Waikato Law Review
|
|
Home
| Databases
| WorldLII
| Search
| Feedback
Waikato Law Review |
|
LORD BROWNE-WILKINSON’S “IDENTIFIABLE
TRUST PROPERTY” PRINCIPLE
In Westdeutsche Landesbank Girozentrale v Islington Borough Council,[1] Lord Browne-Wilkinson, before embarking on a detailed examination of the plaintiff Bank’s arguments, conveniently identified “[t]he relevant principles of trust law”. These principles, described by His Lordship as “uncontroversial”, “basic” and “propositions fundamental to the law of trusts”, apply to express, implied and constructive trusts alike. They can be summarised as follows:
(i) equity operates on the conscience of the owner of the legal interest;
(ii) the owner of the legal interest cannot be a trustee of the trust property until aware of the facts alleged to affect his conscience;
(iii) in order to establish a trust there must be identifiable trust property (unless dishonest assistance is involved).[2]
(iv) once the trust is established, a trust beneficiary has an equitable proprietary interest in the trust property enforceable against subsequent holders other than the bona fide purchaser of the legal interest.
While considerable academic interest has been excited by the second principle, which appears to make an addition to the traditionally accepted requirements for a trust to arise, this article will examine the third principle (and incidentally suggest that the second principle forms part of the third principle correctly understood). It will focus on the words “identifiable trust property” to ascertain what His Lordship meant by this phrase.
The need to establish the exact meaning of these words arises from the fact that “identifiable” is used in more than one sense in the law of trusts and there are passages in His Lordship’s judgment in which it is not readily apparent which usage he has in mind. Accordingly, part II of this article outlines the different usages of “identifiable”, along with two other matters[3] which similarly provide a background context to His Lordship’s formulation of his third principle. This will be followed in part III by a summary of the facts in Westdeutsche and an analysis of Lord Browne-Wilkinson’s application of the third principle in his judgment. It will be suggested that Lord Browne-Wilkinson’s identifiable trust property principle goes beyond simply requiring “defined” property or semantic certainty of subject matter to establish a trust, and focuses on the issue of the proper constitution of trusts. Part IV will discuss the mechanics of identifying property so as to constitute the trust. A parallel will be drawn between identification necessary to transfer equitable title and ascertainment necessary to transfer legal property in sale of goods cases. It will be suggested that in some cases identification requires appropriation.
It should be noted that Gallen J in MacIntosh v Fortex Group Ltd[4] specifically adopted and applied His Lordship’s third principle but no reference was made to it on appeal. The joint judgment of the Court of Appeal in that case did require “a separately identifiable fund” as a prerequisite of an institutional constructive trust.[5]
A. The Role of Property within the Concept of a Trust
In a passage in Westdeutsche[6] that will be analysed in further detail below, His Lordship was at pains to point out that property is the only proper subject matter of a trust and that “[a] trust can only arise where there is defined trust property”. (emphasis added)
As Professor Hayton puts it:
At the core of the trust concept is a duty of confidence imposed upon a trustee in respect of particular property and positively enforceable in a Court of Equity by a person.[7] (emphasis added)
According to Oakley, the inherent nature of a trust is a relationship in respect of property, under which one person, known as a trustee, is obliged to deal with property vested in him for the benefit of another person, known as a beneficiary.[8] Ford and Lee define a trust as an obligation enforceable in equity which rests on a person (the trustee) as owner of some specific property (the trust property) to deal with that property for the benefit of another person (the beneficiary) or for the advancement of certain purposes.[9]
Clearly then, property in the form of “defined”, “specified” or “particular” property is an essential ingredient of a trust. However, it is not a sufficient ingredient. Before a trust can arise, the property that is to be its subject matter must be linked with the trustee via a binding trust obligation. When the property is so linked it is described as being “impressed” with the trust, and the trust can then be said to be established or completely constituted.
B. The Need for a Trust to be Properly Established or Constituted
The focus on the meaning of “identifiable trust property” should not obscure the fact that, in formulating his third principle, His Lordship was considering what was necessary “in order to establish a trust”. His fourth principle also opens with a reference to the establishment of a trust.
There are two methods of establishing an express inter vivos trust. The settlor can make a declaration of trust in respect of the property (in which case she will remain the legal owner of the property and hold it on trust for the beneficiaries). Alternatively, the settlor can relinquish all interest in the property by conveying it to trustees and declaring the trusts on which it is to be held (in which case the trustees will become the legal owners of the property and hold it on trust for the beneficiaries). Whichever method is used, the declaration must refer to specific, particular or, to adopt Lord Browne-Wilkinson’s terminology, “defined” property and, to be enforceable, must comply with any applicable formality requirements of section 49 A of the Property Law Act 1952. Once that property is impressed with the trust obligation it becomes trust property, and the trust is spoken of as being “properly” or “completely” constituted. It is “a complete and perfect trust”.[10]
The necessity for, and mechanics of, impressing property with the trust obligation in order to establish a valid express trust are well understood and fully set out in Foreman v Hazard:
Property may be impressed with an express trust in one of two ways: either by declaration of trust which involves a change in equitable ownership but not in legal title, or by a transfer of property to be held in trust by the transferee which involves a change in legal title and usually too in equitable ownership. No particular form of words is required but in whatever way the intention is expressed the three certainties must be satisfied: certainty of intention to create a trust - that a trust is definitely intended; certainty of the subject matter of the trust - that specified property is to be bound by the trust and certainty of objects, that is of the persons intended to have the benefit of the trust. In addition, the object of the trust must be lawful and any formalities required by law for constituting the trust must have been complied with.[11]
While it can be seen from the above that the mechanics of “establishing” an express trust are explicit, straightforward and well understood, the same cannot be said in relation to the other two types of trust. It is not usual to speak of “establishing” an implied or constructive trust. Despite this it is nonetheless clear that, when Lord Browne-Wilkinson referred to “establishing” a trust in his third principle, he definitely had constructive trusts in mind.[12] His first and second principles refer specifically to express, implied (resulting) and constructive trusts and it is reasonable to assume that the remaining principles stated in terms of “a trust” are intended to apply to all types of trust equally.
