RESEARCH DOSSIER

Round-Tripping Foreign Direct Investment

Abstract. Round-tripping foreign direct investment (FDI) arises when domestically originating capital returns through a nonresident direct investment relationship. This monograph investigates the economic meaning, accounting treatment, ultimate-investor attribution, empirical identification, institutional incentives, and policy consequences of such investment. It differentiates domestic ultimate control from the original source of funding and from productive activity. The study synthesizes international statistical standards, country evidence and published research, including historical work on China and Russia, official Ukrainian transaction estimates, and United States and European ultimate-investor data. Illustrative accounting exercises establish the implications of different attribution rules, signed flows and financing arrangements. The evidence supports a complementary use of residence-based accounts, ultimate-investor statistics, financing information, and operating indicators. It does not establish a globally comparable round-tripping total or transaction-level proof of unlawful behavior.

Keywords: Round-tripping FDI; foreign direct investment; ultimate investing economy; ultimate controlling parent; international investment statistics; offshore economics; pass-through capital; domestic capital formation; investment incentives; ownership networks

Introduction

Foreign direct investment is an indispensable category of external-sector statistics and an imperfect measure of internationally supplied productive capital. The difference becomes conspicuous when a resident investor owns an enterprise outside the home economy and that enterprise invests back into the home economy. The transactions cross national accounting boundaries, while important elements of ownership and economic decision-making remain domestic. This research examines the resulting phenomenon of round-tripping foreign direct investment (FDI), with particular attention to the economic meaning of domestic ultimate ownership, the financial architecture connecting economies, the limitations of existing data, and the consequences for investment policy.

Round-tripping matters because the classification of capital as foreign frequently carries economic interpretations that go beyond the statistical definition. A country reporting substantial inward FDI may be considered successful in attracting international capital, technology, managerial expertise, access to markets, or confidence from outside investors. Each interpretation needs its own evidence. Recorded foreign investment alone establishes a cross-border financial relationship fitting the relevant statistical criteria. It does not establish the foreign origin of ultimate ownership, fresh external saving, or the additional productive activity attributable to the transaction.

The research question is how to identify and interpret the foreign direct investment that returns to the domestic economy through a nonresident corporate relationship. Three subordinate questions follow. How should compilers distinguish a financial circuit from a change in the geography of ultimate control? Under what circumstances can a transaction increase domestic investment while leaving ultimate ownership domestic? What implications follow for economic measurement, public revenue, capital allocation, and investment incentives?

The research uses international statistical standards, published official releases, historical and empirical literature, and explicit accounting illustrations. The quantitative examples constructed in this study are hypothetical. Published figures retain the original reporting population, date, instrument coverage, valuation, and status as estimates where applicable. The study contributes an integrated interpretation of these materials rather than a new named theory or independent estimate of world round-tripping.

Concepts and Scientific Boundaries

The economic object of round-tripping is the return of domestically originating capital through a foreign direct investment relationship. In operational statistics, the identification of resident ultimate investors offers an observable indicator of this process. The two descriptions are related, but an observation of domestic ultimate control cannot invariably trace the original money. Definitions should therefore make the distinction explicit.

The OECD (2025) round-tripping annex describes capital leaving an economy and subsequently returning as direct investment. Geography is integral to this account. The original investor and the receiving direct investment enterprise are residents of the same economy, while at least one nonresident enterprise participates in the cross-border financial relationship. There may be several intermediary economies, and the initiating and receiving domestic companies may belong to a single group or be different legal entities.

An immediate foreign investor is the nonresident counterpart directly associated with an inward position. An ultimate investor is identified through ownership and control relationships under a specified attribution method. A funding source refers to the party furnishing financial resources. An operating location refers to where production, employment, or services occur. Residence refers to the statistical economy of an institutional unit; citizenship is a separate characteristic.

The scientific boundary of this subject also separates five phenomena that often occur together. Pass-through investment describes the movement of investment through an intermediary economy. Phantom FDI describes investment recorded in entities without substantial local operations. A special purpose entity (SPE) is an organizational and statistical category. Capital flight concerns the departure of capital, for which definitions and measurement approaches vary. Profit shifting concerns the geographic distribution of taxable income. Round-tripping concerns the return relationship and the domestic origin or ultimate ownership of the investment.

A conduit serving foreign owners from a third economy may have large pass-through positions and zero identified domestic round-tripping in the host. Conversely, a commercially active foreign affiliate can invest in a domestic enterprise owned by the same domestic group. Real economic activity and round-tripping can coexist. The category crosses the distinction between operating and holding entities and does not by itself characterize compliance with criminal, tax, or investment law.

This is the particular scientific responsibility of Offshore Economics within the wider Global Offshore System: trace capital relationships and evaluate their consequences. General legal doctrines, institutional classifications of foreign affiliates, and specialized analyses of tax planning and financial secrecy belong to their respective research objects.

Direct Investment and the Architecture of Control

FDI statistics cover a lasting investment relationship under an operational ownership criterion. The OECD's direct investment framework (2025) distinguishes substantial influence, associated with the 10 percent voting-power threshold, from majority control. Associates and controlled subsidiaries occupy different positions in a corporate group, even though both can participate in direct investment. A legal share, a voting share, a cash-flow entitlement, and a right to appoint directors may differ.

The implications for round-tripping begin with a familiar corporate structure. Let H be a home-economy enterprise, F a foreign affiliate, and D a second home-economy enterprise. H controls F and F controls D. An equity claim of H upon F is external from H's viewpoint; F's claim upon D is external from D's viewpoint. At the level of ultimate control the chain places D within a home-controlled multinational group. The receiving enterprise is thus domestically controlled and foreign invested at the same time.

Now replace F with an enterprise whose own ultimate parent is resident in a third economy. All other financial observations could be identical from the receiving enterprise's immediate viewpoint. The ultimate geography has changed. Finally, imagine that F owns only a minority stake in D and that an unrelated domestic company controls D. A statistic assigning all D's inward FDI to D's controlling parent could identify domestic ultimate ownership even though F is foreign-controlled. This illustrates why the unit selected for attribution matters.

