RESEARCH DOSSIER

Taiwan and Offshore Finance

Abstract. This research dossier examines Taiwan's participation in offshore finance through its domestic offshore financial units, international investment positions, corporate ownership networks, foreign-currency intermediation, taxation, and financial regulation. It integrates official institutional and statistical evidence with scholarly research on ownership networks, offshore investment, and corporate responses to economic-substance requirements. The study distinguishes bank balance sheets from external wealth, administrative investment registrations from realized flows, legal issuer residence from ultimate parent exposure, and reportable accounts from estimates of tax evasion. Its findings establish Taiwan as both a provider of specialized international financial services and an economy whose firms and investors participate in geographically distributed financial arrangements. The author develops an economic interpretation of the interaction between industrial internationalization, financial intermediation, ownership geography, and regulatory incentives. The evidence is documentary and descriptive; illustrative accounting examples clarify mechanisms without representing empirical estimates. Particular attention is given to reporting vintages, the boundaries of observable data, and the conditions required for causal inference.

Keywords: Taiwan; offshore finance; offshore banking units; international investment position; foreign direct investment; corporate ownership; offshore securities; economic substance; controlled foreign companies; financial information exchange; foreign currency; financial stability

Introduction and Research Object

Taiwan occupies an economically distinctive position in offshore finance. It is a technologically advanced producing economy, a creditor economy with large foreign financial claims, a participant in international corporate ownership chains, and a jurisdiction that has deliberately established an offshore financial business regime within its own territory. Each of these descriptions captures a different population of transactions. Their combination explains why Taiwan cannot be studied adequately through a geographical list of small offshore islands or through a single measure of capital held abroad.

Offshore finance in this dossier denotes the organization of financial transactions, contracts, and ownership positions across jurisdictional boundaries, including specifically designated domestic units serving international business. The analytical unit can be an account, a bank branch, a legal entity, a financial instrument, a corporate group, or a position on an economy's external balance sheet. Each has a different reporting perimeter. The central question is how these perimeters interact in the Taiwanese economy and how their interaction affects the allocation of capital, intermediation, taxation, and financial risk.

Zorome (2007), in his operational definition of offshore centers, develops a statistical approach using international financial data. That approach establishes the importance of the denominator and the population sampled. A financial center can be important because of its nonresident-oriented services even when its industrial economy is large; a wealthy economy can have substantial foreign assets even if the proportion attributable to offshore intermediation is modest. An empirical classification must therefore specify what it measures.

This research is economic in orientation. The applicable rules enter as institutional constraints on pricing, funding, investment, and information. Detailed doctrinal interpretation belongs to offshore law, and insurance product design and actuarial solvency belong to insurance research. The unit called Taiwan is the economic and statistical territory described by the Taiwanese authorities; institutional source names are retained for bibliographic accuracy. No legal or diplomatic conclusion follows from this statistical convention.

The principal argument developed throughout the dossier is that a satisfactory economic account must connect Taiwan's domestic offshore units to its outward capital and corporate networks without consolidating their numbers indiscriminately. The evidence identifies important links between these domains, while the available aggregates do not identify a single economy-wide stock of offshore wealth. This is an analytical conclusion of the present synthesis rather than a proposed new theory.

Historical Formation of Offshore Banking

The institutional origin of Taiwan's contemporary offshore banking regime lies in the effort to enlarge the international functions of its financial system during industrial development. The Taiwan Academy of Banking and Finance (n.d.) dates the establishment of the first offshore banking unit to 1983 and identifies its bank as the International Commercial Bank of China, a predecessor of Mega Bank. The original institutional change was the creation of a specialized booking arrangement inside Taiwan.

An offshore facility can internationalize the liabilities and assets of a local financial group without requiring a foreign branch for every transaction. Exporters generally receive claims and payments denominated in foreign currencies, while importers have obligations in those currencies. A banking group that intermediates both streams can provide deposits, trade credit, collection, and payment services. Pooling offsetting customer needs may economize on external transactions, although differences in currency, maturity, counterparty credit, and settlement timing prevent perfect matching.

The commercial value of this arrangement extends beyond taxes. An exporter with a supply chain in several economies faces time-zone differences, documentary requirements, and distinct lender preferences. A relationship with a local international banking unit can preserve knowledge of the group's operations while providing a platform for lending and settlement in international currencies. The institution may therefore be valuable even when no preferential fiscal treatment is available.

Historical interpretation requires chronology. The 1983 founding event establishes the beginning of the formal offshore banking structure. The scope and conditions of the current system derive from later legal texts and regulatory changes. Today's provisions concerning securities or insurance cannot simply be projected back into the initial regime. A historical quantitative study would require successive versions of the law, bank-level accounts, changes in the population of reporting institutions, and period-specific cross-border investment data.

Taiwan illustrates a more general feature of financial globalization: the location of a specialized international unit can be the same city and even the same premises as that of domestic banking. A change in the legal booking perimeter may alter the types of customers, transactions, and revenues assigned to a unit without corresponding physical migration of most employees. Institutional history must therefore track accounting and supervisory architecture as closely as the opening of foreign offices.

Institutional Architecture and Economic Incentives

The 2025 version of the Offshore Banking Act assigns the Financial Supervisory Commission an administrative role and the central bank an operational role. It establishes offshore banking, securities, and insurance activities under a differentiated legal framework. Banking units have separate accounting; the banking provisions distinguish their authorized foreign-currency services, funding, transactions, and tax conditions from ordinary domestic business (Banking Bureau, 2025, arts. 1-16).

