## _wp10295

## Source details

**Canonical URL:** [_wp10295](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2010/_wp10295.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2010/_wp10295.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2010/_wp10295.pdf.json)

---

### Related literature: Intra-Euro area and intra-EU imbalances
- No reason why positions of constitutive countries or states should be in balance within a monetary union; anecdotal evidence for the US suggests significant interstate current account imbalances can be resolved without interstate balance-of-payments crises (Greenspan, 2004).
- Absence of an exchange rate premium helps facilitate financing of current account deficits.
- Current account deficits across Euro area and EU countries were for some time viewed as benign: capital flows from wealthier European countries (higher GDP and capital per capita) to catching up economies (lower GDP per capita), facilitating convergence; contrasted with Feldstein and Horioka (1980) and Lucas (1990).
- Literature supporting benign view: Blanchard and Giavazzi (2002), Abiad, Leigh and Mody (2007), Schmidtz and von Hagen (2009), Bakker and Gulde (2010).
- Growing view that external imbalances could reflect internal unsustainable developments (excessive indebtedness of private or public agents), likely to result in painful adjustments: Gourinchas (2002), Ahearne and Pisani-Ferry (2006), Blanchard (2006), European Commission (2006 and following years), Guyon (2007).
- Deleveraging in some EU countries with booms and large persistent current account deficits provided closure to the debate; modalities for incorporating these imbalances into EU policy advice are being agreed at EU level.

### Related literature: Bilateral financial linkages at global level
- Motivation: growth in cross-border claims and global imbalances spurred documentation of external positions to understand financial inter-linkages and contagion.
- Lane and Milesi-Ferretti: IIP database for 145 countries over 1970-2004, updated in 2007 (External Wealth of Nations, Mark II - EWNII).
- Kubelec and Sá (2010): bilateral external assets and liabilities for 18 countries over 1980-2005; sample included 6 large European countries but did not cover key creditor and debtor economies within Europe.
- Gravity models used to estimate missing data in some studies.
- Lane and Shambaugh (2010): database for 1990-2005 on major currency exposures; did not detail intra-EU/EU-area bilateral assets and liabilities.
- Milesi-Ferretti, Strobbe and Tamirisa (2010): bilateral assets and liabilities for about 70 countries covering 15 large countries or country groupings; Euro area and Emerging Europe treated as whole.

### Related literature: Bilateral financial linkages in the EU and the Euro area
- Studies using CPIS portfolio investment data show a “Euro area bias”.
- Indirect measures (e.g., bilateral trade balances as proxy for capital flows) were used in some work (Schmidtz and von Hagen (2009)).
- Shared conclusions: capital appeared to flow from higher GDP per capita economies to lower GDP per capita ones (Blanchard and Giavazzi (2002), Abiad, Leigh and Mody (2007)).
- BIS consolidated bilateral data used to investigate contagion channels (Árvai, Driessen, and Ötker-Robe (2009); Tressel (2010)).
- Identified gap: no single view encompassing various financing instruments describing intra-Euro area and intra-EU bilateral positions.

### Methodology — Main principles
- Dataset coverage: 29 reporting European countries over the 2001 to 2008 period, with bilateral positions reported against over 200 partner countries.
- All data are in US dollars.
- Classification follows the IIP classifications from the 5th revision of the Balance of Payments Manual (IMF, 1993), residency principle.
- Ideal components to capture: Reserve assets; foreign direct investment; portfolio investment; financial derivatives; Other investment.
- Missing bilateral components: no bilateral information on financial derivatives and foreign exchange reserves. Approach: subtract aggregate financial derivatives (from national sources reported to the IMF’s electronic data statistical system) and aggregate total foreign reserves assets (from IMF’s Balance of Payment Statistics (BPTS)) from aggregate IIP figures to be decomposed.
- End-of-year bilateral stock positions documented or estimated for:
  - Foreign direct investment;
  - Portfolio investment, divided into equity and debt securities; and
  - Other investment.
- Data represent stocks of assets and liabilities; variations over time may reflect transactions, revaluation effects, exchange rate changes, or revisions.

### Methodology — Data sources
- Direct investment: main source OECD, FDI positions by partner country; practical use of book values; IMF’s CDIS ongoing improvement.
- Portfolio investment: main source IMF’s CPIS. CPIS provides year-end holdings valued at market prices, cross-classified by issuer country. Participation voluntary; some 75 economies participate. Liabilities largely estimated as “mirror” from creditor positions; incomplete participation causes derived bilateral liabilities to fall short of EDSS aggregates.
- Other investment: BIS Locational Banking Statistics used; to avoid double counting with portfolio instruments, only balance sheet items reported under “loans and deposits” are included.
  - Other investment comprises Other investment from banks and Other investment from non-banks.
  - For BIS reporting countries, reported locational data used directly; for BIS non-reporting countries, mirror data used where available.
  - Mirror data limitations: miss relations between non-banks and non-banks and positions involving banks from BIS non-reporting countries.
- Use of locational BIS data consistent with residency-based IIP; consolidated BIS ultimate-basis data not consistent with residency-based IIP but useful for total bank exposures.
- Restricted nature of BIS figures constrains public reporting for some country groupings.

### Gaps and discrepancies
- Inherent data limitations: uncertainty about effective holder of a claim or liability and the economic nature of the claim when intermediary vehicles are involved; introduces “intermediation” and “geographical veil” in CPIS data.
- Asymmetries between counterparties are common; guiding principle: retain data as reported by country authorities.
- No bilateral information on reserves and financial derivatives; for most countries these aggregates represent less than 10 percent of total assets plus liabilities, with exceptions:
  - Some Member states acceding in 2004 or 2007 where reserve assets are a substantial share of external assets.
  - The U.K., which plays a major role in financial derivatives markets (more than a third of its claims and liabilities).
- Sum of reported bilateral positions in FDI, portfolio, and other investment does not add up to gross asset and liability data stripped of financial derivatives and foreign reserve assets; gaps stem from incomplete CPIS and BIS reporting and mirror data limitations.
- At end-2008, unallocated assets and liabilities by country:
  - For most EU advanced countries, gaps are around 15 percent of total assets and liabilities.
  - Gaps are particularly small (as a share) for the U.K.
  - Gaps are very large for Luxembourg (another reason for treating Luxembourg separately).
  - Switzerland’s unallocated total assets are negative, primarily because the sum of other investment bank bilateral claims exceeded reported aggregates.
  - Luxembourg reports minimal bilateral foreign direct investment and has large portfolio equity gaps (may reflect incomplete data on cross border mutual fund industry claims).
  - Data gaps in EU countries that joined in 2004 and 2007 vary: limited gaps in Estonia, Bulgaria and Romania; above-average discrepancies in Lithuania, Latvia, Cyprus and Malta.
  - Discrepancies are also large in Norway.
  - In absolute amount, the most significant discrepancies are for Luxembourg and the U.K.

