RESEARCH DOSSIER
The Economics of Financial Secrecy
Abstract. This Research Dossier examines the economic production, value, distributional consequences, measurement and institutional governance of financial secrecy. It distinguishes confidentiality, information deficits and deliberate concealment in relation to the decision-maker and the legitimate purpose of access. Drawing on published economic research, cross-border banking statistics and international institutional documents, it evaluates evidence on information-exchange reforms, ownership verification, provider incentives, wealth statistics and data governance. The analysis distinguishes recorded financial responses from tax compliance and welfare outcomes, and explains why incomplete policy coverage and heterogeneous measurement complicate causal inference. The central conclusion is that effective financial accountability depends on accurate, securely accessible information for legitimate decisions, with proportionate costs and protection against misuse. This structured narrative synthesis provides original analytical interpretation without asserting a new named theory, independent empirical estimate or scientific priority.
Keywords: financial secrecy; offshore economics; financial confidentiality; information asymmetry; beneficial ownership; automatic exchange of information; cross-border deposits; institutional economics; tax compliance; data governance
The scientific object of financial secrecy
Financial secrecy is an institutional condition in which economically relevant information about financial assets, transactions, ownership, control, or obligations is unavailable to a particular party that could otherwise act upon it. The relevant unit of analysis is therefore an information relationship, not a geographical label. A regulated bank can know its customer's identity while a tax administration lacks information about that customer's foreign holdings. A lender can possess reliable information about a borrower while another creditor cannot establish the borrower's aggregate encumbrances. These are different information relationships, with different consequences for resource allocation.
The distinction between confidentiality and concealment follows from the allocation of legitimate access. Confidentiality protects information against unauthorized or unnecessary disclosure. Concealment describes the deliberate obstruction of information that another party is entitled to obtain. Financial secrecy encompasses both actual restrictions and institutional arrangements that produce opacity; its economic consequences depend on who lacks information, the decision affected, and the purpose for which the information would be used. A foreign account declared to the competent authority may be confidential without being hidden from it. An undisclosed domestic obligation can be economically opaque even when no cross-border transaction exists.
Offshore Economics studies the international distribution of these relationships. The cross-border dimension matters because legal jurisdiction, fiscal residence, financial booking location, and ultimate economic ownership can differ. A claim may be recorded correctly under one statistical or legal concept and still be uninformative for a different decision. The scope of the present dossier is the economic value, production, incidence, measurement and governance of financial secrecy. Household portfolio choices, the legal content of banking confidentiality, the estimation of illicit financial flows and the economics of tax evasion remain neighboring subjects; they enter here only where they illuminate information allocation.
The central inquiry concerns the compatibility of private confidentiality with verifiability of obligations. Treating every confidential arrangement as evidence of misconduct would erase legitimate demand and confuse legal status with information quality. Treating every properly incorporated legal entity as economically transparent would make the inverse mistake. The relevant analytical question is whether the parties responsible for evaluating an economic claim can obtain trustworthy information at a cost proportionate to its importance.
Information as an economic resource
Information affects the terms on which assets are exchanged and contracts enforced. A creditor who cannot establish the distribution of liabilities must assess risk using incomplete evidence. Additional screening, collateral or contractual restrictions can partially compensate for that uncertainty, but they also consume resources and may restrict financing for solvent borrowers. An investor evaluating a company needs reliable information about its claims on future cash flows, related-party exposures and conflicts of interest. A fiscal authority requires information sufficient to establish the tax base under applicable law.
The literature synthesized by Acquisti, Taylor and Wagman (2016) in The Economics of Privacy treats disclosure and privacy as economically consequential choices whose welfare effects vary by setting. Financial information is a particularly clear instance of that problem. Disclosure can reduce asymmetric information and enable transactions while also transferring sensitive information to parties able to misuse it. A person whose assets become publicly searchable may experience unwanted exposure regardless of tax compliance. An entrepreneur may lose bargaining power when competitors acquire knowledge of a commercially sensitive transaction. These costs belong in economic evaluation.
Information is unevenly distributed along multiple axes. The person who knows an asset exists may not know its current value. The party that records title may not know who enjoys the benefits. A reporting institution may know the immediate counterparty but have limited ability to identify the group-level allocation of risks. Timeliness matters as much as availability: a formally accurate ownership record can become economically useless if changes in control are discovered only after a loss has occurred.
A further difficulty is that producing information is costly. Identity verification, valuation, data reconciliation and dispute resolution require labor, infrastructure and judgment. Some costs are fixed and others increase with the complexity or frequency of transactions. A marginal disclosure requirement can have a different effect on a small firm than on a global intermediary operating a reusable compliance platform. Efficient institutional design must ask which information is decision-relevant, who can verify it at lowest cost, how long it should be retained and which users need access.