To discover what Lord Browne-Wilkinson had in mind when he spoke of establishing a trust, particularly in relation to constructive trusts, it is helpful to consider Oakley’s basic definition of the three types of trust:
An “express” trust is said to arise where the settlor expressly creates a relationship of trustee and beneficiary; a “resulting” trust is said to arise where the settlor carries out some other transaction from which the court infers a relationship of trustee and beneficiary; and ...a “constructive” trust is said to arise where the court imposes upon certain persons a relationship of trustee and beneficiary. In other words, an express trust arises out of the intentional creation of the relationship, a resulting trust arises out of some other intentional act of the settlor and a constructive trust arises totally independently of the intention of anyone.[13]
Replacing Oakley’s notion of a trust “arising” with that of Lord Browne-Wilkinson’s notion of a trust being “established”, it is clear that whereas express and resulting trusts are established by the intentional[14] act of the settlor, constructive trusts are established or imposed by the court in response to the constructive trustee’s conduct. As Oakley puts it:
Constructive trusts arise by operation of law. Unlike all other trusts, a constructive trust is imposed by the court as a result of the conduct of the trustee and therefore arises quite independently of the intention of any of the parties.[15]
The conduct which triggers the imposition of the constructive trust has to be unconscionable. However, the relevant conduct does not exist in a vacuum. It is unconscionable conduct in relation to specific property that gives rise to an institutional, as opposed to a remedial, constructive trust.[16]
As Oakley points out, although most constructive trusts are imposed because of the circumstances in which the property that forms the subject matter of the trust has been acquired, there are cases in which the relevant conduct occurs after the property has been acquired.[17] In both types of case it is the combination of specific property with conduct in relation to that property which provides the basis for the imposition of a constructive trust.[18] It seems therefore that in the case of the constructive trust the unconscionable conduct performs the same function as the declaration of trust or the conveyance on trust does in the case of the express trust - the conduct links property with obligation. The conduct, in concert with the court order to which it leads, has the effect of impressing the property with the constructive trust. The trust is established or constituted at the date of the unconscionable conduct.
Support for the above explanation of the role of unconscionable conduct in establishing or constituting a constructive trust can be found in Pooley v Budd:
Trusts, however may be constituted not merely by direct declaration of trust, but also by constructive operation of the consequence flowing from the acts of the person themselves. Thus, equity will not merely enforce the execution of a trust against the trustees themselves but against all persons who obtain possession of the property affected by the trust, provided they had notice of the trust.[19]
In passing, it can now be suggested that if, as will be argued below, Lord Browne-Wilkinson by requiring “identifiable trust property” to establish a trust was requiring defined property to be impressed with the trust obligation, his separate second principle, that the holder of the legal interest must have knowledge of the facts alleged to affect his conscience to be a trustee, is superfluous as far as the constructive trustee is concerned. This is because a constructive trust is not “established” or (to use more appropriate language given that we are dealing with a constructive trust rather than an express or resulting trust ), cannot be “imposed”, in the absence of unconscionable behaviour. Unless the would-be constructive trustee has knowledge of the facts which are alleged to affect her conscience, her behaviour in relation to the property cannot be “unconscionable” and no trust is established or imposed. Accordingly, as far as the constructive trust is concerned, it could be argued that there is no separate requirement for knowledge over and above what is already required by the third principle.
Similar points apply to the express trust. If the express trust is established by a declaration of trust it goes without saying that knowledge is present as a settlor cannot make a declaration of trust without being “aware that he is intended to hold the property for the benefit of others”. His declaration is in effect a statement of his intention, as from that date, now to hold property that he had previously owned absolutely for the benefit of others. If the express trust is established by the transfer of property in trust the trustee is again inevitably “aware that he is intended to hold the property for the benefit of others”.
Leaving aside further discussion of the role of knowledge in the resulting trust to others, it is suggested that, if the third principle is understood as suggested above, the second principle is superfluous in relation to express and institutional constructive trusts.
C. “Identifiability” arises in Two Distinct Aspects of Trust Law
The need for “specified”, “particular” or “defined” property for a trust to exist has been referred to above in the context of the role of property in the trust. This requirement for the establishment of a valid trust is traditionally referred to as “certainty of subject matter”.[20]
The classic certainty case arose in situations where the would-be settlor failed either:
(a) adequately to define, specify or describe the property which she intended to impress with a trust so that either the amount or the nature of the property available for the trust was vague or unclear (for example, “the bulk of my estate” or “my blue chip securities”); or
(b) clearly to specify the share that each beneficiary was to receive.
The problem in such cases is purely semantic and arises from the use of relative words having no specific core of meaning.
