## wp17261

## Source details

**Canonical URL:** [wp17261](https://www.imf.org/-/media/files/publications/wp/2017/wp17261.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2017/wp17261.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2017/wp17261.pdf.json)

---

### I. Introduction / Scope
- CDIS: worldwide IMF-led statistical data collection to improve availability and quality of direct investment data by counterpart economy.
- CDIS database: detailed annual data on inward direct investment positions (since end-2009) cross-classified by economy of immediate investor, and on outward direct investment positions cross-classified by economy of immediate investment.
- Mirror data (counterpart reporting) are used to highlight data gaps/errors and guide follow-up.
- STA projects to address large bilateral asymmetries:
  - May 2016 project based on the 2015 CDIS release for end-2014 CDIS data contacting 47 CDIS participants.
  - 2013 project based on the 2012 CDIS release covered 28 CDIS participants.
- Paper objectives (as presented): (i) examine global asymmetries; (ii) examine asymmetries between CDIS reported and derived data for individual economies; (iii) analyze bilateral economy-level data comparing reporting economies with counterparts; present lessons and proposed actions.

### II. CDIS results for the total world — key findings
- Instrument concentration:
  - Equity represents around 80 percent of total direct investment.
  - Equity concentration highlights importance of consistent valuation methods for unlisted equity to reduce bilateral asymmetries.
- Total outward vs total inward direct investment:
  - In principle totals should match, but differences occur because:
    - not all economies participate in CDIS;
    - not all CDIS reporting economies provide outward data;
    - bilateral asymmetries between reporting economies.
  - Number of reporting economies generally increases over time and is always larger for inward than for outward.
  - 2013: maximum participation observed (76 outward; 110 inward).
- Instrument-level patterns (preserve source wording):
  - "for outward, the weight of equity ranges from 84 to 89 percent, and for outward it ranges from 77 to 81 percent".
  - Global asymmetries in net equity are always positive: total outward-equity > total inward-equity for the seven years covered by the CDIS.
  - Global asymmetries in net debt instruments are always negative: total outward-debt < total inward-debt for the seven years covered by the CDIS.
  - Because inward reporters > outward reporters, asymmetries cannot be explained by missing reporters alone; likely due to different valuation methods by compilers.
  - Implication: investment in equity abroad (outward equity) tends to be overestimated; investment in debt abroad (outward debt instruments) tends to be underestimated.
  - Differences at instrument level are larger than total asymmetries because positive differences in equity are compensated by negative differences in debt.
- IIP vs CDIS:
  - CDIS differs from IIP because IIP coverage (more economies), IIP uses gross assets/liabilities while CDIS presents net (directional principle), and IIP allows six valuation methods for unlisted equity while CDIS requires unlisted equity at own funds at book value.
  - IIP results consistent with CDIS: equity ≈ 72 percent of assets and 65 percent of liabilities in the IIP; equity asymmetries positive; debt asymmetries negative.

### III. Reported vs derived data for individual economies — overview and top asymmetries
- Annex I Purpose: identify ten economies with largest asymmetries between reported and derived data for inward and outward DI, analyze selected economies, identify reasons for asymmetries.
- Causes of asymmetries: missing data by counterparts (nonparticipation, confidentiality, below threshold), bilateral asymmetries; mirror data useful to cross-check and guide bilateral analysis.
- Interpretation rule: reported < derived may suggest undercoverage; bilateral-level analysis recommended.

- Top Ten Differences — Inward (2015, US Dollars, millions; columns: Direct Investment Inward (reported); DI Inward Derived; Inward − Inward Derived; %)
  - China, P.R.: Mainland — 2,579,564; 1,149,305; 1,430,259; 55%
  - Netherlands — 3,939,415; 2,567,917; 1,371,498; 35%
  - China, P.R.: Hong Kong — 1,394,580; 199,413; 1,195,167; 86%
  - Luxembourg — 3,271,769; 2,429,298; 842,471; 26%
  - Singapore — 869,572; 577,405; 292,167; 34%
  - Mexico — 509,292; 279,969; 229,324; 45%
  - Mauritius — 269,105; 72,144; 196,961; 73%
  - Belgium — 455,366; 662,433; -207,067; -45%
  - United States — 3,134,199; 3,886,610; -752,411; -24%
  - United Kingdom — 1,554,303; 2,736,859; -1,182,557; -76%
- Illustrative note: China, P.R. Mainland reported inward US$2,580 billion at end-2015 while counterparts reported outward to Mainland of US$1,149 billion (44 percent of the inward reported).

- Top Ten Differences — Outward (2015, US Dollars, millions; columns: Direct Investment Outward (reported); DI Outward Derived; Outward − Outward Derived; %)
  - Netherlands — 4,680,387; 2,863,149; 1,817,238; 39%
  - Luxembourg — 3,829,107; 2,439,494; 1,389,613; 36%
  - United States — 5,040,648; 4,058,409; 982,239; 19%
  - Germany — 1,376,181; 1,131,120; 245,061; 18%
  - Ireland — 887,510; 645,808; 241,702; 27%
  - Australia — 396,432; 161,945; 234,487; 59%
  - Russian Federation — 286,583; 102,821; 183,762; 64%
  - Italy — 467,300; 306,831; 160,469; 34%
  - Belgium — 446,237; 616,707; -170,471; -38%
  - United Kingdom — 1,563,865; 2,226,308; -662,443; -42%

### IV. Analysis of selected economies — key diagnostic findings
- China, P.R.: Mainland inward difference drivers:
  - 44 percent of the inward difference explained by China, P.R. Hong Kong (Hong Kong reports outward to Mainland US$608 billion vs Mainland inward US$1,239 billion).
  - 41 percent explained by main counterparts not reporting outward (British Virgin Islands US$328 billion; Singapore US$111 billion — 23% and 8% of total difference).
  - Small opposite-direction differences account for small shares (e.g., South Africa 4%).
- Netherlands inward:
  - Differences spread across many counterparts: United Kingdom US$189 billion; Luxembourg US$187 billion; Switzerland US$134 billion (14%, 14%, 10%).
  - 29% explained by missing outward reporters (Bermuda US$291 billion; Cayman Islands US$49 billion).
- China, P.R.: Hong Kong inward:
  - Around 95% of inward difference explained by main counterparts not reporting outward data (British Virgin Islands US$552 billion; China, P.R.: Mainland US$349 billion; Cayman Islands US$85 billion; Bermuda US$74 billion).
- United Kingdom (opposite sign for inward):
  - Reported inward much lower than derived inward; main negative differences: Luxembourg -US$593 billion (50% of asymmetry), Netherlands -US$263 billion (22%), United States -US$160 billion (13%).
  - Missing-counterpart contributions smaller shares (Bermuda, Cayman Islands, British Virgin Islands).

- Outward examples:
  - Netherlands outward:
    - 74% of outward difference explained by differing figures from counterparts; 23% by missing reporters.
    - Main counterpart differences: United Kingdom US$263 billion; United States US$232 billion; Luxembourg US$188 billion (14%, 13%, 10%).
    - Missing inward reporters: Bermuda US$165 billion; Puerto Rico US$109 billion.
  - Luxembourg outward:
    - 49% explained by United States reporting lower mirror inward (US reports inward from Luxembourg US$328 billion vs Luxembourg outward to US US$1,011 billion).
    - United Kingdom accounts for 43% of difference (UK inward from Luxembourg US$105 billion vs Luxembourg outward to UK US$698 billion).
    - 12% explained by nonreporting counterparts (Gibraltar, Bermuda ~5% each).
  - United States outward:
    - Missing inward reporters (Bermuda and Cayman Islands) represent 27 and 26 of the difference respectively.
    - Counterpart lower figures explain: United Kingdom US$160 billion (16%), Canada and Netherlands 7% each.
    - Opposite-sign where counterparts report higher mirror data: Luxembourg -US$233 billion (24%), Mexico -US$130 billion (13%).
  - United Kingdom outward (opposite sign):
    - Reported outward much lower than derived outward.
    - Main negative differences: Luxembourg -US$347 billion (52%), Netherlands -US$189 billion (28%), United States -US$130 billion (20%).
    - Opposite-direction contributions from nonreporting counterparts: Jersey, Cayman Islands, British Virgin Islands (4%, 3%, 2%).

### V. Bilateral comparisons and illustrative bilateral asymmetries
- Smallest bilateral inward asymmetries (end-2015, US Dollars, millions; inward reported; mirror inward; diff; %)
  - United States — Japan: 411,201; 413,194; 1,993; 0%
  - United States — Germany: 255,471; 254,286; 1,185; 0%
  - Belgium — Luxembourg: 160,004; 155,240; 4,763; 3%
  - South Africa — Netherlands: 37,476; 36,626; 850; 2%
  - Spain — Italy: 37,452; 37,722; 270; 1%
  - Poland — Germany: 29,889; 30,196; 307; 1%
  - Norway — Sweden: 26,274; 26,335; 60; 0%
  - Canada — United Kingdom: 24,759; 25,536; 777; 3%
  - Austria — Russian Federation: 20,976; 21,054; 79; 0%
  - Turkey — Netherlands: 18,854; 18,798; 56; 0%
  - (additional small pairs listed in source)
- Largest bilateral asymmetries (top twenty, end-2015, US Dollars, millions; selection)
  - United States — Luxembourg: 328,400; 1,011,038; 682,638; 208%
  - China, P.R.: Mainland — China, P.R.: Hong Kong: 1,238,823; 607,960; 630,863; 51%
  - United Kingdom — Luxembourg: 104,733; 698,015; 593,281; 566%
  - Luxembourg — United Kingdom: 498,010; 150,782; 347,228; 70%
  - United Kingdom — Netherlands: 231,565; 494,631; 263,067; 114%
  - Luxembourg — United States: 736,318; 502,998; 233,320; 32%
  - United States — Netherlands: 282,525; 514,344; 231,819; 82%
  - (other large bilateral entries listed in source, including United States — Ireland: 13,455; 109,186; 95,731; 711%)

### VI. Main reasons for bilateral asymmetries (Section V)
- Root causes often at national level: inadequate/partial data sources, lack of information, noncompliance with recommended guidelines.
- Seven main causes for CDIS bilateral asymmetries:
  - (a) recording positions between fellow enterprises when UCP is nonresident (extended directional principle effects);
  - (b) different coverage and lack or partial coverage (SPEs, fellow enterprises, nonfinancial private sector, real estate);
  - (c) different geographic allocation arising from immediate vs ultimate counterpart economy;
  - (d) unallocated or confidential data (“Not specified (including confidential)”);
  - (e) methodological differences in valuation for listed/unlisted equity and debt instruments and identification/exclusion of debt between affiliated financial corporations;
  - (f) use of different data sources and estimation techniques;
  - (g) criteria for identifying entities in a direct investment relationship.