A financial claim is an enforceable or recognized economic relationship, subject to the relevant instrument and valuation rules. Control is an organizational relationship. It is possible to control a company with less than full participation in its returns, and it is possible to supply financing without controlling management. When control, financial risk, cash-flow rights, and location are used as interchangeable indicators, the interpretation of cross-border capital becomes unstable.

Round-tripping is especially sensitive to this conceptual error because the same group can be organized across several legal units while operating as one economic decision-making organization. The national accounts cannot simply replace institutional units with corporate groups: the distribution of assets, obligations, income, and residence would then be lost. A control-based supplement recovers another analytical dimension without abolishing the original accounts.

Accounting for the Return Circuit

Consider a hypothetical economy H with a resident company A, a second resident company C, and a controlled foreign subsidiary B in economy F. A contributes 100 monetary units of equity to B and B contributes 100 units of equity to C. C builds productive capacity using the funds. The example assumes both positions satisfy direct investment criteria, no other cross-border transactions, no valuation changes, and no pre-existing holdings.

Reporting economy Direct investment assets Direct investment liabilities
H 100 100
F 100 100

The table records the new direct investment claims and obligations rather than all changes in the balance of payments. H's net direct investment asset position changes by zero. Its gross outward and inward positions each increase by 100. F also records equal new direct investment assets and liabilities. The sum of the two economies' inward positions is 200, although the group has transferred the same initial 100 through two ownership links. These are two valid financial relationships and one initial financing amount. Treating their sum as independent additions to productive capital would double-count the initial financing in this stylized example.

Suppose C acquires newly produced machinery for 100. Domestic capital formation may rise by 100 under the stipulated assumptions. Suppose instead C purchases already existing shares in a domestic operating company for the same sum. The transaction changes ownership claims and may create no immediate addition to fixed capital. The round-trip structure is identical in both cases, but the real outcomes differ. Whether domestic saving finances the project in the ultimate macroeconomic sense depends on the rest of the accounts and the financing circumstances.

A third case changes the financial source. A contributes 40 to B, and B borrows 60 from an unrelated foreign lender before providing 100 to C. C's immediate investor and the ultimate controlling parent remain as in the first example. Yet the project incorporates 60 of external financing. Under a control-based inward attribution, all 100 may appear domestic-origin. The group's funding mix and its control-based direct investment classification therefore answer different questions. The 60 borrowing also creates a separate external claim whose instrument and institutional counterpart must be recorded appropriately.

Reverse investments and positions among fellow enterprises complicate the conversion of assets and liabilities into inward and outward directional statistics. The preceding examples deliberately exclude those complications. Actual reconciliation must follow the selected statistical presentation, especially when two resident affiliates interact with the same foreign group. The distinction between a domestic group's consolidated financial exposure and the economy's external accounts is essential: cross-border intragroup claims can disappear under group consolidation while remaining valid entries in residence-based accounts.

A return circuit can arise from multiple transactions across several dates. A position at the end of a year measures the outstanding financial relationship, not the complete history of those transactions. It reflects cumulative transactions, exchange-rate changes, price changes and other volume adjustments. The researcher must not infer a recent circular transfer solely from a year-end position.

Ultimate Investor Attribution

The geographical disaggregation chapter of OECD BD5 (2025) makes an important distinction between two methods. Under winner-takes-all (WTA), a direct investment enterprise is assigned a single ultimate controlling parent. Under proportional ownership (PO), attribution is based on the ultimate controllers of its immediate direct investors, permitting several ultimate investing economies. The standard explicitly warns that WTA may overstate round-tripping where a resident-controlled enterprise has foreign minority investors.

The consequences can be demonstrated without empirical assumptions. Consider a company D in economy H whose domestic parent owns 60 percent. One foreign direct investor, entirely foreign-controlled, owns 40 percent. Suppose its inward equity position is valued at 40. Under a WTA attribution based on D's domestic controlling parent, the inward position is assigned to H. Under PO, the 40 is assigned to the foreign investor's ultimate economy. In this example there is no identified returning domestic capital in the 40. A difference of 40 between the methods is produced solely by attribution.

A second example makes the difference less extreme. A domestic shareholder controls D with 60 percent of its voting rights. Foreign investors B and C each hold 20 percent. B is controlled by a domestic owner and C by a foreign owner. For an inward equity position of 40, with equal-valued stakes, WTA allocates all 40 to H. PO allocates 20 to H and 20 to the foreign economy. Both methods can be consistently described, but their domestic entries measure different constructs. The latter is closer to tracing who ultimately controls each immediate investor.

Neither exercise traces the financing of B or C dollar by dollar. Debt, retained profits, foreign minority investors upstream, and financing secured from independent lenders can alter who supplies or bears financial exposure. A declaration that a position is ultimately controlled domestically is a statement about the relevant control chain.

A further difficulty concerns changes in residence and ownership without new investment. A domestic group acquires an existing foreign company which already owns an investment in the home economy. Ultimate attribution can change from foreign to domestic without a new payment to the home enterprise at that moment. Similarly, corporate redomiciliation or a sale of an intermediate company may change geographical assignments. Such events should be separately identified wherever the statistical framework permits.

For researchers, the central requirement is methodological transparency: identify the attribution rule, the starting unit of ownership tracing, how dispersed ownership is handled, how the data treat unresolved chains, and whether the output records positions or transactions. Comparisons between countries require these details. A common label may conceal different statistical procedures.

International Statistical Standards and Data Sources

External-sector accounts connect the financial account of the balance of payments, investment income, and the international investment position. The IMF BPM7 release (2025) marked the publication of updated international macroeconomic statistical standards on March 20, 2025. Historical country observations frequently remain compiled under earlier standards. A researcher must record the version employed by each series and separate changes in classification from changes in economic behavior.