The distinction has economic implications. An offshore banking unit (OBU) is best understood first as an authorized business and accounting unit within the banking system. It is not equivalent to a corporation incorporated in another jurisdiction, nor does it necessarily establish a new banking group. Its balance sheet records a set of claims and obligations placed within the legally recognized business perimeter. A foreign financial institution may participate in the same system, while a Taiwanese financial group can also maintain separate overseas operations.

A deposit-account arrangement has several parties: the contractual depositor, the benefiting economic person or organization, the unit that owes the deposit, and possibly a correspondent bank holding settlement balances. Their countries of residence and locations of assets need not coincide. A loan has an analogous structure. When its funds are ultimately used by a related operating company outside the borrowing company's jurisdiction, the immediate counterparty and the underlying credit risk may be concentrated in different places.

Incentives act on the choice of booking and service production. Restrictions on currency conversion and customer access affect the addressable market; exemptions from some taxes or funding requirements can change marginal profitability. Compliance, collateral, operating expenses, and group capital costs must enter the same profitability calculation. A favorable tax provision does not establish the magnitude of a bank's actual economic advantage, because the relevant counterfactual is the same portfolio of business booked or delivered under an alternative arrangement.

The statutory capacity to use existing bank infrastructure produces economies of scope. The relationship banker may know the operating history of a borrower, while the international unit has access to a different funding and settlement infrastructure. Consolidated risk management remains essential precisely because a formal accounting separation need not imply independent capitalization or funding. An intra-group transaction can be commercially useful at unit level yet cancel at group level.

Scale, Composition, and Balance-Sheet Measurement

The central bank's Annual Report 2025 reports 59 OBUs at the end of 2025. Their assets totaled USD 327.54 billion. The corresponding offshore securities units (OSUs) numbered 18 with assets of USD 7.91 billion, while 21 offshore insurance units (OIUs) reported USD 0.80 billion (CBC, 2026a, pp. 92-93). The banking figure was 18.4 percent greater than its preceding year-end level. These are reporting-unit balance-sheet aggregates.

Reporting perimeter Units at end-2025 Assets, USD billion
Offshore banking units 59 327.54
Offshore securities units 18 7.91
Offshore insurance units 21 0.80

Source: CBC (2026a), cited above. Rounded figures preserved as reported. The annual report's OBU income-change sentence contains an apparent inconsistency between amount and percentage; that observation is excluded from the quantitative analysis.

The ratios among the three asset totals demonstrate that banking is the largest of the three measured balance-sheet populations by a substantial margin. They do not measure the relative value of services delivered to clients. A broker can arrange a large volume of trades while retaining little principal risk. An asset manager can advise on investments far exceeding the assets recognized on its own corporate balance sheet. An insurance unit holds assets principally in support of claims, reserves, and other contractual obligations, so premium volume and risk-bearing capacity are relevant metrics alongside assets.

A numerical example clarifies the problem of aggregation. Suppose a domestic bank provides 80 units of foreign-currency funding to its OBU, which uses those funds and an additional 20 units of customer deposits to extend an external loan of 100. The OBU records assets of 100 and liabilities of 100. The domestic funding unit records an inter-unit claim of 80. At the consolidated group level, the internal claim and corresponding obligation are eliminated. Adding 100 and 80 to report 180 of external exposure would be economically incorrect. These amounts are hypothetical, not estimates of Taiwanese funding patterns.

The same discipline applies to statistical growth. An increase in OBU assets might reflect new loan origination, portfolio purchases, customer deposits invested in liquid instruments, valuation effects, exchange-rate translation, or changes in reporting composition. The published year-end aggregate cannot isolate these channels. The appropriate next empirical step is component-level reconciliation of assets and liabilities, followed by maturity, currency, counterparty sector, and geography analysis. An aggregate balance sheet is a scale measure, not a diagnosis of financial vulnerability.

Trade Finance and Corporate Treasury

International production relies on financial transactions at stages that do not coincide with the production of final output. A supplier requires funding before settlement; a buyer may promise payment after delivery; a distributor maintains inventory; a manufacturing subsidiary may borrow for machinery and working capital. Offshore banking provides one possible institutional venue for organizing these obligations.

Consider a Taiwanese parent with an Asian production subsidiary and a European sales affiliate. If the parent arranges a foreign-currency facility for group purchases, the subsidiary can receive imported components before its final sales are settled. The financing decision involves the price of bank credit, collateral, expected operating cash flows, exchange-rate risk, internal transfer prices, and creditor rights. A group treasury can centralize payment administration, but the lender must still evaluate the economic substance of the borrower and the ability of the operating entities to repay.

Two financial relationships deserve particular attention. A trade-related loan accompanies a commercial transaction and is often supported by documentation or anticipated receipts. A group funding facility may be supported by a broader set of corporate commitments rather than a single shipment. The same offshore booking mechanism can support both. Their default probabilities and liquidity characteristics depend on different economic information.

A foreign-currency deposit used for working cash is also different from long-term private wealth. Corporate cash may vary sharply with the settlement calendar, taxes, acquisitions, and debt maturities. A balance observed at year end is not an estimate of household savings. Conversely, payment turnover can be very high while average balances remain low. Interpretation requires matching each statistic to its economic function.

