## 1. List of Low-Tax Economies (Content unit: wp17258)

## Source details

**Canonical URL:** [1. List of Low-Tax Economies (Content unit: wp17258)](https://www.imf.org/-/media/files/publications/wp/2017/wp17258.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2017/wp17258.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2017/wp17258.pdf.json)

---

### Introduction: FDI decoupling and motivation for a new global FDI network
- FDI is used to analyze globalization of production, attractiveness of an economy, long-term relationships between economies, technology transfer, and real economic activity generated by foreign companies and investors.
- Central analytical challenge: geographical decoupling in FDI due to offshore financial centers, special purpose entities (SPEs), and intangibles that are easily moved across economies.
- Three main manifestations motivating a new FDI network that removes SPEs and breaks down inward FDI by the ultimate investing economy (UIE):
  - Bilateral asymmetries between reported inward and outward FDI for many economy pairs.
  - Small economies (e.g., Netherlands, Luxembourg) appearing disproportionately large in global FDI because they host many SPEs and MNEs.
  - Complexity of ownership chains producing immediate counterpart economies that differ from the UIE which bears ultimate risks and rewards.
- Purpose of the new FDI network:
  - Remove SPEs and present inward FDI broken down by the UIE for a large set of economies.
  - Combine detailed OECD UIE data with broad coverage of the IMF’s Coordinated Direct Investment Survey (CDIS).
  - Rationale: inward FDI is generally of higher quality than outward FDI.

### Bilateral asymmetries in official FDI data: magnitude, sources, and implications
- Data context:
  - IMF CDIS published annually since 2010 with global coverage of 116 economies.
  - CDIS breaks FDI positions down by the immediate counterpart economy.
  - CDIS follows BPM6 and OECD BMD4; CDIS Guide clarifications referenced.
- Key empirical findings on asymmetries (CDIS end-2015, initial release December 2016):
  - For 44 percent of the 1,805 published bilateral economy pairs, one economy’s reported number is at least twice as high as the counterpart economy’s number.
  - For almost 10 percent of the pairs, one number is at least 10 times higher than the mirror number.
  - Average discrepancy between inward and outward FDI for the 1,805 economy pairs is USD 5.9 billion.
  - Average inward FDI position: USD 12.3 billion.
  - Average outward FDI position: USD 13.0 billion.
  - Coverage in the initial CDIS release: total FDI, equity, and debt instruments cover 1,805, 1,403, and 906 pairs, respectively.
  - FDI equity accounts for approximately 80 percent of total FDI and FDI debt instruments for 20 percent.
- Identified methodological and compilation sources:
  - Valuation/estimation choices: unlisted FDI equity commonly valued by Own Funds at Book Value (OFBV); 18 economies use other valuation principles for unlisted equity; 52 economies deviate from recommended valuation for listed equity.
  - Presentation principles: directional vs asset-liability principles and differing geographical presentation can create asymmetries.
  - Practical compilation/coverage issues (examples): differing country treatments of territories (Channel Islands example) and sample differences.

### Effects of SPEs and the immediate vs ultimate investing economy (UIE)
- Nature and role of SPEs:
  - SPEs typically focus on group financing, financial holding activities, or tax planning and often have little real economic activity where domiciled.
  - SPEs inflate FDI statistics because they have large gross foreign positions but small net foreign positions.
  - Financial centers hosting many SPEs (e.g., Netherlands, Luxembourg) appear more important in immediate-counterpart based FDI statistics than when FDI is allocated by UIE.
- Immediate counterpart economy vs UIE:
  - Immediate counterpart economy measures direct exposures but not the ultimate owner bearing risks and rewards nor the investments’ final destination.
  - OECD encouragement: break down inward FDI by UIE.
  - New FDI network differences:
    - Total inward FDI positions reduced by approximately one-third compared to CDIS.
    - Financial-center dominance diminishes when allocating by UIE.

### Box 1: Fellow enterprises and geographical asymmetries (key points)
- Presentation principles and fellow-enterprise implications:
  - Directional principle (recommended by CDIS) records reverse investment as negative positions and avoids certain inflations that arise under the asset-liability principle.
  - Loans between fellow enterprises can generate bilateral geographical asymmetries, particularly for FDI debt instruments.
  - Fellow enterprises primarily involve FDI in debt instruments, contributing to larger asymmetries for debt than equity.
- Data compilation issues:
  - Differences in survey coverage, grossing up practices, and changing company demographics complicate geographic symmetry.
  - Confidential information sharing among compilers and planned IMF sharing aim to improve data and address SPE issues.
- SPE identification criteria (from statistical manuals/BMD4, Box 6.2):
  - Formally registered legal entity; ultimate owners not residents of territory of incorporation; few or no employees; little or no production; little or no physical presence; most assets and liabilities vis-à-vis non-residents; core business of group financing or holding activities.
- Tax/regulatory drivers and risks:
  - Incentives in offshore financial centers drive SPE expansion; cited estimate of annual U.S. tax revenue loss from offshore tax exploitations around USD 100 billion (U.S. Senate Permanent Subcommittee on Investigations, 2008).
  - SPE funding and location instability: small legislative changes can shift investment patterns causing capital outflows.