The OECD (2025) standard and supplementary series specify alternative presentations of direct investment designed for distinct analytical purposes. Immediate-partner positions sustain bilateral accounting; ultimate-investor positions inform ownership geography. Supplementary series extend these questions toward the ultimate host and parent residence. Implementation and reporting completeness matter as much as the formal availability of the concepts.

The IMF Direct Investment Positions dataset (DIP, formerly CDIS) provides bilateral direct investment positions according to immediate counterpart. It is a valuable infrastructure for investigating the geography of claims. A large symmetric pair of positions can flag potential intermediation, but aggregate bilateral balance sheets alone do not identify which group, if any, completes a return circuit.

The following workflow is appropriate for scholarly comparison. First define the desired population: enterprises with nonresident direct investment relationships in reporting economy H. Then choose a reference period, presentation, currency, and valuation. Establish whether positions include both equity and intercompany debt. Record the treatment of SPEs and the statistical classification of investment funds. Obtain corresponding immediate and ultimate distributions, reconcile their totals when the coverage is identical, and carry unidentified investors explicitly. For empirical use, retain the original publication date and release vintage, rather than combining revisions without documentation.

A country's ultimate ownership dataset should be interpreted with its investment activity data. Employment, capital expenditure, turnover, value added, research activity, and imports and exports help establish the operational consequences of a financial position. Even a fully reconciled external position does not provide these indicators automatically. Their statistical units may differ, and legal-entity linkages are required before attributing activity to a specific financial group.

The absence of ultimate data is an identification problem, not a zero observation. Disclosure thresholds, privacy rules, insufficient international identifiers, and compilation capacity restrict the observable population. Missing observations must remain missing, especially in rankings of economies by the magnitude of domestic-origin foreign investment.

Financial Intermediation and Phantom FDI

The structure of multinational investment frequently includes entities whose financial activity exceeds the resources dedicated to local operations. Such entities can centralize ownership, borrow within a group, or hold equity in affiliates resident elsewhere. Both operational companies and financial holding companies participate in the intermediation network.

Damgaard, Elkjaer, and Johannesen (2019), in What Is Real and What Is Not, estimate that investment into entities with negligible local substance could constitute nearly 40 percent of global FDI. This is a model-based estimate of phantom investment in their research population. It is not a round-tripping percentage, a measure of global criminal wealth, or an observation of the share of all financial capital that is unproductive. The study highlights the analytical cost of assigning productive investment geography directly from immediate positions.

Borga and Caliandro (2018) investigate eliminating pass-through from FDI statistics to better identify economic and financial linkages. This research direction is consistent with retaining the original cross-border accounts while supplementing them with statistics that assign financial relationships to the economic locations of interest.

The distinction has implications for empirical design. A study classifying a large intermediary position as phantom investment investigates the relationship between financial claims and local activity. A study classifying inward claims according to the ultimate owner investigates their control geography. A study identifying round trips through ownership graphs investigates a particular closed geographic relationship. These classifications intersect but are not identical partitions of a single dimension.

For example, a domestic corporation can use a foreign regional headquarters with substantial staff and activity to organize investments in its home market. The return circuit may meet a round-tripping indicator while the foreign unit is a substantial enterprise. A conduit company connecting two foreign jurisdictions may have negligible local employment but perform no round trip to its ultimate investor's economy. A third case can satisfy both descriptions. Published phantom-FDI shares cannot legitimately be added to or subtracted from published round-tripping shares without knowing the overlap, denominator and valuation.

For an intermediary economy, financial intermediation is itself an economic function. Legal, administrative, accounting and financial services involve real labor and capital. Their scale and quality should be measured separately from the underlying assets managed or held through clients' corporate structures. This provides a better account of the contribution and risks of intermediary jurisdictions than treating the entire stock of recorded FDI as a measure of their real productive capital.

Measuring Round-Tripping in Aggregates

Let I(H) denote the total inward direct investment position of reporting economy H under a specified directional presentation. Let I(H,H) denote the portion attributed to H as the ultimate investing economy under a specified method. An elementary domestic-ultimate position share is R(H) = I(H,H) / I(H), provided that the denominator is positive and that the numerator belongs to the same measurement population. This ratio is a descriptive statistic, not a general structural parameter.

Suppose a dataset reports inward positions of 1,000, of which 180 are assigned to domestic ultimate investors, 720 to foreign ultimate investors, and 100 remain unassigned. The reported domestic-ultimate share is 18 percent of the total. The share among identified positions is 20 percent. Under the additional assumption that all unassigned components are nonnegative and represent unknown ownership categories, a simple sensitivity interval for the domestic share is 18 to 28 percent. It is a classification interval, not a sampling confidence interval. With negative net positions or nonadditive published categories, the interval would require a different treatment.

The corresponding flow ratio is not generally comparable to the position share. Let F(H) represent net inward direct investment during a period and F(H,H) the domestic-ultimate portion of comparable transactions. The ratio F(H,H) / F(H) can be unstable as the denominator approaches zero. If domestic-origin net inflows equal 50 and foreign-origin net outflows equal 40, the aggregate net inflow is 10 and the ratio is 500 percent. All figures are hypothetical. The ratio expresses offsets and does not imply that five times the recorded investment consists of returning cash in a gross sense.

The remedy is to present the signed quantities as well as ratios, and to differentiate gross acquisitions from disposals when data support that distinction. In severe downturns a negative share may convey little useful information; the separate amounts and economic context become more informative.

An inward position by ultimate investing economy supplies an important answer about ownership. An outward position by ultimate host economy supplies a complementary answer about investment destination. They should not be automatically matched dollar for dollar. Existing positions can differ because of valuation methods, reinvested earnings, leverage, upstream ownership, timing, or intercompany netting. Aggregate equality in the simplified accounting illustration depends on assumptions that do not characterize a typical multinational group.

