Capital routes through the jurisdiction of least regulation. Individually compliant in every jurisdiction. Collectively non-compliant in the aggregate.
Regulatory arbitrage exploits jurisdictional boundaries. A transaction that is regulated in Jurisdiction A and unregulated in Jurisdiction B will route through Jurisdiction B — not to evade the regulation's purpose but to ensure that no single jurisdiction's regulatory authority encompasses the full structure. The transaction is individually compliant in every jurisdiction it touches. The aggregate structure achieves an outcome that no single jurisdiction would permit if it had authority over the whole.
This is the Accountability Firewall (AF-001 through AF-005) applied to nations rather than departments. The same structural principle — separating knowledge of harm from authority to act on it — operates through jurisdictional boundaries rather than organisational boundaries.
The FinCEN Files (September 2020) were the most comprehensive disclosure of the global regulatory arbitrage architecture to date. The leaked documents — over 2,100 Suspicious Activity Reports (SARs) filed by banks with the US Financial Crimes Enforcement Network between 1999 and 2017 — documented approximately $2 trillion in transactions that the filing banks had flagged as suspicious but continued to process.
The structural finding: banks operating in multiple jurisdictions were individually compliant with each jurisdiction's anti-money laundering (AML) requirements. Each branch filed the SARs required by its local regulator. Each branch conducted the customer due diligence required by its local law. Each branch, evaluated by its local regulator, was in compliance.
The systemic finding: the same capital flows — flagged as suspicious by the banks themselves — moved through the global banking system for years after being flagged, because no single jurisdiction had the authority, the information, or the mandate to stop them. The US received the SARs. It did not share them with the jurisdictions where the suspicious capital was moving. Those jurisdictions did not know the transactions had been flagged. The transactions continued.
The Jurisdictional Void is not an absence of regulation. It is a gap between regulations — a gap that capital flows are designed to exploit.
The most structurally stable form of regulatory arbitrage is the multinational tax architecture — the configuration of corporate structures across jurisdictions to minimise the aggregate tax burden while maintaining technical compliance in every jurisdiction touched.
The mechanism is documented: a US technology company establishes an Irish subsidiary. The Irish subsidiary licenses the company's intellectual property. Revenue generated in European markets is booked through the Irish subsidiary. Ireland's corporate tax rate (12.5%) applies to the revenue rather than the US rate (21%) or the rates of the European countries where the revenue was generated. The arrangement is legal in every jurisdiction.
The "Double Irish with a Dutch Sandwich" — the specific tax architecture used by major technology companies through 2020 — added additional layers: the Irish subsidiary was managed from a jurisdiction with no corporate income tax (Bermuda, in most implementations), making it not tax-resident in Ireland. Royalty payments flowed through a Dutch entity (exploiting the Netherlands' favourable tax treaty network) before reaching the tax-free jurisdiction. Each step was individually compliant with the laws of Ireland, the Netherlands, and the tax-free jurisdiction. The aggregate effect — technology companies paying effective tax rates of 2-4% on non-US income — was not the product of tax evasion. It was the product of tax architecture.
The OECD's Base Erosion and Profit Shifting (BEPS) framework (2015) and the subsequent Global Minimum Tax (Pillar Two, 2023) represent the regulatory response. The response is slow — years from proposal to implementation — and the architecture adapts faster than the regulation.
The cryptocurrency ecosystem has produced the most rapid regulatory arbitrage cycle in financial history, because the technology creates new transaction types faster than any jurisdiction can classify and regulate them.
When China banned cryptocurrency exchanges in 2017, the exchange operators moved to Malta, Singapore, and the Seychelles. When the US SEC began enforcement actions against initial coin offerings (ICOs) in 2018, issuers moved to Switzerland, Gibraltar, and Estonia. When the EU's Markets in Crypto-Assets (MiCA) regulation was proposed in 2020, crypto businesses began establishing operations in jurisdictions not covered by MiCA (UAE, Bahamas, El Salvador).
The pattern is structurally identical to the traditional financial arbitrage documented above — capital routes through the jurisdiction of least regulation — but the cycle time is compressed from years to months. The regulatory response (classify the new instrument, draft the regulation, comment period, implementation) requires 2-5 years in most jurisdictions. The arbitrage response (reincorporate in a new jurisdiction, redirect capital flows) requires weeks.
Regulatory arbitrage exploits the Jurisdictional Void — but the Jurisdictional Void is not just a structural gap in regulatory architecture. It is a cognitive gap in the evaluator's mental model.
A citizen, legislator, or journalist evaluating whether a company is "paying its fair share" of taxes must hold in working memory simultaneously: the company's corporate structure (which entities exist in which jurisdictions), the tax rates and treaty networks of each jurisdiction, the flow of intellectual property, revenue, and royalty payments through the structure, and the accounting treatment that determines where profit is reported. This evaluation exceeds the working memory capacity documented in AOA-006 for the vast majority of the population — including the legislative staffers who draft tax legislation and the journalists who report on it.
The consequence: the public debate about corporate taxation operates at the level of effective tax rates (a single number) rather than at the level of the structures that produce those rates (a multi-jurisdictional architecture). The single number is debatable. The architecture is invisible — not because it is hidden, but because the cognitive capacity required to hold it in view exceeds the working memory of the population debating it.
The Jurisdictional Void cannot be closed by any single jurisdiction acting alone. It can only be closed by cross-jurisdictional information sharing agreements that allow regulators in one jurisdiction to see what regulators in another jurisdiction see — and by harmonised standards that prevent capital from exploiting definitional differences between jurisdictions.
The OECD's Common Reporting Standard (CRS, 2014) is the most comprehensive attempt: automatic exchange of financial account information between participating jurisdictions, covering bank accounts, investment accounts, and certain insurance products. As of 2025, over 100 jurisdictions participate. The CRS has demonstrated that cross-jurisdictional information sharing is technically feasible at global scale.
The CRS also demonstrates the limits of the approach: the United States does not participate in the CRS. The US requires foreign banks to report information on US account holders (through FATCA — the Foreign Account Tax Compliance Act) but does not reciprocally share information on foreign nationals holding accounts in the US. The US is simultaneously the world's largest financial centre and the largest participant in the Jurisdictional Void it nominally opposes.
The Jurisdictional Void — the regulatory gap between national jurisdictions that permits capital structures to be individually compliant in every jurisdiction while being collectively non-compliant in the aggregate, because no single jurisdiction has authority over the whole. The Void is not an absence of regulation — it is a gap between regulations, designed to be exploited by structures that route capital through the jurisdiction of least regulatory friction.
Internal: This paper is part of The Obfuscation Economy (OE series), Saga VIII. It draws on and contributes to the argument documented across 55 papers in 12 series.
External references for this paper are in development. The Institute’s reference program is adding formal academic citations across the corpus. Priority papers (P0/P1) have complete references sections.