Chapter 7: Informal Trust

Introduction

You can use this workbook in a number of ways - but the main idea is that each section takes you through a series of questions, not all of which have clear yes/no answers, but where generally I have a preferred view which I express. The workbook should take about an hour if you just try the questions, or if you are using the workbook for revision purposes.

You can also obtain additional reference material if you require it, while working through the questions. There is quite a bit of reference material, connected by hypertext links, and if you give the questions the thought that I hope they deserve, and browse through the reference material, I would set aside rather more than an hour.

You can always take shortcuts if you prefer, though, by going back to the immediate contents page, or indeed the main contents page.

There are also for each section a number of true/false questions. These questions do have a right and wrong answer, and are actually quite difficult to set. Some are fairly trivial, but some require quite detailed knowledge. I use questions like this to ensure that students have covered the ground prior to tutorials. That is all such questions can do - they cannot test understanding. Nevertheless, it is obviously important that you have covered the main ground of the area. As with the other types of question, you can always skip back to the contents page. You can also cross-link into the reference material if you want further particulars on the question, or my answer to it.

Finally, there is a reference section, accessed via the main contents page. This contains references to the main cases, legislation and articles in the area, with links to further details where the workbook contains them, and additional links, as appropriate, to the resource book. There is also a bibliography. The reference section is structured in hypertext format, which is great for browsing, but probably not all that testing.

You may also care to look at the author's website containing complementary trusts (and other) materials - like the reference section, the website is structured in hypertext format, and cannot therefore be as structured as an IOLIS workbook - but it is good for browsing:

http://www.swan.ac.uk/law/staff/pntodd/index.htm

This workbook deals with informally-created trusts. By this I have in mind trusts that arise otherwise than by express declaration by the settlor - for example where the settlor's intention is inferred rather than expressed, or where a trust is imposed by operation of law.

The workbook covers quite a broad range of areas, including matrimonial property, and beneficial interests in homes shared by persons who are unmarried, and Quistclose trusts. There is also coverage of constructive trusts arising from fraud, and some aspects of resulting trusts, but a comprehensive coverage of all aspects of constructive and resulting trusts will not be found in this workbook, primarily because it would otherwise overlap with other workbooks.
Would you say that the categorisation "informally-created trusts" is arrived at on functional or conceptual grounds? Or maybe you think the categorisation can be arrived at only on some other basis, or indeed no basis?

It is Functional. There is a fuller explanation on the next page.

It would be difficult to justify the categorisation of "informally-created trusts" on conceptual grounds, since the workbook covers areas which are conceptually quite disparate.

As has already been observed, a large section of the workbook is devoted to matrimonial property, and beneficial interests in homes shared by persons who are unmarried. In what is often regarded as the leading case, Lloyds Bank plc v Rosset, Lord Bridge considered two fundamentally different situations where beneficial interests might be acquired by someone who does not have legal title. The first category, if properly analysed in trusts terms at all (explanation), requires the existence of a constructive trust. The second category can usually, but probably not always (as will become clear as we proceed), be analysed in terms of a presumed resulting trust, or category A Westdeutsche trust.

It might therefore be supposed that this workbook covers constructive trusts, and presumed resulting trusts. This is a conceptual grouping, but it is not in fact a good description of the coverage of the workbook.

First, although presumed (i.e, Westdeutsche category A) resulting trusts are covered in their entirety, not all constructive trusts are covered, since knowing receipt, and constructive trusts imposed on those in breach of fiduciary duty, are covered in other workbooks. In the second place, Quistclose trusts, which are most certainly informally created, are arguably examples of what, prior to the House of Lords decision in Westdeutsche, were called "automatic" resulting trusts. Yet the bulk of the law on "automatic" resulting trusts is covered in another workbook.

The coverage of this workbook is therefore presumed resulting trusts, constructive trusts which arise from fraud, and Quistclose trusts. In conceptual terms this is somewhat miscellaneous.

You may take the view that the grouping is no more than a leftover of what cannot be fitted in elsewhere, but I would argue that there is enough factual similarity between informally-created trusts for it to be regarded as a satisfactory functional grouping.

The recent House of Lords decision in Westdeutsche v Islington BC, which categorises resulting trusts and sets out the fundamental requirements for all trusts, impacts upon every section of this workbook. Beneficial interests in shared homes are category A Westdeutsche resulting trusts, whereas Quistclose trusts are (apparently) within category B. Westdeutsche contracts the ambit of resulting trusts, but probably expands the ambit of constructive trusts.

We have also seen how beneficial interests in shared homes may be either resulting or constructive trusts. There is therefore a clear link between every section of this workbook, which in my view justifies treating these areas together, albeit that they are conceptually rather disparate.
Resulting Trust

A resulting trust arises where although legal title is vested in a trustee, equitable title becomes vested in the settlor.The definition in A.J. Oakley, Parker & Mellows, The Modern Law of Trusts, 6th ed, Sweet & Maxwell (1994), at p. 189, is where "the beneficial interest 'results' to the settlor or his estate". (There is a similar definition at p. 27 of the same book.)

As has been explained in the introduction, the reason for including resulting trusts in this workbook is that they do not need to be expressly declared, and can arise informally. The circumstances in which they arise are quite constrained, however, and we examine those circumstances in this section. Really there are two main situations: first, where the settlor transfers legal title but does not intend also to transfer the equitable title; secondly, where the settlor transfers legal title to a trustee, intending to create a trust, but fails to do so, perhaps because the certainty or formality, or other requirements for a trust are not met.

The next page shows you a diagrammatic representation of a normal express trust (legal and equitable titles), and then you are asked some questions which enable us to build up a similar representation of a resulting trust (for comparison). After that, you will be taken back to the resulting trust menu. From there we examine the two main categories of resulting trust, and the differences made by the important recent House of Lords decision in Westdeutsche v. Islington B.C. We can also examine, in detail, the arguments advanced in the case itself.

One of the difficulties in Westdeutsche, it will be suggested, is that the two categories of resulting trust are defined much too narrowly, and unnecessarily so for the decision in the case itself. Moreover, the narrow categorisation causes problems, which do not appear to have been within Lord Browne-Wilkinson's contemplation. Obviously, if the narrow definitions in the categories formed part of the ratio of the case, we would have to treat that as conclusive, because you cannot reasonably argue with House of Lords ratio, but it is reasonable to ask whether they are part of the ratio.

So have a look at the next three pages, which enable us to build up two diagrams.



The next page will contain a similar type of diagram for resulting trusts, but first, have a go at the following questions. The questions are about what happens when a resulting trust is created.


(a) Legal title moves from the settlor to the trustee

o ü Yes No

(b) Equitable title starts with the settlor and is retained by him throughout

o û Yes ü No

(c) Equitable title is created on the transfer of legal title to the trustee, and vests in the settlor

o ü Yes No

Answer (a): YES

Legal title moves in exactly the same way as with an express trust. It is quite important to appreciate that the settlor starts off with legal title to property which is transferred. The relevance of this will become clearer when we consider the second category in Lloyds Bank plc v Rosset, not all of which cannot (for this reason) be explained in resulting trust terms.

Answer (b): NO

This was commonly accepted to be the case, and was central to the bank's argument in Westdeutsche v Islington BC. However, the retention of title idea was rejected by Lord Browne-Wilkinson, who took the view that in the absence of separation of legal and equitable titles, the notion of equitable title had no meaning, since all the rights it carried were encompassed within the legal title.


Answer (c): YES

That is indeed the consequence of the view taken by Lord Browne-Wilkinson in Westdeutsche on the retention of title point. Equitable title springs up for the first time when legal title is transferred, the definition of a resulting trust necessarily requiring that it vest in the settlor.




Categories of Resulting Trust

There are two main ways in which resulting trusts can arise. Until recently, the definitive categorisation was thought to be that of Megarry J. in Vandervell. On the basis of that categorisation, what do you think of the following?

The next page contains questions specifically on Hodgson v Marks.


(a) The categorisation is intended to be exhaustive ü Yes No

(b) The automatic category depends on the conscience of the recipient ü Yes No

(c) The presumed category depends on the conscience of the recipient Yes ü No

Answer (a): YES

I would suggest that Megarry J. envisages that these are the only varieties of resulting trust. Immediately before the description of the two categories, he says:

"Where A effectually transfers to B (or creates in his favour) any interest in any property, whether legal or equitable, a resulting trust for A may arise in two distinct classes of case."

He describes no other categories, and it may be supposed that these categories are therefore intended to be exhaustive.

Answer (b): YES

For the second (automatic) category in Vandervell, there must be a transfer on trust. It is difficult to see how this can occur without the recipient's conscience being affected. The description of this category as "automatic" resulting trusts is intended to convey that it does not depend on the intention of the transferor.

Answer (c): NO

The first category in Vandervell is not automatic, but is based on the presumed intention of the transferor. This presumption can be rebutted. However, nowhere in this categorisation is there any reference to the state of mind of the recipient.

Now there are some specific questions concerning Hodgson v Marks, where the Court of Appeal held that Mrs Hodgson had an interest under a resulting trust, but which is not really the standard type of case.

The reason for considering this case is because although it is easy to classify on the basis of the Vandervell categorisation, it is much more difficult on the basis of the narrower classification in Westdeutsche.

These questions ask you to classify Hodgson v Marks on the basis of the Vandervell categorisation.

The next page asks you about the Westdeutsche categorisation of resulting trusts.


(a) Hodgson v Marks is a presumed resulting trust ü Yes No Maybe

(b) Hodgson v Marks is an automatic resulting trust ü Yes No Maybe

Answer (a): YES

I would say that it can, at least if the transfer was truly voluntary. There is no restriction in the first category in Vandervell to any particular type of property, so it can presumably apply to land just as well as to anything else. The presumption, then, is that Mrs. Hodgson intended to transfer only the bare legal estate, and to retain equitable title.

Answer (b): YES

I would say that it can, if it is accepted that the transfer was expressed to be on trust (this would of course be a bare trust for Mrs. Hodgson). Any declaration of trust was oral, so would fail (under section 53(1)(b) of the Law of Property Act 1925), but in that event, there would be a resulting trust, for Mrs. Hodgson herself. A resulting trust of land does not need to be in writing because of s. 53(2). It might be objected that the effect of this is exactly the same as if the failed trust had been valid, but that does not seem to be a valid objection. If the trust fails, the beneficial interest must go somewhere, so where better than the provider, Mrs. Hodgson herself.