### Low-tax economies list (as provided in source)
- Asia: Hong Kong SAR, Macao SAR, Maldives, Singapore
- Caribbean: Anguilla, Antigua and Barbuda, Aruba, Bahamas, Barbados, British Virgin Islands, Cayman Islands, Dominica, Grenada, Montserrat, Netherlands Antilles, St. Kitts and Nevis, St. Lucia, St. Vincent and Grenadines, Turks and Caicos Islands, U.S. Virgin Islands
- Central America: Belize, Costa Rica, Panama
- Eastern Africa: Mauritius, Seychelles
- Europe: Andorra, Cyprus, Gibraltar, Guernsey, Ireland, Isle of Man, Jersey, Latvia, Liechtenstein, Luxembourg, Malta, Monaco, San Marino, Switzerland
- Northern America: Bermuda
- Middle East: Bahrain, Jordan, Lebanon
- Oceania: Cook Islands, Marshall Islands, Nauru, Niue, Samoa, Vanuatu
- Western Africa: Liberia
- Note: Economies in the list appeared in at least one of: (i) OECD's list of committed jurisdictions and uncooperative tax havens (no jurisdictions included since 2009), (ii) Dharmapala and Hines (2006) tax haven list, (iii) IRS list of offshore haven or financial privacy jurisdictions. Economies in bold in the original report report to the CDIS.

### Tax optimization, transfer pricing, corporate inversions, and FDI intensity
- MNE tax optimization through SPEs/operating units involves profit shifting to low-tax jurisdictions via debt allocation, transfer pricing, or corporate inversions.
- Transfer pricing enforcement is difficult for intellectual property and intangibles.
- U.S. intra-group trade shares (Lanz and Miroudot, 2011):
  - Intra-group trade in goods accounts for 48 percent of total imports and 30 percent of exports.
  - Intra-group services: 22 percent of imports and 26 percent of exports.
- High-profile EU state-aid rulings: Ireland, Luxembourg, Netherlands cases (Apple, Fiat, Starbucks) where transfer prices were found not to reflect economic prices.
- Corporate inversions can move parent headquarters to low-tax jurisdictions without new FDI funding and can affect national statistics (significant impact on Irish GDP; OECD, 2016).
- BEPS implementation may increase near-SPE structures because MNEs will need more presence in low-tax jurisdictions to claim permanent establishment.
- FDI intensity examples (inward FDI-to-GDP):
  - Luxembourg: including SPEs 5,658 percent of GDP; excluding SPEs 393 percent of GDP.
  - Netherlands: including SPEs 525 percent of GDP; excluding SPEs 97 percent of GDP.
  - Luxembourg, Mauritius, Malta, Cyprus largest by FDI intensity; Netherlands, Hong Kong SAR, Ireland, Singapore, Switzerland also near top.

### Inward FDI by UIE: reporting patterns and stylized facts
- UIE definition per BMD4 (para. 610–611): proceed up ownership chain through controlling links (ownership >50 percent of voting power) until an enterprise not controlled by another is reached; country of that ultimate investor is the UIE.
- By January 2017, 12 OECD countries had reported inward FDI positions by UIE: Austria, Czech Republic, Estonia, Finland, France, Germany, Hungary, Iceland, Italy, Poland, Switzerland, United States.
- Observed patterns:
  - Financial-center economies (e.g., Luxembourg, Netherlands) are more dominant on immediate-investor basis than on UIE basis.
  - United States and Germany are more dominant by UIE, consistent with hosting parents of MNEs investing through chains abroad.
  - Ireland more dominant by UIE than immediate investor partly due to U.S. corporate inversions.
  - Round-tripping: Italy reported round-tripping exceeds 10 percent of inward FDI; average for reporting OECD countries is 5 percent.