However, as time went by, cases arose where the problem was not whether the property alleged to form the subject matter of the trust was described with sufficient semantic specificity, but whether property so described had been sufficiently identified for equitable title to pass. The issue typically arose in sale of goods cases where the vendor’s insolvency supervened after payment and prior to delivery, making it crucial to determine whether title had passed to the purchaser. The Sale of Goods Act 1908 contemplates two types of goods: specific and unascertained. Specific goods pose no identification problems as by definition they are “goods identified and agreed on at the time a contract of sale is made”.[21] Property in specific goods passes when intended to pass.[22] On the other hand, no property passes in unascertained goods[23] “unless and until the goods are ascertained,” usually by appropriation to the contract.[24] Accordingly, in cases where the lack of appropriation prevented the passing of legal title under the Act, it has been argued that equitable title has passed so that the insolvent seller held the goods on trust for the purchaser. In what appears to be the first of these cases, Re London Wine Company (Shippers) Ltd,[25] decided in 1975 but not reported until 1986, Oliver J held that “...any such trust must fail on the ground of uncertainty of subject-matter”.[26] However, in the context of that case, involving a specified number of cases of a particular wine which had not been segregated from the dealer’s stock or in any other way appropriated to the contract, when Oliver J said, “to create a trust it must be possible to ascertain with certainty not only what the interest of the beneficiary is to be but to what property it is to attach”, it was identifiability that he had in mind rather than semantic uncertainty. The alleged trust lacked subject matter as it was impossible to tell which of the cases in the dealer’s hands were “trust property” and which were not.
When similar issues arose in Re Goldcorp Exchange Ltd,[27] the trust argument was again rejected by the Privy Council, and it was held that the same reasons which prevent passing of legal title in unascertained goods apply also to the creation of a title in equity.[28] There was in that case “no identifiable property to which any trust could attach”.[29] Their Lordships’ advice does not contain any reference to certainty of subject matter as such.
Nonetheless, identifiability is discussed in connection with certainty in the texts. For example, under the heading of certainty of subject matter, McCormack states that “[t]he property which is said to form the subject-matter of the alleged trust must be identifiable”.[30] According to Moffat,
[a] purported trust will be void if the property intended to form the subject matter cannot be clearly identified. Expressions such as ‘the bulk of my residuary estate’... are too uncertain. ... It is only existing property - eg negotiable instruments, money, chattels, interests in land whether in possession or remainder - that can form the subject matter of a trust.[31]
The latest edition of Pettit opens its discussion of certainty of subject matter by quoting Lord Browne-Wilkinson’s third principle before going on to say,
but the requirement of certainty of subject is somewhat ambiguous for the phrase may mean that the property subject to the trust must be certain, or that the beneficial interests of the cestuis que trust must be certain.[32]
As is apparent from the above, the absence of identifiable trust property is fatal to the creation of a trust. What is also apparent is that the learned authors are treating identifiability as if it were synonymous with certainty. As this article will suggest that Lord Browne-Wilkinson’s identifiability principle goes beyond simply requiring “specified” “particular” or “defined” property (semantic certainty) to establish a trust and focuses on the need for defined property to be impressed with the trust, it is appropriate to examine here whether in fact certainty and identifiability are always employed to cover precisely the same ground. Clarke, discussing property, certainty and identifiability, does distinguish between the latter two concepts:
...[while]... certainty of subject matter is essential in the sense that you cannot have a property interest in a thing that does not exist ...[h]owever, does it necessarily follow that the thing must also be identifiable ? If the thing in question is in existence, and is in the hands of A, is it possible for B to claim a proprietary interest in the thing if it is impossible to distinguish it from other like things also in existence and also in the hands of A?[33]
She goes on to discuss three cases in which in effect all the plaintiffs were claiming that they owned x out of the x +y items held by the defendant, but could not say of any one item held by the defendant whether it was or was not their property. She poses the question: “Does this difficulty - essentially one of identification - rule out a proprietary claim?” In Hunter v Moss,[34] both certainty and identification are mentioned. Here the owner of 950 identical shares in Moss Electrical Co Ltd had orally declared a trust of 50 shares in favour of the plaintiff without indicating which of the 950 shares were to form the subject matter of the trust. The defendant argued that there was no certainty as there was no identification of the 50 shares. The Court of Appeal referred to the need for certainty of subject matter and held that a declaration of trust in respect of a specific number of shares in a named company was sufficiently certain “without any further identification of their numbers”.[35] While this case has been criticised[36] in light of the subsequent judgment in Re Goldcorp Exchange,[37] the distinction that is made in it between certainty and identification has not been challenged.
Both the judgment in Hunter v Moss and Clarke’s article clearly treat certainty and identification as separate matters, which raises the question as to why the same distinction has not been made by text writers. It is suggested that this distinction has been overlooked because semantic uncertainty or the failure to define the subject matter of the trust with sufficient specificity, and the failure to identify which items out of a larger quantity of identical items are to be subjected to the trust both cause the alleged trust to fail for lack of subject matter (if Hunter v Moss is overruled by Goldcorp). However, whereas semantic uncertainty is fatal to the establishment of a trust, identification problems can be overcome. The problem is basically one of proper constitution or forging the link between the obligation and the property. As Oliver J put it in Re London Wine:
I cannot see how, for instance, a farmer who declares himself to be a trustee of two sheep (without identifying them) can be said to have created a complete and perfect trust...and it would seem to me to be immaterial that at the time he has a flock of sheep out of which he could satisfy that interest.[38] (emphasis added)
Identification problems seldom arise where the settlor creates an express trust by transferring property to trustees since “necessarily it must be clear what property has been transferred so as to be held on trust”.[39] Any problems as to whether the trust has been properly constituted usually arise from non-compliance with the formalities prescribed for transfer of certain types of property.
If, however, the trust is to be constituted by a declaration of trust, and the settlor intends to create a trust out of a larger bulk of assets owned by him as in Oliver J’s example above,[40] identification problems can arise as to which of his property the settlor intended the trustee obligation to relate.[41] In such cases the declaration of trust does not of itself transfer the equitable title; the identification process is also necessary. Initial identification problems can be cured if the property is subsequently identified so as to create “a complete and perfect trust”.
