The Trust You Never Wrote Down.

On July 24, 2026, the English Commercial Court demonstrated how that proposition fails. In Soprim Construction SARL v The Republic of Djibouti, Mr Justice Picken granted a final charging order over roughly US $42 million sitting in Standard Chartered accounts in London — accounts held in the name of a company, not the debtor. Soprim Construction SARL v The Republic of Djibouti & ors [2026] EWHC 1850 (Comm) (Picken J, judgment handed down 24 July 2026), reported in the Essex Court Chambers case note Commercial Court Judgment on Charging Orders and State-Controlled Assets (the chambers’ page was not directly retrievable at the time of writing and is relied on as reported). The court’s route to that money is the part every planner should read twice. It did not need to find a sham. It did not need to pierce a corporate veil. It found that a bare trust existed — inferred from conduct, never documented — under which the account-holding company held the funds beneficially for the judgment debtor. And once that trust was found, the money fell squarely within the statute.

The uncomfortable lesson is this: a trust can be created against your interests, by your own behavior, and used as the instrument of your defeat. Most planning worries about whether a creditor can break a trust you wrote. This case is about a trust nobody wrote at all.

Soprim Construction SARL built the Doraleh Container Terminal in Djibouti between 2006 and 2008 — a major port facility, and the source of the funds later at issue. Disputes followed. Soprim obtained arbitral awards against the Republic of Djibouti in a principal sum of approximately US $56 million, which with interest and costs now exceeds US $135 million; the LCIA arbitration is catalogued by italaw as Soprim Construction v. Republic of Djibouti. Enforcement proceedings were commenced in England on February 26, 2019.

Seven years of enforcement later, Soprim identified approximately US $42 million in Standard Chartered Bank accounts in London. The accounts stood in the name of Doraleh Container Terminal SA (“DCT”) — a company, not the Republic. The funds derived from operation of the terminal.

Soprim’s case was not that DCT was a fiction. It was that DCT held those particular funds beneficially for the Republic, and that this beneficial interest was chargeable. Picken J agreed, and granted a final charging order over the entirety of the accounts.

The court held that the bare trust was established. It did so not from a trust deed — there was none — but from the factual matrix: the Republic’s systemic control over DCT, and the conduct of the administrators, which the court found aligned entirely with the State’s interests. From that pattern the court inferred an agreement under which DCT held the funds on bare trust for the Republic. The full reasoning is set out in the judgment; this note does not pincite paragraphs of the transcript.

Pause on the mechanism, because it is more dangerous to a planning structure than veil-piercing is.

Veil-piercing is an equitable remedy of last resort, deployed sparingly, and English courts have narrowed it considerably. It requires something close to abuse. A planner can advise around it: respect formalities, keep books, hold meetings, document decisions, keep the entity genuinely separate.

An inferred bare trust is different in kind. It is not a remedy at all — it is a finding of fact about what the parties actually agreed. The court is not saying “we will disregard this company because its use was improper.” It is saying “on the evidence, this company was always holding this money for someone else, and everyone involved knew it.” Formalities do not help, because the finding is drawn from conduct notwithstanding the formalities. Indeed, the more completely the beneficial owner directs the entity’s dealings with the fund, the more evidence there is of exactly the arrangement the court is looking for.

The second half of the reasoning is a jurisdictional trap that offshore structures walk into constantly.

The Republic’s position depended on the trust question being governed by a law under which no such trust would be found. Picken J held that the SCB accounts were situated in England, and that under Article 7 of the Hague Convention on the Law Applicable to Trusts and on their Recognition (The Hague, 1 July 1985) — given effect in the United Kingdom by the Recognition of Trusts Act 1987 (c. 14) — English law was accordingly the law applicable to the alleged trust. Article 7 supplies the governing law where no choice has been made: the law with which the trust is most closely connected, having regard to the place of administration, the situs of the assets, the trustee’s residence or place of business, and the objects of the trust and the places where they are to be fulfilled.

Where no one drafted a trust, no one chose a law for it. So the default rule ran — and it ran to the place the cash was sitting.

This is worth stating plainly for clients who think in terms of where their entities are registered. A company incorporated in a protective jurisdiction, banking in London, holds its cash under English law’s reach. The registered office does not travel with the money. The situs of a bank account is where the account is maintained, and that situs can supply the law under which an undocumented beneficial interest is found to exist.