### Constructing new global FDI estimates: key methodological highlights
- Overall approach:
  - Build estimates excluding SPEs and broken down by UIE using OECD detailed data and CDIS coverage.
  - Use inward FDI positions as basis because ultimate ownership information is currently available only for inward FDI and is generally higher quality.
- Data inputs (mostly latest available year 2015):
  - Inward FDI (CDIS) for 116 CDIS countries broken down by immediate counterpart economy.
  - Aggregate SPE/non-SPE breakdown for 30 reporting OECD countries.
  - Nominal GDP (IMF World Economic Outlook Database) for CDIS countries.
  - Detailed non-SPE breakdown for 25 reporting OECD countries by immediate counterpart economy.
  - Detailed UIE breakdown for 12 reporting OECD countries.
- SPE adjustment model (Step 2) specification and behavior:
  - Univariate regression: dependent variable = non-SPE inward FDI-to-total inward FDI ratio; explanatory variable = total inward FDI-to-GDP ratio.
  - Log-log transformation for linearity.
  - Sample: economies with FDI intensity above 30 percent (none below that threshold reported hosting SPEs).
  - Model behavior:
    - Generates SPE adjustment factors below 1 for economies with FDI intensity higher than 46 percent.
    - Adjustment factors above 1 for economies with FDI intensity lower than 46 percent; for FDI intensity below 46 percent no SPE adjustments are made.
    - Statistical result: slope significantly negative (p-value<0.001), 95 percent confidence interval from -0.639 to -0.458. Data refer to end-2015.
    - Example: adjustment factor 0.8 for an economy with inward FDI-to-GDP ratio of 70 percent (adjusts inward FDI down by 20 percent).
    - Observed structural non-SPE ceiling: except for Luxembourg, no OECD country reported non-SPE inward FDI exceeding 114 percent of GDP.
- UIE adjustment factors (a_c):
  - Constructed from UIE-to-immediate-counterpart relationships for the 12 UIE-reporting OECD countries.
  - Formula presented: a_c = (Σ_i FDI(UIE)_{i,c}) / (Σ_i FDI(IIE)_{i,c}) where c is counterpart economy and i indexes reporting economies.
  - Example: Netherlands aggregation for the 12 OECD reporting countries: USD 326 billion on UIE basis and USD 906 billion on immediate counterpart basis → a_NLD = 0.36.
  - Caps: adjustment factors generally capped between 0.33 and 3.
  - For economies on the low-tax list: adjustment factors capped between 0.20 and 1 so FDI from these economies can never be higher under UIE than under immediate counterpart.
- Round-tripping adjustment factor (b):
  - Calculated as the simple average for reporting economies: b = (1/n) Σ_i (FDI(UIE)_{i,ci} / totalFDI_i).
  - Implementation: effectively allocates 5 percent of inward FDI back to its own economy (average reported).
- Construction steps summary:
  - Step 3: build non-SPE global estimates using reported non-SPE breakdowns where available; apply SPE adjustment factors to others.
  - Step 4: calculate UIE and round-tripping adjustment factors and apply caps.
  - Step 5: produce final estimates excluding SPEs and broken down by UIE; for 12 UIE-reporting OECD economies use reported data directly; for others apply a_c and b; constraint applied so adjusted FDI cannot exceed reported total inward FDI for an economy.
- Verification and fit:
  - Applying SPE adjustment factors to SPE-reporting OECD countries yields generally small discrepancies.
  - UIE estimations are more uncertain because large adjustments were required for all reporting economies.
  - Joint model test (UIE estimations based on estimated SPE data) shows highest discrepancies.
  - Discrepancies between estimated and reported inward FDI are smaller than bilateral discrepancies in reported data.
  - Caution: estimates based on a subset of reporting OECD countries are uncertain; more reporting of SPE and UIE breakdowns would improve estimates.

### Main empirical findings, key statistics, and stylized facts from the new global FDI estimates
- Removing SPEs reduces total inward FDI positions by one-third (34 percent) compared to the CDIS.
- Round-tripping averages 5 percent of inward FDI.
- Financial centers remain important even after removing SPEs, indicating either active portfolio/management roles for some entities in financial centers or incomplete identification of SPEs in reported data.
- Network implications:
  - CDIS-based network (including SPEs and immediate counterpart): United States, Netherlands, Luxembourg dominate; reporting economies typically receive inward FDI from 60–90 different economies.
  - New network (SPEs removed and UIE breakdown): United States still dominates; role of Netherlands and Luxembourg much smaller; some low-tax economies drop out of top 40 and are replaced by more traditional FDI economies (examples in source: Czech Republic, Finland, Israel, Saudi Arabia, United Arab Emirates).
  - Hong Kong SAR’s inward FDI significantly lower when SPEs removed; strong link with China (Mainland) persists.
  - Major offshore financial centers not reporting to CDIS (British Virgin Islands, Bermuda, Cayman Islands, Gibraltar, Jersey) are important in the global network and would be even more so if they reported.
- Representative numeric thresholds and examples used in methodology:
  - SPE adjustment sample cutoff: 30 percent FDI intensity.
  - FDI intensity threshold for SPE-adjustment direction: 46 percent.
  - Observed non-SPE ceiling (excluding Luxembourg): 114 percent of GDP.
  - Example UIE adjustment factor: a_NLD = 0.36 based on USD 326 billion UIE vs USD 906 billion immediate counterpart for the Netherlands (12 reporting OECD countries).
  - Adjustment factor caps: 0.33 to 3 generally; 0.20 to 1 for low-tax economies.

### Policy and data implications (as presented in source)
- To describe a globalized world, macroeconomic statistics should supplement national-territory-based information with data on global interconnectedness.
- Financial statistics, including the CDIS, could be supplemented with ultimate counterpart economy (UIE) information for a more comprehensive picture of ultimate cross-economy financial linkages and risks.
- Improved reporting of SPE and UIE breakdowns across more countries would reduce estimation uncertainty and improve global FDI network construction.

*Source: IMF Working Paper (Box 1. Fellow Enterprises and Geographical Asymmetries), content unit wp17258.*

### 1.      List of Low-Tax Economies ......................................................................................

