An ownership chart is useful only when the bank can explain what each connection means and what supports it. Complex groups often combine shareholdings, voting arrangements, trusts, nominees and management rights. Treating every line as an identical ownership relationship makes a neat chart unreliable. The objective is a defensible understanding of the people who ultimately own or control the customer under the applicable framework.

The FATF guidance on legal-person beneficial ownership supports obtaining accurate, current information from complementary sources. UK CDD requirements in regulation 28 provide the relevant banking starting point. The workflow below is an analytical method for organising evidence, rather than a universal legal ownership test.

Describe each relationship before calculating percentages

Start with the actual customer and work outward. For each connected entity or person, record jurisdiction, legal form, identifier, relationship type and evidence date. Distinguish economic ownership from voting rights and other forms of control. A company registration number identifies an entity; it does not explain who controls it.

Use an evidence record for every material link. It should identify the source, the date retrieved, any relevant reporting period and the reviewer’s conclusion. If a customer provides a chart, treat it as an assertion to verify proportionately, not as independent confirmation of itself.

Percentage calculations can help trace indirect economic interests, but they are not a complete control analysis. Joint arrangements, appointment rights and other mechanisms may matter under applicable law. Document the legal test being applied and why particular facts satisfy or fail that test. Different compliance regimes may require separate conclusions from the same underlying facts.

Resolve contradictions explicitly

Suppose a registry extract shows one shareholder while a signed customer declaration shows another. Do not quietly select whichever source is easier to process. Establish whether the difference reflects timing, an unregistered transaction, an error or a more substantive concern. Retain both records and document how the discrepancy was resolved.

Evidence quality should be assessed in context. A recently obtained document may describe an old position. A formal document may omit a relevant side agreement. A credible independent source may still have coverage limitations. The reviewer should explain what remains uncertain and the consequence for the relationship decision.

Where a legal framework permits a fallback when no beneficial owner is identified, the bank should document why that route is available and what reasonable measures were attempted. A fallback designation should never conceal an unresolved ownership question or replace the analysis required by law.

A hypothetical layered corporate customer

Imagine a corporate customer held by two intermediate companies in different jurisdictions. One individual has an indirect economic stake through one branch, while another has contractual rights affecting the appointment of directors. The account team initially records only the indirect shareholder.

An effective review records the distinct rights and investigates their relevance under the governing legal tests. It assesses the documentary support, identifies any missing agreement and reaches a reasoned decision. It does not assume that a mathematical shareholding calculation settles every question of control.

The approved record should explain both the result and its limitations. It can identify information requiring refresh after a restructuring, a change in directors or an expiry of an agreement. This makes subsequent reviews more efficient because the next reviewer can see which facts were decisive.

Keep AML ownership and sanctions control connected but distinct

Sanctions analysis may require a different legal test from AML beneficial ownership. OFSI’s UK guidance explains that relevant restrictions can extend to entities owned or controlled by designated persons even when those entities are not themselves listed. Passing an AML percentage threshold is therefore not proof that sanctions analysis is complete.

Use one controlled factual record to support the separate legal assessments. This avoids duplicate research while preserving differences in scope and decision logic. Record who approved each conclusion and what changes would require reassessment.

Quality assurance should challenge the most consequential connections, not merely check whether a chart exists. Useful tests include tracing a sampled ownership link to its source, checking unresolved discrepancies and confirming that all relevant parties reached downstream screening. The result is a customer record that another competent reviewer can understand without reconstructing the entire investigation.