See also notes on Vandervell.

The categorisation in Westdeutsche v Islington BC differs slightly from that in Vandervell.

Westdeutsche v Islington BC [1996] 2 All ER 961 also suggests that it is not correct to talk in terms of equitable title remaining with the settlor (although previous analyses had often been in these terms). The equitable title springs up, for the first time, on the creation of the trust, it be meaningless to talk of equitable title at all prior to the creation of the trust.

Here are some questions on the Westdeutsche categorisation.

The next page invites you further to examine Hodgson v Marks.


(a) The categorisation is intended to be exhaustive ü Yes No

(b) Category B depends on the conscience of the recipient ü Yes No

(c) Category A depends on the conscience of the recipient ü Yes û No


Answer: ALL YES

This must surely be the case. Lord Browne-Wilkinson was concerned to conclude that no resulting trust arose in favour of the bank, and one of the arguments was that there are only two categories of resulting trust, and that the bank did not fall within either. This argument works only if the two categories are exhaustive.

In any case, he begins with the observation that "Under existing law a resulting trust arises in two sets of circumstances ...", and then sets out the (A) and (B) categories. After setting out the categories, he says: "Applying these conventional principles of resulting trust to the present case, the bank's claim must fail." He explains that the case falls into neither of the two categories. The clear implication is that there are only two categories, and that because the bank could not bring the case within either of them, it must fail.

This is the same as the "automatic" category in Vandervell, although Lord Browne-Wilkinson did not accept that this category was independent of the transferor's intention. However, just as in the Vandervell categorisation, the restriction to where there has been a transfer on trust implies that the conscience of the recipient must necessarily be affected.

See further Westdeutsche.

Lord Browne-Wilkinson implied that all trusts depend on the conscience of the recipient being affected, so by implication this must apply to category A resulting trusts: see the fundamentals of the trust in Westdeutsche.

Again, we have some questions specifically on Hodgson v Marks. This is much more difficult to classify within the much narrower Westdeutsche categorisation, and that is what you are now being asked to do.

We can see then that the Westdeutsche categories differ from those in Vandervell. The differences may not seem to be very great, but they can be significant, as we can see from our attempt correctly to place Hodgson v Marks into its correct category. The Westdeutsche categories, which are narrower than those in Vandervell, are examined in greater detail in the next sections.

Category A

We have seen from the introduction that this category is based on the presumed intention of the settlor, and that one of the effects of Westdeutsche is to make it difficult to explain Hodgson v Marks within this category.

Now, we are going to look in slightly more detail at category A. First, a simple question, which is whether it is in principle capable of covering at least some of the cases in the second category in Lloyds Bank plc v Rosset.

Try the question on the next page.


(a) Yes it is

(b) No it isn't

If you have a look at the categorisation in Westdeutsche, I would suggest that the language of category A is almost designed expressly to include the second category in Lloyds Bank plc v Rosset.

In addition to redefining the categories of resulting trusts, Westdeutsche also suggests an overriding conscience requirement for all trusts. That being so, have a go at the following question.

Mr. and Mrs. Omega purchased a house, legal title to which was conveyed to Mr. Omega alone, but with Mrs. Omega providing half the purchase price. Mrs. Omega intended thereby to obtain a beneficial interest in the property, but Mr. Omega thought that the money was a gift.

Assume that Omega's belief is genuine and not fanciful, and is based on a reasonable interpretation of the situation.

There are further questions on the next page, and the following page concludes this discussion.

(a) Does Mrs. Omega have an interest in the house? û Yes ü No

(b) Does your conclusion follow from the ratio of Westdeutsche? ü Yes No

Answer (a): NO

There is nothing in category A in Westdeutsche which suggests that the conscience of the recipient is relevant, and all previous cases had concentrated entirely on the state of mind of the transferor, but the conscience requirement would seem in Lord Browne-Wilkinson's view to form the basis of all trusts, including those in category A, and I do not see how Mr. Omega's conscience is affected here.

The conclusion seems very unfair on Mrs. Omega, and may not be what Lord Bridge envisaged when he clarified the second category in Lloyds Bank plc v Rosset, but seems necessarily to follow from the fundamental basis of all trusts in Westdeutsche being conscience. There is no reason why Mr. Omega's conscience should be affected, and if Lord Browne-Wilkinson is correct, then it surely follows that Mrs. Omega has no interest in the house.

There is also, of course, no common intention in this situation, but in principle, the relevant intention is Mrs. Omega's (had she intended a gift, that would be effective), but we must now assume that that is subject to the overriding requirement that the conscience of the recipient (i.e., Mr. Omega) be affected.


Answer (b): YES

I would say that the aspect of Westdeutsche which most clearly forms part of the ratio is the requirement that the conscience of the recipient be affected. The primary reason for there being no resulting trust in the case was that the local authority did not know that the contract was void, and that therefore its conscience was unaffected. If it had discovered the true position while there was still any traceable property that was still capable of being the subject matter of the trust, then surely there would have been a trust of such trust property as remained identifiable. So I would argue that the conclusion follows from the ratio of the case.

An argument the other way might be that the bank did not come within either category of resulting trust anyway, and failed on that ground, whatever view had been taken on the conscience issue. Certainly, it did not come within category B on that ground. It also looks as though it would not have come within category A, because the payment was not voluntary, and any presumption about equitable title would be rebutted by clear evidence that the bank intended it to pass to the LA.

There seem to be two problems with this.

First, consideration in Westdeutsche being void, the transfer was arguably voluntary, and in those circumstances the bank did not intend to pass equitable title. So maybe it could be argued to fall within category A, even as redefined.

Secondly and more importantly, however, it seems pretty clear that the fundamental requirement for all trusts, that equity imposes on the conscience of the legal owner, is the central reasoning in the case, the redefinition of resulting trusts being secondary. In that case, the conclusion in the last question is definitely part of the ratio.

The conclusions from the last two pages are quite startling, and must surely have been unintended by Lord Browne-Wilkinson. To some extent, the consequences are mitigated by the conclusions on the last page.

We also concluded that the ratio of the case probably demands the results on the last two pages. While it is true that House of Lords decisions are occasionally overruled, this is very rare, and I think we must conclude that the law relating to matrimonial property, and property of cohabitees, has been radically, albeit perhaps unintentionally, altered by the decision in Westdeutsche.

We have also seen that the narrowing of category A makes it more difficult to categorise Hodgson v Marks. Again, this change was probably unintended, and was quite unnecessary (since the property in Westdeutsche was money). This time, however, it is probably not part of the ratio of the case.

The next page returns you to the resulting trusts categories menu page. From there we can go on to look at some of the problems with category B.

Category B

Category B in Westdeutsche requires the property to be transferred on express trusts which fail, or do not exhaust the whole beneficial interest. The case of Re Ames' Settlement [1946] Ch. 217, relied on by the bank, was distinguished by Lord Browne-Wilkinson on this ground - property transferred on express trusts in a marriage settlement was held on resulting trust when the marriage was later rendered void ab initio. Unfortunately, however, it is less clear that other cases, which Lord Browne-Wilkinson clearly assumed would come within category B, are in fact distinguishable from Westdeutsche on the same grounds. In the following pages, we consider two such cases, Re West Sussex Constabulary, and Quistclose.

Have a go at the questions on the next page.


Here are some questions on the categorisation in Westdeutsche. We are considering category B, and in particular the two cases, Re West Sussex Constabulary, and Quistclose.

Have a go at the questions.

The next page contains further questions on category B.


(a) Lord Browne-Wilkinson described West Sussex as a category B case
o ü Yes No

(b) There were expressly declared trusts in West Sussex
o û Yes ü No

(c) Lord Browne-Wilkinson described Quistclose as a category B case
o ü Yes No

(d) There were expressly declared trusts in Quistclose
o Yes ü No

Answer (a): YES

He did so expressly, at p. 991b of the All England report. Recall the definition of category B.

Answer (b): NO

In West Sussex, there was a resulting trust of the identifiable donations only, the unidentifiable donations, and the proceeds from raffles, etc., going to the Crown as bona vacantia. However, the identifiable donations were merely given to the association, and there does not appear to have been any kind of express trust declared. The same is true of most unincorporated association cases. West Sussex ought not therefore to be within category B at all, if Lord Browne-Wilkinson's categorisation is correct.

Answer (a): YES

He did so expressly, at p. 990j of the All England report. Recall the definition of category B.


Answer (b): NO

Not if what is meant by expressly declared in that the words "on trust" were used. Moreover, we may well conclude at the end of the Quistclose section of this workbook that the terms of the primary trust were not the same as the conditions on which the money was transferred to Rolls Razor, since if they were the same it is difficult to see why the primary trust failed.

I would admit, though, that you could argue that the trust was express, but it seems to be an abuse of language to say that the property was transferred on expressly declared trusts, as it undoubtedly was in Re Ames.

Two final questions on the B categorisation in Westdeutsche:

(a) The categorisation is part of the ratio û Yes ü No

(b) The categorisation is correct (in your view) û Yes ü No

Answer: Both NO

Question (a), I can't see why. Surely all that is required is that the recipient is aware, at the time of receipt of the property, of all the relevant facts, and that he or she is not intended to have the beneficial interest himself or herself. I suppose you could describe this as a form of express trust, but both West Sussex and Quistclose differ from Ames, where there were express words of trust used.

Question (b), Not unless either West Sussex and Quistclose are wrong, or you take a very wide view of what is meant by transferring property on express trusts - a wider view, for example, than transferring them and expressly declaring the trusts.

We see, then, that category B has (like category A) been redefined so as to exclude the bank, but that this redefinition causes problems with two of the cases cited expressly by Lord Browne-Wilkinson. However, it is difficult to see that this is part of the ratio, if it is accepted that the central aspect of the case is the lack of knowledge by the local authority, and hence of any reason to impose upon its conscience.

The next page returns you to the resulting trusts categories menu page. From there, if you go up to the main resulting trusts menu, there is some further material on Westdeutsche.

