RESEARCH DOSSIER

Netherlands and Offshore Finance

Abstract. The Netherlands is a major intermediary in international corporate ownership, direct investment, and financial income flows. This Research Dossier examines the historical development, institutions, statistical measurement, economic incidence, and changing policy environment of its conduit sector. It synthesizes historical scholarship, multinational ownership-network studies, research using Dutch special purpose entity records, and government and central bank evidence. The analysis distinguishes gross positions from local productive capacity; ownership-chain geography from ultimate investment; and payment volumes from potential fiscal losses. Official data document a substantial reduction in multinational financial holding companies and in their assets relative to GDP over 2017–2025, alongside a large remaining financial sector. Targeted cross-border income flows have also fallen substantially, but policy attribution requires counterfactual analysis and attention to anticipatory adjustment, changes in reporting and relocation. The dossier develops an integrated economic interpretation of the Dutch role, evaluates domestic and international incidence, and identifies evidence requirements for assessing regulatory effectiveness. It presents original analytical synthesis without claiming a new theory or an independently estimated causal effect.

Keywords: Netherlands; offshore finance; conduit economy; special purpose entities; foreign direct investment; corporate ownership; international taxation; withholding taxes; financial intermediation; tax competition; public revenue; financial statistics

Introduction and Research Design

The Netherlands occupies a central position in the international organization of corporate capital. Its importance arises from the combination of a productive national economy, a sophisticated financial and administrative infrastructure, and a set of institutions that have historically made the jurisdiction attractive as an intermediate location for ownership and financing. This structure gives economic reality to the distinction between the location of a financial claim and the location of the activity generating its return.

The research question is how the Dutch conduit function emerged, how it can be measured, what it contributes to the host economy, how it affects other economies, and how it responds to domestic and international reforms. The unit of analysis is the resident institutional entity and its relationships with the nonresident units that own, finance, or receive income from it. The country is treated as a node in a network rather than as a homogeneous financial instrument. The analysis concerns the Netherlands as a statistical economy; separate Caribbean jurisdictions enter only when historical links illuminate the emergence of Dutch financial intermediation.

The study combines historical interpretation, institutional economics, macroeconomic accounting, comparative tax economics and a critical synthesis of published empirical research. The evidence cutoff is November 18, 2026, with individual observations dated by reporting periods and source vintages. The most recent directly retrieved DNB holding-sector release used here covers year-end 2025. The study does not estimate a new causal coefficient or imply that its literature search exhausts all research. The analytical contribution consists of relating several different indicators of financial intermediation to one another and drawing implications for measurement, welfare and policy evaluation.

Three distinctions guide the inquiry. First, an ownership position is a balance-sheet relationship, whereas an income flow represents a return accruing over a period. Second, the institutional residence of an entity locates a claim statistically, whereas the ultimate controlling enterprise and its productive assets may be elsewhere. Third, the private return from locating an entity in a jurisdiction and the public welfare effects on that jurisdiction or another jurisdiction are distinct quantities. The later chapters apply these distinctions to the Dutch case.

Historical Formation of the Dutch Offshore Function

The postwar history of Dutch offshore finance demonstrates that offshore arrangements developed through connections between established metropolitan financial institutions and territorially distinct economies. Van Beurden and Jonker (2021), in A Perfect Symbiosis, trace the development of Curaçao's offshore sector to initiatives involving Dutch banks, multinational enterprises and official support for extending treaty connections. Their historical account locates offshore development within corporate and state institutions rather than treating island jurisdictions as self-contained inventions.

The relevance of that history extends beyond Curaçao. A jurisdiction gains financial significance when a legal relationship in one territory is recognized by parties in another. Recognition requires contractual confidence, predictable administration and the expectation that intermediaries can complete transactions repeatedly. Treaty relations supply another component of this infrastructure by organizing overlapping claims to tax the same income. The economic advantage of a particular location therefore depends on its connections. A provision that is attractive for one bilateral relationship may be irrelevant to another.

This history also explains why the Dutch case cannot be read solely through the fiscal rate applicable to locally generated business profit. International groups require organizational centers, and these centers may serve domestic and foreign subsidiaries simultaneously. The postwar growth of international corporations expanded the volume of financial relationships requiring administration. A legal entity that previously existed to simplify control can subsequently acquire an additional tax-related role; a structure originally chosen for tax reasons can later serve commercial funding purposes. The functions present at any one moment must be established by evidence rather than inferred from the historical origin of the structure.

