The three certainties

The three certainties are the threshold conditions for an express trust. A settlor may intend benevolence, generosity, family provision, tax planning, insolvency protection, or commercial segregation; but equity will not impose trusteeship unless it can identify, with sufficient precision, three matters: an intention to create a binding trust, the trust property, and the beneficiaries or purposes for whom the property is to be held. The traditional formulation comes from Knight v Knight, though the language of three certainties is a later doctrinal distillation.

The doctrine is elementary but not simple. Its difficulty lies in the fact that certainty is not one idea. Certainty of intention asks whether the words and conduct, construed objectively in context, impose an obligation rather than express a hope, confidence, wish, request, or moral expectation. Certainty of subject matter asks both whether the trust property is identified and whether each beneficiary’s entitlement is sufficiently defined. Certainty of objects asks whether the court can determine who is entitled to enforce the trust or, in the case of a discretionary trust, whether any given claimant is or is not within the class.

For Cambridge purposes the topic repays close attention because it sits at the junction of Week 1’s conceptual account of the trust and the later doctrines of constitution, formalities, resulting trusts, discretionary trusts, and breach. The three certainties tell us when equity recognises a trust in the first place. They also reveal the institutional character of trusts law. The court is not merely interpreting private language. It is deciding whether the legal system can administer, supervise, and enforce a proprietary fiduciary regime.

The examination danger is to recite Knight v Knight and then treat all three certainties as if they posed the same kind of problem. They do not. Intention is a question of construction and objective commitment; subject matter is a question of proprietary identification and beneficial quantum; objects is a question of enforceability, standing, and administrative possibility. Nor should students treat uncertainty as invariably fatal to the claimant. If an intended express trust fails, the consequences differ. Property may remain with the settlor; it may result back to the settlor or estate; an outright gift may be construed; or, in commercial contexts, the absence of segregation may defeat proprietary priority altogether. These remedial consequences often determine the practical importance of the issue.

The three certainties developed from equity’s insistence that conscience can be bound only where the obligation is sufficiently definite. The early use and trust were practical devices for managing land, family settlements, and succession. Equity intervened because the common law title-holder’s conscience was affected by the terms on which he held property. But that intervention necessarily required limits. A court cannot compel a person to perform an undefined moral duty, nor can it administer a trust where the property, beneficiaries, or dispositive terms cannot be known.

Knight v Knight is the conventional starting point. Lord Langdale MR stated that, for a valid trust, the settlor’s words must be imperative, the subject must be certain, and the objects must be certain. The case was concerned with testamentary language, but the formulation became a general organising principle for express trusts. Nineteenth-century equity was especially concerned with the distinction between binding trusts and precatory words. Victorian wills frequently used language of hope, confidence, desire, and recommendation. Courts initially sometimes construed such words generously as creating trusts. Later authority became more cautious, reflecting anxiety that to treat every moral wish as a trust would impose obligations that testators had not clearly assumed and would distort beneficial ownership.

The historical movement is therefore from moralised confidence towards objective legal obligation. Re Adams and the Kensington Vestry is the classic corrective. A gift to a wife “in full confidence” that she would do what was right by the children was held to be an absolute gift rather than a trust. The case marks a refusal to translate domestic confidence automatically into legal duty. Comiskey v Bowring-Hanbury shows the other side: precatory-looking words may create a trust where the instrument, read as a whole, contains a sufficiently imperative scheme. The modern law has not abolished the significance of language; it has displaced formula with construction.

The certainty of objects has a separate history. Fixed trusts traditionally required a complete list of beneficiaries because trustees had to distribute according to defined shares and the court had to be able to execute the trust if the trustees failed. Powers of appointment were tested more leniently: it was enough to say whether a particular person did or did not fall within the class. McPhail v Doulton collapsed much of that distinction for discretionary trusts by adopting the same “is or is not” test for discretionary trusts as for fiduciary powers. That development reflected modern settlement practice and the prevalence of wide discretionary family and employee trusts. It also provoked the difficult Re Baden litigation on semantic uncertainty, evidential uncertainty, administrative unworkability, and capriciousness.