- Asymmetries from fellow-enterprise recording (detailed):
  - When UCP is nonresident for both fellows and extended directional principle applied, both economies may record assets/liabilities between fellows as inward (one with negative sign), producing equal-size asymmetries of opposite sign; remedy is to compare total net DI positions (outward minus inward) between the two economies hosting fellows to eliminate that asymmetry.
  - Numerical illustration: example loan $80 between fellows shown as Inward = - $80 and Inward = $80 in counterpart economies (resulting bilateral mismatch in inward/outward components).

- Netting total direct investment to reduce asymmetries (Tables 5-A and 5-B):
  - Method computes Total Diff = (Outward_A_with_B − Inward_B_with_A) + (Outward_B_with_A − Inward_A_with_B).
  - Examples where netting reduces large bilateral asymmetries:
    - Luxembourg — Netherlands:
      - Outward reported by Luxembourg in Netherlands: 534,763; Inward reported by Netherlands from Luxembourg: 721,968; Diff (A): -187,205
      - Inward reported by Luxembourg from Netherlands: 494,221; Outward reported by Netherlands in Luxembourg: 682,642; Diff (B): 188,421
      - Total Diff (A)+(B): 1,217
    - Brazil — Netherlands:
      - Outward reported by Brazil in Netherlands: -21,417; Inward reported by Netherlands from Brazil: 63,964; Diff (A): -85,381
      - Inward reported by Brazil from Netherlands: 110,210; Outward reported by Netherlands in Brazil: 199,106; Diff (B): 88,896
      - Total Diff (A)+(B): 3,515
  - Note: to apply total DI comparison, countries must report both inward and outward; some CDIS reporters do not provide outward data.

### VII. Coverage issues: SPEs, fellows, nonfinancial corporations, real estate
- Special Purpose Entities (SPEs):
  - SPEs should be included in CDIS (see Annex IV).
  - CDIS metadata shows uneven coverage:
    - Fully covered examples: Netherlands.
    - Partial coverage: Luxembourg, Mauritius.
    - Not covered: Seychelles, Curacao.
    - Negligible/not relevant: Italy, South Africa.
  - Some economies report SPEs under “Not Specified (including Confidential)” with no counterpart detail → significant bilateral asymmetries though little aggregate impact.
  - OECD split by resident SPEs (Table 6, 2015, US Dollars, millions) shows high SPE shares in some economies (e.g., Luxembourg Inward resident SPEs 3,044,784 of 3,271,864 = 93%; Netherlands Inward resident SPEs 3,239,357 of 3,956,821 = 82%).
- Nonresident SPEs:
  - Investments concentrated in few economies; top 15 non-reporters account for 89% and 94% of total non-reported data for inward and outward respectively.
  - Top non-reporters include Virgin Islands, British; Bermuda; Cayman Islands; Gibraltar; Puerto Rico.
  - Aggregate impact: non-reporters/total reported = 10% (Inward) and 12% (Outward).
- Fellow enterprises:
  - Around 85 percent of CDIS reporters include positions with fellow enterprises in totals; nearly half provide separate data.
  - 50 percent record fellows using extended directional principle; ~30 percent apply assets liabilities principle.
  - Importance varies: inward fellows represent 40% of total inward in Belgium, 0.6% in Switzerland.
- Nonfinancial private corporations:
  - Low survey participation and lack of legal mandate impede coverage; many compilers use samples.
  - Census used by 33% for inward and 20% for outward.
  - Data collection improvements (online forms, meetings, company visits) recommended.
- Real estate:
  - Around 40% of CDIS reporters include real estate in direct investment.
  - Real estate typically involves household sector and difficult to survey; compilers use public registers.
  - Counterpart allocation may be incorrect due to transit countries.

### VIII. Geographic allocation: immediate vs ultimate investing economy
- CDIS guideline: report by immediate counterpart economy.
- Some countries allocate inward positions to ultimate investing economy instead; looking-through nonresident SPEs is rare.
- Differences in allocation produce bilateral asymmetries at country level but not at global level.
- Ultimate-investor breakdowns are useful and some economies compile them as additional series.

### IX. Unallocated/confidential data
- Data with unidentified/confidential counterpart reported under “Not specified (including confidential)”.
  - Global weights: ≈ 6% for inward and ≈ 7% for outward; decreased in last two years (Figure 8 referenced).
  - 16 economies report > 15% of inward as not specified; 10 economies > 15% for outward.
- More granular breakdowns increase confidentiality risk and hence not-allocated reporting.

### X. Methodological issues — valuation and instruments
- Equity valuation:
  - CDIS Guide requires unlisted equity valued at own funds at book value (OFBV); listed equity can be valued at market prices.
  - OFBV defined as sum of (a) paid-up capital; (b) all types of reserves identified as equity; (c) cumulated reinvested earnings; (d) holding gains or losses included in own funds.
  - Frequent revaluation of books yields closer proxy to market values; lack of revaluation can cause large bilateral asymmetries.
  - Metadata shows most reporting economies follow listed=market, unlisted=OFBV; however a significant number use OFBV for listed equity—could produce asymmetries when revaluations are not frequent.
  - Other methods in use include historic cost (BPM6 para 3.88(f)).
- Box 1 — United States case (historic cost):
  - Table 8 (Aggregated values, US dollars, billions) — Inward Equity Positions (Net) reported by the United States:
    - 2011: 1,550
    - 2012: 1,728
    - 2013: 1,847
    - 2014: 1,933
    - 2015: 2,087
  - Outward Equity Positions (Net) reported by counterparts:
    - 2011: 2,120
    - 2012: 2,343
    - 2013: 2,763
    - 2014: 2,834
    - 2015: 3,131
  - Difference (United States minus Counterpart Economies):
    - 2011: -570
    - 2012: -615
    - 2013: -916
    - 2014: -901
    - 2015: -1,044
  - Observation: aggregate asymmetries always negative (US reports smaller amounts than counterparts for selected 22 economies).
  - Compiler action recommendation: compilers should try to estimate market or fair value of direct investment equity.
- Debt valuation:
  - Debt instruments classified as debt securities (market prices) and other debt instruments (nominal value).
  - Metadata: >40 percent of reporters follow international standards (debt securities at market value; other debt instruments at nominal).
  - In 2015, 80 percent of inward debt instrument values reported by economies are larger than corresponding outward by counterparts.
  - Potential reasons: overestimated inward data and/or underestimated outward data due to coverage differences; compilers have better access to inward data; intercompany debt compilation issues (not tracking repayments) may lead to underestimation of outward debt.
- Debt between affiliated financial corporations:
  - Should be excluded from direct investment and included in portfolio/other investment; many reporters exclude this debt; asymmetries arise when counterparts cannot exclude due to data source limitations.

### XI. Data sources, estimation, and identification criteria
- Primary sources: enterprise surveys and financial statements.
- Around 70 percent of reporting economies apply estimates to fill nonresponse gaps; most compilers report high survey response rates.
- Some use accumulation of transactions to estimate positions (few), which requires adjustments for price and exchange rate changes to avoid asymmetries.
- Identification of direct investment relationships:
  - FDIR recommended; some economies use simplified methods that can misclassify DI as portfolio/other investment.
  - >50 percent of CDIS reporters apply FDIR for both inward and outward.

### XII. Addressing bilateral asymmetries — recommended approaches
- Compliance with recommended guidelines:
  - Uniform standards in coverage, geographic counterparty allocation (immediate economy), valuation, and identification criteria.
  - Coverage: include resident SPEs, positions between fellow enterprises, direct investment involving nonfinancial corporations, and real estate.
  - Valuation: market value for listed equity and debt securities; OFBV for unlisted/other equity; nominal value for debt instruments other than debt securities.
  - Exclude debt between selected affiliated financial corporations where possible.
  - Use position (stock) source data; avoid accumulation of transactions unless adjusted for price/exchange rate changes.
  - Apply appropriate statistical techniques for survey frames and estimation (grossing up, thresholds, non-response).
  - Evolve from simplified identification methods to FDIR.
- Reduce confidential/unallocated reporting:
  - Effort to allocate investments to specific counterpart economies and minimize “Not specified (including confidential)”.
  - Revisit and revise past time series as data systems improve and delayed information is received.
  - Use past allocations, transaction data, mirror information for counterpart allocation estimates.
- Data and metadata exchange:
  - Bilateral micro-level data exchange between compilers is most efficient to identify causes (subject to confidentiality/legal constraints).
  - Aggregated data and metadata exchange useful and pose fewer confidentiality issues.
  - CDIS metadata questionnaire is valuable for assessing practices and deviations.
  - Existing mechanisms: FDI Network (Eurostat/ECB), bilateral exchanges.
  - Bilateral microdata exchange considered good practice even if not all asymmetries can be reconciled.

### XIII. Institutional initiatives and planned actions
- STA consulted 47 participants in 2016 on plans to reduce asymmetries. Planned actions include:
  - continued aggregated-level analysis;
  - more active bilateral consultations and interaction with counterparts;
  - more active participation in the FDI Network (European countries);
  - development of European Group Register;
  - revisions to CDIS data;
  - enhancements in reporting systems and data sources.
- International initiatives:
  - OECD bilateral asymmetry meetings (WGIIS).
  - Eurostat facilitates bilateral asymmetry platforms.
  - IMF (with Bundesbank) to conduct a workshop for CDIS compilers in November 2017.
  - Follow-up on possibility of expanding FDI Network infrastructure world-wide.
  - Planned enhancements to CDIS metadata questionnaire to add specific bilateral-asymmetry questions and encourage metadata updates.
- Recognition: steps to increase awareness of asymmetries and understand causes are important achievements even if reduction is not always feasible.

*Source: wp17261 (Annexes and material extracted from IMF’s CDIS, 2016 release)*

### References .............................................................................................................

### wp17261 - References .............................................................................................................