Trade and investment are related without being identical. The financing of a shipment is a financial service; ownership of the factory constitutes a direct investment relationship where the reporting criteria are satisfied; the sale of its product is a trade transaction. A transaction-oriented database that conflates the three would count different claims as if they represented one economic event. A serious study of Taiwan's integration into offshore finance needs parallel trade, investment, treasury, and bank-level records.

External Wealth and the International Investment Position

Taiwan's external balance sheet provides a macroeconomic context for offshore finance. The 2025 international investment position reports external assets of USD 3,267.0 billion and external liabilities of USD 1,917.4 billion. Their difference is a positive USD 1,349.6 billion (CBC, 2026b). The release attributes much of external asset growth to residents' foreign securities purchases and valuation gains, and much of liability growth to appreciation of local equities held by nonresidents.

End-year position 2024, initial release 2025, June 2026 release
External assets, USD billion 3,034.87 3,267.0
External liabilities, USD billion 1,478.65 1,917.4
Net international investment position, USD billion 1,556.23 1,349.6

Sources: CBC (2025), 2024 international investment position; CBC (2026b). The two columns are published vintages rather than a harmonized time series. In particular, the 2026 release gives a decline in net external assets of USD 162.6 billion relative to its revised 2024 baseline, not the difference between the two separately published headlines. This distinction prevents false year-on-year calculations.

Net international investment position (NIIP) is the difference between residents' claims on nonresidents and nonresidents' claims on residents. Its change reflects transaction flows and other changes, including valuation and currency effects. A positive NIIP can fall during a period of continued accumulation of foreign assets if external liabilities increase more rapidly through valuation. The appreciation of local equity held abroad is especially important: it increases measured liabilities without necessarily requiring new external borrowing or a cash payment by the issuing company.

Several institutional sectors contribute to external wealth. The central bank owns reserve assets; firms own foreign equity and related-party claims; households and investment intermediaries can hold foreign securities; banks maintain foreign financial claims. The IIP includes their external positions under residence-based statistical concepts. It neither identifies all beneficial owners of offshore entities nor measures how much resident wealth is concealed.

The scale of Taiwan's external assets therefore establishes international financial integration, but an offshore-specific estimate must isolate its intended population. Taking all external assets as the numerator of a supposed offshore wealth measure would include foreign reserve assets, legitimate direct investment, and ordinary institutional foreign securities holdings. The NIIP is also a net position: an institution exposed to gross external obligations may face liquidity stress even when the economy as a whole is a creditor. The distribution of assets across institutions matters for resilience.

Direct Investment Geography and Corporate Chains

Administrative approvals and registrations describe a different object from the IIP. The Ministry of Economic Affairs' December 2024 investment statistics report approximately USD 44.932 billion in registered outbound investment outside Mainland China for 2024, alongside USD 3.654 billion of separately approved Mainland-bound investment. Approved foreign investment amounted to approximately USD 7.858 billion, with Mainland-source investment reported separately at USD 0.297 billion (MOEA, 2025). These amounts refer to administrative reporting and should not automatically be treated as actual cross-border cash flows.

The Ministry of Foreign Affairs' 2024 trade and investment profile identifies the British territories in the Caribbean as both an important registered outward destination (USD 8.20 billion) and a registered inward source (USD 2.48 billion). These categories represent substantial amounts next to the published outward and inward totals. On the rounded outward denominator of USD 44.93 billion, the Caribbean category accounts for about 18.25 percent; this is a calculated share of the administrative profile, not a measured share of the ultimate geographical allocation of Taiwanese productive capital. The grouping cannot be disaggregated into specific islands from this table.

There are at least three economically plausible explanations for a high intermediate-jurisdiction investment figure. First, an entity in the intermediate economy may operate locally and employ capital there. Second, it may serve as a holding or financing company for assets located in third economies. Third, its legal identity may be used to organize an investment whose ultimate owner or destination is elsewhere, potentially including the originating economy. The published geographic figure cannot choose among these explanations by itself.

A hypothetical chain illustrates the attribution problem. A Taiwan-resident group invests 100 in the equity of an overseas holding company, which invests 80 in a third-economy manufacturer and retains 20 for treasury operations. Immediate outward FDI statistics may record 100 against the holding company's jurisdiction. A look-through ownership analysis may attribute 80 toward the manufacturing economy while retaining 20 in the holding jurisdiction, depending on the allocation method and underlying assets. Treating the 100 and 80 as independent investments of the originating group would double-count the downstream component. The example is illustrative and does not identify an actual company.

Round-tripping is a particular case in which funds return through foreign entities to the economy of origin. It requires tracing ultimate ownership and destination, not merely observing flows in two directions. The fact that the same regional category appears in Taiwan's inward and outward statistics is a research lead and cannot establish that these amounts belong to the same chains.

The economics of direct investment should further distinguish financial transactions during a year from cumulative positions. Reinvested earnings, acquisitions, corporate reorganizations, changes in valuation, and divestments can generate different measures. A longitudinal comparison must keep the definition, denomination, population, and treatment of Mainland investment constant, and report series revisions explicitly.

Corporate Ownership Networks and Classification

Garcia-Bernardo, Fichtner, Takes, and Heemskerk (2017) apply a corporate ownership network analysis to global offshore finance. Their historical network measurement identifies Taiwan as a prominent sink under their methodology and relates this finding to technological companies linked through Hong Kong and Caribbean ownership chains. This is direct scholarly evidence that conventional bilateral statistics can miss important portions of Taiwan's corporate geography.

