[2025] EWCA Civ 1211

The conclusion that an absence of sufficient writing does not affect the validity of the trust that has been declared is also supported by the decision in Gardner.

In that case, Lord Mount Edgecumbe granted a lease to George Wilkinson in 1812, permitting him to dig for tin and other minerals on land called the Wheal Regent Sett in Cornwall. In 1813, Wilkinson executed a deed, declaring that he was trustee of the lease for Mr Rowe. In the intervening period, Wilkinson had committed an act of bankruptcy, the consequence of which was that his property passed to his assignees in bankruptcy. The assignees brought an action seeking a declaration that the declaration of trust had been executed fraudulently with a view to protecting Rowe from the effects of Wilkinson’s insolvency, and that the lease was part of the bankruptcy estate.

Rowe’s defence was that he was the beneficial owner of the lease, it having been purchased in the name of Wilkinson for his benefit. The court directed a trial of the issue “whether the name of Wilkinson was used in the lease of 23d of January 1812, as trustee for Rowe”. The jury determined that it was.

The question then arose whether the trust “could not prevail because there was no written declaration of trust within the Statute of Frauds other than the indenture of 24th August 1813, which being executed by the bankrupt after his bankruptcy could not operate to defeat the claim of his assignees”.

The report of the decision of the Vice-Chancellor (Sir John Leach) is at (1825) 2 Sim. & St. 346. Having observed that there was no previous authority on the point, the Vice-Chancellor identified the question (at p.353) as being: “can the bankrupt be said to have any interest in this mine at the time of his bankruptcy?”. Although Wilkinson could have recovered possession of the mine by force of his legal title, “but he would then have recovered, not in respect of his interest, but by converting a statute made for the prevention of fraud into an instrument of his own fraud”. It was not disputed that, if the deed of August 1813 had been executed before bankruptcy, it would have prevailed against the assignees. The Vice-Chancellor continued:

“This deed, therefore, in respect of the moral obligation on the trustee to give effect to his trust, would not in such case have been considered a mere voluntary deed. If, in respect of the moral obligation affecting the trustee, this declaration of trust would have prevailed against the assignees if executed the day before the bankruptcy without any other consideration, I cannot find a principle why it should not prevail against the assignees if executed the day after bankruptcy, especially when it is considered that a trust does not pass by assignment in the bankruptcy.”

While the result was clear, the reasoning – particularly as to where the beneficial interest in the lease lay at the time of Wilkinson’s bankruptcy – is equivocal. On the one hand the Vice-Chancellor referred to the “moral” obligation on Wilkinson. On the other hand, whatever language was used to describe the obligation, it prevented the declaration of trust from being a mere voluntary deed and enabled it to prevail against the assignees, on the basis that “a trust” does not pass by assignment in the bankruptcy. It also enabled the Vice-Chancellor to answer the question he posed (did Wilkinson have any interest in the mine at the time of his bankruptcy?) in the negative.

The judgment on appeal of the Lord Chancellor (Lord Lyndhurst) (1828) 5 Russ. 258, at pp.261-262 is clearer on the point:

“Assuming the bankrupt to have been a trustee for Mr Rowe, there was nothing, I think, to prevent him from making a valid declaration of trust, notwithstanding his bankruptcy. It is true that the property of a trader cannot be assigned by him after his bankruptcy: the property is no longer his; it is vested in his assignees. But property held in trust, is not the property of the bankrupt. It does not pass to his assignees. The only question therefore, as it appears to me, in this case is, whether the declaration contained in the deed was founded upon a previous trust, or was altogether fraudulent. That question, however, has been decided in substance by the jury upon the trial of the issue; for they have found that the name of Wilkinson was used in the original deed as a trustee for Rowe.”

It is true that, in Gardner, by the time the matter came before the Court, Wilkinson had signed a declaration sufficient to satisfy the Statute of Frauds. That does not detract from the critical conclusion in the decision, however, that before such writing was produced there was a trust over the lease which prevented it from passing to the trustee’s assignees upon bankruptcy. The Lord Chancellor, in contrast to the Vice-Chancellor at first instance, referred to this as a trust, not merely a moral obligation. It is not correct, in my view, to characterise the subsequent writing as having “perfected” the trust (as CGC contends in its supplemental skeleton at §23). That would turn s.53(1)(b) from a rule of evidence into a rule of validity.

Gardner is binding on us. Mr McQuater reserved the right to challenge its correctness in the Supreme Court, but submitted that it is in any event distinguishable, since it was a 3-party case, such that the absence of writing could have been overcome by applying the principle in Rochefoucauld. There is some support for that view in the first instance judgment, where the Vice-Chancellor referred to the fact that if Wilkinson had recovered possession of the mine by force of his legal title, he would have recovered “by converting a statute made for the prevention of fraud into an instrument of his own fraud.” That was not, however, the basis upon which the case was decided on appeal. The basis of the Lord Chancellor’s decision was that there was a trust over the lease (as found by the jury as a matter of fact) prior to Wilkinson’s bankruptcy, and that the existence of that trust prevented the lease passing to Wilkinson’s assignees, notwithstanding that the trust had not been evidenced in writing at the time of the bankruptcy. The later execution of the written deed therefore satisfied the Statute of Frauds.

Mr McQuater also contended that it is impermissible to rely on authorities concerned with the Statute of Frauds when interpreting s.53(1)(b), citing Grey v Inland Revenue Commissioners [1960] AC 1. That case concerned the interpretation of s.53(1)(c) LPA 1925. At p.13, Viscount Simonds noted the presumption, when construing a consolidating Act, that the Act is not intended to alter the law, but that this “must yield to plain words to the contrary”. The predecessor to s.53(1)(c) in the Statute of Frauds was s.9, which referred to “grants and assignments” of trusts, whereas s.53(1)(c) used the word “dispositions” of trusts. Viscount Simonds concluded that, although the LPA 1925 was a consolidating Act, it consolidated various pieces of prior legislation, including the Law of Property Acts of 1922 and 1924 which had themselves been pieces of amending legislation. Lord Radcliffe, at p.17 found that these earlier enactments contained numerous amendments to the Statute of Frauds, which were “avowedly changes”. Accordingly, there was no direct link between s.53(1)(c) and s.9 of the Statute of Frauds. As such “it is inadmissible to allow the construction of the word ‘disposition’ in the new Act to be limited or controlled by any meaning attributed to the words ‘grant’ or ‘assignment’ in section 9 of the old Act.”

