Leaving this course?

Lesson 1 of 8

What a trust is, and why it is not a company

A trust is not an entity. That one fact explains most of the difficulty it causes in financial crime work, and it is the fact most often got wrong on a file.

The arrangement, not the thing. A company exists: it is registered, it has a number, it can own property and be sued. A trust is a relationship. A settlor transfers assets to a trustee, who holds the legal title and is bound by duty to administer them for beneficiaries. There is no entity holding the assets. There is a person holding them on terms.

Which means the usual questions do not fit. Asking who owns a trust is a category error; nobody owns it. Asking for its registration number assumes a register that may not exist. Asking for its directors confuses it with a company. The questions that work are: who created this, who controls it, who benefits, and on what terms.

The parties, and what each one tells you:

  • The settlor put the assets in. The source of those assets is the source of the structure's wealth, and if you do not know the settlor you do not know it.
  • The trustee holds legal title and exercises control. They may be an individual, a professional firm, or a company in a jurisdiction chosen for the purpose.
  • The beneficiaries receive, or may receive. This is where most of the opacity lives, for reasons the next lesson sets out.
  • The protector or enforcer, where one exists, holds powers over the trustee: to remove and replace them, to veto distributions, sometimes to amend the terms. A protector can be the person really in control while appearing nowhere in the ownership chain.

Foundations sit between the two. A foundation is an entity, registered and with legal personality, but it is used for the same purpose as a trust: assets dedicated to a purpose or a class of people, administered by a council rather than owned by shareholders. Treat them as the arrangement they resemble in substance, not as the company they resemble on paper.

Legitimate uses are the majority. Succession planning across generations, holding assets for a minor or a person who cannot manage their own affairs, charitable endowments, pension arrangements, employee share schemes, and keeping a family business intact through a death. A structure being a trust tells you nothing about intent.

But the property that makes trusts useful is the property that makes them attractive to abuse. Legal ownership and the benefit of ownership sit with different people, in a document that is usually private, often governed by the law of a jurisdiction neither party lives in. That is a feature for the family and a problem for anybody trying to establish who is behind a payment.

So the standard treats them as a transparency problem of their own. The international standards address legal persons and legal arrangements separately, because what works for a company register does not work for a private relationship, and the next seven lessons are about the difference.

Lesson 2 of 8

Where the opacity comes from

Opacity in these structures is rarely the result of anybody lying. It is usually the result of arrangements that are entirely lawful and that happen to leave your question unanswerable.

Discretionary interests are the first source. In a discretionary trust the beneficiaries have no fixed entitlement. The trustee decides who receives what, and when, among a class of people. So the honest answer to "who benefits from this trust?" may be "the trustee has not decided", and the class may be defined as widely as the settlor's descendants and any person the trustee adds later.

Letters of wishes are the second. A side document, usually not part of the trust deed and usually not disclosed, in which the settlor sets out how they would like the trustee to exercise that discretion. It has no legal force. It is frequently followed. A structure can therefore be controlled in substance by somebody who holds no power on paper.

Nominees are the third. A nominee shareholder holds shares for somebody else; a nominee director acts on instruction. Both are lawful in many places and both place a name in the register that is not the name of the person deciding. The signal to look for is a person or firm appearing as shareholder or director across many unconnected structures.

Layering adds the fourth. A company owned by another company owned by a trust whose trustee is a company in a third jurisdiction. Each layer may have a defensible reason. The effect of four of them is that establishing control requires cooperation from four sets of records, in four legal systems, with four different disclosure rules.

Then there is the choice of law. The trust deed states which law governs it. That choice determines what must be registered, who can demand information, and whether the arrangement is even recognised. A trust governed by the law of a place with no registration requirement and strong confidentiality is not evidence of wrongdoing, and it is a material fact about what you will be able to verify.

Powers reserved to the settlor are worth hunting for. A settlor who can revoke the trust, replace the trustee, direct investments or add themselves as a beneficiary has not really given the assets away. Such a reserved power is the clearest single indicator that the apparent structure and the real control differ.

And the purpose trust has no beneficiaries at all. Some arrangements exist for a purpose rather than for people, with an enforcer appointed to hold the trustee to it. They are legitimate and they defeat a control framework whose only question is who the beneficiaries are.

None of this is a red flag by itself. Every mechanism in this lesson appears in ordinary family planning. They matter because they are the tools available to somebody who wants distance between an asset and themselves, and because knowing they exist is what lets you ask the question that resolves them.

