What are they?
It seems most sensible to define resulting trusts as follows –
A resulting trust exists when a trustee holds some right or interest on trust for the person from whom he received that right or interest.
On this view, what marks out a resulting trust as ‘resulting’ is who the trustee holds on trust for – he holds on trust for the person from whom he received the trust property. Under a resulting trust, the beneficial interest in the trust property has ‘jumped back’ (in Latin, resalire) to the person from whom the trust property has derived.
On this definition, it’s a mistake to think that a given trust can be classified as ‘express’ or ‘constructive’ or ‘resulting’. That’s like saying that a jumper can be classified as ‘hand-made’ or ‘machine-made’ or ‘red’. ‘Express’ and ‘constructive’ refer to how a trust came into existence (either it came into existence because the settlor intended to create it, or because it arose under some rule of law); ‘resulting’ refers to what it looks like. A given trust can be classified as ‘express’ or ‘constructive’ when we ask how it arose; it can be classified as ‘resulting’ or ‘non-resulting’ depending on for whom the trust property is held on trust. So it’s conceptually possible for a given trust to be: ‘express resulting’, ‘express non-resulting’, ‘constructive resulting’, or ‘constructive non-resulting’.
Because of s 53(2) of the Law of Property Act 1925, which says that ‘This section does not affect the creation of resulting, implied or constructive trusts’, many people think that when we should not refer to a trust as being ‘resulting’ if it is also ‘express’ in nature. But, as we will see, there are trusts that are routinely referred to as ‘resulting’ which some argue are also ‘express’ in nature.
When do they arise?
There seem to be three situations where a resulting trust will arise:
(1) Where A transfers a right or interest to B and B agrees to hold that right or interest on trust for A.
(2) Where A transfers to B, or purchases for B, an interest in land in circumstances where the law presumes that A did not intend to make a gift of that interest in land to B. (The trust arising in this situation is known as a ‘presumed resulting trust’.)
(3) Where A transfers to B a right or interest, intending that B should hold that right or interest on trust, and that trust fails. (The trust arising in this situation is known as an ‘automatic resulting trust’ – though as people such as Peter Birks and Lord Browne-Wilkinson have pointed out, there is nothing automatic about a resulting trust arising on trust failure. A might have intended that B should keep the right or interest for herself if it could not be held in trust; and in such a case, there will be no trust back to A if the trust on which A intends B to hold the right or interest fails.)
(There is a possible fourth situation (though some would say that it is simply an example of (1) and I would say that it is an example of (3)), where A transfers money to B but specifies that that money is only to be used for a non-charitable purpose. It is generally accepted that in this situation B will hold the money on trust for A, but will be allowed to use the money for the purpose specified by A. This trust is known as a Quistclose trust, and is dealt with in a separate note on this website.)
Why do they arise?
There is no problem with understanding why the resulting trust in (1) arises – it arises because that was what A and B intended.
I also think there is no real problem with understanding why presumed resulting trusts arise – they arise because the law presumes, when A transfers the interest in land to B that A intended that that interest should be held on trust for him, and the law gives effect to that intention. It might be objected that such an explanation is inconsistent with s 53(1)(b) of the Law of Property Act 1925, which provides that:
‘a declaration of trust respecting any land or any interest therein must be manifested and proved by some writing signed by some person who is able to declare such trust…’
As there will be no writing evidencing A’s intention that the interest in land that was transferred to B should be held on trust for him, it might be argued that the courts would violate s 53(1)(b) if they gave effect to that intention. However, Bill Swadling has convincingly argued that all that s 53(1)(b) says is: ‘If you want to base your claim on someone’s making a declaration of trust over land, then you have to provide some written evidence that that declaration of trust was made.’ But in a presumed resulting trust situation, A has no need to base his claim against B, that B holds on trust for him, on an allegation that he intended that B should hold on trust for him. All A has to do, to make out his claim, is to show that he transferred an interest in land to B, and the courts will do the rest. They will presume that A intended that B should hold that interest on trust for him, and then ask B to provide evidence that that was not A’s intention.