### 1.      List of Low-Tax Economies

### Introduction: FDI decoupling and motivations for a new global FDI network
- Foreign direct investment (FDI) is a key link in global economic interconnectedness and is used to analyze globalization of production, attractiveness of an economy, long-term relationships between economies, technology transfer, and real economic activity generated by foreign companies and investors.
- Central analytical challenge: geographical decoupling in FDI — traditional FDI data often do not reflect real economic integration because of offshore financial centers, special purpose entities (SPEs), and intangibles (e.g., intellectual property rights) that are easily moved across economies.
- Three main ways FDI measures are geographically decoupled, motivating construction of a new FDI network that removes SPEs and breaks down inward FDI by the ultimate investing economy (UIE):
  - Bilateral asymmetries between reported inward and outward FDI for many economy pairs.
  - Small economies (e.g., Netherlands, Luxembourg) can appear disproportionately large in global FDI because they host many SPEs and MNEs, inflating financial FDI measures relative to real activity.
  - Complexity of ownership chains means the immediate counterpart economy often differs from the UIE that bears ultimate risks and rewards.

- Purpose of the new FDI network constructed in the paper:
  - Remove SPEs and present inward FDI broken down by the UIE for a large set of economies.
  - Combine detailed OECD UIE data with broad coverage of the IMF’s Coordinated Direct Investment Survey (CDIS).
  - Rationale: inward FDI is generally of higher quality than outward FDI. The new network aims to provide a clearer picture of real economic integration and ultimate financial linkages.

### Bilateral asymmetries in official FDI data: magnitude, sources, and implications
- Coverage and data context:
  - IMF CDIS published annually since 2010 with a global coverage of 116 economies.
  - CDIS breaks FDI positions down by the immediate counterpart economy, allowing comparison of economy-pair inward/outward records.
  - Data for a reference year are published in December of the following year; some economies report with a longer time lag.
  - The CDIS follows BPM6 and the OECD BMD4, with further clarifications in the Coordinated Direct Investment Survey Guide – 2015 (CDIS Guide).

- Key empirical findings on asymmetries (CDIS end-2015, initial release December 2016):
  - For 44 percent of the 1,805 published bilateral economy pairs, one economy’s reported number is at least twice as high as the counterpart economy’s number.
  - For almost 10 percent of the pairs, one number is at least 10 times higher than the mirror number.
  - Average discrepancy between inward and outward FDI for the 1,805 economy pairs is USD 5.9 billion.
  - Average inward FDI position: USD 12.3 billion.
  - Average outward FDI position: USD 13.0 billion.
  - Conclusion: no systematic overstatement of inward FDI compared to outward FDI, or vice versa, in aggregate.

- Instrumental coverage in the CDIS initial release:
  - Total FDI, equity, and debt instruments cover 1805, 1403, and 906 pairs, respectively, reflecting data availability at the initial release of the 2015 CDIS in December 2016.
  - FDI equity accounts for approximately 80 percent of total FDI and FDI debt instruments for 20 percent.

- Identified methodological and compilation sources of asymmetries:
  - Valuation and estimation methods:
    - Unlisted FDI equity (most common) lacks market prices; fair valuation methods are used.
    - Choice of valuation method can have a significant impact (Damgaard and Elkjaer (2014) evidence cited).
    - Own funds at book value (OFBV) is the most common valuation method and promotes cross-economy comparability but may not approximate current market values when assets/liabilities are at historical costs or intangibles are limited in accounting.
    - CDIS Guide recommends: unlisted FDI equity at OFBV and listed equity at market value.
    - Compliance deviations: 18 economies use other valuation principles for unlisted equity for inward and/or outward FDI; 52 economies deviate from the recommended valuation principle for listed equity.
  - Presentation principles and treatment choices:
    - The methodology allows FDI to be presented under different principles, which can generate asymmetric geographical data.
    - Some economies report geographical breakdown on the UIE basis; sometimes no counterpart economy is allocated to SPEs.
  - Practical compilation and coverage issues:
    - Example: United States includes the Channel Islands as part of the United Kingdom, but the United Kingdom does not.
    - Differences in coverage and sample uncertainties contribute to asymmetries.

### Effects of SPEs and the immediate vs ultimate investing economy (UIE)
- Nature and role of SPEs:
  - SPEs typically focus on group financing, financial holding activities, or tax planning and do not necessarily reflect stable investment motives.
  - SPEs often have no or very limited real economic activity in the economy where they are domiciled but can significantly inflate FDI statistics.
  - Financial centers that host many SPEs (e.g., Netherlands, Luxembourg) appear much more important in immediate-counterpart based FDI statistics than when FDI is allocated by UIE.

- Importance of distinguishing immediate counterpart economy vs UIE:
  - Immediate counterpart economy provides a good measure for direct exposures but lacks information about the ultimate owner who bears risks and rewards or the investments’ end destination.
  - OECD countries are encouraged to break down inward FDI by UIE to close this data gap.
  - The global geographical FDI network differs substantially when based on immediate counterpart economy versus UIE; financial centers hosting SPEs are much less important as ultimate FDI economies.

- Impact of the new FDI network constructed in the paper:
  - Total inward FDI positions in the new network are reduced by approximately one-third compared to the CDIS.
  - Economies hosting financial centers play a less dominant role in the new network.
  - The new network provides a clearer picture of real economic integration and ultimate financial linkages than current available data.