### IV. Stylized facts — Overview and key metrics
- Database constructed with data from BIS, OECD, ECB and national sources; and WEO.
- Data exclude financial derivatives and reserve assets.
- Analysis covers primarily 1999–2008 (financial integration charts) and up to 2009 for some regions.
- Financial integration measured as sum of cross border assets and liabilities expressed as a percentage of GDP (excludes intra-zone claims in plotted series).
- Euro Area external position: -17.7 percent of its GDP at end-2008, down to -12.4 percent at end Q2 2010.
- Reintegrating financial derivatives would attenuate the 2008 inflexion: the increase in the value of total assets and liabilities in the form of financial derivatives roughly doubled in 2008 compared with 2007 (to around +550 bn Euros).
- IMF and ECB data broadly consistent; overall difference limited to 2%-6% from 2001 to 2008 on the asset side and 1%-5% on the liabilities side in percent of total EA assets/liabilities as reported by the ECB.

### Internal financial integration and partners
- Other EU countries constituted the first financial partner of both EU countries and Euro area countries, with around half of total assets and liabilities.
- For Euro area countries, assets and liabilities vis-à-vis other Euro area countries represented 40-45 percent of total assets or liabilities.
- After correcting for un-tracked assets:
  - EU assets and liabilities vis-à-vis EU countries represented around 2/3 of the allocated assets and liabilities.
  - Other Euro area countries accounted for a bit more than half of the tracked assets and liabilities of Euro area countries.
- Correlation metrics:
  - Correlation between net positions vis-à-vis EU countries (percent of GDP) and vis-à-vis Euro area countries: 0.80.
  - Correlation between global IIPs and net positions vis-à-vis the EU at end-2008: 0.75.
- Examples at end-2008:
  - Countries with significant positive global IIP: Luxembourg, Belgium, Germany; outside EU: Switzerland and Norway.
  - Countries globally debtor at end-2008 despite positive positions vis-à-vis EU: Austria, France, Finland, Denmark and the UK.
- France and the UK acted as financial intermediaries:
  - France: financing from other financial centers including the US, largely in deposits and loans to its financial institutions; large debt bonds vis-à-vis other Euro area countries; extended loans to Spain and Italy.
  - UK: large net assets on the US; funding through deposits and loans was more important.

### Capital flows within EU/EA and GDP per capita correlation
- Dispersion of net external positions vis-à-vis the rest of the EU and Euro area increased between 2002 and 2008.
- Between 2002 and 2008, correlation between IIPs vis-à-vis the group and GDP per capita increased.
- Observation: strong correlation between IIPs vis-à-vis the rest of the zone and GDP per capita indicates capital appears to have flowed from wealthier countries to catching up economies.
- Prior reference: correlation of 0.43 between IIPs and GDP per capita at end 2004 (Lane and Milesi-Ferretti (2007)).

### Partner mismatch and financing concentration
- Bilateral trade relations are a poor proxy for bilateral financing relationships (France example: largest accumulated trade surplus with the UK yet UK was France’s second largest net creditor at end-2008).
- For most advanced EU economies, main gross financial partners concentrated among France, Germany, Luxembourg, Netherlands, the UK and the US.
- Emerging Europe showed more diversity among financial partners (e.g., Austria for Central and Eastern European Countries; Nordic countries for Baltic financing).

### Intra-Euro area dependency (2002–2008)
- Relative importance of intra Euro area net financing increased between 2002 and 2008.
- Among Euro area countries with large financing needs:
  - Ireland and Spain: role of intra Euro area net financing increased.
  - Greece and Portugal: intra Euro area financing slightly decreased but remained predominant.
- Specific country shares:
  - Ireland: in 2008 intra-EU and intra-Euro area net financing represented more than twice its global financing needs.
  - Spain: share of EU and intra-Euro area net financing roughly doubled between 2002 and 2008 (from less than 1/3 to around 2/3 for the Euro area; 40-45 percent to 80 percent for the EU).
  - Greece: share of Euro area financing slightly decreased to around 80 percent in 2008 (88 in 2002).
  - Portugal: share evolved from 2/3 to slightly below 60 percent.
- Creditor-side observations: several Euro area countries have IIPs vis-à-vis the rest of the Euro area equal to or exceeding their global IIPs (Germany broadly equal; France and Belgium exceed).

### Country groupings and divergence of net positions
- Country grouping definitions used in analyses:
  - (EACC): EA creditor countries — Austria, Belgium, Finland, France, and Germany.
  - (EALFN): EA countries with limited financing needs — Italy and the Netherlands.
  - (EASFN): EA countries with significant financing needs — Greece, Ireland, Portugal and Spain.
  - (LUXG): Luxembourg.
  - (CYMAS): Cyprus, Malta and Slovenia.
  - (EUCC): Non Euro area other EU Creditor countries — Denmark, Sweden and the UK.
  - (NMS): Non Euro area EU countries acceded in 2004 or 2007 — Poland, Romania, Lithuania, Estonia, Latvia, Bulgaria, Czech Republic, Slovak Republic, Hungary.
  - (SN): Other countries — Switzerland and Norway.
- IIPs vis-à-vis rest of EU show large—and growing—positions for EACC, EASFN (debtor side), and NMS (debtor side).
- Persistence of current account deficits in NMS and EASFN indicates IIPs vis-à-vis rest of EU increased as a result of net flows.

### Financing patterns and financial center specialization
- Patterns reflect geographic and instrument specializations; locational data caveats noted (overstates UK and Finnish banking exposure).
- Euro area creditor countries: strong links with other EU countries; largest debtors were EASFN financed via loans and debt bonds; EALFN financed mainly via debt.
- Luxembourg: largest total assets and liabilities as share of GDP; dense funding linkages; significant activity in debt and equity markets; largest net assets held outside EU (US); limited FDI — consistent with fund industry role.
- Non Euro area creditor countries (Denmark, Sweden, UK): UK as global financial intermediary with large financing from Rest of World; cross border loans predominate.
- CYMAS: net debtors to rest of EU; specialize as banking intermediaries receiving funds from Euro area (Austria, Greece, Germany) and Russia, channeling funds into the UK.
- NMS: more limited financial linkages and accumulated assets; importance of FDI financing, then deposits and loans, limited role of portfolio debt instruments.
- EASFN: limited role of FDI; predominant cross border funding via deposits and loans (mainly from Euro area and other EU creditor countries), followed by debt portfolio liabilities.
- EALFN: relative importance of debt portfolio financing larger; less important cross border loans financing.

