150 of the report as authority for the proposition that “the Privy Council has expressed the opposite view” is Arthur, in which, in our view, the Privy Council did not in fact do so. However, the Commission also relied in support of its view on Haque v Raja [2016] EWHC 1950 (Ch), where, in the course of a judgment in which he concluded that an interim proprietary freezing order should not be continued, Henderson J said:

Henderson J thus accepted that the availability of a knowing receipt claim was “not dependent upon the survival of the claimant’s beneficial interest” and so was “unaffected by the technicalities of overreaching and land registration”. However, the point was conceded and does not appear to have been the subject of argument. As the Judge put it in paragraph 101 of the Judgment, counsel for Mr Khan “had better arguments to pursue on an interim hearing”.

While it may be legitimate to refer to knowing receipt as a species of equitable wrongdoing, it is not based exclusively on fault. For liability to arise, the defendant must also have received trust property or, as Hoffmann LJ put it in El Ajou v Dollar Land Holdings plc, “assets which are traceable as representing the assets of the plaintiff”.

As Mitchell and Watterson pointed out, receipt of trust property cannot be required just to establish that the defendant has been enriched. Were that the sole concern, someone who had benefited from trust property in some other way (say, because the defendant’s spouse had wrongfully spent trust money on a holiday for them both) could be expected to be vulnerable to a knowing receipt claim, but that is not the case. In fact, even a person who has received trust property and himself used it to fund a holiday will not be liable for knowing receipt if he had no knowledge of any breach of duty until after the money had been expended. Unconscionability must coincide with possession of trust property for liability to arise.

Of course, there might be said to be an ambiguity in the proposition that, for a knowing receipt claim to succeed, the defendant must have received trust property. It might mean either that the property in question has to have been subject to a trust when in the defendant’s hands or that it suffices that the property was so subject up to the point it was transferred to the defendant. However, it is hard to see why such significance should be attached to the requirement for receipt of trust property if it were enough that the property should have been subject to a trust before it reached the defendant. If all that mattered was that the claimant should once have had an interest in the property, why should a defendant be liable for knowing receipt if he still had it when he learned that it had been transferred to him in breach of trust but escape liability entirely, regardless of any issue as to unconscionability, if he had already exhausted it through personal expenditure? Why should it be crucial that the defendant had possession of property that had been subject to the relevant trust if it never had to be so subject when held by the defendant?

Take a case in which there has been overreaching. Property was formerly comprised in a trust, but on being sold to the defendant it was freed from the trust, which instead attached to the proceeds of sale. Mr Chapman argued that the defendant could potentially be liable for knowing receipt. However, it is not obvious why the fact that the property transferred to the defendant happened to be trust property in the past should expose him to the possibility of a knowing receipt claim when (a) the overreaching principle is that the property in the hands of the recipient is freed from any trust encumbrance and that substitute property has become subject to the trust, (b) the defendant could not have been liable for knowing receipt if he had benefited from the property without receiving it and (c) the price paid by the defendant might have been a full one.

A transferee will also take free of equitable interests if he is a bona fide purchaser for value without notice. Mr Chapman accepted that a knowing receipt claim is not available against such a person, but, echoing Dixon, he attributed that to the transferee not being “knowing” rather than to the absence of a proprietary basis for such a claim. The cases, however, show otherwise. In Macmillan Inc v Bishopsgate Investment Trust plc (No 3), bona fide purchase for value without notice was considered to give the defendants a good defence to any claim by Macmillan, including one for knowing receipt, not because it would mean that their retention of the shares was not unconscionable, but because the banks would have taken free of Macmillan’s interest. In Millett J’s view, a “continuing proprietary interest” was crucial.

That a “continuing proprietary interest” is a prerequisite of a knowing receipt claim is also indicated by Lightning, Akers v Samba (SC) and Courtwood. As we have said, in Lightning the Court of Appeal took it that the English Court could not grant relief against a transferee if under the lex situs the claimant’s equity was extinguished by the transfer. In Akers v Samba (SC) Lord Mance, giving the leading judgment, saw Macmillan Inc v Bishopsgate Investment Trust plc (No 3) as establishing that, “where under the lex situs of the relevant trust property the effect of a transfer of the property by the trustee to a third party is to override any equitable interest which would otherwise subsist, that effect should be recognised as giving the transferee a defence to any claim by the beneficiary” and Lightning as recognising that, as a result, the lex situs can “have a significance in the case of a third party transfer”. In Courtwood, Nugee J considered the foundation of a knowing receipt claim to be that “the assets do not belong in equity to the recipient” and that “what gives the equity to the claimants” is “the fact that the transaction which is impugned is not one which transfers a good title to the recipient”.

The cases in which the Courts have accepted that knowing recipients have “custodial” obligations, including in particular an obligation to restore the property, also tend to suggest that, to succeed in a knowing receipt claim, a claimant must have had a proprietary interest in the property when it was in the hands of the defendant. It makes sense to think of a knowing recipient owing such duties in circumstances where the property is subject to an interest having priority to the recipient’s. It is much more difficult to see why a recipient should be bound to “restore” property or otherwise to have “custodial” responsibilities in respect of it if he has an unimpeachable title to it.