This does not help CGC when it comes to the consequences of non-compliance with s.53(1)(b). There is indeed an important difference between s.7 of the Statute of Frauds and s.53(1)(b), namely the absence in the later provision of the words “or else they shall be utterly void and of none effect” found at the end of s.7. If anything, however, as observed by Professor T.G. Youdan in Formalities for Trusts of Land, and the Doctrine of Rochefoucauld v Boustead (1984) 43(2) CLJ, p.306, at fn77 on p.321, this can be taken to have confirmed the interpretation of s.7 in the previous cases on the Statute of Frauds. Looking at s.53 in isolation, the orthodox view is supported by the stark contrast between sub-paragraphs (a) and (c) which require the creation or disposition of an interest in land, and the disposition of a subsisting equitable interest or trust, to be in writing, and s.53(1)(b), which requires only that the declaration of trust be evidenced in writing. That clear distinction would be of no effect if the absence of writing rendered a declaration of trust invalid.

In my judgment, the orthodox view is therefore clearly established in law: if a trust has been declared, but there is insufficient evidence to satisfy s.53(1)(b), the trust is nevertheless valid.

The three more recent authorities on which CGC relies do not, on analysis, detract from that conclusion. In Gissing v Gissing, Lord Diplock referred, at p.905A-B, in a passage explaining the difference between express trusts and implied, resulting or constructive trusts, to it being necessary “to constitute a valid declaration of trust”, for the declaration of trust to be “in writing”, citing s.53(1)(b). This is a misstatement of s.53(1)(b) and, in circumstances where the distinction between an invalid trust and one which has been validly declared but not yet evidenced in writing was not addressed at all, it cannot be taken as a rejection of the orthodox view. The same is true of the passage relied on in Lloyds Bank Plc v Rosset, at p.129C, where Lord Bridge referred obiter and in passing to the need for a valid declaration of trust of land to be “in writing”.

In the passage relied on by CGC from Armitage v Nurse, at p.253H, Millett LJ said:

“I accept the submission made on behalf of [the plaintiff] that there is an irreducible core of obligations owed by the trustees to the beneficiaries and enforceable by them which is fundamental to the concept of a trust. If the beneficiaries have no rights enforceable against the trustees there are no trusts.”

The plaintiff’s submission in that case had, however, been as to the content of the duties that formed the irreducible core of obligations owed by a trustee: see the passage from the plaintiff’s argument at p.245D-E. Millett LJ was not concerned with the different question as to the nature of the obligations arising from a declaration of trust of land, where that trust was insufficiently evidenced by writing.

CGC relies in support of its contention that the orthodox view is wrong on an article by Dr David Wilde, Formalities for Declaring Trusts of Land, [2021] 3 Conv 263. Dr Wilde argues that the “conventional interpretation” is wrong. As he notes at p.271, however, there are two recent decisions of the Court of Appeal that adopt the orthodox approach: Sandhar v Sandhar & Kang Ltd [2008] EWCA Civ 238, at §26; and Ong v Ping [2017] EWCA Civ 2069. Although he criticises these cases on the basis that the reasoning is not compelling, they are binding on us. In Ong v Ping, a valid declaration of trust by Madam Lim was found to have been made in April 1986, even though it was not evidenced sufficiently in writing signed by her until two years later (in a letter in which she in fact expressed a wish to revoke the trust): see Sir Colin Rimer’s judgment at §59 and §62.

The fact that, in contrast to Dr Wilde’s article, the leading textbooks (including Snell on Equity, 34^th ed at §22-036; The Law of Trusts by Thomas and Hudson, 2^nd ed., at §5.12; The Law of Trust and Trustees by Underhill and Hayton, 20^th ed., at §14.14; and Lewin on Trusts, 20^th ed., at §3-013 and 3-014) endorse the orthodox view (as recorded by the judge at §203 to §206 of his judgment) further supports the conclusion in favour of this view.

In its supplemental written submissions, CGC’s argument, assuming the orthodox view to be correct, is first expressed as follows: “a non-compliant declaration of trust creates no enforceable obligation requiring the putative trustee to hold the land on trust for the putative beneficiary”, with the consequence that the whole of the beneficial interest in NIOC House remained vested in NIOC, and available to meet claims of its creditors.

In my judgment, there is no material difference between the argument expressed in this way, and the argument that the orthodox view is wrong, such that the lack of writing means there is no valid trust. For all the reasons I have set out above in support of the orthodox view, I reject this way of putting the argument. It fails to meet the point that the lack of writing does not affect the validity of the trust which has been declared, and fails to give any meaning to the conclusion that the trust exists from the moment of its declaration.

It is important in my view to define precisely what is meant in saying that a trust of land which is not yet evidenced sufficiently in writing is “unenforceable”. S.53(1)(b) does not itself say anything about the enforceability or non-enforceability of the trust. While it is true that without the necessary evidence a beneficiary could not enforce the trust against the trustee, that is the indirect consequence of the inability to prove the existence of the trust, not a direct consequence of non-compliance with the section. I agree, in this respect, with the view expressed by Professor Swadling in The Nature of the Trust in Rochefoucauld v Boustead (above) at pp.107 to 108, that s.7 of the Statute of Frauds (and, by implication s.53(1)(b)) is “no more a rule of ‘enforceability’ than validity”.