### I. INTRODUCTION
- The CDIS is a worldwide statistical data collection effort led by the IMF to improve availability and quality of data on direct investment, both at overall level in the international investment position (IIP) and by immediate counterpart economy.
- CDIS database presents detailed annual data on inward direct investment positions starting from data since end-2009 cross-classified by economy of immediate investor, and on outward direct investment positions cross-classified by economy of immediate investment.
- Mirror data (obtained from counterpart reporting economies) can highlight data gaps or errors and guide follow-up efforts.
- IMF Statistics Department (STA) conducted projects to raise awareness of and address large bilateral asymmetries:
  - May 2016 project based on the 2015 CDIS release for end-2014 CDIS data, contacting 47 CDIS participants.
  - A 2013 project based on the 2012 CDIS release covered 28 CDIS participants.
- The paper analyzes asymmetries in direct investment positions reported in CDIS using a top-down approach and (i) examines global asymmetries; (ii) examines asymmetries between CDIS reported and derived data for individual economies; and (iii) analyzes bilateral economy-level data comparing reporting economies with counterparts. Lessons and proposed actions to reduce bilateral asymmetries are presented in later sections.

### II. CDIS RESULTS FOR THE TOTAL WORLD
- Direct investment positions are concentrated in equity, representing around 80 percent of the total direct investment.
- The concentration in equity highlights the importance of applying consistent valuation methods for unlisted equity to reduce bilateral asymmetries.

A. Total Outward Direct Investment vs. Total Inward Direct Investment
- In principle, total outward direct investment positions should equal total inward direct investment positions, but differences occur because:
  - not all economies participate in the CDIS;
  - not all CDIS reporting economies provide outward data;
  - there are bilateral asymmetries between reporting economies.
- Reporting counts and a key observation:
  - In most cases, the number of reporting economies increases over the years and is always larger for inward than for outward.
  - The global difference is particularly small in 2013, when the maximum number of participants is reached (76 for outward and 110 for inward).
- Instrument-level findings (as presented in Figure 3):
  - For all periods, the weight of the net equity component in total direct investment is much higher than the weight of net debt instruments.
  - Exact wording preserved from source: "for outward, the weight of equity ranges from 84 to 89 percent, and for outward it ranges from 77 to 81 percent".
  - Global asymmetries in net equity are always positive: total outward-equity > total inward-equity for the seven years covered by the CDIS.
  - Global asymmetries in net debt instruments are always negative: total outward-debt < total inward-debt for the seven years covered by the CDIS.
  - Because the number of inward reporters is larger than outward reporters, these asymmetries cannot be explained by missing reporters alone; differences are likely due to different valuation methods applied by compilers.
  - Implication: investment in equity abroad (outward equity) tends to be overestimated; investment in debt instruments abroad (outward debt instruments) tends to be underestimated.
  - Differences at instrument level are larger than total asymmetries because positive differences in equity are compensated by negative differences in debt.

B. Total Direct Investment Assets vs. Total Direct Investment Liabilities in the IIP
- Values reported in the CDIS differ from values reported in the direct investment component of the IIP due to:
  - the number of economies reporting IIP data to STA is larger than those reporting CDIS data;
  - CDIS presents direct investment on a net basis (directional principle) while IIP presents gross assets/liabilities;
  - IIP allows six different valuation methods for unlisted equity (as proxies for market value, para. 7.16 of BPM6), whereas CDIS requires unlisted equity to be valued at own funds at book value.
- The number of economies compiling IIP data is larger than those participating in the CDIS because IIP data are compiled on an aggregated basis while CDIS requires counterpart-economy breakdowns.
- IIP results (Figures 4 and 6) are consistent with CDIS findings:
  - Equity represents approximately 72 percent of assets and 65 percent of liabilities in the IIP.
  - Asymmetries for equity are always positive; asymmetries for debt instruments are always negative.

### III. CDIS REPORTED VS CDIS DERIVED DATA FOR INDIVIDUAL ECONOMIES
- This section focuses on asymmetries between reported and derived data for individual economies and discusses the usefulness and limitations of derived measures of direct investment.
- (Text continues beyond supplied excerpt detailing methods, comparisons, and implications for individual economies.)

*Source: wp17261 - References*

### Annex I).

### Annex I)

### Overview
- Purpose: identify the ten economies with the largest asymmetries between reported and derived data for inward and outward direct investment, analyze selected economies, and identify reasons for asymmetries.
- Key points:
  - Asymmetries can be spread among several counterparts or concentrated in a few.
  - Causes include missing data by counterparts (nonparticipation in CDIS, confidentiality, below threshold) and bilateral asymmetries.
  - Mirror data are useful to cross check and verify countries’ estimates, highlight data gaps/errors, and identify areas for improvement.
  - If reported data by a given economy are lower than derived data, it can suggest under coverage; bilateral-level analysis is recommended.

### Top Ten Differences — Inward (Table 1)
- Source: IMF’s CDIS, 2016 release. 2015, US Dollars, millions. Columns: Direct Investment Inward (reported); DI Inward Derived (reported by counterparts); Inward-Inward Derived; %
- Top ten inward differences:
  - China, P.R.: Mainland — 2,579,564; 1,149,305; 1,430,259; 55%
  - Netherlands — 3,939,415; 2,567,917; 1,371,498; 35%
  - China, P.R.: Hong Kong — 1,394,580; 199,413; 1,195,167; 86%
  - Luxembourg — 3,271,769; 2,429,298; 842,471; 26%
  - Singapore — 869,572; 577,405; 292,167; 34%
  - Mexico — 509,292; 279,969; 229,324; 45%
  - Mauritius — 269,105; 72,144; 196,961; 73%
  - Belgium — 455,366; 662,433; -207,067; -45%
  - United States — 3,134,199; 3,886,610; -752,411; -24%
  - United Kingdom — 1,554,303; 2,736,859; -1,182,557; -76%
- Illustrative interpretation:
  - China, P.R.: Mainland reported inward of US$2,580 billion at end-2015 while counterparts reported outward in Mainland of US$1,149 billion (44 percent of the inward reported).

### Top Ten Differences — Outward (Table 2)
- Source: IMF’s CDIS, 2016 release. 2015, US Dollars, millions. Columns: Direct Investment Outward (reported); DI Outward Derived (reported by counterparts); Outward-Outward Derived; %
- Top ten outward differences:
  - Netherlands — 4,680,387; 2,863,149; 1,817,238; 39%
  - Luxembourg — 3,829,107; 2,439,494; 1,389,613; 36%
  - United States — 5,040,648; 4,058,409; 982,239; 19%
  - Germany — 1,376,181; 1,131,120; 245,061; 18%
  - Ireland — 887,510; 645,808; 241,702; 27%
  - Australia — 396,432; 161,945; 234,487; 59%
  - Russian Federation — 286,583; 102,821; 183,762; 64%
  - Italy — 467,300; 306,831; 160,469; 34%
  - Belgium — 446,237; 616,707; -170,471; -38%
  - United Kingdom — 1,563,865; 2,226,308; -662,443; -42%

### Analysis of Selected Economies — Inward Differences (selected highlights)
- China, P.R.: Mainland:
  - 44 percent of the inward difference explained by China, P.R. Hong Kong (Hong Kong reports outward to Mainland of US$608 billion versus Mainland’s inward of US$1,239 billion).
  - 41 percent explained by main counterparts not reporting outward data (British Virgin Islands US$328 billion; Singapore US$111 billion — representing 23 and 8 percent of total difference respectively).
  - Some smaller opposite-direction differences (e.g., South Africa) representing 4 percent of total difference in absolute terms.
- Netherlands:
  - Differences spread among many countries: United Kingdom US$189 billion, Luxembourg US$187 billion, Switzerland US$134 billion (14, 14, and 10 percent of total difference respectively).
  - 29 percent of difference explained by main counterparts not reporting outward data (Bermuda US$291 billion; Cayman Islands US$49 billion).
  - Opposite-direction difference from United States representing 5 percent of total difference in absolute terms.
- China, P.R.: Hong Kong:
  - Around 95 percent of the inward difference explained because main counterparts do not report outward data (British Virgin Islands US$552 billion; China, P.R.: Mainland US$349 billion; Cayman Islands US$85 billion; Bermuda US$74 billion).
- United Kingdom (opposite sign):
  - Reported inward much lower than derived inward.
  - Main negative differences: Luxembourg -US$593 billion (50 percent of asymmetry), Netherlands -US$263 billion (22 percent), United States -US$160 billion (13 percent).
  - Opposite-direction contributions from nonreporting counterparts: Bermuda, Cayman Islands, British Virgin Islands representing 6, 3, and 2 percent of total difference in absolute terms.

### Analysis of Selected Economies — Outward Differences (selected highlights)
- Netherlands:
  - 74 percent of outward difference explained by differing figures from counterparts; 23 percent explained by missing reporters.
  - Main counterpart differences: United Kingdom US$263 billion, United States US$232 billion, Luxembourg US$188 billion (14, 13, and 10 percent of total).
  - Missing inward reporters: Bermuda US$165 billion; Puerto Rico US$109 billion (9 and 6 percent).
- Luxembourg:
  - 49 percent of outward difference explained by the United States reporting much lower mirror inward (US reports inward from Luxembourg US$328 billion versus Luxembourg’s outward to US US$1,011 billion).
  - United Kingdom accounts for 43 percent of difference (UK inward from Luxembourg US$105 billion versus Luxembourg outward to UK US$698 billion).
  - Only 12 percent of difference explained by nonreporting counterparts (Gibraltar and Bermuda each ~5 percent).
  - Some opposite-sign differences from counterparts reporting higher mirror data (Netherlands 13 percent; Germany 6 percent).
- United States:
  - Significant part explained by main counterparts not reporting inward data: Bermuda and Cayman Islands represent 27 and 26 of the difference respectively.
  - Counterpart lower figures also explain difference: United Kingdom US$160 billion (16 percent), Canada and Netherlands 7 percent each.
  - Opposite-sign cases where counterparts report higher mirror data: Luxembourg -US$233 billion (24 percent), Mexico -US$130 billion (13 percent).
- United Kingdom (opposite sign):
  - Reported outward much lower than derived outward.
  - Main negative differences: Luxembourg -US$347 billion (52 percent), Netherlands -US$189 billion (28 percent), United States -US$130 billion (20 percent).
  - Opposite-direction contributions from counterparts not reporting inward data: Jersey, Cayman Islands, British Virgin Islands representing 4, 3, and 2 percent respectively; also from countries reporting lower mirror data (e.g., Belgium 3 percent).

