The interest in ambulatory trusts is that some of the practicalities remain in a state of flux – however the reader needs fi rst to have some idea of what they are. Classically ambulatory trusts occur in the mutual will situation:

“The doctrine…is to the effect that where two individuals have agreed to the disposal of their property and have executed mutual wills in pursuance of the agreement, on the death of the fi rst (“the fi rst testator”) the property of the survivor (“the second testator”), [comprising] the subject matter of the agreement [i.e. now both estates], is held on an implied trust for the benefi ciary named in the wills. The survivor may thereafter alter his will, because a will is inherently revocable, but if he does his personal representatives will take the property subject to the trust.”1

So far so good, but during the lifetime of the survivor how can they deal with the combined estates given that some sort of trust is now in place? What are the parameters of their permitted activities?

“The purpose of an arrangement [of this sort is] to enable the survivor during his life to deal as absolute owner with the property passing under the will of the party fi rst dying…the object of the transaction is to put the survivor in a position to enjoy for his own benefi t the full ownership, so that, for instance, he may convert it and expend the proceeds if he choose.”2

We shall see later some further comment by Dixon J as to the fi duciary duties imposed upon the survivor as to the assets in his hands but the ambulatory nature of the trust can be seen where he continued3:

“...I do not see the diffi culty in modern equity in attaching to the assets a constructive trust which allowed the survivor to enjoy the property subject to a fi duciary duty …crystallising on his death and disabled him only from voluntary dispositions inter vivos.”

Notice that what changes during the lifetime of the survivor is the corpus of the trust estate but what is not in issue is what will happen to the corpus on death4. The trust estate therefore morphs with time.

A species of trust would appear to arise at the agreement stage but either can, apparently, resile until the death of the fi rst so there is no immediate and irrevocable trust at that stage. On the death of the fi rst a trust crystallises permitting utilisation by the trustee or, put another way, with a floating aspect over the joint estates. The trust only fi nally settles in a fixed manner on death.

Ambulatory trusts occur not only in mutual wills but, as we shall see, also commonly in proprietary estoppel cases. More recently they seem to have been accepted as underpinning some joint property cases.

It is useful at this stage to clear away a source of potential confusion – the requirement for certainty in express private trusts. We all remember from our student days that one of the three required certainties of a private trust was certainty as to the subject matter of the trust5. So a valid trust can be as to “my collection of Picasso engravings” as much as it can be to “the contents of my deeds box deposited with XYZ Bank”. Equally, as with an ambulatory trust arising from a mutual will, the trust defi ned as at date of agreement may be as vague as “my estate at my death” and the fi nal trust will then be fi xed at date of death upon whatever is left6.

Correspondingly one can see why Lord Walker in Thorner v Majors 7 did not have any diffi culty with the possibility that between the date of (1) the promise of the farm by Peter to David at Peter’s death coupled with detrimental reliance, and (2) Peter’s demise, that the extent of the farm might change. After all the promise was effectively “whatever the farm comprises at my death”.

So certainty of subject matter is not the problem it might be thought to be. The defi ning difference between this and other trusts appears to be the fl oating nature of the trust which one might say is another way of observing that certain common fi duciary duties do not apply or are modifi ed. For instance compare the ambulatory trust with a discretionary trust where the trustees are themselves within the class of benefi ciaries. In such a case the trustees can make dispositions in their own favour but are still bound by a duty to consider all the objects of the trust when exercising their discretions, they have a duty to act impartially and with an eye to the interests of successive benefi ciaries. In a mutual will case the survivor is not so hidebound on the statements in the authorities, whoever may be entitled on their death the survivor appears to be fully entitled to focus their utilisation of the trust property on themselves (rather than remaindermen).

Joint property cases

Before proceeding to the two main issues arising on ambulatory trusts in practice let us look briefl y at joint property. In Kernott v Jones 8 it was remarked:

“It was also accepted [in Stack v Dowden9] that the parties’ common intentions might change over time, producing what Lord Hoffmann referred to in the course of argument as an “ ‘ambulatory’ constructive trust””:

For present purposes this meant that looking back in time the court could determine that as at the time of acquisition of a couple’s home they might hold it for themselves upon a constructive trust in certain proportions. Then, as in Kernott the couples’ circumstances changed and they formed (in effect)

a new intention10 that Mr Kernot’s share would become fi xed whilst Ms Jones’ share would grow as the equity in the house grew. The result was that we can identify a constructive trust which at purchase is in maybe equal benefi cial shares but at a later stage those shares may change to either (i) different but fi xed shares or (ii) (as in the case itself) shares that would continually fl uctuate with the housing market.

Wider Issues – esp. Mutual Wills and Proprietary Estoppel

The interesting issues – more stark in mutual wills and proprietary estoppel cases are:

  1. How far can the ultimate benefi ciaries11 control the “trustee” during their dominion over the estate(s)?
  2. What can the ultimate benefi ciary hope to achieve if the trust is broken prematurely?