Lesson 3 of 8

Who you have to identify

The international standard requires you to identify the beneficial owner of a legal arrangement. Applying that to a trust means identifying several people, and the list is not optional.

The required set is wide on purpose. For a trust, it covers the settlor, the trustee or trustees, the protector where one exists, the beneficiaries or the class of beneficiaries, and any other natural person exercising ultimate effective control. Not one of these alone, and not whichever is easiest to obtain.

A class is not an excuse to stop. Where the beneficiaries are defined by characteristic rather than by name, you take enough information to establish the identity of each at the time of a distribution or when they exercise their rights. The obligation does not disappear; it attaches to the point at which somebody actually receives value.

Control and entitlement are different tests, and you need both. A beneficiary may receive a great deal and control nothing. A protector may control everything and receive nothing. A framework that records only entitlement will miss the person who matters, and one that records only control will miss the person the money reaches.

The trustee has duties of their own. A trustee is required to hold adequate, accurate and current information on the identity of the parties, and to disclose their status when forming a business relationship or carrying out an occasional transaction above the threshold. A trustee who declines to say they are acting as trustee is not meeting the standard, and that refusal is itself information.

Which gives you a question to ask on every account. Are you acting on your own behalf, or for somebody else? It is a simple question and it is routinely not asked, with the result that an account is opened in the name of a professional who is not the customer in any meaningful sense.

Beware the professional intermediary account. A lawyer, accountant or trust company operating pooled or client accounts may hold money for many underlying parties. Treating the firm as the customer and stopping there leaves you with no view of the people whose money it is, which is exactly the view the obligation exists to produce.

One national example, where the obligation runs through several instruments. The Proceeds of Crime and Anti-Money Laundering Act carries the customer due diligence (CDD, which much of the industry calls KYC) duty; the Companies Act requires a company to keep a register of its beneficial owners; and the Companies (Beneficial Ownership Information) Regulations set out what must be recorded and lodged. A trust holding shares in a Kenyan company therefore meets a registration requirement at the company layer even where the arrangement itself is not registered.

And record what you could not get. Where a party cannot be identified, the file should say so explicitly, with what was asked and what was refused. An incomplete file that admits what is missing supports a decision. One that looks complete because the gaps were left blank does not.

Lesson 4 of 8

The gatekeepers

Most structures in this subject are created and administered by professionals. Understanding what those firms do, and what they know, tells you where the information actually is.

Trust and company service providers do the work. They form companies, act as or provide trustees, provide a registered office and an address, supply nominee shareholders and directors, and act as an authorised signatory. Taken together, a single provider can supply every visible element of a structure while the client remains invisible.

They are therefore the control point. The provider knows who instructed the formation, who pays the fees, who gives directions, and which other structures the same client holds. No bank in the chain knows any of that. The standard puts obligations on these firms for exactly this reason, and supervision of them varies enormously between jurisdictions.

What to ask when one appears on your file. Which jurisdiction licenses them and under what regime. Whether they are supervised for anti-money-laundering purposes and by whom. Whether they will confirm, in writing, that they hold current information on all the parties. Whether they are prepared to disclose it on request or only under compulsion. A provider that will not answer these questions has told you something.

Lawyers and accountants hold the same position with an added complication. Professional privilege genuinely protects legal advice. It does not cover holding funds, forming entities or acting as a nominee, which are not advice. The distinction is frequently asserted more broadly than it holds, and a privilege claim over the identity of a party to a transaction deserves a careful second look.

Fees tell you about the client, not just the structure. Who pays the provider is a direct indication of who is driving the arrangement. Fees paid by a party who appears nowhere in the structure is one of the more useful signals available.

Provider churn matters. A structure that changes administrator every year or two may be shopping for one that asks less. The pattern is visible from the file history if anybody looks, and it rarely appears in a periodic review that only checks whether documents are current.

And the risk-based approach cuts both ways. A large, regulated provider in a well-supervised jurisdiction, with its own controls and a reputation to lose, is a lower risk than an unlicensed formation agent with a website and a mailbox. Treating all of them as one category wastes effort on the first and under-weights the second.

The practical position. When a professional intermediary stands between you and the client, your due diligence has two subjects: the structure, and the firm that built it. Most frameworks document the first and assume the second, which is the wrong way round when the firm is where the knowledge lives.

Lesson 5 of 8

What the warning signs actually look like

Red flags for structures are not exotic. They are mismatches, and the useful ones are those a reviewer can see from the documents in front of them.