### CDIS Results by Compiling Economy — Bilateral Comparisons
- Approach: compare data bilaterally for individual economies and their counterparts using inward/outward reported vs counterpart mirror reported.
- Smallest bilateral asymmetries (Table 3): inward differences not exceeding three percent of total (end-2015, US Dollars, millions). Selected entries:
  - United States — Japan: 411,201; 413,194; 1,993; 0%
  - United States — Germany: 255,471; 254,286; 1,185; 0%
  - Belgium — Luxembourg: 160,004; 155,240; 4,763; 3%
  - South Africa — Netherlands: 37,476; 36,626; 850; 2%
  - Spain — Italy: 37,452; 37,722; 270; 1%
  - Poland — Germany: 29,889; 30,196; 307; 1%
  - Norway — Sweden: 26,274; 26,335; 60; 0%
  - Canada — United Kingdom: 24,759; 25,536; 777; 3%
  - Austria — Russian Federation: 20,976; 21,054; 79; 0%
  - Turkey — Netherlands: 18,854; 18,798; 56; 0%
  - Canada — Japan: 15,896; 16,062; 166; 1%
  - Singapore — Germany: 14,559; 14,188; 371; 3%
  - Australia — Germany: 12,592; 12,206; 386; 3%
  - United States — China, P.R.: Hong Kong: 11,102; 11,036; 66; 1%
  - Spain — Portugal: 10,552; 10,374; 179; 2%
  - Russian Federation — Switzerland: 8,552; 8,543; 9; 0%
  - Slovak Republic — Netherlands: 8,536; 8,487; 50; 1%
  - Turkey — Spain: 7,088; 7,032; 56; 1%
  - Turkey — Russian Federation: 6,040; 6,013; 27; 0%
  - Canada — Australia: 6,025; 6,213; 188; 3%
- Largest bilateral asymmetries (Table 4): top twenty bilateral asymmetries as of end-2015 (US Dollars, millions). Selected entries:
  - United States — Luxembourg: 328,400; 1,011,038; 682,638; 208%
  - China, P.R.: Mainland — China, P.R.: Hong Kong: 1,238,823; 607,960; 630,863; 51%
  - United Kingdom — Luxembourg: 104,733; 698,015; 593,281; 566%
  - Luxembourg — United Kingdom: 498,010; 150,782; 347,228; 70%
  - United Kingdom — Netherlands: 231,565; 494,631; 263,067; 114%
  - Luxembourg — United States: 736,318; 502,998; 233,320; 32%
  - United States — Netherlands: 282,525; 514,344; 231,819; 82%
  - Netherlands — United Kingdom: 364,574; 175,967; 188,607; 52%
  - Luxembourg — Netherlands: 494,221; 682,642; 188,421; 38%
  - Netherlands — Luxembourg: 721,968; 534,763; 187,205; 26%
  - United Kingdom — United States: 432,987; 593,028; 160,041; 37%
  - Luxembourg — Canada: 177,920; 36,252; 141,668; 80%
  - Netherlands — Switzerland: 256,832; 122,829; 134,003; 52%
  - United States — United Kingdom: 483,841; 353,390; 130,451; 27%
  - Mexico — United States: 222,476; 92,812; 129,664; 58%
  - Ireland — Luxembourg: 151,641; 264,261; 112,620; 74%
  - Switzerland — Luxembourg: 207,398; 316,342; 108,944; 53%
  - Switzerland — Netherlands: 183,523; 289,527; 106,005; 58%
  - United States — Ireland: 13,455; 109,186; 95,731; 711%
  - Brazil — Netherlands: 110,210; 199,106; 88,896; 81%

### Main Reasons for Bilateral Asymmetries (Section V)
- Root causes often at national level: inadequate/partial data sources, lack of information, noncompliance with recommended guidelines.
- Bilateral asymmetries mainly due to:
  - Limitations in compilation process.
  - Use of methodologies deviating from standards.
  - In some cases (especially fellows) asymmetries can arise even when following international methodological standards.
- Seven main reasons for CDIS bilateral asymmetries:
  - (a) way of recording positions between fellow enterprises when the ultimate controlling parent (UCP) is nonresident;
  - (b) different coverage and lack or partial coverage;
  - (c) different geographic allocation arising from the use of the immediate vs ultimate counterpart economy;
  - (d) unallocated or confidential data;
  - (e) methodological differences in the valuation for listed and unlisted equity and debt instrument or the identification and exclusion of debt between affiliated financial corporations;
  - (f) use of different data sources and estimation techniques;
  - (g) criteria for identifying entities in a direct investment relationship.

### Asymmetries from Recording Positions Between Fellow Enterprises (detailed)
- Cause: treatment of direct investment positions between fellow enterprises when UCP is nonresident for both fellows; arises under the extended directional principle.
- Effect: both economies record assets and liabilities between the fellows as inward (one with a negative sign), producing equal-size asymmetries of opposite sign when comparing inward vs outward across counterparts.
- Remedy: compare total net direct investment positions (outward minus inward) between the two economies hosting the fellows; this eliminates the asymmetry caused by fellow recording.
- Example scenario (textual):
  - Fellow enterprise A resident in economy A; fellow enterprise B resident in economy B; both apply extended directional principle.
  - If UCP is resident in a third economy C (nonresident for both fellows):
    - Economy A records assets and liabilities between fellows A and B in inward (because UCP is nonresident).
    - Economy B records assets and liabilities between fellows A and B in inward (because UCP is nonresident), and with opposite sign.
  - Result: bilateral asymmetry between inward reported by one economy and outward reported by counterpart (both record under inward).
- Numerical illustration in text:
  - Example loan $80 between fellows shown as Inward = - $80 and Inward = $80 in counterpart economies.

### Netting Total Direct Investment to Reduce Asymmetries (Tables 5-A and 5-B)
- Method: compare total net direct investment positions using outward and inward reported by each economy with the counterpart and compute Total Diff = (Outward reported by A with B − Inward reported by B with A) + (Outward reported by B with A − Inward reported by A with B).
- Table 5-A (US Dollars, millions). Two examples where netting reduces large bilateral asymmetries:
  - Luxembourg — Netherlands:
    - Outward reported by Luxembourg in Netherlands: 534,763
    - Inward reported by Netherlands from Luxembourg: 721,968
    - Diff. = Outward − Inward (A): -187,205
    - Inward reported by Luxembourg from Netherlands: 494,221
    - Outward reported by Netherlands in Luxembourg: 682,642
    - Diff. = Outward − Inward (B): 188,421
    - Total Diff. (A) + (B): 1,217
  - Brazil — Netherlands:
    - Outward reported by Brazil in Netherlands: -21,417
    - Inward reported by Netherlands from Brazil: 63,964
    - Diff. (A): -85,381
    - Inward reported by Brazil from Netherlands: 110,210
    - Outward reported by Netherlands in Brazil: 199,106
    - Diff. (B): 88,896
    - Total Diff. (A) + (B): 3,515
- Table 5-B presents same information as total DI (net) by reporting economy and counterpart:
  - Luxembourg — Netherlands:
    - Luxembourg: Outward 534,763; Inward 494,221; Total DI (net) 40,542
    - Netherlands: Outward 682,642; Inward 721,968; Total DI (net) -39,326
    - Diff: 1,217
  - Brazil — Netherlands:
    - Brazil: Outward -21,417; Inward 110,210; Total DI (net) -131,627
    - Netherlands: Outward 199,106; Inward 63,964; Total DI (net) 135,142
    - Diff: 3,515
- Note: to apply total direct investment comparison, countries must report both inward and outward positions; some CDIS reporting economies do not provide outward data.

*Source: IMF’s CDIS, 2016 release (Annex I).*

### Annex 3 includes additional examples and analyzes asymmetries that could arise in one of

### wp17261 - Annex 3 includes additional examples and analyzes asymmetries that could arise in one of

### B. Different Coverage, Lack or Partial Coverage
- Different coverage, lack of coverage, or partial coverage for special purpose entities (SPEs), fellow enterprises, nonfinancial private sector, and real estate may create large asymmetries.
- Capacity to compile data from these institutional units greatly varies across economies, including the systems used to collect them which could imply different coverage.

Special Purpose Entities (SPEs)
- Direct investment positions involving SPEs should be included in the CDIS (see Annex IV).
- CDIS metadata questionnaire reveals SPEs are not evenly covered:
  - Fully covered examples: Netherlands.
  - Partial coverage examples: Luxembourg and Mauritius.
  - Not covered examples: Seychelles and Curacao.
  - Negligible/not relevant examples: Italy and South Africa.
- Some economies include SPEs’ data but cannot allocate SPEs investments to any specific country or consider these investments confidential; such data are reported under “Not Specified (including Confidential)” with no counterpart economy detail.
  - Effect: significant asymmetries at bilateral economy level, no impact at aggregate level.

Resident SPEs in Direct Investment Data
- CDIS does not separately identify data for SPEs; OECD requests separate reporting for resident SPEs.
- Table 6 (OECD Data, 2015) — Total Direct Investment Positions with split by resident SPEs (US Dollar (millions)):
  - Austria: Inward All resident units 262,641; Resident SPEs 91,430; % 35% — Outward All resident units 311,507; Resident SPEs 101,837; % 33%
  - Belgium: Inward 455,379; 28,165; 6% — Outward 446,250; 13,207; 3%
  - Chile: Inward 221,986; 1,252; 1% — Outward 101,791; 2,331; 2%
  - Denmark: Inward 113,888; 13,759; 12% — Outward 183,201; 12,937; 7%
  - Hungary: Inward 197,551; 113,090; 57% — Outward 151,893; 117,033; 77%
  - Iceland: Inward 11,293; 3,442; 30% — Outward 11,079; 3,442; 31%
  - Korea: Inward 169,659; 736; 0% — Outward 276,153; 53; 0%
  - Luxembourg: Inward 3,271,864; 3,044,784; 93% — Outward 3,829,324; 3,622,873; 95%
  - Netherlands: Inward 3,956,821; 3,239,357; 82% — Outward 4,714,195; 3,592,423; 76%
  - Norway: Inward 143,422; 1,771; 1% — Outward 165,330; 981; 1%
  - Poland: Inward 185,177; 1,308; 1% — Outward 23,589; 1,308; 6%
  - Portugal: Inward 118,986; 13,580; 11% — Outward 56,733; 8,961; 16%
  - Spain: Inward 547,265; 38,919; 7% — Outward 490,322; 29,297; 6%
  - Sweden: Inward 302,525; 21,933; 7% — Outward 374,280; 16,892; 5%
  - Switzerland: Inward 839,827; 122,956; 15% — Outward 1,129,768; 104,591; 9%
- Data extracted on 03 Oct 2017 16:17 UTC (GMT) from OECD.Stat (Dataset: FDI positions, main aggregates BMD4).
- Note: zero values (resident SPEs not present or not significant) were excluded.

