Who Counts as a Customer?

Abstract Global anti-money laundering (AML) regimes produce a structurally stable Hollow Win (C=0, A=1, B=1) because the legal definition of 'customer' terminates at the legal person rather than the beneficial owner. The flawed bilateral game G between financial institutions (A) and the state (B) excludes system welfare C—financial integrity, rule of law, and market trust—from the payoff space. Both parties maximize private gain: banks earn fee revenue from opaque legal entities, and regulators claim enforcement credit through audits and fines, while illicit financial flows continue unabated. The system welfare beta is bounded at β W ≥ 5.0, meaning each dollar of global AML compliance industry revenue ($180B annually) is associated with at least five dollars of system welfare destruction ($700B+ in un-interdicted laundering). This is an Intractability Theorem domain: the constraint is institutional, not physical. The game change R shifts the identification anchor from legal personhood to beneficial control, with Denmark's Central Business Register—featuring integrated civil-register verification and administrative dissolution powers—as proof-of-concept. Policy implication: AML customer due diligence must be redefined as beneficial-owner verification against authoritative central registers, eliminating simplified due diligence for legal entities and replacing box-checking...

6. Decision Accounting: The 2015 EU 4th AML Directive The 16-field Decision Accounting (DA) record (Postnieks, 2026a) provides a structured framework for evaluating regulatory decisions. Fields 1–15 derive from regulatory convergence across 16 regimes (α=0.91), while Field 16 (SYSTEM WELFARE) is the framework's structural addition—it is non-derivable from the bilateral payoff space and is the heart of the contribution. This section applies the DA record to the European Union's decision, in the 4th AML Directive (2015/849), to adopt a 25% beneficial ownership threshold. The decision is the foundational regulatory choice that determines the game structure for 27 member states and, through the EU's market power, for much of the global financial system. Field 1: WHO. The European Commission (Directorate-General for Financial Stability, Financial Services and Capital Markets Union), the European Parliament (Committee on Economic and Monetary Affairs), and the Council of the European Union (Economic and Financial Affairs Council). The trilogue negotiations involved approximately 50 officials from the three institutions. Field 2: WHAT. The decision to define "beneficial owner" as any natural person who ultimately owns or controls more than 25% of the shares or voting rights of a legal entity. The threshold was adopted over alternatives including 0% (no threshold), 10%, 15%...

contribution here is to prove that the boundary between legal personhood and beneficial control is the structural flaw that produces the Hollow Win, to bound the system welfare beta for the AML compliance paradigm, and to identify the game change that transforms the equilibrium. Section 2 identifies the flawed game and maps it to the 8-outcome taxonomy. Section 3 states the theoretical framework formally, proves the boundary result, and bounds β W. Section 4 reviews empirical evidence spanning scandals, field experiments, and enforcement data. Section 5 identifies the game change and presents real-world proof-of-concept. Section 6 applies the 16-field Decision Accounting record to the foundational threshold decision. Section 7 derives implications. Section 8 concludes. ---

7. Implications If the AML Boundary Collapse Theorem is accepted, three things change in regulatory design, political economy, and the structure of global finance. Regulatory design. The default AML architecture must shift from legal-person customer identification to beneficial-owner verification against authoritative registers. Customer due diligence should not be a box-checking exercise directed at corporate vehicles; it should be a verification process directed at natural persons. The three components of the Danish model—integrated register, automatic verification, and dissolution power—should become the global minimum standard. Simplified due diligence, which permits reduced verification for low-risk customers, should be eliminated for legal entities because legal entities are inherently high-risk vehicles for boundary exploitation. Political economy. The Hollow Win persists because it is politically efficient. It allows governments to claim action against money laundering while preserving the revenue that financial centers earn from handling opaque flows. It allows banks to claim compliance while preserving fee income from customers they know to be suspect. It creates a compliance industry—lawyers, consultants, software vendors, training providers—whose revenue depends on the continued production of compliance artifacts that do not change the equilibrium. The $180 billion...

