Protection Has an Expiry Date.

Every developed legal system contains a trap-door beneath transfers made by a debtor sliding toward insolvency. Move value out of reach too late — to a friendly creditor, to a related company, to a structure — and the law reserves the power to pull it back. The doctrine wears two closely related faces. One is the unfair preference: a payment that favours one creditor over the general body on the brink of collapse. The other is the fraudulent disposition: a transfer designed to put assets beyond the reach of creditors altogether. They are cousins, not twins — but they share a spine. Each turns on the same fatal facts: a transfer made while insolvent, or by design on the eve of trouble, to defeat a class of claimants.

For anyone who builds protective structures, that shared spine is the whole lesson. A new judgment out of the British Virgin Islands illustrates the preference limb with unusual clarity — and, by contrast, shows why the timing and the jurisdiction of genuine planning are everything.

In Almond v Linxens, decided in June 2026, the BVI court ordered a clawback of US$125.9 million. The facts read like a textbook. A BVI finance vehicle within a large Chinese conglomerate made an unrequested repayment of a US$125.9 million loan to a connected company — a loan not due for another nine months. Two days later the vehicle announced that it could not meet some US$463 million in bond obligations. The cash had been pushed to a related entity on the very threshold of default.

Because the recipient was a connected person, the transaction fell squarely within the unfair-preference regime of the BVI Insolvency Act 2003. Under § 245, a transaction entered into when a company is insolvent — or that tips it into insolvency — which leaves a creditor better off than it would have been in liquidation, is an unfair preference. Crucially, BVI law does not require proof of an intention to prefer; effect is enough. And where the counterparty is connected, the vulnerability period stretches to two years, and the law presumes both that the transaction was a preference and that it was not made in the ordinary course of business. The burden flips onto the recipient to prove otherwise.

The recipient’s defence ran along two lines, and both failed. First, it argued the company was solvent at the time, so no insolvency transaction arose. But it could not assemble the basic building blocks of solvency — it was unable to produce bank statements evidencing the hundreds of millions in cash it claimed to hold. A solvency case that cannot be documented is not a solvency case at all.

Second, it argued the repayment was an ordinary treasury movement. Here the court drew the distinction that gives the judgment its teeth: the difference between the type of a transaction and its design. Repaying a loan is, in the abstract, an ordinary thing to do. But the design of this repayment — early, unrequested, to a related party, days before a public default, moving cash where the group wanted it to land — revealed preference-seeking conduct. A preference obtained by design is a different animal from the incidental consequence of ordinary trading. The routine label could not survive contact with the surrounding facts.

It would be a mistake to read Almond v Linxens as a story about offshore weakness. The British Virgin Islands is the jurisdiction where the creditor won. The real lesson sits one level up, and it is the oldest lesson in this field: protection has an expiry date that runs backwards. The doctrine does not punish the structure; it punishes the timing. Value moved while solvent, for proper reasons, long before any creditor was on the horizon, is value moved cleanly. Value moved in the twilight zone — that legally ambiguous dusk before insolvency when a debtor starts rearranging the furniture — is value the law can follow and recover.

This is why the firm says, without elegance but with conviction, that you build the ark before the rain. A trust settled, or a company funded, on the eve of a claim is not asset protection. It is a confession with a date stamp. The protective power of any structure is fixed at the moment the assets go in — and that moment must come while the sky is clear.

Timing is half the answer. The other half is the law that governs the structure once a creditor comes calling. The same transfer that the BVI preference regime unwound would meet a very different reception inside a properly settled Nevis or Belize trust — not because those jurisdictions bless fraud, but because they make the attacker’s road far steeper.

Nevis keeps a fraudulent-disposition regime, but on terms that favour the prudent settlor. The operative provision — § 26 of the Nevis International Exempt Trust Ordinance, Cap. 7.03(N) (the section long known to practitioners by its original number, 24) — puts the burden on the creditor and fixes it at the criminal standard:

Two features of § 26 reward the settlor who acts early. First, even where a creditor clears that formidable bar, the disposition is not void or voidable: the trust survives, and is liable to satisfy the claim only to the extent of the interest the settlor transferred into it. The structure is not destroyed; it is, at worst, made answerable up to the value that went in. Second, the section closes the window on timing. A disposition is not fraudulent against a creditor whose cause of action had not yet accrued when it was made (§ 26(4)), nor against one whose claim accrued more than a year before it (§ 26(3)). And § 26(5) forbids the obvious shortcut: intent to defraud may neverbe imputed to a settlor merely because he settled the trust within two years of a creditor’s cause of action accruing, or because he kept powers or benefits, or is himself a beneficiary. Timing alone proves nothing.

Belize is more uncompromising again. Rather than impose a short limitation period, it removes the cause of action. § 7 of the Belize Trusts Act, Cap. 202 walls the trust off from foreign claims and foreign judgments entirely — a Belize court will not vary the trust, set it aside, or recognise any claim against the trust property founded on the law or court order of another jurisdiction. There is no two-year clock to wait out because there is no clawback mechanism to invoke. The only true exception is fraud in the making of the trust itself — a forged deed, a lie to the trustee.

The contrast is not one of degree but of architecture. The following sets the BVI preference regime that undid Almond v Linxens against the two firewall jurisdictions the firm favours.

None of this is a licence to cheat a creditor, and the firm would not frame it as one. A Nevis or Belize structure is not a machine for laundering a debt incurred yesterday. It rewards foresight, not flight. What these regimes do — and do well — is protect the settlor who acted while solvent, for legitimate reasons, before any claim arose, by making the attacker prove a real fraud to a real standard within a real and short window. The settlor who waits until the twilight zone, as the debtor in Almond v Linxens effectively did, has forfeited the only advantage that mattered.

That is the discipline at the heart of the practice. The BVI judgment is a useful reminder of the downside, written by a court in real numbers: US$125.9 million, recovered, because the planning came last instead of first. Build the structure early, in the right jurisdiction, and the same doctrine that recovered that money works for you rather than against you. Build it late, and no jurisdiction on earth will save it.

The BVI analysis follows Harneys, Twilight zone: treasury payments and unfair preferences, summarising Almond v Linxens(BVI, June 2026) and the unfair-preference provisions of the BVI Insolvency Act 2003, § 245. The Nevis passage is quoted verbatim from § 26(1) of the Nevis International Exempt Trust Ordinance, Cap. 7.03(N) (originally § 24), in the consolidated text published by the Nevis Financial Services Regulatory Commission (Revision Date 31 December 2017); the Belize provision is the Belize Trusts Act, Cap. 202, § 7.