This accords with the mental element required to establish liability in knowing receipt, which, as the passage quoted above from Akindele makes clear, is that the state of knowledge of the recipient makes it unconscionable for him to retain the property. If the law treats the receipt of the property as conferring unencumbered title on the recipient, it is difficult to see why retention should be regarded as unconscionable. It is not the receipt which must be unconscionable; indeed the constructive trust duty may arise where the recipient has done nothing which could be described as unconscionable, as for example where he receives property in ignorance of any trust and is only later informed of it whilst still in possession: at that very moment of after-acquired knowledge, when possession and knowledge coincide, the duty as a constructive trustee immediately arises notwithstanding that the recipient has at that stage done nothing unconscionable. That is so if and because subsequent retention by him (or disposal) would be unconscionable. This illustrates the fallacy in Mr Chapman’s submission that the equitable wrong which gives rise to liability for knowing receipt is made out merely by receipt of property in breach of trust coupled with the requisite knowledge that the transfer is such a breach: the submission wrongly treats the knowledge element as unconscionability of receipt, whereas it is unconscionability of retention.

Turning to the scope, if any, for knowing receipt claims to be brought in the context of registered land, Arthur illustrates the importance of the particular terms of the relevant statutory regime. In England, that regime is to be found in the Land Registration Act 2002 and the arguments advanced by, on the one hand, Conaglen and Goymour and, on the other, the Law Commission in part turn on sections 26 and 29 of that statute, on which we do not need to comment. What we would say is that the views expressed in the last sentence of paragraph 5.150 of the Law Commission’s 2018 report do not have a sound basis in authority.

In all the circumstances, it seems to us that the Judge was right to conclude that a knowing recipient “must have held trust property, not property to which from the moment of receipt he had good title” and that “a claim in knowing receipt, where dishonest assistance is not alleged, will fail if, at the moment of receipt, the beneficiary’s equitable proprietary interest is destroyed or overridden so that the recipient holds the property as beneficial owner of it”. That conclusion is, as the Judge said, borne out by “a consistent line of case law” in which it has either been decided that “a claim in knowing receipt cannot succeed unless the claimant has a continuing proprietary interest following the impugned transfer” or that has been assumed to be correct.

In short, a continuing proprietary interest in the relevant property is required for a knowing receipt claim to be possible. A defendant cannot be liable for knowing receipt if he took the property free of any interest of the claimant. It follows that, as the Judge held, “absent a continuing proprietary interest in the Disputed Securities at the time of registration, the claim in knowing receipt as pleaded will fail”.

The Disputed Securities were shares in five Saudi Arabian joint stock companies, all of them large banks, whose transfer to Samba was effected by registration. The shares in four of them were publicly listed on the Tadawul, the Saudi stock exchange. Shares in the fifth, National Commercial Bank (“NCB”), were not so listed at the time of the September Transfer, although they were by the time of the Judgment. As unlisted securities, the NCB shares were governed by different registration arrangements. In the light of his conclusion on the Knowing Receipt Issue, the question for the Judge was whether as a matter of Saudi Arabian law, as the lex situs of the shares, the transfer to Samba left SICL with an interest in the shares which the English Court would characterise as a subsisting proprietary interest.

The Judge found that although the Saudi Arabian Courts would recognise a form of ownership interest in SICL prior to the transfer, the registration of the shares in Samba’s name was conclusive of full and exclusive ownership rights thereupon vesting in Samba. On the appeal, the claimants challenge the second part of this finding, contending that the Judge was wrong to find that the registration left SICL without the form of ownership interest enjoyed prior to the transfer. By a respondent’s notice, SNB challenges the first part of this finding, contending that the Judge was wrong to find that prior to the transfer SICL had any form of ownership interest which Saudi Arabian law would recognise. Although the respondent’s notice point is logically prior to the point in the appeal, the parties addressed their arguments first and primarily to the argument in the appeal on the effect of registration, recognising that on the Judge’s findings on the Knowing Receipt Issue, the claimants must succeed on this issue for the appeal as a whole to succeed. We shall adopt the same approach.

The Judge explained the reasons for his conclusions in a careful and detailed section of the Judgment running to 88 paragraphs. The following summary is sufficient for the purposes of this appeal.

The Courts of the Kingdom of Saudi Arabia do not apply foreign law. They would seek to give effect to the Six Transactions in accordance with the Saudi Arabian Courts’ view of Saudi Arabian law. There is no distinction in Saudi Arabian law between legal and beneficial ownership as such. Therefore the beneficial interest of SICL under English/Cayman law would not have been recognised as such by the Saudi Arabian Courts.