There is no universal optimum expressed as a percentage of all financial data made public. The economically useful quantity is verifiable information delivered to a competent decision-maker for a legitimate purpose. Its benefit includes improvements in decisions and deterrence; its cost includes both production and unwanted exposure. The two dimensions may be improved together when institutions reduce duplication, maintain secure records and permit narrowly justified access.
The demand for confidentiality and opacity
Demand for financial confidentiality arises from several distinct motives. Households protect sensitive wealth information; organizations preserve negotiating positions and operational security; fiduciaries administer assets on behalf of others; owners of legitimate foreign investments seek predictable treatment and secure custody. A person may rationally prefer a financial service whose information controls reduce exposure to fraud or coercion. The strength of such preferences depends on circumstances and does not by itself reveal intent.
Demand can also reflect an attempt to reduce the probability that an enforceable claim will be discovered. A taxpayer who underreports assets, a debtor who conceals property from an entitled creditor, or an agent who hides a conflict of interest expects a private benefit from the other party's information deficit. The distinctive distributional feature is that the benefit is secured partly through costs imposed on others. This creates an externality unless the affected parties can detect, contract around or otherwise address it.
A single observed transaction generally does not identify which motive dominates. Cross-border financial services can jointly provide custody, portfolio diversification, legal certainty, currency choice, specialist expertise and confidentiality. Measuring the price of a service cannot isolate the value of secrecy if these attributes are bundled. Nor does an individual's movement of funds following a regulatory change establish an unlawful purpose. Compliant investors may respond to new paperwork, fees, restrictions or declining service quality. Identification of unlawful conduct requires additional evidence about relevant duties, behavior and facts.
Differences in willingness to pay also matter. Complex financial arrangements have setup costs, ongoing service fees and organizational demands. Fixed costs imply that a benefit proportional to the size of a portfolio can be economically more attractive for holders of large portfolios. This is a general scale argument, not an estimate of the prevalence of evasion. It provides an economic explanation for why research on opacity should examine wealth groups and service access rather than relying exclusively on mean behavior.
The distinction is operational for policy evaluation. A restriction that reduces the attractiveness of unlawful concealment while preserving secure confidentiality may improve aggregate welfare more than an indiscriminate expansion of public disclosures. Conversely, a privacy exemption that blocks legitimate accountability can benefit concealment. The consequences depend on actual access and enforcement, not the label attached to a regime.
The supply of secrecy and intermediary incentives
Financial opacity is produced within a network of institutions rather than by a single anonymous actor. Financial intermediaries, corporate service providers, fiduciaries and public registries each hold or generate information. They face revenue incentives, costs of due diligence, legal obligations and the risk of reputational or regulatory loss. Where transactions span several institutions, each may have a partial view of the relationship. The social challenge is to reconcile those records securely and efficiently.
A provider can regard thorough verification as a cost of accepting a customer, a means of reducing future loss, or a source of trust in the service. These incentives depend on whether the provider bears downstream costs. If deficiencies result in a penalty imposed on another institution rather than the originating provider, private incentives to invest in information quality can be weaker than social incentives. Contractual monitoring, professional responsibility and supervision can help align them.
Findley, Nielson and Sharman (2015) examined actual provider responses through field experiments on anonymous incorporation. Their experiments contacted incorporation services in many jurisdictions and a separate sample of firms in the United States. The reported differences in responses illustrate that declared national policy and observed provider conduct can diverge. The measured outcome is a response to a standardized inquiry, not a census of completed incorporations or a count of crimes. Response selection, the time of the experiment and the population of providers delimit generalization.
The economic lesson is that supervision based on the text of formal rules alone overlooks performance at the point where information is generated. A formally stringent requirement with weak verification, inadequate resources or inconsistent treatment can deliver limited effective transparency. A provider that invests in verification may have a competitive disadvantage if rivals avoid equivalent costs, especially when customers cannot distinguish quality. This helps explain the case for comparable obligations and oversight across providers.
The legitimate market for confidentiality is also a market for trust. Institutions can differentiate services through secure handling, accuracy, confidentiality guarantees, responsiveness and a reputation for compliance. Private information protection is valuable partly because customers expect providers to resist unauthorized disclosure. The resulting commercial incentive can support lawful accountability when the rules clearly regulate competent access and misuse. Regulatory design should preserve this legitimate source of reputation while preventing firms from profiting from information deficiencies that undermine enforceable claims.
Jurisdictional competition and network externalities
Financial systems interact across jurisdictional boundaries. A jurisdiction can specialize in services that attract nonresident clients, collecting domestic wages, fees and fiscal receipts. The service may also affect tax bases, creditor recoveries and supervisory information in other jurisdictions. Benefits and costs therefore need not accrue to the same population.
This is a familiar coordination problem. If each jurisdiction evaluates its information policy solely by domestic revenue and domestic compliance cost, it can underweight effects elsewhere. International cooperation changes the feasible allocation of information and can reduce mutually harmful competition over opacity. The cooperative mechanism has its own costs, including legal negotiation, data exchange, dispute handling and security safeguards.