### Organization of the paper (content coverage)
- Section II: analyzes bilateral asymmetries between inward and outward FDI reported by economy pairs and provides explanations for these asymmetries.
- Section III: discusses the role and treatment of SPEs in FDI.
- Section IV: compares inward FDI broken down by the immediate counterpart economy and the UIE.
- Section V: develops new global FDI estimates and tests them against reported data.
- Section VI: analyzes global FDI networks based on CDIS data and the new FDI estimates.
- Section VII: summarizes the key conclusions.

*Source: IMF staff paper (chapter/excerpt provided).*

### Box 1. Fellow Enterprises and Geographical Asymmetries

### Box 1. Fellow Enterprises and Geographical Asymmetries

### Statistical sources of bilateral asymmetries
- Two presentation principles for FDI: the directional principle (recommended by the CDIS) and the asset-liability principle (gross positions).
- Under the directional principle, reverse investment from an FDI enterprise to its parent is recorded as negative outward FDI in the parent's economy and negative inward FDI in the FDI enterprise's economy, providing symmetric recording.
- Loans between fellow enterprises (related through a common investor) can generate bilateral geographical asymmetries. Example structure in the source: an FDI enterprise in economy B lends to a fellow enterprise in economy C while the ultimate controlling parent resides in economy A; this results in negative inward FDI in economy B and positive inward FDI in economy C.
- The directional principle avoids inflating aggregated FDI as the asset-liability principle would for certain economies (e.g., economy B in the example).
- Fellow enterprises primarily have FDI in debt instruments; this contributes to larger bilateral asymmetries for FDI debt instruments than for FDI equity (referenced Figure 1).

### Data compilation and reporting issues
- Some economies survey all FDI enterprises while others gross up; changing company demographics (new FDI, mergers, splits, bankruptcies) complicate maintaining the reporting population.
- Confidential information sharing among compilers (European Central Bank and Eurostat, 2014) and planned IMF information sharing based on the CDIS aim to improve data and address SPE issues.

### The role of Special Purpose Entities (SPEs)
- SPEs break the direct link between the receiving economy and the ultimate owner and “inflate” FDI because SPEs have large gross foreign positions but very small net foreign positions.
- Typical SPE identification criteria (from statistical manuals/BMD4, Box 6.2) include: formally registered legal entity subject to national law; ultimate owners not residents of the territory of incorporation; few or no employees; little or no production; little or no physical presence; most assets and liabilities vis-à-vis non-residents; core business of group financing or holding activities.
- Empirical observations:
  - Blanchard and Acalin (2016): high positive correlation between quarterly FDI inflows and outflows in several economies, suggesting transitory passage of FDI through economies.
  - Lane and Milesi-Ferretti (2017): FDI positions have continued to expand post-financial crisis, primarily from financial centers and corporate structure complexity.
- Tax/regulatory/confidentiality incentives in offshore financial centers drive SPE expansion. Example: annual U.S. tax revenue loss from offshore tax exploitations estimated to be around USD 100 billion (U.S. Senate Permanent Subcommittee on Investigations, 2008).
- SPE funding and location are likely less stable; small legislative changes can shift investment patterns and cause capital outflows.

### Low-tax economies (Table 1 as listed in source)
- Asia: Hong Kong SAR, Macao SAR, Maldives, Singapore
- Caribbean: Anguilla, Antigua and Barbuda, Aruba, Bahamas, Barbados, British Virgin Islands, Cayman Islands, Dominica, Grenada, Montserrat, Netherlands Antilles, St. Kitts and Nevis, St. Lucia, St. Vincent and Grenadines, Turks and Caicos Islands, U.S. Virgin Islands
- Central America: Belize, Costa Rica, Panama
- Eastern Africa: Mauritius, Seychelles
- Europe: Andorra, Cyprus, Gibraltar, Guernsey, Ireland, Isle of Man, Jersey, Latvia, Liechtenstein, Luxembourg, Malta, Monaco, San Marino, Switzerland
- Northern America: Bermuda
- Middle East: Bahrain, Jordan, Lebanon
- Oceania: Cook Islands, Marshall Islands, Nauru, Niue, Samoa, Vanuatu
- Western Africa: Liberia
- Note (from source): Economies in the list appeared in at least one of: (i) OECD's list of committed jurisdictions and uncooperative tax havens (no jurisdictions included since 2009), (ii) Dharmapala and Hines (2006) tax haven list, (iii) IRS list of offshore haven or financial privacy jurisdictions. Economies in bold report to the CDIS (bolding in original source).

### Tax optimization, transfer pricing, and corporate inversions
- MNE tax optimization through SPEs or operating units often involves shifting profits to low-tax jurisdictions via debt allocation, transfer pricing, or corporate inversions.
- Transfer pricing issues: arm's-length principle is difficult to enforce for transfers of intellectual property and intangibles.
- U.S. intra-group trade shares (Lanz and Miroudot, 2011): intra-group trade in goods accounts for 48 percent of total imports and 30 percent of exports; 22 percent and 26 percent of services imports and exports, respectively.
- High-profile EU state-aid rulings found Ireland, Luxembourg, and the Netherlands allowed transfer prices that did not reflect economic prices (Apple, Fiat, Starbucks), with countries instructed to collect additional taxes and appeals filed.
- Corporate inversions move parent headquarters to low-tax jurisdictions without new FDI funding; can function as near-SPEs and affect national statistics (example: significant impact on Irish GDP data; OECD, 2016).
- BEPS implementation may increase near-SPE structures because MNEs will need more presence in low-tax jurisdictions to claim permanent establishment.