So, for example, in Hodgson v Marks (1971), Hodgson owned a house in which Evans stayed as a lodger. Evans started expressing concerns that Hodgson’s son would force him out of the house. Hodgson sought to allay those concerns by transferring title to the house to Evans. It was understood by both Hodgson and Marks that the house would be held on trust for Hodgson, but nothing was put in writing. Evans later sold the house to Marks. The Court of Appeal held that when Hodgson sold the house to Evans, Evans held it on trust for Hodgson. In so holding, they did not act inconsistently with s 53(1)(b). In order to establish that Evans held on trust for her, Hodgson did not need to rely on their agreement that he would hold on trust for her (which agreement would have had to have been evidenced in writing in order to be relied on under s 53(1)(b)). All she had to do was show that she had conveyed her house for nothing to a comparative stranger and then sit back. The courts would then presume that Hodgson had intended Evans to hold on trust for her and switch their attention to Evans and ask whether there was any evidence that she had intended to make an outright gift to her. As no such evidence could have been offered, the courts would find that the house had been transferred to Evans to be held on trust for Hodgson and give effect to that trust. And that’s exactly what happened in Hodgson v Marks. So, in effect, the resulting trust over the house in Hodgson v Marks was an express trust, where Hodgson did not attempt to prove directly that she intended that the house should be held on trust for her (which would have required that that intention be evidenced in writing) but simply set up a presumption (which went unrebutted) that she intended that the house be held on trust for her by proving that she conveyed it for nothing to a comparative stranger. And all presumed resulting trusts over interests in land can be analysed in a similar manner.
The real difficulty is with so-called ‘automatic’ resulting trusts – trusts arising in situation (3), above. A number of different theories have been advanced as to why such trusts arise. In order to consider them, let’s see how they apply to a stock situation – call it Dead Man – where A transfers money to B to be held on trust for C and C is already dead at the time the transfer is made. The trust for C having failed to come into existence, B ends up holding the money on trust for A. Why?
The first theory – endorsed by Lord Browne-Wilkinson in Westdeutsche Landesbank v Islington LBC (1996) – is that we find that B holds on trust for A in Dead Man because we presume that A intended that if B could not hold on trust for C, then she should hold on trust for A. The big problem with this theory is Vandervell v IRC (1967). In that case, Vandervell transferred 100,000 shares in his company to the Royal College of Surgeons (RCS). The idea was the RCS would receive premiums on the shares, and once they had received £150,000 in premiums (the exact amount they needed to endow a chair in Vandervell’s name), the shares would be purchased back from the RCS for £5,000. The option to repurchase the shares was vested in a trustee company. At the time the shares were transferred and the option to repurchase the shares vested in the trustee company, Vandervell never said for whom the option to repurchase should be held on trust for. It was suggested that it be held on trust for Vandervell’s children or on trust for the employees working in Vandervell’s company. Vandervell was happy with either suggestion, but did not say which option he wanted to go for. The House of Lords held that at the time the option to repurchase the shares was retained, it was held on trust for Vandervell: if A transfers property to B to be held on trust, and does not specify for whom, then it is to be held on trust for A. Now – the decision in Vandervell v IRC creates a problem for this first theory as to why resulting trusts arise in situation (3) because it’s not possible to presume in Vandervell that Vandervell intended that the option to repurchase be held on trust for him. Whatever else we know, we know for certain that the last person in the world Vandervell wanted the option to repurchase to be held on trust for, was him. That’s because if he retained an interest in the shares while they were held by the Royal College of Surgeons he would remain liable for tax on the premiums on those shares. And that’s what happened as a result of the decision in Vandervell v IRC – Vandervell became liable for tax on the dividends paid on the shares before they were repurchased. So it’s hard to say that the resulting trust over the option to repurchase the shares was giving effect to an intention we presume that Vandervell had, that that option to repurchase should be held on trust for him.