The result has a specific empirical meaning. It is a classification generated from ownership links, network centrality, and the parameters of a particular study. It neither establishes a permanent jurisdictional label nor determines the conformity of individual holdings with tax and financial regulation. Corporate networks evolve when firms reorganize, new laws take effect, and reporting data improve. A new classification must rerun an appropriately documented method rather than simply repeat a historical result.

The international literature helps specify alternatives. Damgaard, Elkjaer, and Johannesen (2019) distinguish between vehicles interposed in financial ownership chains and real investment in their global FDI network study. Their findings concern global estimation and are not a Taiwan-specific percentage. The model's treatment of unmatched or indirectly measured jurisdictions matters for applying it to Taiwan, whose coverage in international datasets cannot be assumed identical to that of other economies.

A graph representation provides a practical description of corporate groups. Nodes represent legal entities and edges represent ownership or financing relationships. Edge types must distinguish equity, debt, guarantees, and other contractual exposures. Assigning each node its legal jurisdiction makes it possible to describe immediate routes. Assigning employees, physical assets, production, operating revenues, or ultimate ownership requires additional observations. An ownership graph with only legal addresses cannot reveal the full geography of economic activity.

A researcher should record the dates and sources of each relationship. Shareholdings can change in the middle of a reporting year, complex structures may contain cycles, and minority stakes may have disproportionate influence through shareholder agreements. Nominee arrangements and incomplete beneficial ownership can create missing edges. Replacing missing observations with assumptions may be necessary for a sensitivity test, but inferred ownership must remain separate from observed ownership.

Taiwan's industrial production is particularly relevant to the interpretation of ownership networks. A group with substantial foreign manufacturing may legitimately maintain intermediate companies for financing, governance, acquisitions, or regional coordination. A network path passing through a low-tax jurisdiction is a structural fact, while its economic effect and motivation require further evidence. This principle applies equally to Taiwanese and non-Taiwanese groups.

Portfolio Finance and the Location of Issuance

Portfolio securities create a different mapping problem. The issuer can be a financing subsidiary incorporated in a jurisdiction distinct from the operating parent. The holder can be a fund incorporated in another location and administered through a third. The security can be denominated in a currency associated with none of those jurisdictions.

Coppola, Maggiori, Neiman, and Schreger (2020), in Redrawing the Map of Global Capital Flows, associate traded securities with ultimate issuer parents to improve interpretation of international portfolios. The significance for Taiwan is methodological: a portfolio allocated by legal issuer residence can give a different picture of economic sector and parent-country exposure than one reassigned by corporate links. The cited working paper's results do not provide a new Taiwan-specific portfolio reconstruction in this dossier.

Suppose a Taiwanese group issues a USD bond through an overseas financing subsidiary. A Taiwanese investor purchases the bond through a foreign investment fund. The respective statistical accounts could identify a foreign issuer, a foreign fund share, and a liability issued offshore, while important components of the underlying business risk relate to Taiwan. If the bond is denominated in USD and the investor measures wealth in NTD, the investor also bears currency risk unless it is transferred or hedged. Reclassification by parent changes the economic interpretation of counterparty exposure; it does not change the legal issuer or contractual currency.

A useful empirical analysis would therefore present residence-based positions together with a separately constructed ultimate-parent map. It would distinguish fund domicile from underlying portfolio allocation; custody location from beneficial ownership; and derivative hedges from the gross securities positions being hedged. At every transformation, the researcher should preserve totals where accounting identities require it and document unresolved securities.

This also explains why an apparent increase in offshore portfolio holdings can occur without a new wave of offshore entity creation. Valuation changes increase securities positions. Funds may change their asset allocations while maintaining their legal domicile. Companies can issue additional debt through an existing vehicle. Careful interpretation requires data on quantities, prices, currency translation, and issuance structures.

Cross-Strait Finance and Regional Interdependence

Cross-strait financing adds institutional and geopolitical dimensions to the study of offshore banking. The Financial Supervisory Commission's English cross-strait banking regulations explicitly address offshore banking units and third-area branches. The selected historical regulatory text illustrates why the ultimate destination of credit may matter for supervision even when its immediate borrower is located elsewhere. This dossier uses the published provisions to identify that mechanism rather than assert the current numerical ceilings applicable to every institution.

In a regional corporate chain, the location of a holding entity can be affected by several considerations: accessibility of banking services, relations with lenders, corporate governance, local licensing, investment approvals, and the feasibility of remittance across different legal systems. Existing intermediary links may be inherited from earlier patterns of commercialization. Their continued use can reflect path dependence as well as current cost advantages.

Country and jurisdiction exposures must therefore be traced in parallel with contractual borrowing relationships. A bank that provides a facility to a holding vehicle can face deterioration in its loan portfolio when the operating affiliate that supplies repayment cash suffers a disruption in another region. A bank may have direct guarantees, enforceable collateral, or diversified group income as sources of protection. Quantifying the exposure requires those details.

Geopolitical shocks act through trade, financial claims, operational services, and expectations. Payment interruptions affect liquidity before they necessarily affect the net worth of a firm. A decline in demand may impair the underlying business value; a disruption of market access may make refinancing more costly. The diversification of legal entities by geography can reduce some operational exposures while introducing coordination and enforcement costs. It does not automatically diversify the productive assets that support their revenues.