Nonresident SPEs in Direct Investment Data
- Some countries are main destinations of SPEs (example: Netherlands — pass-through funds); others are characterized by outward investment in SPEs (example: Russia).
  - Russia: no resident SPEs; Russian enterprises use SPEs registered abroad for buyback operations, holding companies, and conduit companies. Popular jurisdictions for SPEs investing in Russia include Netherlands, Bahamas, Bermuda, British Virgin Islands.
  - Russian companies choose off-shore countries for SPE registration aimed at optimization of tax payments (IMF, BOPCOM 16/21).
- Challenges in SPE data collection include lack of statistical capacity, unavailability due to confidentiality, and unwillingness to invest in compilation given limited domestic economic impact (IMF, BOPCOM 16/18).

Lack of Participation in CDIS of Economies Hosting SPEs
- Derived mirror data are a useful source for economies that do not compile direct investment data.
- For 2015, inward and outward derived data for non-reporters represent 10% and 12% of the total reported inward and outward positions respectively.
- Investments concentrated in few economies: top 15 non-reporters represent 89% and 94% of total non-reported data for inward and outward respectively.
- Most significant non-CDIS reporters are economies that host SPEs (British Virgin Islands, Bermuda, Cayman Islands).

Table 7. Top 15 Economies that Did Not Report 2015 Inward/Outward CDIS (US Dollars, millions)
- Inward Derived / Outward Derived, 2015; columns: Total / Equity (Net) / Debt (Net) — sample entries:
  - Virgin Islands, British: Inward Total 642,764; Equity (Net) 492,090; Debt (Net) 125,419 — Outward Total 1,134,319; Equity (Net) 961,062; Debt (Net) 85,922
  - Bermuda: Inward 632,105; 607,018; -32,119 — Outward 710,520; 510,981; 93,013
  - Cayman Islands: Inward 554,091; 494,080; 25,431 — Outward 568,772; 365,098; 142,119
  - Gibraltar: Inward 121,606; 76,451; -8,475 — Outward 189,199; 171,050; 7,488
  - Puerto Rico: Inward 109,951; 91,103; 18,848 — Outward 270; 202; 47
  - Bahamas, The: Inward 82,774; 26,730; 2,958 — Outward 98,304; 64,051; 13,253
  - Barbados: Inward 73,772; 17,896; -1,926 — Outward 16,062; 9,518; 3,372
  - United Arab Emirates: Inward 73,420; 69,179; 3,207 — Outward 88,372; 35,631; 31,583
  - Argentina: Inward 67,945; 61,978; 4,795 — Outward 12,992; 13,087; -781
  - Jersey: Inward 66,162; 110,264; -44,639 — Outward 249,294; 207,614; 25,497
  - Egypt: Inward 61,602; 39,895; 6,076 — Outward 416; 594; -125
  - Saudi Arabia: Inward 47,933; 25,325; 7,150 — Outward 14,858; 13,477; 770
  - Taiwan Province of China: Inward 45,477; 25,272; 898 — Outward 70,049; 53,656; 2,616
  - Colombia: Inward 37,586; 36,544; -1,603 — Outward 18,331; 17,974; 436
  - Peru: Inward 35,374; 26,401; 1,440 — Outward 1,439; 1,884; -1,052
- Aggregates:
  - Total Top 15 non-reporters: Inward 2,652,560; Equity (Net) 2,200,224; Debt (Net) 107,459 — Outward 3,173,198; 2,425,878; 404,157
  - % Top 15/total non-reporters: 89% 90% 76% — 94% 94% 91%
  - Total non-reporters: Inward 2,968,281; Equity (Net) 2,448,574; Debt (Net) 140,525 — Outward 3,378,389; 2,574,071; 443,130
  - % Non-reporters/total reported: 10% 11% 4% — 12% 10% 20%
  - Total World (as reported): Inward 28,405,605; Equity (Net) 23,075,491; Debt (Net) 3,750,588 — Outward 28,244,188; 25,165,725; 2,200,781
- Source: IMF’s CDIS, 2016 release.

### Fellow Enterprises
- Equity and debt instrument positions between fellow enterprises are to be classified as direct investment and therefore included in the CDIS.
- Some economies do not cover these positions under direct investment, mostly because they cannot identify them; they may classify these positions under portfolio or other investment, causing asymmetries.
- About half of the 107 economies reporting 2015 inward data separately identify investment between fellow enterprises.
- For outward CDIS, 32 out of 73 economies separately identify investments for fellows.
- Economies using the local enterprise group (LEG) as reporting unit cannot separately identify data between fellow enterprises (example: United States).
- Importance of investments between fellows varies:
  - Inward positions with fellow enterprises represent 40% of total inward direct investment in Belgium.
  - Inward positions with fellow enterprises represent 0.6% in Switzerland.
- Annex V contains positions between fellow enterprises relative to total inward and outward for CDIS participants that separately identify fellows.

### Nonfinancial Private Corporations
- Low participation of nonfinancial private corporations in direct investment surveys is a key challenge; they often comprise the bulk of target respondents.
- Gaps in coverage for the nonfinancial sector can result in bilateral asymmetries if counterpart coverage levels differ.
- Data collection approaches:
  - Most economies collect data from a sample of the target population.
  - Census of enterprises used by 33% of reporting economies for inward and 20% for outward.
- Low response rates often due to absence of legal mandate; survey participation may be voluntary.
- Legal mandate alone may not suffice; effective measures include:
  - Online user-friendly questionnaires.
  - Consultative meetings with industry associations.
  - Company visits to nonresponding corporations.
- Example reference: Philippines’ Experience (BOPCOM 16/27).

### Real Estate
- Nonresident investments in real estate may be significant for some countries but may not be covered, causing asymmetries with counterpart economies that do cover them.
- Around 40% of CDIS reporting economies include data on real estate in their direct investment (CDIS metadata responses).
- Most real estate positions involve the household sector, which is difficult to cover because household surveys are costly and cannot be conducted with required periodicity.
- Compilers primarily collect real estate data through public registers.
- Some countries cover real estate but counterpart allocation may be inaccurate or not provided (example: Austria).
- Russia: direct investments of individuals in real estate are included in CDIS; main data source is banks’ reports on personal transfers for real estate purchases (international transaction reporting system). Counterpart economy breakdown may be incorrect because recipients may be transit countries (example: Switzerland being used as transit).

### C. Different Geographic Allocation: Immediate vs. Ultimate Investing Economy
- CDIS guidelines: report direct investment by immediate (first) counterpart economy for both inward and outward positions.
- Some countries allocate inward positions to the ultimate investing economy instead.
- When first enterprise is a nonresident SPE, few countries look through the nonresident SPE to the next enterprise in the ownership chain.
- Differences in geographic allocation produce bilateral asymmetries at country level, not at global level.
- Inward positions by ultimate investing economy are useful for analysis; some economies compile this additional breakdown.
- Ultimate investor definition: enterprise that has control over the investment decision; country of residence of the ultimate investor is the ultimate investing country (Paragraphs 610 and 611 of BD4).

### D. Unallocated and Confidential Data
- When counterpart country cannot be identified or is confidential, data reported under “Not specified (including confidential)”.
  - Effect: does not affect global asymmetries, but prevents accurate bilateral comparisons and may affect analytical value if weight is relevant.
- Figure 8 (not reproduced here) shows weights of “Not Specified (including confidential)” over total direct investment:
  - Values steady over years except increases in 2012 for inward and 2012/2013 for outward.
  - Global level: not allocated/confidential ≈ 6% for inward and ≈ 7% for outward; have decreased in last two years.
- Annex VI lists top economies reporting unallocated or confidential data > 15% of total inward or outward:
  - For inward: 16 economies > 15%.
  - For outward: 10 economies > 15%.
- More granular breakdowns increase confidentiality risk; the more granular the data, the higher the risk of confidentiality.
- Example: SPEs covered in CDIS but limited in number may be subject to confidentiality restrictions and reported under "Not specified (including confidential)".

### E. Methodological Issues — Valuation Methods
- Direct investment is mainly concentrated in equity, representing around 80% of total direct investment.
- Different valuation methods for unlisted equity could produce large bilateral asymmetries.
- International standards: financial positions should be valued at market value (BPM6, paragraph 3.84).
- CDIS equity categories: listed equity, unlisted equity, and other equity.
  - Listed equity: can be valued at market prices.
  - Unlisted/other equity: may lack observable market prices.
- When market values unavailable, CDIS Guide requires valuation at own funds at book value (OFBV) (paragraph 3.12 of CDIS Guide).
  - OFBV definition: sum of (a) paid-up capital; (b) all types of reserves identified as equity in the enterprise’s balance sheet; (c) cumulated reinvested earnings; and (d) holding gains or losses included in own funds in the accounts, whether as revaluation reserves or profits or losses.
- Frequent revaluation of assets/liabilities yields closer approximation to market values; lack of revaluation for several years may poorly reflect market values.
- CDIS Guide prefers OFBV from the books of the direct investment enterprise to promote comparability and consistency between economies.
  - Compiler difficulties: access to books of nonresident direct investment enterprises for outward data may be limited.
  - Books should include revaluations to serve as accurate proxy of market values.
- CDIS metadata shows not all countries follow international valuation standards:
  - Most reporting economies follow standards: listed equity at market value and unlisted equity at OFBV.
  - Significant number of economies use OFBV to value listed equity — may not cause significant asymmetries if books frequently revalued, but could produce large asymmetries when books are not revalued for several years.
  - Other valuation methods used by some economies include historic cost (BPM6, paragraph 3.88 (f)): “Historic cost, in its strict sense, reflects the cost at the time of acquisition, but sometimes it also may reflect occasional revaluations.”