The same cautions apply to deflation and normalization. An absolute nominal dollar amount records financial scale at a vintage. A ratio to GDP compares that scale with domestic output. A ratio to inward positions compares it with a different financial universe. All three can be useful; none supplies a measure of additional productive investment without further evidence.

Ownership Graphs and Statistical Identification

The identification of individual returning relationships requires linking domestic and foreign legal entities into an ownership graph. Each node represents an identifiable institutional unit, and each edge has a type, direction, value or ownership weight, and reference date. A graph can record voting rights and funding claims as separate edge types. The geographic path of control and that of finance are then evaluated without merging the two into a single network.

Where legal entities are identified and ownership information is complete, an ownership chain can reveal whether the controlling investor resides in the receiving economy. The observed relationship may be studied alongside financial transactions, providing a stronger account of capital origin than ownership alone. In practice, equivalent entity identifiers may be absent across registries, names may change, mergers may create discontinuities, and legal documentation may report control relationships differently from accounting surveys.

Casella (2019) proposes a probabilistic approach using absorbing Markov chains to estimate ultimate-investor distributions across more than 100 recipient economies. This is a useful strategy when direct tracing of every ownership chain is infeasible. A transition probability is an estimated relationship in a statistical construction. It does not establish that a named company transferred a particular sum along an identified route.

Network inference should document the assumptions used to convert available stocks into estimated ultimate distributions. Different matrices can reproduce similar observations, particularly when intermediate locations have extensive cross-holdings. Validation therefore requires comparison with independently observed ultimate-investor information for reporting economies and sensitivity testing under alternative assumptions.

A minimum audit retains the original source of every ownership link, its date, the rule for following or stopping a chain, and the handling of unresolved circular ownership. A researcher should distinguish unknown ultimate owners from domestic ultimate owners. The problem becomes especially difficult if the analyst encounters a network of companies controlling one another without a uniquely identifiable external controller under the selected operational rule.

This produces a hierarchy of research confidence. Documented ownership and transactions are stronger evidence for particular return circuits than an ownership-based classification alone. Official aggregate ultimate-investor data are useful for population estimates. Model-based estimates can broaden coverage but require explicit uncertainty. Immediate bilateral positions and offshore-location proxies are useful clues. They are not sufficient evidence to identify specific ultimate residents or their conduct.

Economic Incentives and Institutional Conditions

Round-tripping changes the corporate and jurisdictional organization of investment. A domestic owner can have commercial or institutional reasons to hold an enterprise through a foreign affiliate. The economic task is to explain these reasons and distinguish observed behavior from hypothetical mechanisms.

A preferential regime for foreign investment creates a difference between the value of a domestic investment held directly and that of an economically similar investment held through a foreign enterprise. The difference may concern taxation, administrative procedures, access to financial services or other aspects of the investment environment. A differential benefit can induce firms to change their legal ownership structures even when the underlying project and ultimate controlling investor remain unchanged. The costs of reorganizing and maintaining the corporate relationship determine whether the change is commercially worthwhile.

A second mechanism is access to capital. A company may require specialist financial intermediaries, an international shareholder base, a regional treasury center, or a funding market unavailable domestically. The resulting organizational structure may change the timing, currency, distribution, and cost of financing. In this case, identifying domestic ultimate control is compatible with access to genuine external resources.

A third mechanism concerns property rights and the institutional environment. Corporate groups may seek a stable legal setting and predictable contract enforcement. The resulting ownership structures can serve legitimate commercial interests. Under other circumstances, opaque ownership may be associated with evasion or illicit activity. Classification of specific conduct must follow verified facts and applicable law; a financial pattern alone is insufficient.

The OECD's 2025 special-cases analysis also identifies mergers, regional headquarters, and commercial coordination within supply chains as possible circumstances producing round-tripping. This matters for causal analysis: if a returning relationship emerges through the acquisition of an already multinational enterprise, it cannot automatically be explained as an effort to acquire a preference specifically reserved for foreign investors.

These mechanisms can operate together. One company might benefit from both a wider borrowing market and improved contractual reliability. Another may change its structure in response to a fiscal reform. The researcher needs observable implications, such as a policy discontinuity, differences across eligible firms, changes in independently observed financing arrangements, or shifts in ownership structure after identifiable corporate events.

The choice of jurisdiction is consequently endogenous to business characteristics and institutional conditions. Economies with large volumes of returning investment may have internationally active enterprises, distinctive investment incentives, limitations in domestic financial markets, or combinations of these features. Their experience cannot be inferred from the volume of round-tripping alone.

Historical Evidence from China

Chinese FDI provides a major historical application of the distinction between immediate foreign capital and returning domestic investment. In his revised round-tripping study published by the Asian Development Bank Institute in 2005, Xiao examines the scale and measurement of the phenomenon and discusses the limitations of earlier estimates. The publisher describes an earlier estimate of as much as one-quarter of total inward flows, while noting the absence of a transparent method for that earlier figure. This is evidence of a historically significant measurement debate, not a verified current share for China.

The historical setting involved changing incentives toward inward investment, cross-border financial administration, and close links to Hong Kong. These conditions generated hypotheses about why domestic capital would be held through nonresident companies. A measure based on investment from Hong Kong would, however, include genuine Hong Kong investors and investors ultimately controlled elsewhere. It could also miss transactions passing through other jurisdictions.

The econometric problem is especially acute when researchers attempt to estimate domestic-origin investment by comparing mirror flows. The discrepancy between a host's inward report and the counterpart's outward report can have different causes: geographical allocation, investor coverage, exchange-rate conversion, timing, reporting error or the position of intermediate companies. A residual estimate requires assumptions about the part of the discrepancy attributable to each cause.

The historical debate demonstrates the importance of using ownership and investment records together. It also shows why an estimate developed for one phase of Chinese investment reforms cannot be projected into a later period without accounting for changes in industrial organization, international finance, taxation and statistical coverage.

