## _wp12284 - Executive Summary

## Source details

**Canonical URL:** [_wp12284 - Executive Summary](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp12284.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp12284.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp12284.pdf.json)

---

### Methodology and dataset
- Tracks US$ 42 trillion of sovereign debt across 24 major advanced economies, quarterly from 2004 to 2011.
- Investor classes: domestic central banks, domestic banks, domestic nonbanks, foreign official sector, foreign banks, foreign nonbanks.
- Coverage: Represents 98 percent of the general government debt of all advanced economies as of end-2011, based on October 2012 WEO figures.
- Common sovereign debt definition: general government gross debt on a consolidated basis (currency and deposits; securities other than shares; loans).
- Valuation principles:
  - Tradable securities valued at market prices in most international data sources; where necessary converted to face value using country- and investor-specific information (Table 4 examples).
  - Country-specific valuation practices preserved exactly (selected entries include: Australia: Book value / Book value / Market value; Belgium: Book value / Book value / Market value; United Kingdom: Market value / Market value / Market value; United States: Book value / Book value / Face value).
  - All U.S. dollar statistics converted into respective national currencies using end-quarter exchange rates from the IFS for BIS and QEDS.
- Primary data sources: BIS International Locational Banking Statistics, IMF International Financial Statistics (IFS), IMF/World Bank Quarterly External Debt Statistics (QEDS); additional national sources where noted.
- Estimation approach by investor type preserved:
  - Foreign holdings from QEDS (general government securities held externally + general government loans held externally), with valuation adjustments where required.
  - Foreign official sector: foreign central bank holdings (COFER and CPIS based assumptions), SMP holdings allocated proportionally, and foreign official loans from QEDS.
  - Foreign banks: BIS creditor-side claims with share of public sector claims estimated via consolidated statistics; reported without ad-hoc valuation adjustments.
  - Foreign nonbanks: residual = total foreign holdings − foreign official − foreign banks.
  - Domestic holdings: total debt − foreign debt holdings; domestic central bank and domestic bank holdings from IFS or national sources; domestic nonbanks = residual.
- Robustness checks:
  - Comparison with national comprehensive holder data for 12 advanced economies shows alignment where national data are general government at face value.
  - For euro area, 70 percent of domestic and foreign nonbank estimates explained by ECB insurance/pension fund and investment fund statistics; remainder likely non-euro holdings and household/corporate holdings.
  - Consistency checks and adjustments applied to ensure logical relations (e.g., foreign central bank holdings ≤ total foreign holdings).

### Key findings on investor trends (post-crisis, 2004–11)
- Aggregate foreign ownership of advanced economy sovereign debt rose from US$ 5 trillion to US$ 12 trillion between 2004 and 2011.
- Foreign investors increased their share for most advanced economies, except the euro area periphery.
- Euro area periphery (high-spread euro area) foreign exposure fell by about US$ 400 billion between mid-2010 and end-2011, with outflows across foreign central banks, foreign banks, and foreign nonbanks.
- Foreign bank ownership remained roughly stable at around US$ 2 trillion over 2004–11; increases driven mainly by foreign central banks and foreign nonbanks.
- Share changes (sample averages 2004–11):
  - Low-spread euro area: foreign share increased from 50 to 62 percent.
  - Other advanced economies: foreign share increased from 20 to 31 percent.
  - Traditional reserve countries: foreign share increased from 14 to 21 percent.
  - High-spread euro area: foreign share rose until end-2009 then fell from 60 percent to 35 percent during the turmoil.
- Investor-type flows mid-2010–end-2011:
  - Foreign central banks: outflows ~US$ 60 billion (mainly Ireland and Spain); increased holdings of low-spread euro area by ~US$ 240 billion.
  - Foreign banks: reduced exposure to high-spread euro area by ~US$ 170 billion; increased exposure to low-spread euro area (notably Germany) by ~US$ 50 billion.
  - Foreign nonbanks: cumulative net outflows ~US$ 160 billion (most in late 2011); inflows to low-spread euro area ~US$ 110 billion partly offset outflows.
- Domestic response:
  - Domestic banks filled most of the financing gap when foreign private outflows occurred.
  - Euro area domestic bank holdings of own government debt rose from 15 percent to about 20 percent of GDP.
  - Home bias for euro area banks rose from 57 percent (end-2007) to 69 percent (end-2011); in Greece, Italy, Spain own-government share approaches 100 percent of overall euro government holdings.
  - Drivers of bank demand include reduced private loan demand, deleveraging, regulatory incentives (zero percent risk weights), Basel III liquidity demand, and central bank operations (e.g., ECB LTROs).