Jurisdictions differ in economic size, institutional capacity, financial specialization and legal history. A rule that creates a moderate information gap in a large center can have greater aggregate consequences than a similar gap in a small one. A small center can nevertheless be highly relevant to an individual case. Aggregate systemic contribution and case-specific exposure answer separate questions.
Interdependence also means that a change in one jurisdiction may affect outcomes in another. A reduction in recorded external liabilities after reform may coexist with a rise elsewhere. A study that uses unaffected jurisdictions as control units must consider whether those jurisdictions are indirectly affected. This is a problem for inference as well as for policy: a collection of local changes does not establish an equivalent global change.
The appropriate objective is the reliability of cross-border accountability. Coordination that establishes equivalent outcomes while allowing jurisdictions to choose locally appropriate administrative means can decrease compliance fragmentation. Capacity differences remain relevant: formal reciprocity is less useful when one party lacks the systems to receive or act upon information. Mutual assistance should include the ability to consume and verify information, not merely the authority to request or transmit it.
Economic incidence and distribution
The private beneficiary of an opaque relationship is not necessarily the final beneficiary of the associated economic surplus. Service providers may receive fees; skilled intermediaries may earn wages or rents; other parties may incur increased information costs. In the case of illegal tax concealment, the initial effect includes the difference between tax legally due and tax actually paid. The subsequent incidence depends on adjustments in public expenditure, borrowing, enforcement and taxes on other bases. A fiscal shortfall is not automatically an equivalent measure of deadweight loss.
Credit markets exhibit another channel. When creditors cannot distinguish reliable information from inaccurate representations, they may charge broader risk premia or demand additional security. The resulting costs may extend to transparent counterparties that are pooled with opaque ones. Conversely, creditworthy borrowers who can selectively prove their financial condition may obtain better terms without revealing their entire financial position to the public. The value of institutional verification is partly the ability to separate those groups.
Alstadsæter, Johannesen and Zucman (2019), in Tax Evasion and Inequality, combine leaked offshore customer information with Scandinavian administrative wealth records. Their reported estimate for the wealthiest 0.01 percent is approximately 25 percent of taxes evaded, substantially above the rates detected by stratified random audits. This result concerns their setting and measurement strategy. It should not be universalized to all wealth holders or treated as a direct contemporary estimate of global tax evasion.
The contrast between evidence sources has a deeper methodological significance. A random sample of taxpayers does not guarantee uniform detection across forms of conduct. If an audit procedure detects some reporting omissions more readily than others, estimated rates depend on the detection technology. Administrative records may accurately describe declared income while remaining systematically incomplete about activities outside the reporting system.
A distributional evaluation should separate the direct tax effect, changes in measured wealth, service-provider income and the real resources consumed in concealment and enforcement. Weighting those components involves social judgments that are not determined by accounting alone. An economy can have accurate aggregate asset measures and still lack information about individual compliance; conversely, it can improve case-level compliance without immediately discovering all the missing assets relevant to macroeconomic statistics.
Macroeconomic visibility and international balance sheets
Cross-border positions are measured under specific accounting and statistical conventions. An asset of one party generally corresponds to a liability of another, although recorded international positions may be incomplete or allocated differently. An asymmetry between aggregates can indicate a reporting gap, valuation inconsistency or classification difference. A discrepancy is evidence of a problem requiring explanation, not a self-contained estimate of unlawful activity.
Zucman's (2013) Missing Wealth of Nations uses portfolio statistics and Swiss evidence to infer historically unrecorded household financial assets. The paper estimates that about eight percent of global household financial wealth was held in havens, with roughly three quarters absent from official records, and shows how adjustments can change measured net foreign asset positions. These are findings of that historical study conditional on its methods and assumptions, not a 2026 wealth stock or a direct measure of unpaid tax.
The distinction between residence and ownership is central. The BIS Locational Banking Statistics organize positions from the perspective of the booking institution and distinguish cross-border positions by residence. A recorded liability to a resident entity is not automatically a liability to the entity's ultimate controlling individual. Residence-based statistics are appropriate for analyzing some questions about cross-border banking and currency exposure; they are less suitable by themselves for measuring the final allocation of household wealth.
Stocks and flows must also be distinguished. A stock is measured at a date; a transaction flow refers to an interval. A difference in stocks can reflect exchange-rate changes, asset revaluation, changes in coverage or other statistical breaks as well as transactions. The BIS glossary explicitly describes adjustments for foreign-exchange movements and breaks. Even an adjusted movement remains a banking statistic rather than a direct observation of taxable income or unlawful transfers.
Public statistical agencies therefore need information suited to the economic question. Useful improvements include reconciled instrument definitions, consistent reporting residence, explicit counterpart-sector coverage and transparent documentation of discontinuities. Ultimate ownership attribution may require supplementary information, protected microdata and secure matching. The feasibility of those improvements depends on resources and legal authority. Accuracy should be judged by the intended statistical concept rather than an expectation that one database answer every question about global capital.