The relevant legislation in Saudi Arabia governing the ownership of shares in joint stock companies at the time of the September Transfer was article 102 of the Companies Regulation 1965 (“CR”) which provided that shares are transferred by registration in the shareholders’ register maintained by the company (other than bearer shares with which we are not concerned). Article 102 of CR applied the Capital Market Regulation 2003 (“CMR”) to the transfer of shares listed on the Tadawul. Chapter four of CMR provides for the establishment of the Securities Depository Centre (“the SDC”), which by article 26 is the only entity in the Kingdom of Saudi Arabia authorised to operate transfers and registration of ownership of securities traded on the Tadawul. Article 25 of CMR requires the Capital Market Authority (“CMA”) to establish a committee, known as the Committee for the Resolution of Securities Disputes (“CRSD”) which has jurisdiction over disputes falling under CMR and the other rules, regulations and instructions issued by CMA and the Tadawul with respect to public and private actions. Article 21 of CMR provides that, subject to authorised exceptions, listed securities can only be traded in the SDC. Article 27 provides that all transfers of listed securities are to be registered in the records of the SDC, and that such registration serves as “conclusive evidence and proof of ownership” against all third parties, subject only to the provisions of article 27(d). Article 27(d) permits an application to be made to the SDC to correct or amend the register if there is “an error in the information entered into the registry”. The ability of SICL to challenge Samba’s exclusive ownership of the listed shares in four of the five banks depended entirely on whether it could obtain rectification of the register pursuant to article 27(d) of CMR.

At paragraph 173, the Judge divided the relevant questions into three:

The Judge said that as the legislation governing share ownership is part of the law of Saudi Arabia and is to be interpreted consistently with Islamic law, it might be artificial to separate out issues (ii) and (iii), but that doing so was nevertheless a convenient way of analysing the difference between the experts.

On the characterisation issue, the Judge preferred the evidence of the claimants’ expert, Professor Mallat, that a Saudi Arabian judge would regard the Six Transactions as more than just a contract between SICL and Mr Al-Sanea, and instead as making SICL the underlying owner of the trust property. Saudi Arabian law has embraced the concepts of true underlying ownership, and registration in the name of a front or nominee, in the context of capital markets and banking legislation, and takes account of the fact that these same questions arise in the context of trusts and similar ‘offshore’ legal structures. A Saudi Arabian judge would be able to use a variety of imprecise analogues in Islamic law to understand the intended effect of the Six Transactions, namely that SICL would be the real owner and Mr Al-Sanea the front with fiduciary obligations. A Saudi Arabian judge would not be likely to treat SICL’s interest as being only a matter of contract between it and Mr Al-Sanea: he would characterise SICL's interest under lawful Cayman Islands trusts as an ownership interest, not merely a contractual right.

As regard the second issue, remedy against third party, the Judge addressed it as a matter of Islamic law ignoring for the time being the statutory provisions for registration of shares. As such it was a broad question of Islamic law about the extent to which an owner’s rights can be vindicated where he has suffered harm, not against the person who directly caused the harm but against someone who received the property with knowledge of the lack of authority. Basic norms of Islamic law suggest that there should be a valid claim, and a Saudi Arabian Court would probably provide a remedy to recover the property against the party with knowledge.

In relation to the third issue, the Judge rejected the evidence of SNB’s expert, Mr Haberbeck, in his written report, that only a ‘clerical error’ would be corrected under article 27(d) of CMR; but accepted his evidence that article 27(d) could only be used to correct something that had gone wrong within the SDC system itself; it could never be used to give effect to a transaction that had taken place outside the system, since the whole purpose of the SDC register was to ensure that only authorised and registered transactions take place affecting title to shares, in order to maintain the reliability and transparency of the register. To admit claims based on extraneous dealings would undermine the whole basis of the authorised registration system, reflected in its purposes set out in article 5 of CMR.

Mr Haberbeck supported his view that article 27(d) could only be used to correct something that had gone wrong within the SDC system by reference to eight cases where the CRSD had declined to accept jurisdiction over disputes about shares unless the disputed transactions had been conducted in accordance with the applicable rules of CMR and delegated legislation issued thereunder. The Judge recognised that a party seeking to challenge ownership could in principle, as a matter of jurisdiction, bring a claim before a Saudi Arabian Court, rather than the CRSD; but held that there was no persuasive evidence that any Court in Saudi Arabia would have allowed the purposes of the Tadawul to be compromised by directing Samba to transfer the Disputed Securities to SICL's nominees or otherwise by recognising SICL's ownership; and that it was probable that no Court or committee other than the CRSD would accept a claim to assert SICL’s ownership, since, as was common ground, a claim to upset Samba’s title had to be brought under article 27(d).

As to the NCB shares, under article 102 of the CR registration is prima facie evidence of ownership, and until displaced is conclusive as to ownership. There is nothing to prevent a claim for rectification of the share register if the claimant has a sufficient basis for it; and before the September Transfer, as between SICL and Al-Sanea, the CR would have permitted SICL to have Mr Al-Sanea replaced as registered owner by another nominee. However that could not occur after transfer to a third party. Registration of title was conclusive in favour of the third party purchaser, just as it was for listed shares under CMR.

The Judge also considered whether under Saudi Arabian law there would be a personal claim for compensation against Samba after the September Transfer (i.e. a claim mirroring in its essentials a knowing receipt claim under English law), describing this as uncharted territory. He said that he was unable to decide that it was probable that a case against Samba for compensation would succeed; but in any event such claim would be premised on the loss of SICL’s property, not an assertion of continuing proprietary rights.