The mechanics of “identifying” property so as to constitute the trust and transfer the equitable title will be discussed in part V below.
2. Identifiability in the context of tracing
Tracing in the trust context is a process used to identify the plaintiff beneficiary’s original property or its exchange product in the hands of the defendant or third party, to enable the plaintiff to bring a proprietary claim against a particular item of property or fund. It is the process of following property through changes in form that enables a beneficiary to identify misapplied trust property through one or more substitutions.[42]
A proprietary claim is particularly desirable if the misapplying trustee has since become insolvent, thus rendering a personal claim futile. Tracing would be appropriate if, for example, a trustee who has since become bankrupt had misappropriated trust funds to buy a valuable painting for her husband. If he still has the painting the beneficiary can trace the trust funds into this substitute asset in his hands and make a proprietary claim in respect of it. The link between the funds and the painting is fairly direct, but what if, instead of buying her husband a painting with the misappropriated funds, the trustee had simply deposited the funds in her husband’s bank account? No proprietary claim can succeed unless a plaintiff can identify his property in the defendant’s hands,[43] so the question arises as to how one identifies funds in an account which may, at the time the trust funds are deposited, be either in credit (that is, where the bank owes money to the account holder) or in debit (that is, where the account holder owes money to the bank). The answer lies in equity’s rules and presumptions which enable the trust beneficiaries to identify their property in the hands of the trustee or an innocent volunteer.[44] These tracing rules are set out at length in all standard trusts texts. However, one such rule, particularly relevant to a discussion of Westdeutsche, is that, if the plaintiff’s property or its exchange product disappears, as for example money would if deposited in a bank account which was or subsequently became overdrawn, the right to trace is lost.[45]
3 Comparing the different uses of the concept of identifiability in trust law
As has been seen, the concept of identifiability performs two distinct functions in trust law. First, identifiability determines whether there is the requisite property for a trust to be established and as such is relevant to the proper constitution of the trust. Secondly, in misapplication cases, equity’s identification principles determine whether there are any surviving trust assets (either in their original form or in exchange products) which can be traced and made the subject matter of a proprietary claim.
Whereas lack of identifiability in the first sense means that no trust comes into existence, lack of identifiability in the second sense has no bearing on the existence of the trust but limits the remedy available for breach of a properly constituted trust to the personal claim.
In practical terms this means that a claimant who wishes to assert a proprietary claim in respect of particular assets, on the basis that they are held for her on trust, might first use the concept of identifiability to establish that that property was impressed with a trust. Secondly, and quite independently of the previous use, if that property had either changed hands or changed form, or both, she would use the concept to establish whether any identifiable property remained. If there is no identifiable property in the first sense then there is no need to discuss identifiability in the second sense. This seems obvious, but it is submitted that in both Westdeutsche and in Gallen J’s judgment in Fortex the two steps seem to flow into each other.
Westdeutsche Landesbank Girozentrale (the Bank) had entered into an interest rate swap agreement with Islington Borough Council (the Council) involving an up front payment to the Council on 18 June 1987 of 2.5 million pounds. The money was credited to a bank account containing other Council money which became overdrawn later that month. Such transactions were subsequently found to be ultra vires local authorities. The Bank sought restitution of the balance of the up-front payment, minus payments made, plus compound interest. The sole issue in the House of Lords was the availability of compound interest. The majority of their Lordships refused to depart from the traditional position that compound interest was only available in equity. They examined whether a claim in equity was available on the basis that the payments made under the ultra vires contracts gave rise to a resulting trust and concluded that it was not.
As mentioned in the introduction, Lord Browne-Wilkinson prefaced his judgment with a general discussion of trust principles. The purpose of this part of the article is to analyse the application of the third principle within its context to establish what His Lordship meant when he referred to “identifiable trust property”. It will be suggested that this aspect of the judgment is far from clear. First, there is a hint of tautology about the statement that “[i]n order to establish a trust there must be identifiable trust property”. Could his Lordship have simply said “In order to establish a trust there must be identifiable property” or does the repetition of “trust” add something? Secondly, it is not totally clear which of the two senses of “identifiable” discussed above that His Lordship has in mind in some passages of the judgment. Thirdly, the judgment also refers to “defined trust property”. Is this phrase a synonym for “identifiable trust property”?[46]
Lord Browne-Wilkinson’s first reference to identifiability in the context of the facts of the case came when, having established his fundamental principles, he said:
Those basic principles are inconsistent with the case being advanced by the bank. The latest time at which there was any possibility of identifying the ‘trust property’ was the date on which the moneys in the mixed bank account of the local authority ceased to be traceable when the local authority’s account went into overdraft in June 1987. At that date, the local authority had no knowledge of the invalidity of the contract but regarded the money as its own to spend as it thought fit. There was therefore never a time at which both (a) there was defined trust property and (b) the conscience of the local authority in relation to such defined trust property was affected. The basic requirements of a trust were never satisfied.[47] (emphasis added)
At first sight the juxtaposition of “identifying” with “traceable” in the second and third sentences of the above passage suggests that His Lordship is equating identifiable trust property with traceable property and using identifiability in the second of the two senses discussed above. However, the repeated reference to “time” in the second and penultimate sentences of that paragraph suggests that His Lordship regards “trust property” as “identifiable” when two factors coincide: (a) there is defined trust property; and (b) the conscience of the alleged trustee in relation to the defined trust property is affected.