Certainty of subject matter has also changed in importance. In older land and chattel trusts, identification of the property was usually straightforward. Modern commercial trusts often involve fluctuating funds, unsegregated goods, securities, customer money, and insolvency. The decisions in Re London Wine, Hunter v Moss, and Re Goldcorp show that certainty is not merely semantic: it determines whether a claimant has proprietary priority over general creditors. This is why the three certainties cannot be treated as a quaint historical checklist. They remain central to the allocation of risk, entitlement, and insolvency priority in contemporary private law.

1. Certainty of intention. The question is whether, objectively construed, the settlor intended to impose binding equitable obligations on the alleged trustee. No technical words are required. “Trust”, “trustee”, and “beneficiary” are powerful but not conclusive labels. Conversely, a trust may arise without the word “trust” if the transaction shows that the holder of property is not to take beneficially and must deal with the property for another.

The modern approach is objective. The court asks what a reasonable person, with the relevant background knowledge admissible on ordinary principles of construction, would understand the words and conduct to mean. Subjective undisclosed intention is not enough. In family and testamentary contexts, the difficult line is between an enforceable trust and an expression of hope. In commercial contexts, the question often turns on whether money or assets were segregated, whether the recipient was free to use them as his own, and whether the arrangement is better characterised as debt, agency, bailment, charge, or trust.

Paul v Constance illustrates informality. Repeated statements that money in an account was shared, coupled with the parties’ dealings, sufficed to create a trust. Re Kayford shows commercial intention: a company protecting customer prepayments by placing them in a separate account created a trust even before insolvency. Re Farepak, by contrast, demonstrates that a trust is not created merely because customers have paid money which the company morally ought to safeguard. The trust analysis depends on intention manifested by the company’s acts and arrangements, not on sympathy for disappointed creditors.

1. Certainty of subject matter. This has two dimensions. First, the trust property itself must be identified. A trust cannot attach to an uncertain mass unless the law can say which property is held on trust. Secondly, the beneficial interests must be defined: the court must know what each beneficiary is to receive, or the mechanism by which entitlement is to be determined.

The leading cases should be arranged by function rather than chronology. Knight v Knight supplies the canonical taxonomy: intention, subject matter, objects. It is not a mechanical rule, but a vocabulary for distinguishing binding equitable obligation from imperfect disposition. Re Adams and Comiskey then demonstrate the modern method for intention. The court asks whether the instrument, construed as a whole, imposes obligation. Words of confidence may be merely precatory; equally, apparently soft language may be made imperative by the surrounding dispositive scheme.

Paul v Constance and Re Kayford show that certainty of intention may arise informally. Paul is useful because it prevents an overly formal account of trusts: ordinary language and conduct may be enough. Kayford is useful because it prevents an overly domestic account: a company may create a trust over customer money if its conduct objectively segregates that money for customers rather than leaving them as ordinary creditors. These cases should be contrasted with situations where money is paid into a general account and the recipient remains free to use it. In such cases, debt is usually the better analysis.

The subject-matter cases are indispensable in commercial trusts. Re London Wine and Re Goldcorp deny trusts over unascertained bulk goods where no specific property has been appropriated. They reflect both property principle and insolvency policy: proprietary claims should not be conferred upon selected creditors unless the relevant assets can be identified as trust property. Hunter v Moss complicates the picture. It upholds a trust of part of a holding of identical shares without segregation. The decision may be defended on the basis that shares of the same class in the same company are indistinguishable choses in action. It may be criticised for privileging intangible property without a fully convincing account of why unascertained tangible fungibles should be treated differently.

For objects, McPhail v Doulton and Re Baden (No 2) dominate. McPhail transforms discretionary trust analysis by adopting the “is or is not” test. Re Baden exposes what McPhail leaves unresolved: the difference between conceptual and evidential uncertainty, the extent to which a class must be practically ascertainable, and whether a court can uphold broad dispositive discretions without being able to list beneficiaries. An excellent Tripos answer does not merely state McPhail. It asks why the test changed, what it enables trustees to do, and how the court retains supervisory control.

Use the cases to show the underlying policies. Certainty protects trustees from indeterminate obligations; protects beneficiaries by making rights enforceable; protects third parties and creditors by identifying proprietary rights; and protects the court from assuming administrative tasks it cannot perform. The cases are not isolated illustrations but attempts to police the boundary of the trust as a private-law institution.

See practice questions section below.

See further reading section below.