*Source: IMF’s CDIS, 2016 release*

### Box 1. A Case of Valuation at Historic Cost: The United States

### Box 1. A Case of Valuation at Historic Cost: The United States

### Summary of valuation issue
- Equity investment valued at historic cost produces smaller estimates than valuation at market value or OFBV because historic cost does not include revaluations.
- The United States is one country using historic cost for equity valuation in the CDIS context.
- Analysis selects 22 counterpart economies that provided outward equity positions in the United States (different than null or confidential) and for which the United States also provided inward equity positions received from them (different than null or confidential) during the last five years.

### Table 8: Aggregated values reported (US dollars, billions)
- Inward Equity Positions (Net) reported by the United States:
  - 2011: 1,550
  - 2012: 1,728
  - 2013: 1,847
  - 2014: 1,933
  - 2015: 2,087
- Outward Equity Positions (Net) reported by Counterpart Economies:
  - 2011: 2,120
  - 2012: 2,343
  - 2013: 2,763
  - 2014: 2,834
  - 2015: 3,131
- Difference (United States minus Counterpart Economies):
  - 2011: -570
  - 2012: -615
  - 2013: -916
  - 2014: -901
  - 2015: -1,044
- Observation: Asymmetries are always negative for the aggregate of the 22 selected economies—the total amount reported by the United States is smaller than the total amount reported by its counterpart economies for all years in the period.

### Explaining positive bilateral asymmetries at individual-economy level
- Some bilateral asymmetries at individual economy level are positive; notably with Belgium and the United Kingdom these asymmetries are positive for the five years analyzed.
- Possible explanations:
  - Counterpart economies (Belgium and the United Kingdom) use OFBV to value listed equity instead of market value.
  - Other factors such as coverage or geographical allocation.

### Compiler action recommendation on equity valuation
- Compilers should address these asymmetries and try to estimate the market or fair value of direct investment positions on equity.

### Valuation methods for debt instruments in the CDIS
- Debt instruments consist of:
  - (i) debt securities, which should be valued at market prices; and
  - (ii) other debt instruments (loans, deposits, insurance, pension, and standardized guarantee schemes, trade credit and advances, and other accounts payable/receivable), which should be valued at nominal value.
- Nominal value is defined as the amount the debtor owes to the creditor, which comprises the outstanding amount of principal and any accrued interest not yet paid.
- Metadata-based findings:
  - More than 40 percent of the CDIS reporting economies follow the international standards and value debt securities at market value and other debt instruments at nominal value.
  - In principle, bilateral asymmetries produced on direct investment debt instruments due to different valuation methods are not expected to be relevant, as market value is not likely to significantly differ from nominal value.
  - In 2015, 80 percent of the values reported under inward debt instrument positions by individual economies are larger than the corresponding values reported under outward debt instrument positions by their counterpart economies.
- Potential reasons for observed positive asymmetries in debt:
  - Overestimated inward debt data and/or underestimated outward debt data mainly due to different data coverage and, to a lesser extent, valuation method differences.
  - Compilers usually have access to more accurate data to compile inward direct investment than outward direct investment.
  - Debt instrument liabilities (including intercompany debt) are more closely monitored in external debt contexts.
  - Methods to compile intercompany debt may focus on initial disbursements and may not properly track debt repayments or may account repayments through other means (repatriating benefits, under invoicing, etc.), tending to underestimate outward debt instruments.

### Debt between affiliated financial corporations
- Debt between selected affiliated financial corporations (deposit-taking corporations, investment funds, and other financial intermediaries except insurance corporations and pension funds) should be excluded from direct investment and included in portfolio or other investments as appropriate.
- For some economies, data sources may lack sufficient detail to identify these positions and exclude them, leading to asymmetries when counterpart economies follow the recommendation and exclude them.
- CDIS metadata shows that most CDIS reporters exclude this debt from direct investment; therefore, in principle, this exclusion should not generate significant bilateral asymmetries.

### Data sources and estimation techniques
- Main data sources: enterprise surveys and financial statements of companies; these are the most adequate and accurate for compiling direct investment positions.
- Complementary data sources may be used to fill gaps or until survey results are available.
- Around 70 percent of reporting economies apply some estimates to fill gaps for nonresponding companies.
  - Most compilers indicate high survey response rates; in most cases, estimated amounts are not relevant for the economy.
- Use of non-enterprise-survey/financial-statement sources and different estimation methods for nonresponding units could produce asymmetries between counterpart economies, but are not expected to be relevant provided information related to key companies is compiled.
- Accumulation of transactions:
  - Few economies use accumulation of transactions as a data source to estimate direct investment positions in the CDIS.
  - Some economies use accumulation of transactions to estimate data for nonresponding units.
  - Positions on equity and debt securities are more likely to be affected by price changes; estimation from balance of payments transactions should consider exchange rate and price changes to avoid bilateral asymmetries.

### Criteria for identifying entities in a direct investment relationship
- Economies should use the framework for direct investment relationships (FDIR) to identify entities in a direct investment relationship.
- Some economies follow simplified methods (direct influence/indirect control, or participation multiplication method), which can misclassify direct investment as portfolio or other investment.
- More than 50 percent of CDIS reporters apply the recommended standards and identify direct investment relationships based on the FDIR for both inward and outward positions.
- Positions between related companies could be misclassified for economies using other methods.

### Addressing bilateral asymmetries — key approaches
- Compliance with recommended guidelines:
  - Apply uniform standards in coverage, geographical counterparty allocation, valuation, and criteria for identifying direct investment relationships.
  - Coverage should include investments in or from resident SPEs, positions between fellow enterprises, direct investment involving nonfinancial corporations, and real estate.
  - Geographical allocation should be based on the immediate (first) counterpart economy for breakdowns by counterpart.
  - Valuation: market value for listed equity and debt securities; OFBV for unlisted equity and other equity; nominal value for debt instruments other than debt securities.
  - Exclude debt between selected affiliated financial corporations from direct investment where possible.
  - Use position (stock) source data; avoid relying on accumulation of transactions unless adjustments for price and exchange rate changes are made.
  - Use appropriate statistical techniques for survey frames and estimation (grossing up, thresholds, non-response).
  - Aim to evolve from simplified identification methods to FDIR.
- Reducing reporting of confidential and unallocated data by counterpart economies:
  - Do best efforts to allocate investments to specific counterpart economies and minimize use of confidential/unallocated counterparts.
  - Revisit and revise past time series as data collection systems improve and delayed information is received.
  - Estimations for counterpart allocation could use past allocations, transaction data, mirror information, etc.
- Data and metadata exchange:
  - Bilateral data exchange at enterprise level is the most efficient way to identify and address asymmetries, though enterprise-level exchange may be limited by confidentiality.
  - Data exchange at more aggregated levels and metadata exchange are useful and pose fewer confidentiality issues but may be less effective for reducing asymmetries.
  - The CDIS provides valuable information on aggregated data levels and metadata.
  - Existing mechanisms for micro-level bilateral exchange include the FDI Network (Eurostat and ECB) and various regular bilateral exchanges between statisticians.
  - Bilateral exchange of micro data between statisticians is considered good practice to reduce asymmetries and identify their causes, even if it does not reconcile all transactions and positions.

*Source IMF’s CDIS, 2016 release*

### Box 2. The FDI Network

### Box 2. The FDI Network

### Purpose and operation of the FDI Network
- Transmission and exchange of confidential data in the context of the FDI Network is foreseen exclusively for statistical purposes and in particular for increasing the consistency/quality of the EU/euro area balance of payments statistics.
- Transmission of confidential data is enabled pursuant to the specific regulations of the European Parliament and of the Council.
- Exchange of micro level data between direct investment compilers is based on a secure data exchange between countries which are parties to the financial transaction/position; concerned countries then try to reconcile their data. (ECB/Eurostat BOPCOM 14/20)

### Confidentiality constraints and data granularity
- Confidentiality at microdata level (enterprise) or at higher levels of aggregation (for some types of investments or sectors) may hamper analyzing bilateral asymmetries with counterpart economies.
- For some economies, sharing micro data with third parties or even with statistical agencies of counterpart economies may not be feasible due to confidentiality or legal constraints.
- Increasing data flow with more granular information requires substantial work on confidentiality and legal issues.
- The new recommendation on promotion of data sharing of the Second Phase of the G-20 Data Gaps Initiative (DGI-2) (recommendation II.20) was welcomed by the G-20 economies to facilitate sharing experiences with exchanges of granular data and identifying ways to overcome existing barriers.
- Some economies conduct bilateral exchanges of micro data through bilateral agreements or contracts to preserve confidentiality; some economies request prior authorization from reporters.
- Data exchange at more aggregated levels poses fewer confidentiality issues; CDIS granularity with breakdowns by:
  - instrument (net equity and net debt),
  - further breakdowns of net debt instruments into gross assets and gross liabilities,
  - separate identification of positions between fellow enterprises,
  provides a wide range for analysis.

### Metadata exchange and its role
- Metadata exchange is important for identifying and analyzing reasons for asymmetries.
- Economies collaborate with main counterparts via professional exchanges among officials responsible for compilation and dissemination of direct investment statistics.
- Results of the CDIS metadata questionnaire (available at the CDIS website) provide detailed information on collection and compilation practices adopted by CDIS reporting economies and are an important source for assessing data quality and deviations from recommended standards.
- The structured design of the metadata questionnaires allows easy cross-economy comparisons and facilitates identification of methodological differences; questionnaires include compilers’ contact data to facilitate information exchange.
- Methodological comparability is improving with implementation of BPM6 and the 2015 CDIS Compilation Guide, but areas of practice variation remain (see Annex VII).
- Footnote: OECD’s FDI database contains valuable information on bilateral DI positions, transactions, and income (See OECD.Stat).

### Planned actions by CDIS participants to reduce asymmetries
- STA consulted with 47 CDIS participating economies as part of the second exercise of the CDIS bilateral asymmetries conducted in 2016 on plans to implement specific actions to reduce asymmetries.
- Main categories of planned actions reported include:
  - continue analyzing reasons for discrepancies on an aggregated level;
  - more active bilateral consultations and interaction with counterparts;
  - more active participation in the FDI Network (European countries);
  - continued work to develop the European Group Register;
  - revisions to CDIS data;
  - enhancements in reporting systems and improvements in data sources.