The Judge therefore concluded that SICL had no continuing proprietary interest in the Disputed Securities after the September Transfer capable of supporting a claim against Samba in knowing receipt.

The Judge’s findings of the effect of Saudi Arabian law were explained by reference to the evidence which he heard. The reports of Professor Mallat and Mr Haberbeck ran to some 300 pages, with some 2,500 pages of exhibits, including numerous cases and other Saudi Arabian materials. The experts were cross-examined for two court days, during which the Judge also asked questions of them from time to time. The Judge explained his assessment of Professor Mallat and Mr Haberbeck as witnesses, and his approach to their evidence, as follows. The Judge said that having heard both experts being cross-examined at length, and seen the way in which they answered questions, and explained some of the cases referred to, had given him significant help in assessing the credibility of their opinions. There was also, as regards CMR and associated legislation, an inherent likelihood of certain evidence about it being correct because the purpose of the legislation was not in dispute: it was designed to support an efficient, transparent, secure and reliable stock exchange on a modern western model, and to facilitate competition with other such markets, to which objectives uncertainty or insecurity would be inimical.

Professor Mallat was immensely knowledgeable about Islamic law and fundamental tenets of English, French and Roman law; he gave evidence about Islamic law with some authority. However he did not have relevant knowledge of litigation in Saudi Arabia and was not well-placed to speak with real authority on the application and effect of CR and CMR and related legislation, or on practice in the CRSD or the Commercial Court or relevant committees in Saudi Arabia. Further, the Saudi Arabian case law on which he relied was not always to the point or fully analysed. The Judge felt uneasy about relying on Professor Mallat's opinions on the approach that would likely be taken by the CRSD in particular, and Saudi Arabian Courts more generally, to claims adverse to registered shareholders and to the operation of the SDC, CR, CMR and associated statutes, because he had no relevant experience of those matters.

Mr Haberbeck was not an Arab and did not speak or read Arabic fluently. He was not qualified as a lawyer in Saudi Arabia or any Middle Eastern country and his Arabic was not good enough to act as an advocate or to understand everything that was said in hearings, although he was able to follow them. On the other hand he had been a practitioner in Saudi Arabia for 32 years, with a practice largely involving commercial cases on behalf of international clients. He had considerable practical experience and knowledge of Saudi Arabian law as practised in the Courts of that country including the Board of Grievances, the Commercial Court, the Committee for Resolution of Banking Disputes and the CRSD. He himself drafted pleadings in cases before these tribunals, except the CRSD, and was currently involved in 25 cases in the CRSD. He had experience in the attitudes and approach of judges in those tribunals and experience of cases in Saudi Arabia raising issues of ownership and share registration. He had a detailed and complete understanding of the cases which were relied upon by both expert witnesses in support of their opinions.

In respect of both Professor Mallat and Mr Haberbeck, there were inconsistencies between their evidence at trial and what each had said in reports submitted for the purposes of the s. 127 jurisdictional dispute. Those inconsistencies led the Judge to exercise caution about what parts of their evidence he felt able to accept. Subject to that caution, he would give more weight to the evidence of the expert who was likely to have greater experience and expertise in relation to the particular matter or question. In general on the basic principles of Islamic law and the way in which they would be understood and applied by a Judge in a particular context, Professor Mallat, as a scholar and Arabic speaker with wider experience of Islamic law in practice across the Middle East, was likely to be more reliable in his assessment; whereas Mr Haberbeck was much better placed to provide an informed and reliable opinion on how in practice the CRSD, the Commercial Court and various judicial committees in Saudi Arabia would apply the capital markets and companies legislation.

On the appeal, SNB submitted that this Court should be very slow to interfere with these findings of fact by the trial judge, based on the expert evidence which he had heard, whilst arguing that this did not apply to its respondent’s notice point. Conversely, the claimants contended that such an approach was applicable to prevent the Court from interfering with the Judge’s findings on the respondent’s notice point, but inapplicable to its arguments on the appeal. We were referred to a large number of authorities, but the principles are not really in doubt.

The general approach to the review of a trial judge’s findings of fact on an appeal to this Court, and the reasons for it, were set out in the well-known judgment of Lewison LJ in FAGE UK Ltd v Chobani UK Ltd [2014] EWCA Civ 5 at paragraph 114:

In Henderson v Foxworth Investments Ltd [2014] UKSC 41, [2014] 1 WLR 2600, Lord Reed said at paragraph 67:

Similar caution applies on an appeal against a trial judge’s assessment or evaluation of expert evidence: Wheeldon Bros Waste Ltd v Millenium Insurance Co Ltd [2018] EWCA Civ 2403, [2019] 4 WLR 56 at paragraph 11.