A structure disproportionate to the wealth. Three companies in two jurisdictions and a trust, to hold a single modest property. Complexity costs money to create and maintain, so a structure more elaborate than the assets justify was built for a reason other than administration.

A purpose that nobody can state plainly. Succession across two countries is a reason. Asset protection from a specific identified risk is a reason. "Tax and privacy", offered without detail, is not a purpose; it is a label over one.

The settlor who is also everything else. Settlor, protector, principal beneficiary and signatory, with a professional trustee who follows instructions. The assets have not left the settlor in substance, whatever the deed says.

Jurisdictional geography with no connection to anybody. A trust governed by the law of a place where no party lives, holding a company registered in a second, with a bank account in a third, and assets in a fourth. Each hop is a place your enquiries have to travel to.

Trustee or provider changes without explanation, particularly shortly before a large transaction or after a request for information.

Distributions that do not look like distributions. Round amounts at regular intervals to a beneficiary who then passes them on. Payments to third parties on a beneficiary's instruction. Loans to beneficiaries that are never repaid and never pursued, which is a distribution wearing a different name.

Assets that are hard to value. Art, jewellery, unlisted shares, intellectual property, or a loan to a connected party. Each permits a value to be asserted rather than observed, which matters when the question is how much went in and where it came from.

Reluctance about the deed itself. A client who will provide a certificate of incorporation but not the trust instrument is withholding the only document that establishes the terms. Summaries prepared by the provider are not the same thing.

And the strongest signal is a changed answer. Where the parties described at onboarding differ from those described a year later, with no documented event in between, something has moved that nobody told you about. Comparing today's declaration with the one on file is a cheap control that few frameworks actually run.

Lesson 6 of 8

Verifying what you have been told

Identification is a declaration. Verification is evidence. For arrangements the gap between the two is wider than anywhere else in due diligence, because the central document is private.

Registers help, within limits. Many jurisdictions now register beneficial ownership of companies, and some register trusts with a local connection. A register entry is evidence that somebody made a declaration, which is not the same as evidence that the declaration is true. Treat it as a cross-check, not as a conclusion.

The trust instrument is the primary document. Obtain it, and read the operative clauses rather than the summary: who may be added as a beneficiary, who may remove the trustee, what powers the settlor reserved, and whether any of it may be amended. Where only an extract is offered, record that an extract is what you have.

Corroborate from unconnected sources. A declaration from the trustee, confirmed by the deed, supported by a register entry and consistent with an independent search, is verified to a reasonable standard. The same declaration repeated by the provider who drafted it three times is one source.

Discrepancy reporting is now part of the work in several regimes. Where you find that what a register says differs from what your own enquiries show, there may be a duty to report the discrepancy, and in any case there is a reason to resolve it before relying on either.

Expect to use the company layer. An arrangement that holds shares in a company falls within that company's own ownership disclosure, so the arrangement surfaces at the corporate layer even when it is not registered itself. In Kenya, for example, a company's register of beneficial owners, and the information lodged under the beneficial ownership regulations, is one of the few routes by which a trust standing behind a local company becomes visible.

Verify the trustee as a party in their own right. A professional trustee is an entity with owners, a licence and a supervisor, and the standing of that firm is part of the risk of the structure.

Document the standard you applied. Which parties were verified, on what evidence, which remain declared only, and why that was acceptable given the risk. A file that distinguishes verified from asserted supports every later decision, including the decision not to act.

And accept that some structures cannot be verified to your standard. That is a finding, not a failure. The question it raises is whether to proceed, and that is a decision for a person with authority to take it, recorded with reasons.

Lesson 7 of 8

Keeping up with a structure that changes

Structures are not static, and almost nothing in a standard monitoring programme is designed to notice when one changes.

The events that matter are changes of party. A new trustee, a new protector, a beneficiary added or removed, a power transferred. Each alters who controls the arrangement, and none of them generates a transaction.

So ask on a cycle, and ask specifically. An annual confirmation that lists the parties you hold and asks whether any has changed is worth more than a general request for updated documents. Specific questions produce specific answers; a request to confirm that nothing has changed produces a signature.

Distributions are the transaction risk. A payment out of a structure is the point at which a beneficiary becomes identifiable and at which value reaches a person. Monitoring that treats distributions as routine internal movements misses the only moment the arrangement touches the real economy.

Watch for the structure's purpose drifting. A trust established to hold a family home that starts receiving third-party business receipts is doing something it was not described as doing. The original stated purpose is the baseline, which is a reason to record it in a form a monitoring team can read.