There is an immediately apparent difficulty with this passage in that both (a) and (b) refer to defined “trust” property. With reference to that expression in (a), if there is defined trust property then in the very nature of things the conscience of the owner of the legal title is affected, otherwise it would only be defined property. Similarly, in (b), if the conscience of the legal title owner is affected in relation to trust property it must be defined. If however the reference to “trust” is omitted from (a) and (b),[48] this difficulty disappears. It can then be suggested that what Lord Browne-Wilkinson meant in his third principle is that there is identifiable trust property so that a trust is established or constituted when the link between defined property and the conscience of the alleged trustee is present. This link could be, and in the case of an express trust would be, expressed in terms of property being “impressed” with the trust or the trust being constituted. In other words, what looks like a single requirement “identifiable trust property” seems to be comprised of two elements: (a) defined property; and (b) some knowledge or conduct in relation to the defined property which has the effect of “impressing” it with the trust so that the trust is properly constituted.
Accordingly, the identifiability principle could be expressed that, to establish a trust, defined property must be impressed with the trust. This formulation makes it explicit that the principle contains the requirement for both semantic certainty of subject matter and proper constitution of the trust. It also explains why, instead of simply stating, “[i]n order to establish a trust there must be identifiable property”, His Lordship required “identifiable trust property”, as this phrase underlines the requirement that the property has to be identifiable as trust property. The only property identifiable as trust property is property which has been impressed with the trust. He is not using “identifiable” simply as a synonym for “defined”. To be identifiable, property must be defined but that alone is not sufficient. As used by Lord Browne-Wilkinson in his third principle, “identifiable” comprehends “defined” but has a wider meaning as explained above.
Applying this to the facts of Westdeutsche, defined property (in the form of a specific sum of money from the bank which was deposited into the local authority’s account) had existed but no trust was ever established in respect of it. No trust could be established because, before the defined property could be impressed with any trust obligation (that is, before the local authority realised the loan was illegal), it had ceased to exist. This happened because the bank account into which the local authority had originally deposited the loan money went into overdraft shortly after the money was deposited and before the local authority knew anything which could affect its conscience (that is, before it could become a trustee). No opportunity for the local authority to act unconscionably in relation to the money arose. The money had gone before the local authority had sufficient knowledge to attract the label of unconscionable conduct to anything that it may have done with the money had it still been available. The defined property was never impressed with a trust. It was never transformed into “identifiable trust property”. No trust was ever constituted in respect of it.
Against the above interpretation of the third principle it might be argued that Lord Browne-Wilkinson was using “identifiable” to mean “defined” and nothing more. This interpretation overlooks the fact that “identifiable” and “defined “ are not exact matches as are “identifiable” and “definable” or “identified” and “defined”. Nor are the words exact synonyms. Property is “defined” when its nature, properties or essential qualities are described or fixed with precision or when its boundary or extent is determined.[49] Property is “identifiable” when it can be proved or recognised as being a certain thing or when its identity can be established.[50]
Property that is not defined with sufficient certainty can never become trust property. A trust cannot be established in respect of “most of my blue chip shares” because the phrase “blue chip shares” has no precise meaning, and even if it had the extent of the trust property is rendered vague by the phrase “most of”. On the other hand, defined property which is not currently identifiable trust property can become so. For example, if I have 20 cases of 1995 Cloudy Bay Sauvignon Blanc in my wine cellar and declare on the birth of my first child that I will hold 10 cases on trust for her until she is 21, those 10 cases will only become identifiable trust property when I take the appropriate steps to impress them with the trust. Precisely what those steps are will be discussed in greater detail in part IV below.
His Lordship’s first reference to “defined trust property” has already been discussed above in the context of the argument that identifiable has a wider meaning than defined. Later in the judgment he refers to “defined trust property” in the context of a restitutionary argument to extend the definition of the resulting trust:
The search for a perceived need to strengthen the remedies of a plaintiff claiming in restitution involves, to my mind, a distortion of trust principles. First, the argument elides rights in property (which is the only proper subject matter of a trust) into rights in ‘the value transferred’. A trust can only arise where there is defined trust property: it is therefore not consistent with trust principles to say that a person is a trustee of property which cannot be defined.[51] (emphasis added)
There are two references to trust principles in this paragraph. Does this mean that, if one read backwards from this passage to His Lordship’s original statement of “[t]he relevant principles of trust law”, principle (iii) could now equally well read “[i]n order to establish a trust there must be defined trust property” or as His Lordship puts it “[a] trust can only arise where there is defined trust property”.[52]
It is suggested that it should not be so interpreted. His Lordship’s sole object in the passage above is to dispel any notion that there can be a trust in the absence of property. He is not concerned here with the requirements for establishing or constituting a trust as he was when he stated his third principle. He is focussing instead on the permissible subject matter of a trust. His sole concern is to limit trusts to “defined trust property” so as to exclude the possibility of a trust arising in respect of “the value transferred”. He is stressing the role of property within the trust concept and making the same point as the authors cited in Part I made, namely, that defined, specified or particular property is essential.
