UBO Mapping Under AMLR: The Complete Guide

From 10 July 2027, the Anti-Money Laundering Regulation (Regulation (EU) 2024/1624, AMLR) changes how beneficial ownership is identified across the EU. Ownership is no longer a single trace to the largest owner at each level. AMLR asks you to calculate ownership by accumulation, assess control in parallel, and name anyone who reaches 25%.

For Nordic teams the change runs deeper than the maths: national registers can no longer stand alone as the verification step. You determine ownership yourself, then check the register against your result. This guide covers what changes, the accumulation method in practice, why control is the part most processes miss, how the combination rule catches mixed structures, what replaces the register shortcut, and what it means for your team.

Key takeaways

• Ownership uses the accumulation method: multiply percentages down each chain, then add what each person holds across all chains (Art. 52).
• The threshold is 25% or more, not "more than 25%" (Art. 52). Exactly 25% now qualifies.
• Control is assessed separately (Art. 53); a person can be a beneficial owner through control with little or no equity.
• A combination rule catches people who own at one level and control at another.
• Registers are no longer your verification step : verify yourself, compare, and report discrepancies within 14 calendar days of detection.
• No owner found means identify and verify all senior managing officials, not one stand-in.
• Net effect: more named individuals per company, each flowing into identity verification, screening, and ongoing monitoring.

What AMLR changes about UBO identification

Most processes today follow the chain to the largest holder at each level and stop. AMLR makes that insufficient. Beneficial ownership is assessed through ownership, control, and the combination of the two, and a person caught by any one of them is a beneficial owner. A process that only multiplies ownership chains will under-identify beneficial owners against the regulation.

Ownership: the accumulation method and the 25% line

Under Art. 52, indirect ownership is calculated by multiplying the shares or voting rights held through each intermediate entity in a chain, then summing a person's holdings across every chain that involves them, against a threshold of 25% or more. The calculation is cumulative, so you cannot stop following a chain because one holding looks small. And the wording moved from "more than 25%" to "25% or more", so exactly 25% is now in scope. For certain higher-risk sectors, a delegated act may set the threshold below 25%.

Worked example. Shady Sharks Ltd. is the company being onboarded. Its ownership runs through two chains:

• Mammal Money Laundering Ltd. owns 30% of Shady Sharks. Mammal is owned by Corruption Carlson (80%) and Bribery Barnes (20%, a PEP).
• Llama Tax Havens Inc. owns 40% of Shady Sharks. Llama is owned 100% by Octopus Offshore Holdings, which is in turn owned 100% by Corruption Carlson.

Running the accumulation method on Corruption Carlson:

• Via Mammal: 80% × 30% = 24%.
• Via Octopus and Llama: 100% × 100% × 40% = 40%.
• Accumulated: 24% + 40% = 64%. Beneficial owner, and the ultimate beneficial owner of Shady Sharks.

The point is the accumulation. Assessed one chain at a time, Carlson looks like a 24% holder through Mammal and a separate 40% holder through the offshore chain. Only by adding the chains do you see the true 64%, sitting behind two layers of holding companies. Bribery Barnes, by contrast, accumulates to just 6% (20% × 30%) and falls below the 25% ownership threshold, but as a PEP still needs screening if named on the structure. None of this runs without first mapping the full structure: every holding company and individual with capital or voting rights. You cannot multiply across chains you have not mapped.

Control: the part most processes miss

Control is assessed independently of ownership (Art. 53). Someone can be a beneficial owner through control alone, with little or no equity, which is exactly why it is missed: these people rarely appear in ownership data. Triggers typically include a majority of voting rights, the right to appoint or remove most of the board or management body, and veto, profit, or asset-disposition rights. Softer forms (nominee arrangements, voting agreements, family influence) are a documented, case-by-case judgement.

Worked example. Dana holds 8% of TargetCo but, through a shareholders' agreement, has the right to appoint and remove a majority of the board. On ownership Dana is far below 25%; on control, Dana is a beneficial owner. It appears in no register, so it has to be asked directly at onboarding: voting agreements, appointment rights, nominee arrangements.

Combination: owning at one level, controlling at another

AMLR also catches mixed structures [confirm Art.]. Worked example. HoldCo controls TargetCo through a 55% stake. Priya owns 40% of HoldCo. Her indirect ownership of TargetCo is 22% (40% × 55%), below the threshold. But because HoldCo controls TargetCo and Priya holds 25% or more of that controlling entity, the combination rule names her a beneficial owner. Ownership maths alone would have stopped at 22% and missed her.

Why the register stops being your verification step

Consulting the register stays a requirement; being your verification step does not [confirm Art.]. Today, for many Nordic teams, a match between the company's declaration and the register is treated as enough. From July 2027 you determine and verify beneficial ownership through your own means first, then compare against the register.

Verification should rest on methods you can stand behind: a document check, qualified electronic identification, or other reliable sources, applied on a risk-proportionate basis with the reasoning recorded. This favours original documents: when your result comes from registry filings, articles of association, and shareholder lists, you can show exactly how you reached it and defend it. When it comes from a data provider whose method you cannot see, you cannot.