### Implications for refinancing risk and domestic financial stability
- Higher reliance on foreign investors raises government refinancing risk: foreign private investors less likely to roll over during confidence shocks or global risk aversion.
- Domestic banks stepping in to absorb outflows deepens sovereign-bank linkages, increasing potential for negative feedback loops:
  - Market concerns about sovereign risk can undermine bank confidence if banks hold large sovereign portfolios or if bank funding costs are tied to sovereign yields.
  - Domestic banks may require sovereign support, reinforcing sovereign-bank interdependence and raising the risk of self-fulfilling debt crises.
- Policy relevance: investor-base dynamics should complement standard debt sustainability analyses (DSA) to capture market-driven sovereign risk components.

### Sovereign Funding Shock Scenarios (FSS) — framework and scenarios
- Objective: assess vulnerability to sudden withdrawal of foreign private investors over a one-year horizon; to be used alongside DSA.
- Behavioral and domestic-absorption assumptions:
  - Trigger: foreign private investors (foreign banks and foreign nonbanks) pull out.
  - Domestic institutional investors assumed able to reallocate 1 percent of their assets under management to domestic sovereign debt.
  - Remaining gap filled entirely by domestic banks.
  - Domestic central bank assumed not to buy government debt during the stress episode.
  - Projected fiscal deficit and projected supply of debt held constant across scenarios.
  - Three EU/IMF program countries are excluded from scenarios.
- Three scenarios (in increasing severity):
  - Scenario 1 (100 percent rollover): α = 0, β = 100 percent, γ = 0 — foreign private investors stop financing their share of the deficit but roll over all maturing debt over the next year.
  - Scenario 2 (0 percent rollover): α = 0, β = 0, γ = 0 — foreign private investors stop financing their share of the deficit and do not roll over maturing debt.
  - Scenario 3 (0 percent rollover plus 30 percent sale): α = 0, β = 0, γ = 30 percent — foreign private investors do not roll over and sell 30 percent of remaining exposure (30 percent based on experience in Greece, Ireland, Portugal).
- Simulation outputs and thresholds:
  - Metrics: additional sovereign debt domestic banks would need to absorb (percent of bank assets); resulting sovereign debt held by domestic banks (percent of bank assets).
  - Countries most sensitive (additional absorption > 5 percentage points of bank assets under severest shock): Belgium, Czech Republic, Italy, Slovenia.
  - Countries highly resilient: Australia, Denmark, Sweden, Switzerland.
  - Countries where domestic banking system would end up with > 15 percent of bank assets in own sovereign debt under severest shock: Italy, Slovenia.
  - Countries already with high bank exposure and thus vulnerable: Japan, Czech Republic.
- Limitations:
  - May underestimate shocks if foreign investors also withdraw from non-sovereign domestic assets.
  - Assumes foreign private share of short-term debt proportional to share in total debt; deviations can amplify or reduce shocks.
  - Does not differentiate maturity structures across foreign banks vs nonbanks; data limitations constrain refinement.

### Investor Base Risk Index (IRI) — methodology and findings
- Purpose: capture likelihood that different investor types will suddenly outflow; index range 0–100 (higher = more prone to sudden outflow).
- Construction:
  1. Calculate historical correlations between log changes in investor holdings and log changes in 10-year sovereign bond yields over 2008Q3–2011Q4.
  2. Transform correlations linearly so highest positive correlation → risk score 0; highest negative correlation → risk score 100.
  3. Aggregate IRI: weighted average of investor risk scores using investor shares in total debt stock.
- Risk scores by investor type (from Table 6):
  - Domestic central bank: correlation 0.148 → Risk Score 0
  - Domestic bank: correlation 0.056 → Risk Score 26
  - Domestic nonbank: correlation 0.093 → Risk Score 16
  - Foreign central bank (excluding SMP holdings): correlation 0.097 → Risk Score 14
  - Foreign bank: correlation -0.162 → Risk Score 87
  - Foreign nonbank: correlation -0.207 → Risk Score 100
- Key IRI findings and stylized classification (example as of end-2009):
  - High investor-base-risk countries (Quadrant II: high debt, high IRI): high-spread euro area countries, Austria, France.
  - High debt, low IRI (Quadrant I): Germany, Japan, U.K., U.S.
  - Low debt, high IRI (Quadrant IV): Finland, Netherlands, New Zealand, Slovenia.
  - Low-risk investor base countries: Australia, Japan, Switzerland, U.S.
  - IRI limitations: based on correlations during 2008Q3–2011Q4; correlations can change; very low IRI may signal excessive reliance on domestic investors (domestic banks) which is a separate vulnerability.