### V. Conclusion — database construction and uses
- Constructed a database describing bilateral external assets and liabilities, excluding reserves and financial derivatives, for a number of European economies between 2000 and (text truncated in source).
- Data limitations and coverage:
  - No bilateral information on reserve assets and financial derivatives, accounting for around 10 percent of total assets and liabilities for most, but much larger for the UK.
  - Gap in coverage of portfolio investment, foreign direct investment, and other investment was around 15 percent of total assets and liabilities for most countries.
  - Real-time data not available due to reporting lags; data used were mostly available until end 2008.
  - New data sources and more participating countries should enable improvements over time.
- Dataset uses and potential improvements:
  - Constructed and used to map Euro area imbalances; other uses possible.
  - Accounting for portfolio and direct investment exposures, together with BIS consolidated data, could improve understanding of transmission channels across Euro Area countries.
  - As updated, dataset could help track consequences of ongoing deleveraging in a number of countries.
- Stylized facts summary:
  - EU and Euro Area are financially very integrated zones.
  - More than half of the allocable assets and liabilities of EU (respectively EA) countries are vis-à-vis other EU (respectively EA) countries.
  - Individual country positions vis-à-vis the rest of the EU only partially reflected their global IIP positions; some countries act as financial creditors within the EU while having net liabilities vis-à-vis the rest of the world.
  - Bilateral linkages only weakly related to bilateral trade positions.
- Correlation with GDP per capita:
  - Strong correlation within both EU and EA between accumulated IIP positions vis-à-vis the zone and relative GDP per capita — confirming capital flows from higher GDP per capita countries to lower GDP per capita countries.
- Differences across country groups:
  - Notable differences between financing of Euro area countries with significant financing needs and of EU 2004/2007 acceding Member states, both in origins and in nature.
  - Financing of Euro area countries with significant financing needs was less exclusively ensured by other EU or Euro area countries.
  - Inter-dependency across Euro area countries increased over time as persistently large current account positions translated into an increasing share of assets/liabilities held vis-à-vis other countries in the area.

### Policy implications and recommendations
- Excessively large accumulated current account imbalances are a matter of common interest among Euro area countries even absent contagion effects.
- Correcting these imbalances is likely to be painful not only for deficit countries but also for their partners within the Euro area, as they imply a negative correction in the value of their accumulated assets.
- Closer cooperation among countries to avoid the buildup of both public and private sector imbalances would be in all countries’ best interests.

### Appendices and data diagnostics (selected figures and tables)
- Appendix I: summary of data construction and gaps; Figures I-1 through I-5 document unallocated assets/liabilities by instrument and country, and shares of Reserves and Financial Derivatives (2001–2008 and 2008 snapshots).
- Appendix I — United Kingdom comparison (Table I-2) — IMF Database vs Pink Book (USD Billions) and Coverage (Percent):
  - Belgium — 176 — 181 — 97
  - France — 462 — 541 — 85
  - Germany — 500 — 602 — 83
  - Ireland — 338 — 399 — 85
  - Italy — 139 — 160 — 87
  - Luxembourg — 167 — 198 — 84
  - Netherlands — 383 — 422 — 91
  - Spain — 243 — 282 — 86
  - Norway — 54 — 48 — 5
  - Switzerland — 225 — 238 — 94
- Appendix II: locational versus consolidated BIS data — definitions, key differences quoted verbatim (e.g., “Inter-office positions are netted out of consolidated statistics, while they are included in locational data.”).
- Empirical comparison (Table II-1: end-2007, values in billions of US$):
  - Austria — 483 — 540 — 57
  - Belgium — 1,162 — 1,353 — 192
  - Finland — 102 — 8 — -94
  - France — 2,814 — 3,552 — 738
  - Germany — 3,561 — 4,257 — 696
  - Greece — 124 — 91 — -33
  - Ireland — 1,030 — 743 — -287
  - Italy — 648 — 1,111 — 464
  - Netherlands — 1,326 — 2,442 — 1,116
  - Portugal — 139 — 142 — 3
  - Spain — 613 — 1,220 — 607
  - Sweden — 337 — 691 — 354
  - Switzerland — 1,539 — 2,544 — 1,005
  - United Kingdom — 6,844 — 4,005 — -2,839
- Tables and figures in the source cover top financial and trade positions (2008), bilateral trade balances accumulated 1998–2008, top asset/liability partners, and Net Investment Positions with EU and Euro Area for 2002 and 2008 (percent of GDP).

*Section V concludes. Source: _wp10295 - Section V concludes.*

### Section V concludes.

### Section V concludes.

### Related literature: Intra-Euro area and intra-EU imbalances
- No reason why positions of constitutive countries or states should be in balance within a monetary union; anecdotal evidence for the US suggests significant interstate current account imbalances can be resolved without interstate balance-of-payments crises (Greenspan, 2004).
- Absence of an exchange rate premium helps facilitate financing of current account deficits.
- Current account deficits across Euro area and EU countries were for some time viewed as benign: capital flows originated from wealthier European countries (higher GDP and capital per capita) to catching up economies (lower GDP per capita), facilitating convergence; this contrasted with Feldstein and Horioka (1980) and Lucas (1990).
- Literature supporting benign view: Blanchard and Giavazzi (2002), Abiad, Leigh and Mody (2007), Schmidtz and von Hagen (2009), Bakker and Gulde (2010).
- Growing view that external imbalances could reflect internal unsustainable developments (excessive indebtedness of private or public agents), likely to result in painful adjustments: Gourinchas (2002), Ahearne and Pisani-Ferry (2006), Blanchard (2006), European Commission (2006 and following years), Guyon (2007).
- Deleveraging in some EU countries that experienced booms and large persistent current account deficits provided a form of closure to the debate; modalities for incorporating these imbalances into EU policy advice are being agreed at EU level.

### Related literature: Bilateral financial linkages at global level
- Motivation: growth in cross-border claims and global imbalances spurred documentation of external positions to understand financial inter-linkages and contagion.
- Lane and Milesi-Ferretti constructed an IIP database for 145 countries over 1970-2004, updated in 2007 (External Wealth of Nations, Mark II - EWNII).
- Kubelec and Sá (2010) constructed bilateral external assets and liabilities for 18 countries over 1980-2005; their sample included 6 large European countries but did not cover key creditor and debtor economies within Europe.
- Gravity models were used to estimate missing data in some studies.
- Lane and Shambaugh (2010) built a database for 1990-2005 focusing on major currency exposures; did not detail intra-EU or intra-Euro area bilateral assets and liabilities.
- Milesi-Ferretti, Strobbe and Tamirisa (2010) computed bilateral assets and liabilities for about 70 countries covering 15 large countries or country groupings; Euro area and Emerging Europe treated as whole.