FDI and export-led industrialization may be linked through production networks, management and technology transfers. A returning domestic investment placed in a new factory can contribute to employment and productive capacity even if ultimate control is domestic. Conversely, foreign-controlled ownership of an existing firm need not generate an immediate addition to physical capital. Evaluating the role of investment in China's economic development requires separate observation of these outcomes.

The China case is therefore primarily a warning against attributing all recorded foreign investment to outside investors, and against dismissing all returning investment as devoid of economic significance. A careful historical account differentiates statistical classification from the contribution of the investment to production.

Russian Regional Evidence and Identification Limits

Ledyaeva, Karhunen, and Whalley (2013) investigate Russian round-trip investment in relation to offshore jurisdictions and corruption. The later journal article by Ledyaeva, Karhunen, Kosonen, and Whalley (2015), Capital Round-Tripping and Corruption, examines firm-level regional distributions using offshore-origin investment as a proxy for much of the returning investment. Their empirical work relates the geography of offshore-linked investment to regional institutional conditions.

The underlying methodological issue is that the observed immediate origin of the investor is an imperfect indicator of its ultimate owner. Misclassification can affect estimates in both directions. Offshore-linked firms with genuine foreign ultimate owners may enter the proxy category, while domestic-origin investment through other immediate economies may be missed. If misclassification varies across regions and sectors, the relationship between the proxy and institutional indicators need not equal the relationship for actual domestic-origin investment.

Regional corruption may influence the choice of ownership structure, but it can also be associated with other characteristics that influence investment: natural-resource composition, regional wealth, local infrastructure, market size, political networks, or the organization of privatized enterprises. An empirical association motivates investigation of mechanisms; proof of the source or unlawful use of particular funds requires transaction-specific evidence and applicable legal findings.

A further limitation concerns external validity. Historical regional results cannot be assumed to describe subsequent periods marked by new geopolitical, investment, regulatory, or reporting regimes. The appropriate use of the studies is to demonstrate an empirical approach and its interpretative boundaries.

The evidence also establishes a general principle for research design in transition economies: the institutional conditions that influence capital leaving an economy can also influence investment flowing back. Investment in a home market may remain profitable even when the investor prefers an ownership or financial arrangement in another jurisdiction. The economic explanation should account for both the original outward relationship and the subsequent domestic investment decision rather than treating the two as unrelated events.

A researcher seeking stronger identification could combine linked ownership data, independently measured financial flows, dated institutional measures, and transparent comparison groups. Confidential transaction evidence should remain appropriately protected. Public aggregate research does not require naming individual investors where doing so contributes nothing to scientific identification.

Ukraine: A Direct Official Measurement Case

Ukraine supplies unusually explicit official reporting of transactions categorized by the residence of the ultimate controlling parent. The National Bank of Ukraine (2026), in its 2010–2025 round-tripping estimates, reports USD 11.9 billion of cumulative returning investment, equivalent to 25.2 percent of USD 47.3 billion in FDI inflows over the period after excluding reinvested earnings. The OECD's 2026 infrastructure policy review corroborates the same percentage and denominator qualification.

The denominator is crucial. This is a cumulative transaction-based comparison with reinvested earnings excluded. It is not a share of year-end inward FDI positions, a share of all external financing, or an estimate of gross capital flight. The reported rounded totals produce approximately the reported percentage; the published percentage should be retained as the authoritative result because the underlying unrounded figures are unavailable in this study.

The NBU presentation reports negative net returning transactions during 2022, around USD 0.6 billion, and a subsequent recovery. In 2025 it identifies USD 603 million of returning transactions, equal to 38.3 percent of the defined inward inflow. The published presentation also notes debt reorganization into authorized capital within the broader 2025 inflow. That fact requires caution in associating the financial-account measure with new cash entering a business.

This case is particularly informative for the economics of adjustment. A domestic-ultimate investment position can exist alongside sharp reductions in current returning transactions. A group may reduce an exposure, dispose of assets, restructure debt, or postpone new projects when uncertainty rises. Negative net transactions may therefore coexist with large outstanding positions. The economic response is better described using both measures.

Events affecting Ukraine since 2014 and the full-scale Russian invasion in 2022 provide important context. They altered uncertainty, operating conditions and financial relationships. The official narrative connects those shocks with observed changes in returning investment. This monograph does not attempt to isolate their causal magnitudes or estimate counterfactual capital flows.

The sectoral evidence offers another useful qualification. NBU reports that during 2010–2013 a large fraction of returning investment was directed to real-sector enterprises. The economic classification of the recipient points toward businesses outside the financial sector in that reporting framework. It does not guarantee that each transaction directly financed new fixed assets.

The Ukrainian material demonstrates the analytical benefits of separately publishing domestic-ultimate transactions, sign conventions, institutional context, and instrument detail. It also underscores that a meaningful measure can be constructed at the national level without extrapolating it into a global share.

United States and European Evidence

Round-tripping is relevant to economies with mature corporate and financial markets as well as economies undergoing major institutional transition. Statistics published by the U.S. Bureau of Economic Analysis illustrate the difference between immediate and ultimate ownership. In its 2024 release, Direct Investment by Country and Industry, 2023, BEA reports inward FDI positions of approximately USD 5.39 trillion at historical cost at the end of 2023. The Netherlands led on the basis of foreign parent residence; Japan led on the ultimate-beneficial-owner basis. These are release-vintage observations rather than a claim about the latest available revised figures.

The terminology of ultimate beneficial owner in this BEA release refers to a statistical chain of majority ownership. It should not be assumed identical to beneficial ownership as defined for all legal and regulatory purposes. What matters for comparing the published geography is the operational rule applied to the foreign parent's ownership chain.

Eurostat's 2025 ultimate investing economies release covers 13 EU reporting economies for reference year 2024. Both geographical distributions sum to EUR 3,892 billion for this same reporting cohort. Luxembourg accounted for the largest immediate share, while the United States accounted for the largest ultimate share. The equivalent aggregate demonstrates reallocation of a common position universe. It does not measure a disappearance of cross-border claims. The cohort is not the entire European Union.