### Ongoing and planned international initiatives
- OECD organizes bilateral asymmetry meetings adjacent to the Working Group on International Investment Statistics (WGIIS) meetings.
- Eurostat facilitates platforms for bilateral asymmetry meetings.
- The IMF, with support of Bundesbank, will conduct a workshop for CDIS compilers aimed at reducing asymmetries in November 2017.
- Work ahead includes following up on the possibility of Eurostat sharing the infrastructure of the FDI Network to expand it at world level (see IMF, BOPCOM 16/18).
- Planned enhancements to the CDIS metadata questionnaire:
  - add specific questions on bilateral asymmetries (main reasons that explain asymmetries for specific counterpart countries, instruments (equity/debt), type of investments, sectors involved; future or current work to address/understand/reduce them; links to related publications);
  - encourage economies to update and fine-tune metadata information as needed to reflect up-to-date methodological differences.
- The steps taken to increase awareness of asymmetries and understand their causes are an important achievement even if reduction is not always feasible.

### Annex I — Mirror data: usefulness and limitations
- Usefulness:
  - Mirror data (direct investment positions by counterpart economy at CDIS) allow comparison between data reported by one economy and mirror data reported by counterpart CDIS reporting economies.
  - Mirror information can be used to calculate derived data for direct investment and can help economies without counterpart detail to estimate inward and outward direct investment positions by counterpart economy.
  - Comparison between reported and derived data can cross-check and verify estimates, highlight data gaps or errors, or identify areas for improvement.
- Limitations:
  - Differences between derived measures and actual data arise not only from bilateral asymmetries; some counterparts may not participate in CDIS, may not provide data due to confidentiality, or may fall below a reporting threshold.
  - Use of derived data should be done with caution.

### Annex II — Fellow enterprises: guidance and statistics
- Definition: Fellow enterprises are enterprises under the control or influence of the same immediate or indirect investor but neither fellow enterprise controls or influences the other fellow enterprise (BPM6. Paragraph 6.17 c).
- International guidelines:
  - Direct investment positions between fellow enterprises should be recorded in the CDIS based on the extended directional principle; only when the residence of the UCP is unknown should these positions be recorded on an assets liabilities basis.
- Compilation methods (CDIS Guide, paragraph 4.10):
  - Extended directional principle: “both asset and liability positions between fellow enterprises are to be recorded in outward direct investment when the ultimate controlling parent is a resident, and in inward direct investment when the UCP is a nonresident”.
  - Assets liabilities principle: “to record assets of a resident fellow enterprise on a nonresident fellow enterprise in outward direct investment, and to record liabilities of a resident fellow enterprise to a nonresident fellow enterprise in inward direct investment”.
- Reporter statistics:
  - Most CDIS reporters (around 85 percent) include direct investment positions with fellow enterprises in their total inward and outward direct investment positions.
  - Nearly half provide separate data on fellows.
  - CDIS metadata questionnaire shows:
    - 50 percent of CDIS reporters record direct investment positions with fellow enterprises abroad using the extended directional principle.
    - around 30 percent of the economies apply the assets liabilities principle.

### Annex III — Asymmetry scenarios and empirical examples
- Scenario (a): When the Ultimate Controlling Parent (UCP) is nonresident for both economies
  - Table 9 examples (As of end-2015, US Dollars, millions; Source IMF CDIS, 2016 release):
    - Poland / Spain:
      - Outward DI (Net) with Fellows, reported by the reporting economy: 883
      - Inward DI (Net) with Fellows, reported by the counterpart economy: 3,395
      - Diff. = Outward - Inward: -2,512
    - Spain / Poland:
      - Outward DI (Net) with Fellows, reported by the reporting economy: -385
      - Inward DI (Net) with Fellows, reported by the counterpart economy: -3,122
      - Diff. = Outward - Inward: 2,737
    - Total Difference: 225
    - Germany / Sweden:
      - Outward DI (Net) with Fellows, reported by the reporting economy: 146
      - Inward DI (Net) with Fellows, reported by the counterpart economy: -1,235
      - Diff. = Outward - Inward: 1,381
    - Sweden / Germany:
      - Outward DI (Net) with Fellows, reported by the reporting economy: 114
      - Inward DI (Net) with Fellows, reported by the counterpart economy: 1,372
      - Diff. = Outward - Inward: -1,258
    - Total Difference: 123
  - Observations:
    - In examples, countries record larger absolute amounts under inward than outward because UCPs are resident in a third economy.
    - Large component differences (e.g., $2.5 billion) can largely offset when comparing total net direct investment (outward minus inward); bilateral asymmetry drops to smaller amounts (e.g., $0.2 billion or $0.1 billion in the examples).
- Scenario (b): When one economy collects consolidated data from Local Enterprise Groups (LEGs)
  - Description:
    - Some economies collect consolidated direct investment data from resident parent companies as Local Enterprise Groups (LEGs); consolidated data include affiliated units resident in the compiling economy and do not identify separate fellow enterprise flows, so values under “direct investment between fellow enterprises” can be null.
    - When counterpart collects data from individual enterprises, asymmetries can arise in the “direct investment between fellow enterprises” component, while aggregated totals remain consistent.
  - Example (as of end-2015, US Dollars, millions; Source IMF CDIS, 2016 release):
    - Netherlands / Slovak Republic:
      - Outward with Fellows, reported by the reporting economy: -1,334
      - Inward with Fellows, reported by the counterpart economy: 1,334
      - Diff. = Outward with fellows – Inward with fellows: 8,487 1,334 (table shows component asymmetry: -1,334 vs 1,334)
      - Outward reported by the reporting economy: 8,487
      - Inward reported by the counterpart economy: 8,536
      - Diff. = Outward - Inward: -50
  - Observation:
    - In the Netherlands–Slovak Republic example, large asymmetry arises in the fellows component (US$1.4 billion) but aggregated total outward vs total inward difference decreases to almost zero.

### Annex IV — Special Purpose Entities (SPEs): inclusion and identification
- Rationale for inclusion:
  - SPEs channel a large amount of funds between entities outside the economy in which they are located and have little presence in the host economy.
  - SPEs often located in countries offering tax or other advantages generate cross-border transactions and positions between subsidiaries in different countries.
  - For SPEs most assets and liabilities are with nonresidents; they typically have few or no employees, little or no production and limited physical presence.
  - SPEs should be included in direct investment statistics because their funds are an integral part of a direct investor’s financial transactions and positions; excluding them would distort and substantially understate direct investment financial flows and positions at aggregate levels.
  - Inclusion promotes symmetry and consistency among economies.
- Identification indicators for SPEs (OECD Benchmark Definition- BD4):
  - The enterprise is a legal entity, formally registered with a national authority and subject to fiscal and other legal obligations in the economy in which it is resident;
  - The enterprise is ultimately controlled by a nonresident parent company, either directly or indirectly;
  - The enterprise has few or no employees, little or no production in the host economy, and little or no physical presence;
  - Almost all the assets and liabilities of the enterprise represent investments in or from other economies;
  - The core business consists of group financing or holding activities (they channel funds from nonresidents to other nonresidents) and managing/directing plays only a minor role in daily activities.

*Source: wp17261 - Box 2. The FDI Network (IMF CDIS material extracted from the supplied PDF content)*

### ANNEX V. ECONOMIES REPORTING DATA ON FELLOW ENTERPRISES TO THE CDIS

### ANNEX V. ECONOMIES REPORTING DATA ON FELLOW ENTERPRISES TO THE CDIS

### Inward direct investment — key findings
- For all economies reporting fellow enterprises data for inward, positions with fellow enterprises represent 2 percent of total inward direct investment.
- Highest and lowest shares (in absolute terms):
  - Belgium: 40 percent
  - Mexico: 12.7 percent
  - Switzerland: 0.6 percent
  - Mauritius: 0.8 percent