### SPEs, FDI intensity, and country examples
- Major absolute FDI recipients include United States, China (Mainland), United Kingdom, Germany, and France; smaller economies with financial centers and high FDI: Netherlands, Luxembourg, Hong Kong SAR, Singapore, Ireland, Switzerland (Figure 2 in source).
- FDI intensity (inward FDI-to-GDP): Luxembourg largest by wide margin, followed by Mauritius, Malta, and Cyprus; Netherlands, Hong Kong SAR, Ireland, Singapore, and Switzerland remain near the top.
- SPE impact on reported FDI-to-GDP:
  - Luxembourg inward FDI position excluding SPEs: 393 percent of GDP; including SPEs: 5,658 percent of GDP.
  - Netherlands inward FDI position excluding SPEs: 97 percent of GDP; including SPEs: 525 percent of GDP.

### Inward FDI by Ultimate Investing Economy (UIE)
- UIE definition per BMD4 (para. 610–611): proceed up ownership chain through controlling links (ownership of more than 50 percent of voting power) until an enterprise not controlled by another is reached; the country of that ultimate investor is the ultimate investing country (UIC/UIE).
- By January 2017, 12 OECD countries had reported inward FDI positions by the UIE for the OECD's BMD4 Partner Country Database: Austria, Czech Republic, Estonia, Finland, France, Germany, Hungary, Iceland, Italy, Poland, Switzerland, United States.
- Patterns:
  - Economies with significant SPE presence (e.g., Luxembourg, Netherlands) are more dominant when measured by immediate investing economy than by UIE, suggesting pass-through investments originating elsewhere.
  - United States and Germany (few/no resident SPEs) are more dominant when measured by UIE, indicating they host parents of MNEs investing through chains abroad.
  - Ireland is more dominant by UIE than immediate investing economy in part due to U.S. corporate inversions and resulting UIE links in U.S. data.
- Round-tripping (ultimate investor from same economy as direct investment enterprise) “inflates” FDI; reported instance: Italy round-tripping exceeds 10 percent of inward FDI; average for reporting OECD countries is 5 percent (Figure 4).

### Constructing new global FDI estimates (methodology highlights)
- New estimates exclude SPEs and break down FDI by UIE; built from OECD detailed data and broad CDIS coverage.
- Rationale for using inward FDI positions: generally better quality than outward FDI and ultimate ownership information currently available only for inward FDI.
- For OECD countries reporting non-SPE and UIE breakdowns, those data are included directly.
- For remaining economies, CDIS data are adjusted for SPEs and then transformed from immediate counterpart to UIE.
- SPE adjustment model:
  - Empirical tendency: economies with high total inward FDI-to-GDP ratios are more likely to host SPEs (Annex I).
  - Observed structural limit: Except for Luxembourg, no OECD country has reported an inward FDI position for non-SPEs that exceeds 114 percent of GDP.
  - Model specification: univariate regression with non-SPE inward FDI-to-total inward FDI ratio as dependent variable and total inward FDI-to-GDP ratio as explanatory variable; log-log transformation for linearity.
  - Sample: economies with FDI intensity above 30 percent (since none below that threshold host SPEs).
  - Model behavior: generates SPE adjustment factors below 1 for economies with FDI intensity higher than 46 percent; adjustment factors above 1 for economies with FDI intensity lower than 46 percent. For FDI intensity below 46 percent, no SPE adjustments are made.
  - Statistical result: slope is significantly negative (p-value<0.001), with the 95 percent confidence interval ranging from -0.639 to -0.458. Data refer to end-2015.
  - Example: model generates an adjustment factor of 0.8 for an economy with an inward FDI-to-GDP ratio of 70 percent (adjusting down inward FDI by 20 percent); only seven OECD countries have ratios above that threshold.
  - Alternative multivariate approach noted but not used due to overfitting risk.
- UIE geographical adjustment factors (a_c):
  - Constructed from UIE to immediate counterpart relationships for the 12 UIE-reporting OECD countries.
  - Formula in source: a_c = (Σ_i FDI(UIE)_{i,c}) / (Σ_i FDI(IIE)_{i,c}) for c_i ≠ (notations as in source).
  - Example: aggregate inward FDI position from the Netherlands for the 12 OECD reporting countries: USD 326 billion on UIE basis and USD 906 billion on immediate counterpart basis → adjustment factor a_NLD = 0.36.
  - Adjustment factors capped between 0.33 and 3 to avoid extreme adjustments; for economies on the low-tax list (likely high SPE presence), adjustment factors capped between 0.2 and 1 so FDI from these economies can never be higher under UIE than under immediate counterpart.
- Round-tripping adjustment factor (b):
  - Calculated as the simple average for reporting economies: b = (1/n) Σ_i (FDI(UIE)_{i,ci} / totalFDI_i) for ci = (notations as in source).
  - Applied to total non-SPE inward FDI for each economy, effectively allocating 5 percent of inward FDI back to its own economy (the 5 percent figure reported in source).
- Final constraints: adjustments constrained so adjusted FDI cannot exceed reported total inward FDI for an economy.
- The new global FDI estimates (full economy-by-economy matrix) are published along with the paper.