The OECD's 2023 Portugal investment study identifies Portugal itself among significant ultimate investors in the country. This provides another institutional example of why domestic ultimate ownership can be present within a cross-border investment universe. The study considers financial and institutional motives and the influence of attribution methods. No present-day Portugal round-tripping percentage is inferred here.

Together these cases demonstrate that changes in investor-country rankings may reflect the geographic architecture of international business rather than the physical relocation of production. An analyst studying competition, economic dependence or investment relationships should explicitly decide whether the relevant partner is the immediate creditor or equity holder, the controlling investor, or the ultimate location of operating activity.

The U.S. and European sources also caution against treating a single ultimate-investor concept as perfectly harmonized across statistical releases. Reporting coverage, direct investment definitions, valuation, reference dates, and ownership allocation practices determine whether cross-national comparisons are meaningful. A statistical term acquires comparable meaning only through its operational metadata.

India and the Limits of Cross-Country Comparison

The Reserve Bank of India's 2024 foreign liabilities and assets census documents a survey population encompassing entities with inward or outward direct investment. The coexistence of inward and outward positions is necessary for certain return-circuit structures, but it cannot establish any numerical share of round-tripping in the absence of suitable ownership attribution.

The methodological contrast with Ukraine is instructive. A census of cross-border assets and liabilities can be extensive and economically valuable without identifying the ultimate residence of investors for each inward relationship. National banking and investment statistics organized by immediate counterpart may document financial exposure accurately while leaving the ownership question unanswered.

One might observe large inward investment from a particular jurisdiction and large outward investment to that jurisdiction. The two measurements need not relate to the same firms. Inferring a circular transaction from aggregate symmetry creates an ecological inference problem. An unrelated partnership of outward and inward investment patterns could produce the same aggregate observations.

For India, as for jurisdictions without a verified ultimate-investor measure in the present evidence corpus, this research leaves the magnitude open. A comparative argument gains accuracy when cases with direct measurement are distinguished from cases offering evidence only about the broader external investment relationship. The selection of countries for interpretation should be based on available measurement, not on their reputation as offshore users or investment centers.

A consistent international comparison would request comparable definitions, valuations, ultimate-investor allocation rules, years, and coverage, with a separate description of data gaps. It would then compare like-for-like indicators rather than fill gaps with unsupported estimates.

Investment, Productivity, and Domestic Capital Formation

The real-economic implications of round-tripping depend on what the investment finances, who ultimately supplies resources, how risk is allocated, and whether the project would have occurred without the foreign corporate relationship. Domestic ultimate ownership is compatible with both a highly productive and an unproductive investment.

Consider a domestic company undertaking a capacity expansion. If an international corporate structure improves the terms of external borrowing, the investment may become viable. The company's project can increase labor demand and productivity even though its controller remains resident. The economic gain arises from the availability or terms of financing and from the project itself, not from changing the statistical label of the owner.

A second company changes its legal structure solely to qualify for a fiscal preference, while the underlying investment would have occurred anyway. In this hypothetical case domestic capital formation is unchanged. The policy transfers value to the recipient and introduces organizational cost. The fiscal preference can be economically costly when the additional investment it induces is small.

A third transaction involves the acquisition of an existing operating company. The acquisition may change governance, integration with supply chains, or access to markets. These effects can be economically substantial, although the transaction need not create fixed capital at the acquisition date. This distinguishes the transaction's financial classification from subsequent operating performance.

Such examples show why common macroeconomic regressions using total inward FDI as an explanatory variable for productivity must be interpreted carefully. A coefficient may mix the effects of external finance, foreign ownership, domestic ownership reorganized across borders, sectoral allocation, and changes in asset valuation. If the mixture varies systematically with national institutions, the estimated relationships can be misleading.

The broader FDI environment reinforces the point. UNCTAD's World Investment Report 2026 estimates global FDI of USD 1.6 trillion in 2025, 6 percent higher than the prior year. This is a global FDI figure from the cited report, not a round-tripping estimate. An international headline cannot identify the individual contributions of corporate restructuring, productive projects, or returning ownership arrangements without more detailed breakdowns.

A welfare assessment therefore requires explicit counterfactual reasoning. Which projects would occur independently of the offshore relationship? Which would be enabled by the improved availability of financing or contract enforcement? Who gains from any preference, and who bears its fiscal cost? What obligations are created, and how will they respond to adverse shocks? These questions give economic content to the analysis of returning investment.

Public Revenue, Investment Incentives, and Governance

Investment incentives seek to influence business decisions. Their economic merit depends on whether they address a well-specified market failure or deliver additional public benefits at a reasonable cost. A rule differentiating firms by immediate foreign corporate ownership may reward organizational rearrangement as readily as external investment.

At a minimum, evaluation should distinguish the eligible transaction from the intended public outcome. If the objective is fixed capital formation, investment expenditure and operational milestones are appropriate evidence. If the objective is employment, measured job creation and persistence matter. If the objective is access to foreign knowledge, evidence about technology, personnel, and supply relationships matters. Ownership geography may be relevant, but it cannot substitute for the outcome measure.

From a public-finance perspective, a subsidy or exemption to a domestic-ultimate investor is not intrinsically inefficient. It may induce additional activity with significant spillovers. The same subsidy becomes difficult to justify if the investment would have taken place on equivalent terms without the preference. The relevant comparison is incremental social value relative to fiscal cost and distortion.

Tax analysis should separate domestic ownership, the residence and substance of corporate entities, the economic allocation of taxable income, and the applicable legal obligations. A round-tripping indicator establishes none of the detailed legal conditions that determine liability in a particular case. Evidence of prohibited conduct requires a legally relevant factual foundation. Legitimate commercial arrangements and contested or unlawful conduct must remain analytically distinguishable.