Conclusion The AML customer-identification game produces a stable Hollow Win because "customer" is defined at the legal-personhood level, structurally excluding system welfare from the bilateral payoff space. Financial institutions earn fees, regulators claim enforcement credit, and the system absorbs the welfare cost of unchecked illicit finance. The theorem is an Intractability result: the constraint is institutional, not physical, and a transformation exists that produces a better equilibrium. The empirical evidence is overwhelming and convergent. The Danske Bank Estonia case processed €200 billion through legal entities whose beneficial owners were invisible to the compliance system. The Panama Papers revealed 214,000 offshore entities designed to exploit the boundary between legal personhood and beneficial control. The FinCEN Files showed banks filing suspicious activity reports on transactions they continued to process, generating regulatory cover while preserving fee income. The Global Shell Games field experiment proved that anonymous shell companies are readily available in virtually all jurisdictions, including those with the strictest AML rules. FATF mutual evaluations consistently identify beneficial ownership as the critical deficiency across decades of standard-setting. The enforcement data confirm the futility: only 0.2% of laundered proceeds are seized...

structures, wasted compliance spending, and a false sense of progress. The difference between right and wrong was approximately $100–200 billion annually in continued illicit flows through EU financial centers. Field 12: CONSTRAINTS. The decision-makers faced constraints: (1) political opposition from member states with large corporate services sectors; (2) lobbying by the financial services industry, which argued that lower thresholds would drive business to non-EU jurisdictions; (3) the absence of a centralized EU beneficial ownership register, which made verification of any threshold difficult. The decision-makers did not consider the constraint of threshold-evasion strategies, despite their well-documented existence. Field 13: UNCERTAINTY. The decision-makers were uncertain about: (1) the distribution of beneficial ownership stakes in EU companies; (2) the elasticity of illicit flows with respect to threshold levels; (3) the compliance costs of lower thresholds; (4) the welfare consequences of the 20% gap. The impact assessment acknowledged these uncertainties but did not model them quantitatively. The decision was made under conditions of uncertainty that favored the status quo. Field 14: COMMUNICATION. The decision was communicated through the Official Journal of the European Union and standard legislative channels. The European Commission published a press release...

Palan, R., Murphy, R., & Chavagneux, C. (2010). Tax havens: How globalization really works . Cornell University Press. Picciotto, S. (2011). Regulating global corporate capitalism . Cambridge University Press. Platt, S. (2015). The money laundering machine: How the financial system enables crime . Palgrave Macmillan. Postnieks, E. (2026a). The Missing System Theorem. SAPM Working Paper No. 1. Postnieks, E. (2026b). Decision Accounting. SAPM Working Paper No. 2. Postnieks, E. (2026c). Conflictoring. SAPM Working Paper No. 3. Postnieks, E. (2026d). The Game-Change Conjecture. SAPM Working Paper No. 4. Postnieks, E. (2026e). Reform Dividend. SAPM Working Paper No. 5. Sharman, J. C. (2011). The money laundry: Regulating criminal finance in the global economy . Cornell University Press. Sharman, J. C. (2017). The despot's guide to wealth management . Cornell University Press. Shaxson, N. (2011). Treasure islands: Tax havens and the men who stole the world . The Bodley Head. Shaxson, N. (2018). The finance curse: How global finance is making us all poorer . The Bodley Head. Sikka, P. (2003). The role of offshore financial centres in globalization. Accounting Forum, 27 (4), 365–399. Singer, P. W. (2008). Corporate warriors: The rise of the privatized military industry . Cornell University Press. Slemrod, J., & Wilson, J. D. (2009). Tax competition with parasitic tax havens. Journal...

Falsification Conditions The AML Boundary Collapse Theorem is falsifiable. The following observations would refute it: Falsification Condition 1 (FC1): A jurisdiction with θ = 0 that achieves a system welfare beta β W < 1.0 for a sustained period of at least five years, with independently verifiable data showing that illicit financial flows are interdicted at rates exceeding 10% of estimated flows. This would demonstrate that the legal-personhood boundary does not necessarily produce the Hollow Win, contradicting the theorem's claim that θ = 0 structurally excludes system welfare. Falsification Condition 2 (FC2): A randomized controlled trial in which financial institutions assigned to θ = 0 with enhanced disclosure requirements (additional SAR fields, more frequent audits, mandatory training) achieve statistically significant reductions in illicit financial flows compared to a control group with standard θ = 0 requirements, with effect sizes exceeding 20% and p-values below 0.01. This would refute the Disclosure Futility theorem by showing that information-only interventions can shift the equilibrium. Falsification Condition 3 (FC3): Evidence that the 25% beneficial ownership threshold in the EU 4th AML Directive has been ineffective at enabling evasion—specifically, that corporate service providers cannot engineer ownership structures below the threshold, or that such...