The other way CGC puts the case, on the face of it, avoids that problem, since it focuses on s.53(1)(b) as an evidential requirement. It contends that, as s.53(1)(b) requires that a declaration of trust “must” be proved by the requisite writing, it follows that a “non-compliant” declaration should not be capable of proof for any purpose unless and until the section is complied with. Accordingly, NIOC is itself unable to prove that it had declared a trust over NIOC House, and cannot rely on the fact that it had done so to show that the Transfer effected a transfer of only the legal estate.

Although more promising, I do not accept this argument. In the absence of any authority in its favour, it is necessary to consider the argument from first principles. In my judgment, it still fails to give effect to the conclusion that the absence of writing does not affect the validity of the trust that has been declared, and it is contrary to (or at least not supported by) a purposive construction of s.53(1)(b).

In addressing the first point, it is important to keep in mind that what matters for the purposes of s.423 (as it does for s.238 or s.339) is whether the judge was correct to find that the absence of writing at the time of the Transfer meant that the beneficial interest in NIOC House remained vested in NIOC. Viewing this from the perspective of NIOC (in accordance with MC Bacon, above), it is difficult to escape the logic that a validly declared trust by A in favour of B has the consequence of separating the legal and beneficial interest in the property, vesting the latter in B, and subjecting A to the obligations of trustee. The effect of s.53(1)(b) is merely that B is unable to prove its existence, which matters if, but only if, A denies that it made the declaration (a point to which I will return when considering the purposive interpretation of the section).

The necessary logic of CGC’s argument, that without the requisite evidence in writing the beneficial interest remains in A, is that it is the provision of the writing that causes the beneficial interest to vest in B (and thus to cause the economic value in the property to be transferred to B). I do not accept that the provision of evidence as to a pre-existing fact (the declaration of trust) has that effect. That is particularly so given the range of circumstances in which the requirement for sufficient writing might be satisfied (see, for example, Thomas and Hudson, The Law of Trusts (2^nd ed) at §5.10; Underhill and Hayton, Law of Trusts and Trustees (20^th ed) at §14.14; and Lewin on Trusts (13^th ed, 1928) at p.49): the writing can take any form, including a letter, any memorandum or a recital in another document; it need not be addressed to the beneficiary; it is effective however long after the creation of the trust, and may be provided in a pleading or affidavit after proceedings have been commenced.

It is also unnecessary that the writing is accompanied by any particular intention on the part of the trustee. In Ong v Ping (above) for example, the requisite evidence was contained in a letter, dated two years after the original (but insufficiently evidenced) declaration of trust, in which the trustee expressed the wish to her solicitors that the trust be cancelled. The validity of the trust – as from the time of its declaration, not merely from it being evidenced in writing – is emphasised by the fact that the original declaration was held to be irrevocable, notwithstanding the absence of writing. The requisite writing might later be found without any positive action on the part of the trustee at all, for example if it is discovered after his death in the safe where he locked it away.

In Hudson v Hathway [2022] EWCA Civ 1648; [2023] KB 345, the Court of Appeal held that an email subscribed with a person’s first name was sufficient to constitute the signing of a document for the purposes of s.53(1)(c)). Accordingly, a simple email exchange in which the trustee answers “yes” to a question as to whether they had made the earlier declaration of trust will suffice.

Accordingly, I do not accept that the later provision of written evidence confirming the existence of the previously declared trust has any substantive impact on the location of the beneficial interest. It is not, in particular, the cause of A (in my above example) being divested of the beneficial interest in the property.

If that is so, I do not see how the transfer of the legal interest from A to B can have the consequence of divesting A of the beneficial interest in the property or of transferring that beneficial interest, or its value, to B. A could have achieved the same result in two steps, by producing a signed memorandum of the trust before executing the transfer. In that case if (which I consider to be correct) the first step did not transfer the beneficial interest to B, then the second step takes place in the context of a valid and evidenced trust and (as is common ground) involves no transfer of any value to B (see §‎135 above). A written transfer of the legal estate may well cover both steps, where it recites the reason for the transfer, viz the pre-existing trust. Where there is no issue, so far as A is concerned, that it has declared a trust, then I do not see how transferring the legal estate without having first provided (or at the same time providing) written confirmation of the trust makes any difference.

This conclusion accords with the purpose of s.53(1)(b). As I have noted under Ground 1, the purpose of the Statute of Frauds was identified in the preamble, as the prevention of fraudulent practices commonly endeavoured to be upheld by perjury and subornation of perjury. The purpose of s.7 was to protect the owners of land from the loss of their land through the perjury of others making false claims as to the existence of a trust over the land. As the Master of the Rolls said in Forster v Hale (above at §‎144), once there is written evidence from the trustee, which may be long after the declaration of trust, the danger of parol declarations, against which the Statute was intended to guard, is entirely taken away.

Mr McQuater submitted that the purpose of s.7 and of s.53(1)(b) went further, and was also intended to provide protection to the creditors of landowners. He submitted that the original intended purpose of s.7 included the protection of creditors and that, although this was watered down so that s.7 became only an evidential requirement, it ought not to be watered down further “to say that you don’t even need to comply with that vis a vis creditors”. This was supported, he said, by the legislative history of s.7, which is helpfully described in the paper by Professor Youdan (above) at pp.307-314. Professor Youdan points out that the Act was passed after much debate and revision. He cited the concerns of Lord Nottingham, who it appears was responsible for drafting the statute, as to the effects of informal trusts on third parties such as creditors, which was reflected in his draft of the provisions relating to the formalities of trusts: “All declarations or creations of trusts or confidences by parol shall be utterly void and of no effect…” Professor Youdan notes, however, that while this would have provided protection for creditors, it was not the effect of the provision eventually passed – at least as interpreted by the courts. Mr McQuater submitted that just because the protection for creditors had been watered down to an evidential requirement does not mean that it should be watered down further to say that s.53(1)(b) need not be complied with “vis-à-vis creditors”.