Westdeutsche v Islington [1996] 2 All ER 961 (House of Lords)


We have seen that category A resulting trusts in Westdeutsche are defined more narrowly than presumed resulting trusts in Vandervell, although it is not clear that this is necessarily intentional. Category B in Westdeutsche has also been narrowly defined, perhaps in a way that excludes some cases that were previously thought to come within the category. Again, it was not clear whether this was intentional. The purpose of this section is to examine exactly which aspects of these categorisations were necessary to rebut the bank's argument that there was an automatic resulting trust in Westdeutsche, and hence clearly part of the ratio.

In Westdeutsche, the bank argued that although it had necessarily parted with legal title to the money (because once it had been paid into the local authority's mixed account it could no longer be traced at common law), it did not intend (knowing now that the contract was void) to part with the equitable title. It therefore retained the equitable title, which never left the bank. The next page is a diagrammatic representation of this argument. There then follow a number of questions on Westdeutsche.

Here are some questions about the bank's argument in Westdeutsche.

(a) The bank argued that it had started with legal and equitable title
o ü Yes No

(b) The bank argued that it had not intended to part with equitable title
i. ü Yes No

(c) The bank argued that it had therefore retained equitable title
o ü Yes No

Answer: All YES

Question (a): The conventional view was probably that where no trust existed, the owner of the property had both legal and equitable to it, the equitable title being the valuable title, giving the owner the right to enjoy the property.

Question (b): At least, not in the circumstances that now prevailed, that the contract was void ab initio. It had, of course, intended to part with both legal and equitable title to the money originally, but was then in ignorance of the true facts.

Question (c): It argued that this followed automatically from the previous answer, there being no need for the conscience of the recipient to be affected.

Some further questions on the bank's argument (you may need to refer to Re Ames):

Have a go at these.

The next pages consider the refutation of the bank's argument.


(a) The bank relied on Re Ames [1946] Ch 217 ü Yes No

(b) The bank relied on Sinclair v Brougham [1914] AC 398 ü Yes No

(c) The bank relied on Chase Manhattan Bank [1981] Ch 105 ü Yes No

Answer: All YES

Question (a): The similarities were that property was transferred on the assumption that a marriage was valid but it later turned out to be void ab initio, and the property was held on resulting trust.

Question (b): The similarities were that property was transferred on the assumption that a banking business carried on by a building society was valid but it later turned out to be ultra vires and void ab initio, and the property was held on resulting trust.

Question (c): By mistake, money was paid to the defendant bank twice (or so Goulding J. was asked to assume), and the defendant bank was required to hold the second payment on trust for the plaintiff (also a bank), although it was unaware at the time of receipt that it had already been paid.

The bank's arguments were all refuted in Westdeutsche, Lord Browne-Wilkinson taking the view that to describe the bank as having equitable title prior to the transfer was meaningless in the absence of separation of title, so that it was not correct to say that the bank had retained equitable title. Nor was a trust created later, because the LA did not know (until the trust property had been exhausted) that the contract was void, and therefore its conscience was unaffected.

The next page shows a diagrammatic representation of Lord Browne-Wilkinson's view, and then there are further questions for you to try.

Now for some questions about Lord Browne-Wilkinson's refutation of the bank's arguments (you may need to refer to Re Ames):

Have a go at these questions.

The next pages examine Lord Browne-Wilkinson's reductiones ad absurdum arguments.


(a) Re Ames was wrongly decided û Yes ü No

(b) Sinclair v Brougham was wrongly decided ü Yes No

(c) Chase Manhattan was wrongly decided û Yes ü No

(d) The LA was not a trustee because it was unaware of the facts that would affect its conscience ü Yes û No

(e) If the LA had become aware of the true facts before the account became overdrawn, it would have been trustee of the remaining sum ü Yes û No

Answer (a): NO

It was held to be correctly decided, but distinguishable on the grounds that the property had been transferred on express trusts, which was not of course the case in Westdeutsche itself. Because it was transferred on express trusts, the conscience of the recipient must necessarily have been affected.


Answer (b): YES

The case was overruled, an unusual case where the HL overruled one of its own previous decisions.

Answer (c): NO

The case was overruled, an unusual case where the HL overruled one of its own previous decisions.

Answer (d): YES

This is (in my view) the ratio of the case.

Answer (e): YES

Yes, because its conscience would then have been affected.


Lord Browne-Wilkinson advanced the following reductiones ad absurdum to show that the bank's resulting trust argument was wrong (this is just a summary):

(i) Suppose T gives £1m, and 100 shares, to R1 under a contract which both suppose valid but which is in fact void. R1 pays £1m into a mixed account.

R1 pays £50,000 from the mixed the account to R2, not for value - R2 is some kind of business - R2 goes insolvent with debts to trade creditors who have paid R2 for goods which have not been supplied.

R1 transfers the shares to R3 on an equitable charge in return for a loan.

Under the bank's argument, T takes in priority to all R2's trade creditors; R2 as volunteer takes subject to T's equitable interest, and hence becomes trustee for T; in tracing the £50,000 into the account of R2, R2 will be treated as having drawn out his own moneys first, leaving T's in the account (see Hallett's Estate). T also takes in priority to R3 (note that R3 has only equitable not legal title).

Neither R2 nor R3 know anything about T, or T's dealings with R1 - also they are not in any meaningful sense unjustly enriched at T's expense - the result is absurd - therefore the bank's argument was wrong.

(ii) If the contract had been valid T would have had purely personal rights against R1 - it was absurd that he should be so much better off just because the contract was void.

He also argued against the importation of equitable principles into commercial transactions.

The next page asks you a question about this argument, the idea being to test whether the absurd conclusion really does follow from the argument adopted by the bank.

The question is, does Lord Browne-Wilkinson's argument on the previous page prove too much?

Suppose, for example, T instead transferred the money to R1 on a Quistclose trust, but that R1 did not pay the money into a special account, but into a mixed account, as in the example. The remainder of the example is as before, but let us just concentrate on R2, and R2's trade creditors.

Surely the result is the same? Therefore, if Lord Browne-Wilkinson's reasoning here is correct, we should not use Quistclose trusts in commerce. Since it is clear that we do, Lord Browne-Wilkinson's reasoning must be incorrect; indeed, we have just proved this by another reductio ad absurdum.

The next page draws some conclusions about the categories of resulting trust.


(a) I am impressed by the above argument (that the reasoning is incorrect)

(b) I am unimpressed by this argument

Answer: (b)

There is a flaw, at least regarding R2 (though possibly not in respect of R3). Although T's title is not defeated by the transfer of the £50,000 to R2, R2 never becomes trustee on Lord Browne-Wilkinson's view, since this is one of the cases where although there is a separation of legal and equitable title (legal title in R2, equitable title in T), there is no trust. So it will not follow that in tracing the £50,000 into the account of R2, R2 will be treated as having drawn out his own moneys first (since Hallett's Estate applies only where a fiduciary mixes money with his own). So there is no problem with Lord Browne-Wilkinson's reasoning, at least in this regard.

After defining the two categories of resulting trust, Lord Browne-Wilkinson said that the case clearly did not fall within category B because there was no transfer of money to the local authority on express trusts. Nor did the case fall within category A, because any presumption of resulting trust was rebutted by clear evidence that the bank intended that the money paid should become the absolute property of the local authority. Since there are only two categories of resulting trust, the claim by the bank must fail.

We have seen in this section that Hodgson v Marks can no longer be explained as a category A case, but surely there was no need for this category to be redefined so as to exclude the case - after all, Westdeutsche itself involved the payment of money, so there was no need to exclude other forms of property. We have also seen that it is difficult to fit West Sussex or Quistclose into category B, but again the category seems to be unnecessarily narrowly defined - surely all that is required is for the recipient either to take on trust, or to be aware at the time of receipt of all the circumstances giving rise to a trust. There is surely no need to go further, and demand that the transfer is on express trusts.

The next page contains further questions on Westdeutsche.

Here are some true / false questions on resulting trusts (there are some more on the next page):

The decision of the Court of Appeal in Westdeutsche Landesbank Girozentrale v. Islington London Borough Council [1994] 4 All ER 890 was reversed in its entirety by the House of Lords: [1996] 2 All ER 961.


(a) True

(b) False

Answer: False

The answer is false - the only issue in the House of Lords was whether compound interest was payable, the LA having by then conceded that the capital had to be repaid, with simple interest.
Constructive Trust

We do not consider all constructive trusts in this workbook, only those that relate to the other areas covered in this workbook, and which are not covered in other workbooks (there are other workbooks, for example, on knowing receipt, knowing assistance, and breach of fiduciary duty). The basis of all the trusts considered here is fraud, in the sense that the conscience of the trustee is affected.

In the justification for grouping together the areas considered in this workbook in the first place, I suggested that there was no general principle applying to all constructive trusts. Here I shall suggest that constructive trusts are imposed in a few discrete areas, but that the law has not yet arrived at a general theory. It is not dissimilar to the law of tort before Donoghue v Stevenson [1932] AC 562, where liability was imposed in discrete areas, but before a general formulation of the law of negligence had been arrived at.

Like resulting trusts, constructive trusts differ from their express counterparts in that the formality provisions of s. 53(1) of the Law of Property Act 1925 are avoided. We will also consider the possibility, later in this section, that they do not need to comply with the normal rules for certainty of subject matter.

The next page returns you to the constructive trusts menu. From there, the questions are aimed at getting you to consider when constructive trusts will be imposed, but also to consider the limits of the fraud-based doctrine.


Obtaining Property by Fraud

In this section we examine the nature of equitable jurisdiction where property is obtained, or retained, by fraud. We shall start with the old case of McCormick v Grogan (1869) LR 4 HL 82, which was cited by Lord Browne-Wilkinson in support of the proposition in Westdeutsche that a thief who steals a bag of coins and mixes them with his or her own holds the stolen money as constructive trustee for the victim. McCormick v. Grogan was a leading case on secret trusts, which are probably best described as constructive trusts, and whose rationale is similar to the other trusts considered in this section.

We shall then consider Rochefoucauld v Boustead, Bannister v Bannister and Binions v Evans, all of which have various features in common with McCormick v Grogan, and the secret trusts cases which followed it. At the end of the section, we shall ask whether this line of cases in fact supports Lord Browne-Wilkinson's reasoning in Westdeutsche. We may possibly conclude, either that his reasoning represents a new departure for constructive trusts, in which case the consequences are as yet unknown, or that the stolen bag of coins problem would have been better analysed in resulting trust terms.