### Related literature: Bilateral financial linkages in the EU and the Euro area
- Several efforts mapped financial inter-linkages among EU/Euro area countries using CPIS portfolio investment data (Lane and Milesi-Ferretti, 2005; Lane, 2006; Coeurdacier and Martin, 2006) showing a “Euro area bias”.
- Other studies used indirect measures (e.g., bilateral trade balances as proxy for capital flows — Schmidtz and von Hagen (2009)).
- Shared conclusion in some papers: capital seemed to flow from higher GDP per capita economies to lower GDP per capita ones (Blanchard and Giavazzi (2002), Abiad, Leigh and Mody (2007)).
- BIS consolidated bilateral data used in some studies (Árvai, Driessen, and Ötker-Robe (2009); Tressel (2010)) to investigate contagion channels.
- Gap identified: no single view encompassing various financing instruments and describing intra-Euro area and intra-EU bilateral positions.

### Methodology — Main principles
- Dataset coverage: 29 reporting European countries over the 2001 to 2008 period, with bilateral positions reported against over 200 partner countries.
- All data are in US dollars.
- Classification: follows the IIP classifications from the 5th revision of the Balance of Payments Manual (IMF, 1993), which follows the residency principle; external assets and liabilities are claims between a country’s residents and non-residents.
- Ideal components to capture: Reserve assets; foreign direct investment; portfolio investment; financial derivatives; Other investment.
- Missing bilateral components: no bilateral information on financial derivatives and foreign exchange reserves. The approach subtracts aggregate financial derivatives (from national sources reported to the IMF’s electronic data statistical system) and aggregate total foreign reserves assets (from the IMF’s Balance of Payment Statistics (BPTS)) from aggregate IIP figures to be decomposed.
- End-of-year bilateral stock positions are documented or estimated for:
  - Foreign direct investment;
  - Portfolio investment, divided into equity and debt securities; and
  - Other investment.
- Data represent stocks of assets and liabilities; variations over time may reflect transactions, revaluation effects, exchange rate changes, or revisions.

### Methodology — Data sources
- Direct investment: main source is the OECD, FDI positions by partner country. OECD Benchmark Definition recommends market value valuation, but in practice book values from balance sheets are generally utilized. Ongoing improvement via IMF’s Coordinated Direct Investment Survey (CDIS).
- Portfolio investment: main source is IMF’s Coordinated Portfolio Investment Survey (CPIS). CPIS provides year-end holdings of portfolio investment securities valued at market prices, cross-classified by issuer country. Participation is voluntary; some 75 economies participate. Participating countries report asset positions and are encouraged to report liabilities but most do not; liabilities are estimated in CPIS as a “mirror” from creditor positions. Incomplete participation leads to the sum of derived bilateral liabilities falling short of reported EDSS aggregate.
- Other investment: taken from BIS Locational Banking Statistics. To avoid double counting of portfolio instruments, the balance sheet items included are those reported under “loans and deposits”.
  - Other investment comprises Other investment from banks and Other investment from non-banks.
  - Other investment from banks: loans and deposits made by banks to non-residents in all currencies, including interbank borrowing and loans and inter-office balances. For BIS reporting countries, reported data is directly used. For BIS non-reporting countries, mirror data are used (country A banks’ claims vis-à-vis B provided through country B banks’ reporting, if B reports).
  - Other investment from non-banks: trade credit claims, financial leases, insurance and pension claims; derived using incomplete mirror data (limited to information provided by country B banks reporting liabilities vis-à-vis country A non-banks). If B is non-BIS reporting, no information on A non-banks’ claims vis-à-vis B; relations between non-banks and non-banks are missed.
- Use of locational BIS data (based on residency) is consistent with balance of payments data and IIP; consolidated BIS data on an ultimate basis is not consistent with residency-based IIP but represents the best snapshot of total bank exposures.
- The restricted nature of BIS figures constrains public reporting: bilateral Other investment positions of, or vis-à-vis, groupings of countries must be reported in a number of cases (notably for EU Member states that joined in 2004 or 2007 that do not document BIS locational statistics).

### Gaps and discrepancies
- Inherent data limitations: uncertainty about effective holder of a claim or liability and the economic nature of the claim when intermediary vehicles (mutual funds, trustees) are involved; described as introducing an “intermediation” and a “geographical veil” in CPIS data.
- Asymmetries between positions reported by counterparties are common (liabilities reported by country A vis-à-vis B may not match assets reported by country B vis-à-vis A), especially for FDI data. Guiding principle: for each country, retain data as reported by the country authorities.
- No bilateral information on reserves and financial derivatives; for most countries, the sum of these aggregates represents less than 10 percent of total assets plus liabilities, but exceptions include:
  - Some Member states that acceded in 2004 or 2007 where reserve assets are a substantial share of external assets.
  - The U.K., which plays a major role in the financial derivatives market (more than a third of its claims and liabilities).
- The sum of reported bilateral positions in FDI, portfolio, and other investment does not add up to gross asset and liability data stripped of financial derivatives and foreign reserve assets; gaps stem in particular from incomplete CPIS and BIS reporting and mirror data limitations.
- At end-2008, unallocated assets and liabilities by country:
  - For most EU advanced countries, gaps are around 15 percent of total assets and liabilities.
  - Gaps are particularly small (as a share) for the U.K.
  - Gaps are very large for Luxembourg (another reason for treating Luxembourg separately when dividing countries into groupings).
  - Switzerland’s unallocated total assets are actually negative, primarily because the sum of other investment bank bilateral claims exceeded reported aggregates.
  - Luxembourg reports minimal bilateral foreign direct investment and has large portfolio equity gaps (may reflect incomplete data on cross border mutual fund industry claims).
  - Data gaps in EU countries that joined in 2004 and 2007 vary: limited gaps in Estonia, Bulgaria and Romania; above-average discrepancies in Lithuania, Latvia, Cyprus and Malta.
  - Discrepancies are also large in Norway.
  - In absolute amount, the most significant discrepancies are for Luxembourg and the U.K.

*Section V concludes. Source: _wp10295 - Section V concludes.*

### Appendix I provides a summary of the data sources country by country and further

### _wp10295 - Appendix I provides a summary of the data sources country by country and further

### IV. STYLIZED FACTS — Overview
- Database constructed with data from BIS, OECD, ECB and national sources; and WEO.
- Data exclude financial derivatives and reserve assets.
- Analysis covers primarily 1999–2008 (financial integration charts) and up to 2009 for some regions.