Deaths and divorces move control. Succession events are exactly when a structure is most likely to be restructured, and they are rarely reported to a financial institution by anybody.

Keep the records retrievable for the retention period. For arrangements the test is whether you could reconstruct, years later, who the parties were at a given date. That requires a dated history rather than a current snapshot, and it is a design decision in the record-keeping system.

Reviews should be triggered as well as scheduled. A change of provider, an unexplained distribution, adverse information about any party, or a jurisdictional change are each better triggers than a calendar date, and a framework that relies only on periodicity will always learn late.

And decide in advance what would make you exit. A structure that cannot be explained, a party who cannot be identified, or an answer that keeps changing. Writing the exit condition before it arises makes the decision an application of policy rather than an argument under pressure.

Lesson 8 of 8

One structure, followed through

A private bank is asked to open an account for a company. The documents are in order. What follows is the ordinary version of this work.

The first layer. The applicant is a company registered offshore, whose sole shareholder is another company, whose shares are held by a trustee company. The customer is therefore a trust, two layers down, and nothing on the application form says so.

The questions asked. Who settled the trust, with what assets, and where did those assets come from. Who are the beneficiaries, or how is the class defined. Is there a protector. Who pays the provider. What is the purpose of the arrangement, stated as a sentence a person could repeat.

What came back. A trust deed extract, naming a class of beneficiaries as the descendants of a named settlor. A protector, who turned out to be the settlor's business partner. A letter from the provider confirming it held current information. No source of wealth beyond a description of a family business.

The first finding. The settlor had reserved the power to add beneficiaries and to remove the trustee. The bank recorded that the settlor retained effective control notwithstanding the structure, and treated him as the beneficial owner for every purpose including screening.

The second finding. Screening the protector produced an adverse media match connected to public procurement in a third country. Not a sanctions hit, not a conviction, but a documented allegation.

What the bank did. It did not decline. It took the file to a committee with the structure diagram, the control analysis and the adverse information, and the committee set conditions: source of wealth evidenced to a documentary standard, no third-party receipts, distributions reported to the bank in advance, and an annual confirmation of parties with a named person accountable for obtaining it.

What happened eighteen months later. A distribution was made to a beneficiary who immediately transferred it to a company connected to the protector. That was visible only because the bank had required advance notice of distributions and had recorded who the protector was. It filed a report and exited the relationship.

The point of the example. Nothing was concealed by a forged document. Everything turned on reading the deed, identifying the person with reserved powers, and designing a monitoring condition at onboarding that would surface the one transaction that mattered. The work that produced the outcome was done before there was anything to find.

Knowledge check

Thirty questions

Answer all thirty questions, then check your answers. You need 27 out of 30 to pass and receive a certificate. If you score less, you will see which answers were right and wrong, and then go through the course again before you retake the check. Your answers, progress and times are kept only in this browser.

Standards and frameworks covered

The published frameworks this course teaches toward. Saolix is not affiliated with, endorsed by or accredited by any of these bodies; naming a standard is a citation, not a claim about its author.

Sources

The official documents this course relies on. Laws and guidance change, so check the current version.

  1. The FATF Recommendations, Recommendations 24 and 25 on transparency of legal persons and legal arrangements · Financial Action Task Force (FATF)
  2. Guidance on Beneficial Ownership of Legal Persons · Financial Action Task Force (FATF)
  3. Concealment of Beneficial Ownership · FATF and the Egmont Group of Financial Intelligence Units
  4. Guidance for a Risk-Based Approach for Trust and Company Service Providers · Financial Action Task Force (FATF)
  5. Guidance for a Risk-Based Approach for Legal Professionals · Financial Action Task Force (FATF)
  6. The Puppet Masters: How the Corrupt Use Legal Structures to Hide Stolen Assets · Stolen Asset Recovery Initiative, World Bank and United Nations Office on Drugs and Crime
  7. Regulation (EU) 2024/1624 on the prevention of the use of the financial system for money laundering, including the beneficial ownership of express trusts · Official Journal of the European Union, via EUR-Lex
  8. Proceeds of Crime and Anti-Money Laundering Act, 2009 (No. 9 of 2009) · National Council for Law Reporting, Kenya
  9. Companies Act, 2015 (section 93A, register of beneficial owners) · National Council for Law Reporting, Kenya
  10. Companies (Beneficial Ownership Information) Regulations, 2020 · National Council for Law Reporting, Kenya
  11. Register a trust as a trustee · HM Revenue and Customs, United Kingdom