In my judgment, this legislative history provides no support for the proposition that it is any purpose of s.53(1)(b) to protect creditors. Whatever may have been Lord Nottingham’s purpose, the language of his draft did not survive into the provision as enacted. Whether or not the “utterly void and of none effect” language at the end was intended to protect creditors, the courts had clearly established long before 1925 that those words were themselves of no effect, and they do not appear in s.53(1)(b). A requirement that a declaration of trust be evidenced by some writing is incapable in itself of providing any protection for creditors: there is no requirement of publicity; the section is complied with where the trustee signs a written note of the trust then locks it away in a safe; and it is complied with where the trustee only produces a written note much later.

Mr McQuater’s complaint that NIOC’s argument in this case turns the purpose of the statute on its head, such that it is being used as an instrument of fraud, confuses two different things: the risk of fraud perpetrated on the trustee (NIOC) as the owner of the land, and the risk of the trustee (NIOC) perpetrating a fraud on its creditors. It is the former which is addressed by s.7 and s.53(1)(b). The latter is addressed (today) by s.423. Gardner demonstrates the difference between the two. There was a question in that case about a fraud being perpetrated on creditors, but that related wholly to the original creation of the trust, and was answered by the jury’s finding that the mine was acquired by Wilkinson at the outset for the benefit of Rowe. That was not the mischief with which the Statute of Frauds was concerned. Similarly, in the present case, there is no suggestion that the declaration of trust via the Mortgage in 2019 was a fraud on NIOC’s creditors. It is CGC’s interpretation which turns the purpose of the statute on its head, because it seeks to prevent reliance on the trust by NIOC, the person for whose protection the writing requirement is imposed.

The fact that the evidence necessary to prove a declaration of trust can post-date it, without any limit of time, and after proceedings have commenced, reinforces this view. There is long-standing authority for the last proposition: see Lewin on Trusts, 13^th ed., (1928) at p.49: “The statute will be satisfied, if the trust can be manifested and proved by any subsequent acknowledgement by the trustee, as by … his answer in Chancery, or by an affidavit…”. Mr McQuater accepted this proposition, noting it was established, for example, by Nab v Nab (1717) 10 Mod 404, one of the cases cited in that passage from Lewin.

A colourful example is provided by another of the cases cited by Lewin, namely Cottington v Fletcher (1740) 2 Atk 155. The plaintiff, while a Roman Catholic, assigned land to the defendant. He subsequently conformed to the Church of England and brought proceedings for a re-assignment of the property, contending that he had only assigned it in trust for himself (so as to avoid the penalties of a statute barring Roman Catholics from owning property). The defendant resisted the claim, claiming reliance on the Statute of Frauds, but in his answer admitted that the property had been assigned to him for the purposes stated by the plaintiff. The Lord Chancellor held that had the defendant’s demurrer to the bill stood alone, then his reliance on the Statute of Frauds would have succeeded, but by admitting the trust in his answer he could no longer rely on the Statute of Frauds.

Having regard to the purpose of s.53(1)(b), and the fact that it is a rule of evidence not validity, there is in my judgment a simple justification for this. If the putative trustee is before the Court as a party to proceedings in which the question of the existence of the trust arises, and admits its existence, then the mischief at which the section is directed is removed. There is no need for any proof at all, as to the existence of a trust, if the trustee admits its existence, and a provision which goes only to the nature of the evidence required to prove the trust is simply irrelevant. As the appellants submitted in their post-hearing submissions, evidence is only strictly admissible in proceedings in relation to matters which are in issue, citing Akhtar v Boland [2014] EWCA Civ 872; [2015] 1 All E.R. 644, per Sir Stanley Burnton at §16:

“Where an allegation made by one party in proceedings is admitted by the other party in unqualified terms, that other party must not seek to adduce evidence or raise arguments to the effect that that admission is not binding on him. The court has no jurisdiction to investigate a fact that has been admitted, unless the party making the admission obtains the permission of the court under CPR 14.1(5) to withdraw the admission and does so.”

There is no reason, in my judgment, why the same is not true under s.53(1)(b). This is to be distinguished from the position under s.4 of the Statute of Frauds (which among other things provided that before an action could be brought on a guarantee, there must be a signed memorandum or note of the contract). Because this required the existence of writing before the action could be brought, it has been held that an admission made after proceedings were commenced is insufficient: see James Williams, The Statute of Frauds, Section Four (Cambridge University Press, 1932) citing, for example Lucas v Dixon (1889) 22 QBD 357. Even then, however, an admission in one action may provide the necessary evidence in a second action following discontinuance of the first.

CGC submitted, in its reply submissions, that the conclusion in Nab v Nab was ripe for reconsideration, citing a decision of the trial division of the Supreme Court of Queensland, Livin the Dream v Attorney-General [2025] QSC 21, at §17-21. In that case, the applicant was entitled to rely upon an admission of the existence of a trust contained in an affidavit signed by the respondent (as sufficient evidence for the purposes of the Australian equivalent of s.53(1)(b)), where that affidavit had been made before the applicant commenced its action. But the view was expressed that an affidavit signed after the commencement of the action would not suffice. That case, in turn, cited three authorities. The first two of these (Dudgeon v Chie (1954) 55 SR (NSAW) 450, and Fletcher v Burns (1997) 12 BPR 22937), cited at §16 and §17 of the judgment, related to the Australian equivalent to s.4 of the Statute of Frauds, concerning contracts. As I have noted above, the statutory language is materially different. The third case, Oliver v Renwick Street Pty Ltd [2024] NSWSC 346, did concern the Australian equivalent of s.53(1)(b), but – as the appellants submitted in their reply post-hearing submissions – the reasoning there was also based on contract cases, without noting the difference in language. The Australian cases do not, in my judgment, provide a persuasive reason for reconsidering the English law position so far as s.53(1)(b) is concerned.