### A. External financial integration of the EU and the EA is high
- Financial integration measured as sum of cross border assets and liabilities expressed as a percentage of GDP (excludes intra-zone claims in plotted series).
- Financial integration of the EU and the Euro area increased since Euro adoption.
- Euro Area external position: -17.7 percent of its GDP at end-2008, down to -12.4 percent at end Q2 2010.
- Euro Area assets and liabilities are relatively high compared with the United States and Japan despite a relatively small negative external position.
- Strong inflexion in Euro Area assets and liabilities in 2008 reflects a strong slow-down in transactions and revaluation effects due to price, non-exchange rate related, adjustments.
- Reintegrating financial derivatives would attenuate the 2008 inflexion: the increase in the value of total assets and liabilities in the form of financial derivatives roughly doubled in 2008 compared with 2007 (to around +550 bn Euros).
- Note on data consistency: IMF and ECB data broadly consistent; main difference arises from unallocated intra-EA financial derivatives and reserves. Overall difference limited to 2%-6% from 2001 to 2008 on the asset side and 1%-5% on the liabilities side in percent of total EA assets/liabilities as reported by the ECB.

### B. Internal financial integration among EU and EA countries is large
- Other EU countries constituted the first financial partner of both EU countries and Euro area countries, with around half of total assets and liabilities.
- For Euro area countries, assets and liabilities vis-à-vis other Euro area countries represented 40-45 percent of total assets or liabilities.
- Coverage is more complete on the asset side than on the liabilities side, resulting in a large negative unallocated IIP position for both EU and Euro area.
- After correcting for un-tracked assets:
  - EU assets and liabilities vis-à-vis EU countries represented around 2/3 of the allocated assets and liabilities.
  - Other Euro area countries accounted for a bit more than half of the tracked assets and liabilities of Euro area countries.

### C. Differences between EU/EA IIPs and global IIPs
- Correlation between net positions vis-à-vis EU countries (percent of GDP) and vis-à-vis Euro area countries: 0.80.
- Correlation between global IIPs and net positions vis-à-vis the EU at end-2008: 0.75.
- Examples of observed positions at end-2008:
  - Countries with significant positive global IIP: Luxembourg, Belgium, Germany; outside EU: Switzerland and Norway.
  - Countries that were globally debtor at end-2008 despite positive positions vis-à-vis EU: Austria, France, Finland, Denmark and the UK.
- Germany: limited financing from abroad (Japan, Switzerland and France); large savings channeled to UK, Spain, Ireland and the US, often as credits and loans.
- France and the UK acted as financial intermediaries:
  - France: financing from other financial centers including the US, largely in deposits and loans to its financial institutions; held large debt bonds vis-à-vis other Euro area countries; extended loans to Spain and Italy.
  - UK: large net assets on the US; funding through deposits and loans was more important.

### D. Capital flows within EU/EA: from advanced to lower GDP per capita countries
- The database provides net financial asset positions vis-à-vis the rest of the EU and vis-à-vis the rest of the Euro area.
- Analysis focuses on EU and Euro area countries in 2002 and in 2008 (excludes Luxembourg and Ireland as outliers in some comparisons).
- Dispersion of net external positions vis-à-vis the rest of the EU and Euro area increased between 2002 and 2008.
- Between 2002 and 2008, correlation between IIPs vis-à-vis the group and GDP per capita increased.
- Observation: strong correlation between IIPs vis-à-vis the rest of the zone and GDP per capita indicates capital appears to have flowed from wealthier countries to catching up economies.
- Correlation between global IIP and GDP per capita among EU (EA) countries also exists but is weaker; referenced prior overall correlation of 0.43 between IIPs and GDP per capita at end 2004 (Lane and Milesi-Ferretti (2007)).

### E. Top net creditors/debtors ≠ top trade partners or gross financial partners
- Bilateral trade relations are a poor proxy for bilateral financing relationships.
  - Example: France’s largest accumulated trade surplus (1998–2008) was with the UK and largest deficit with Germany; UK was France’s second largest net creditor at end-2008 while Germany was the fifth largest debtor.
  - Germany showed more similarity between trade and financial links in some cases but not consistently.
- For most advanced EU economies, main gross financial partners concentrated among France, Germany, Luxembourg, Netherlands, the UK and the US.
- Emerging Europe showed more diversity among financial partners (e.g., Austria frequently among largest partners for Central and Eastern European Countries; importance of Nordic countries for Baltic financing).
- Database contains banking data on a locational basis:
  - Claims and liabilities of a foreign bank operating in the UK are reported as UK claims and liabilities, overstating the UK role relative to ultimate consolidated risk.
  - For risk mapping, consolidated BIS data ideally limited to deposits and loans should be combined with bilateral portfolio and FDI positions.

### F. Intra-Euro area dependency is high and increased for several countries (2002–2008)
- Relative importance of intra Euro area net financing increased between 2002 and 2008 (Table 4 evidence).
- Among Euro area countries with large financing needs:
  - Ireland and Spain: role of intra Euro area net financing increased.
  - Greece and Portugal: intra Euro area financing slightly decreased but remained predominant.
- Ireland: in 2008 intra-EU and intra-Euro area net financing represented more than twice its global financing needs; in 2002 Euro area financing only covered its then smaller financing needs.
- Spain: share of EU and intra-Euro area net financing roughly doubled between 2002 and 2008 (from less than 1/3 to around 2/3 for the Euro area; 40-45 percent to 80 percent for the EU).
- Greece: share of Euro area financing slightly decreased to around 80 percent in 2008 (88 in 2002).
- Portugal: share evolved from 2/3 to slightly below 60 percent.
- Creditor side: several Euro area countries have IIPs vis-à-vis the rest of the Euro area equal to or exceeding their global IIPs (Germany broadly equal; France and Belgium exceed), reflecting net accumulated assets vis-à-vis the rest of the Euro area comparable or exceeding global financing capacities.
- Euro nominal effective exchange rate appreciated by slightly less than 20 percent between end 2002 and 2008, reducing the relative share of liabilities not denominated in Euros.