Have a go at the questions on the next page.


We start, then, with some questions on McCormick v Grogan, the earliest authority cited by Lord Browne-Wilkinson in Westdeutsche in support of the proposition that the thief of money holds it on constructive trust for the victim.

In McCormick v Grogan then:


(a) Was there a fraud? û Yes ü No

(b) If there had been a fraud, would Grogan have been required to return the property? û Yes ü No

(c) If there had been a fraud, would Grogan have been required to carry out his promise? ü Yes û No

Answer (a): NO

No - that is why no secret trust was imposed.

Answer (b): NO

With secret trusts, this is usually the last thing the testator intends (why?), and the court instead requires the legatee to carry out the trust.


Answer (c): YES

This is indeed the way in which secret (and half-secret trusts) work. The property is not held on resulting trust for the estate, not least because that is not what the testator would have wanted. The resulting trust solution would certainly prevent the legatee from fraudulently benefiting (unless, of course, he or she happened also to be residuary legatee), but that is not what equity regards as the fraud - it is instead failing to carry out the promise upon which the property was obtained.

In Blackwell v Blackwell [1929] AC 318, Viscount Sumner observed (regarding a half secret trust), that

"For the prevention of fraud equity fastens on the conscience of the legatee a trust, a trust, that is, which otherwise would be inoperative; in other words it makes him do what the will in itself has nothing to do with; it lets him take what the will gives him and then makes him apply it, as the Court of conscience directs, and it does so in order to give effect to wishes of the testator, which would not otherwise be effectual."

This makes it clear that in order to prevent the fraud, the court requires the legatee to give effect to the wishes of the testator. It is also fairly clear that the mechanism by which this is done is by imposing a constructive trust upon the legatee.

From the next page onwards, the discussion moves away from secret trusts, to become more general.

McCormick v Grogan was an early authority on secret trusts, which are probably constructive trusts based on fraud.

In the secret (and half-secret) trusts cases, of which Ottaway v Norman [1972] Ch 698 perhaps represents the culmination, where the nature of the fraud is discussed, the property is transferred to the legatee only because he or she agrees to hold it on certain terms. The legatee is then bound by those terms, but does not hold the property on resulting trust for the estate. It is important to observe the role of the legatee's representation, and the result.

The mechanism by which these cases operate is probably to impose a constructive trust on the legatee, the constructive trust operating independently of the will and hence avoiding requirements of the Wills Act 1837, as amended now by the the Administration of Justice Act 1982 (or indeed in Ottaway v Norman, writing requirements of LPA 1925, s. 53). The constructive trust reasoning may not be necessary, however, merely to avoid these formality provisions. In Rochefoucauld v Boustead [1897] 1 Ch. 196, the defendant purchased property expressly as trustee for the plaintiff, but claimed that the trusts, being oral, were unenforceable because s. 7 of the Statute of Frauds 1677, the precursor of LPA s. 53(1)(b). Although even then, the section contained the equivalent of s. 53(2), the Court of Appeal did not rely on this, imposing a constructive trust, but was nonetheless unimpressed by the lack of writing defence, taking the view that:

"The Statute of Frauds does not prevent proof of a fraud, and it is a fraud for a person to whom land is conveyed as a trustee, and who knows it was so conveyed, to deny the trust and claim the land as his own. Therefore a person claiming land conveyed to another may prove by parol evidence that it was so conveyed on trust for the claimant, and may obtain a declaration that the grantee is a trustee for him."

Nowhere can mention of a constructive trust be found at all in Rochefoucauld v Boustead, but the principle is surely the same as in the secret trust cases: the defendant having taken the land on terms was bound by those terms, and could not rely on a formality statute, whose purpose was to prevent fraud.

Most of the cases considered in this section are similar to secret trusts cases, but in the cases considered next, it may not be correct to say that the trustee(s) would not have obtained the property at all but for the conditions on which he agreed to take it, but he (or they) would not have got the property on such favourable terms in the absence of the conditions.

Have a go at the questions on the following pages (those on the next page are about Bannister v Bannister).

We now move away from the secret trust cases, but to cases where a similar principle appears to be operating.

The questions on this page are all about Bannister v Bannister:

Bannister v Bannister was followed in the case considered on the next page.

Anyway, in Bannister v Bannister:


(a) The plaintiff obtained the property by fraud. û Yes ü No

(b) It would have been a fraud for the plaintiff to retain the property with vacant possession ü Yes No

(c) The plaintiff was required to reconvey the property to the defendant. û Yes ü No

(d) The plaintiff was required to give effect to his promises. ü Yes No

Answer (a): NO

The evidence would have to show that he always intended to claim possession from the defendant, and this view was not accepted in the case.

Answer (b): YES

He probably would not have obtained the cottages at all but for the promise to allow the defendant to remain, and certainly would not have obtained it at the price he paid, so for him to keep the cottages but to renege on his promise would have been a fraud.

Answer (c): NO

No - that would have been an unsuitable result because:

ii. there was no reason to suppose that the defendant wanted them back, at any rate if she was also required to repay the purchase money to the plaintiff;

iii. he had not obtained the conveyance by fraud;

iv. the fraud was not so much retaining the property, but doing so without performing the promises made to the defendant.

Answer (d): YES

This was the result, as it normally is in cases of this type.


In Binions v Evans, Megaw and Stephenson L.JJ. decided the case on the basis that the original agreement (between Mrs. Evans and the trustees of the Tredegar Estate) created an equitable life tenancy, which bound the purchaser with notice. There is nothing particularly special about this reasoning, except that Megaw L.J.'s view, that the life tenancy was justified on the basis of Bannister v Bannister, looks like an unjustifiable extension of that decision, because the trustees did not obtain any property on the strength of a representation.

Lord Denning M.R.'s view is different, and much more interesting. It has also been followed, albeit without great enthusiasm, in later cases, such as Ashburn Anstalt v Arnold.

The next page asks you questions on Lord Denning M.R.'s reasoning in Binions v Evans.

Diagrammatic representation of Binions v Evans.

On Lord Denning's view in Binions v Evans (the constructive trust reasoning, rather than his views on contractual licences):


(a) The plaintiffs obtained the property cheaply on the strength of their undertaking ü Yes No

(b) The plaintiffs were required to return the property to the trustees û Yes ü No

(c) The plaintiffs were kept to the terms of their undertaking ü Yes No

Answer (a): YES

Lord Denning M.R. certainly proceeded on the assumption that they had obtained the property more cheaply, having taken expressly subject to Mrs. Evans' rights.

Answer (b): NO

This would seem on the face of it to be an absurd solution, although it would have been perfectly reasonable from Mrs. Evans' viewpoint, because she could then have enforced the original contract against the trustees of the Tredegar Estate (cf. Esso v Kingswood Motors [1974] Ch 142). But it would hardly have been reasonable, from the viewpoint of the trustees, to unravel the transaction, especially if they were also required to return the purchase price.

In none of the cases considered in this section was the property returned, nor in any of the cases was a resulting trust regarded as an appropriate solution.

Answer (c): YES

Exactly so, the mechanism being a constructive trust which was imposed upon them. Yet although this is described as a constructive trust, if you take property expressly subject to conditions, is that really so different from what you are doing when you undertake express trusteeship? Is the constructive trust in Binions v Evans really so different from the express trust in Rochefoucauld v. Boustead?

There is further consideration of Binions v Evans on the next page.


All the cases considered in this section have the following in common:

1. The trustee has either obtained property, or obtained property on more favourable terms than would otherwise have been the case, on the strength of a representation.

2. In none of the cases has the remedy had the effect of returning the property, or unravelling the transaction.

3. In each case, the trustee has been kept to the terms of the representation.

In cases such as Bannister v Bannister and Binions v Evans, the plaintiffs obtained property expressly subject to conditions, and it is difficult to see why they could not therefore be regarded as express trustees. In this case, of course, these cases do not extend the law of constructive trusteeship very much. Binions v Evans could simply be regarded as a case like Eves v Eves considered in the next section. All the elements of an express trust are present, but the fraud basis allows you to avoid the writing requirements of express trusts of land.

It is also clear from Ashburn Anstalt v Arnold that the principles in these cases are unlikely to be extended. In Bannister the plaintiff's representations were oral, in which case it might be objected that express trust reasoning would fall foul of s. 53(1)(b) of the LPA, but Rochefoucauld v Boustead shows us that equity will not all that provision to be used as a cloak for fraud, whether the trust is categorised as express or constructive.

There is, moreover, a debate about whether secret trusts might also in reality be properly categorised as express trusts - in short, therefore, none of the cases considered here support a very wide view of constructive trusts. It is also noteworthy that in each case the recipient had made (and was held to) a representation, and that in none of the cases was the recipient of the property required to return it.

None of these cases therefore justifies Lord Browne-Wilkinson's conclusion that a thief who steals a bag of coins and mixes the coins with his own holds the money on constructive trust for the victim. There, there is no representation, and the property is returned to the victim. It is difficult to see why the stolen bag of coins problem could not be decided on resulting trust principles, even consistently with the decision in Westdeutsche itself, except that an unnecessarily narrow categorisation of resulting trusts was adopted in that case. If on the other hand, constructive trusts have been extended to cover a new situation, we do not know what the extent of this new jurisprudence is.

In Stocks v Wilson [1913] 2 KB 235, a fraudster was required to return the property (or rather, the proceeds of its sale) to the victim, but the fraud was of an infant representing to a seller that he was an adult, and would pay for the goods. Obviously, he could not be held to the representation that he was an adult, and to require him to pay the price of the goods would have been for equity to enforce by the back door a contract that was for public policy reasons unenforceable. Therefore, the only way of preventing the fraudster from benefiting from his fraud was to require him to return the property (or its proceeds) to the victim. The case is unusual, however, and it is not easy to generalise from it.

In McCormick v Grogan, Bannister v Bannister and Binions v Evans, a constructive trust was imposed upon the recipient, which avoided the formality requirements of the Wills Act 1837, as amended now by the Administration of Justice Act 1982 (in McCormick) and s. 53 of the LPA in the other cases). However, you might consider whether any of these cases differ materially from Rochefoucauld v Boustead, where the defendant was held to have taken the property on terms which made him an express trustee, the writing requirement being avoided on the principle that equity will not allow a statute to be used as a cloak for fraud.

