The three certainties

The three certainties are the entry conditions for an express trust. A settlor may use the language of generosity, family expectation, commercial convenience or testamentary hope; equity will not treat that language as creating a trust unless three matters are sufficiently certain. First, there must be certainty of intention: the alleged settlor must have manifested an intention to impose binding fiduciary obligations, not merely to express a wish, motive or moral confidence. Secondly, there must be certainty of subject matter: the trust property must be identified, and the beneficial shares or interests must be sufficiently defined. Thirdly, there must be certainty of objects: the persons or purposes in whose favour the trust is to operate must be ascertainable by the appropriate legal test.

These certainties perform distinct functions. Certainty of intention protects autonomy. It prevents courts from converting loose domestic or testamentary language into enforceable obligations merely because a donee was expected to behave decently. Certainty of subject matter protects the proprietary character of the trust. A trustee cannot hold property on trust unless it is clear what property is burdened by the trust and what beneficial entitlement has been carved out of it. Certainty of objects protects administration and enforceability. Equity will not recognise a trust unless there is someone, or in the case of limited exceptions a valid purpose, by reference to whom the court can supervise performance.

For Durham Year 2 students, the topic is a bridge from Week 1. Week 1 established that a trust divides legal title from beneficial entitlement, and that equity imposes conscience-based obligations on the holder of legal title. Week 2 asks when that equitable machinery is switched on. It is therefore a topic about thresholds: when a promise becomes a trust; when words of hope remain outside equity; when property is sufficiently identified; and when a class of beneficiaries is sufficiently workable for judicial control.

The leading formulation derives from Knight v Knight. The modern law, however, is not a simple application of a nineteenth-century formula. It has developed through testamentary family cases, commercial insolvency disputes, employee pension litigation, and the law of discretionary trusts and powers. Particular care is needed with certainty of objects: fixed trusts, discretionary trusts and powers of appointment are governed by different tests, although the boundary between them is often blurred in problem questions.

In examination terms, the best answers avoid treating the certainties as a ritual incantation. They ask, in order: what juridical device is alleged; what words or conduct are said to create it; what property is affected; what beneficial interests are claimed; and what test of object certainty applies to that particular device. The conclusion should then explain the consequence of failure. Failure of intention normally leaves an outright gift or no trust. Failure of subject matter may leave the intended trustee holding beneficially, or produce a resulting trust. Failure of objects will normally invalidate the trust, unless the case concerns a power or a valid exceptional purpose trust.

The three certainties emerged from the Chancery tradition as a disciplined response to a recurring problem: people often transfer property while expressing expectations about how the recipient should behave. Equity needed a method for distinguishing enforceable trusts from gifts accompanied by advice, hope, affection or confidence. That distinction was particularly important in testamentary dispositions, where the alleged settlor was dead and the court had to construe a document without further clarification.

Knight v Knight is usually treated as the canonical source. Lord Langdale MR stated that a valid trust requires certainty of words, certainty of subject and certainty of object. The terminology has since changed. Modern lawyers usually speak of certainty of intention rather than certainty of words, because no particular form of words is required. The point is substantive: did the settlor intend to impose an obligation which equity will enforce? The historical language remains significant because many early cases concerned precatory words. Victorian testators frequently gave property to a spouse or relative with phrases such as in full confidence, hoping, desiring or requesting that the donee would benefit others. The older cases sometimes treated such language as sufficient to create a trust. Later authority moved against that approach, partly because it strained ordinary language and partly because it undermined the security of gifts.

The shift is visible in Lambe v Eames and Re Adams and the Kensington Vestry. These decisions reflected a more exacting attitude to intention. A donor who gives property absolutely, while expressing a hope that the recipient will use it generously, has not necessarily imposed a trust. The court must construe the instrument as a whole, but it will not lightly deprive the recipient of beneficial ownership by converting moral language into a fiduciary office. This development is also consistent with the broader nineteenth-century movement towards doctrinal clarity in private law: property rights should not depend upon vague sentiment.

The certainty of subject matter developed along a related path. Equity could enforce a trust only if the property and the beneficial interests were ascertainable. This requirement was straightforward where a testator gave Blackacre to T on trust for A for life, remainder to B. It was far harder where a testator gave the bulk of an estate, a reasonable income, or such parts of property as another might select. The cases show the court resisting trusts where it cannot identify either the fund or the shares. At the same time, equity sometimes supplies machinery where the settlor has supplied a standard capable of objective application, for example by authorising trustees to select items or allocate shares.