Institutional persistence is possible even after an advantage disappears. A group that changes its organizational design must replace contracts, update governance, reassess debt covenants and sometimes obtain regulatory or shareholder consent. Service providers may redirect expertise to more substantive corporate activity. The Dutch economy's accumulated institutional infrastructure thus makes the response to reform a question of changing functions as well as changing numbers of companies. This interpretation follows from the historical evidence but does not assert that a single historical route caused the contemporary balance-sheet totals.

Financial Geography and the Conduit Function

A conduit economy intermediate in corporate ownership transmits financial claims and associated income between other economies. Its significance can be assessed at several levels: the distribution of corporate entities, the value of their cross-border positions and the income attributable to these positions. Garcia-Bernardo, Fichtner, Takes and Heemskerk (2017), using a large ownership graph in Conduits and Sinks, identified the Netherlands with the United Kingdom, Ireland, Singapore and Switzerland as five major intermediate locations in their observed network.

Ownership-network centrality captures the relational position of the country. It measures a different phenomenon from banking-center size or equity-market capitalization. A company database registers direct ownership relationships; paths constructed from those relationships can identify intermediate jurisdictions. Even when the underlying ownership data are accurate, a path describes legal control or stakes as represented in the database. The valuation of an ownership chain depends on the data source, the treatment of indirect ownership and the method used to distribute value among branches.

Three forms of location must therefore be tracked. The first is where the entity is resident; the second is where the direct counterparty resides; the third is where the ultimate investor, final host enterprise or productive assets reside. These dimensions can coincide for an ordinary direct investment but diverge for a multilayered group. A Dutch intermediate link may connect two foreign economies without directing their economic activity toward the Netherlands.

The network classification establishes a historical empirical role, not a timeless rank or legal designation. It cannot certify that all Dutch-resident financial institutions have an offshore function. It also does not show that every chain identified in the original database persists after substantial tax reforms. A longitudinal replication would need harmonized ownership records, entity identifiers and valuation methods. The observed position of the Netherlands supplies a reason to undertake those measurements, not substitutes for them.

The Statistical Perimeter of the Conduit Sector

The boundaries of the measured sector determine the substantive question answered by any reported total. In its pass-through entity definitions, DNB distinguishes internationally defined special purpose entities (SPEs) from a broader group of multinational-owned captive financial institutions within sector S.127. Some Dutch holdings that fail the SPE test still participate in cross-border intra-group funding. Conversely, historical special financial institution classifications do not map exactly to the modern sector.

Under the OECD's 2025 special cases chapter, an SPE is a resident institutional unit with minimal local physical presence, a limited workforce of at most five employees and nonresident control, among the other specified characteristics. The broader Dutch statistical category includes entities that can perform important holding functions while falling outside this narrower designation. Population differences matter when analysts compare series called shell companies, conduit companies, financial holdings or SPEs.

A conduit-sector asset figure is the aggregate financial size of the covered institutional population. It includes the financial claims actually recorded for that population. Total inward foreign direct investment (FDI), by contrast, describes investments defined by the direct investment relationship across the entire resident economy. Total outward FDI represents the corresponding direction of investment influence outside the economy. Neither is restricted to Dutch conduit companies. Consequently, dividing a conduit-asset total by all inward FDI produces a ratio of nonidentical populations rather than a valid estimate of the proportion of inward investment that is phantom.

Likewise, cases in which operational firms participate in intragroup financial transactions are economically relevant even if the firms are not classified within the conduit sector. A manufacturer with employees, factories and a research division may hold foreign subsidiaries and perform financial functions. Removing all its cross-border positions from measures of genuine activity would be just as problematic as counting all the positions of a lightly staffed intermediary as new Dutch production.

Researchers should preserve the original sector name and time vintage for each observation. In particular, a revised central bank series should take precedence over an unrevised series for comparisons over the overlapping period, while the earlier series can be studied as evidence about statistical revision. Switching silently between entity populations can generate a false acceleration or reversal of sectoral change.

Direct Investment, Balance Sheets, and Domestic Production

Foreign direct investment denotes an investment relationship involving a lasting interest and influence. DNB's 2024 FDI and income analysis uses the conventional threshold of at least 10% of voting power. The OECD (2025), in the fifth FDI Benchmark Definition, develops separate reporting dimensions for investment purpose, pass-through funding and ultimate investor and host economies.

These dimensions permit a sharper distinction between investment in a productive enterprise and investment that establishes an intermediate financial claim. A foreign parent can contribute 100 monetary units to a Dutch holding company; the holding can use those funds to acquire shares in an enterprise outside the Netherlands. In the stylized case, the Dutch entity records an equity liability of 100 and a corresponding equity asset of 100. Gross external positions attributable to the entity are substantial even though it has created no new productive capacity at its location. If the final enterprise invests 100 in productive capacity, counting both international positions as independent additions to world productive capital would double count the underlying allocation.