The next page returns you to the constructive trusts menu, from where you can consider a different type of fraud-based case, where (for different reasons) a representor is held to the representation he (at least, it was he in the cases we consider) has made.


Representations and Reliance

We are now going to consider other fraud-based cases where there is a representation, but the nature of the fraud is different from the previous section - the difference will become obvious as we proceed through the questions.

The questions on the next page ask you to consider Eves v Eves, which is one of the leading cases of this type of equitable fraud.

We begin with some questions on Eves v Eves:

Diagrammatic representation of legal and equitable titles in Eves v Eves.

There are more questions about this case on the next page.


Stuart Eves obtained property by fraud Yes No

Answer: NO

He had paid the entirety of the purchase price, and made no representations to the vendor, so this case differs from those in the previous section, where property was obtained on the strength of the representation, or the terms on which property was obtained were affected by the representation.



This case, and others which are within the first category in Lloyds Bank v Rosset, show that there is a different type of constructive trust case, which is based on fraud, but where the fraud would be reneging on a promise which has been relied on. As in the cases considered in the previous section, the solution is to hold the representor to the representation.

The next page returns you to the constructive trust menu page. However, this line of cases is considered again in detail in the beneficial interests in shared homes section of the workbook, under the first category in Lloyds Bank v Rosset, and again when we consider quantification.

An observation that I would make about these cases is this. In Eves v Eves, for example, all the elements of an express trust are present, with the exception of writing. As we will see in the section on beneficial interests in shared homes, Stuart Eves had done everything necessary to declare himself trustee for himself and Janet, except make the declaration in writing. The only effect of Janet's reliance was to get around the writing requirement.

Two further questions that we will need to consider at some stage are, first, if that is the correct analysis of these cases, are the cases considered in the last section, where property is transferred on the strength of a representation, really a different category of case, or are they basically the same? Secondly, is it true that reliance avoids only the writing requirement, or is there anything else it can avoid?

Floating Trust

In all the cases considered so far, the trust property has been known from the outset. In Bannister v Bannister and Binions v Evans, for example, the trust property was that which was conveyed to the plaintiffs. In Eves v Eves, the property was the bungalow purchased by the couple. In the cases in this section, the trust property is not known from the outset, but crystallises later, as for example where the trust is of all the property owned by the trustee at his or her death.

Usually a trust of this type would be void for uncertainty of subject matter, but there is now quite good authority for the existence of floating constructive trusts. Most of the cases involve mutual wills, but there are authorities outside this area of law (albeit rather inconclusive). In that case the question becomes, how far does the floating trust concept extend?

The questions on the next page and on page 3 are intended to encourage you to consider when a floating trust is needed.

Let us begin, then, with a simple mutual wills case. Let us suppose that an agreement is made between A and B, whereby they each agree that the first to die will leave everything that he or she owns to the survivor, on the condition that the survivor leaves the same property to their children on the death of the survivor. Suppose B is the survivor. B accordingly receives property under A's will. Which of the following statements, in your view, is correct?

The next page considers a variation on this arrangement.


(a) B does not become a constructive trustee û True ü False

(b) B becomes a trustee, but the property is identified at the outset (no need for a floating trust) ü True False

(c) B becomes a constructive trustee under a floating trust crystallising on his or her death True ü False

Answer (a): False

All the requirements for enforceable mutual wills are present, so there will be a trust. There is an abundance of authority that B takes the property as trustee. The precise nature of the trust is what is examined next.

Answer (b): True

We saw from the last question that there will be a trust. However, B has agreed to leave to the children only the property received under A's will. Therefore the trust property is identified as soon as B receives it, and there is no need for a floating trust. Cases such as Re Oldham [1925] Ch 75 and the land in Ottaway v Norman [1972] Ch 698 are probably best explained in these terms. No new jurisprudence is required; they are simply examples of cases where property is conveyed on the strength of a representation, and the representee is required to give effect to his or her undertaking.
Answer (c): False

As we saw in the last question, there is no need for a floating trust, because the property is identified from the outset.


Now let us change the facts slightly. This time, each agrees that he or she will leave all his or her property on his or her death to the other. Again, B is survivor, and therefore obtains all of A's property under A's will. Again, the question is which of these statements is correct?

The next page begins to draw some conclusions on the extent (if any) to which floating trusts should be recognised.


(a) B is required by contract to leave all his/her property on B's death to A's estate ü True False

(b) This contract can be enforced by A's next of kin directly û True ü False

(c) There is a trust, but no need for a floating trust û True ü False

(d) There is a floating trust ü True False

Answer (a): True

This must be correct - there is no reason why there should not be an enforceable contract here, enforceable by A's estate. The problem is that B will often in practice be the executor or administrator, in which case the contract will in practice be unenforceable.

Answer (b): False

They are not party to the contract and are not within the marriage consideration - they are not in the position of the children in Pullan v Koe [1913] 1 Ch 9, who could enforce a contract to which they were not party because they came within the marriage consideration.

Answer (c): False

Unlike the previous example, B's undertaking extends to all the property owned at his or her death, and therefore the trust property cannot be identified in advance of then.

Answer (d): True

Unlike the previous example, B's undertaking extends to all the property owned at his or her death, and therefore the trust property cannot be identified in advance of then.


A possible reason why the trust is used to enforce mutual wills is that the contractual remedies are inadequate, as we saw in the last question.

There has long been authority that the trust attaches to all the survivor's property, as long as the survivor accepts the legacy under the other's will (see, e.g., Re Hagger [1930] 2 Ch 190, but cf. Re Oldham [1925] Ch 75). This is an extension of the reasoning we have seen so far, in that the obligation extends beyond the property actually received by B. It also suggests that unlike express trusts, constructive trusts can cover future as well as existing property.

This floating trust concept originated with the Australian case of Birmingham v Renfrew, was accepted in principle in Ottaway v Norman, applied in Re Cleaver, but possibly not as part the ratio, since it was not clear whether any additional property was covered, apart from that acquired by the survivor under the will.

The authorities were not really conclusive until Re Dale, Proctor v Dale [1994] Ch 31; Morritt J applied the mutual wills doctrine where the second testator had received no benefit at all under the first testator's will, the mutual agreement having been that their children should share equally. There the trust clearly applied to the whole of the widow's estate, since the widow had received no property to which it could apply. There must have been a floating trust, crystallising on the death of the widow.

Equally, however, we know that express trusts cannot include future property: Re Ellenborough [1903] 1 Ch 697. It follows that not all trusts can float. On the next page we examine where the limits of the doctrine might lie.

One of the central difficulties is the floating trust; does it just apply to mutual wills, and if so why, or does it apply more generally to any constructive trust? If it applies more generally, what difference does this make?

It is possible to argue to limit the doctrine to mutual wills; mutual wills are a special case, in that there are difficulties with their contractual enforcement, which are not present with other types of trust.

This does not seem terribly convincing, however. The cases on mutual wills may be better regarded as an extension of the principles from such cases as Binions v Evans, where property is transferred on the strength of a representation, and the representor is held to his representation. From cases such as Lyus v Prowsa, it seems not to matter very much how closely the representation is connected with the property; the extension in mutual wills cases is that the constructive trust can attach to any property, and is not necessarily limited to the property transferred.

This however suggests that the floating trust depends upon property being transferred on the strength of a representation, but this does not seem terribly convincing either. What is so special about the transfer of property? Another way of looking at the Binions v Evans line of cases is to reason that taking property expressly subject to a promise is essentially the same as taking property on express trust. In other words, these cases are just further examples of the principle that equity will not allow a statute to be used as a cloak for fraud, or the type of trust we looked at in Eves v Eves. All the requirements for an express trust are present. As express trusts they would fall foul of the writing requirement for land, but because the representation has been relied upon the writing requirement is avoided, since otherwise it could be used to perpetrate a fraud. The extension with mutual wills is that the certainty of subject-matter requirement is also avoided.

If this is correct, it is arguable that the floating trust concept can apply to any fraud-based constructive trust. In that case the reasoning in Stokes v Anderson is correct. It would follow, if this analysis were correct, that category one Rosset trusts are also exempt from the certainty of subject-matter requirement. It would not be necessary therefore, to have identified the house before the representation is made. The constructive trust would then have two roles: to get around the formality requirement, and also to avoid the certainty of subject-matter requirement.


A General Discretion

We have seen three situations where constructive trusts are imposed, all of which are fraud-based, but where the equitable jurisdiction appears to be fairly narrowly-confined. There are also areas dealt with in other workbooks, such as where there is a breach of fiduciary duty, and a trust is imposed to prevent a fiduciary from profiting from his or her wrong, or where trust property is received by someone with the knowledge requisite for knowing receipt.

It is quite common for students to take the view that the constructive trust can give the courts a more generalised discretion to do what they conceive to be just. The purpose of this section is to examine whether this is true.

The cases of Cleaver v Mutual Reserve Fund Life Association [1892] 1 KB 147 and Re Crippen [1911] P. 108 establish that anyone obtaining property by a criminal act holds that property on constructive trust for the person from whom he or she obtained it.

(The next pages further examine the limits of the fraud doctrine currently under discussion).

(a) Yes

(b) No


Answer: (b)

The cases both relate only to felonious killing, and establish that the killer, or anyone claiming through the killer, cannot claim the property. That is justified on grounds of public policy. The person who has legal title cannot retain it, however, so holds it on trust for the estate of the person killed. This is probably correctly categorised as a Westdeutsche category B resulting trust, not a constructive trust at all.

In any case, neither case establishes anything that does not relate to felonious killing.

We have seen that equity will not allow a statute to be used as a cloak for fraud, and that constructive trusts based on fraud need not comply with formality statutes. The questions on the next pages invite you to consider what is meant by fraud in this context, and whether there is any general principle which operates against anybody who might be said to be acting in bad faith.

Here are some questions on Midland Bank Trust Co Ltd v Green, which may suggest limits to the fraud-based doctrine.

Have a go at them.

The next page draws some conclusions on the limits of the doctrine examined so far.


