Shell companies, nominees and the corporate veil

Students are sometimes surprised that I refuse to call shell companies illegal. Half the group structures any of us have audited contain them for perfectly ordinary reasons. The investigative question is never whether a shell exists, but what work it is doing. A shell companyShell companyLegal entity with no significant operations, used as a holding or transaction vehicle. is a legal entity with no significant operations, employees or physical presence. Shells are not per se illegal, they are used lawfully as holding vehicles in group structures, as joint-venture SPVs (special-purpose vehicles) to ring-fence project risk, as intellectual-property licensing hubs, as securitisation issuers, and as dormant reserves ready to be activated when a commercial opportunity arises. Their misuse in laundering derives from three inherent properties. First, they have full legal personality: they can hold assets, contract, sue and be sued, open bank accounts, take title to real property, and be counterparty to financial instruments. Second, they can be incorporated cheaply and quickly, particularly in jurisdictions with light-touch beneficial-ownership rules — often in under 24 hours, at fees under USD 2,000, and with no meaningful physical presence requirement. Third, their apparent ownership can be layered behind further shells, trusts, foundations and nominee arrangements, so that the beneficial-ownership chain terminates in a jurisdiction that will not disclose to the investigating state.

“THE ANATOMY of a shell companyShell companyLegal entity with no significant operations, used as a holding or transaction vehicle. built for opacity typically has six distinguishable components.

RELATED VARIANTS

A SHELF COMPANY is a shell deliberately AGED by a CSPCSPCorporate Service Provider, professional firm that incorporates and administers companies for clients. so it can be sold with a history. The CSPCSPCorporate Service Provider, professional firm that incorporates and administers companies for clients. incorporates the entity, files annual returns (often nil), sometimes opens a bank account, and holds the entity dormant for one, three or ten years. On sale, the buyer acquires an entity with an "established" registration date, filed accounts, and — in some cases — a bank account already opened, which materially accelerates its use for onward transactions. Shelf companies are legal in most jurisdictions but their use to disguise novelty is itself a red-flag pattern.

A FRONT COMPANY differs from a shell in that it does conduct real business, but the real business is a cover for the illicit activity. Cash-intensive front businesses (car washes, restaurants, taxi fleets, laundromats) commingle illicit cash with genuine takings; front consultancies invoice for "strategic advice" that never occurs. Fronts are harder to detect than shells because they present operational documentation (invoices, payroll, VAT returns), but sector benchmarking (revenue per employee, revenue per square metre, utility consumption per unit revenue) reliably identifies outliers.

A SHELF TRUST or FOUNDATION extends the shell logic to non-corporate structures. Discretionary trusts (common law) and foundations (civil law) can hold assets, transact, and shield beneficial identity behind trustee/foundation-council opacity. The protector; a party with veto or removal rights over the trustee, is frequently the true controller and is not always disclosed on public documents.

THE INVESTIGATOR'S TASK is to look BEHIND the register: to identify the CSPCSPCorporate Service Provider, professional firm that incorporates and administers companies for clients., the nominee's principal, the ultimate beneficial owner (UBOUBOUltimate Beneficial Owner — the natural person who ultimately owns or controls an entity.). Techniques include: cross-referencing the CSPCSPCorporate Service Provider, professional firm that incorporates and administers companies for clients. against known typologies (some CSPs have been publicly named repeatedly in leaks); matching signatures across filings to identify a single hand behind multiple entities; using leaked datasets alongside registry data to overlay declared and undeclared connections; requesting information from the jurisdiction's FIU via Egmont Secure Web, and — where a live commercial relationship exists — bringing a "know your customer" pressure through counterparties who will require BO evidence as a condition of continued dealings.

I have watched investigators lose good cases by presenting an ownership chart as if every line on it were proved. Mark the inferred edges differently from the confirmed ones, in your own working papers first. Beneficial-ownership investigation is a discipline of TRIANGULATION. No single source suffices. The investigator's craft is to accumulate corroborating evidence from as many independent channels as possible, to weight each channel by its reliability, and to distinguish confirmed edges in the ownership graph from inferred ones. The output is not a certainty but a defensible reconstructed chain from the entity of interest to the natural person(s) who ultimately own or control it.

CORE PUBLIC SOURCES.