### G. Divergence of intra-European net positions across country groupings
- Country groupings defined by patterns vis-à-vis rest of EU (net positions as a share of GDP):
  - (EACC): EA creditor countries — Austria, Belgium, Finland, France, and Germany (positive IIPs vis-à-vis rest of EU, excluding Luxembourg).
  - (EALFN): EA countries with limited financing needs — Italy and the Netherlands (moderately negative IIP vis-à-vis other EU countries).
  - (EASFN): EA countries with significant financing needs — Greece, Ireland, Portugal and Spain (large negative IIP vis-à-vis rest of EU).
  - (LUXG): Luxembourg.
  - (CYMAS): Cyprus, Malta and Slovenia (joined EU in 2004; part of Euro area).
  - (EUCC): Non Euro area other EU Creditor countries — Denmark, Sweden and the UK.
  - (NMS): Non Euro area EU countries acceded in 2004 or 2007 — Poland, Romania, Lithuania, Estonia, Latvia, Bulgaria, Czech Republic, Slovak Republic, Hungary.
  - (SN): Other countries — Switzerland and Norway.
- IIPs vis-à-vis rest of EU show large—and growing—positions for:
  - EA Creditor Countries (EACC).
  - Euro Area countries with significant Financing Needs (EASFN) — debtor side.
  - Non Euro Area Member States that acceded in 2004 and 2007 (NMS) — debtor side.
- Persistence of current account deficits in NMS and EASFN indicates IIPs vis-à-vis rest of EU increased as a result of net flows.

### H. Financing patterns reflect specialization of financial centers and debtor needs
- Patterns reflect geographic and instrument specializations; locational data caveats noted (overstates UK and Finnish banking exposure; understates others).
- Euro area creditor countries: strong links with other EU countries; little instrument specialization; largest debtors were EASFN countries financed via loans and debt bonds, and EALFN countries mainly via debt.
- Luxembourg: largest total assets and liabilities as share of GDP; dense funding linkages (cross border deposits and loans, balanced across zones); significant activity in debt and equity markets; largest net assets held outside EU (US); limited FDI — consistent with fund industry role.
- Non Euro area creditor countries (Denmark, Sweden, UK): reflect UK as global financial intermediary with large financing from Rest of World; largest creditor position vis-à-vis United States; cross border loans predominate.
- CYMAS (Cyprus, Malta, Slovenia): net debtors to rest of EU; second largest share of assets and liabilities to GDP; specialize as banking intermediaries receiving funds from Euro area (Austria, Greece, Germany) and Russia, channeling funds notably into the UK.
- NMS: more limited financial linkages and accumulated assets; importance of FDI financing, then deposits and loans (including intra-group), limited role of portfolio debt instruments — consistent with Gulde and Bakker (2010) for EU9.
- EASFN: limited role of FDI; predominant cross border funding via deposits and loans (mainly from Euro area and other EU creditor countries), followed by debt portfolio liabilities.
- EALFN: relative importance of debt portfolio financing larger; less important cross border loans financing.

### V. CONCLUSION — Database construction
- Constructed a database describing bilateral external assets and liabilities, excluding reserves and financial derivatives, for a number of European economies between 2000 and (text truncated in source).

*Source: _wp10295 - Appendix I provides a summary of the data sources country by country and further (IMF PDF content provided).*

### 2008. We documented some inherent limitations to the data and estimated the size of our

### _wp10295 - 2008

### Data limitations and coverage
- No bilateral information on reserve assets and financial derivatives, accounting for around 10 percent of total assets and liabilities, but much larger for the UK.
- Gap in coverage of portfolio investment, foreign direct investment, and other investment was around 15 percent of total assets and liabilities for most countries.
- Real-time data not available due to reporting lags; data used in constructing the database were mostly available until end 2008.
- New data sources and a growing number of participating countries should enable this dataset to be improved over time.

### Dataset uses and potential improvements
- Constructed and used to map Euro area imbalances; other uses are possible.
- Accounting for portfolio and direct investment exposures, together with BIS data on a consolidated basis, could improve understanding of possible transmission channels across Euro Area countries.
- As it is updated, this dataset could help track the consequences of the ongoing deleveraging in a number of countries.

### Stylized facts on financial integration
- The EU, and the Euro Area, are both financially very integrated zones.
- More than half of the allocable assets and liabilities of EU (respectively EA) countries are vis-à-vis other EU (respectively EA) countries.
- Individual country positions vis-à-vis the rest of the EU only partially reflected their global IIP positions; some countries act as financial creditors within the EU while having net liabilities vis-à-vis the rest of the world.
- Bilateral linkages were related to bilateral trade positions only to a very limited extent.

### Correlation with GDP per capita
- Strong correlation, both within the EU and within the EA, between accumulated IIP positions vis-à-vis the countries of the respective zone and relative GDP per capita—confirming that capital has indeed flown within both zones from countries with higher GDP per capita to countries with lower GDP per capita.

### Differences across country groups
- Notable differences between financing of Euro area countries with significant financing needs and of EU 2004 or 2007 acceding Member states, both in origins and in nature.
- Financing of Euro area countries with significant financing needs was less exclusively ensured by other EU or Euro area countries.
- Database suggests inter-dependency across Euro area countries increased over time, as persistently large current account positions translated into an increasing share of assets of Euro area creditor countries, or liabilities of Euro area debtor countries, held vis-à-vis other countries in the area over time.

### Policy implications and recommendations
- Excessively large accumulated current account imbalances are a matter of common interest among Euro area countries even absent contagion effects.
- Correcting these imbalances is likely to be painful not only for deficit countries but also for their partners within the Euro area, as they imply a negative correction in the value of their accumulated assets.
- Closer cooperation among countries to avoid the buildup of both public and private sector imbalances would be in all countries’ best interests.

*Source: _wp10295 - 2008*

### References

### _wp10295 - References

### Bibliographic sources included
- Extensive reference list of works cited, including IMF working papers, articles, books, ECB/BIS/OECD/European Commission publications, and mimeos.
- Representative entries (authors and years) include:
  - Abiad, A, D. Leigh, and A. Mody, 2007
  - Ahearne, A., and J. Pisani-Ferry, 2006
  - Árvai, Z., K. Driessen, and İ. Ötker-Robe, 2009
  - Bakker, B., and Gulde, A,-M., 2010
  - Barrios, S., Iversen, P., Lewandowska, M., and R. Setzer, 2009
  - Berger, H. and V. Nitsch, 2010
  - BIS, 2008
  - Blanchard, O. and F. Giavazzi, 2002
  - Coeurdacier, N., and P. Martin, 2006
  - Decressin, J., and E. Stavrev, 2009
  - European Central Bank, 2005
  - European Commission, 2006, 2008, 2009, 2010
  - Feldstein M., and C. Horioka, 1980
  - Felettigh A., and P. Monti, 2008
  - Fidora, M., M. Fratzscher and C. Thimann, 2006
  - Gourinchas, P., 2002
  - Guyon, T., 2007
  - International Monetary Fund, 1993; 2003; 2005
  - Kubelec, C. and F. Sá, 2010
  - Lane, P and G. M. Milesi–Ferretti (multiple entries: 2001, 2005, 2007a, 2007b, 2008)
  - Lucas, R, 1990
  - Milesi-Ferretti, G.M., F. Strobbe and N. Tamirisa, 2010
  - Shadler, S., Drummond, P., Kuijs, L., Murgasova, Z, and R. Van Elkan, 2005
  - Sinn, S., 1990
  - Tesar, L., and I. Werner, 1995
  - Tressel T., 2010