### Foreign Investor Position Index (FIPI) — methodology and findings
- Purpose: measure foreign investors’ country allocation relative to a benchmark portfolio; focuses on foreign investor holdings excluding foreign official loans and SMP holdings of foreign central banks.
- Calculation: FIPI = country’s share in foreign investor holdings − country’s share in benchmark portfolio (benchmarks constructed separately for euro area and non-euro area groups).
- End-2011 patterns:
  - Euro area: foreign holdings overweight Germany and France; underweight Italy and Spain.
  - Non-euro area: foreign holdings overweight U.K. and Australia; underweight Korea and Switzerland.
  - Outliers noted: U.S. overweight by 2555 basis points; Japan underweight by -2822 basis points (figure notes preserved).
- Limitations: sensitive to benchmark choice; does not distinguish loans vs securities exposures.

### Policy implications and recommendations
- Integrate sovereign investor-base analysis into:
  - Public debt management: monitor investor relations, conduct scenario analysis, maintain flexible issuance strategies, consider investor asset-allocation behavior.
  - Financial sector supervision: emphasize macro-financial risks from bank sovereign holdings; strengthen banking system backstops; enhance loss-absorbing capital; perform systemic stress tests with credible sovereign scenarios; coordinate with public debt managers.
  - Sovereign risk measurement: complement standard DSA with investor-base risk indicators (IRI, FIPI) and sovereign funding shock scenarios.
- Data and transparency:
  - Reduce data gaps in securities holding statistics and holder classification, consistent with the G20 Data Gaps Initiative and BIS-ECB-IMF Handbook on Securities Statistics (Part 2); support ECB Centralized Securities Database initiative for the euro area.
- Further research areas:
  - Empirically examine links between sovereign yield levels/volatility and investor holdings.
  - Analyze effects of investor-base diversity on sovereign market liquidity.
  - Incorporate investor-base dynamics into unwinding of crisis-related monetary and financial policies.
  - Assess impacts of new financial regulations on investor holdings and sovereign debt markets.

### Limitations, caveats, and data-constraints (selected)
- COFER currency breakdown covers about half of foreign exchange reserves; unallocated reserves assumed to have same currency composition as allocated reserves.
- SMP country decomposition and distribution across Eurosystem central banks not publicly disclosed—country-level SMP estimates based on simplifying assumptions.
- Foreign bank estimates may be overstated for countries with large international financial centers due to reliance on consolidated BIS statistics.
- Foreign and domestic nonbank estimates are residuals and may include measurement errors; foreign nonbanks may include holdings of foreign banks resident in non-reporting countries (e.g., China, Russia).
- Changes in holdings reflect transactions, price revaluations, exchange rate changes, and statistical reclassifications; conversion to face value mitigates price-revaluation effects but relies on top-down approximations.
- Exchange rate effects considered minor for sample given that 80 percent or more of outstanding debt is denominated in local currency as of end-2011.

*Source: _wp12284 - Executive Summary (authors' calculations and figures/tables as presented in the source).*

### Executive Summary ......................................................................................................

### _wp12284 - Executive Summary

### I. Introduction
- Identifies the beginning of the document at page 5.
- Sets up the scope that follows in subsequent sections (see II–VI).

### II. Motivation
- Located on page 6.
- Presents the rationale for examining sovereign investor bases and related risks.

### III. Methodology for Sovereign Investor Base Estimates
- Spans pages 8–21 with four subparts:
  - A. Overview (page 8)
    - Figure 1: Summary of Methodology (listed under Figures).
    - Table 1: Sample of Countries (listed under Tables).
    - Table 2: Summary of Data Sources (listed under Tables).
  - B. Estimation Methodology for Each Investor Type (page 14)
    - Table 4: Valuation of Debt Securities in Various Databases (listed under Tables).
    - Table 5: Estimation of Foreign Central Bank Holdings of Government Debt, end-2011 (listed under Tables).
  - C. Robustness Checks and Adjustments (page 19)
  - D. Limitations and Extensions (page 20)
- Additional methodological material referenced:
  - Table 3: Construction of General Government Gross Debt on a Consolidated Basis for Non-European Countries, end-2011 (listed under Tables).