In order to overcome this problem, in Re Vandervell’s Trust (No 2) (1974), Megarry J came up with a second theory as to why resulting trusts arise on the failure of a trust. This theory led to these resulting trusts being called ‘automatic’ – a label that has stuck on them to this day, despite academic and judicial denials that such trusts arise automatically. The reason why Megarry J called them ‘automatic’ was because he thought resulting trusts arising out of the failure of a trust arose automatically, irrespective of the intention of the person attempting to set up the trust. They had to arise automatically, Megarry J thought, if we are to explain why there was a resulting trust in Vandervell v IRC. Megarry’s theory as to why resulting trusts arise on the failure of a trust helped to explain – he thought – why they arose automatically. On his theory, the reason why B ends up holding on trust for A in Dead Man is that at the start of the story A has a legal and beneficial interest in the money he wants to transfer to B. A successfully transfers the legal interest in the money to B, but is unsuccessful in transferring the beneficial interest to C. So the beneficial interest remains in A: you keep what you don’t give away. At the end of the story then, B has legal title to the money, and A has the beneficial interest. Hey presto: B holds on trust for A. We can explain the trust in Vandervell v IRC in the same way. At the start of the story, Vandervell has a legal and beneficial interest in the option to repurchase. He transfers the legal title to his trustee company but dithers over who should get the beneficial interest and ends up giving it to no one. So it remains in him: he keeps what he does not give away. At the end of the story then, Vandervell’s trustee company has legal title to the option to repurchase, but Vandervell (horror of horrors from his point of view) still has the beneficial interest. Hey presto: the trustee company holds the option to repurchase on trust for Vandervell.
This seems very neat, but it suffers from a huge problem. The problem is that the entire theory depends on our accepting that at the unencumbered legal owner of property has both a legal and beneficial interest in that property. This is untrue. All the unencumbered legal owner of some property has is legal title to that property. He is able to enjoy the property beneficially, true – but that is not because he has a beneficial interest in the property. It’s because no one else has a beneficial interest in that property that could stop the legal owner enjoying the property himself. Ironically, Vandervell v IRC itself refutes Megarry’s theory as to why resulting trusts arise on trust failure in cases like Vandervell. While the Inland Revenue’s argument that Vandervell had retained an interest in the shares conveyed to the RCS by virtue of the fact that the option to repurchase was held on trust for him, it had also tried to run an argument that the shares that were transferred to the RCS were held by the RCS on trust for Vandervell. The Revenue’s argument was that before the shares were transferred to the RCS, they were held on trust for Vandervell, and Vandervell’s oral instruction to his trustees to transfer the shares to the RCS so that they RCS would become absolute legal owners of the shares was incapable of divesting him of his beneficial interest in the shares. This is because – the Revenue argued – Vandervell’s instruction amounted to an attempted to dispose of his beneficial interest in the shares to the RCS and that could only be done in writing under s 53(1)(c) of the Law of Property Act 1925. The House of Lords rejected this argument. What Vandervell was trying to do in instructing that the shares held on trust for him should be transferred to the RCS was destroy his beneficial interest, not transfer it to the RCS. Compliance with Vandervell’s instruction would result in a situation where the RCS was the legal owner of the shares and no one would have a beneficial interest in the shares: not Vandervell, and not the RCS.
Megarry J’s explanation of why resulting trusts arise on trust failure is now generally rejected. However, Lord Millett might be taken to have attempted to revive it in Air Jamaica Ltd v Charlton (1999), when he observed that:
‘A resulting trust…arises whether or not the transferor intended to retain a beneficial interest – he almost always does not – since it responds to the absence of any intention on his part to pass a beneficial interest to the recipient.’
So, Lord Millett might be taken to be arguing here that in Dead Man,B holds on trust for A because the beneficial interest in the money cannot vest in C, and A does not intend that B should have the beneficial interest. Given this, the only person who can have a beneficial interest in the money is A. However, the problem with this is that it assumes someone must have a beneficial interest in the money. The possibility that no one has a beneficial interest in the money, and B is left free to walk away with the money as its legal owner is overlooked.