Certainty of objects underwent the most dramatic twentieth-century development. Traditional doctrine required a complete list of beneficiaries for a fixed trust. That requirement made sense because fixed beneficiaries have proprietary shares, and the trustee must distribute according to those shares. Discretionary trusts and powers created greater difficulty. In IRC v Broadway Cottages, the Court of Appeal treated a discretionary trust much like a fixed trust and required the possibility of drawing up a complete list. That position was repudiated in McPhail v Doulton, where the House of Lords adopted the test already used for powers in Re Gulbenkian: it must be possible to say of any given person whether he or she is or is not within the class.

This historical movement matters for exams. It shows that the three certainties are not merely formal requirements. They reflect changing judicial views about autonomy, family provision, administrative feasibility and the institutional competence of courts. Equity is willing to uphold trusts where the settlor has created a real fiduciary structure; it is not willing to rescue imprecise benevolence at the cost of proprietary coherence.

1. Certainty of intention

Certainty of intention asks whether the alleged settlor intended to create a trust, not whether the word trust was used. The inquiry is objective. The court construes words and conduct in context. A person may create a trust without technical language; equally, the use of elevated or fiduciary-sounding language will not create a trust if, properly construed, it expresses only confidence, hope or a non-binding arrangement.

The central distinction is between imperative and precatory language. Imperative language imposes an obligation: T is to hold the shares for A; T must apply the income for B; T is not to use the fund except for C. Precatory language expresses a wish or moral expectation: I hope; I desire; I am confident; I request. No single word is conclusive. In Comiskey v Bowring-Hanbury, language which began with confidence was held, in the context of the will as a whole, to impose a trust because the instrument went on to prescribe a scheme of division. In Re Adams, by contrast, the phrase in full confidence that the widow would do what was right by the children did not create a trust.

Intention may be inferred from conduct. Paul v Constance is the standard example. Repeated statements that money was as much the claimant's as the defendant's, combined with their treatment of the account, were sufficient to show a trust despite the absence of formal language. The case is not authority for the proposition that casual words always create trusts. It is authority for a contextual approach where conduct confirms that the alleged settlor treated property as beneficially shared.

Commercial context may alter the analysis. In Re Kayford, a mail-order company opened a separate bank account for customers' advance payments. The court held that a trust was created. The segregation of money, the purpose of protecting customers, and the company's conduct pointed to an intention to create fiduciary obligations. In commercial cases, separation of funds is often powerful evidence of intention, though it may also be relevant to subject matter.

1. Certainty of subject matter

Certainty of subject matter has two dimensions: certainty of the trust property and certainty of the beneficial interests. The trust property must be identifiable. If the settlor says that T is to hold some of my wine, part of my estate, or the bulk of my residuary property on trust, the court may be unable to identify the fund. In Palmer v Simmonds, a trust of the bulk of the estate failed because bulk was too uncertain. A trustee cannot be compelled to administer an unidentified portion of property.

The cases should be organised by the certainty they illuminate, while remembering that a single case may involve more than one certainty.

On intention, Knight v Knight supplies the classical formulation. It is not a mechanical formula but a framework for analysis. The more difficult authorities concern precatory words. Re Adams and the Kensington Vestry demonstrates the modern reluctance to convert words of confidence into a trust where the donee is given property absolutely. Comiskey v Bowring-Hanbury shows the opposite possibility: words of confidence may be imperative when the instrument as a whole creates a binding scheme. Paul v Constance confirms that no technical vocabulary is required and that conduct can establish intention. Re Kayford adds a commercial dimension: segregation of customer money and the purpose of protection can evidence a trust.

On subject matter, Palmer v Simmonds is the classic case on uncertain trust property. A trust of the bulk of an estate failed because the subject matter could not be identified. Re London Wine and Re Goldcorp show the strict approach to tangible assets forming part of a larger stock: unless the specific goods have been appropriated, the claimant cannot assert a proprietary trust. Hunter v Moss is the main qualification. A trust of a specified number of identical shares within a larger holding was valid without segregation. The case is regularly examined because it sits uneasily with the tangible goods cases and invites discussion of policy in insolvency.

On objects, IRC v Broadway Cottages represents the old complete list requirement for discretionary trusts. Re Gulbenkian adopted the is or is not test for powers. McPhail v Doulton extended that test to discretionary trusts and is the leading modern authority. Re Baden (No 2), the sequel to McPhail, shows the difficulty of applying the test. The Court of Appeal agreed in result but offered different accounts of conceptual and evidential certainty. The phrase relatives survived, but the reasoning is fractured. That fracture is often more important than the result: it reveals that certainty of objects is partly about language, partly about proof, and partly about administrative feasibility.

A sound case discussion avoids two errors. First, do not overstate Hunter v Moss. It does not abolish the need for certainty of subject matter; at most it relaxes segregation for identical intangible property. Secondly, do not apply McPhail to every trust. Fixed trusts still require a complete list. McPhail concerns discretionary trusts, where the beneficiary has a right to be considered rather than a fixed aliquot share.

See practice questions section below.

See further reading section below.