1. Introduction Between 2007 and 2015, Danske Bank's Estonian branch processed approximately €200 billion in transactions through non-resident accounts held primarily by UK-registered limited liability partnerships and anonymous shell companies with Russian beneficial owners. The bank complied with every applicable AML rule. Each account holder was a legal entity with a registration certificate, a nominal director, and a business address. The bank conducted customer due diligence, verified legal existence, and filed suspicious activity reports when transactions exceeded internal thresholds. The Danish Financial Supervisory Authority inspected the branch repeatedly and found no material deficiencies. Yet the system degraded: the flows laundered through those accounts have been linked to the Russian Laundromat, the Azerbaijani laundromat, and transnational corruption networks spanning the former Soviet Union. In 2018, Danske Bank's own investigation, conducted by Bruun & Hjejle, concluded that the branch had functioned as a conduit for suspicious transactions because the "customer," as defined by AML rules—the legal entity—bore no resemblance to the economically meaningful party (Bruun & Hjejle, 2018). This is not a compliance failure in the conventional sense. It is a design failure rooted in a boundary problem. The boundary separates legal personhood from beneficial control...

in the Estonian financial system, the reputational damage to Nordic banking, and the enabling of kleptocratic regimes. A more comprehensive estimate, including the multiplier effects of corruption on governance quality, places the system welfare cost at approximately €5–10 billion. Using the midpoint: €7.5 billion. β W calculation: Π = €1.5 billion (Danske's fee revenue from the portfolio). W = €7.5 billion (conservative system welfare destruction). β W = 7.5/1.5 = 5.0. This is consistent with the paper's headline bound. Axiom demonstrated: Axiom 3 (Competitive Pressure). Danske Bank's internal documents revealed that the bank feared losing the non-resident portfolio to competitors—primarily Swedbank and SEB, which operated similar programs in Estonia. The bank's compliance officers warned that rejecting opaque customers would shift the business to competitors who would accept them. The competitive pressure made unilateral deviation irrational. The Hollow Win was stable because any bank that attempted to verify beneficial owners would lose revenue to banks that did not. Commentary: The Danske case is the canonical empirical demonstration of the AML Boundary Collapse Theorem. The bank complied with every applicable rule. The regulator inspected and found no deficiencies. The system welfare destruction was massive. The boundary between legal personhood and beneficial control was...

The 25% threshold was a political compromise that maximized the appearance of progress while minimizing disruption to the corporate services industry. The system welfare cost was structurally excluded from the analysis. The decision was a Hollow Win: the EU claimed credit for transparency, the financial industry avoided costly verification requirements, and the system continued to degrade. The DA record also demonstrates the value of Field 16 as a structural addition to regulatory decision-making. If the 2013 impact assessment had been required to estimate the SYSTEM WELFARE of the 25% threshold—including the welfare cost of evasion strategies, the opportunity cost of misdirected compliance spending, and the dynamic effects on illicit flows—the decision might have been different. The Five-Minute Test (Postnieks, 2026a) asks whether a reader unfamiliar with the situation can reconstruct what was decided, by whom, why, under what authority, and what would trigger reconsideration. The DA record for the 25% threshold decision passes the test for Fields 1–15 but fails on Field 16: the SYSTEM WELFARE is not reconstructible from the available documentation. This failure is not accidental; it is structural. The bilateral game G excludes C by design.

Case 1: Danske Bank Estonia (2007–2015) — The €200 Billion Hollow Win Entity: Danske Bank A/S, Estonian branch; Danish Financial Supervisory Authority (DFSA); Republic of Estonia Period: 2007–2015 (peak non-resident portfolio operations); 2018 (Bruun & Hjejle report publication) Facts: The Estonian branch processed approximately €200 billion in transactions through non-resident accounts held by UK-registered limited liability partnerships, Scottish limited partnerships, and British Virgin Islands companies. The branch employed approximately 400 staff, of whom roughly 30 were in compliance roles. The non-resident portfolio generated approximately €1.5 billion in fee income over the period, representing roughly 15% of Danske Bank's total group profit during those years. The bank's own investigation found that 6,200 of 10,000 non-resident customers were suspicious, with 2,000 classified as "high risk" and 1,200 as "very high risk." Despite this, the bank continued the relationships because the legal-entity customers passed shallow due diligence checks: registration certificates, nominal directors, and business addresses were all verifiable. Bilateral payoff analysis (A = Danske Bank, B = DFSA): Party A (Danske Bank) gain: €1.5 billion in fee revenue (Π A = revenue, not profit). The bank's compliance cost for the non-resident portfolio was approximately €50 million over the...