National corporate registries. Disclosure varies enormously. UK Companies House is exemplary — free, structured, machine-readable, with directors, secretaries, addresses, filed accounts, and (since 2016) the Persons with Significant Control (PSC) register, and it maintains full history. Singapore's ACRA and New Zealand's Companies Office are similarly rich. Many jurisdictions publish only directors and registered office and lock accounts, UBO or share history behind fees or in-person applications. Some jurisdictions publish nothing at all beyond incorporation confirmation. An investigator maps early which of the registries in a jurisdictional chain will yield what.

Beneficial-ownership registers. FATF Recommendation 24 as amended in 2022 requires jurisdictions to ensure adequate, accurate and up-to-date BO information is accessible to competent authorities in a timely manner. Implementation is highly uneven. The EU's mandated central registers (AMLD4/5) had, until the 2022 CJEU judgment in Luxembourg Business Registers, been generally public — that judgment held that unrestricted public access disproportionately interfered with fundamental rights to privacy and data protection and struck down the general-public-access provision, though competent-authority and legitimate-interest access remain. The U.S. Corporate Transparency Act (2021, reporting from 2024) delivers a FinCEN-held register accessible to law enforcement, tax authorities in defined circumstances, and financial institutions with customer consent; though FinCEN's March 2025 interim final rule removed domestic entities from the reporting population, so the register now captures foreign-formed entities registered in the U.S. rather than U.S.-formed shells. Africa's OGP (Open Government Partnership) commitments have driven Nigerian, Ghanaian and Kenyan BO registers with varying access rules. The investigator must know, for each jurisdiction in the chain, whether they can access the BO register, on what legal basis and after what request.

Tax administration data. Rich, high-integrity, but access is tightly constrained by tax-secrecy law. Access requires either (i) a domestic lawful-basis authority (a court order, an FIU-to-tax gateway, a specific inter-agency memorandum) or (ii) an international exchange-of-information channel (Common Reporting Standard, Exchange of Information on Request, or Country-by-Country Reporting under BEPS Action 13). CRS is passive: information arrives annually on a self-report basis from participating foreign financial institutions. EOIR is targeted: the investigating state formulates a specific request under a bilateral treaty or the MAAC and receives a response, typically within 90 days.

Court filings. Litigation exposes ownership in dispute. Insolvency proceedings, shareholder disputes, divorce petitions, and enforcement of foreign judgments often surface UBO information not otherwise on public record. Court files are geographically dispersed and typically not centrally indexed; investigators develop a local network of court-record retrievers or use commercial services (PACER in the U.S., Companies House-linked litigation trackers in the UK, LexisNexis in multiple jurisdictions).

Land and property registries. In many jurisdictions still the single richest source for integration-stage evidence. UK Land Registry data is machine-readable and searchable by owner; French, Spanish, and German cadastres are querable through professional channels. Real-property title chains reveal purchase price, funding source (mortgage lender), and, increasingly — beneficial ownership behind corporate title holders where UBO transparency has been extended to property.

Sanctions and adverse-media databases. OFAC, EU consolidated list, UK OFSI, UN sanctions committees. Commercial screening providers (World-Check, Dow Jones Risk & Compliance, LexisNexis Bridger) aggregate PEP lists, adverse media, and enforcement actions. These databases are not authoritative for ownership per se but frequently surface network connections that trigger further investigation.

LEAK-DERIVED DATASETS (use with care and respect for source jurisdiction laws; leaked data may be inadmissible in some proceedings but is universally useful for investigative lead-generation):

ICIJ Offshore Leaks Database aggregating Panama Papers (2016), Paradise Papers (2017), Pandora Papers (2021), FinCEN Files (2020) and predecessors. Free, searchable, structured.

“OpenCorporates aggregates registry data across ~140 jurisdictions into a single searchable graph.

OpenOwnership publishes structured beneficial-ownership data using the BODS (Beneficial Ownership Data Standard) schema.

Aleph (by OCCRP) integrates leaks, registries, sanctions and court filings across the OCCRP investigative network.

Sayari, Kharon, Refinitiv provide commercial equivalents with deeper coverage and analyst-grade tooling.

INVESTIGATIVE MOVES.

1. Start with the target entity; extract every named individual (director, secretary, agent, address) and every named entity (parent, shareholder, related company). This becomes your initial node set.