(a) Did the mother buy the farm in good faith? û Yes ü No

(b) Did the mother successfully rely on a statutory provision? ü Yes No

(c) Was the mother considered fraudulent? Yes ü No

(d) Did the mother obtain the property by making a representation? û Yes ü No

(e) Did the mother make any representation which was relied upon by the son? û Yes ü No

Answer (a): NO

Certainly not - the transaction was a sham at an undervalue, whose only purpose was to defeat the son's unregistered estate contract.

Answer (b): YES

Yes, she did - LPA 1925, s. 199, the effect of which was to render void against her (as purchaser) the son's unregistered land charge.

Answer (c): NO

No - Lord Wilberforce did not think it was a fraud to rely on a statutory provision. Fraud must therefore mean something different from merely acting in bad faith.

Answer (d): NO

No - she bought the property from a willing vendor (who was indeed party to the scheme), albeit at an undervalue.

Answer (a): NO

No - she made no representation to the son - and it seems that fraud within this section of the workbook requires a representation to be made.


We can perhaps sum up by saying that there seems to be no general principle operating against anybody who is acting in bad faith, and no general principle that requires anybody who acquires property by a criminal act to return it. It is true that a knowing receiver is required (in effect) to return the trust property to the beneficiary, but there a fiduciary duty has already been established.

It would seem, therefore, that Lord Browne-Wilkinson's view that a thief who mixes stolen coins with his own is required to hold them on constructive trust for his victim is not justified on the basis of previous authority, and therefore represents an extension to the field of constructive trusts. It remains to be seen how far this extension is likely to be taken.

There is, however, another possible answer to the stolen coins problem. Lord Browne-Wilkinson thought that the reasoning in Chase Manhatten Bank v Israel-British Bank [1981] Ch 105, was wrong, where Goulding J held that where a mistakenly paid a recipient bank the same amount of money twice over, the recipient bank held the second payment for the paying bank as resulting trustee, since the recipient bank was unaware of the mistake at the time of the receipt, and its conscience therefore unaffected. However, he thought that the result in the case may have been justified on the grounds that the recipient bank became a trustee once it had discovered the mistake, because then its conscience was affected. It is also clear that the LA would have become trustee in Westdeutsche itself if the decision in Hazell v Hammersmith had occurred before the account into which the money had been paid had become overdrawn. The thief could be liable on a similar principle, because he would have been aware of the facts from the outset. It probably makes little difference whether this is categorised as a resulting or a constructive trust.

Here are a few questions that lead to a conclusion with which not everyone will agree.

You may need to refer to Rochefoucauld v Boustead, Eves v Eves, LPA 53, part performance (quote from Steadman v Steadman).

The next page is a concluding question on part performance.


(a) Rochefoucauld v Boustead was an oral express trust of land ü Yes No

(b) Eves v Eves would have been decided the same way even in s. 53(2) had never been enacted. ü Yes û No

(c) The part performance doctrine is fraud-based ü Yes No

(d) The part performance doctrine does not derive its authority from any enactment ü Yes No

(e) The part performance doctrine has been abolished û Yes ü No

Answer (a): YES

It was - there is no constructive trust reasoning anywhere in the case.

Answer (b): YES

That must be right - you would simply adopt the reasoning from Rochefoucauld v Boustead.

Answer (c): YES

Certainly - see the quote from Steadman v Steadman

Answer (d): YES

Indeed it did not - it was based on the principle that equity will not allow a statute to be used as a cloak for fraud.

There was no relevant statute when the doctrine was originally developed, but s. 40(2) of the Law of Property Act 1925 recognised its existence.

Answer (e): NO

No, but s. 40 of the LPA - including s. 40(2) - was repealed by the Law of Property (Miscellaneous Provisions) Act 1989, s. 2(8).


So can we conclude that the part performance doctrine has been abolished?

(The next page draws some conclusions from this discussion.)


(a) Yes

(b) No

Answer: NO

I don't see how - if equity will not allow a statute to be used as a cloak for fraud, why should not that principle apply equally to the Law of Property (Miscellaneous Provisions) Act 1989?

Quite a startling conclusion, perhaps, with which not everyone will agree, In Seggs v Biggs (1995) 71 P & CR 120, Neill LJ observed (obiter, since on the facts, there was no act of part performance):

"I would, with respect, doubt the judge's conclusion that the doctrine of part performance has been 'abolished'. It is true that it is provided by section 2(8) of the 1989 Act, that section 40 of the Law of Property Act 1925 will cease to have effect, but the doctrine is an equitable doctrine, and it may be that in certain circumstances the doctrine could be relied on".

The next page returns you to the constructive trusts menu page.

Beneficiary Interest in Shared Homes

The question that we are (typically) addressing in this section is where A and B share a house, where legal title to the home is vested in A, and B claims an equitable title. We need to consider how B can acquire an interest, and if B does so, the basis upon which it is quantified.

The House of Lords has thrice considered beneficial interests in matrimonial and other shared property, in Pettitt v Pettitt, Gissing v Gissing and Lloyds Bank v Rosset.

Pettitt v Pettitt and Gissing v Gissing both established that the ordinary law of property applied, and that there were no special principles applying only to this area. For married (and engaged) couples, both decisions have been overturned by legislation, but the general law continues to apply where parties are unmarried, and in all other circumstances where the MPPA 1970, s. 37 and the MCA 1973, ss. 24-25 do not apply. It follows (in my view) that the principles considered in the resulting and constructive trusts sections of this workbook apply here also.

In the 70s and 80s the law was developed by a number of Court of Appeal decisions, Gissing v Gissing being applied in Burns v Burns, where as with the earlier case, no beneficial interest was acquired. There appeared, however, to be a different type of case, exemplified by Eves v Eves and Grant v Edwards, to which different principles applied.

The existence of two categories of case was conclusively established by the third House of Lords decision, in Lloyds Bank plc v Rosset, which ought probably now to be regarded as the leading case. The first category involves discussions between the parties, such as occurred in Eves v Eves and Grant v Edwards. In the second category the interest is inferred from the contributions alone, as in Gissing v Gissing and Burns v Burns.

There are some general questions on the next page.

Here are some general questions on the main House of Lords cases on beneficial interests on matrimonial homes. It is important to establish how far the House of Lords cases take us, and also what issues they leave open. You may also wish to look at:

Pettitt v Pettitt;
Gissing v Gissing;
Lloyds Bank plc v Rosset, which is the leading recent authority on beneficial interests in shared homes;
the first and second categories from that case;

The questions on the next page relate specifically to the matrimonial legislation - which does not always apply, so there is still plenty of scope for the general equitable principles.


(a) The two categories in Lloyds Bank v Rosset are exhaustive ü Yes No

(b) Lord Bridge's intention in Lloyds Bank v Rosset was to restate rather than change the law ü Yes û No

(c) Quantification was not an issue in Lloyds Bank v Rosset ü Yes No

(d) Quantification was not an issue in Gissing v Gissing ü Yes No

(e) Quantification was not an issue in Pettitt v Pettitt ü Yes û No

Answer: All YES

(a) In Lloyds Bank v Rosset, (in this passage) Lord Bridge clearly distinguished between two types of case. It is fairly clear that he had in mind an exhaustive categorisation.

(b) In Lloyds Bank v Rosset [1991] 1 AC 107, the House of Lords clarified what had become a confused area of law. However, I would not say that there was any intention to alter the law. If Lloyds Bank plc v Rosset has not changed the law, it follows that authorities prior to the case continue to have as great an authority now as they have ever had. This will become relevant when we consider precisely what types of contribution and conduct fall within the second category.

(c) Only acquisition was at issue - because Mrs. Rosset did not acquire an interest, there was no need to quantify it. It may be that different principles apply to quantification: see, e.g., Midland Bank plc v. Cooke.

(d) Like Mrs Rosset, Mrs Gissing had no interest, so there was no need to quantify it.

(e) This time the claim was brought by the husband, who failed to acquire any interest at all. It follows that in none of the HL cases was quantification an issue.

Some general questions. Further references you may care to consider are:

Matrimonial Proceedings and Property Act 1970, s. 37
Matrimonial Causes Act 1973.

The next page returns you to the menu, from where we can consider the two categories in Lloyds Bank v Rosset in more detail.


1. MCA 1973, ss. 24-25, applies only where a court grants a decree of divorce, a decree of nullity of marriage or a decree of judicial separation. ü Yes No

2. Sec 37 of the MPPA 1970 can apply to couples who are neither married nor engaged. û Yes ü No

3. Where interests in matrimonial property are at issue, the provisions of either the MCA 1973 or the MPPA 1970 always apply. û Yes ü No

Answer (a): YES

This is correct - see s. 24.

Answer (b): NO

It can only apply to these, engaged couples being included by virtue of s. 2(1) of the Law Reform (Miscellaneous Provisions) Act 1970.

Answer (c): NO

These statutes only apply to improvements, and where the marriage has broken down. Where the dispute is between husband and wife, the marriage will often have broken down, but there is no reason why this should be so where third parties are involved - good examples of where ordinary equitable reasoning applied are Williams & Glyn's Bank v Boland and Lloyds Bank plc v Rosset itself.


Rosset Category 1

The first category differs from the second mainly in that there are discussions between the parties, and usually (but I shall suggest not always) it is also necessary for the party claiming the interest to have relied on the discussions. The contributions made by the party claiming the interest have a totally different function from under the second category.

Lord Bridge clearly thought that Eves v Eves and Grant v Edwards were examples of this type of case (i.e., they would now be categorised within the first category), whereas Burns v Burns is an example of the second category. It is clear from Springette v Defoe that discussions are required, but the conceptual basis of the category was not made clear in Rosset. It was not, for example, made clear whether the interest arose by constructive trust or proprietary estoppel, and whether these were regarded as equivalent or not. That is an issue that we can examine through the questions.

Another issue that we will need to consider is the basis upon which interests under the first category are quantified. Usually this does not cause any particular problems, but it is necessary to stop and think a little about Eves v Eves and Drake v Whipp.

Have a go at the questions on the following pages. The questions on the next page ask you about the role of the discussions necessary under this category.

In Lloyds Bank v Rosset, Lord Bridge insisted for the first category on an "agreement, arrangement or understanding" reached between the parties "that the property is to be shared beneficially". He then said that "the finding of an agreement or arrangement to share in this sense can only, I think, be based on evidence of express discussions between the partners, however imperfectly remembered and however imprecise their terms may have been."

What would you say is the role of the discussions?