### IV. Key Trends in The Post-Crisis Period
- Covered on pages 22–33 with four subparts:
  - A. How Did Foreign Demand Change? (page 23)
    - Figure 3: Foreign Holdings of Government Debt by Country of Origin, 2004–11.
    - Figure 4: Advanced Economies: Foreign Investors as an Investor Class, 2004–11.
    - Figure 5: Advanced Economies: Foreign Share in Government Debt Stock, 2004–11.
    - Figure 6: Advanced Economies: Foreign Net Purchases and Sales of Government Debt, 2008–11.
  - B. How Did Foreign Demand Change by Investor Type? (page 28)
    - Figure 9: Euro and Non-Euro Area Countries: Foreign Net Purchases of Government Debt.
    - Figure 10: Advanced Economies: Country Allocation of Foreign Investors, 2004–11.
    - Figure 11: Advanced Economies: Cumulative Net Foreign Purchases of Government Debt.
  - C. How Did Domestic Investors React to Falls in Foreign Demand? (page 31)
    - Figure 13: Foreign Investors Position and Sovereign Downgrades, 2004–11.
  - D. How Did Domestic Bank Demand Change More Generally? (page 32)
    - Figure 14: Advanced Economies: Bank Holdings of Own Government Debt, 2004–11.
    - Figure 15: Euro and Non-Euro Area: Bank Holdings of Own Government Debt, 2008–11.
    - Figure 16: Euro Area: Bank Holdings of Own Government Debt Securities, 2007–11.
- Box 1: Country Groups (appears under Boxes and linked to these analyses).

### V. Sovereign Investor Base Risk Indicators
- Covered on pages 34–46 with three subparts:
  - A. Sovereign Funding Shock Scenarios (page 35)
    - Figure 17: Advanced Economies: Sovereign Funding Shock Scenarios.
    - Box 2: Methodology of Sovereign Funding Shock Scenarios.
  - B. Investor Base Risk Index (page 40)
    - Figure 18: Advanced Economies: Investor Base Risk Index, 2008–11.
    - Table 6: Risk Scores by Investor Type.
    - Figure 19: A Stylized Framework for Sovereign Risk Analysis.
    - Figure 20: Advanced Economies: Application of Investor Base Risk Index, end-2009.
  - C. Foreign Investor Position Index (page 45)
    - Figure 21: Advanced Economies: Foreign Investor Position Index (FIPI), end-2011.

### VI. Conclusion
- Located on page 47.
- Summarizes findings and implications drawn across the preceding analyses.

### Supplementary Materials Listed in the Source
- References (page 59).
- Annex Figures:
  - Annex Figure 1: Advanced Economies: Holders of Government Debt, 2004–11.
  - Annex Figure 2: Advanced Economies: Cumulative Purchases of Government Debt, since end-2007.
- Tables and Figures (as enumerated above) supporting the analysis and methodology.

*Source: _wp12284 - Executive Summary (pages and figures/tables as listed in the source table of contents).*

### EXECUTIVE SUMMARY

### EXECUTIVE SUMMARY

### Methodology and dataset
- Introduces a methodology to compile internationally comparable estimates of sovereign debt holdings.
- Dataset characteristics:
  - Tracks US$ 42 trillion of sovereign debt.
  - Covers 24 major advanced economies.
  - Investor classes: domestic central banks, domestic banks, domestic nonbanks, foreign official sector, foreign banks, foreign nonbanks.
  - Period: 2004 to 2011 on a quarterly basis.
  - Represents 98 percent of the general government debt of all advanced economies as of end-2011, based on October 2012 WEO figures.
  - Uses a common definition of sovereign debt: general government gross debt on a consolidated basis (currency and deposits; securities other than shares; loans).
  - Data compiled either in face value or adjusted for valuation changes, as appropriate.
  - Foreign investor holdings estimated separately for foreign official sector, foreign banks, and foreign nonbanks.