There is a separate accounting distinction between a stock and a flow. An investment position at the end of a period equals the opening position plus net transactions, price and exchange-rate revaluations, and other changes in volume. The value of a claim can fall when exchange rates change although an entity remains in the same jurisdiction. An exit by one group can be offset by an increase in the value of positions held by surviving groups. It follows that changes in the number of registered entities and the stock of financial assets need not move proportionally.

An analogous problem arises when financial income is compared with GDP. Suppose a foreign subsidiary pays its Dutch parent a dividend of 8 units and the parent distributes most of it to a foreign shareholder. The dividend is property income; the locally produced services supporting the chain may have much smaller value. A ratio of 8 to the Dutch entity's local wages would measure scale relative to an input, not a rate of productive return. National accounts provide separate categories for production, property income and cross-border assets and liabilities.

Damgaard, Elkjaer and Johannesen (2019), in What Is Real in FDI, estimate that almost 40% of global FDI could be phantom, with the statistical location of investment obscuring the geography of underlying activity. Their adjusted network is an analytical estimate with assumptions about allocation, not an observation of every ultimate asset. The result cannot automatically determine the phantom fraction of contemporary Dutch direct investment. Its value for this dossier lies in demonstrating why gross FDI geography has to be interpreted with supplementary statistics.

Corporate Tax Institutions and Commercial Incentives

A Dutch intermediary can be useful because an international group needs a stable administrative center, an efficient treasury or a predictable corporate ownership structure. A country can support those uses through ordinary company law, tax administration, accounting capabilities, available credit and skilled professional services. These features are complements: a theoretically advantageous fiscal position may remain unused if organizations find the institutional environment unreliable.

Tax design nevertheless matters. According to the Dutch government's corporate income tax summary, the 2026 statutory rates are 19% up to €200,000 of taxable profit and 25.8% for the portion above that threshold. Its participation exemption applies to qualifying interests, with a stated minimum ownership threshold of 5% and additional conditions. The mechanism limits cascading taxation of profit distributed within corporate groups; its policy rationale is related to the avoidance of repeated taxation of the same corporate income.

The economic question is how this treatment interacts with rules in the other economies connected to a chain. A participation exemption at the intermediary level can be neutral with respect to real investment if it avoids double taxation. The same rule can permit a reduction in the overall effective taxation of income when combined with divergent treatment in other territories. An institutional mechanism can therefore have both an investment-neutrality effect and a tax-base allocation effect depending on the international environment.

Several margins of choice deserve distinction. The location of a factory may be influenced by workforce skills, transport and market access. The ownership of that factory may be held by a subsidiary located elsewhere. Financing of the subsidiary may be organized in a centralized treasury. Decisions about control, funding and productive location need not be taken together. Policy measures that raise the cost of one type of financial arrangement can alter an ownership or financing layer while leaving the productive geography unchanged.

A study of tax incentives must therefore consider both alternative arrangements and the conditions for lawful access to the relevant treatments. The statutory rate alone does not determine the effective burden; a comparison should consider the applicable base, credits, exemption eligibility, contractual rights and compliance costs. This monograph analyzes those economic mechanisms rather than recommending any particular tax structure or determining the treatment of individual taxpayers.

Treaty Networks and the Geography of Investment

Weyzig (2013), studying Dutch SPE microdata in Tax Treaty Shopping, identifies the treaty network and dividend withholding rates as determinants of the routing of FDI. His estimate of approximately €1.6 trillion of FDI routed through the Netherlands in 2009 documents the scale of the function in that historical dataset. The reported result supports the importance of treaty incentives rather than a claim that every observed investment used an improper treaty benefit.

Van ’t Riet and Lejour (2018), in Optimal Tax Routing, use a network formulation of the corporate tax system. Their model places the Netherlands among the major intermediate locations and reports an average potential reduction in the modeled burden on repatriated dividends of about six percentage points. The model demonstrates the importance of international links over isolated national rates. Its results are conditional on the specified fiscal rules, assumptions concerning access and the network at the time of measurement.

The distinction between potential and realized outcomes is central. A network can make an alternative route appear attractive, yet a group may lack eligibility for a treatment, incur costly compliance obligations or face business restrictions that invalidate the modeled comparison. Conversely, a route selected for operational reasons may benefit from a treaty without the benefit being the cause of the ownership decision. Evidence about a tax incentive is stronger when it is accompanied by observations of changes in behavior after the incentive changes.

