The Door That Only Opens One Way.

The matter is reported as Jersey court rescues £15m trust from inheritance bill after tax planning mistake, Bailiwick Express, August 2026, describing a Royal Court of Jersey judgment cited in the report as [2026] JRC 171, before Deputy Bailiff Mark Temple sitting with Jurats Dulake and Gardener, the parties anonymised as “AB” and “BB”. At the time of writing the judgment had not yet appeared in the Jersey Legal Information Board’s public database, and the facts, figures and citation here are as reported; anyone relying on the decision should obtain the judgment text itself.

It is a good result, and a well-earned one. It is also one of the most instructive cases of the year for anyone who thinks about protecting family wealth, because of what it says about the shape of the door the court opened.

That door opens for mistakes. It does not open for creditors. Understanding why is the whole lesson.

Begin with the map, because the map is the point. The settlor and his wife were South African by birth. They had lived in the United Kingdom long enough to matter. The trust was settled in Jersey. The asset that caused the trouble was shares in a British Virgin Islands company. And the tax that nearly landed was British.

Four jurisdictions, and the only one with no personal connection to the family at all was the one whose court eventually rescued them. This was not a Jersey family. It was an internationally mobile family with a Jersey trust — which is the ordinary case, not the exception, and which is precisely why the failure happened. The report does not state that the settlor or his family were resident in Jersey; the Jersey connection is the situs and administration of the trust and the jurisdiction of the Court.

The structure was, on its face, exactly what a careful family builds. A Jersey trust settled in 2003 — more than two decades ago — holding an investment portfolio reported at around $15 million, established for straightforward intergenerational succession. Settled early, held offshore, professionally administered. Nothing improvised, nothing reactive, nothing done in the shadow of a claim.

Then two things happened, several years apart, that nobody joined up.

First, in April 2017, the settlor and his wife became UK deemed domiciled. Under the rules then in force — Inheritance Tax Act 1984, s. 267, as amended by the Finance (No. 2) Act 2017 — an individual resident in the UK for at least 15 of the previous 20 tax years is treated as domiciled in the UK for inheritance tax purposes, whatever their actual domicile, and a South African domicile of origin is no answer to it. That status does not announce itself. There is no letter, no filing, no moment of decision. It simply arrives with the passage of a fifteenth tax year, and it changes the tax character of everything the person subsequently does.

Second, in 2018, in what the report describes as a “tidying up” exercise, shares in a BVI company were transferred into the trust under two donation agreements — without UK tax advice being taken on the transfers themselves.

The consequence was expensive and, in retrospect, entirely predictable. A settlement of non-UK assets by a settlor who is not UK-domiciled is generally excluded property for inheritance tax; the same settlement made by a settlor who is domiciled or deemed domiciled is not. See IHTA 1984, s. 48(3), and ss. 58–65 (the relevant-property regime of entry, ten-year anniversary and exit charges). The 2018 transfers therefore landed inside the relevant-property regime rather than outside it. Per the report: an immediate inheritance tax charge of roughly £350,000, a further charge of about £144,000 falling due later in 2026, and ongoing annual UK income tax and capital gains tax of roughly £20,000 and £13,000 respectively — a total exposure north of £500,000 on a structure that had been designed, fifteen years earlier, precisely to avoid it. The precise computation of those charges is not set out in the published report and is not asserted here.

The trustee discovered the problem only when contemplating further transfers. The settlor’s evidence was that he “would never have proceeded with the transfer… in 2018… had I been aware of all the material issues.” His wife had died in 2023. Every adult beneficiary supported the application.

The Court found a genuine mistake, declared the two donation agreements voidable and of no effect, and authorised disclosure of the judgment to HMRC if required. The 2018 transfers are treated as though they never happened. The trust reverts to what it was in 2017.

The jurisdiction the Court exercised is not improvised equity. It is codified. Articles 47B to 47J of the Trusts (Jersey) Law 1984, inserted by the Trusts (Amendment No. 6) (Jersey) Law 2013 and in force since 25 October 2013, give the Royal Court an express statutory framework for setting aside transfers into trusts and exercises of trust powers.

The operative provision for a settlor is Article 47E. On the application of the settlor, or a personal representative or successor in title, the Court may declare a transfer of property into a trust voidable — with such effect as the Court determines, or of no effect from the time it was made — where a mistake was made in relation to the transfer, the transfer would not have been made but for that mistake, and the mistake was of so serious a character as to render it just to grant relief.