In Walter Lilly & Co Ltd v Clin [2021] EWCA Civ 136, [2021] 1 WLR 2753, Carr LJ referred to these and subsequent authorities and summarised the approach at paragraph 85:

Foreign law is a question of fact which the trial judge is required to determine on the basis of the evidence deployed by the parties. The task for the judge is to determine what the highest available Court in the foreign jurisdiction would decide if the point had come before it: Dexia Crediop SpA v Comune di Prato [2017] EWCA Civ 428 at paragraph 34. The FAGE approach remains the starting point in any appeal from such a determination of foreign law (ibid paragraph 36). However it may be qualified, because a question of foreign law has been described as “a question of fact of a peculiar kind”: Parkasho v Singh [1968] P 233, 250. In Macmillan Inc v Bishopsgate Investment Trust (No 4) [1999] CLC 417, Evans LJ giving the judgment of this Court said:

Mr Chapman relied on this qualification to submit that this Court is as well placed as the Judge to decide what article 27(d) of CMR meant, because there was no issue between the experts as to the Saudi Arabian principles of construction. We cannot accept this submission. Where the foreign law is in the form of a provision in a code, statute or other written source, the task of the Court remains one of determining how the foreign Courts would interpret and apply it, based on the evidence of the expert witnesses. Generally speaking the Court’s task is not to address how it would itself interpret and apply the provision; the wording of the provision is to be considered only as part of the evidence and as a help to decide between conflicting expert testimony: see A/S Tallinna Laevauhisus v Estonian State Steamship Line (1946) 80 Ll L Rep 99 at 107; Lazard Brothers & Co v Midland Bank Ltd [1933] AC 289 at 298; and Dallah Estate and Tourism Holding Company v Ministry of Religious Affairs of the Government of Pakistan [2009] EWCA Civ 755, [2010] 2 WLR 805 at paragraph 69. There is a qualification to this general principle, recognised in Macmillan Inc v Bishopsgate Investment Trust (No 4), and applicable to a first instance Court as well as on appeal, where the nature of the foreign law issue means that the English Court’s expertise approaches that of any foreign law expert, for example where the foreign law is written in the English language and involves concepts similar to English law and familiar to English judges; or where the foreign Courts would be influenced by the English Courts’ decisions on the issue: see King v Brandywine Reinsurance Co (UK) Ltd [2005] EWCA Civ 235, [2005] 2 All ER (Comm) 1 at paragraph 68.

However, there is no scope for applying any such qualification to the general principle in this case. The only authorised texts of CMR and CR are in Arabic and the trial judge had to work from an agreed translation, albeit that in the case of CMR it was one published by the CMA. Saudi Arabian law is an Islamic system of law, whose concepts and principles are far removed from the common law familiar to English judges. The interpretation and application of CR and CMR fell to be determined not only under this unfamiliar system of law, but also against the background of practice and culture in the capital markets in Saudi Arabia, with which again the English Courts have no inherent familiarity. There were, moreover, numerous Saudi Arabian Court decisions prayed in aid by both experts in support of their contentions as to how the provisions would be interpreted and applied to the facts of the present case. The Judge needed the assistance of extensive expert evidence to explain and explore these cases in order to determine the foreign law issues before him, and referred to a number of them in the Judgment as lending support, or not, to the views of the witness who sought to rely on them. For example his critical finding that article 27(d) of CMR could not apply to “errors” arising outside the CRSD system was based in part on his acceptance that Mr Haberbeck’s view was supported by the eight cases upon which he relied in this respect. It would have been both wrong in principle, and impossible in practice, for the Judge to approach the task as one of interpreting the provisions for himself as a matter of construction. It would be equally wrong for this Court to seek to do so. This is a case in which the FAGE approach applies: this Court should be slow to interfere with the Judge’s findings of fact on Saudi Arabian law and should only do so in accordance with the principles applicable generally to findings of fact made by a trial judge who has based his findings on evidence from witnesses.

Mr Chapman first submitted that because it was common ground that Saudi Arabian law did not recognise the concept of an equitable interest, the effect of a transfer of shares by registration was logically incapable of extinguishing it: Saudi Arabian law could not extinguish that which it did not recognise in the first place. This is to confuse the respective roles of English and Saudi Arabian law in the issue at stake. The Saudi Arabian law question required the Court to determine what ownership rights SICL had in the shares, if any, as a result of the transfer to Samba by registration. That is the role of the lex situs governing property rights. The English law question is then whether such ownership rights as SICL had after the transfer were properly to be characterised as a sufficient proprietary interest to support the personal cause of action in knowing receipt. The question was not, as is inherent in Mr Chapman’s submission, whether as a matter of Saudi Arabian law the transfer extinguished an English law concept of an equitable interest. The role of the lex situs is to identify what rights exist, not how they would be characterised under English law, as Millett J explained in Macmillan v Bishopsgate Investment Trust plc (No 3) at 1001B-C:

Mr Chapman next submitted that in assessing what rights SICL had following, and as a result of, transfer by registration of the shares, the Judge fell into error in treating the issue as determined by what remedies SICL might have against Samba in respect of the shares. The submission was, essentially, that the focus should have been on the rights, not the availability of remedies. In our view this is an unfair criticism of the Judge’s reasoning and conclusions. His essential reasoning was that the registration of the shares in Samba’s name was conclusive of Samba’s ownership unless and until the register was rectified; that SICL could not have the register rectified; and that if it could not have the register rectified, it had no ownership rights. That is not to conflate rights and remedies: it is to decide that the absence of any available remedy for SICL to rectify the register left the register as conclusive of ownership in Samba’s favour. The Judge found in terms that after the transfer SICL had no property interest in the shares as a matter of Saudi Arabian law.