Treaty formation itself is endogenous. Countries with intense existing investment relationships may be more likely to negotiate treaties; geographically and historically related economies may also have similar business networks. Regressions relating FDI to treaty presence cannot automatically establish the causal magnitude of an individual treaty. Firm-level observations help by identifying routed ownership, but research still needs an identification strategy when the aim is to measure causal effects.

The economic rationale of bilateral treaties includes avoiding double taxation and reducing uncertainty. A treaty's social assessment must consider these gains alongside potential reductions in the source country's tax base. The source of income, residence of the investor and intermediate location can be in three different economies, making the aggregate impact of a bilateral change dependent upon relationships outside the pair. That is the substantive lesson of the Dutch experience for international public finance.

Income Channels and the Evidence on Tax Gains

Dividends, interest and royalties arise from different economic relationships. Dividends distribute equity returns; interest remunerates debt; royalties compensate for rights to use intellectual property. Different accounting and tax treatments mean that an observed change in one category cannot be extrapolated mechanically to the other categories.

Lejour, Möhlmann and van ’t Riet (2022), in The Immeasurable Tax Gains, examine bilateral origins and destinations of income passing through Dutch SPEs. Their data contain €140 billion of flows for 2016. Combining flow data with tax information, the authors calculate a potential royalty-related tax gain of almost €3 billion for a single year. They do not measure comparable savings for the other two categories and explain the limits of the available information. This finding is more discriminating than an aggregate assertion that every payment through a conduit generates a fiscal loss.

A lower tax charge on a payment can reflect a treaty provision, a difference in the characterization of income, a credit available to the recipient or other components of the tax system. The economic loss to a particular treasury is the difference between its observed receipts and the receipts under a relevant counterfactual that takes business behavior into account. The tax saving estimated for a firm need not equal the fiscal loss of a single country. Payments might be reorganized, the tax liability might shift between countries or the underlying activity might change.

The study of Dutch conduit structures should accordingly separate the face value of payments, the taxable portion of those payments, tax liabilities under the actual arrangement and counterfactual liabilities. An estimate obtained by applying a statutory rate to all gross cross-border returns implicitly assumes that the payments would occur unchanged, that they would be taxable at that rate and that no relief would apply. Those assumptions require independent evidence.

There is also a timing issue. Reinvested earnings may be recorded as direct investment income even when no corresponding cash distribution has occurred. A dividend may distribute profit that accrued in an earlier period. Royalty payments can be large in relation to current local production if they concern intellectual property developed elsewhere. Consequently, a financial-income time series cannot by itself measure current profit generation or a cash tax base. These distinctions become particularly important when comparing income data before and after withholding tax reform.

The Scale and Transformation of Dutch Financial Holdings

DNB's statistical reporting documents both the extraordinary scale of resident multinational financial holdings and a substantial contraction relative to the national economy. In the 2024 statistical revision, the bank reports a 2016 sector total of €4,832 billion, equivalent to 671% of GDP. The revised 2021 position was €4,567 billion, or 512% of GDP; the 2023 estimate reported in that release was €4,648 billion, or 435% of GDP. The revision incorporated adjustments to loans and population coverage.

A later DNB release (2026), Financial Holding Companies Almost Halved, reports that the 2017 population exceeded 14,000 and the 2025 population was just below 8,000. The associated balance-sheet ratio declined from 652% to 401% of GDP in this reporting vintage, with a year-end 2025 sector total of €4,732 billion.

Observation Balance-sheet assets (€ billion) Assets/GDP Reporting vintage
2016 4,832 671% DNB 2024
2021 4,567 512% DNB 2024
2023 4,648 435% DNB 2024
2017 — 652% DNB 2026
2025 4,732 401% DNB 2026

This table preserves the two vintages rather than splicing them into a newly constructed homogeneous series. The ratio in the later vintage declined by 251 percentage points from 2017 to 2025, or approximately 38.5% relative to its starting ratio. The magnitude of the 2025 position demonstrates that a major sectoral contraction is compatible with a large residual role in the global economy.

The entity counts show a decline of substantial magnitude but are rounded thresholds, so calculating an exact percentage from 14,000 and 8,000 would be spurious precision. The position-to-GDP ratio can decrease because assets fall, because nominal GDP increases, or through both effects. For example, a nominal financial position of 400 relative to GDP of 100 yields 400%; if GDP increases to 125 while assets remain at 400, the ratio falls to 320%. A relative contraction need not imply the same percentage reduction in nominal claims.

These observations have implications for policy evaluation. The number of establishments counts participation in a particular organizational arrangement. The financial size weights those establishments by their balance sheets. The ratio to GDP measures the scale of financial intermediation relative to production. A decline in the first two indicators with a smaller or larger movement in the third would be entirely possible. They should be reported together, but never as if they were interchangeable observations of one latent variable.