between legal personhood and beneficial control. When a UK limited liability partnership opens an account at Danske Bank Estonia, the bank's customer is the LLP. The LLP has a legal existence, a registered address, and a filing history at Companies House. The bank can verify all of this. What the bank cannot see—what the AML rulebook does not require it to see with sufficient rigor—is the natural person who placed the money, who controls the transactions, and who benefits from the proceeds. The beneficial owner may be a Russian government official, a drug trafficker, or a sanctions target. The legal entity is a mask. The bank's compliance system is designed to inspect the mask, not the face behind it. This is not an information problem in the ordinary sense. The information exists in principle: someone knows who the beneficial owner is. The problem is that the game structure does not require the information to be produced, verified, or acted upon. The bilateral payoff functions Π A and Π B are both maximized at a low level of verification. Deep verification is costly for the bank and politically costly for the regulator, because it reveals the scale of the problem that neither party has the institutional capacity to solve. The conflictoring concept (Postnieks, 2026b) illuminates an additional structural feature: the financial institution and the corporate services provider...

outputs are produced regardless of whether beneficial ownership is actually identified. The first-order condition for the state is: ∂E[u B]/∂b = ∂Π B/∂b - ∂c B/∂b + ∂E[credit B]/∂b = 0 At b = low enforcement: ∂Π B/∂b ≈ 0 (low political credit), ∂c B/∂b ≈ 0 (low cost), ∂E[credit B]/∂b ≈ 0 (few visible outputs). At b = medium enforcement: ∂Π B/∂b > 0 (audits produce statistics), ∂c B/∂b > 0 (budget allocated), ∂E[credit B]/∂b > 0 (fines publicized). At b = high enforcement: ∂Π B/∂b ≈ 0 (diminishing returns to enforcement statistics), ∂c B/∂b >> 0 (industry lobbying, legal challenges, capital flight), ∂E[credit B]/∂b ≈ 0 (marginal credit from additional audits is low). The state's optimum is at b = medium enforcement, where visible outputs are maximized relative to political and budgetary costs. By Axiom 1, neither party's optimization includes ∂W/∂a or ∂W/∂b. The system welfare term is not in the bilateral optimization. The action profile (a = accept shallow, b = medium enforcement) satisfies both parties' first-order conditions. This is the boundary-collapse equilibrium. Step 2: The outcome is (C=0, A=1, B=1). Under the equilibrium (accept shallow, medium enforcement), the financial institution earns positive revenue r - c s > 0, so A=1. The state earns positive political credit and budget allocation, so B=1. System welfare W is degraded because illicit financial flows continue...

2. The Flawed Game The game that produces the AML boundary collapse is a bilateral regulatory compliance game G = (A, B, Π A, Π B) between a financial institution A and the state B. The financial institution chooses whether to accept a customer relationship, how deeply to investigate the customer's ownership structure, and whether to file a suspicious activity report. The state chooses whether to audit, whether to fine, and whether to prosecute. Nature draws a state ω ∈ Ω describing whether the customer's beneficial owner is a criminal, a politically exposed person, a sanctions target, or a legitimate actor. The parties observe the legal personhood of the customer but not necessarily the beneficial ownership. The private payoffs are defined as follows. Π A includes transaction fees, deposit balances, correspondent banking revenue, advisory fees for structuring complex ownership chains, and the cost of compliance labor and software. In equilibrium, Π A is positive: the bank earns more from the customer relationship than it spends on compliance. Π B includes political credit for enforcement statistics, budget allocations to regulatory agencies, career advancement for officials, and international reputation from FATF mutual evaluations. In equilibrium, Π B is also positive: regulators publish enforcement actions, collect fines, and report statistics that justify their budgets...

Theorem 1: AML Boundary Collapse Theorem Statement. Under Axioms 1–5, the AML customer-identification game with θ = 0 produces a stable boundary-collapse equilibrium with outcome (C=0, A=1, B=1), the Hollow Win. The equilibrium is an Intractability Theorem domain: there exists a transformation R that sets θ = 1, producing a structurally better equilibrium. Proof. Step 1: Existence of the boundary-collapse equilibrium. Consider the game Γ with θ = 0. By Axiom 5, the financial institution's due diligence obligation terminates at the legal entity. By Axiom 2, the institution observes legal personhood with certainty but beneficial ownership only with probability p 0. The financial institution's expected utility under action a ∈ A is: E[u A(a, b, ω)] = Π A(a) - c A(a, b) + E[penalty A(a, b, ω)] where c A is compliance cost and penalty A is expected regulatory sanction. Under θ = 0, the compliance cost for shallow acceptance is c A(accept shallow) = c s, where c s is low (verification of registration documents, nominal director identity check, sanctions screening). The compliance cost for deep acceptance is c A(accept deep) = c d, where c d >> c s (beneficial ownership tracing, multi-jurisdictional investigation, legal analysis of ownership chains). The expected penalty for...