Have a go at these questions.

The next page asks you questions about the role of detrimental reliance.


(a) A declaration of trust

(b) A representation for estoppel purposes

(c) A contract

(d) To found a Bannister v Banniser type of fraud


Answer: (a) & (b)


Lord Bridge also required that it is also necessary under the first category "for the partner asserting a claim to a beneficial interest against the partner entitled to the legal estate to show that he or she has acted to his or her detriment or significantly altered his or her position in reliance on the agreement".

What in your view is the function of the detrimental reliance?

(a) To quantify the interest

(b) To get around the writing requirement of s. 53(1)(b) of the Law of Property Act

(c) To found an estoppel

(d) To establish the interest in the first place

Answer: B & C

Here are some general questions on category 1, which can be answered either true or false.

Links you may find useful for the first five questions (which all relate in some way to the first category) are:

Grant v Edwards; Eves v Eves; Lloyds Bank plc v Rosset; Drake v Whipp.

For Lloyds Bank plc v Carrick, see the author's casenote at http://www.ncl.ac.uk/~nlawwww/1996/issue4/todd4.html .

Have a go at the questions here and on the next page.


(a) Grant v Edwards [1986] Ch 638 is a first category Rosset case. ü True False

(b) Eves v Eves [1975] 1 WLR 1338 is a first category Rosset case. ü True False

(c) Mrs Carrick had an interest under the first Rosset category in Lloyds Bank plc v Carrick [1996] 2 FCR 771. û True ü False

(d) Linda Grant was awarded a half share. ü True False

(e) The woman was held entitled to a 19.4% share in Drake v Whipp [1996] 1 FLR 826. û True ü False

Answer (a): True


Answer (b): True


Answer (c): False


Answer (d): True


Answer (e): False


Now for some questions on the need for, and relevance of, discussions between the parties (remember that Lord Bridge in Lloyds Bank v Rosset regarded evidence of discussions as essential for claims under this category). Further references you may care to consider are:

Springette v Defoe; Ungurian v Lesnoff; Grant v Edwards; Hammond v Mitchell.


Springette v Defoe [1992] Fam Law 489 depended on the discussions about beneficial interest which had taken place. True False


Answer: False

There had been no discussions, and the decision depended on that. The shares therefore depended on the respective contributions of the parties, and the case is difficult to reconcile with Midland Bank plc v Cooke.


Rosset Category 2

The following issues arise under the second category.

1. What comes within it? Lord Bridge's view in Rosset is different from Burns v Burns, but the tenor of Lord Bridge's speech is that he was intending to clarify, rather than alter the law.

2. What about improvers? Lord Bridge does not mention them, but the possibility that they might obtain an interest was left open in Pettitt v Pettitt.

3. What is the conceptual basis of the category? Most of the cases can be explained as resulting trusts, but the discount cases, such as Springette v Defoe, cannot (or at least cannot easily).

4. How do we quantify interests? In principle only the contributions themselves ought to be relevant, but in that case how can Midland Bank plc v. Cooke explained?

Quantification and improvements are considered elsewhere within the workbook. The next two pages contain questions on what is in this category.

In Burns v Burns, Fox LJ said, in relation to a case where there had been no discussions between the parties about beneficial interests (at p. 327) that:

"What is needed, I think, is evidence of a payment or payments by the plaintiff which it can be inferred was referable to the acquisition of the house. ... If there is a substantial contribution by the woman to the family expenses, and the house was purchased on a mortgage, her contribution is, indirectly, referable to the acquisition of the house since, in one way or another, it enables the family to pay the mortgage instalments. Thus, a payment could be said to be referable to the acquisition of the house if, for example, the payer either (a) pays part of the purchase price or (b) contributes regularly to the mortgage instalments or (c) pays off part of the mortgage or (d) makes a substantial financial contribution to the family expenses so as to enable the mortgage instalments to be paid."

Have a go at the following questions about this passage.

Like this page, the questions on the next page are about what is in the second category.


(a) Lord Bridge would include headings (a) to (c) in the second category ü Yes No

(b) Lord Bridge would include heading (d) in the second category û Yes ü No

(c) Lord Bridge intended to change the law and exclude heading (d) û Yes ü No

(d) We must assume that Fox LJ's views in Burns are wrong û Yes ü No

Answer (a): YES

Answer (b): NO

Answer (c): NO

Answer (d): NO

Answer (e): NO

One way of resolving the issues considered on the previous page would be to consider whether heading (d) is out of line with what we know to be in the second category. For example, if everything else were explicable as resulting trusts, but heading (d) was not, that might suggest that in principle, heading (d) ought to be excluded.

Note that the headings, taken from Fox LJ's judgment in Burns v Burns are where the payer:

(a) pays part of the purchase price or
(b) contributes regularly to the mortgage instalments or
(c) pays off part of the mortgage or
(d) makes a substantial financial contribution to the family expenses so as to enable the mortgage instalments to be paid.

You may also need to refer to Springette v Defoe.

The next page sets out some conclusions to this discussion.


(a) Heading (a) is explicable in resulting trusts terms ü Yes No

(b) Heading (d) is explicable in resulting trust terms û Yes ü No

(c) Headings (b) and (c) are explicable in resulting trust terms û Yes ü No

(d) The discount in Springette v Defoe is explicable in resulting trust terms û Yes ü No

(e) We can clearly conclude that heading (d) is not explicable on the same basis as the rest of the second category Yes No

Answer (a): YES

Answer (b): NO

Answer (c): NO

Answer (d): NO

Answer (e): NO

We see then, that in principle, there is no reason why Burns v Burns heading (d) contributions should not come within this category.

We also see that the second category cannot be explained entirely in terms of resulting trusts.

Elsewhere in the workbook we consider whether the second category might also encompass improvements, and also how interests under this head are quantified.


Specific Issues – Presumption of Advancement

The resulting trust analysis of the second category in Rosset depends on there being no presumption of advancement. It is clear from Pettitt v Pettitt that any presumption will be very weak in this area of law, and in no recent cases has the presumption been decisive (see also McGrath v Wallis). However, if even a weak presumption operated it could be decisive in cases such as Tinsley v Milligan, where a fraud has been perpetrated - if in that case there had been a presumption of advancement from Milligan to Tinsley, (e.g., perhaps if Tinsley had been Milligan's daughter), then it is difficult to see how Milligan would have been able to rebut it, without pleading her fraud. As there was no presumption of advancement, Milligan could rely on the presumption of resulting trust, and did not need to plead her fraud.

See also Tribe v Tribe [1995] 4 All ER 236, where the presumption was rebutted, but no fraud had actually been perpetrated.

The next page asks you some questions on the presumption of advancement.

These are true/false questions, which in principle can be answered right or wrong. They are quite useful for ensuring that the basic groundwork has been adequately covered - if you can answer these, you will be better equipped to deal with more interesting and difficult questions.

Point of true/false questions

Some of these questions are themselves quite difficult, though, and you may want to jump to links elsewhere and then return to them later.

Further references you may need are:

Tribe v Tribe; McGrath v Wallis.

Have a go at the questions.

There are more questions on the presumption of advancement on the next page.


(a) The presumption of advancement from a mother to her children, if any, is less strong than that from a father to his children. ü True False

(b) The presumption of advancement was rebutted in Tribe v Tribe [1995] 4 All ER 236. ü True False

(c) The fraud had been successfully perpetrated in Tribe v Tribe [1995] 4 All ER 236. û True ü False

(d) The presumption of advancement was rebutted in McGrath v Wallis [1995] 2 FLR 114. ü True False

(e) The presumption of advancement was described as a judicial instrument of last resort in McGrath v Wallis [1995] 2 FLR 114. True False


Answer (a): True


Answer (b): True


Answer (c): False


Answer (d): True

Answer (e): True



In McGrath v Wallis [1995] 2 FLR 114, Nourse L.J. based his views on the presumption of advancement on the decision of the House of Lords in Pettitt v Pettitt [1970] AC 777.

Have a go at the question.


(a) True

(b) False

Answer: True

Special Issues: Improvements

Lord Bridge's description of the second category in Lloyds Bank v Rosset does not include the case where the person claiming the beneficial interest has improved the property, but not otherwise contributed to it. Nor would it be easy to justify including them on resulting trust principles, because the person claiming the beneficial interest has not transferred legal title to anything. Moreover, cases such as Thomas v Fuller-Brown suggest that improvements will not generally give rise to a beneficial interest in property.

However, the issue was left open in Pettitt v Pettitt, and other cases, such as the discount cases, which are clearly within the second category, also cannot easily be explained on resulting trust principles. So the issue is still open, I would suggest.

Another way of taking improvements into account might be under the first category - see Passee v Passee and Drake v Whipp, both of which may be explicable on a similar basis.

A full discussion of these issues can be found in Nicola Glover and Paul Todd: The Myth of Common Intention (1996) 16 L.S. 325.

The next page returns you to the beneficial interests sub-menu page.

Quantification Issues


Under the second category, the contributions determine both the existence of the interest and its extent (after all, there being no discussions, there is nothing else from which it can be determined). This is in any case clear from cases such as Springette v Defoe and Huntingford v Hobbs.

There is a problem with Midland Bank plc v Cooke, but if Midland Bank plc v Cooke is right then Springette v Defoe and Huntingford v Hobbs are probably wrong, and the law that is operating here is different from that which operates elsewhere in the law of trusts, contrary to the views expressed in the House of Lords in Pettitt v Pettitt and Gissing v Gissing. Alternatively, I would suggest, Midland Bank plc v Cooke is wrong.

Under the first category, the contributions ought not to be relevant to the quantification, which should therefore be determined from the representations. Sometimes, as in Hammond v Mitchell, there is a clear statement that the trust will be a half share, but usually the representation is silent, in which case the courts usually infer a half share, in the absence of evidence to the contrary.

There is a problem with Eves v Eves, where a quarter share was awarded, and with Drake v Whipp, where the share was 30 per cent. Eves is very difficult to justify (but the principles may not have been fully developed by the time of that case), but in Drake v Whipp, it seems that the parties initially agreed that the share was to be determined by her contributions. There seems to be no obvious reason why the parties should not do this, subject to certainty of subject matter requirements being satisfied.

Have a go at the questions on the next few pages. The questions on the next page ask how far this issue has been determined by the House of Lords, and also on the extent to which the courts have a discretion to determine how much they award.


