Apologies to Andrew. I was actually replying to the query about parent-child bank accounts. This turns out to be based on future amendments to legislation as confided secretly to ATT and perhaps others.
HMRC’s new exclusion example is worse than not particularly useful. It is at very best narrow, at worst obscure and misleading. No one should be expected to puzzle over the meaning of general public guidance, still less to have to consult a specialist adviser about it.
A true stand-alone resulting trust is just not registrable as it is not an express trust at all (see Underhill, Snell et al). A classic occasion is a settlor creating an express trust but failing to dispose of the beneficial interest. If the failure is total there is never any express trust. If it is partial the trust is partly express and partly resulting. “To A for life” or “to my children on attaining 18” with no gift over on A’s death or if no child takes. Something every drafter and precedent studiously avoids. I do not think a specific reverter to settlor provision is properly described as a “resulting trust”; it is a specific limitation of an express trust e.g. a vested remainder, a contingent one that vests in the event, or one appointed under a power. HMRC at some level purport to get the difference: TSEM 9500 and 9600.
There are key points on substance and timing.
1 HMRC mention “mixed trusts”, not in guidance but in TRSM 10030: " the trust must register if any part of the trust is required to register". Literally, this is nonsense. Trusts are required to register, not parts. My example above of an express trust with a resulting trust within it may be what they mean and I think it is registrable in principle.
2 A resulting trust must exist from creation, if only potentially in the case of a partial failure as its actual operation may depend on future events e.g. all members of a class failing to take. The assets do not suddenly “go back” to the settlor if none does. The settlor had some species of contingent or defeasible equitable interest from the outset.
3 The obligation to register adheres at a precise time. For an “existing trust” that is (apparently) 1 September 2022 and, if subsequently created, the date of its creation. Another view is it is 1 September 2021 when the register opened! The longer permitted time within which to register surely does not affect this. The funny trust in HMRC’s example would not be registrable if at the key date all the beneficiaries had died provided the settlor then owned the assets absolutely or by bare trust. Although bare trusts are registrable this would be a resulting bare trust not an express one.
4 Once a trust has become registrable it stays registered for between 5-10 years. There is authority to require updating of prescribed information (save on assets) but no provision for a registration to expire except through time. So if HMRC’s Settlor becomes entitled later than the due date for registration the trust would have had to be registered and stay there until time-expired. If the Settlor becomes entitled before that date the trust was never registrable. Some who register trusts early may thus find they did not need to, if the correct date is 2022! As trusts do not “end”, at least not without a precise definition of that term, it is not a blanket answer to say this has already occurred. No doubt HMRC will not seek penalties for trusts already “ended”, whatever that means, prior to the due date (whatever that is).
Why is legislation on trusts not drafted by someone who understands the law and why is guidance not written in plain English?
Jack Harper