Nothing in the available data quantifies how much of Taiwan's offshore banking balance sheet would be impaired under a specific cross-strait scenario. Such a scenario requires granular sectoral exposures, maturity profiles, currency positions, legal guarantees, collateral, stress parameters, and assumptions about policy responses. The study treats geopolitical risk as an economically relevant mechanism and refrains from numerical forecasts without those inputs.

Taxation, Retained Earnings, and Economic Substance

A multinational group's location decisions respond in part to the timing and level of tax on corporate income. An entity that retains earnings abroad can affect the timing of taxation at the owner level under some tax systems. Controlled foreign company (CFC) rules change that result by defining circumstances in which foreign-company earnings are attributed to a resident taxpayer without requiring an ordinary distribution.

Taiwan's Ministry of Finance (2022) announced CFC commencement in 2023 for enterprise and individual regimes. The policy altered the domestic treatment of qualifying foreign structures. An economic evaluation must distinguish the effective reporting and tax rules applicable to enterprises from those applicable to individuals rather than assume that a single threshold or calculation governs all taxpayers.

The Ministry of Finance's 2026 explanation of the substantial operating activities exemption identifies the relevance of premises, local staff, and a passive-income criterion for the enterprise regime. Economic substance is thus an observable compliance question, not an inference that can be made solely from the jurisdiction of incorporation. Regulatory exemptions and thresholds must be evaluated under the complete applicable law and facts of a taxpayer's case.

The economic incentives can be illustrated without tax advice. A parent holds a foreign entity that earns 100 units, pays 5 of local tax, and retains 95. Under one hypothetical regime, resident-owner taxation arises only on distribution. Under another, some proportion of current foreign income enters the resident owner's current tax base, subject to the rules governing tax credits and exemptions. The expected present value of taxation can change even when the location and productivity of the underlying enterprise remain unchanged. Actual liability requires a legal calculation and is not supplied by this example.

Foreign jurisdictions have also changed their own institutional conditions. Chen, Wang, and Tsai (2025), in their research on economic substance and offshore investment, analyze Taiwan-listed firms over 2016-2021. Their abstract reports changes after economic-substance legislation took effect in certain offshore jurisdictions, including reductions in some subsidiary structures and higher effective cash tax rates. The findings connect foreign regulatory developments to corporate behavior among the sampled Taiwanese companies.

The research design raises several important issues for further investigation. Global economic-substance reforms overlapped with other transparency measures and ultimately the pandemic. Listed firms are more observable than unlisted companies, and firms with particular international structures differ systematically from those without them. Changes in effective tax rates also reflect income composition and tax-base changes. This dossier accepts the published findings as attributed, but has not re-estimated the regressions or inspected their full robustness specifications.

A central lesson is that the impact of reform is multidimensional. Lower incentives to maintain certain vehicles may lead to fewer intermediate entities, different locations, new staffing arrangements, altered dividend policies, or changes in taxes paid. These outcomes do not necessarily occur together. Evaluating compliance through entity counts alone, or success through nominal tax rates alone, would miss the organizational response.

Financial Information Exchange and Enforcement

Regulation of offshore finance increasingly depends on access to trustworthy information about account ownership and tax residence. An institution must determine a customer's reportable status under the applicable rules; authorities must match the information to taxpayers and identify gaps. Financial reporting provides a basis for risk evaluation, while tax liability is established through the relevant assessment procedures.

The Ministry of Finance's 2025 CRS reporting statistics identify 225,222 reportable accounts held or controlled by tax residents of Australia, Japan, or the United Kingdom, with an aggregate year-end 2025 balance of NTD 202.5 billion. The July 2026 release scheduled information exchange with those partners for September 2026. It does not itself verify that the scheduled exchange was completed.

The published amount cannot be interpreted as all foreign-account balances in Taiwan. Its coverage is the reportable population specified in the release. It also cannot be interpreted as Taiwanese residents' wealth abroad or as evidence that the reported assets were untaxed. A reportable account is not equivalent to an individual; one person or entity can have several accounts, and an account can involve controlling persons.

Comparison with the 2017 corporate ownership literature illustrates why temporal attribution matters. A paper identifying offshore characteristics in a historical sample is evidence about that sample and date. Later exchange arrangements change the information environment. Neither historical limitations nor present administrative reporting suffice to demonstrate comprehensive transparency across every partner economy and instrument.

The shift toward information exchange changes the expected return to concealment by increasing the possibility that inconsistencies between reported income and financial positions can be discovered. This expected return depends on how data are collected and matched, the scope of enforcement, and the reliability of ownership classifications. Administrative reporting statistics are consequently an output of a compliance system, not an estimate of its entire deterrent benefit.

The social economics of information also requires protecting legitimate private data. Financial privacy supports personal autonomy and security; regulated confidentiality limits the disclosure of sensitive information to unauthorized parties. The relevant public-policy problem is to enable lawful access, accurate reporting, correction of errors, and accountability while minimizing unnecessary dissemination. The existence of a confidential account is not proof of an offense.

Financial Integrity and the Evidentiary Boundary

Illicit use of offshore finance can involve concealed beneficial ownership, proceeds of crime, fabricated business relationships, or violations of reporting and payment rules. These mechanisms concern specific conduct and require evidence. The economic value of a financial center cannot be inferred simply from the presence or absence of allegations about particular institutions.

The Asia/Pacific Group's 2019 mutual evaluation concerns the anti-money-laundering and counter-terrorist-financing system assessed during the November 2018 visit. The FATF publication distinguishes the assessment of technical compliance from observed effectiveness. This is an important separation: enactment of a rule and successful application of that rule are different outcomes. The historical evaluation is not a current comprehensive audit of all Taiwanese offshore institutions.