Governance measures that improve the reliability of ownership and transaction reporting support both statistical accuracy and compliance oversight. Data collection should employ appropriate risk assessment, confidentiality safeguards, and proportional reporting obligations. The public interest in reliable aggregate information can often be satisfied through protected survey data and published statistics without exposing private financial details.

Investment agencies can publish two complementary descriptions of their results: the recorded value of qualifying foreign financial transactions, and the additional activity supported under the program. Where available, ultimate-investor information can add a third description of ownership. This approach prevents the success of an investment policy from being evaluated exclusively by a category of corporate finance.

Empirical Research Design

A credible empirical investigation begins by specifying an estimand. The share of inward positions under domestic ultimate control, the amount of net returning transactions, and the causal effect of preferential treatment upon organizational change are separate quantities. Each requires a different design.

For descriptive measurement, the researcher can compare immediate and ultimate-owner distributions within a common dataset, with attention to missingness and confidentiality. Where individual company accounts and ownership information are linkable, the analysis can classify investments according to the immediate investor's ultimate controlling parent and compare that result with the direct investment enterprise's controller.

For a causal question involving the introduction or withdrawal of a preference, a difference-in-differences design may be considered when there are credible comparison firms or jurisdictions. The identifying assumptions include the plausibility of parallel pre-treatment trends, the absence of simultaneous changes affecting groups differentially, and a well-defined treatment. Anticipation, selective reorganization, and spillovers can undermine these assumptions. Researchers must analyze the timing and incentives before interpreting any estimated coefficient.

A discontinuity in eligibility could sometimes support a regression-discontinuity design. Its usefulness depends on whether businesses can manipulate the eligibility variable and whether observations around the threshold are comparable. Corporate ownership thresholds may themselves be strategic choices, making this concern especially serious.

Firm-level panels can be used to investigate changes around acquisitions or financing events. The analysis should distinguish pre-existing inward positions, new equity contributions, earnings reinvestment, intercompany loans, restructurings and exits, and should track both the financial event and subsequent operating outcomes. Failure to link the investment event to the correct economic enterprise can introduce systematic measurement error.

For network models using inferred paths, researchers should report alternative parameterizations, sensitivity to unassigned counterparties, and cross-validation against official ultimate-investor series. For historical case analysis, version control of statistical releases and original institutional records is essential. A later revised series may be statistically preferable for descriptive history yet inappropriate if the research concerns information available to decision-makers at an earlier date.

Results should be presented with the level of inference warranted by their design. Descriptive associations need not be causal; a statistically precise finding can still rest on an incorrect proxy. The present monograph contains no original microdata regression and no independent estimate of causal effects.

Macroeconomic Exposure and Financial Stability

Round-tripping can alter the distribution of financial obligations across borders without transferring ultimate corporate control. Cross-border ownership positions and associated debt can affect the transmission of shocks through resident enterprises and their affiliates.

The simplified accounting example demonstrated offsetting external positions. In actual capital markets, assets and liabilities can differ in currency, maturity, valuation and liquidity. A domestic group may have an overseas holding company that owes money to external creditors and holds illiquid equity in the domestic economy. A deterioration in the domestic business may weaken the foreign unit's ability to service obligations. A disruption in offshore financing may constrain its domestic affiliate. The balance-sheet structure determines exposure.

From a financial stability perspective, aggregate netting can conceal asymmetric obligations. A banking system or enterprise sector may have considerable foreign-currency payments due even when the economy has offsetting external assets in aggregate. The residence of the ultimate owner is insufficient to identify whether available liquidity is accessible to the entity that needs it.

Equity positions also differ from borrowing. A loss on equity may be absorbed by owners; an inability to make a contractual debt payment can trigger a default or restructuring. Intragroup loans can play a commercial financing role, while specific contractual arrangements may transmit interest-rate, exchange-rate, and refinancing shocks through a group. Analysis should consider the instrument and contractual risk in addition to the ownership chain.

The distinction between external accounts and consolidated group exposures is useful in designing macroprudential research. Residence-based accounts provide an economy-wide account of claims and obligations. Group-level data can illuminate how shocks propagate among affiliated entities. Even within a domestic-ultimate group, legal barriers, insolvency arrangements, and liquidity constraints may determine whether resources are accessible across its entities.

The policy implication is that supplementary ownership statistics improve the diagnosis of cross-border financial exposures. They do not remove the need to examine the claims and obligations recorded in conventional external accounts. Both views remain necessary because the institutional structure gives financial instruments their consequences.

Scientific Synthesis and Research Frontiers

The study supports an integrated interpretation of round-tripping FDI as a geographically circular investment relationship whose economic consequences depend on the underlying financing, ownership, and use of resources. Immediate partner statistics, ultimate-investor statistics, group financing data, and measures of enterprise activity form complementary bodies of evidence.

Three distinctions organize the substantive conclusions. The first separates foreign residence of the immediate investor from domestic ultimate control. The second separates control-based attribution from the historical and contractual source of finance. The third separates the transfer or creation of a financial claim from additional productive investment. All three distinctions are necessary for a meaningful interpretation of reported return circuits.

The examined empirical materials reveal different observational strategies. Historical Chinese research emphasizes the difficulty of separating returning investment from other cross-border relationships. Russian regional research explores institutional correlates through firm-level proxies. Ukraine's official series offers a directly reported estimate of domestic-ultimate transactions under a specified denominator. U.S. and European statistics demonstrate reallocation across immediate and ultimate investor geographies.

These examples are analytically complementary rather than directly comparable. They involve different populations, years, instruments and classifications. A worldwide total cannot be constructed by combining the selected percentages and positions. A global estimate would require a common reference period, careful handling of incomplete reporting, transparent ultimate-ownership rules, and a strategy to distinguish control from funding.

Future research should strengthen the connection between ownership graphs and investment-accounting records. Internationally compatible enterprise identifiers could help, subject to confidentiality. Progress in identifying intermediate entities and publishing ultimate-host measures could improve the geographical interpretation of outward investment. Longitudinal data would better distinguish restructuring from finance supplied to new activity.

