The pitch arrives the same way in Jersey, Cayman, Hong Kong and Singapore: settle your assets into a trust and they pass outside your estate, beyond the reach of probate, divorce and disorder. The pitch is not false. It is incomplete in a way that costs families dearly, because a trust is not a vault. It is a transfer of ownership wrapped in obligations, and it holds only as well as the settlor's willingness to let go.

What the structure actually does

A trust separates legal title from benefit. The trustee — a licensed company, in most serious cases — owns the assets and must manage them for the beneficiaries under the trust deed and the law of the chosen jurisdiction. Done properly, this delivers real things: continuity across death without probate in every country where assets sit, professional administration of wealth a family cannot yet manage itself, protection for vulnerable beneficiaries, and a measure of confidentiality in an era when little else provides it. Depending on where everyone lives, it may also be tax-neutral or tax-efficient, though any adviser who leads with tax in 2026 is selling yesterday's product.

What a trust does not do is make assets untouchable regardless of the settlor's behaviour. Every serious jurisdiction has fraudulent-transfer rules that unwind settlements made to defeat existing creditors. Courts can and do look through structures where the trustee never genuinely exercised ownership — the so-called sham analysis, or its quieter cousin, the finding that the trust assets remain effectively the settlor's own. The wall between settlor and assets is real only when the settlor respects it in practice, year after year.

The control paradox

Here lies the recurring error. The people wealthy enough to need trusts are usually people for whom control was the entire point of getting wealthy. They settle the assets, then keep directing the investments, ordering the distributions and treating the trustee as an obedient clerk. Each act of retained control chips at the wall. When the divorce or the creditor or the feuding child eventually arrives, the other side's lawyers subpoena the correspondence, and the emails read like instructions from an owner, not wishes from a settlor.

Modern statutes have tried to square this circle. Hong Kong and Singapore permit settlors to reserve investment powers by design; Jersey and Cayman practice accommodates reserved powers and protectors with broad authority. These tools are legitimate and often wise — a founder should not have to hand the family company's votes to a trust company's investment committee. But every reserved power is a calculated trade: governance retained today against protection weakened tomorrow. The right question for the settlor is not how much control can I keep, but how much am I willing to lose.

Letters of wishes and the protector problem

Two instruments sit beside the deed and cause most of the confusion. The letter of wishes is the settlor's non-binding guidance — who should benefit, when, why. Trustees who rubber-stamp it turn the trust into the settlor's alter ego; trustees who ignore it entirely find themselves in court or out of a job. The competent ones treat it as evidence of intent, weigh it against circumstances the settlor never foresaw, and minute their reasoning.

The protector is the trust's designated sceptic, typically holding power to veto distributions and replace the trustee. Appointed well — an independent professional with no economic interest — the protector is the structure's immune system. Appointed badly, the protector becomes the control problem in a new costume. A protector who is the settlor's oldest friend, or worse the settlor personally, re-imports ownership through the side door, and courts have noticed.

Why trustees get fired

Trustee relationships fail for boring reasons, which is why they are preventable. Silence: quarterly reports that arrive half a year late, beneficiaries who cannot get answers. Fees: escalation clauses exercised without conversation. Conflicts: a trustee embedded in a banking group that keeps selling the group its own products. And distance — the relationship manager who knew the family leaves, and the file passes to someone reading it for the first time at the meeting.

The families who avoid this treat the trustee as a counterparty to be managed, not a vault to be forgotten. They insist on meeting the actual administrators, not just the sales team. They review fees against the deed annually. They write the letter of wishes carefully and update it when the family changes, because a stale letter describing a family that no longer exists is a gift to future litigants.

The transparency era

The selling point is shifting under the industry's feet. Automatic exchange of tax information and beneficial-ownership registers have narrowed what secrecy a trust can legitimately offer, and the jurisdictions that built their franchises on discretion are rebuilding them on administration. Expect more litigation testing reserved powers, more courts asking whether protectors were truly independent, and more families discovering that the trust's durability was decided at the moment of settlement — by how much the settlor was honestly prepared to surrender. The trusts that survive the next decade will be the boring ones: well-minuted, independently administered, and respected by the people who created them.