However, my view is that Lord Millett misspoke in Air Jamaica and did not intend to endorse Megarry J’s theory as to why resulting trusts arise on trust failure. When Lord Millett said that a resulting trust ‘responds to the absence of any intention on [the transferor’s] part to pass a beneficial interest to the recipient’ he really meant to say that in Dead Man, B holds the money on a resulting trust for A because A didn’t intend that B should benefit from that money. So the focus is not on who A intended should have a beneficial interest (a proprietary concept) in the money, but on who A intended should benefit (a straightforward, non-legal concept) from the money. If A did not intend that B should benefit from the money that he has transferred to B, then B would be unjustly enriched if B were allowed to walk off with the money and enjoy the benefit of it. In order to prevent this happening – as it might if we simply say that B is the legal owner of the money and doesn’t hold on trust for anyone – we find that B holds the money on trust for A, the person at whose expense B would be unjustly enriched if he were allowed to walk off with the money.
This is the Birks-Chambers theory of why resulting trusts arise. It is named after Peter Birks, who first advanced the theory in his Introduction to the Law of Restitution (1989), and Robert Chambers, who developed the theory in a PhD (supervised by Peter Birks) that was turned into a book, Resulting Trusts (1997). On this theory, resulting trusts arise on trust failure to prevent someone enjoying the benefit of property that he was never intended to benefit from. So, on this theory, in Dead Man, B holds the money on trust for A not because (as the first theory would suggest) A had a positive intention that B should hold on trust for A, but because A did not have a positive intention that B should benefit from that money.
(The fact that Lord Millett says in Air Jamaica that resulting trusts arising on trust failure respond to an ‘absence of intention’ on the part of the transferor makes me think that he was actually meaning to endorse the Birks-Chambers thesis, but screwed up by focussing on the absence of an intent to pass a beneficial interest to the recipient instead of focussing on the absence of an intent that the recipient should benefit from the money. He did a bit better in the subsequent case of Twinsectra Ltd v Yardley (2002), where he said that: ‘The central thesis of Dr Chambers’s book is that a resulting trust arises whenever there is a transfer of property in circumstances in which the transferor…did not intend to benefit the recipient.’ But in the very next sentence, he can’t stop himself sliding back into Air Jamaica type language: ‘It responds to the absence of an intention on the part of the transferor to pass the entire beneficial interest, not to a positive intention to retain it.’)
The fact that the Birks-Chambers theory as to why resulting trusts arise on trust failure focuses on the fact that in a case like Dead Man, A did not intend that B should benefit from the money rather than (as with the first theory) arguing that A did intend that B should hold on trust for him, means that the Birks-Chambers theory can handle Vandervell v IRC. They can argue that the trustee company held the option to repurchase on trust for Vandervell because, whatever else he intended, he did not intend that the trustee company should benefit from that option to repurchase; so the trustee company was held to hold the option to repurchase on trust for Vandervell in order to stop it exercising that option to repurchase for its own benefit.
So – should we accept the Birks-Chambers theory as to why resulting trusts arise on trust failure? There are two problems with it.
The first problem is that if the Birks-Chambers theory were correct, we would expect to find resulting trusts arising whenever A transfers money to B and does not intend B to benefit from that money. In particular, we would expect a resulting trust to arise in a situation where A transfers money to B by mistake. But there is no authority that a resulting trust will arise in such a situation. The furthest the courts have been willing to go is that where A pays B money by mistake, B may hold the money on trust for A if B was aware that the money was transferred to him by mistake: Chase Manhattan Bank v Israel-British Bank (1981), as explained by Lord Browne-Wilkinson in Westdeutsche Landesbank v Islington LBC (1996). And even then some judges have been unwilling to go that far: see Shalson v Russo (2005), per Rimer J at .