The possibility that Lord Browne-Wilkinson was using identifiable in the tracing sense in the passage quoted above[53] has already been discussed and dismissed. However, when he next speaks of an identifiable trust fund in a later and separate section of the judgment, he again seems to be doing so in the tracing context:
[Counsel for the Bank’s specific disavowal of a constructive trust] was plainly right because the local authority had no relevant knowledge sufficient to raise a constructive trust at any time before the moneys, upon the bank account going into overdraft, became untraceable. Once there ceased to be an identifiable trust fund, the local authority could not become a trustee: Re Goldcorp Exchange Ltd (in receivership) [1994] 2 All ER 806... [54] (emphasis added)
This is a puzzling passage. The last sentence would make more sense if the reference to an identifiable trust fund was replaced with the words identifiable fund. The reference to there ceasing to be an identifiable trust fund suggests that there had been a trust fund at some stage which, as a result of the tracing rules relating to trust funds deposited into accounts which are or subsequently become overdrawn, had ceased to exist. This must be incorrect. As discussed above there had been a defined fund in the form of the loan moneys. That fund had ceased to exist (upon the account going into overdraft) before it could be impressed with any trust. The facts which potentially could have given rise to a constructive trust in respect of that defined fund were not known until after it had been dissipated. There was no subject matter onto which any trust obligation could fasten. As no identifiable trust fund ever existed His Lordship’s reference to an identifiable trust that had ceased to exist is inexplicable.
In summary, the meaning of identifiable trust property used by His Lordship in formulating and applying his trust principles is not clear. Three interpretations are possible:
1. Identifiable is synonymous with traceable. There are passages in the judgment where the word seems to be used in this sense. However, this cannot be the case where it appears in the third principle which relates to the establishment of a trust. Tracing has nothing to do with the establishment of a trust and can only take place in relation to property in respect of which a trust has previously been established.
2. Identifiable means no more than defined in the sense that trust property must be defined or described in terms which have precise and certain meaning. It requires no more than semantic certainty. Against this it has been argued that identifiable and defined are not synonyms.
3. Identifiable means that, not only must trust property be defined, it must also be impressed with the trust so that it is identifiable as trust property.
It is submitted that the third interpretation is correct. Not only does it explain the otherwise tautologous repetition of trust in the phrase “identifiable trust property”, but it is also consistent with the recent comments of the Court of Appeal in Fortex.[55]
As has already been discussed in section 3 of Part II above, identification difficulties arise most frequently in trust law where the settlor has attempted to constitute the trust by making a declaration of trust in respect of a specified number of items in circumstances where she owns a larger number of identical items. Similar attempts to establish an equitable title to property by invoking the trust concept in sales of unascertained goods have failed for the very same reason that it is impossible to establish legal title under the Sale of Goods Act 1908 (“it is impossible to have a title to goods when nobody knows to which goods the title relates”).[56] In these circumstances, “precise identification is just as much a requirement of equity as of law”.[57]
The analogy between the transfer of property in sale of goods cases and the constitution of trusts has been drawn by Sealey and Hooley:
Just as property cannot be transferred in unascertained goods, so also we may note that there cannot be a valid trust of unidentified property.[58]
Accordingly, the mechanics by which property passes under the Sale of Goods Act in relation to an agreement to sell unascertained goods provide a guide as to how unascertained goods can become identifiable trust property.
The Sale of Goods Act 1908 contemplates two types of goods, specific and unascertained. Property in specific goods passes when intended to,[59] but no property passes in unascertained goods until they are ascertained by appropriation to the contract.[60] Where the contract is for specific goods, there is no identification problem as by definition specific goods are “goods identified and agreed on at the time a contract of sale is made”.[61]
Unascertained goods are not defined by the Act but logic suggests that they are goods which have not been identified and agreed on at the time of the contract so that the contract is in respect of generic rather than specific items. Nor does the Act distinguish between goods that are wholly unascertained, in that the parties have not even designated a source of supply in their contract, and those goods that are partially identified as a result of agreement between seller and buyer that they shall be supplied from an identifiable bulk.[62]
Professor Goode calls partially identified goods “quasi-specific” goods and confines his use of unascertained goods to goods which are wholly unascertained and notes:
Goods can move (figuratively) from an unascertained to an ascertained state directly by an act of appropriation sufficient for precise identification. Alternatively there can be a staged progression, from unascertained to quasi specific (as the result of a post-contract agreement that the goods shall be supplied from an identifiable source or bulk) and thence to ascertained goods through appropriation from that bulk.[63]
He is suggesting that where unascertained goods rather than specific goods are involved precise identification requires appropriation:
Where the contract is for the sale of unascertained or quasi-specific goods, two stages of identification separated by an interval of time are involved. At the moment the contract is made, the parties must, as terms of the contract, agree upon the characteristics by which the goods to be supplied are to be identified. This means at least some verbal description, whether in writing or by word of mouth, but in many cases the difficulty of expressing in words all the relevant characteristics of the required article makes it desirable for the description to be supplemented by other methods, eg a sample, photographs, drawings...(etc). At the second stage, goods possessing the specific characteristics must be set aside and appropriated to the contract, so that the seller ceases to be entitled to proffer or the buyer to take other goods, even if having all the designated characteristics.[64]
The importance of appropriation is stressed by Oliver J in Re London Wine Company (Shippers) Ltd:
Section 16 [section 18 of the New Zealand statute] states quite clearly that the property is not transferred to the buyer unless and until the goods are ascertained but it is not a necessary corollary of this that the property does pass to the buyer when they are ascertained. To produce that result one either has had to find an appropriation (from which an intention to pass the property will be inferred) or one has to find an intention manifested in some other way.[65]
In that case the company was an insolvent wine dealer which had stocks of wine in several warehouses. Some of the wine had been sold to customers for laying down or investment purposes. Although it was clearly contemplated that the wine would belong to the purchasers and would be stored by the company, no appropriation from the bulk of the wine in storage had been made to answer any particular contracts. In the absence of appropriation by earmarking or otherwise setting aside each purchaser’s wine, legal property did not pass under the Sale of Goods Act, nor had the company created a completely constituted trust sufficient to pass the equitable title.