### Key findings on investor trends
- Foreign investors continue to increase their share in sovereign debt markets for most advanced economies, apart from the euro area periphery.
- In the euro area periphery, foreign investors reduced exposure to sovereign debt by about US$ 400 billion between mid-2010 and end-2011.
  - Outflows occurred across foreign central banks, foreign banks, and foreign nonbanks.
  - Outflows varied in scale, speed, and timing, reflecting different risk characteristics.
- In all cases with large foreign private outflows, domestic banks stepped in to fill most of the financing gap not filled by the foreign official sector.
- Advanced economy banks’ exposure to own government debt has been on the rise since 2008.

### Implications for refinancing risk and domestic financial stability
- Higher reliance on foreign investors can increase government refinancing risk because foreign private investors may be less willing to roll over holdings during confidence shocks or global risk aversion episodes.
- Domestic banks mitigating short-term refinancing gaps can deepen sovereign-bank linkages, potentially jeopardizing domestic financial stability over the long term.
  - Market concerns about sovereign risk could undermine confidence in banks if banks hold large amounts of government debt or if their funding costs are closely tied to sovereign yields.
  - Domestic banks may require sovereign support, reinforcing sovereign-bank interdependence and raising the risk of negative feedback loops and, in extreme cases, self-fulfilling debt crises.

### Risk framework, scenarios, and indices
- Proposes sovereign funding shock scenarios (FSS) to assess sovereigns’ vulnerability to sudden investor outflows; FSS can be used alongside standard debt sustainability analyses (DSA).
- Introduces two risk indices to assess and monitor vulnerability to shifts in investor behavior:
  - Investor Base Risk Index (IRI).
  - Foreign Investor Position Index (FIPI).
- The framework and indices aim to measure a country’s ability to manage sudden outflows of foreign investors by relying more on the local investor base, notably domestic banks.

### Motivation and analytical focus
- Motivated by the need to understand demand-side dynamics for sovereign debt as general government gross debt of advanced economies rose from 74 to 105 percent of GDP over 2007–11 and, based on WEO projections, will stabilize at around 110 percent of GDP in 2014.
- Emphasizes cross-country comparability to track global shifts in demand, including “safe haven flows” linked to the euro periphery turmoil.
- Focuses on:
  - Governments’ refinancing risks due to shifts in investor base.
  - Potentially harmful sovereign-bank linkages threatening domestic financial stability.
- Does not address diversity of the investor base or debt sustainability issues directly.

### Policy implications and recommendations
- Calls for integrating robust sovereign investor base analysis into:
  - Public debt management.
  - Financial sector supervision.
  - Sovereign risk measurement.
- Highlights the need to reduce data gaps in securities holding statistics, in line with the G20 Data Gaps Initiative.

*Source: _wp12284 - EXECUTIVE SUMMARY.*

### 2011. Annex I provides the full picture of investor holdings since 2004. Annex II provides a

### _wp12284 - 2011. Annex I provides the full picture of investor holdings since 2004. Annex II provides a

### Valuation principles and adjustments
- Tradable securities are, in principle, valued at market prices in most international data sources (external debt statistics, BIS international banking statistics, IFS), while national accounting rules may require different valuation methods.
- Based on metadata, the valuation method used by each country in its reporting is identified and, where appropriate, securities holdings in market value are converted to face value (Table 4).
- Valuation adjustment for debt securities: convert market value to face value where necessary using country- and investor-specific information when available (e.g., investor-specific market-to-face ratios for Belgium, Italy, Netherlands, Portugal; transactions-based estimates for Greece and U.K.).
- For EU countries: market-to-face ratio calculated by comparing government financial accounts (market value, from ECB) with Maastricht debt definition (face value, from Eurostat).
- For non-EU countries: use Barclays Capital Global Treasury Index, which provides both market and face values.
- BIS and IMF compilation guides are referenced for valuation principles.