The picture is complicated on the facts of this appeal because the judge found that NIOC entered into the Transfer with the requisite purpose under s.423, and there is no appeal against that finding. That does not, however, mean that it was the Transfer that in fact had the effect of putting NIOC House beyond the reach of its creditors. The point is best tested in the context of s.238, where a transaction at an undervalue – if entered into within two years of the onset of liquidation – is avoidable in the liquidation without the need to prove any particular state of mind on the part of the company. The only question, therefore, under s.238, is whether the provision of signed writing proving the existence of the trust, or the transfer of the legal estate to the beneficiary before such writing has been provided, is the cause of the beneficial interest in the asset being vested in the beneficiary. That would be critical in any case where the declaration of trust was outside the two-year period, but the signed evidence was within it (see s.240(1)(a) of Insolvency Act 1986). For the reasons set out here, I do not think that it is.

NIOC cited certain authorities from the United States, which have firmly rejected the notion that third party creditors are entitled to take the benefit of the equivalent in the US to s.53(1)(b). Bogert’s The Law of Trusts and Trustees at §62 refers to the fact that a majority of American States have enacted a form of the Statute of Frauds that closely tracks s.7 of the English statute. At §69, the authors make the point that where the American statutes include wording similar to that in s.7 of the Statute of Frauds, to the effect that the absence of writing renders the trust void, this has generally been construed as meaning “not enforceable over the objection of the trustee”. They point to “many cases” where creditors’ attempts to attack the trust on the ground that it is unenforceable have been rejected. Mr Thanki took us to one such case, in the Supreme Court of California, Cardoza v White (1933) 27 P.2d 639. In that case (where the facts closely mirrored those on this appeal) property was transferred by a Mr Nunes to a Mr Callaghan, subject to an oral declaration of trust in favour of a third person. Callaghan carried out the terms of the trust. Creditors of Mr Nunes argued that the transaction could not be upheld as a trust because it was not evidenced in writing. Langdon J rejected that argument, holding:

“But an oral trust in real property cannot be held wholly void; it is merely unenforceable when, in an action brought to compel performance of its terms, the party to be charged asserts its invalidity. Callaghan, the party obligated, did not refuse to recognize the trust. On the contrary, he carried out its terms. A creditor of the trustor has no right to challenge the voluntary completed performance by a trustee in such a situation.”

Neither this case, nor any of the others cited in the US textbooks to which we were referred, is binding on us. The reasoning, however, accords with the view I have set out above as to the purpose and effect of s.53(1)(b).

That leaves the possibility that because A, in my example, has the practical ability to prevent B enforcing the trust by refusing either to acknowledge the existence of the trust, or to admit it in the context of proceedings, and thus take advantage of s.53(1)(b), the later admission or provision of evidence, or the performance by A of the trust, constitutes the giving up of something of value. This might be thought of as ‘ransom value’: a claim by B to enforce the trust against A depends on A providing a signature confirming the existence of the trust, and A can in practice require some measure of payment from B for doing so. In fact, CGC’s only case as to the value of what was given up by NIOC (being that determined by the judge) is that it was the whole value of the beneficial interest in NIOC House. I do not accept this argument (whether it is based on giving up value equal to the whole of the beneficial interest, or some lesser ransom value).

CGC’s argument in this respect follows from the conclusion (which I accept, as noted above) that the principle in Rochefoucauld does not apply in a self-declaration case. It follows, as CGC contends, that a person who has declared a trust over land is able to prevent the beneficiary from enforcing the trust, provided that (and for as long as) the beneficiary is unable to point to evidence in writing of the trust’s existence and terms, signed by the trustee, even if the trustee knows full well that he did declare a trust. The fact that the law would countenance such behaviour is recognised as the price to pay for ensuring the protection of landowners generally against the risk of fraudulently claimed trusts.

In some sense, therefore, it would be correct to say, in a self-declaration case, that the consequence of the absence of writing is that the putative trustee can deal with the trust property inconsistently with the trust, for example by taking the income for himself or disposing of it to a third party. I do not accept, however, that this means that as a matter of law the trustee is “absolutely entitled” to the property, as suggested by Davies and Virgo, Equity and Trusts: Text, Cases and Materials (3^rd ed), at p.120, or that it follows that if the trustee does give effect to the trust he is transferring something of value to the beneficiary.

For the reasons given above, that would be to ignore the consequence of the orthodox view explained above that, although the beneficiary is as yet unable to prove the existence of the trust with the requisite writing, there is nevertheless a valid trust. If A acts inconsistently with the trust, for example by appropriating part of the property, or its income, to himself, that is still a breach of trust, even if B would not succeed in a claim for breach of trust unless and until sufficient writing is provided by A or obtained by B. (A would remain indefinitely vulnerable to a claim for breach of trust: all that would be required would be a written admission – even if inadvertently made – as to the existence of the trust.)

It is one thing to say that A can in practice prevent B from enforcing the trust, by failing to admit or provide the requisite signed evidence, but another to say that the law should treat A, in a case where A does not dispute that it has created a trust, as retaining the beneficial interest in the property (or something else of significant value in money or money’s worth) unless and until it does so. The mischief at which the section is directed is the risk of fraud played on the trustee, and there is no such risk where the trustee accepts the existence of the trust and acts in accordance with it. Accordingly, having regard to the fact that s.53(1)(b) is a rule of evidence, intended to protect a landowner from false claims as to the existence of a trust, I do not accept CGC’s argument that the lack of sufficient writing means that the trust cannot be proved “for any purpose”. The section, instead, precludes someone who is asserting a trust against a trustee who disputes it, from proving the trust. Specifically, it does not preclude the trustee from relying on a trust which it in fact declared.

This is not to extend the rule in Rochefoucauld that the statute should not be used as an instrument of fraud, because I accept that the trustee in fact remains able to dispute the trust, even if to do so is fraudulent, and also remains able to rely upon the statute as protection against a claim to assert the trust, even if to do so would be a breach of that trust. It is, instead, to prevent reliance on the statute for a purpose for which it was not intended. To do otherwise would be to construe the section in a manner against the interests of the person (the landowner who had declared a trust) for whose protection it was enacted, on the premise that they could choose to act either dishonestly or in breach of that trust, or both. That would not, in my view, be an appropriate basis for construing the section.