External Positions and Income-Flow Adjustment

DNB (2024) reported in its external income statistics that total direct investment into Dutch companies at the end of 2023 amounted to around €4.8 trillion, and outward direct investment exceeded €5.6 trillion. The inward position attributable to low-tax jurisdictions declined from €449 billion in 2016 to €165 billion in 2023. This bilateral grouping has a policy-related meaning that differs from the institutional classification of conduit companies.

In DNB's 2025 income-flow release, corresponding end-2024 direct investment positions are €4,946 billion inward and €5,759 billion outward. The ratios to GDP are 436% and 508%. The difference between the two published directional stocks is €813 billion. That arithmetic difference should not be presented as the country's net international investment position, as the latter includes other functional investment categories and may be presented under a different accounting principle.

The same release describes a profound fall in the combined selected income flows to low-tax jurisdictions: €37 billion in 2019, €5 billion in 2020 and more than €6 billion in 2024. This is a shift in a particular set of cross-border payments, not the value of all foreign investment or the disappearance of the entire intermediary function. The composition of the payments also changed, with dividends gaining prominence relative to earlier royalty flows.

A useful inference is that the contraction of one type of income channel can occur more quickly than the contraction of the existing financial stock. A parent can stop making royalty payments through a particular country while it maintains control of subsidiaries located there. A firm can likewise change the contractual location of its intellectual property without immediately retiring a finance company. The adjustment of an annual flow and the retirement of a legal entity have different economic and administrative costs.

The observed persistence of large inward and outward FDI positions after the income-flow decline is consistent with institutional adaptability. It may also reflect valuation, exchange rates, growth of unaffected activity and differences in data coverage. The statistical evidence alone cannot quantify the contribution of each explanation. A complete decomposition would require longitudinal data on firms, positions, transactions and income across the reform years.

Domestic Value Creation and Fiscal Incidence

The economic assessment of the Dutch conduit function requires accounting for the local services supplied to multinational enterprises and the way in which those services are produced. Incorporation, auditing, administration, legal advice and financial management can generate employment, fee revenue and knowledge accumulation. The associated financial assets cannot serve as an estimate of their output.

The 2021 Conduit Companies Committee's findings, summarized by the government in High Number of Shell Companies, place the 2019 sector at about 12,400 companies with approximately €4.5 trillion in assets. The committee estimates direct employment at 3,000–4,000 and treasury receipts at approximately €650 million for 2019. It reports average annual flows of €170 billion for 2015–2019. These are estimates for a particular identified population and historical period.

The ratio of estimated direct tax receipts to the reported assets is approximately 0.0144% (650 million divided by 4.5 trillion). This descriptive calculation is not a corporate income tax rate, a tax avoidance rate or the return of an asset portfolio. Its limited interpretation is that the direct fiscal receipts reported by the committee were small relative to the financial positions of the population it studied. Likewise, the committee's direct employment estimate does not include every employee in the professional-service ecosystem. A rigorous estimate of indirect employment would need to isolate the demand attributable to the sector.

Domestic public welfare depends on service-sector value added, income accruing to residents, fiscal receipts, administrative and oversight costs and the opportunity cost of inputs. A broader international welfare evaluation also includes effects on foreign tax revenues and other parties. The sector's private benefits may accrue to international group shareholders or clients rather than Dutch residents. The location of the intermediary therefore cannot be used as a proxy for the location of ultimate gains.

When a conduit company exits, the full amount on its balance sheet cannot constitute lost Dutch GDP. The relevant change in production arises from the domestic contribution and the response of linked service providers. If their resources are redirected toward other activities, gross employment losses and longer-term effects can differ. The policy decision is about the value of legitimate financial and administrative services under rules that account for their effects on other economies.

Reforms of the Dutch and International Tax Environment

The Dutch reform agenda has targeted specific mechanisms associated with profit relocation and tax-base erosion. The government's tax avoidance policy summary records conditional withholding taxes on certain group interest and royalty payments beginning in 2021, with dividends included from 2024. The coverage relates to specified low-tax destinations and relevant arrangements; the measure should not be described as a uniform withholding tax on all international group income.

The government's 2024 tax avoidance monitoring also discusses the earnings-stripping restrictions adopted under ATAD and changes intended to eliminate advantages arising from hybrid mismatches. It acknowledges that the consequences of all more recent measures could not yet be fully assessed at the date of that monitoring report. This is significant because one reform can influence multiple margins simultaneously, including the amount of debt used, the classification of payments and the location of intermediate entities.