### Data tables and scope (tables 2A–3F and table 4)
- Coverage:
  - Top financial and trade positions for European and selected non-European countries, primarily for 2008.
  - Bilateral trade balances accumulated from 1998 to 2008.
  - Top seven positive (negative) net foreign asset positions at end-2008.
  - Top seven foreign asset positions and top seven foreign liability positions at end-2008.
  - Net investment positions with the European Union and the Euro Area for 2002 and 2008 (percent of GDP).
- Table groupings and focal country sets:
  - Table 2A. Top Financial and Trade Positions: Euro Area Creditor Countries, 2008
  - Table 2B. Top Financial and Trade Positions: Euro Area Countries with Limited Financing Needs, 2008
  - Table 2C. Top Financial and Trade Positions: Euro Area Countries with Significant Financing Needs, 2008
  - Table 2D. Top Financial and Trade Positions: Other Euro Area Countries, 2008
  - Table 2E. Top Financial and Trade Positions: Other European Union Creditor Countries, 2008
  - Table 2F. Top Financial and Trade Positions: Switzerland and Norway, 2008
  - Table 3A–3F. Top Gross Financial Partners by country groups, 2008 (Total Assets 1/; Total Liabilities 2/)
  - Table 4. Net Investment Positions of European Countries, 2002 and 2008 (Percent of GDP)
- Notes and definitions provided in tables:
  - Visegrad Countries = Czech Republic, Hungary, Poland and Slovakia.
  - Baltics = Estonia, Latvia and Lithuania.
  - Other Offshore Centers exclude the Cayman Islands, Guernsey and Jersey (where bilateral data are available).
  - Rest of the World is composed of BIS non-reporting countries excluding the Visegrad Countries, the Baltics, Other Offshore Centers and International Organizations.
  - 1/ Net foreign asset positions at end-2008 (top seven positive/negative).
  - 2/ Top seven bilateral trade balances, accumulated from 1998 to 2008.
  - Table 3 notes: 1/ Top seven foreign asset positions at end-2008. 2/ Top seven foreign liability positions at end-2008.
  - Table 4 notes: 1/ Net investment position with the European Union (based on corresponding year's members). 2/ Net investment position with the Euro Area (based on corresponding year's members). 3/ Includes financial derivatives and reserve assets.

### Figures and visual analyses (Figures 3A–4F)
- Figures present graphical decomposition of net foreign assets and gross positions (assets and liabilities) by country groups for 2008, with component breakdowns:
  - Components shown: Direct Investment; Portfolio Equity; Portfolio Debt; Other Investment.
  - Country/group labels and legends used consistently across figures:
    - EACC - Euro area creditor countries
    - EALFN / EACLFN - Euro area countries with limited financing needs
    - EASFN / EACSFN - Euro area countries with significant financing needs
    - LUX - Luxembourg
    - CYMAS - Cyprus, Malta and Slovenia
    - EUCC - Other EU creditor countries
    - NMS - Non-Euro area countries that acceded in 2004 or 2007
    - SN - Switzerland and Norway
    - USA - United States
    - JPN - Japan
    - BRIC - Brazil, Russia, India and China
    - ROW - Rest of the World
- Specific figures:
  - Figure 3A: Net Foreign Assets and Gross Positions: European Union Countries, 2008 (Percent of GDP)
  - Figure 3B: Net Foreign Assets and Gross Positions: Euro Area Countries, 2008 (Percent of GDP)
  - Figure 3C–3H: Country-group and individual-country breakdowns (e.g., Austria, Belgium, Finland, France, Germany; Italy and Netherlands; Greece, Ireland, Portugal and Spain; Luxembourg; Cyprus/Malta/Slovenia; Denmark/Sweden/United Kingdom)
  - Figure 3I: Non-Euro Area EU Countries that Acceded in 2004 or 2007, 2008 (Percent of GDP)
  - Figure 3J: Norway and Switzerland, 2008 (Percent of GDP)
  - Figures 4A–4F: Net Foreign Positions by country groups and specific countries, 2008 (Percent of GDP), with components labeled ID (net direct investment), IPE (net portfolio equity), IPD (net portfolio debt), IO (net other investment)

### Representative numeric and temporal markers in the unit
- Year references and ranges explicitly preserved:
  - End-2008 (net foreign asset positions and other end-2008 statistics)
  - Accumulation period for bilateral trade balances: 1998 to 2008
  - Comparative years for net investment positions: 2002 and 2008
- Data sources explicitly stated:
  - IMF database constructed with data from BIS, OECD, ECB and national sources; and WEO.

*Content unit: References, tables, and figures from _wp10295 (data and bibliographic listings as presented in the source PDF).*

### Appendix I.   Summary of Data Construction and Gaps

### Appendix I.   Summary of Data Construction and Gaps

### Data sources and construction
- Reserves and Financial derivatives: “no bilateral information on Reserves and Financial derivatives– which represent less than 10 percent of total assets and liabilities for most, but not all, countries (see Figure I-5).”
- International sources used for International Investment Positions components:
  - Bilateral direct investment: the OECD, FDI positions by partner country.
  - Portfolio investment (equity and debt): the IMF’s Coordinated Portfolio Investment Survey (CPIS).
  - Other investment: BIS locational statistics.
- Structure of direct investment data:
  - OECD used in most cases (Czech Republic, Finland, France, Germany, Greece, Ireland, Italy, Luxembourg, Norway, Netherlands, Poland, Portugal, Slovakia, Spain, Sweden and Switzerland).
  - National sources used (Austria, Denmark, Estonia, Hungary, Latvia, Lithuania, UK).
  - ECB or Eurostat sources used (Bulgaria, Cyprus, Malta, Romania and Slovenia).
  - For Belgium, total direct investment data is derived from mirror positions.
- CPIS participation and limitations:
  - Participation to CPIS is voluntary, with 75 participants so far.
  - Some key players are missing (in particular China and several oil exporters), creating a gap between portfolio liabilities as reported in EDSS and liabilities the authors are able to allocate.
- Other investment complications:
  - For Other investment, BIS reporting countries’ locational data are used directly for bilateral banking claims and liabilities.
  - For BIS non-reporting countries, mirror data are used on both asset and liability sides.
  - Missing elements when relying on mirror data for non-reporting countries:
    - The concerned country’s bilateral bank claims and liabilities on banks from other non-reporting countries.
    - Bank claims and liabilities on all (non-resident) non-banks.
  - Mirror data for Other investment, non-banks capture claims and liabilities of non-resident banks on resident non-banks but miss:
    - Positions involving non-resident banks from BIS non-reporting countries.
    - Relations between non-banks and non-banks.
  - The authors choose not to impute or assume structures for the missing parts and classify missing parts (or excesses) as “unallocated”.
- UK alternative source note:
  - The Annual Pink Book published by the Office of National Statistics covers banks and non-banks financial positions; bilateral coverage is limited to a smaller number of countries and total coverage is roughly similar to data derived from BIS sources. Comparison provides an indication of the size of non-banks to non-banks positions for a selected group of countries.
- BIS Guidelines: “Guidelines to the international locational banking positions, BIS, updated in December 2008.”