CGC contends that the judge was wrong to find that the Mortgage Documents constituted declarations of trust.

There was no real dispute between the parties as to the relevant legal principles. The question whether a trust has been declared is an objective one: whether a reasonable person, armed with the background knowledge available to NIOC, would have understood that NIOC was manifesting, through the Mortgage Documents, an intention to create a trust: Ong v Ping [2017] EWCA Civ 2069, per Sir Colin Rimer at §40.

The judge applied that test in concluding that the Mortgage was a declaration of trust. He noted, at §187 of the judgment, that while not expressly referring to a trust, clause 1.4 of the Mortgage stated that NIOC was the legal owner of NIOC House, and that the Fund was its sole beneficial owner, the necessary consequence of which, in English law, is that NIOC is the trustee of NIOC House for the Fund. He found, at §188, that “A reasonable person would surely conclude that NIOC had created a trust in favour of the Fund.”

At §189 the judge referred to various objections raised by CGC, to the effect that (1) it was implausible that NIOC would have declared a trust in a transaction with a third party; (2) there was no evidence from the signatories as to their intentions; (3) although Blackstone Solicitors acted in connection with the Mortgage, Mrs Nawaz of that firm made no reference to the Mortgage in her witness statement; and (4) the Land Registry refused to register the Mortgage because they had not seen the trust deed. His answer to each of these points was that they were either irrelevant to, or did not detract from, the objective meaning of clause 1.4. He reached a similar conclusion in relation to the Certificate of Title at §191.

CGC contends that in reaching this conclusion the judge wrongly ignored a number of relevant background circumstances:

The question whether there has been a declaration of trust is a question of fact: Paul v Constance [1977] 1 WLR 527, per Scarman LJ at p.531H to p.532B. Unless the judge made an error in determining what constitutes, in law, a declaration of trust, the question on appeal therefore is whether there was sufficient evidence to justify the judge’s conclusion of fact.

Points (1) to (3) and (6) boil down to the submission that the judge was wrong to ignore the fact that NIOC, the Fund and their advisers were acting under the mistaken belief that NIOC House already belonged beneficially to the Fund. The fourth point (that the appellants were unaware that the Fund lacked separate legal personality) goes to the same issue, since it negates the possibility that the appellants knew that the pre-2019 declarations of trust relied on were ineffective.

CGC contends that this is inconsistent with the need for a “present intention” to declare a trust. It contends that the requirement for a present intention means that a person cannot be taken to have declared a trust which they believe already exists. It is necessary to distinguish, it says, between (i) an erroneous prior and continuing belief that ownership was vested in the Fund and (ii) the need for a positive act to ‘correct’ that erroneous belief by language signalling a clear, unequivocal and irrevocable intention to immediately constitute a trust at a particular point in time (i.e. on the date when the Mortgage Documents were executed).

The appellants contend, on the other hand, that insofar as the authorities refer to a requirement for a “present intention” to constitute a trust, this means no more than that the three certainties required to create a trust (certainty of intention, certainty of object and certainty of subject) must coincide. Specifically, the reference to a present intention precludes a declaration of trust arising where the intention is that it should arise only upon the occurrence of some further step to be taken in the future.

The appellants contend, therefore, that it is not necessary to show that the declaration of trust was intended to effect a change in the relationship between trustee and beneficiary and that, conversely, an effective declaration of trust can be found even where the settlor believes he is confirming a pre-existing arrangement.

The appellants pointed to five authorities in support of that proposition.

First, Re Northcliffe [1925] Ch 651. In 1911, Lord Northcliffe settled on trust certain existing property, and property he may acquire in the future. This was an ineffective declaration of trust, insofar as it referred to freehold property not then owned by Lord Northcliffe. Subsequently, having acquired further properties, Lord Northcliffe made a will in which he stated “I also confirm the undermentioned deeds and settlement” (referring to the 1911 settlement). It was contended by a beneficiary of the will that Lord Northcliffe had, by his will, merely confirmed that the 1911 settlement was to be regarded as valid and binding, but he had no intention to declare himself trustee of the freeholds. Russell J found, however, that the will constituted a valid declaration of trust over the later acquired properties, reading it as “a declaration by the testator that he holds the Kent freeholds upon the trusts of the settlement”.

Second, Grey v IRC [1958] Ch 690. On 1 February 1955, a settlor transferred shares in a company to the appellants (trustees of existing settlements in favour of his grandchildren) as his nominees. On 18 February 1955, the settlor orally directed the appellants to hold the shares on the trusts for his grandchildren. On 25 March 1955, the appellants executed declarations of trust, the operative parts of which stated that the trustees acknowledged that they had been since the preceding 18 February “and are now” holding the shares upon the specified trusts. The Court of Appeal held that the direction of 18 February 1955 was a purported disposition of an equitable interest in the shares, and failed for lack of writing by reason of s.53(1)(c) LPA 1925. The declaration of 25 March 1955 was, however, held to be valid. Lord Evershed MR, at p.707, said this:

“Whatever might be or have been the effect (if any) of the trustees' acknowledgment or declaration that they had so held the shares since the preceding February 18, it clearly follows, in my judgment, that (on the hypothesis that the oral directions on February 18 had no legal effect) the instruments of March 25, 1955, must have effectively established or constituted the relevant trusts.”

This aspect of the decision was not addressed or challenged when the case went on appeal to the House of Lords.