The Netherlands participates in the international minimum tax framework. The Dutch government's 2024 review confirms the introduction of the Minimum Tax Rate Act 2024. The EU's minimum taxation directive establishes a 15% minimum effective rate for qualifying multinational and large domestic groups with consolidated revenues of at least €750 million in at least two of the four preceding fiscal years (Council of the European Union 2022, arts. 2–3). The scope, exclusions, timing and computation of top-up obligations require separate legal analysis.

From an economic perspective, the policy measures operate through distinct mechanisms. A conditional withholding tax raises the cost of a covered payment at a particular point. Interest restrictions change the tax consequences of financing through debt. Hybrid mismatch rules restrict gains that depend on inconsistent classifications across jurisdictions. The minimum tax reduces some incentives associated with low effective taxation for groups within its scope. The effect of the combination depends on the remaining opportunities, the cost of adjustment and the availability of substitutes across countries.

These reforms pursue base protection while preserving legitimate cross-border investment and financing. The design problem is particularly demanding because aspects of corporate structures that facilitate lawful financing may resemble those used in aggressive profit allocation. Entity characteristics and flows can inform risk assessment, but the actual treatment depends on the governing conditions and facts. Financial intermediation, tax avoidance and illegal tax evasion are separate categories of activity and require distinct evidence.

Evaluating Reform Effects and Causal Identification

The first large fall in the selected low-tax-jurisdiction income series occurred between 2019 and 2020, before the conditional interest and royalty withholding tax took effect in 2021. The date ordering precludes treating that fall as an unqualified mechanical effect of tax collection under the new rule. Anticipatory restructuring is one plausible mechanism; contemporaneous changes in global economic activity and other tax reforms are additional candidates.

Observed outcomes establish the historical association between a changing institutional environment and the reconfiguration of Dutch financial holdings. A causal evaluation would require measuring what would have occurred in the absence of the policy. Several empirical strategies merit consideration. With detailed data one could compare affected payment categories and destinations with similar but unaffected ones, examining pre-reform trends. Firm-level information could distinguish entities that disappeared from the population from those that altered the classification or geographic distribution of income. A network analysis could trace whether a decline in Dutch positions coincided with a corresponding increase elsewhere.

Each proposed strategy has limitations. Destination-level classification is itself related to tax policy and can change during the observation period. Firms can anticipate legislation before the formal treatment date. Cross-country comparisons are complicated by the international scope of the reforms. An untreated jurisdiction may in fact receive diverted activity from the treated economy, violating the assumption that comparison units are unaffected. A panel of firms may suffer from selection if firms exiting the population disappear from the data.

A policy impact study should therefore specify its estimand before selecting a method. The effect on income directly paid from Dutch entities to the targeted territories is one estimand. The effect on the worldwide tax liability of an international group is another. The effect on public revenue in each source country is yet another. These quantities can move differently, and the first is much more readily observed than the others.

The evidence supports a qualified finding of major Dutch-sector and targeted-payment transformation. It does not establish the independent effect of each law or the worldwide welfare consequences of all reorganizations. That conclusion is substantive: it identifies exactly which policy questions aggregate Dutch data can answer, and what additional evidence is needed for decisions about international tax cooperation.

Developing Economies and International Revenue Allocation

The conduit role has consequences for countries where multinational enterprises operate. These countries may rely heavily on taxes associated with business profits and cross-border income payments. A reduction in withholding associated with treaty access can benefit investment under some conditions while also affecting source-country revenues. The outcome depends on whether it changes productive activity, reduces double taxation or mainly changes the legal geography of payments.

The Dutch Policy and Operations Evaluation Department (IOB), in its tax system support evaluation of policies in 2012–2020, identifies potential revenue risks for developing countries and recommends considering how the treaty network could affect withholding tax receipts before new treaties are concluded. It also highlights the importance of administrative capacity and the country-specific distribution of priorities in international tax cooperation.

An appropriate framework for interpreting these findings distinguishes the private cost of capital from the host government's fiscal resources. Reducing excessive taxation of a genuine new investment can encourage projects with positive social returns. By comparison, reducing tax on an existing return through a change in ownership may alter public receipts with little change in productive activity. The two outcomes call for different economic assessments even when the initial observed transaction appears in the same FDI category.

Public revenue can support infrastructure, public health, education and the administration of property rights. Its value depends on the state's use of funds and the distribution of the burden across alternative tax instruments. A fiscal loss in a capital-importing country cannot be assessed only by comparison with the fee revenue generated in the intermediate jurisdiction. The relevant comparison involves benefits and costs accruing to different populations.