Mr Chapman next submitted that there was an inconsistency between the Judge’s finding on the issues identified in paragraph 173(i) and (ii) on the one hand, and paragraph 173(iii) on the other: having found that SICL had a form of ownership interest recognised by Saudi Arabian law prior to the transfer, and that Islamic law principles would in principle provide a remedy to SICL against Samba after the transfer, it necessarily followed that Saudi Arabian law must recognise a property interest in the shares vested in SICL after the transfer. We can see no such inconsistency in the Judge’s findings. He recognised that issues (ii) and (iii) were really aspects of a single question but addressed them separately as a convenient way of analysing the expert views. Issue (ii) addressed the position in respect of property generally and as a matter of principle, and was always subject to the more particular examination in relation to shares, and the Saudi Arabian provisions affecting transfer of ownership in shares, which the Judge addressed extensively under the heading of issue (iii). There was nothing inconsistent in the Judge’s process of reasoning and conclusion that, in respect of property generally and in principle, SICL might have a remedy against Samba after transfer, but that there could be no such ownership rights in respect of registered shareholdings.

Mr Chapman further submitted that the Judge’s conclusion on article 27(d) was inconsistent with the plain language of the provision: once it was accepted that “error” was not confined to clerical errors, there was no warrant for reading into the provision words which were not there, namely that the error had to be within the SDC system. This is to treat the exercise for the Judge, and for this Court, as one of freestanding construction applying English law principles, which is the wrong approach. The Judge was bound to ask how the provision would be interpreted and applied by Saudi Arabian Courts; he did so and reached his conclusions by reliance on the testimony of an expert witness experienced in the practice of those Courts who supported his view with case law and recourse to the statutory purposes of the system.

Mr Chapman further criticised the Judge’s reasoning in relation to the interpretation of article 102 of CR, which he held would not permit rectification any more than would article 27 of CMR in relation to listed shares. However, the Judge’s conclusion on the evidence was that registration of unlisted shares was conclusive as to ownership absent rectification, just as with listed shares. He was entitled to accept Mr Haberbeck’s view that rectification would not have been available against Samba after transfer, and to reject Professor Mallat’s contrary view, which as the Judge observed was supported by nothing other than assertion that the position was the same as for listed securities under CMR. This could provide no support once the Judge had rejected Professor Mallat’s view that that was the position under CMR for listed securities, for the reasons he gave.

Mr Chapman made a number of more granular criticisms of certain paragraphs in the Judgment. For example he sought to argue that the Judge had failed to give effect to an alleged concession by Mr Haberbeck; and had unfairly treated what Professor Mallat had said as a concession. We were referred to a few isolated passages in the transcript of their evidence. We are not persuaded that the passages we were shown support the criticisms. But in any event, we were shown only a fraction of the evidence adduced at trial, and did not have the advantage which the trial judge enjoyed of hearing the witnesses give evidence and explain the material. Even a transcript does not fully convey the nuance, tone and full meaning of what can be conveyed orally in the witness box. The claimants’ reliance on snippets of transcript in this way was impermissible “island hopping” in the memorable phrase of Lewison LJ.

In summary, Mr Chapman’s points, even taken cumulatively, do not come close to satisfying the criteria for this Court to interfere with a judge’s findings of fact on foreign law in a case of this kind. The conclusions of the Judge in this case were reasonably open to him on the evidence he heard, and there is nothing in his clear and detailed reasoning which suggests he was wrong in his conclusions.

In the light of our conclusion that the Judge cannot properly be criticised for the findings which he determined adversely to the claimants, the respondent’s notice point does not arise, and we do not propose to address it.

The final question in this appeal was termed the “Valuation Issue” and it was concerned with whether the Judge erred as a matter of principle in his conclusion that a block discount should be applied when determining the objective value of the Disputed Securities for the purposes of an account on the footing of knowing receipt, or when valuing the benefit obtained and retained by SNB. In the light of our conclusions in relation to the Knowing Receipt Issue and the Saudi Arabian Law Issue, the question of valuation does not arise. However, for the sake of completeness we consider it in outline below.

To be clear, the claimants’ claim was not for the return of the Disputed Securities but for an account of their value on either the date of the September Transfer or the date of judgment, together with income derived from them and interest.

Before the Judge and before us, SNB accepted that the claimants’ claim, therefore, is for substitutive performance and Mr Onslow, on behalf of SNB, argued that this requires it to pay the objective value of the Disputed Securities being their market value on a sale which, in the light of the size of the holdings and the expert evidence which the Judge accepted, would be reduced by a block discount. Mr Chapman, on behalf of the claimants, on the other hand, contended that SNB is required to reconstitute the trust fund as if there had been no misappropriation. As a result, he says that the objective value of the Disputed Shares is what it would cost to purchase them in the market rather than their value on a hypothetical sale. It is accepted that a block discount would not arise in relation to such a purchase.