Third, Paul v Constance (above). In that case, the deceased had set up a bank account into which he paid the proceeds of a damages claim following an injury at work. He was then living with the plaintiff as man and wife. At various times he said to the plaintiff: “the money is as much yours as mine.” The Court of Appeal held that this was sufficient to constitute a valid declaration of trust. Scarman LJ,with whom Cairns and Bridge LJJ agreed, said (at p.531G) that “there must be clear evidence from what is said or done of an intention to create a trust”. The judge at trial had concluded that there was. On appeal, as Scarman LJ noted at p.532A-B, the question was whether there was sufficient evidence to justify the judge’s conclusion. The judge had accepted the submission of counsel for the plaintiff that the words used by the deceased on more than one occasion conveyed “clearly a present declaration that the existing fund was as much the plaintiff’s as his own”, and Scarman LJ considered that the judge had been right to do so. He said, at p.532E-F:

“It might, however, be thought that this was a borderline case, since it is not easy to pin-point a specific moment of declaration, and one must exclude from one's mind any case built upon the existence of an implied or constructive trust, for this case was put forward at the trial and is now argued by the plaintiff as one of express declaration of trust. It was so pleaded and it is only as such that it may be considered in this court. The question, therefore, is whether, in all the circumstances, the use of those words on numerous occasions as between the deceased and the plaintiff constituted an express declaration of trust. The judge found that they did. For myself, I think that he was right so to find. I therefore would dismiss the appeal.”

Fourth, Rowe v Prance [1999] 2 FLR 787. However, this is simply an example – similar to Paul v Constance – of the court finding an express declaration of trust (of a boat) from various statements and the conduct of the ‘trustee’, including repeated use of the word “our” in relation to the boat.

Fifth, Grant v Grant (1865) 34 Beav. 623, where a husband was found to have declared a trust over certain chattels for his wife. The case established that, while there must be clear, unequivocal and irrevocable words declaring a trust, it was not necessary to use technical words, such as “I hold the property in trust for you”. Sir John Romilly MR said, at p.625: “Any words that shew that the donor means, at the time he speaks, to divest himself of all beneficial interest in the property are, in my opinion, sufficient for the purpose of creating the trust.” He continued, at p.626, that, although it was not sufficient to say “I intend to give it to you”, because in that case “the thing is not complete”:

“But if the donor makes an express declaration that “I do now give it”, I am opinion that is sufficient. I am also of opinion that it is sufficient if he makes a declaration “I have already given it”.

In my judgment, these cases make good the proposition advanced by the appellants. There is no indication in either Northcliffe or Grey that the settlor was aware of the invalidity of the original declaration (and so was consciously correcting an erroneous belief) or that this was a necessary ingredient before the later declaration – confirming what was believed to be an existing trust – could be effective as the declaration of trust. The fact, as confirmed by Paul v Constance, that a sufficient intention to declare a trust can be inferred from words and conduct repeated over a period of time is inconsistent with a requirement that the declaration of trust must coincide with a point in time at which there is intended to be a change in the relationship. That is also confirmed by Grant v Grant which shows that words indicating an intention to declare a trust in the future are insufficient, but words confirming having done so in the past will suffice.

It may in a particular case be necessary to pinpoint the moment that a trust arises (for example where the timing of the creation of a trust is relevant for tax purposes), but that does not mean that no trust can be held to exist unless it is possible to point to a single occasion which constitutes the declaration of trust. The inquiry, as Ong v Ping shows, is a broader one than that: would the reasonable observer understand from a person’s words and conduct that they were manifesting an intention to hold property on trust for another.

In fact, I understood CGC to accept that a declaration of trust need not effect an intended change in the relationship. In its supplemental skeleton dealing with Respondent’s Notice Point 1, it accepted “a present intention – which may be manifested alongside a statement of past belief as to the existence of ownership and/or a trust … – is a necessary ingredient of a declaration of trust.” The problem, it contends, for NIOC is that its erroneous statement (in the Mortgage Documents) of prior and continuing belief that ownership was vested in the Fund was not coupled with the requisite present intention to declare a trust.

That, however, is an overly restrictive interpretation of the Mortgage Documents. It would be obvious, to the reasonable person viewing the Mortgage Documents, with the knowledge that NIOC believed that NIOC House had long been beneficially owned by the Fund, that NIOC was at the very least confirming, as at the date of the Mortgage Documents, that the property was held on trust. In my judgment, the judge was entitled to find that this suffices to establish the requisite intention for the declaration of a trust.

CGC was unable to identify any authority for the proposition that there must also be a present intention to create a trust, in the sense that this precluded a declaration of a trust which the trustee already believes to exist. In its skeleton argument it cited Re Cozens [1913] Ch 478, at p.487, for that proposition. All that Neville J said there, however, was that he was not satisfied by the evidence that there was ever any present and irrevocable intention on the part of Cozens to declare himself a trustee, as the evidence suggested instead that he intended to create a charge.

CGC placed some reliance on an article by Sinead Agnew and Simon Douglas, entitled Self-Declarations of Trust (2019) 135 LQR p.67. The authors point out the important role played by a declaration, in a self-declaration case (as opposed to a “3-party” case), because the declaration is the only legally relevant transaction. They suggested, at p.79, that two things are required for an express trust to be effectively declared: “(a) an intention to hold rights for the benefit of another and (b) an intention that by the settlor’s very words or action, the trust relationship should be immediately constituted.” As is made clear over the following two pages, however, what they mean by this is that there must be an intention that the putative trustee is from that moment – as opposed to some point in the future – divested of its beneficial interest in the property. In each of the cases referred to by the authors in this section of the article, the person seeking to establish the trust was found to have the intention of doing so pursuant to some further step, such as by a formal declaration of trust drawn up by lawyers (as in Rabin v Gerson [1986] 1 WLR 526). Where that step did not take place, it was insufficient to fall back on the general intention to create a trust, manifested by the instructions to lawyers, because that intention was not that the person declaring the trust was immediately divested of their beneficial interest in the property. Nothing in this article undermines, in my view, the proposition that a declaration of trust can be effective even if it is understood (whether rightly or wrongly) to confirm an existing position.

As to the second of CGC’s criticisms of the judge, that the parties believed that the property was subject to an Iranian law amanat, as Mr Mumford submitted, a settlor need not understand the legal consequences of their intentions. What matters is an intention that the beneficial interest in property held at law by the settlor should vest in someone else. That is precisely the effect, in English law, of an intention that property should be subject to an amanat under Iranian law. The fact that Iranian law does not recognise a split title is irrelevant, when the question is as to the consequence, in English law, of the settlor’s intention.