- Top 15 inward economies reporting 2015 data for fellow enterprises (US Dollars, millions). Columns shown: Total Inward (a), Of Which Inward (Net) with Fellow Enterprises (b), Equity (Net) (d), Debt (Net) (e), Equity plus debt with fellow enterprises — Liabilities (Gross) (g), Assets (Gross) (h), % (f/a). (Preserve table arithmetic as presented.)
  - United Kingdom: Total Inward 1,554,303; Of Which, Inward (Net) with Fellow Enterprises 1,426,215 128,088; Equity (Net) 246,925; Debt (Net) 118,837; Equity plus debt with fellow enterprises Liabilities (Gross) 16,989; Assets (Gross) 18,965 1,977; % 1.1
  - China, P.R.: Hong Kong: Total Inward 1,394,580; Of Which, Inward (Net) with Fellow Enterprises 1,369,491 25,089; Equity (Net) 148,281; Debt (Net) 123,192; Equity plus debt with fellow enterprises Liabilities (Gross) -32,108; Assets (Gross) 61,560 93,669; % 2.3
  - Ireland: Total Inward 866,218; Of Which, Inward (Net) with Fellow Enterprises 594,503 271,714; Equity (Net) 656,477; Debt (Net) 384,763; Equity plus debt with fellow enterprises Liabilities (Gross) 37,513; Assets (Gross) 314,684 277,171; % 4.3
  - Switzerland: Total Inward 862,624; Of Which, Inward (Net) with Fellow Enterprises 846,048 16,576; Equity (Net) 166,100; Debt (Net) 149,523; Equity plus debt with fellow enterprises Liabilities (Gross) -4,749; Assets (Gross) 86,123 90,872; % 0.6
  - Germany: Total Inward 786,941; Of Which, Inward (Net) with Fellow Enterprises 589,849 197,092; Equity (Net) 372,738; Debt (Net) 175,647; Equity plus debt with fellow enterprises Liabilities (Gross) 99,318; Assets (Gross) 208,770 109,452; % 12.6
  - France: Total Inward 660,107; Of Which, Inward (Net) with Fellow Enterprises 626,559 33,548; Equity (Net) 179,829; Debt (Net) 146,281; Equity plus debt with fellow enterprises Liabilities (Gross) 20,322; Assets (Gross) 124,052 103,730; % 3.1
  - Spain: Total Inward 546,803; Of Which, Inward (Net) with Fellow Enterprises 456,936 89,867; Equity (Net) 144,207; Debt (Net) 54,338; Equity plus debt with fellow enterprises Liabilities (Gross) 54,312; Assets (Gross) 87,835 33,522; % 9.9
  - Mexico: Total Inward 509,292; Of Which, Inward (Net) with Fellow Enterprises 397,520 111,772; Equity (Net) 162,861; Debt (Net) 51,089; Equity plus debt with fellow enterprises Liabilities (Gross) 64,570; Assets (Gross) 103,779 39,209; % 12.7
  - Brazil: Total Inward 460,381; Of Which, Inward (Net) with Fellow Enterprises 392,577 67,804; Equity (Net) 85,658; Debt (Net) 17,854; Equity plus debt with fellow enterprises Liabilities (Gross) 40,283; Assets (Gross) 40,283 (blank); % 8.7
  - Belgium: Total Inward 455,366; Of Which, Inward (Net) with Fellow Enterprises 624,393 -169,027; Equity (Net) 283,088; Debt (Net) 452,115; Equity plus debt with fellow enterprises Liabilities (Gross) -181,194; Assets (Gross) 174,700 355,895; % 39.8
  - Italy: Total Inward 337,083; Of Which, Inward (Net) with Fellow Enterprises 302,579 34,505; Equity (Net) 86,592; Debt (Net) 52,088; Equity plus debt with fellow enterprises Liabilities (Gross) 4,065; Assets (Gross) 33,264 29,199; % 1.2
  - Sweden: Total Inward 289,518; Of Which, Inward (Net) with Fellow Enterprises 237,229 52,288; Equity (Net) 93,310; Debt (Net) 41,022; Equity plus debt with fellow enterprises Liabilities (Gross) 28,275; Assets (Gross) 38,666 10,391; % 9.8
  - Mauritius: Total Inward 269,105; Of Which, Inward (Net) with Fellow Enterprises 169,065 100,041; Equity (Net) 100,053; Debt (Net) 12; Equity plus debt with fellow enterprises Liabilities (Gross) -2,040; Assets (Gross) 10,628 12,668; % 0.8
  - Austria: Total Inward 261,349; Of Which, Inward (Net) with Fellow Enterprises 250,833 10,516; Equity (Net) 41,241; Debt (Net) 30,725; Equity plus debt with fellow enterprises Liabilities (Gross) 9,089; Assets (Gross) 31,244 22,154; % 3.5
  - Russian Federation: Total Inward 257,287; Of Which, Inward (Net) with Fellow Enterprises 201,342 55,945; Equity (Net) 65,021; Debt (Net) 9,075; Equity plus debt with fellow enterprises Liabilities (Gross) 25,833; Assets (Gross) 26,580 747; % 10.0
  - All other economies reporting fellows’ data: Total Inward 2,225,455; Of Which, Inward (Net) with Fellow Enterprises 1,845,340 380,111; Equity (Net) 592,015; Debt (Net) 211,908; Equity plus debt with fellow enterprises Liabilities (Gross) 48,891; Assets (Gross) 163,129 114,242; % 2.2
  - Total reporting fellows’ data: Total Inward 11,736,412; Of Which, Inward (Net) with Fellow Enterprises 10,330,478 1,405,929; Equity (Net) 3,424,397; Debt (Net) 2,018,469; Equity plus debt with fellow enterprises Liabilities (Gross) 229,369; Assets (Gross) 1,524,262 1,294,898; % 2.0
  - World: Total Inward 28,405,605; Of Which, Inward (Net) with Fellow Enterprises 23,075,491 3,750,588; Equity (Net) 5,991,193; Debt (Net) 2,701,123; Equity plus debt with fellow enterprises Liabilities (Gross) 229,369; Assets (Gross) 1,524,262 1,294,898; % 0.8

- Source note in annex: Percentages for inward and outward data are calculated based on net figures (following the directional principle and netting liabilities and assets). Weights for gross liabilities and assets would be higher.

### Outward direct investment — key findings
- For all economies reporting fellow enterprises data for outward, positions with fellow enterprises represent 0.7 percent of total outward direct investment.
- Highest and lowest shares (in absolute terms):
  - Spain: 4.4 percent
  - Korea, Republic of: 4.3 percent
  - Russian Federation: 4.2 percent
  - Lowest percentages (rounded to 0.0 percent): Cyprus, Switzerland, China, P.R.: Hong Kong

- Top 15 outward economies reporting 2015 data for fellow enterprises (US Dollars, millions). Columns shown: Total Outward (a), Of Which, Outward (Net) with Fellow Enterprises (b), Equity (Net) (d), Debt (Net) (e), Equity plus debt with fellow enterprises — Assets (Gross) (g), Liabilities (Gross) (h), % (f/a).
  - China, P.R.: Hong Kong: Total Outward 1,383,586; Of Which, Outward (Net) with Fellow Enterprises 1,207,413 176,173; Equity (Net) 228,656; Debt (Net) 52,482; Equity plus debt with fellow enterprises Assets (Gross) 559; Liabilities (Gross) 3,435 2,876; % 0.0
  - Germany: Total Outward 1,376,181; Of Which, Outward (Net) with Fellow Enterprises 1,473,916 -97,735; Equity (Net) 294,267; Debt (Net) 392,002; Equity plus debt with fellow enterprises Assets (Gross) -6,805; Liabilities (Gross) 31,173 37,978; % 0.5
  - France: Total Outward 1,198,770; Of Which, Outward (Net) with Fellow Enterprises 1,156,723 42,047; Equity (Net) 264,103; Debt (Net) 222,056; Equity plus debt with fellow enterprises Assets (Gross) 32,814; Liabilities (Gross) 180,356 147,542; % 2.7
  - Switzerland: Total Outward 1,113,081; Of Which, Outward (Net) with Fellow Enterprises 990,120 122,961; Equity (Net) 349,219; Debt (Net) 226,257; Equity plus debt with fellow enterprises Assets (Gross) -356; Liabilities (Gross) 21,974 22,331; % 0.0
  - Ireland: Total Outward 887,510; Of Which, Outward (Net) with Fellow Enterprises 794,189 93,321; Equity (Net) 217,694; Debt (Net) 124,373; Equity plus debt with fellow enterprises Assets (Gross) 14,864; Liabilities (Gross) 21,728 6,864; % 1.7
  - Spain: Total Outward 490,332; Of Which, Outward (Net) with Fellow Enterprises 529,242 -38,910; Equity (Net) 57,527; Debt (Net) 96,437; Equity plus debt with fellow enterprises Assets (Gross) -21,688; Liabilities (Gross) 17,294 38,982; % 4.4
  - Italy: Total Outward 467,300; Of Which, Outward (Net) with Fellow Enterprises 464,199 3,101; Equity (Net) 70,932; Debt (Net) 67,832; Equity plus debt with fellow enterprises Assets (Gross) -2,696; Liabilities (Gross) 8,091 10,787; % 0.6
  - Belgium: Total Outward 446,237; Of Which, Outward (Net) with Fellow Enterprises 414,148 32,089; Equity (Net) 76,470; Debt (Net) 44,381; Equity plus debt with fellow enterprises Assets (Gross) 8,630; Liabilities (Gross) 38,548 29,921; % 1.9
  - Sweden: Total Outward 357,133; Of Which, Outward (Net) with Fellow Enterprises 329,042 28,092; Equity (Net) 89,887; Debt (Net) 61,795; Equity plus debt with fellow enterprises Assets (Gross) 2,158; Liabilities (Gross) 21,647 19,489; % 0.6
  - Austria: Total Outward 309,993; Of Which, Outward (Net) with Fellow Enterprises 281,495 28,498; Equity (Net) 35,934; Debt (Net) 7,436; Equity plus debt with fellow enterprises Assets (Gross) 10,809; Liabilities (Gross) 15,241 4,432; % 3.5
  - Russian Federation: Total Outward 286,583; Of Which, Outward (Net) with Fellow Enterprises 282,099 4,484; Equity (Net) 79,322; Debt (Net) 74,839; Equity plus debt with fellow enterprises Assets (Gross) 12,128; Liabilities (Gross) 13,201 1,074; % 4.2
  - Korea, Republic of: Total Outward 271,581; Of Which, Outward (Net) with Fellow Enterprises 233,033 38,548; Equity (Net) 41,361; Debt (Net) 2,813; Equity plus debt with fellow enterprises Assets (Gross) 11,758; Liabilities (Gross) 12,031 273; % 4.3
  - Mauritius: Total Outward 221,313; Of Which, Outward (Net) with Fellow Enterprises 173,915 47,397; Equity (Net) 57,792; Debt (Net) 10,395; Equity plus debt with fellow enterprises Assets (Gross) 2,013; Liabilities (Gross) 12,639 10,626; % 0.9
  - Denmark: Total Outward 178,893; Of Which, Outward (Net) with Fellow Enterprises 155,092 23,801; Equity (Net) 51,663; Debt (Net) 27,863; Equity plus debt with fellow enterprises Assets (Gross) 515; Liabilities (Gross) 660 145; % 0.3
  - Cyprus: Total Outward 174,029; Of Which, Outward (Net) with Fellow Enterprises 159,235 14,794; Equity (Net) 15,331; Debt (Net) 538; Equity plus debt with fellow enterprises Assets (Gross) 25; Liabilities (Gross) 25 0; % 0.0
  - All other economies reporting fellows’ data: Total Outward 665,829; Of Which, Outward (Net) with Fellow Enterprises 648,656 17,172; Equity (Net) 121,716; Debt (Net) 104,544; Equity plus debt with fellow enterprises Assets (Gross) 2,074; Liabilities (Gross) 31,894 29,817; % 0.3
  - Total reporting fellows’ data: Total Outward 9,828,352; Of Which, Outward (Net) with Fellow Enterprises 9,292,519 535,832; Equity (Net) 2,051,875; Debt (Net) 1,516,042; Equity plus debt with fellow enterprises Assets (Gross) 66,801; Liabilities (Gross) 429,938 363,136; % 0.7
  - World: Total Outward 28,244,188; Of Which, Outward (Net) with Fellow Enterprises 25,165,725 2,200,781; Equity (Net) 4,682,433; Debt (Net) 2,742,392; Equity plus debt with fellow enterprises Assets (Gross) 66,801; Liabilities (Gross) 429,938 363,136; % 0.2

*Source IMF’s CDIS, 2016 release*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2017/wp17261.pdf_