Two features of the statute matter to planners.

First, “mistake” is defined broadly. Article 47B extends it to mistakes as to the effect, consequences or advantages of a transfer, to mistakes of fact existing before or at the time of the disposition, and expressly to mistakes of law — including the law of a foreign jurisdiction. A UK tax consequence misunderstood in Jersey is squarely within the definition. That is not an accident of drafting; it is the point of it, and it is the provision that made this cross-border rescue possible at all.

Second, Jersey deliberately charted its own course. Where the UK Supreme Court in Pitt v Holt; Futter v Futter [2013] UKSC 26 narrowed the equitable doctrine of mistake and held that a trustee’s decision within the scope of its powers is not void or voidable merely because it was taken on inadequate deliberation or mistaken tax advice, the Jersey courts had already declined to follow the English Court of Appeal’s approach — see In re the S Trust (Royal Ct. of Jersey, 21 June 2011) — and the States then legislated the position into statute. Jersey is, by design, a jurisdiction where an honest error in trust administration can be corrected.

But — and this is the part clients tend to hear only as good news — it is a jurisdiction where an honest error can be corrected. Not a change of mind.

Strip the case to its structural features and the reasons for the outcome are visible:

• There was no adverse creditor. The only party whose position improved from the transfers was a revenue authority, and the Court expressly contemplated disclosure to HMRC. Nobody was ambushed.
• The beneficiaries consented. Every adult beneficiary supported the application. There was no contest about who should benefit.
• The mistake was genuine ignorance, not calculated risk. This distinction is doing enormous work. The Jersey courts have drawn a developing line between a settlor who was genuinely unaware of a risk — and is therefore mistaken about it — and a settlor who understood the risk, took it on advice, and simply misjudged it. The second gets no relief. A structure that was built to run close to a line and then crossed it is a bet that lost, not a mistake.
• The correction was sought promptly on discovery, and the whole history was put before the Court. Full disclosure is the price of discretionary relief.

Change any one of those facts and the application looks very different.

Here is where the case earns its place in this series.

A client reading the headline hears something dangerous: a court will unwind a transfer if it turns out badly for me. It will not. Jersey’s mistake jurisdiction runs in exactly one direction — it lets a transferor undo a transfer that was made in ignorance and that prejudiced no one but the transferor’s own tax position. It is not a mechanism for adjusting who gets paid.

Against creditors, Jersey law runs the other way, and has for centuries. Jersey’s customary-law Pauline action (the actio Pauliana, inherited from Roman law) allows a creditor to reverse a debtor’s transfer of assets — including a transfer into a Jersey trust. The modern statement is In re Esteem Settlement and the No. 52 Trust 2002 JLR 53 (Royal Ct. of Jersey), where a judgment creditor pursued transfers the debtor had made into two Jersey trusts and succeeded in part. The elements are the ones any American practitioner would recognise from voidable-transactions law wearing different clothes: the plaintiff must be a creditor of the transferor; there must be a transfer made when the transferor was insolvent or which rendered him so; the plaintiff must establish an intention to defraud creditors; and the plaintiff must show actual prejudice.

Those are opposite inquiries, and no amount of careful drafting converts one into the other. A settlor who moves assets into a trust because a claim is coming has not made a mistake within the meaning of Article 47B. He has made a transfer — and given a creditor a second cause of action to bring alongside the first.

The settlor in this case could ask a court to rewind 2018 precisely because nobody was chasing him in 2018. That is the sequence that made relief available. It is not available in reverse.

This is the practical lesson worth carrying into client conversations, and it is not the one the headline suggests.

The 2003 trust was fine. Twenty-three years old, settled long before any of the trouble, exactly the kind of seasoned structure this series has argued for repeatedly. The problem was a transfer made in 2018 — and the law assessed that transfer by reference to 2018 facts, not 2003 ones.

A seasoned structure does not season new contributions. Every subsequent transfer into an existing trust is a fresh disposition, judged against the settlor’s status, solvency, residence, and creditor landscape on the day it is made. This is true for tax, as this settlor learned. It is equally true — and considerably less forgiving — for creditor exposure. A twenty-year-old trust that receives a transfer three weeks after a lawsuit is filed offers the settlor no seasoning benefit on those assets whatsoever. The age of the vessel does not launder the timing of the cargo.