Beneficiary control

As to benefi ciary control Dixon J in Birmingham v Renfrew 12 said:

“No doubt gifts and settlements, inter vivos, if calculated to defeat the intention of the compact, could not be made by the survivor and his right of disposition, inter vivos, is, therefore, not unqualified.”

Notice the judge is taking as his reference point for what the survivor can do, the terms of the original compact between the parties. He continues:

“But, substantially, the purpose of the arrangement will often be to allow full enjoyment for the survivor’s own benefi t…” (my underlining).

Again he is demonstrating that any limitation upon the freedom of action will normally spring from what was agreed when the original parties were alive. Nourse J13 added in at this point:

“…Dixon J was there clearly referring to (lifetime dispositions) calculated to defeat the intention of the compact. No objection could normally be taken to ordinary gifts of small value.”

However fi nancially the world has moved on. The costs of long term care for the elderly can have a massive impact on an estate. Simple living costs of, say, an active widow who in a long widowhood still drives, wants to travel for enjoyment and to visit relatives and so on may be justifi ed by asking what her late husband would have wanted - yet be seen by the benefi ciaries as a spiteful attempt to live the high life and ensure they get nothing at the end of the day. Just to gauge where the line falls by putting small gifts on one side and obvious waste on the other is no real help. We can, indeed must, resort to the compact and perhaps also to the difference between what one can infer was agreed and in default what one might imply was agreed14. If one was to have asked husband and wife about their assumptions at the compact one can see the diffi culties with this task. The wife might expect if she dies fi rst that her widower might lead a similar life, even remarry. The husband’s prospective attitude might be in that scenario that he would not remarry (“been there, done that”) but rather fancies enjoying being footloose and fancy free.

Similarly with a proprietary estoppel (after the point where there has been detrimental reliance) the promisor retains his estate until his death and may anticipate he can do what he likes with his own assets. However Lord Scott suggested in Thorner v Major 15 that if Peter had tried to sell the farm in his lifetime that David could have obtained “consequential relief” – leaving open the tantalising issue as to what that might have been.

Clearly the promisee has achieved at this earlier point a real fl oating trust that he or she may well see as something that he can protect from certain dissipating acts by the promisor. Perhaps even a dealing can be prevented in certain circumstances16. But the question remains, is the court going to be wholly fi xated on the compact when deciding what actions by the survivor/ promisor are permissible, is it going to adopt an “absolute owner” stance17 as the default position or is it going to also stand back and apply a reasonableness or proportionality test? This remains an unknown in modern circumstances.

Premature determination

Gillett v Holt 18 was a case of premature determination of the fl oating trust – Mr Holt having indicated he no longer felt bound by his – as one might say – engagement, with Mr Gillett. Similarly with the fl oating trust as with a mutual will – suppose the survivor of mutual wills is left in possession of the contents of a wine cellar and the survivor is an alcoholic intent on drinking their way through it. What can the ultimate benefi ciary hope to achieve at this early point in time in terms of material benefi ts compared with the assets promised for a future event (usually death)? Gillett saw this issue being grappled with head on. There the remedy was fashioned by Robert Walker LJ (as he then was) in part taking account19 of the accelerated receipt of benefi ts that would otherwise only have been received later when Mr Holt died. In argument in Jennings v Rice 20 the writer had the temerity to suggest to the same judge that another key aspect must have been the certainty of receipt rather than the “risk of the game” that Mr Holt might have dissipated his estate for whatever reason by his death – which did not evoke any words of disagreement.

Notice that what the Court plainly did not do was establish the life expectancy of Mr Holt, assess his likely ultimate estate and carry out a net present value calculation to award to Mr Gillett. That course would as near as possible have carried out their engagement in a contractual manner. A variation would be to then discount for the “risk of the game” dissipation aspect21. One cannot work it out from the reported fi gures but the impression is that Mr Gillett received a respectable share but not an overly generous one – especially given that Mr Holt had passed away by the time Jennings v Rice was argued in 2002.

This question remains difficult to advise upon but relevant factors are bound to be discounts from what was agreed or promised, the life expectancy of the survivor/promisor (and indeed the ultimate beneficiary).

An important sub-plot though is the time gap between when the promisee acted to his detriment and the breach in a non-premature determination case (e.g. the promisor dies). The writer takes the view that the Court should follow what the parties themselves agreed. So if the agreement was “you come and look after me for the rest of my days and the house is yours” it should not matter whether the promisee looks after the promisee for 3 months or 30 years. Again it is a risk both sides take that the promisor will live a short or long time – their choice. However this was not the view of the Court of Appeal in argument in Jennings nor in the result22. That illustrates that the Court will go outside the engagement between the parties in assessing what remedy to fashion even in a complete fulfi lment case. If this is the case in this scenario it must also affect the starting point for a premature determination case.