Economic research can organize enforcement information into distinct categories: suspicious transaction reporting, administrative findings, formal investigations, adjudicated offenses, asset recovery, and institutional remediation. A report of suspicious activity is a screening signal. Administrative sanctions may concern technical compliance without an underlying proven criminal offense. A conviction establishes a different legally recognized set of facts. Aggregating these categories into a single purported measure of illicit capital would be misleading.

Selection also affects the interpretation of enforcement data. More reports can arise from better detection or increasing suspicious activity, or both; fewer reports can reflect genuine compliance or inadequate detection. A jurisdictional comparison must account for institution counts, transaction volume, legal thresholds, and differences in reporting requirements. Similar ambiguities affect the use of prosecution and confiscation figures.

For Taiwan, the research priority is to connect institution-specific regulatory information to the relevant population of business activities. Cross-border corporate ownership and large bank positions justify careful supervisory attention. Neither is sufficient to attribute wrongdoing. The present dossier does not name individuals or banks as offenders or estimate the aggregate amount of illegally held Taiwanese wealth because it lacks the corresponding claim-specific evidence.

Private Wealth, Securities, and Insurance Services

Private cross-border asset management has a different economic structure from corporate treasury and banking intermediation. A household may seek diversification across issuers, currencies, asset classes, and custody arrangements. The intermediary may provide advice, product distribution, execution, custody, or discretionary investment management. Each function produces fees and risks that can be allocated differently within a financial group.

Taiwan's official program description records the establishment of the Kaohsiung asset management zone in July 2025. Its description of business trials includes services for high-net-worth offshore clients. The initiative places wealth management within a broader strategy for increasing the local capacity to supply international financial services. A program authorization documents an institutional opportunity; it does not establish future client demand, market share, or realized revenue.

The unit of measurement in wealth management is decisive. Assets under management can be much larger than a service company's assets. Assets under custody need not be subject to investment discretion. Advisory relationships may involve portfolios held elsewhere. Fee income measures a stream rather than a stock, and gross flows into a platform can include transfers of previously owned securities rather than new household savings. The same portfolio might appear in several commercial measures.

Insurance adds the function of contingent risk transfer and long-duration saving. The CBC's annual report records a statutory extension of offshore insurance tax incentives through the end of 2035. This establishes a dated institutional development, not an estimate of the resulting competitive advantage. A full economic evaluation would identify premium volumes, customer numbers, distribution expenses, policy denomination, the nature of guarantees, and reinsurance and investment arrangements. These measures cannot be inferred from OIU assets alone.

Domestic insurers' investments abroad belong to a broader institutional portfolio question. Their balance-sheet exposure and liabilities can differ materially from the much smaller explicitly designated OIU perimeter. An insurer holding foreign bonds to support domestic policy obligations may be internationally invested without writing offshore insurance contracts. Understanding this distinction is necessary before interpreting Taiwan's presence in international insurance and asset management.

A comparison with other Asian asset-management markets should use standardized categories and comparable periods. It should separately measure local service production, booking of client assets, ultimate investment allocation, regulatory compliance, and the residence of beneficial owners. A large pool of financial assets creates potential demand for local management services; it is not a sufficient measure of competitiveness in attracting foreign clients.

Foreign Currency, Liquidity, and Financial Stability

Offshore financial relationships can redistribute rather than eliminate risk. An institution that borrows and lends in the same currency may avoid part of its direct currency mismatch while remaining exposed to maturity gaps, credit deterioration, and the cost of funding. A customer whose revenues are denominated in NTD and debt is denominated in USD bears a different currency exposure from one with matching USD revenue.

Hedging transfers an exposure through a contractual relationship with a counterparty. A forward contract or swap may offset future changes in currency value while generating collateral, counterparty, replacement, and rollover risks. The availability of a hedge at the required maturity and size matters as much as the nominal hedge ratio. In times of stress, market liquidity and margin calls can create cash demands before the underlying assets mature.

The CBC's May 2026 financial stability report release describes sound bank asset quality and capitalization in 2025 while noting increased market risks in life insurance and continued external uncertainties. The assessment covers the broader financial system. It cannot be used as a stand-alone certification that every offshore unit has low risk or that its aggregate balance sheet is resilient to every scenario.

A hypothetical bank balance sheet can clarify the transmission mechanism. A unit has USD 100 million in long-maturity loans funded by USD 70 million in short-term liabilities and USD 30 million of stable group funding. Matching currencies does not solve its refinancing exposure: if holders of USD 40 million of short-term liabilities refuse renewal, the institution must replace funding, draw liquidity, or sell assets. Whether it can obtain liquidity from the parent depends on group capacity, legal arrangements, and the timing of stress. The example is not a measured Taiwanese institution.

For institutional portfolios, changes in the NTD/USD exchange rate can affect local-currency asset value. If an institution holds a USD asset worth 100 and has no corresponding foreign-currency liability or hedge, appreciation of the NTD reduces the asset's NTD value at unchanged USD market price. The simultaneous movement of interest rates, credit spreads, and asset prices can reinforce or offset the currency effect. A complete stress test must consider these joint movements.