The Flawed Game G: Payoff Structure The AML customer-identification game G is a bilateral regulatory compliance game between financial institutions (A) and the state (B). The payoff structure is defined by three elements: 1. Party A (Financial Institution) Payoff (Π A): Revenue from customer relationships (fees, deposits, correspondent banking) minus compliance costs. Under the current regime, compliance costs are low because due diligence terminates at the legal person. The bank verifies a registration certificate, checks a nominal director, and screens the entity against sanctions lists. The marginal cost of accepting an additional opaque customer is near zero. The marginal revenue is positive. Therefore, Π A is maximized at "accept shallow"—the bank accepts the customer, collects fees, and files SARs when transactions exceed thresholds. 2. Party B (State) Payoff (Π B): Political credit from enforcement statistics, budget allocations to regulatory agencies, career advancement for officials, and international reputation from FATF mutual evaluations. The state's payoff is maximized at a level of enforcement that generates visible outputs—audits completed, fines levied, SARs received—without imposing requirements so stringent that the industry relocates or the financial center loses competitiveness. The state does not maximize system welfare; it maximizes its own institutional...

Theorem 2: System Welfare Beta Bound Statement. Under the boundary-collapse equilibrium, the system welfare beta β W = -dW/dΠ satisfies β W ≥ 5.0, placing the AML customer-identification paradigm in the strong intractability range. Proof. Let Π denote global AML compliance industry revenue. Let W denote system welfare destruction attributable to the legal-personhood boundary. By the boundary-collapse equilibrium, the marginal effect of compliance spending on welfare is negative: each dollar of compliance spending under θ = 0 produces compliance artifacts (SARs, audits, training) that do not interdict illicit flows, while the system continues to absorb welfare costs. Define W = W 0 + ΔW(θ), where W 0 is baseline welfare destruction from illicit finance and ΔW(θ) is the incremental destruction attributable to the boundary problem. Under θ = 0, ΔW(0) > 0 because the boundary prevents interdiction. Under θ = 1, ΔW(1) < ΔW(0) because beneficial owners are identified. From the empirical evidence (Section 4.7), illicit financial flows are approximately $2.8 trillion annually (UNODC estimate scaled to 2024 GDP). The World Bank found that opaque legal structures are used in over 70% of large corruption cases (Van der Does de Willebois et al., 2011). Attributing 60% of illicit flows to the boundary problem yields W ≥ 0.60 × $2.8 trillion = $1.68 trillion. However, not all of this...

that UK-registered entities were involved in approximately £80 billion in suspicious transactions between 2016 and 2020. The welfare cost includes: (1) continued laundering through UK shell companies—estimated at £20 billion annually; (2) reputational damage to the UK as a financial center—the 2022 Economic Crime Survey found that 40% of compliance professionals believed the UK was a "safe haven" for illicit finance; (3) the opportunity cost of a missed reform—the UK could have implemented a Danish-style verification system but chose a cheaper, less effective alternative. The cumulative system welfare cost over the 2016–2024 period is estimated at £100–200 billion. β W calculation: Π = £200 million (Companies House revenue). W = £100 billion (conservative system welfare destruction). β W = 100,000/200 = 500. This extreme value reflects the fact that Companies House's revenue was tiny relative to the welfare destruction enabled by the unverified register. Axiom demonstrated: Axiom 1 (W-Exclusion). The UK government's decision to create an unverified register was made within a bilateral payoff space that excluded system welfare. The government's objective was to demonstrate transparency at minimal cost. Companies House's objective was to process filings efficiently. Neither party's objective function included the system welfare consequence of an unverified register. The decision...