Measuring the degree and scale of secrecy
There are at least three empirical objects: institutional opportunity for secrecy, observed opacity in specific relationships, and economic harm attributable to opacity. They have different denominators and evidentiary demands. A legal assessment describes the rules governing information. An audit of records describes the quality of implementation. An outcome evaluation estimates consequences for behavior or welfare. Confusing these objects generates false precision.
The Tax Justice Network Financial Secrecy Index methodology combines a score for institutional opportunities for opacity with a weight related to the scale of services provided to nonresidents. The producer's overview describes scores across twenty indicators and a global scale measure derived from financial-service activity. Since 2025 the producer has moved to rolling indicator updates, with annual scale-weight updates. A reported ranking is consequently version- and date-dependent.
The distinction between opportunity and scale is useful. A jurisdiction offering comparatively broad opacity to a small financial market and a jurisdiction with narrower opacity but a vast financial market pose different systemic questions. Neither score alone is a measure of observed criminal conduct. The choice of indicators, aggregation and weights embeds measurement objectives and judgment. An index is best used as a structured description of institutional exposure, not as a substitute for transaction-level investigation.
The present research does not compute new scores, reproduce the index's weighting formula or rank individual jurisdictions. Such exercises require a frozen dataset version and sensitivity analysis. An appropriate empirical comparison would report both the standalone score and scale component, examine changes in underlying indicators, and distinguish updated observations from actual regulatory change. A jurisdiction's relative rank can change because another jurisdiction changes, even when its own institutions remain constant.
Coverage limitations also affect administrative data. The universe of accounts under an exchange standard, the universe of beneficial-ownership records and the universe of financial institutions are not identical. Subsets can overlap and cannot simply be added. A sound reporting table identifies the observation unit, period, population, currency, treatment of missing values and source of legal or statistical definitions. Where the denominator is incomplete, uncertainty must be preserved rather than replaced by an unjustified global total.
Bilateral exchange and the first transparency shocks
The early international policy response emphasized the capacity to exchange tax-related information upon request. Economically, requests change the expected accessibility of information to a competent authority. Their effect depends on the ability to identify relevant cases and obtain assistance, as well as on the incentives of financial actors.
Johannesen and Zucman (2014), in The End of Bank Secrecy?, investigate bilateral deposit data after the G20-era information-exchange crackdown. They report a pattern consistent with movement of deposits toward jurisdictions without a corresponding bilateral treaty, rather than a general return of funds to residence jurisdictions. The empirical result establishes that responses in the observed deposit network can offset the effect of partial coverage. It does not establish the legal status of every deposit.
A treaty changes several margins simultaneously. The perceived probability of disclosure can rise before a request is ever made. Compliance expenditure can increase. A service relationship may become less convenient for a customer whose funds are fully declared. Effects on recorded positions combine movements by people with different motives and costs. The policy effect on unlawful concealment is therefore more specific than the policy effect on all deposits.
The central analytical implication is that the unit of reform and the unit of outcome need to match. A bilateral agreement can succeed at producing information for those two parties while leaving the wider network incompletely observed. Its direct effect is informative, yet a global effectiveness claim requires evidence about indirect effects and changes in reporting beyond the treated pair. A local decline must be evaluated in conjunction with counterparties and other asset categories.
This does not imply that partial compliance improvements are worthless. Information received under bilateral arrangements may have immediate investigative or revenue value even when other deposits relocate. The proper evaluation tracks the quality and use of the exchanged information as well as the aggregate financial response. An instrument can be effective at a defined purpose without eliminating every form of opacity.
Automatic exchange and the evidence from deposits
Automatic exchange changes the acquisition of information from a primarily case-triggered process into a recurring administrative process for a defined reporting population. Standardization can lower the marginal cost of supplying information because institutions compile recurrent records in a comparable form. The benefit depends on reliable identification of the relevant reporting and receiving jurisdictions and on their capacity to match data to taxpayers.
Casi, Spengel and Stage (2020), in their study of the Common Reporting Standard, estimate an 11.5 percent reduction in cross-border deposits located in tax havens following CRS adoption in the short run. The authors also report evidence of geographical redistribution. That percentage concerns the studied deposit outcome and design; it cannot be read as 11.5 percent of all hidden wealth disclosed.
Beer, Delgado Coelho and Leduc (2019) report in Hidden Treasure an estimated average 25 percent reduction in foreign-owned deposits in offshore jurisdictions associated with automatic exchange frameworks. Their broader observation period is 1995–2018; the dataset includes 39 reporting countries and more than 200 counterparty jurisdictions. This is an IMF working paper rather than an institutional guarantee of the result. The estimates in the two papers should neither be added nor treated as direct replications, because the reporting groups, periods, treatment definitions and specifications differ.