### Model verification and fit
- Verification approach: apply SPE adjustment factors to CDIS data for SPE-reporting OECD countries and UIE adjustment factors to reported non-SPE data for UIE-reporting OECD countries.
- Discrepancies for SPE estimations are generally small because only a few OECD countries have large SPE presence (adjustments modest in most cases).
- UIE estimations are more uncertain because large adjustments are made for all reporting economies and investment patterns vary across economies.
- Joint model test (UIE estimations based on estimated SPE data) shows highest discrepancies.
- Discrepancies between estimated and reported inward FDI are smaller than bilateral discrepancies in reported data (Figure 1).
- Caution: model estimates based on a subset of reporting OECD countries are inevitably uncertain; more reporting of SPE and UIE breakdowns would improve estimates.

### Main empirical findings and stylized facts from new global FDI estimates
- Removing SPEs reduces total inward FDI positions by one-third compared to the CDIS (total inward FDI positions are 34 percent lower).
- “Traditional” major economies become more dominant in the adjusted global FDI network.
- Financial centers remain important even after removing SPEs, indicating either active portfolio/management roles for some entities in financial centers or incomplete identification of SPEs in reported data.
- Round-tripping is on average 5 percent of FDI.
- Example caps and adjustments in methodology:
  - SPE adjustment thresholds and model behavior: 30 percent sample cutoff, 46 percent FDI intensity threshold for adjustment direction, 114 percent non-SPE ceiling observed (excluding Luxembourg).
  - UIE adjustment factor example: a_NLD = 0.36 for the Netherlands based on 12 reporting OECD countries (USD 326 billion UIE vs USD 906 billion immediate counterpart).
  - Adjustment factor caps: 0.33 to 3 generally; 0.2 to 1 for low-tax economies.

### Network implications
- CDIS-based network (including SPEs and immediate counterpart): United States, Netherlands, Luxembourg dominate; very high connectedness with reporting economies typically receiving inward FDI from 60–90 different economies.
- New network (SPEs removed and UIE breakdown): United States still dominates; role of Netherlands and Luxembourg much smaller; some economies on low-tax lists drop out of top 40 and are replaced by more traditional FDI economies (examples: Czech Republic, Finland, Israel, Saudi Arabia, United Arab Emirates).
- Hong Kong SAR’s inward FDI significantly lower when SPEs removed, but strong link with China (Mainland) persists in the adjusted network because no UIE adjustment was made for Hong Kong based on reporting OECD patterns.
- Some major offshore financial centers (British Virgin Islands, Bermuda, Cayman Islands, Gibraltar, Jersey) are important in the global network despite not reporting to CDIS; their importance would be larger if they reported.

### Conclusions (as presented in source)
- SPEs and offshore financial centers can hide ultimate bilateral linkages; adjusted global FDI data accounting for SPEs and ultimate ownership provide a different and arguably more informative picture of long-term relations and final investment patterns.
- Key takeaways:
  - Removing SPEs reduces total inward FDI positions by one-third (34 percent).
  - Traditional major economies become more dominant in the adjusted network.
  - Financial centers remain important even after SPE removal.
  - Round-tripping averages 5 percent of FDI.
- Policy/data implication: to describe a globalized world, macroeconomic statistics should supplement national-territory-based information with data on global interconnectedness; financial statistics, including the CDIS, could be supplemented with ultimate counterpart economy information for a comprehensive picture of ultimate cross-economy financial linkages and risks.

*Source: IMF Working Paper (Box 1. Fellow Enterprises and Geographical Asymmetries) — content unit wp17258 as provided.*

### ANNEX I: SPE BREAKDOWN

### ANNEX I: SPE BREAKDOWN

### Table I.1. Inward FDI Positions Broken Down by SPEs and Non-SPEs
- Table columns (as presented): FDI (percent of GDP) — Total, SPEs, Non-SPEs ; FDI country rank — Total, Non-SPEs ; Change in rank
- Country entries (Total SPEs Non-SPEs   Total  Non-SPEs Change):
  - Luxembourg: 5,658 5,266 393   1 1 0
  - Netherlands: 525 428 97   2 3 -1
  - Hungary: 161 92 69   3 7 -4
  - Switzerland: 132 19 114   4 2 2
  - Belgium: 100 6 94   5 4 1
  - Chile: 87 1 85   6 5 1
  - Estonia: 84 2 82   7 6 1
  - Austria: 70 24 46   8 14 -6
  - Iceland: 66 19 47   9 13 -4
  - Sweden: 61 4 57   10 9 1
  - Czech Republic: 60 0 60   11 8 3
  - Portugal: 58 6 51   12 12 0
  - Slovak Republic: 56 0 56   13 10 3
  - Latvia: 55 0 55   14 11 3
  - United Kingdom: 54 14 41   15 16 -1
  - Spain: 46 3 42   16 15 1
  - New Zealand: 39 0 39   17 17 0
  - Poland: 39 0 38   18 18 0
  - Denmark: 38 4 34   19 21 -2
  - Norway: 37 0 36   20 19 1
  - Finland: 35 0 35   21 20 1
  - Slovenia: 29 0 29   22 22 0
  - France: 27 0 27   23 23 0
  - Germany: 23 0 23   24 24 0
  - Turkey: 21 0 21   25 25 0
  - Italy: 19 0 19   26 26 0
  - United States: 17 0 17   27 27 0
  - Korea: 12 0 12   28 28 0
  - Greece: 12 0 12   29 29 0
  - Japan: 4 0 4   30 30 0
- Source and note (as presented):
  - Source: IMF (World Economic Outlook Database) and OECD (BMD4 Partner Country and Main Aggregates Databases).
  - Note: Includes all countries as reported as of end-2015 to the OECD for publication by January 2017. Some OECD countries may have reported zero when no information about SPEs was available.