Where findings differ, you act: report discrepancies to the central register within 14 calendar days of detection, with a derogation allowing you to ask the customer to correct the register first in defined cases [confirm Art.].

When no beneficial owner can be identified

If, after exhausting all means, no beneficial owner can be found, AMLR requires you to identify and verify all of the company's senior managing officials, not a single stand-in. Senior managing officials are the executive members of the management body and the people responsible for day-to-day management. For a company with a large management body, that can mean several identity checks and screening runs where a previous process recorded one.

What this means for compliance teams

Every change points the same way: more named individuals per company. Control adds people ownership maths misses, accumulation pulls in split owners who look small in any one chain, combination catches mixed structures, and the fallback expands from one senior manager to all of them.

Every beneficial owner named carries the same downstream work: identity verification, screening against PEP and sanctions lists, and a place in the ongoing monitoring queue. Run manually or across disconnected tools, that volume compounds across the portfolio, and the result becomes hard to reconstruct when a regulator asks how you got there.

This is the connected-data expectation underneath AMLR. A UBO graph in one tool, screening in another, and monitoring in a third adds a manual handoff at every step. Preparing for AMLR means onboarding, screening, and monitoring on the same customer record, with every named person flowing automatically from the ownership map into screening and monitoring. It is the same operational lesson behind clearing a back book: see our guide to KYC remediation under AMLR.

Where UBO mapping breaks down in practice

The regulation is one thing; running it across a real portfolio is another. A few recurring problems decide whether a UBO process holds up under AMLR.

Cross-border chains do not trace themselves. Each jurisdiction means a different registry, in a different format, with different freshness and access. Chains that cross borders are time-consuming to follow, so under pressure they get traced partially, or not at all. The accumulation method only works if the whole chain is followed, wherever it leads.

Aggregate control slips through individual thresholds. Three shareholders each holding 18% do not individually trip a 25% flag, but together they control 54%. Without cumulative logic across related parties, a process clears the entity and misses the people behind it. The same applies to families with sub-threshold stakes, and to share classes that carry votes out of proportion to equity.

Some chains end at a locked door. Trusts and foundations are not always required to declare their controllers. When a chain ends at one, the answer is not a blank field: AMLR expects a documented, targeted effort to establish control, and a clear record of where the trail ends and why.

Ownership drifts after onboarding. A director is added, a shareholding is transferred, a new entity appears in the structure. None of it surfaces on its own. A UBO mapped at onboarding and left until the next periodic review is, in practice, a UBO you no longer know.

How Strise approaches UBO mapping under AMLR

Strise is built around the way AMLR expects beneficial ownership to be established: mapped from the company number to the ultimate natural persons, calculated, verified across multiple sources, and kept current, with the audit trail assembled as the work runs.

• Map the full structure automatically. From one company number, Strise traces upstream to the ultimate natural persons and downstream to subsidiaries, across jurisdictions and through holding layers, with diluted percentages calculated and every data source cited at each node.
• Cumulative ownership and real control. Direct and accumulated ownership are calculated at every level, so a 60% stake in a company that owns 50% of another shows as 30%, not 60%, against whatever threshold your policy sets. Family members with sub-threshold stakes are aggregated, and share classes are mapped by voting control, not just headline ownership.
• Verify across multiple sources. Registries, AI-read documents, and analyst-verified data are cross-referenced into one ownership record, every source cited and every change timestamped. When a registry hits a wall, Strise generates a targeted document request, then extracts and maps the response with no re-entry.
• Stay current. Changes to ownership, directors, or structure are picked up automatically and trigger a review when they affect risk, rather than waiting for the next cycle. Every named person flows into PEP and sanctions screening and ongoing monitoring on the same record.

In practice that means around 80% of routine cases handled automatically, complex cases resolved in about 30 minutes rather than three hours, and roughly six times more complex cases per analyst with no new hires, drawing on data from more than 250 jurisdictions. See how UBO discovery works.

An AMLR UBO mapping checklist

• Map the whole structure first. Every owner and holding company, every level, ownership and voting rights recorded separately. You cannot calculate across chains you have not mapped.
• Run ownership by accumulation. Multiply down each chain, add across chains, and apply the 25% or more threshold (Art. 52).
• Assess control in parallel. Voting rights, appointment rights, veto and profit rights, and softer arrangements (Art. 53).
• Check for combination. Anyone who owns at one level and controls at another .
• Ask about control at onboarding. Voting agreements, appointment rights, nominee arrangements.
• Verify yourself, then check the register, and report discrepancies within 14 calendar days.
• No owner found? Name every senior managing official, not one stand-in.
• Connect each named person to the same record. Screening and ongoing monitoring, kept current as the structure changes.

‍

-

‍
Published: 26 June, 2026. This guide is published by Strise.

See AMLR-ready UBO mapping in action. Book a demo with Strise →