2. Reverse-search each individual and each entity across every registry and leaked dataset available. Look for clustering: same nominee, same address, same CSP, same accountant, same lawyer, same bank. Clustering is the strongest inductive signal that entities apparently unrelated on any single register are in fact part of a common structure.

3. Map connections in a graph. Paper is fine for small cases; larger cases use Miro or Lucidchart for narrative diagrams, Maltego or i2 Analyst's Notebook for professional link-analysis, Neo4j or Palantir for enterprise-scale. Distinguish CONFIRMED edges (evidenced by primary documents) from INFERRED edges (evidenced by circumstantial clustering) — a distinction that will matter enormously if the chart ever enters evidence.

4. Where the chain terminates in a jurisdiction that will not disclose, formulate an MLATMLATMutual Legal Assistance Treaty, formal instrument for cross-border evidence gathering. or FIU-to-FIU information request via Egmont Secure WebEgmont Secure WebThe Egmont Group's secure FIU-to-FIU information-exchange platform.. FIU-to-FIU requests are informal, fast (often days), but produce intelligence not evidence. MLATs are formal, slow (often 6–18 months), but produce admissible evidence. Where the tax administration is involved, exchange under the MAAC is a third channel; where an international investigation is joint, participation in a Joint Investigation Team (JIT) can dispense with request formalities entirely.

5. Cross-check every step against ALTERNATIVE INNOCENT EXPLANATIONS. Every corroborating fact should be tested against the "boring" hypothesis, the entity is a legitimate holding structure of a legitimate business; before being accepted as evidence of illicit intent.

A COMMON ANALYTICAL TRAP: assume that a common address = common ownership. In many offshore jurisdictions a single CSP's registered-office address is used by thousands of unrelated entities. Common addresses are a starting hypothesis, not a conclusion. Similarly, a common nominee director is a hypothesis of common CSP, not necessarily common UBO, though in practice the two often coincide.

A SECOND TRAP: treating leaked data as complete. The Panama Papers exposed the client files of a single law firm (Mossack Fonseca) and the Pandora Papers exposed 14 offshore service providers. The universe of offshore CSPs is much larger. Absence of a name in a leak does not exclude offshore structuring; presence does not prove wrongdoing.

A THIRD TRAP: overconfidence in a "final" beneficial owner. UBO regimes typically identify the person at ≥25% or with effective control, but genuine control can vest below the threshold (a 24% shareholder with veto rights; a manager with day-to-day authority), and true beneficial ownership can be split across family members below individual thresholds. The investigator's chain should identify not one UBO but the network of persons who benefit from and control the structure.

The output of a well-executed BO investigation is a legally defensible narrative — corroborated at every step by independent sources, honest about the limits of each inference, and directly connectable to the evidential steps that would move it from intelligence to proof. That is the standard the tribunal will apply when the chart is projected on the courtroom wall.

Trusts and foundations are the second great opacity architecture, sitting alongside shells. A trust is a fiduciary relationship in which a settlorSettlorThe person who creates a trust by transferring property to a trustee. transfers property to a trustee who holds it for the benefit of one or more beneficiaries. A foundation (Stiftung, Anstalt) is a hybrid civil-law construct that operates like a legal person but functions like a trust. Both have entirely legitimate uses — estate planning, charitable endowment, asset segregation for minors or vulnerable adults, protection against political-risk expropriation. Both are also, in laundering contexts, the most powerful concealment devices in the professional's toolkit, because they can dissociate legal ownership (the trustee), equitable ownership (the beneficiary), and effective control (the settlorSettlorThe person who creates a trust by transferring property to a trustee. or the protectorProtectorA person with reserve powers over a trust — often to veto distributions or replace the trustee.).

The investigative problem is that the "beneficial owner" concept, designed for corporate structures, does not map cleanly onto a trust. FATF Recommendation 25 (2012, revised 2022) requires jurisdictions to obtain identity information on the settlorSettlorThe person who creates a trust by transferring property to a trustee., trustee, protectorProtectorA person with reserve powers over a trust — often to veto distributions or replace the trustee., beneficiaries or class of beneficiaries, and any other natural person exercising ultimate effective control. In practice, however, this five-part list is where each element hides: the settlorSettlorThe person who creates a trust by transferring property to a trustee. may be a deceased matriarch; the trustee may be a corporate trustee in a secrecy jurisdiction; the protectorProtectorA person with reserve powers over a trust — often to veto distributions or replace the trustee. may be a lawyer bound by privilege; the beneficiaries may be an unnamed class ("descendants of the settlorSettlorThe person who creates a trust by transferring property to a trustee."), and the effective controller may be someone with an informal, unwritten influence over the trustee.