CGC’s fifth point of criticism of the judge is that he ignored the implausibility that NIOC would have intended to declare a trust in documents that related to the creation of a charge in favour of a third party, the bank. This is a straightforward disagreement with the judge’s decision on the facts. Mr Thanki took us to a number of email exchanges prior to the execution of the power of attorney, which indicated that: the bank would have preferred NIOC to be removed from the title documents, so that the bank would be dealing with the underlying reality that the Fund owned the property; but the Fund’s solicitor (Mr Cathcart) preferred that NIOC should retain its position as “custodian trustee”; and Mr Cathcart indicated that he would work with the solicitors acting for the bank “to clarify the issues of ownership and also look at the sort of documentation which will be required by the bank’s solicitors to properly affect [sic] a legal charge over the property”. The implausibility objection was taken before the judge by CGC and clearly taken into account by him (see §189 of the judgment). The evidence to which Mr Thanki took us shows that he was right to reject it. Given the bank’s need for clarity over the ownership position as between NIOC and the Fund, where both entities were identified as mortgagor in the Mortgage Documents, it is entirely plausible that NIOC would have used the Mortgage Documents to declare a trust over the property.

CGC’s seventh point is addressed below under Respondent’s Notice Point 2(b).

As to its eighth point, Eversheds’ own understanding of the position, the subjective understanding of NIOC’s and the Fund’s lawyers is irrelevant, since the test is an objective one.

For these reasons, I reject the argument made in Respondent’s Notice Point 2(a).

I have already referred to the judge’s finding that the Mortgage Documents constituted declarations of trust by NIOC. It is implicit in this finding that those who signed the documents were authorised to do by NIOC. As to the Mortgage the judge said expressly that it was executed “by NIOC, acting by its attorney, NTT.”

CGC contends nevertheless that the judge’s finding that NIOC made declarations of trust cannot stand because the appellants failed to adduce any evidence that any of NTT, Mr Rahgozar or Eversheds were authorised to execute a declaration of trust on NIOC’s behalf. Mr McQuater submitted that it was for the appellants to establish that the Mortgage Documents constituted declarations of trust, that in order to do so they needed to establish that they were signed with NIOC’s authority, and that they failed to do so. Indeed, they did not even seek to establish as much at trial.

It is accepted by the appellants that the judge did not make any actual finding that NTT, Mr Rahgozar or Eversheds were duly authorised to make a declaration of trust on behalf of NIOC. That, however, is because the point was never in issue.

I have referred to the relevant pleadings at §‎113 to §‎114 above, in connection with Ground 2. As there noted, NIOC relied on the Mortgage Documents which were signed by someone purporting to be authorised to do so by NIOC. So far as relevant to Respondent’s Notice Point 2, NIOC pleaded reliance on the Mortgage Documents as constituting declarations of trust and, following CGC’s denial of that proposition, sought further information of that denial. In that further information, CGC put in issue whether signature by an agent was sufficient, as a matter of law, to comply with s.53(1)(b). It did not, however, put in issue that those who made the declarations of trust were authorised to do so.

In my judgment, this is the mirror image of the point raised by Ground 2: whereas it was for NIOC to plead and establish that the signature of Mr Rahgozar, NTT or Eversheds constituted that of NIOC for the purposes of Regulation 4, it was for CGC, faced with a pleading that the Mortgage Documents were declarations of trust by NIOC, and with a request for information of its bare denial of that plea, to put in issue the authority of those who executed those documents to act on behalf of NIOC. The point not having been taken at trial, it is too late to do so on appeal, for similar reasons as those set out above under Ground 2. If the authority of Mr Rahgozar, NTT or Eversheds to declare a trust on behalf of NIOC had been put in issue, NIOC and the Fund would have had the opportunity to call evidence on the point.

Mr McQuater suggested that we could find, on the basis of the documents before us on this appeal, that there was no basis for the contention that either NTT or Eversheds was authorised to declare a trust. He pointed, in the case of NTT, to the fact that the case on NTT’s authority is based on a written power of attorney, and that on its face it is inconsistent with there being any authority to declare a trust over NIOC House.

Pursuant to the power of attorney, dated 25 July 2018, NIOC appointed NTT to be its attorney to “sign, or execute and deliver as deeds, as the case may be, any agreements, leases, deeds of grant, pledge for loan security, mortgages and legal charges (the “Documents”) to which [NIOC] is a party”, and “any other documents which are referred to in the Documents or which are ancillary to or related thereto or contemplated by the provisions thereof”, and to “do and perform any and all of the other acts, matters or things in connection with the Documents”. The power of attorney was expressly stated, however, not to cover the sale or transfer of ownership of NIOC House.

Mr McQuater submitted that this wording was clearly incapable of authorising NTT to declare a trust over NIOC House, since it was limited to executing documents relating to the intended mortgagee in favour of the bank over NIOC House. As with any contract, however (it being common ground that the power of attorney is governed by English law) it must be construed in light of the background context within which it was executed. Mr Thanki highlighted a number of elements of that context which are at least potentially relevant in this regard.

First, it was the common understanding of those involved in executing the power of attorney that the benefit of NIOC House was already, and had for a long time been, vested in the Fund. That tends to suggest that what was envisaged by the prohibition on the sale or transfer of ownership of NIOC House was any sale or transfer of ownership away from the Fund.

Second, the email exchanges prior to the execution of the Mortgage, to which Mr Thanki took us (referred to above at §‎224, and for the same reasons there given) provide an arguably plausible justification for the power of attorney extending to making a declaration of trust over NIOC House.

It is not possible, therefore, to conclude on the basis of the material before us that there is no prospect of NIOC establishing that NTT was authorised to declare a trust over the Property. The point cannot be determined either way without a further trial. For that reason, this is a clear case, in my judgment, where it is too late for CGC to put in issue NTT’s authority to declare a trust over NIOC House on behalf of NIOC.