This problem is not insuperable. It could be argued that if the law doesn’t find resulting trusts when the Birks-Chambers theory indicates that it should, that is a problem that originates in the law’s failure to give full effect to the principles which underlie it, and does not indicate that there is something wrong with the Birks-Chambers theory.
The second problem is much more serious. It is that in a case like Dead Man, it is not clear why the law has to hold that B holds on a resulting trust for A in order to prevent B being unjustly enriched at A’s expense. Why not simply allow A a personal remedy against B, allowing A to sue B for the value of the money that he transferred to B? That would seem to be sufficient to prevent B being unjustly enriched as a result of being allowed to retain the money that A transferred to B. If the Birks-Chambers theory is to satisfactorily explain why resulting trusts arise in a case like Dead Man, it has to explain why a proprietary remedy, rather than a personal remedy, is required to prevent B being unjustly enriched. This is a challenge that has not yet been met.
So – we have looked at three theories so far as to why resulting trusts arise on trust failure and seen that all of them suffer from one flaw or another. The theory that the resulting trust in Dead Man arises to give effect to an intention we presume that A had that B should hold the money on trust for him if it cannot be held on trust for C runs into the rock of Vandervell v IRC. The theory that the resulting trust arises because A has failed to give away his beneficial interest in the money that he transferred to C simply does not work: A never had a beneficial interest in the money to give away. The theory that the resulting trust arises because B would be unjustly enriched if she were allowed to enjoy that money for her own benefit runs into the problem that allowing A to sue B for the value of that money would seem sufficient to prevent B being unjustly enriched. So it seems likely that if we are to explain satisfactorily why resulting trusts arise in cases like Dead Man, we have to look elsewhere.
The fourth theory we should consider is the ‘gap’ theory, first proposed by Lord Denning MR in Re Vandervell’s Trust (No 2) (1974) and revived by John Mee in a recent essay (‘“Automatic” resulting trusts: retention, restitution, or reposing trusts?’ in Mitchell (ed), Constructive and Resulting Trusts (2009)). The idea is that in a case like Dead Man, B won’t be allowed to deny that he holds the money he received on trust, but as C is already dead, we have a problem: we must find that B holds on trust for someone, but for whom? We solve the problem by finding that B holds on trust for A. This theory only works if we accept that B simply won’t be alowed to deny that he holds the money he has received from A on trust. This is plausible: there is something a bit ‘off’ about someone who receives money on the express understanding that they will hold it on trust and then attempts to turn round and argue that they are entitled to enjoy that money for their own benefit. Someone who attempted to play fast and loose with the truth in this way might be held to be acting unconscionably, and Equity could be expected to prevent someone acting in this way. So the ‘gap’ theory has a lot going for it.
An alternative to the ‘gap’ theory is the ‘invalidity’ theory. The idea is that in a case like Dead Man, A’s attempt to create a trust for C has failed. When someone tries to peform a legal act that turns out to to be legally invalid, the natural legal response is to put them back to where they were in the first place. But in Dead Man, Equity cannot do this. A has transferred to B legal title to the money that was meant to be held on trust for C, and Equity cannot reverse that transfer. The best it can do is to find that B holds on trust for A. This is the closest Equity can get to getting A back to where he was before he made his ineffective attempt to create a trust for C. Once B holds on trust for A, A can exercise his Saunders v Vautier rights to get back legal title to his money from B, and that will restore the parties to the position they were in before A attempted to transfer the money to B to be held on trust for C. So Equity finds that B holds on trust for A simply because A has not succeeded in his attempt to create a valid trust for the benefit of C, and finding that B holds on trust for A is the best it can do by way of reversing what A has done.
Either the ‘gap’ or ‘invalidity’ theories seem to me to explain satisfactorily why we find resulting trusts in a situation like Dead Man where a trust has failed. It is very hard to think of a scenario which would determine which theory is ‘better’. But either seems much more satisfactory than the first three theories we considered.