Before turning to the Judge’s approach to the expert evidence, it is helpful to remind oneself of the nature of the Disputed Securities, the way in which they have been dealt with and the definitions which the Judge adopted in relation to the valuation exercise.

The Disputed Securities comprise large holdings of shares in five Saudi Arabian banks, including Samba itself. The holdings, as at the September Transfer, were: Arab National Bank - 889,797 shares; Banque Saudi Fransi – 4,248,146 shares; NCB – 2,129,250 shares; Samba – 7,130,044 shares; and Saudi British Bank – 3,319,346 shares. At the date of the September Transfer, all of the shares, save those in NCB, were listed and publicly quoted on the Tadawul. All of the companies are now publicly quoted on the Tadawul: see paragraph 126 of the Judgment.

As at 31 January 2020, being the valuation date taken by the experts as the date of judgment for the purposes of exposition of the block discount issue, the number of shares in each holding and the percentage of the relevant bank’s issued equity represented by the holding were as follows: Arab National Bank - 2,053,375 shares being 0.14% of the issued equity; Banque Saudi Fransi – 7,080,242 shares being 0.59% of the issued equity; NCB – 4,258,500 shares being 0.14% of the issued equity; Samba – 15,844,541 shares being 0.79% of the issued equity; and Saudi British Bank – 6,638,691 shares being 0.32% of the issued equity: see paragraph 208 of the Judgment.

On Mr Al-Sanea’s transfer of the Disputed Securities to Samba, it credited Mr Al-Sanea’s account with the market value of the Disputed Securities on the date of transfer. That amounted to some 801 million Saudi riyals (paragraph 2 of the Judgment). SNB still holds the Disputed Securities. Mr Chapman informed us that as at 9 December 2021 the Disputed Securities were valued at US$407.7 million and that that figure would be reduced by some US$21.5 million if a block discount were applied.

As mentioned in paragraph 2 above, Samba’s assets and liabilities have now been transferred to SNB. It was not suggested that this affected the claim for substitutive performance and we shall assume that nothing material turns on it.

The Judge noted at paragraph 4(iii) of the Judgment that the only dispute in relation to the Valuation Issue was whether a “block discount” should be applied to the quoted prices for the Disputed Securities on the Saudi Arabian stock exchange on either the date of the September Transfer or the date of judgment. In answering that question he had to determine whether, in fact, the objective value of the Disputed Securities in the circumstances was their market value, their investment value or liquidation value. He rejected both an investment value and a liquidation value. There is no appeal in relation to the liquidation value and, although one of the claimants’ grounds of appeal touched upon the appropriateness of investment value, it was not pursued in oral argument. The Judge’s reasoning in relation to each basis of valuation feeds into his conclusions about the application or otherwise of a block discount and, therefore, it is important to understand it in full.

As the Judge pointed out at paragraph 209 of the Judgment, “market value” is defined in International Valuation Standard (“IVS”) 104 as:

The Judge noted at paragraph 210 of the Judgment that “[w]here small numbers of shares are to be sold on an active public exchange, the price per share can readily be taken to be the quoted reference price on the day in question” but that “where a large block of shares is to be sold on a single day, the reference price is unlikely to be obtainable, at least where the size of the block far exceeds the average daily traded volume (‘ADTV’) for that stock, as is the case for each of the holdings of the Disputed Securities”.

He went on to state that “[t]he market has insufficient liquidity to absorb the sale of the holding without a bespoke transaction or transactions at a negotiated price. If the entirety of a large holding is to be sold on a particular day, a discount may well be required to achieve the sale.” Furthermore, “[i]f the holding is to be sold in smaller parcels of shares over a longer period, there may still be an impact on price from the cumulative volume being sold, depending on the liquidity in the market and the impact (if any) on sentiment from large volumes of sales, and the seller is further exposed to the risk (but not the certainty) of a fall in the share price from other causes before the shares are fully sold” (paragraph 211 of the Judgment). The Judge accordingly concluded, also at paragraph 211, that “[t]he amount of the block … discount is therefore likely to reflect the alternatives that are reasonably available to the seller. The seller will not willingly accept a lower price than they can realise by other means.”

The first question the Judge posed was whether as a matter of law, when seeking to identify the objective value of an asset in connection with a trustee’s liability for substitutive performance, a measure of value based upon what could be realised for the asset in a transaction was inappropriate. He stated that, if that were the case, the proper measure of value would have to be the holding or investment value and that investment value would be a value specific to the trust which may have (or have had) as its objective the holding of the asset, with no intention of selling it: paragraph 224 of the Judgment.