The later study by Bénétrix, Emter and Schmitz (2024), Automatic for the (Tax) People, addresses heterogeneity using more detailed BIS sectoral data for 2014–2019. Household deposits are an important component of the observed decline, while some financial-intermediary positions between haven jurisdictions increase. The authors also analyze portfolio and direct investment; those results are consistent with changes extending beyond deposits, although indirect evidence does not reveal the legal status or complete provenance of individual balances.
The studies collectively strengthen the conclusion that information policy changes recorded financial choices. Their different estimates do not identify a common universal effect. The correct comparison concerns the direction of observed responses, their data coverage and the margins studied. A comprehensive assessment would also measure actual declarations, assessments, payments, matching quality and the costs of both legal and unlawful behavioral responses. Those outcomes cannot be recovered from aggregate deposit changes alone.
Identification, uncertainty and competing explanations
A causal evaluation compares an outcome under a policy with the unobserved outcome that would have occurred without it. An event study or panel comparison attempts to approximate this counterfactual using variation in implementation timing or coverage and observations on relevant comparison groups. The validity of the exercise depends on assumptions about trends, selection and interference.
Selection can arise when governments adopt exchange agreements at times related to their financial conditions. Anticipation can shift responses before a formal implementation date. In their discussion of pre-policy trends, Bénétrix, Emter and Schmitz (2024) explicitly report indications of anticipation and investigate robustness to timing. This illustrates why signature, announcement, legal effect and first actual exchange should be recorded separately. Collapsing them into one date can misstate the treatment.
The use of comparison jurisdictions raises the possibility of interference. Cross-border markets are interconnected; an entity whose observable position changes in one place may generate a change elsewhere. A control observation may therefore reflect some treatment effects. Sensitivity analysis could compare pooled and sector-specific outcomes, include measures of policy coverage, and examine alternative definitions of the network. The proposed checks are analytical recommendations; this dossier does not claim to have implemented them on restricted microdata.
Precision matters as well. In Table 3 of the Bénétrix study, the coefficient for the specified haven-to-haven nonbank-financial-institution subgroup is 0.401, with standard error 0.231 and the authors' ten-percent significance designation. The relatively imprecise subgroup estimate should be presented as suggestive, especially because multiple specifications and outcomes are examined. The reported increase is not a demonstrated count of deliberately concealed arrangements.
Changes in balance sheets can reflect currency movements, investment returns, financial conditions, market consolidation, changes in reporting coverage and unrelated regulatory reforms. A credible study identifies the changes that are absorbed through adjustments or fixed effects and those that remain plausible confounders. Interpretations should follow the measured outcome: the effect on recorded deposits is a different estimand from the effect on previously concealed taxable wealth.
The evidence in this dossier is a structured narrative synthesis, not a systematic review with exhaustively screened databases or a new econometric replication. Most journal evidence is checked at the published abstract and metadata level; the Bénétrix article and Findley author manuscript were inspected selectively in greater depth. The limits of access preclude a fully reproduced numerical audit of original coefficients across all studies. The authorial contribution consists of interpretation, distinctions and policy evaluation; it does not claim a novel dataset or a verified new theory.
Beneficial ownership and the economics of verification
Beneficial ownership information addresses the difference between a registered legal relationship and the natural person or persons who ultimately own or control a legal structure under the relevant framework. The economic significance extends beyond taxation. Ownership and control are relevant to conflicts of interest, credit risk, fiduciary responsibilities and the enforceability of claims. Accurate records can reduce the resources required to establish those relationships.
FATF's 2023 guidance on legal persons describes the revised Recommendation 24 requirements and discusses a multipronged implementation approach. The objective is access by competent authorities to beneficial ownership information of suitable quality and currency through a registry or an alternative effective mechanism. A useful registry forms part of a broader information environment rather than serving as a guarantee that every filed statement is correct.
The complementary 2024 guidance on legal arrangements addresses express trusts and similar structures within Recommendation 25. The distinction matters because legal title, managerial functions and entitlement to benefits can be allocated among different parties. Information collection must respect the relevant legal concepts when mapping those relationships. Treating the conventional shareholder register as the universal ownership record can leave consequential relationships outside the observation frame.
Verification creates a relationship between the quality of initial information and the capacity to correct it after change. A system that permits record updates but fails to reconcile conflicts may accumulate errors. A system that imposes high penalties for innocent inconsistencies without an accessible correction process may discourage legitimate participation. The economic design problem encompasses data accuracy, update incentives, responsibility for verification, safe correction and the actual time required to retrieve sufficient evidence.
Several measures should be evaluated separately: completeness of required records, reconciliation of discrepancies, timeliness, processing burden, availability to authorized users and observed outcomes in appropriate cases. A raw count of registered entities is useful for workload measurement but cannot establish effective transparency. The desired result is credible knowledge for decisions, not the mere existence of a database.