International coordination has to account for differences in information systems and enforcement capacity. Formal treaty protections may require evidence about the ownership and activity of entities across multiple territories. If small administrations face greater costs obtaining and evaluating that information, identical legal rules can have different practical effectiveness. A meaningful evaluation of Dutch anti-avoidance reform therefore needs both Dutch data and information on changes in actual foreign tax receipts and productive investment.

Household Offshore Wealth and the Distribution of Tax Evasion

Corporate conduit statistics do not measure the wealth that Dutch households hold abroad. Households may be ultimate shareholders of international corporations, direct account holders, owners of investment funds or beneficiaries of other arrangements; each position has its own statistical and legal treatment. A gross corporate financial asset cannot be attributed to resident household wealth without tracing ownership and accounting for liabilities.

Leenders, Lejour, Rabaté and van ’t Riet (2023), in Offshore Tax Evasion and Wealth Inequality, exploit datasets on more than 29,000 taxpayers identified as evading taxes. Their evidence places evasion disproportionately among wealthy households. The research provides a different lens on offshore finance from the corporate conduit studies because its population is Dutch taxpayers with documented or identified tax noncompliance rather than financial holding companies.

Observed evasion data are necessarily conditioned by detection, disclosure and enforcement. Tax amnesties and administrative investigations can reveal assets missing from conventional surveys or tax records, yet they may have different coverage across wealth groups and across types of financial arrangements. The proportion of households appearing in such data cannot simply be equated with the true population frequency of undetected evasion. Estimates correcting recorded wealth require assumptions and independent information about unobserved holdings.

The distributional relevance arises from the unequal opportunities to hold and administer wealth across borders, but wealth alone does not establish a person's intent or compliance status. Lawful ownership of foreign assets is part of international financial diversification. The empirical question is whether assets and income are disclosed and taxed according to the applicable obligations.

The household and corporate dimensions intersect through ultimate ownership and financial intermediaries, but the available data reviewed here cannot produce a unified micro-level allocation of Dutch household holdings across the corporate conduit network. Their separation protects the validity of both literatures and prevents a false inference that the decline in corporate holdings measures the success of household tax compliance.

Professional Services, Integrity, and Financial Stability

The operation of complex corporate groups requires services that go beyond incorporation. Entities must maintain governance arrangements, accounts, records and, where applicable, regulatory compliance. International ownership creates demand for expertise in multiple systems at once. The density of specialized banks, accountants, law firms and trust service providers can sustain a country's relevance even when tax advantages diminish.

DNB's trust office supervision identifies governance, integrity, enforcement and market access among its supervisory domains. This institutional setting matters because service providers form points of contact between financial claims and national oversight. Effective oversight depends on the ability to understand customers and the economic functions performed by the entities they administer. The existence of a supervisory framework supplies neither evidence that all supervised entities comply nor evidence of misconduct by every supervised entity.

Systemic risk is a separate question from the extent of tax-base diversion. A holding company with predominantly equity financing and matched intragroup claims can have a very large balance sheet without performing the same maturity transformation as a deposit-taking bank. Another vehicle may depend on external creditor funding or enforceable guarantees and transmit losses to operating subsidiaries. Aggregate assets alone cannot resolve exposure to market, liquidity, credit or operational risk.

Relevant assessment would require consolidation at the group level and disaggregation by instrument, maturity, currency, counterparty and guarantee structure. Gross positions matter because even offsetting claims can create contractual obligations and legal dependencies. Net positions matter because many exposures disappear upon consolidation within a group. Cross-border insolvency or disputes may still affect how and when claims can be realized. A purely netted view can understate gross contractual interconnectedness.

The domestic consequences of a sudden restructuring should also be separated from the consequences abroad. Closure of financial entities may reduce demand for administrative services without eliminating an equivalent amount of Dutch productive capital. However, operating companies or financial institutions with exposures to those entities can be affected by the settlement and transfer of claims. The magnitude of that risk must be established from detailed exposures; the data summarized in this study establish scale, not a quantified stress-test loss.

The Netherlands in the Changing International System

The Dutch case places a developed and productive economy inside the global offshore system. Its financial holdings operate within a larger national institutional environment. This distinguishes the country from an economy whose public revenue and employment depend overwhelmingly on a narrow segment of cross-border corporate administration. It also means that the adjustment of the Dutch conduit function can involve movement toward more substantive financial activity rather than a simple disappearance of the financial sector.

International competition takes place across several dimensions: legal predictability, access to counterparties, taxation, quality of administration, skill availability and regulatory credibility. When one dimension changes, competing jurisdictions can respond. A reduction in Dutch intermediation might correspond to a reduction in the worldwide use of a type of arrangement or to substitution toward a different jurisdiction, a different instrument or a more complex ownership relationship. A national indicator cannot discriminate among these system-wide outcomes without evidence about counterparties.