Another research priority concerns the evaluation of incentives. A systematic comparison of policies and firm behavior before and after changes in eligibility could establish when returning investment reflects additional productive projects and when it largely reflects organizational responses. Such work requires robust identification and should not infer an outcome from the inward FDI category alone.

The original contribution of this dossier is the synthesis of these measurement and economic questions into a coherent research account, supported by transparent hypothetical accounting cases. It proposes no new named theory, no novel statistical standard, and no unsupported global estimate.

Conclusion

Round-tripping foreign direct investment is a meaningful economic and statistical phenomenon because internationally recorded ownership relationships can return investment to the home economy of the ultimate investor. Its existence illustrates the importance of preserving the residence-based financial accounts while supplementing them with information about ultimate ownership, financing and economic activity.

The direct investment label establishes a cross-border investment relationship under statistical rules. Domestic ultimate ownership establishes a distinct fact about who controls the relevant investment or enterprise. The original funding source, the degree of external risk bearing, the productive use of the resources and the additionality of the investment require additional evidence.

The accounting illustrations demonstrate that valid gross cross-border positions can expand without an equivalent expansion of domestic net external claims. A domestic-ultimate investment may nevertheless involve foreign borrowing, productive capacity creation or other commercial benefits. Conversely, the introduction of an international ownership chain may change the classification of an existing financial relationship without creating new fixed capital.

International statistical standards provide a route toward improved interpretation by publishing immediate and ultimate geographical measures. The choice between alternative ultimate-owner allocation methods can materially affect estimated domestic shares. The official evidence from Ukraine, the United States and European reporting economies makes the practical significance of those measurement distinctions clear.

Economic policy should evaluate actual investment outcomes and costs alongside geographical ownership information. Measures that improve statistical attribution, carefully evaluate investment incentives and protect the integrity of financial reporting support more meaningful decisions about cross-border capital. Research on individual misconduct requires appropriately established evidence.

A mature account of the offshore economy recognizes that its geographic distribution consists of claims, corporate relationships, financial services, and productive activities that do not share one location. Round-tripping FDI makes these distinctions visible within the international investment statistics themselves.

References

Borga, Maria, and Cecilia Caliandro. 2018. Eliminating the Pass-Through: Towards FDI Statistics That Better Capture the Financial and Economic Linkages between Countries. NBER Working Paper 25029. Source record.

Casella, Bruno. 2019. Looking through Conduit FDI in Search of Ultimate Investors – A Probabilistic Approach. Transnational Corporations 26(1). DOI: 10.18356/8a8b094c-en. Source record.

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. Source record.

Eurostat. 2025. Top Ultimate Investing Economies: US and Germany. News article, November 11. Dataset bop_fdi6_pos. Source record.

International Monetary Fund. 2025. Release of New Standards for Macroeconomic Statistics (BPM7). IMF Policy Paper, 2025/023. Source record.

International Monetary Fund. n.d. Direct Investment Positions by Counterpart Economy (DIP), formerly Coordinated Direct Investment Survey (CDIS). IMF Data, dataset documentation. Source record.

Ledyaeva, Svetlana, Päivi Karhunen, and John Whalley. 2013. Offshore Jurisdictions (Including Cyprus), Corruption Money Laundering and Russian Round-Trip Investment. NBER Working Paper 19019. DOI: 10.3386/w19019. Source record.

Ledyaeva, Svetlana, Päivi Karhunen, Riitta Kosonen, and John Whalley. 2015. Offshore Foreign Direct Investment, Capital Round-Tripping, and Corruption: Empirical Analysis of Russian Regions. Economic Geography 91(3): 305–341. DOI: 10.1111/ecge.12093. Source record.

National Bank of Ukraine. 2026. Estimation of Round Tripping Transactions for 2010–2025. Statistical presentation, slide 3, Core Points: Results. Source record.

OECD. 2023. The Impact of Regulation on International Investment in Portugal. Paris: OECD Publishing. Chapter, Trends and Impacts of Foreign Direct Investment in Portugal. Source record.

OECD. 2025. OECD Benchmark Definition of Foreign Direct Investment (Fifth Edition). Paris: OECD Publishing. Chapter 2, Main Concepts and Definitions of Foreign Direct Investment. DOI: 10.1787/7f05c0a3-en. Source record.

OECD. 2025. OECD Benchmark Definition of Foreign Direct Investment (Fifth Edition). Paris: OECD Publishing. Chapter 6, Special Cases, Annex 6.A. DOI: 10.1787/7f05c0a3-en. Source record.

OECD. 2025. OECD Benchmark Definition of Foreign Direct Investment (Fifth Edition). Paris: OECD Publishing. Chapter 8, Disaggregation by Geography and by Industry. DOI: 10.1787/7f05c0a3-en. Source record.

OECD. 2025. OECD Benchmark Definition of Foreign Direct Investment (Fifth Edition). Paris: OECD Publishing. Standard and Supplementary FDI Series. DOI: 10.1787/7f05c0a3-en. Source record.

OECD. 2026. Infrastructure Policy Review of Ukraine. Paris: OECD Publishing. Chapter, Macroeconomic Developments and the Enabling Environment for Infrastructure Investment and Financing. Source record.

Reserve Bank of India. 2024. Census on Foreign Liabilities and Assets of Indian Direct Investment Companies, 2023–24: Provisional Results. RBI statistical release. Source record.

U.S. Bureau of Economic Analysis. 2024. Direct Investment by Country and Industry, 2023. News release, July 2024. Source record.

UN Trade and Development (UNCTAD). 2026. World Investment Report 2026: International Investment in a Turbulent Era. Geneva: United Nations. Source record.

Xiao, Geng. 2005. Round-Tripping Foreign Direct Investment in the People's Republic of China: Scale, Causes and Implications. Asian Development Bank Institute, January. Revised December 2004 discussion paper. Source record.