Macro-financial strength and institutional liquidity therefore require different indicators. Taiwan's positive NIIP reflects economy-wide net external claims; it does not guarantee that the holder of the liquid assets is the institution whose obligations are due. Conversely, a gross foreign-currency obligation need not imply major unhedged risk when the debtor has matching income, suitable assets, or enforceable hedges. Sound risk analysis reconciles gross positions, net positions, and the time profile of cash flows.

Economic Development and the Distribution of Benefits

The domestic economic value of offshore finance depends on the services actually produced, the allocation of funding, the costs of maintaining the regulatory environment, and the distribution of rents. Gross assets give only indirect evidence about domestic value added. A banking unit can hold a large portfolio of low-margin claims, while specialized financial advisory services may employ fewer balance-sheet assets relative to their income.

Potential benefits of developed international financial services include reducing transaction costs for exporters, increasing access to diversified lenders and investments, improving corporate treasury services, supporting internationally mobile professional expertise, and producing fee and interest income. Their magnitude depends on the counterfactual. If transactions would otherwise be served by a domestic unit, the incremental benefit of offshore booking may be limited. If transactions would otherwise be served entirely abroad, the retention of services may generate additional local activity.

Potential costs include foregone public revenue where incentives apply, supervision and compliance expenses, heightened contingent demands on financial groups, and incentives to favor formal booking over productive activity. Measuring these requires information beyond unit balance sheets. The fiscal effect of a tax exemption is not the nominal statutory rate multiplied by gross assets; the relevant tax base is taxable income under the alternative regime, adjusted for changes in behavior and operating costs.

Distributional outcomes are likewise empirical. Large firms may have greater access to legal expertise and cross-border treasury arrangements; smaller exporters can benefit from shared banking infrastructure. Professional service providers may capture part of the returns through specialized fees. Household investors with large financial portfolios benefit from broader diversification opportunities, while consumers who lack access to complex products may receive little direct benefit. These mechanisms warrant inquiry but do not yield distributional shares without microdata.

Economic development also affects the demand for offshore finance. As production expands internationally, firms need contracts and funding across jurisdictions. As retained income and financial savings accumulate, institutions seek diversified investment opportunities. Financial services can then support real-sector development, while industrial and household demand supports the financial sector. The relationship is bidirectional, and observational co-movement cannot establish causality.

A policy evaluation would compare local employment and value added, service exports, funding costs, fiscal outcomes, consumer protection, and financial stability before and after relevant reforms, while controlling for external trade cycles and asset prices. It would also test whether institutions have relocated existing business within their groups rather than produced new services. A program capable of generating meaningful local expertise and market infrastructure can succeed even if no particular target for gross offshore assets is met.

Research Methods, Reproducibility, and Limits

The dossier combines a claim-scoped documentary review, comparative institutional analysis, statistical reconciliation, and analytical examples. The evidence cutoff is November 18, 2026. It uses annual official observations with explicitly identified reporting vintages; it does not claim that every monthly series or every legal amendment has been exhaustively reviewed. No original survey, firm-level regression, audit file, or confidential account data was obtained.

A reproducible investigation of Taiwanese offshore finance would preserve five separate classes of evidence. Unit-level financial statements describe banking and other designated activities. Balance-of-payments and IIP releases describe residence-based flows and positions. Administrative investment records describe registrations and approvals. Corporate disclosures establish ownership, activities, and financing relationships. Supervisory and fiscal records describe legal constraints, reporting, and enforcement. The classes can be joined only where definitions, dates, and identifiers are compatible.

The most important statistical tests are straightforward. Compare totals within a single publication vintage; preserve the published accounting unit; identify the denominator for any ratio; separate annual flow data from year-end balances; distinguish group consolidation from unit-level aggregation; and record changes in population or classification. The offshore investment share reported in this dossier is simple arithmetic on the published 2024 outward-investment geographic profile. It is not an econometric estimate.

Evidence is also differentiated by source depth. Official HTML releases and selected statutory articles were read directly. The 2025 CBC annual report and the Ministry of Foreign Affairs investment profile were retrieved as PDF text; visual screenshot verification failed and is explicitly recorded. The review of the Taiwan-specific economic substance article is limited to institutional abstract and bibliography. The NBER portfolio paper is supported by an indexed record and institutional search corroboration; direct page retrieval failed. The IMF working papers are used for their verified abstracts and methodological propositions, not claims about uninspected tables. The historical ownership-network article is used for its accessible selected results. No claim of complete full-text review or replication is made.

For identification of causal effects, a credible design needs measurable intervention timing and a comparison population. A study of the 2023 Taiwan CFC reform might compare firms with pre-reform exposure to qualifying foreign structures against comparable firms with different exposure. It would need to address anticipation, simultaneous foreign reforms, sector differences, and changes in profitability. A study of the asset management zone might examine service activity relative to comparable non-zone institutions, recognizing selection into participation. These are possible empirical designs, not results generated here.

Three gaps remain particularly important. First, the shares of OBU assets attributable to types of counterparties and ultimate users need detailed balance-sheet and exposure records. Second, the extent of intermediate-jurisdiction corporate chains requires firm-level linkage rather than geographic approvals alone. Third, comprehensive estimation of offshore household wealth would require evidence on beneficial owners and foreign financial assets that the sources reviewed here do not provide. These limits are substantive characteristics of the current research evidence, not permission to replace missing values with invented estimates.

The related Science is Offshore Economics. Geographic, legal, insurance, and behavioral research within the wider program must preserve the independently specified questions and evidence of those disciplines. No inactive dossier route is presented as a published source.