The Privy Council expressed its entire agreement with this case in Re Goldcorp Exchange Ltd.[66] Accordingly, where the alleged sale or trust involves goods forming an unidentified part of a bulk of identical goods, specific appropriation is required to transfer legal and equitable title alike.
According to Professor Hayton, in an article strongly critical of Hunter v Moss, the position is the same irrespective of whether tangible or intangible assets are involved. He stated that “[i]ndeed ‘Equity follows the law’, so surely one needs a specific appropriation where concerned with assets forming an unidentified part of a fungible bulk”.[67]
He makes the same point as editor of Underhill and Hayton :
No problems arise as to certainty of subject matter of a declaration of trust where the settlor’s declaration relates to specific property, eg ‘my one and only Picasso painting, all my shares in ICI plc, all my money currently in my Woolwich building society account’. However, if it relates to part of his unascertained property in which he retains an interest, no trust can arise till there has been a segregation or appropriation of specific property out of the larger mass of property, eg where it is intended to create a trust of 20 out of 80 cases of Chateau Latour 1982 in the settlor’s cellar or of 1000 pounds out of 5000 pounds in the settlor’s Woolwich Building Society account.[68]
Professor Martin has suggested that Hunter v Moss could be distinguished from another case[69] involving intangible assets in which the alleged trust failed because no identifiable assets had been impressed with the trust. The basis of the distinction is that, in the former case, the bulk from which the trust was carved was identified (that is, the 950 shares in MEL), whereas in the latter case it was not. If this was the case a declaration of trust in respect of $x out of a designated bank account with a larger balance would be valid.[70]
This cannot be correct. Although we might refer colloquially to our “money in the bank”, no money matching the balance of our account is physically held for us. Instead, money paid into a bank account ceases to be the depositor’s money and becomes the bank’s on the basis that it will repay the money when called upon to do so;[71] the relationship between the bank and its customer is one of debtor and creditor (and vice versa if the account goes into overdraft). The customer, while her account is in credit, now owns a single indivisible asset in the form of a debt (as opposed to a number of units capable of segregation from a bulk) and cannot make a declaration of trust in respect of part of it.
As Professor Goode explains, “[a] participation in a debt is necessarily an interest in the entirety of the debt, not an interest in an undivided part of the debt...”.[72] Accordingly, Professor Hayton’s approach is to be preferred. To constitute properly a trust in respect of “1000 pounds out of 5000 pounds in the settlor’s Woolwich Building Society account”, the settlor would have to segregate or appropriate the 1000 pounds by withdrawing it from that account. Only then could it become identifiable trust property within the meaning of Lord Browne-Wilkinson’s third principle.
This is consistent with the joint judgment in Fortex, which regarded a “separately identifiable fund” as essential to the imposition of an institutional constructive trust.
My submission is that Lord Browne-Wilkinson’s identifiability principle requires defined property to be impressed with the trust so that equitable title to the property passes to the beneficiary and the trust is properly constituted. I submit also that the principle is not satisfied if the property alleged to form the subject matter of the trust is not defined with semantic certainty. Nor is it satisfied if the property alleged to form the subject matter of the trust, although semantically certain, has not been subjected to any trust obligation so as to transfer equitable title and constitute the trust.
The issue of identifiability sufficient to establish a trust has arisen several times in recent years (frequently within a sale of goods context) and has not been consistently dealt with. Lord Browne-Wilkinson’s identifiability principle, stated in wide terms and embracing all types of property and express, implied and institutional constructive trusts alike, as interpreted above, clarifies this issue. The tenor of the joint judgment in Fortex on this point indicates that the Court of Appeal is thinking along similar lines in so far as the institutional constructive trust is concerned.
[*] LLB (Victoria), Lecturer in Law, University of Waikato.
[1] [1996] UKHL 12; [1996] 2 All ER 961, 988.
[2] In full the principle reads: “In order to establish a trust there must be identifiable trust property. The only apparent exception to this rule is a constructive trust imposed on a person who dishonestly assists in a breach of trust who may come under fiduciary duties even if he does not receive identifiable trust property”(at 988).
[3] Namely, the role of property within the trust concept and the need for the proper establishment of a trust.
[4] [1996] NZHC 1575; (1997) 6 NZBLC 102,141, 102,146.
[5] Fortex Group Ltd (in receivership and liquidation) v MacIntosh [1998] 3 NZLR 171. See case note at p 127 of this Review.
[6] [1996] UKHL 12; [1996] 2 All ER 961, 991.
[7] Hayton, “The Irreducible Core Content of Trusteeship” in Oakley, A (ed) Trends in Contemporary Trust Law (1996) 47.
[8] Oakley, A Constructive Trusts (3rd ed, 1997) 2.
[9] Ford, H and Lee, W Principles of the Law of Trusts (3rd ed, 1996) 1000.
[10] Infra note 19, and accompanying text.
[11] [1984] 1 NZLR 586, 594.
[12] Supra note 2.
[13] Oakley, supra note 8, at 29.
[14] In Foreman v Hazard [1984] 1 NZLR 586, 594, the court is speaking in terms of “the intention” to impress property with an express trust. The intention in such trusts takes the concrete form of a declaration of trust.
[15] Oakley, supra note 8, at 1.
[16] As to the difference between the two types of constructive trust: “A remedial constructive trust is one which is imposed by the Court as a remedy in circumstances where, before the order of the Court, no trust of any kind existed. The difference between the two types of constructive trust, institutional and remedial, is that an institutional constructive trust arises upon the happening of the events which bring it into being. Its existence is not dependent on any order of the Court. Such order simply recognises that it came into being at the earlier time and provides for its implementation in whatever way is appropriate. A remedial constructive trust depends for its very existence on the order of the Court; such order being creative rather than simply confirmatory” (supra note 8).