### Gaps, unallocated amounts, and their evolution
- General patterns:
  - Tracking unallocated data provides a benchmark of possible measurement errors and coverage gaps.
  - Figure I-1 reports gaps in assets and liabilities under Portfolio investment, Foreign direct investment, and Other investment for each country, at end-2007.
  - The size of unallocated assets (as a percent of total external assets) tends to decrease over time, reflecting improved coverage and reporting; 2008 marks a rebound consistent with bilateral data not yet fully available and data revisions improving consistency over time.
  - By asset category, a large part of unallocated amounts is due to limited coverage of Other investment.
  - Unallocated liabilities are overall slightly larger than unallocated assets, with a notably larger share of unallocated portfolio liabilities, reflecting uncertainty on the final holder and reliance of CPIS liabilities data on derived mirror data from participating countries’ reported assets.
- Figures referenced (as constructed in the source):
  - Figure I-1: Composition of Total Unallocated Assets (Percent of total assets), weighted by country GDP, 2001–2008; asset categories shown: Other investments, Portfolio investments, Direct investments.
  - Figure I-2: Composition of Total Unallocated Liabilities (Percent of total liabilities), weighted by country GDP, 2001–2008; liability categories shown: Other investments, Portfolio investments, Direct investments.
  - Figure I-3: Unallocated Assets and Liabilities, 2008 (Percent of total) — shows country-level breakdown of unallocated asset and liability shares by instrument (Direct investment; Portfolio investment, equity; Portfolio investment, debt; Other investment; Unallocated).
  - Figure I-4: Unallocated Assets and Liabilities, 2008 (USD Billions) — country-level USD amounts by instrument.
  - Figure I-5: Share of Reserves and Financial Derivatives, 2008 (Percent of total assets and liabilities) — country-level shares showing Reserve assets; Securities held as reserve assets by other countries and international organizations' holdings; Financial derivative assets; Financial derivative liabilities.

### United Kingdom: Comparison of Other Investment Data, 2008 (Table I-2)
- Sources: IMF database constructed with data from BIS, OECD, ECB and national sources; and UK Pink Book from ONS.
- Table fields: IMF Database (USD Billions); Pink Book (USD Billions); Coverage 1/ (Percent)
- Country rows (exact numbers preserved):
  - Belgium — 176 — 181 — 97
  - France — 462 — 541 — 85
  - Germany — 500 — 602 — 83
  - Ireland — 338 — 399 — 85
  - Italy — 139 — 160 — 87
  - Luxembourg — 167 — 198 — 84
  - Netherlands — 383 — 422 — 91
  - Spain — 243 — 282 — 86
  - Norway — 54 — 48 — 5
  - Switzerland — 225 — 238 — 94

### Appendix II — Locational versus Consolidated BIS data (summary)
- Definitions and coverage:
  - BIS locational banking statistics: aggregate international claims and liabilities of all banks (including affiliates of foreign banks) resident in a reporting country; residency principal follows balance of payments guidelines; main items include (i) loans and deposits; (ii) debt securities; and (iii) other assets and liabilities (e.g. equity shares, derivatives).
  - BIS consolidated banking statistics: compiled on a group world-wide (headquarter) basis; include exposures of foreign affiliates (subsidiaries and branches) of the same banking group; collected on both an “immediate borrower” and “ultimate risk” basis; consolidated data on an ultimate basis provide the best snapshot of true cross-border bank exposures.
- Key differences affecting claims (items preserved verbatim where presented):
  - “Inter-office positions are netted out of consolidated statistics, while they are included in locational data.”
  - “Local claims in local currency are large for many banks, and are only included in consolidated data.”
  - “Locational statistics cover banks' offices/affiliates that are located in the BIS reporting countries (43 countries) whereas consolidated banking statistics by nationality cover banks' headquartered in BIS reporting countries, but includes information on the positions of their offices/affiliates in all countries in the World.”
  - “Reporting institutions in locational statistics also include in many cases non-banks (e.g. brokers and dealers) but only banks in the case of consolidated statistics.”
  - “Locational banking statistics include all on-balance sheet items (instruments) but consolidated foreign or international claims do not include on-balance sheet derivatives claims (positive market value). Such positions are reported separately (see BIS Table 9C).”
- Empirical comparison (Table II-1: Comparison of External and Foreign Claims of Reporting Banks, end-2007; values in billions of US$; table source: BIS and staff estimates):
  - Notes: Columns correspond to Locational external claims; Consolidated foreign claims 1/; Consolidated minus locational claims 2/.
  - Country rows (exact numbers preserved):
    - Austria — 483 — 540 — 57
    - Belgium — 1,162 — 1,353 — 192
    - Finland — 102 — 8 — -94
    - France — 2,814 — 3,552 — 738
    - Germany — 3,561 — 4,257 — 696
    - Greece — 124 — 91 — -33
    - Ireland — 1,030 — 743 — -287
    - Italy — 648 — 1,111 — 464
    - Netherlands — 1,326 — 2,442 — 1,116
    - Portugal — 139 — 142 — 3
    - Spain — 613 — 1,220 — 607
    - Sweden — 337 — 691 — 354
    - Switzerland — 1,539 — 2,544 — 1,005
    - United Kingdom — 6,844 — 4,005 — -2,839
  - Table footnotes:
    - 1/ Table 2A of the BIS Locational Banking Statistics; external positions of banks.
    - 2/ Table 9D of BIS Consolidated Banking Statistics; consolidated foreign claims, ultimate risk basis.
- Practical implication for the database construction:
  - To avoid double counting with portfolio (CPIS) data, the authors take only the loans and deposits component in locational statistics.

*Source: Appendix I and Appendix II text as provided in the content unit.*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2010/_wp10295.pdf_