2012, many cartel-related; (2) addiction costs in the United States—approximately $200 billion annually in healthcare, lost productivity, and criminal justice costs; (3) corruption of Mexican institutions—police, military, and government officials compromised by cartel money. Attributing even 5% of these costs to the HSBC-enabled flows yields a system welfare cost of approximately $10 billion annually, or $60 billion over the six-year period. A more conservative estimate, focusing on the direct laundering costs, places the welfare cost at $15–20 billion. β W calculation: Π = $1.2 billion (HSBC Mexico revenue). W = $15 billion (conservative system welfare destruction). β W = 15/1.2 = 12.5. This is in the strong intractability range. Axiom demonstrated: Axiom 4 (Regulatory Budget Constraint). The OCC and FinCEN had limited resources to supervise HSBC's global operations. The OCC's international banking supervision division had approximately 50 examiners for all U.S. banks' foreign operations. The regulators knew that HSBC Mexico's compliance was weak—the OCC had issued a cease-and-desist order in 2003—but lacked the staffing to conduct comprehensive examinations. The deferred prosecution agreement was a compromise: the regulators demonstrated enforcement activity without shutting down a systemically important bank. The regulatory budget constraint made the Hollow Win stable...

that disclosure without verification is insufficient—thousands of companies listed fictitious beneficial owners, including a seven-year-old child. The Danish model completes the transformation by adding the verification layer that the UK omitted. The system welfare implication is measured in trillions of dollars over the relevant time horizon. The United Nations Office on Drugs and Crime estimates criminal proceeds at approximately 3.6% of global GDP, or $2.8 trillion annually. Global Financial Integrity estimates illicit financial outflows from developing countries alone at $1.1 trillion annually. The OECD estimates that base erosion and profit shifting cost governments $100–$240 billion annually in lost tax revenue. The World Bank found that shell companies were used in over 70% of large corruption cases. These are not separate problems; they are manifestations of the same structural flaw—the boundary between legal personhood and beneficial control that the AML regime was designed to police but instead preserves. The system welfare gain from closing this boundary is not merely the reduction in laundering but the restoration of trust in financial institutions as intermediaries rather than conduits. When the beneficial owner is the customer, the financial system ceases to be a vector for the Hollow Win and becomes capable of generating the Win-Win-Win outcome: financial...

Theorem 3: Disclosure Futility Statement. Under Axioms 1–5, disclosure-only interventions that do not change θ from 0 to 1 cannot shift the equilibrium from (0,1,1) to (1,1,1). Proof. Consider an intervention I that increases the information available to the financial institution or the state without changing the legal definition of "customer." Examples include: requiring additional SAR fields, increasing the frequency of audits, publishing lists of suspicious entities, or mandating training programs. Under Axiom 1, system welfare W remains excluded from the bilateral payoff functions regardless of the information set. Under Axiom 2, the information partition expands but the legal obligation terminates at the legal entity. Under Axiom 3, competitive pressure remains: the institution that acts on additional information by rejecting opaque customers loses revenue to competitors who do not. Under Axiom 4, the state's regulatory intensity remains bounded. The intervention I changes the information set from I A to I A' where I A ⊂ I A', but does not change the action set or the payoff functions. The financial institution's first-order condition remains ∂E[u A]/∂a = r - c s > 0 at accept shallow. The state's first-order condition remains at medium enforcement. The equilibrium is unchanged. Empirically, the FinCEN Files (Section 4.3) demonstrate that SAR filing—a disclosure-only...

Why AML identity rules collapse at the boundary between legal personhood and beneficial control Erik Postnieks Independent Researcher, Salt Lake City, UT, USA Working Paper — May 2026 Correspondence: erik@woosterllc.com Companion to: Postnieks (2026a), "The Missing System Theorem" ---

European Banking Authority that surveyed 50 banks on their experience with beneficial ownership identification. The study found that banks struggled to identify beneficial owners even with existing thresholds. The Commission did not commission original research on the distribution of beneficial ownership stakes in EU companies, the prevalence of threshold-evasion strategies, or the welfare consequences of the 20% gap. The evidence base was constructed to support a predetermined policy choice. Field 7: AUTHORITY. The decision was made under the EU's competence in internal market harmonization (Article 114 TFEU). The legal basis was the establishment of a single market for financial services, not the protection of system welfare. The authority derived from the EU's treaty powers, not from any mandate to maximize financial integrity. Field 8: TRAINING. The decision-makers were trade lawyers, financial regulators, and policy generalists. None had training in game theory, welfare economics, or the empirical study of illicit finance. The European Parliament's rapporteur for the directive was a lawyer specializing in financial regulation. The decision-making process did not include experts in beneficial ownership evasion techniques. Field 9: REVIEW. The directive required the Commission to review its operation by June 26, 2019, and every three years thereafter. The first review...