Confidentiality, public access and data governance
The degree of public access to financial information constitutes a separate decision from whether information should be collected and verified. Different users have different legitimate purposes and risks. A tax authority establishes taxable obligations; a supervisor assesses financial institutions; a creditor evaluates contractual exposure; a journalist may investigate matters of public interest. The information appropriate to one purpose may exceed or fall short of that appropriate to another.
The distinction is illustrated by the November 2022 judgment in WM and Sovim v Luxembourg Business Registers. The Court of Justice invalidated the specific EU directive amendment insofar as it imposed access in all cases for any member of the general public. The judgment's operative scope concerns that general-access provision. It does not establish that ownership information should never be collected or made available to competent parties, and it is not a complete description of currently applicable national law.
From an economic perspective, expanded disclosure can create positive information externalities: outside parties may identify inconsistencies, reveal conflicts or help reduce investigative costs. It can also generate negative externalities through unauthorized aggregation, security exposure, mistaken identification and unnecessary intrusion. The magnitude and distribution of these effects depend on information granularity, access procedures, retention, correction and enforcement against misuse.
A well-governed information arrangement must therefore be evaluated on actual access, accuracy and accountability. A formally closed register with effective competent access can achieve some objectives better than an open register with unreliable data. Public or appropriately justified access can complement competent-authority access for other purposes. An evaluation should identify whose information deficit is addressed by each form of disclosure, rather than assuming that public visibility is the sole meaningful dimension.
Operational security is a component of institutional credibility. Participants should know which authority controls data, which processing purposes are authorized and how errors are corrected. Information systems require authentication, appropriate restrictions on access, records of use and remedies against misuse. The aim is to lower justified information barriers while preserving the protection that makes legitimate financial relationships trustworthy.
The implementation gap in international exchange
International standards establish a common design objective, but their benefits depend on implementation. The OECD (2020) guide on financial-institution compliance identifies robust reporting and due diligence processes as important conditions for useful exchange. The information may have travelled through several organizations before it is matched and applied to a tax assessment. Failure at any point can reduce the practical value of otherwise successful transmission.
The OECD's 2025 AEOI peer review update provides a dated illustration of this distinction. It reports that 114 of 118 assessed jurisdictions had legal frameworks rated in place or in place but in need of improvement. Of 108 jurisdictions assessed in the initial effectiveness reviews, 63 percent were rated on track. The denominators differ; these are not successive percentages of one common population. Moreover, legal and implementation ratings address different criteria and cannot simply be subtracted to derive a count of malfunctioning systems.
The gap is economically significant because adoption of a standard can be less costly than maintaining consistent operational performance. Building information systems, training staff, validating records and handling mismatches require continuing expenditure. Jurisdictions with limited administrative capacity may need support to make formally available information usable. A standard that generates information without enabling its correct interpretation can impose costs on providers without yielding its potential fiscal or supervisory benefits.
Evaluation should follow the full information lifecycle: legal authorization, identification, capture, verification, transmission, receipt, secure storage, matching, competent analysis, correction and authorized action. Distinguish transmission failures from matching failures and incomplete source records. A measure such as the proportion of records matched to the intended population is more informative about practical utility than the number of records transmitted alone, provided the denominator is appropriately defined.
Aggregate tax recovery is important but not a sufficient implementation metric. Recoveries depend on the applicable law, the preexisting compliance rate, audit resources and reporting lags. An improved system can deter noncompliance without an observable increase in assessments; conversely, a surge in recoveries may reflect the discovery of historical liabilities. Both current deterrence and retrospective enforcement should be studied using their own outcome definitions.
Digital financial instruments and evolving coverage
Digital financial instruments reorganize the production and distribution of information. Some records can be generated automatically while the economic identity associated with a record still requires verification. The presence of an auditable transaction history does not by itself establish who is responsible for the associated obligations. Conversely, verified identities linked to financial records do not require public disclosure of all personal details.
The OECD (2025) consolidated Common Reporting Standard describes amendments extending coverage to specified electronic money products, central bank digital currencies and indirect investment exposure to crypto-assets, alongside changes in due diligence and reporting. The publication is an unofficial consolidation; jurisdiction-specific obligations and commencement dates require separate legal verification. The expansion illustrates that reporting boundaries must evolve when economically comparable activity is organized through new instruments.
In 2024 the OECD announced CARF and amended CRS transmission schemas and indicated that first exchanges under these frameworks were expected to begin in 2027. This is a reported implementation expectation, not an observed exchange outcome. The present work makes no claim about subsequent jurisdiction-specific commitments or schedules.
Information architecture should evolve with financial practice while preserving distinctions between legal duties, record types and enforcement objectives. An institution-based reporting system and a transactional data source can differ in coverage, valuations and error patterns. Combining them requires a clear identity-resolution method, deduplication and evidence of comparability. A larger volume of data does not automatically increase the number of correctly identified obligations.