He reasoned that he was unable to conclude that the measure of liability for substitutive performance must always be the investment value of the misappropriated asset because, amongst other things, “[t]here are likely to be many cases in which market value at the valuation date is higher than the long-term holding value of the asset” and “[m]any trust assets may be held otherwise than to achieve investment or operational objectives”: paragraph 225 of the Judgment. He continued as follows:

The Judge then stated that he did not accept that market value was automatically the appropriate basis to adopt in order to determine the objective value of the Disputed Securities and that the purpose of the valuation must be borne in mind: paragraph 229 of the Judgment. He determined at paragraph 230:

The Judge concluded at paragraph 236 of the Judgment:

The Judge had already set out his impression of the expert witnesses in relation to valuation at paragraphs 213-216 of the Judgment. He stated that his impression was that Mr Worsnip on behalf of the claimants “was taking rather extreme positions in order to support the argument that an orderly liquidation basis of value was appropriate, and in particular that no adjustment on such a valuation was required for the risk of movement in the market over time adverse to the seller; and further, that even if market value was the right basis of value to take, there could in principle be no block discount on such a valuation either”: paragraph 215 of the Judgment. On the other hand, he found Mr Steadman on behalf of SNB to be an impressive witness and his explanations for the use of the market value basis and the type of put option model to be very persuasive. However, he was left with the impression that not enough objectivity and scepticism had been applied to the results of various models and that too much machinery and not enough objective judgment went into Mr Steadman’s percentage discounts: paragraph 216.

The Judge went on to address the expert valuation evidence in more detail. In summary, his approach was as follows:

He rejected Mr Worsnip’s opinion that there would be no block discount because the tendency of a sale of a large block of shares to depress the market price would be matched by the upwards pressure on price that a willing buyer of a large block will have because there is no requirement to assume, contrary to the known facts, that on the valuation date there is a buyer who is willing to buy a large block of shares at the reference price or above: paragraph 237 of the Judgment;

Having set out the opposing views about the effect of “dribbling out” the shares in each of the holdings in small quantities each day at paragraphs 237-241, the Judge stated that Mr Worsnip for the claimants was wrong to assume that dribbling out shares over a protracted period at a rate of 20-25% of ADTV could not itself impact on price and that Mr Worsnip had not given effect to the liquidation basis because he had assumed that there was no allowance required for the risk of the share value falling after the valuation date before a sale could be concluded: paragraphs 243 and 244. Instead, the Judge accepted Mr Steadman’s evidence that “with listed securities there is always a risk of a fall in price and risk carries with it a cost, which impacts on value”: paragraph 242;

He concluded that in carrying out his fall back valuation, Mr Worsnip had not, in fact, given effect to the liquidation basis of value he purported to apply because he assumed that there is no allowance required for the risk of the share value falling after the valuation date before a sale could be concluded and because he assumed that the dribbling out of shares would have no impact on the share reference price: paragraph 244; and

Instead, he accepted the approach to determining market value adopted by Mr Steadman on behalf of SNB which was to calculate the appropriate discount from the valuation date reference price to reflect the cost of the other alternatives available to the seller having accepted that the only viable alternative was the “dribble out” option and the model suggested by Joseph Estabrook in chapter 7 of Reilly & Schweihs, Handbook of Advanced Business Valuation (2000): paragraphs 245-248, 249 and 250.

The Judge’s reasoning at paragraphs 249 and 250 was as follows:

The Judge concluded that, doing the best he could in the light of all the evidence, he should reduce Mr Steadman’s discounts by 30% across the board: paragraph 263 of the Judgment. On that basis, the Judge set out what he considered to be the appropriate block discount for each of the tranches of shares at paragraph 264.

We have set out some of the Judge’s approach to the expert evidence because it is interwoven with his reasoning. It is not for us to replicate the Judge’s task and evaluate the expert evidence ourselves in order to re-determine the Valuation Issue. Were this issue still to be determined, we would be concerned solely with whether the Judge had made an error of law and principle in his approach to the valuation of the Disputed Securities and to the application of a block discount, in particular.

As we have already mentioned, it is unnecessary for us to decide this issue. We would not, however, wish to be taken to have endorsed the Judge’s conclusions. It seems to us that there is a persuasive argument for saying that, where a trustee has elected to receive the value of an asset rather than its return in specie, the sum which is necessary to restore or re-constitute the trust fund will often at least be best determined by reference to the cost of the asset had it been purchased by the trustee rather than what the asset would have fetched on a sale. That might be said to be the measure most likely to put the trust fund back into the position it would have been in if the misappropriated asset had still been held for the benefit of the beneficiaries and, in the present case, to represent the full monetary equivalent of the trust property. It was accepted that a block discount would not apply to the market value ascertained by reference to the purchase price.

We agree with the Judge that one must adopt a basis of valuation which reflects the nature of the property in question and all of the relevant circumstances. Even so, it can be cogently contended that the Judge was mistaken in thinking the application of a block discount appropriate. Were the Judge’s approach correct, the amount which a person in knowing receipt of trust property in the form of shares would be required to pay by way of compensation for breach of ancillary liability would seem to be less the greater the percentage stake that the misappropriated shares represent in comparison with the company’s issued share capital.

We should also add that we are not necessarily endorsing the view that the relevant securities should be valued as if they were all purchased or, for that matter, were all sold on a single day or should have been subject to “dribble out” or, for that matter, “dribble in”.

The appeal will be dismissed.