The corollary is a discipline: treat every top-up as a new settlement. Re-run the analysis. Confirm status. Confirm solvency. Confirm no claim is pending or reasonably foreseeable. Document the advice. A “tidying up” exercise is exactly the kind of low-ceremony administrative act where nobody thinks to do any of that — which is why it was a tidying-up exercise that cost this family half a million pounds.

The mobility of this family is not incidental colour. It is the mechanism of the loss, and it generalises.

Nothing about the trust changed between 2017 and 2018. Nothing about the BVI company changed. What changed was a fact about the settlor personally, in a fourth jurisdiction, arising automatically from nothing more active than continuing to live where he already lived. The trust deed could not have anticipated it. The trustee’s own jurisdiction had no reason to flag it. The asset’s jurisdiction was irrelevant to it. And it silently converted a well-designed structure into a taxable one.

That is the characteristic failure mode of genuinely international families, and it is worth naming plainly: the risk does not usually sit in the structure. It sits in the person. Residence, domicile, deemed domicile, citizenship, tax residence of a trustee, place of effective management, physical presence day-counts — these are personal attributes that drift, and they drift without generating any document that anyone reviews.

Note too that the trap has recently widened, not narrowed. From 6 April 2025, the UK replaced the deemed-domicile test with a residence-based one: under the Finance Act 2025, Sch. 13, an individual resident in the UK for at least 10 of the previous 20 tax years is a “long-term resident” and within the scope of inheritance tax on worldwide assets, with settled property moving in and out of scope by reference to the settlor’s status from time to time. The old fifteen-year runway is now ten, and settlor status is now a variable rather than a fixture. Structures set up under the previous regime should be re-examined against the new one.

For a creditor analysis the same principle bites harder still, because the relevant personal facts are solvency and foreseeability of claims — which change faster than residence does, and which no one thinks to document until a court asks.

One last observation, offered without criticism of anyone involved, because the report does not tell us where the advice chain broke.

This series argues constantly that protection must be independently administered — that a genuine, independent trustee is the line between a trust and a sham. That is right, but it is only half the requirement. Independent administration also has to mean competent administration: a trustee whose acceptance procedures catch the tax and legal consequences of an incoming asset before the transfer instruments are signed, not when someone later proposes a further transfer. For a cross-border family, that necessarily includes knowing where the settlor has been living and what that has done to his status.

An independent trustee who processes what the settlor sends without asking what it triggers has satisfied the form of the requirement and missed its substance. When a family later needs to prove that a structure was genuinely someone else’s to run, the administrative record is the evidence. A record showing a trustee that asked hard questions at every funding event is worth a great deal. A record showing a trustee that papered whatever arrived is worth considerably less — and in a creditor contest, rather than a tax one, it may be worth nothing at all.

• Jersey is a jurisdiction where honest mistakes can be fixed. Articles 47B–47J are a real and valuable feature, and the June 2026 decision shows the Court using them where the equities are clean — including for mistakes about foreign law.
• The relief is for ignorance, not for regret. A risk understood and taken is not a mistake. Courts distinguish the two, and they are alert to the difference.
• Nothing in this jurisdiction helps against a creditor. The Pauline action points the other way and is very much alive. A transfer made in the shadow of a claim creates evidence, not protection.
• Every transfer into an existing trust is a new transfer. Seasoning attaches to dispositions, not to vehicles. Re-run the whole analysis at each funding event and document it.
• Track the person, not just the structure. The settlor’s status is the variable most likely to move and least likely to be reviewed. Deemed domicile arrived without notice in April 2017; long-term residence now bites at ten years. Structures should be tested against the settlor’s current status in every jurisdiction he touches, not the status assumed when the trust was settled.

The family in this case were fortunate — not because the law is generous, but because their timeline was clean. They made an error in a year when no one was pursuing them, discovered it, disclosed it fully, brought every beneficiary along, and asked a court to correct it. That is the only posture in which this kind of relief is available anywhere.

The clients who will not be so fortunate are the ones who want to undo a transfer because someone is now pursuing them. For them there is no Article 47E, in Jersey or anywhere else. There is only the transfer, the date on it, and a creditor’s lawyer reading that date very carefully.

Build it early. Fund it carefully. Review it every time you touch it — and every time the family moves. This note is general commentary on published developments and is not legal or tax advice for any particular person or structure; outcomes turn entirely on the facts and on the law of the relevant jurisdictions, and any specific situation should be assessed with qualified counsel in each jurisdiction concerned.