The economic objective remains the same across instruments: relevant and trustworthy information for authorized decision-makers with proportionate compliance costs. The choice of technical architecture should follow that objective rather than treating either the use of digital technology or its absence as a measure of financial secrecy.
Welfare evaluation and institutional choice
The welfare consequences of financial secrecy depend on the underlying activity and the distribution of information. The potential benefits of confidentiality include protection from misuse, preservation of bargaining positions and greater willingness to use trustworthy formal finance. The potential costs of opacity include inaccurate risk pricing, hidden fiscal liabilities, distorted statistics, investigative expense and unaccountable conflicts of interest. Effective intervention has its own direct and indirect costs.
A transparent accounting example illustrates the distinction between fiscal revenue and welfare. Suppose an enforcement improvement recovers 100 units of tax that was legally due, uses 12 units of real enforcement resources, and induces 8 units of additional real compliance expenditure among participants. The budgetary gain before other adjustments is 88 units if the enforcement cost is borne by government. The resource cost of implementation is 20 units. The recovery itself transfers purchasing power between taxpayers and the public sector; it is not automatically a 100-unit increase in social surplus. Further welfare analysis requires the value of public expenditure, avoided distortions, deterrence, distributional weights and other behavioral and privacy effects. These hypothetical figures are illustrative, not estimates.
The same reasoning applies to mandatory disclosure. A measure that creates a registry may incur fixed infrastructure costs and recurrent costs per record. If it reduces the cost of each appropriately authorized verification, its benefits may accumulate with repeated uses. Redundant data capture across institutions may increase the burden without increasing the quality of information; narrowly targeted verification may improve accuracy at lower resource cost. The preferred arrangement depends on concrete institutional capabilities and risk.
There is also an issue of incidence between large and small financial-service providers. A fixed compliance platform can produce economies of scale that increase concentration. Small compliant providers may face a higher average cost than incumbents. Standardized interfaces, shared public infrastructure and proportionate obligations can reduce avoidable burden, subject to preserving confidentiality and substantive effectiveness. Concentration and competitive neutrality should be studied independently from the information objective.
Policy should be evaluated against realistic alternatives. Secrecy elimination is not a meaningful operational benchmark because all institutions restrict some information for valid reasons. A regime of unlimited disclosure would produce real security and privacy costs, while unrestricted concealment would undermine legitimate claims. The decision concerns the marginal improvement in verifiability relative to its full costs, including distributional consequences. Rights and legal constraints remain substantive constraints rather than quantities presumed freely tradable for fiscal gains.
Research priorities and limitations
A durable evidence program should link institutional rules to observed information quality and then to economic outcomes. These links define successive questions. A legal comparison can establish whether an obligation exists. An institutional evaluation can establish whether records are generated and accessible. A causal outcome study can assess the effect of a specific change on measured behavior. An economic evaluation can consider the incidence and resource costs of that change.
Future empirical work should distinguish account holders from ultimate owners, residence from citizenship, balances from taxable income and the date of announcement from the date of exchange. It should identify the population under study and account for changes in data coverage. For deposit reforms, pre-policy trends, currency adjustments, network interference and heterogeneity across sectors are essential checks. For privacy and governance reforms, independent measurements of information access, correction delays, security incidents and costs would help evaluate trade-offs.
Research combining leaks with administrative data confronts selection, legal and ethical questions. Leaked files represent the institutions and periods from which they originate. Matching success may vary across individuals and systems. Researchers require appropriate authorization, data protection and secure handling; individual identification is not necessary for publication of aggregated findings. Investigations should preserve the distinction between an observed financial relationship, an allegation and verified unlawful conduct.
The literature used here gives stronger evidence about historical behavior and institutional standards than about a single global welfare quantity. The evidence does not justify deriving such a number. It also does not resolve all contemporary jurisdiction-specific rules, ownership thresholds or implementation dates. No current country rankings, legal advice, estimates of annual global losses or claims about individual conduct are presented.
The original contribution of this dossier is the comparative economic interpretation of information relationships, measurement objects and institutional outcomes. It is a synthesis of established concepts and checked research findings, with explicitly bounded claims. It neither proposes a novel named theory nor claims scientific priority. The documentary evidence ledger records the access depth and unresolved verification tasks for later scholarly review of this precise text.
Conclusion
Financial secrecy is economically consequential because information determines the practicability of financial, contractual, fiscal and supervisory decisions. Its costs and benefits attach to specific information relationships: who possesses relevant information, who needs it, whether access is competent and legitimate, and what resources are required to establish its reliability. This framing explains why financial confidentiality and effective accountability can coexist.
The economic supply of opacity depends on providers' incentives and the coordination of institutions. The demand for protection includes legitimate security and commercial privacy as well as opportunities for unlawful concealment. Observed cross-border positions combine these heterogeneous motives and do not, by themselves, establish the legality of financial activity. Wealth concentration and fixed costs make distributional analysis important, while statistical conventions make the distinction between recorded counterparties and ultimate owners indispensable.