Have a look at the analysis of what is now the first category in Lloyds Bank v Rosset, by Lord Diplock in Gissing v Gissing, and by Lord Bridge in Rosset itself.

Then have a go at these questions.

The next page asks you to consider how to quantify, where the discussion does not expressly make the quantification clear.


(a) The second paragraph in the Diplock extract appears to give the courts a wide-ranging and general discretion ü Yes No

(b) The entire Diplock extract appears to give the courts a wide-ranging and general discretion û Yes ü No

(c) Detrimental reliance is relevant to quantification on the Diplock analysis û Yes ü No

Answer: (a): YES (b): NO, (c) : NO

Would you say that the decision in Stokes v Anderson fits more comfortably into the first or second category in Lloyds Bank plc v Rosset?

The next page asks you some more detailed questions on the quantification in Stokes v Anderson.

(a) First category

(b) Second category

Answer: (a) First Category
In Huntingford v Hobbs [1992] Fam. Law 437, the eventual decision depended (in the view of Sir Christopher Slade) on the contributions of the parties to the purchase price and the mortgage loan.


(a) True

(b) False

Answer: (a) True


Quistclose Trust

If you make an unsecured loan, your position is very precarious should the borrower go into liquidation before repaying it. The debt action, being a personal action, is generally useless against someone who is bankrupt.

If on the other hand, you can retain the beneficial interest in the money loaned, you are protected against the bankruptcy of the borrower - because the money is your property, it never becomes part of the bankrupt's assets, to be shared among his or her creditors. Of course, you need to be able to identify the money, which might be difficult if it has become mixed with the borrower's own money. It is also true, of course, that to the extent that you gain from the situation the other creditors lose, so it is unsurprising that liquidators and trustees in bankruptcy try to resist the implication of a trust in your favour.

Quite often, trusts is this situation are made expressly, a good example being the trust receipt used by banks financing international sale transactions. Quistclose trusts arise without any express declaration of trust, however, or at least without any express words of trust, and can therefore reasonably be described as informal trusts. They are analysed in this part of the workbook.

The next page asks you about which general principles applicable to the law of trusts might be relevant to the Quistclose trust. You may need to look at some of the pop-up links to find sufficient information to attempt them.

Now for a few basic questions on the Quistclose case itself. If you are familiar with this material, you might find these rather easy, in which case it is probably best to proceed directly to the next page, where the questions are slightly more difficult.

(a) Quistclose was a unanimous decision of the House of Lords. ü True False

(b) There was a trust because clear words of trust were used. û True ü False

(c) The money was paid over for a specific purpose. ü True False

(d) The money was paid into a separate account at the bank. ü True False

(e) The bank was aware of the basis upon which the money had been paid. ü True False

Answer: (a) True

Answer: (b) False

Answer: (c) True

Answer: (d) True

Answer: (e) True


Here are some more basic, but slightly more difficult questions on the trust in Barclays Bank v Quistclose Investments Ltd.

Have a go at these.

The question on the next page is more open to argument.


(a) Lord Wilberforce's analysis in Quistclose was that only on the failure of the primary trust (for the shareholders) did a trust arise in Quistclose's favour. ü True False

(b) From Lord Wilberforce's speech alone, it can be deduced that the principles of Quistclose apply only where the money is paid into a special account. û True ü False

Answer: (a) True, (b) False


Here are some general questions on Rowan v Dann

Have a go at these.

The questions on the next page are specifically on Re Kayford.


(a) Scott L.J. took the view in Rowan v Dann that a trust arose in favour of Mr. Rowan by reason of the failure of the joint venture project. ü True False

(b) Scott L.J. took the view in Rowan v Dann that his reasoning was consistent with the Quistclose line of authorities. ü True False

Answer: Both TRUE


Relevance of Separate Account

We saw in the basic issues that the money was paid into a separate account in Barclays Bank in Quistclose, but there is nothing in Lord Wilberforce's speech which expressly states that this is necessary. There now follow a number of questions on the relevance (or otherwise) of the money being paid into a separate account. It will be seen (I suggest) that the authorities do not clearly state whether or not a separate account is required, but it is (I suggest) a matter that is relatively easy to work out from principle.

It will also be seen that there are two issues that need to be kept separate. First, is there a trust at all? This is surely simply a matter of whether sufficient certainty of intention has been shown to create one. Secondly, where the recipient is bankrupt, the lender will need to identify property in his or her hands - it is not sufficient simply that a trust exists.

The special account may be relevant to both these issues.

The next page contains some questions on Quistclose.

The following questions are based on Barclays Bank v Quistclose itself.

Have a go at them or on the next page are some cases on Re EVTR).


(a) The money was paid into a separate account in Barclays Bank ü Yes No

(b) Lord Wilberforce says that a separate account is a requirement of a Quistclose trust û Yes ü No

(c) Rolls Razor had gone into liquidation ü Yes No

Answer: (a) YES, (b): NO, (c): YES

Here are some more questions on Re EVTR [1987] B.C.L.C. 646 :

There are more EVTR questions on the next page. Have a go at these questions.


(a) Was the money placed in a separate account in the solicitors? ü Yes No

(b) Was the issue of the separate account discussed in the Court of Appeal? û Yes ü No

Answer: (a) YES, (b): NO


The conclusion from the last page suggests that the Court of Appeal held that a separate account would be necessary.However, the case may not have depended on that.

Have a go at the following question. The question is:

Would you say that the decision in Anglo Corporation v Peacock depended on:


(a) the failure of AG to accept a condition that the money would be used for a specific purpose (purchase of computers)? ü Yes No

(b) a general requirement that the money be placed into a special account? û Yes ü No

(c) failure to rebut the inference that AG could mix the money with its general funds? ü Yes No

Answer: (a) YES, (b): NO, (c): YES

What would be your view if AG had become aware that the money had been intended by Anglo to be paid over for a specific purpose while some or all of it was still identifiable in equity (for example, if the general account had remained above a certain level for the whole time since the payment had been made)? (refer back to facts of case, if necessary, to remind you of who the parties are)

Have a go at the question.


(a) There would be a Quistclose trust of what was identifiable

(b) There would still be no Quistclose trust

Answer: B

I would suggest the following conclusions are possible on the issue of the separate account:

1. The authorities are inconclusive as to whether it is required;

2. It is difficult to see why it should be required simply to establish a personal action against the recipient (but the inference must be rebutted that the recipient is entitled to pay it into his or her general funds, and the recipient must accept the property on the basis of the conditions attached to it);

3. Where the trustee is bankrupt a personal claim will not suffice, and it is necessary to identify trust property in his hands. A separate account will make this easier, but would not seem to be absolutely necessary in all cases, especially as the principles in Hallett's Estate will normally accord preferential treatment to the Quistclose lender.

The next page arrow takes you back to the Quistclose menu.
Who Can Enforce?

One of the most difficult issues in this area of law, addressed by P.J. Millett Q.C., is who can enforce the primary trust in a Quistclose case (there is no particular difficulty about enforcement of the secondary trust). The following questions are intended to get you to address this question.

You may care to look at a diagrammatic view of the primary and secondary trusts in Quistclose, based on Lord Wilberforce's analysis in the case. You are then asked a number of questions about this, and then (from page 3) questions which take you on to the Carreras Rothmans decision, for which (at least in Millett's view) a similar representation would be entirely inappropriate. An examination of Millett's views begins on page 12.

We can now draw some concluding remarks on the question of enforcement of the primary trust.

Lord Wilberforce's analysis in Quistclose is clear and consistent, as is Peter Gibson J's in CR, but neither is consistent with the decision in Carreras Rothmans.

Millett was forced to adopt reasoning in CR that was extremely tortuous, in order to counter the arguments that the liquidator would have put had Millett used simpler arguments. Millett ends up explaining all the decisions that he mentions, including CR itself, but his reasoning accords with that of neither Lord Wilberforce nor Peter Gibson J. It also appears to allow the lender wide powers of revocation, whether or not the primary purpose has failed.

Let us suppose (for the sake of argument) that we decide that we prefer the conventional Wilberforce analysis. We must overcome two obstacles: first, that it cannot explain the decision in CR, and secondly, why the primary trust had failed in Quistclose itself.

As to the first obstacle, CR is a first instance decision which has not been followed. The decision in the case is inconsistent with House of Lords reasoning in Quistclose, and with the views expressed by the judge himself (in CR). The result (in my opinion) was unfair, giving CR priority over all other creditors, and I would suggest that the case was wrongly decided.

That leaves us with the problem of explaining why the primary trust had failed in Quistclose itself. If the trust was to pay the dividends, this could still be done, albeit that it might no longer suit the settlor's purpose. But was it a trust to pay the dividends? There were no words of trust, and the existence of the trust was inferred from the specific purpose and the separation of the money from Rolls Razor's general assets. Since the trust is inferred from the circumstances, why should not its terms also be inferred from the circumstances? After all, the settlor's purpose was clear enough. If that is possible, the terms of the trust may not have been to pay the dividends, in which case it may be possible to conclude that the primary trust had failed.


Resulting or Express Trust?

The questions on this and the next page address the issue of the classification of Quistclose trusts. This could matter if anyone attempted to set up a Quistclose trust of land, because unless they are either resulting or constructive trusts, trusts of land have to be in writing.


(a) The trust in Re EVTR was described as a resulting trust. ü Yes No

(b) The CA were describing the secondary trust in EVTR. ü Yes No

(c) The trust in Rowan v Dann was described as a resulting trust. ü Yes No

(d) The CA were describing the secondary trust in Rowan v Dann. ü Yes No

(e) The entire arrangement in Rowan v Dann was held to amount to an oral trust of land. ü Yes No

Answer: All YES

Some more questions examining the correct classification of Quistclose trusts.

The next page is the Quistclose menu page.


(a) On Millett's view the primary trust is a resulting trust. ü Yes No

(b) On Lord Wilberforce's view the primary trust is a resulting trust. û Yes ü No

(c) In the CA's view, the primary trust in Rowan v Dann was a resulting trust. û Yes ü No

(d) The primary trust in Rowan v Dann therefore should have fallen foul of LPA s. 53(1)(b). ü Yes No

Answer: (a): YES, (b): NO, (c): NO, (d): YES