[17] Ibid, 5. While Westdeutsche was not of the former type of case, it would potentially have been of the latter type had the property in question not ceased to exist before the local authority had sufficient knowledge to do anything unconscionable in relation to it.
[18] Note too that, unless the court orders otherwise, a constructive trust takes effect from the time of the relevant conduct (Re Sharpe (a Bankrupt) [1980] 1 WLR 219).
[19] [1851] EngR 397; (1851) 14 Beav 34.
[20] See Foreman v Hazard [1984] 1 NZLR 586.
[21] Section 2.
[22] Section 19.
[23] The Act does not define unascertained goods but logic suggests that they are goods which have not been agreed on and identified at the time that the agreement to sell is entered into. They are generic rather than specific.
[24] Section 18 and s 20 Rule 5 (1).
[26] At 137.
[27] [1994] 3 NZLR 385.
[28] At 393.
[29] At 397.
[30] McCormack, G Proprietary Claims and Insolvency (1997) 68.
[31] Moffat, G Trusts Law Text & Materials (2nd ed, 1994) 127 and 129.
[32] Pettit, P Equity and the Law of Trusts (8th ed, 1997) 44. In this article the discussion of certainty is limited to the first of the two usages listed by Pettit.
[33] Clarke, “Property Law” (1995) 48 CLP 113-115.
[34] [1994] 3 All ER 216. (This case is discussed in the article by Clarke, ibid).
[35] At 220.
[36] Hayton, “Uncertainty of Subject Matter of Trusts” (1994) 110 LQR 335, and Ockelton, “Share and Share Alike?” (1994) CLJ 448.
[37] [1994] 3 NZLR 385.
[38] [1986] PCC 121, 137.
[39] Hayton D, The Law of Trusts (2nd ed, 1993) 75-76. There may still be problems in the exceptional cases referred to therein.
[40] No problem arises if the trust is declared in such a way as to make any further identification unnecessary, eg “all my sheep” or “all the sheep on my Te Kuiti farm”, or “my sheep earmarked with the numbers 15 to 45 inclusive”.
[41] Hayton, D (ed) Underhill and Hayton Law of Trusts and Trustees (15th ed, 1995) 125.
[42] McCormack, G Proprietary Claims and Insolvency (1993) 170-171.
[43] Jones, G (ed) The Law of Restitution (4th ed, 1993) 75.
[44] Ibid, 92.
[45] Ibid, 91.
[46] See Friar, “Equity, Restitution and Commercial Commonsense” [1996] NZLJ 448.
[47] [1996] UKHL 12; [1996] 2 All ER 961, 988-989.
[48] Support for editing His Lordship’s words in this fashion can be found in the passage under scrutiny. In the second sentence of that paragraph he has put speech marks around the words “trust property” which suggests that whether the property was or was not trust property was being left undecided for the time being. Used in this way “trust property” seems to be shorthand for “the property which is said to form the subject-matter of the alleged trust”. In parts (a) and (b) of the penultimate sentence of the paragraph the speech marks do not re-appear and this may well be an accidental omission. As suggested, the problem in (a) and (b) can be avoided by omitting the word trust altogether or alternatively, and in keeping with the beginning of the paragraph, by placing that word in speech marks.
[49] Hanks, P (ed) Collins English Dictionary (1979) 389.
[50] Ibid, 728.
[51] [1996] UKHL 12; [1996] 2 All ER 961, 991.
[52] This is the sense in which Friar understood it when he paraphrased the third principle as “there must be defined trust property” (Friar, supra note 46).
[53] Supra note 51.
[54] At 990.
[55] Fortex Group Ltd (in receivership and liquidation) v MacIntosh [1998] 3 NZLR 171, 175, where Tipping J spoke of the need for “a separately identifiable fund”.
[56] Re Goldcorp Exchange Ltd [1994] 3 NZLR 385, 395.
[57] Goode, “Ownership and Obligation in Transactions” 103 LQR 433, 449.
[58] Sealey, L and Hooley, R Text and Materials in Commercial Law (1994) 264.
[59] Section 19.
[60] Section 18 and s 20 Rule 5(1).
[61] Section 2.
[62] The same distinction was drawn by Lord Mustill in Re Goldcorp Exchange Ltd [1994] 3 NZLR 385, 392-393 between “generic goods “ and “goods sold ex-bulk”.
[63] Goode, R Commercial Law (1982) 218.
[64] Ibid, 220.
[65] [1986] PCC 121, 150.
[66] [1994] 3 NZLR 385, 401.
[67] Hayton, “Uncertainty of Subject Matter of Trusts” 110 LQR 335, 336.
[68] Hayton, D (ed) Law of Trusts and Trustees (15th ed, 1995).
[69] MacJordan Construction Ltd v Brookmount Erostin Ltd [1992] BCLC 350. The explicit approval of this case in Re Goldcorp Exchange [1994] 3 NZLR 385, 401 strongly suggests that Hunter v Moss was wrongly decided.
[70] Martin, J (ed) Modern Equity (15th ed, 1997) 97.
[71] Foley v Hill [1848] EngR 837; (1848) 2 HL Cas 28,36.
[72] Goode, supra n 57, at 448.
NZLII:
Copyright Policy
|
Disclaimers
|
Privacy Policy
|
Feedback
URL: http://www.nzlii.org/nz/journals/WkoLawRw/1998/4.html