The empirical literature documents substantial responses in certain offshore deposit categories to international information-exchange reforms. Its findings support scrutiny of geographic and institutional displacement, sectoral variation and the limits of partial coverage. The same literature warns against equating a decline in recorded deposits with a global reduction in tax evasion or an increase in recovered revenue. The measured outcome, its counterfactual and the relevant population must remain explicit.
Standards for automatic exchange and beneficial ownership demonstrate that a credible legal framework is only part of effective transparency. Information quality, verification, secure exchange, administrative capacity, correction and lawful use are decisive. Greater public visibility can have value for some purposes, yet competent access and protection against misuse are distinct requirements. The strongest economic case is for an information infrastructure that makes legitimate accountability dependable while limiting unjustified exposure and unnecessary duplication.
The economic analysis of financial secrecy thus belongs at the center of Offshore Economics. It connects the allocation of information with incentives, market structure, public revenue, wealth measurement and international cooperation. Future progress should be assessed through evidence on the quality of specific information relationships and the outcomes they enable, not through a single undifferentiated measure of secrecy.
References
- Acquisti, Alessandro, Curtis Taylor, and Liad Wagman. 2016. The Economics of Privacy. Journal of Economic Literature 54 (2): 442–492. DOI: 10.1257/jel.54.2.442. Source
- Alstadsæter, Annette, Niels Johannesen, and Gabriel Zucman. 2019. Tax Evasion and Inequality. American Economic Review 109 (6): 2073–2103. DOI: 10.1257/aer.20172043. Source
- Bank for International Settlements. n.d. Locational Banking Statistics: Overview and Statistical Glossary. BIS Data Portal. Accessed November 18, 2026. Source
- Beer, Sebastian, Maria Delgado Coelho, and Sebastien Leduc. 2019. Hidden Treasure: The Impact of Automatic Exchange of Information on Cross-Border Tax Evasion. IMF Working Paper 19/286. Washington, DC: International Monetary Fund. DOI: 10.5089/9781513519975.001. Source
- Bénétrix, Agustín, Lorenz Emter, and Martin Schmitz. 2024. Automatic for the (Tax) People: Information Sharing and Cross-Border Investment in Tax Havens. Economic Policy 39 (120): 853–895. DOI: 10.1093/epolic/eiae041. Source
- Casi, Elisa, Christoph Spengel, and Barbara M. B. Stage. 2020. Cross-Border Tax Evasion after the Common Reporting Standard: Game Over? Journal of Public Economics 190: 104240. DOI: 10.1016/j.jpubeco.2020.104240. Source
- Court of Justice of the European Union. 2022. WM and Sovim SA v Luxembourg Business Registers, Joined Cases C-37/20 and C-601/20, Judgment of November 22, 2022, operative part. Official Journal C 24, January 23, 2023, p. 3 (2023/C 24/03). Source
- Financial Action Task Force. 2023. Guidance on Beneficial Ownership of Legal Persons. Paris: FATF, March 10. Source
- Financial Action Task Force. 2024. Guidance on Beneficial Ownership and Transparency of Legal Arrangements. Paris: FATF, March 11. Source
- Findley, Michael G., Daniel L. Nielson, and J. C. Sharman. 2015. Causes of Noncompliance with International Law: A Field Experiment on Anonymous Incorporation. American Journal of Political Science 59 (1): 146–161. DOI: 10.1111/ajps.12141. Source
- Johannesen, Niels, and Gabriel Zucman. 2014. The End of Bank Secrecy? An Evaluation of the G20 Tax Haven Crackdown. American Economic Journal: Economic Policy 6 (1): 65–91. DOI: 10.1257/pol.6.1.65. Source
- OECD. 2020. Automatic Exchange of Information: Guide on Promoting and Assessing Compliance by Financial Institutions. Paris: OECD Publishing. DOI: 10.1787/7655bed0-en. Source
- OECD. 2024. Crypto-Asset Reporting Framework and Amended Common Reporting Standard: OECD Releases IT Format for Transmitting Information and Issues Interpretative Guidance. Paris: OECD, October 2024. Source
- OECD. 2025. Consolidated Text of the Common Reporting Standard (2025): Standard for Automatic Exchange of Financial Account Information in Tax Matters. Paris: OECD Publishing. DOI: 10.1787/055664b1-en. Source
- OECD. 2025. Peer Review of the Automatic Exchange of Financial Account Information 2025 Update. Paris: OECD Publishing. DOI: 10.1787/bbf150e4-en. Source
- Tax Justice Network. n.d. About the Index. Financial Secrecy Index. Accessed November 18, 2026. Source
- Zucman, Gabriel. 2013. The Missing Wealth of Nations: Are Europe and the U.S. Net Debtors or Net Creditors? Quarterly Journal of Economics 128 (3): 1321–1364. DOI: 10.1093/qje/qjt012. Source