Foundations aggravate the problem because civil-law jurisdictions (Liechtenstein, Panama, Curaçao, the Netherlands Antilles) permit foundation charters that name only classes of beneficiary and reserve extensive powers to a foundation council chosen by the founder. The result is a legal person with no shareholders, no register of beneficiaries, and a governance body that reports to no one.

Investigative technique proceeds along four fronts. First, obtain the trust or foundation constitutional document, the deed, the charter, the letter of wishesLetter of wishesA non-binding document from the settlor to the trustee expressing intended treatment of the trust — often the most probative single document.. In a laundering case this typically requires an MLAT to the jurisdiction of the trustee or fiduciary. Second, obtain trustee-held records: minutes, distributions, correspondence. Third, follow the money out; beneficiary distributions are the moment the fiduciary shield lifts, because the recipient is a natural person receiving an identifiable payment. Fourth, examine the letter of wishesLetter of wishesA non-binding document from the settlor to the trustee expressing intended treatment of the trust — often the most probative single document., a non-binding but almost universal document in which the settlorSettlorThe person who creates a trust by transferring property to a trustee. expresses their intentions to the trustee. Letters of wishes routinely reveal effective control, and are often the most probative document in a trust-laundering case.

Correspondent banking is the plumbing of the international payment system. A correspondent bankCorrespondent bankA bank in Country A that holds an account for a respondent bank in Country B, providing access to Country A's payment system. in Country A holds an account for a respondent bank in Country B, allowing the respondent's clients to send and receive value across jurisdictions and currencies without themselves needing a local presence. The system is indispensable to international trade — and structurally exposed to laundering because the correspondent bankCorrespondent bankA bank in Country A that holds an account for a respondent bank in Country B, providing access to Country A's payment system. sees only its respondent, not the respondent's clients.

The nested correspondentNested correspondentA downstream bank using another bank's correspondent access, hidden from the correspondent's view. problem multiplies the exposure. A respondent bank in Country B may itself act as correspondent for a further downstream bank in Country C. From the correspondent in Country A's perspective, only Country B's respondent is visible; the Country C bank — and, downstream of it, that bank's clients, are hidden. Every laundering enforcement action of the last decade with a "correspondent" fact-pattern (HSBC 2012, Standard Chartered 2012 and 2019, Danske Bank 2018, Deutsche Bank 2020) has featured nested access as an aggravating structural feature.

FATF Recommendation 13 and the Wolfsberg Correspondent Banking Principles set out the response. The correspondent bankCorrespondent bankA bank in Country A that holds an account for a respondent bank in Country B, providing access to Country A's payment system. must undertake enhanced due diligence on its respondent: understand the respondent's AML controls, business type, geographical footprint, regulatory status, and; critically, whether the respondent itself provides downstream correspondent services to other banks. Nested arrangements must be disclosed by the respondent and, where the correspondent judges the downstream risk unacceptable, refused. Payable-through accounts, a related and more dangerous variant in which the respondent's clients transact directly through the correspondent's account, require even stricter controls and are prohibited in some jurisdictions.

De-riskingDe-riskingThe wholesale exit by a correspondent bank from an entire respondent-country market judged high-risk. — the wholesale exit by major correspondents from entire respondent-country markets judged high-risk — is the other face of the same problem. When correspondents withdraw, respondent banks lose access to the international payment system, remittance corridors collapse, and legitimate trade contracts. FATF, the Financial Stability Board and the G20 have repeatedly cautioned against blanket de-riskingDe-riskingThe wholesale exit by a correspondent bank from an entire respondent-country market judged high-risk. and encouraged case-by-case assessment; in practice, the incentives favour exit. The reforming official's task in a country experiencing de-riskingDe-riskingThe wholesale exit by a correspondent bank from an entire respondent-country market judged high-risk. is to make the case for retention through demonstrable AML uplift, enhanced supervision, credible enforcement, transparent reporting; precisely because the alternative, exclusion from correspondent networks, is developmentally disastrous.