The policy goal of limiting arrangements driven by artificial tax advantages can coexist with preserving the efficiency of cross-border investment. The difficulty lies in identifying circumstances in which a financial structure performs valuable coordination or risk-bearing functions and circumstances in which its benefits are mainly redistribution of public revenue. The same institutional forms can be used for distinct purposes; the evaluation belongs at the level of actual relationships and their consequences.

The strongest internationally coordinated monitoring would retain four types of data in parallel: resident-entity statistics, ultimate-ownership statistics, income and tax-payment data, and indicators of real production and employment. The OECD (2025) reporting developments described in the FDI benchmark are directionally consistent with this need. The statistical quantities should share well-documented identifiers where confidentiality and national law allow, while preserving the distinctions between populations.

At a broader level, the Netherlands shows why the global offshore system must be analyzed through the network created by large economies and smaller specialized jurisdictions together. The original network evidence places advanced economies among the key intermediate locations. The modern statistical evidence shows that changing tax conditions can reduce some of their intermediary functions without immediately eliminating their central position in corporate capital relationships.

Evidence Synthesis and Authorial Findings

The evidence supports five connected analytical findings. First, the Dutch conduit function is economically relational. Its measurable importance arises from the concentration of intermediate ownership and financing relationships, which can be established without equating national territory with a tax classification. Second, the financial scale of an intermediary economy is different from its productive contribution. The difference explains the coexistence of multitrillion-euro positions with comparatively modest direct employment and fiscal receipts in the identified holding-company population.

Third, evidence of a tax-influenced ownership geography is strongest when observed structures are linked to specific fiscal incentives and compared against credible alternatives. The Dutch microdata literature on treaties and income flows supports this interpretation while leaving material differences across income categories. Fourth, the Dutch reform period demonstrates that entity counts, balance-sheet ratios and specific cross-border income flows have changed at different rates. Their joint movement provides a stronger description of restructuring than any single indicator.

Fifth, the domestic and international incidence of the sector is distributed asymmetrically. The gains experienced by corporate groups or service providers do not reveal the overall benefits or costs to taxpayers and treasuries across the jurisdictions involved. Measuring the net effect requires explicit counterfactuals, including the possibility that investment and financing would have occurred under a different arrangement.

These findings are an interpretive synthesis of source evidence and transparent accounting reasoning, not a newly named theory. The original contribution is the joint reading of administrative entity counts, stock-to-GDP ratios, chains of corporate ownership and payment-specific tax evidence as observations of distinct economic processes. It explains why reducing one class of payments can be a policy achievement while leaving large recorded financial positions. It also explains why the raw financial scale cannot furnish an estimate of public harm or economic benefit by itself.

Source limitations are integral to that conclusion. Some journal arguments were evaluated through detailed abstracts and bibliographic records rather than complete paywall-restricted texts. Government summaries establish the institutional measures and official interpretation of results rather than the legal incidence of every arrangement. No full microdataset was available for this research. The covered time series are dated and may be revised. A comprehensive empirical follow-up would acquire consistent time-series extracts, establish a firm-level panel, include changes in the definitional perimeter and estimate separate effects on financial income, entity location and public revenue. Those tasks describe additional research rather than results claimed here.

Conclusion

The Netherlands has developed an economically significant role in the international intermediation of corporate capital. The institutional history and ownership-network literature explain the emergence of that role; direct investment accounting clarifies its measurement; research using Dutch firm and payment data identifies particular tax incentives; and recent official statistics document major changes in the size and composition of the conduit function.

The decline in the number of multinational financial holdings and their assets relative to GDP, together with the reduced flow of selected income to low-tax jurisdictions, provides evidence of substantial restructuring. The reported balance sheet of €4,732 billion for year-end 2025 establishes the continuing importance of the sector. Its contraction and persistence belong to the same historical process of adaptation.

For economic policy, the central task is to measure the effects of intermediation at the level where they occur. Productive capacity and employment describe local economic activity. Gross and consolidated positions describe financial relationships and exposure. Payment-specific data and counterfactual analysis describe potential tax effects. Ultimate ownership and household evidence help establish the distribution of gains and obligations. The Netherlands cannot be adequately understood by replacing these measures with a single national label or aggregate financial statistic.

The case therefore contributes to Offshore Economics by demonstrating the importance of distinguishing recorded investment from production and the private advantages of international corporate arrangements from their public incidence. Sound policy combines protection of fiscal bases with economically useful international financing, evaluated against empirical evidence about real investment, reporting quality and the allocation of tax burdens across jurisdictions.

References