Conclusion

Taiwan's role in offshore finance is best explained through the interaction of institutional specialization, international production, accumulated external claims, corporate ownership structures, and changing regulatory requirements. Its offshore banking units provide an established domestic platform for international financial business; overseas holdings and issuance arrangements connect Taiwanese corporations and investors to financial jurisdictions outside Taiwan.

The quantitative evidence demonstrates the importance of keeping statistical perimeters distinct. At year-end 2025, official reports identify a substantial offshore banking balance sheet and an even larger economy-wide positive external investment position. Administrative investment records reveal significant registered relationships with intermediary geographic categories, but they do not trace the final uses or ultimate owners of the invested capital. Corporate network research gives an additional lens through which these relationships can be interpreted.

The central analytical conclusion is that offshore finance can increase the geographical distance between the legal recording of a claim and the economic activity or risk that supports it. In Taiwan, that distance must be investigated at the level of individual units, instruments, counterparties, and corporate groups. The existence of intermediate entities, foreign-currency business, or external wealth does not establish a single motive or a measure of unlawful conduct. A rigorous inquiry assigns such conclusions to the appropriate evidentiary level.

Recent tax and information-reporting developments demonstrate an institutional environment in motion. CFC rules alter incentives concerning retained foreign earnings, and economic-substance requirements and information exchange can change corporate structures and reporting practices. Local asset management initiatives seek to expand financial services capacity. Their real effects on domestic value added, public revenue, ownership structures, and financial stability require continuing empirical evaluation.

Taiwan belongs in the scientific study of the global offshore system because it shows how offshore activities can coexist with an advanced industrial economy and how capital-market geography can diverge from production geography. Its experience strengthens the case for evaluating offshore finance through disaggregated ownership, financial, and institutional evidence. This conclusion is a synthesis of documented findings and established economic reasoning. It does not claim a new named theory, a contemporary universal ranking, or an independent empirical estimate of hidden wealth.

References

Asia/Pacific Group on Money Laundering. 2019. Mutual Evaluation Report of Chinese Taipei. October 2019; FATF assessment landing page.

Asian Asset Management Center. n.d.. Strengthening Asset Management: Kaohsiung Zone of the Asian Asset Management Center. Program description and trials.

Central Bank of the Republic of China (Taiwan). 2025. Taiwan's International Investment Position (2024). Press release, June 13, 2025.

Central Bank of the Republic of China (Taiwan). 2026a. Annual Report 2025. Central Bank Operations, pp. 92-93.

Central Bank of the Republic of China (Taiwan). 2026b. Taiwan's International Investment Position (2025). Press release, June 15, 2026.

Central Bank of the Republic of China (Taiwan). 2026c. The Bank Releases Financial Stability Report 2026. Press release, May 29.

Chen, Ming-Chin, Jui-Chih Wang, and Chen-Yu Tsai. 2025. The Effects of Economic Substance Act on Offshore Investment Structures and Tax Avoidance. NTU Management Review 35(1): 89-140. DOI: 10.6226/NTUMR.202504_35(1).0003.

Coppola, Antonio, Matteo Maggiori, Brent Neiman, and Jesse Schreger. 2020. Redrawing the Map of Global Capital Flows: The Role of Cross-Border Financing and Tax Havens. NBER Working Paper 26855. DOI: 10.3386/w26855.

Damgaard, Jannick, Thomas Elkjaer, and Niels Johannesen. 2019. What Is Real and What Is Not in the Global FDI Network?. IMF Working Paper 19/274. DOI: 10.5089/9781513521527.001.

Financial Supervisory Commission, Banking Bureau. 2025. Offshore Banking Act. English consolidated text, amended July 16, 2025.

Financial Supervisory Commission. 2011. Regulations Governing Banking Activity and the Establishment and the Investment by Financial Institution Between the Taiwan Area and the Mainland Area. English version, selected provisions on offshore and third-area branches.

Garcia-Bernardo, Javier, Jan Fichtner, Frank W. Takes, and Eelke M. Heemskerk. 2017. Uncovering Offshore Financial Centers: Conduits and Sinks in the Global Corporate Ownership Network. Scientific Reports 7: 6246. DOI: 10.1038/s41598-017-06322-9.

Ministry of Economic Affairs, Republic of China (Taiwan). 2025. Taiwan FDI Statistics Summary Analysis (December 2024). Department of Investment Review, annual administrative investment figures.

Ministry of Finance, Republic of China (Taiwan). 2022. CFC Rule Shall Be Enforced from the 2023 Taxable Year for Enterprises and January 1, 2023 for Individuals. Press release, March 16.

Ministry of Finance, Republic of China (Taiwan). 2026a. Conditions for Profit-Seeking Enterprises to Apply for the Substantial Operating Activities Exemption under the CFC Rules. Northern Area National Taxation Bureau, explanatory release.

Ministry of Finance, Republic of China (Taiwan). 2026b. Statistics of 2025 CRS Data Reported by Financial Institutions. Press release, July 30.

Ministry of Foreign Affairs, Republic of China (Taiwan). 2025. 2025-2026 Taiwan at a Glance. Trade and Investment Profile (2024).

Taiwan Academy of Banking and Finance. n.d.. Major Events in Banking. Financial Education exhibition, 1983 entry.

Zorome, Ahmed. 2007. Concept of Offshore Financial Centers: In Search of an Operational Definition. IMF Working Paper 07/87. DOI: 10.5089/9781451866513.001.