### ANNEX II: CONSTRUCTION OF THE NEW GLOBAL FDI NETWORK

- Overview: The new global FDI estimates are developed in sequential steps (data input, SPE adjustment model, construction of non-SPE global estimates, UIE and round-tripping adjustments, and final construction of FDI excluding SPEs and broken down by UIE).

- Step 1. Data input (from latest available year, mostly 2015)
  - Inward FDI (CDIS): Inward FDI positions broken down by immediate counterpart economy for all 116 reporting CDIS countries.
  - Aggregate SPE/non-SPE breakdown (OECD BMD4 Partner Country and Main Aggregates Databases): Inward total FDI positions split by resident SPEs and non-SPEs for the 30 reporting OECD countries.
  - Nominal GDP (IMF World Economic Outlook Database): For all CDIS countries.
  - Detailed non-SPE breakdown (OECD BMD4 Partner Country Database): Inward FDI positions for resident non-SPEs broken down by the immediate counterpart economy for the 25 reporting OECD countries.
  - Detailed UIE breakdown (OECD BMD4 Partner Country Database): Inward FDI positions for resident non-SPEs broken down by the UIE for the 12 reporting OECD countries.

- Step 2. Estimation of SPE adjustment model
  - Purpose: To exclude SPEs for non-reporting economies by estimating an SPE adjustment model based on the split between total SPE/non-SPE data for the 30 reporting OECD countries.
  - Specification: Univariate regression with the non-SPE inward FDI-to-total inward FDI ratio as the dependent variable and the total inward FDI-to-GDP ratio as the explanatory variable.
  - Transformation: A log-log transformation is made to achieve linearity (Figure 5).

- Step 3. Construction of global FDI estimates for non-SPEs
  - For the 25 OECD countries reporting non-SPE data by immediate counterpart economy: use the data directly.
  - For the five OECD countries that only split between total SPE/non-SPE data: use aggregate data to calculate an SPE adjustment factor for each economy and apply it to all its counterpart economy positions.
  - For remaining economies that do not split between SPEs and non-SPEs: use CDIS data with an economy-specific SPE adjustment factor based on the model from step 2.
  - Assumption: For a given economy, the same SPE adjustment factor is applied to all its counterpart economies, equivalent to assuming that investors from each counterpart economy use SPEs proportional to their total FDI in the economy.

- Step 4. Calculation of UIE adjustment factors
  - Two types of adjustment factors are calculated to move from immediate counterpart economy breakdown to UIE and to adjust for round-tripping.
  - (i) UIE adjustment factor calculation (based on the 12 UIE-reporting OECD countries):
    - Formula (as presented): a = ∑ FDI(UIE) / ∑ FDI(IIE) for ci≠, where a denotes the UIE adjustment factor, c the counterpart economy, i the reporting economy, n the number of reporting economies, FDI(UIE) the inward FDI position according to the UIE breakdown, and FDI(IIE) the inward position according to the immediate investing economy breakdown.
    - Caps to avoid extreme adjustments: adjustment factors are capped between 0.33 and 3.
    - For economies on the list of low-tax economies (likely to have a significant SPE presence): adjustment factors are capped between 0.20 and 1 so that FDI from these economies can never be higher under the UIE breakdown than under the immediate counterpart economy breakdown.
  - (ii) Round-tripping adjustment factor, b:
    - Calculated as a simple average across reporting economies using the formula (as presented): b = (1/n) ∑ (FDI(UIE)_ci / totalFDI_i) for ci = .
    - Implementation note: round-tripping is effectively allocating 5 percent of the inward FDI back to its own economy (applied later in construction).

- Step 5. Construction of the new global FDI estimates
  - The new estimates contain FDI excluding SPEs and broken down by the UIE.
  - For the 12 UIE-reporting OECD economies: data are used directly.
  - For remaining economies: data from the global estimates for non-SPEs (step 3) are adjusted to UIE by applying the UIE adjustment factor, a_c, to each inward position broken down by the immediate counterpart economy.
  - Round-tripping: calculated for each economy by applying the adjustment factor, b, to total non-SPE inward FDI, effectively allocating 5 percent of the inward FDI back to its own economy.
  - Constraint: Adjustments are constrained so that on an economy level, adjusted FDI data cannot exceed reported total inward FDI.

*Source: wp17258 - ANNEX I: SPE BREAKDOWN (IMF).*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2017/wp17258.pdf_