With a beneficial interest identified and English law applicable, the final step was mechanical. Section 2(1)(a)(ii) of the Charging Orders Act 1979 (c. 53) permits a charge to be imposed on “any interest held by the debtor beneficially under any trust.” The judgment addressed the meaning of that phrase and applied it to the interest the court had found. An earlier interlocutory decision in the same proceedings, on security for costs, is reported as Soprim Construction SARL v The Republic of Djibouti [2026] EWHC 418 (Comm) (27 February 2026).

Note the breadth of the drafting. “Any interest.” “Any trust.” The statute does not require the trust to be express, or written, or intended by the parties to be a trust at all. A beneficial interest under an inferred bare trust is an interest held beneficially under a trust. The statutory hook was drafted wide enough that the difficult work was entirely evidential — prove the beneficial interest, and the remedy follows.

This should recalibrate how planners think about the English charging-order remedy. In the LLC context we are accustomed to the charging order as a limiting device — the creditor’s exclusive and often frustrating remedy, reaching distributions if and when they are made and nothing more. Here it operated as an acquisitive one, fixing a charge on identified cash in an identified account. The label is the same; the effect is not remotely the same. The character of a charging order depends entirely on the statute that creates it and the asset it attaches to.

Four points follow, and none of them is about sovereign debt.

1. Nominee and interposed-entity ownership is the weakest form of separation. If an entity holds an asset while the beneficial owner directs its every dealing with that asset, the arrangement is not protection. It is a documented beneficial interest waiting to be found. The structures that work are the ones where the separation is real — where the settlor genuinely cannot direct the disposition of the fund, because an independent trustee holds a genuine discretion. The Republic’s problem was not that it used an entity. It was that it never stopped controlling the money.

2. Do not confuse an entity’s situs with an asset’s situs. Protective jurisdictions protect what is within their jurisdiction. Cash in a London correspondent account is in London. Securities held through a New York custodian are subject to New York process. When a client asks whether their structure is “offshore,” the honest follow-up is: where is the money actually sitting, and whose courts have process over that place? Structuring the holder while leaving the asset in a creditor-friendly forum solves the easier half of the problem and leaves the harder half untouched.

3. Documentation is a double-edged instrument. Planners rightly insist on contemporaneous records. Soprim is a reminder that records document whatever actually happened. Board minutes recording that the entity’s directors acted on the beneficial owner’s instruction; correspondence in which administrators treat the fund as the principal’s; a pattern of dealings in which the entity never exercises independent judgment — all of that is evidence of a bare trust. The remedy is not to document less. It is to arrange the affairs so that the honest record shows genuine independence, because the record will show the truth either way.

4. The seasoning point holds, with a twist. Our standing theme is that protection must be established years before any claim, because a structure assembled under the shadow of a dispute produces the badges of fraud rather than a defence. Soprim adds a corollary about the years after. The Republic’s exposure came not from a hurried transfer but from years of consistent conduct that revealed who really owned the fund. Seasoning is necessary, but it is not sufficient. A structure that is old and continuously undermined by the settlor’s control is not a seasoned structure; it is a long evidentiary record of beneficial ownership.

The right-hand column is not exotic. It is the ordinary discipline of a structure built properly and then left alone. The failure in the left-hand column was not a failure of jurisdiction selection or of legal sophistication. It was a failure to actually let go.

The most quoted sentence in asset protection is that a structure is only as good as the day it was created. Soprim v. Djibouti suggests a harder truth: a structure is only as good as every day since.

The Republic did not lose because a court broke its trust. It lost because a court found one — assembled out of seven years of conduct, governed by the law of the place where the cash happened to sit, and delivered to the creditor through a statute drafted wide enough to catch any beneficial interest under any trust. There was no deed to attack because there was no deed. There was only a pattern of behaviour, and it spoke clearly enough for a judge to read.

For clients, the practical question is not “is my structure well drafted?” It is “if a court examined a decade of how I have actually dealt with these assets, what arrangement would it infer?” Those two questions have the same answer only when the separation is genuine — when an independent trustee really decides, and the settlor really cannot. Everything else is a trust waiting to be found by someone who is not on your side. This piece is general commentary on a reported English judgment and is offered as information, not as legal advice for any person or situation; outcomes turn on the specific facts, the governing law, the situs of the assets, and the jurisdiction in which enforcement is sought, and anyone with a structure that may present these features should take advice from qualified counsel in the relevant jurisdictions.