### Table of valuation practices (excerpted sources and treatments)
- Data sources: IFS / National Sources, BIS QEDS, External debt.
- Selected country treatments as reported (exact entries preserved):
  - Australia: IFS/National Sources = Book value; BIS QEDS = Book value; External debt = Market value
  - Austria: Market value; Book value; Market value
  - Belgium: Book value; Book value; Market value
  - Canada: Book value; Book value; Face value
  - Czech Republic: Book value; N/A; Market value
  - Denmark: Market value; Market value; Market value
  - Finland: Book value; Book value; Market value
  - France: Book value; Book value; Market value
  - Germany: Book value; Book value; Market value
  - Greece: Book value; Market value; Market value
  - Ireland: Book value; Book value; Face value
  - Italy: Book value; Book value; Market value
  - Japan: Book value; Book value; Market value
  - Korea: Book value; Book value; Market value
  - Netherlands: Book value; Book value; Market value
  - New Zealand: Book value; N/A; Face value
  - Portugal: Book value; Book value; Market value
  - Norway: Book value; Book value; Market value
  - Slovenia: Book value; N/A; Market value
  - Spain: Book value; Book value; Market value
  - Sweden: Book value; Book value; Market value
  - Switzerland: Book value; Book value; Market value
  - United Kingdom: Market value; Market value; Market value
  - United States: Book value; Book value; Face value
- Note: “Book value” indicates an accounting treatment based on national accounting standards; trading portfolio generally at market value, banking portfolio at nominal value.
- All U.S. dollar statistics are converted into respective national currencies using end-quarter exchange rates from the IFS for BIS and QEDS.

### Primary data sources and overall methodology
- Primary data sources: BIS International Locational Banking Statistics, IMF International Financial Statistics (IFS), IMF/World Bank Quarterly External Debt Statistics (QEDS).
- All sources are based on the residency principle of the investor, include comparable definitions of general government, and use similar definitions of debt instruments.
- Valuation of securities may differ across sources and require adjustments as detailed above.

### Estimation methodology by investor type
- Foreign holdings:
  - Estimated from QEDS as sum of (i) general government securities held externally (money market instruments; bonds and notes) and (ii) general government loans held externally.
  - QEDS records debt securities at market value (except Canada, Ireland, U.S.) and are converted to face value based on valuation adjustments (Tables 2 and 4).
  - New Zealand external debt obtained from Reserve Bank of New Zealand (not in QEDS).
- Foreign official sector holdings:
  - Consist of (i) government securities held by foreign central banks as foreign exchange reserves; (ii) government securities held by foreign central banks within the Eurosystem as part of the Securities Markets Program (SMP) (only SMP holdings of foreign central banks included); and (iii) foreign official loans.
- Foreign central bank holdings estimation:
  - United States: U.S. Treasury International Capital (TIC) reporting system used; TIC is hybrid of market and face value (bonds at market value; bills at face value). Valuation adjustment applied only to Treasury bond component using QEDS (face value) and TIC (market value).
  - Other countries: COFER and CPIS databases used. Assumptions:
    - Unallocated COFER reserves have same currency composition as allocated reserves.
    - 80 percent of foreign exchange reserves in a currency consist of government securities of the issuing country(ies).
    - CPIS distribution used to allocate COFER EUR and “other currencies” to individual euro area and other advanced economies.
- Securities Markets Program (SMP) holdings:
  - Total SMP size obtained from ECB (Consolidated financial statements of the Eurosystem: Asset item 7.1).
  - Country composition of SMP assumed proportional to outstanding Maastricht debt of SMP countries.
  - Distribution of SMP purchases within Eurosystem assumed in line with national central banks’ contribution to ECB capital base.
  - SMP initially targeted Greece, Ireland, Portugal (starting 2010Q2), then Italy and Spain (starting 2011Q3).
  - SMP holdings recorded on cost basis; valuation adjustment applied to estimate face value.
- Foreign official loans:
  - Include bilateral official loans and multilateral loans (EU/IMF program loans for Greece, Portugal, Ireland after 2010; project loans from IFIs such as EIB and World Bank).
  - External loan statistics in QEDS used to capture these loans.
- Foreign bank holdings:
  - Primarily estimated from BIS International Locational Banking Statistics (creditor-side claims of reporting banks).
  - Share of external claims on government sector estimated using BIS International Consolidated Banking Statistics on an immediate borrower basis: ratio = total international claims on public sector (immediate borrower) divided by sum of total international claims on public sector (immediate borrower) and nonbank private sector (immediate borrower).
  - Valuation: BIS recommends loans at face value and securities at market value; national practices differ. Given banking book vs trading book distinctions and limited information on shares, BIS statistics reported as provided without ad-hoc valuation adjustments.
- Foreign nonbank holdings:
  - Estimated as residual: total foreign holdings minus foreign official sector holdings minus foreign bank holdings.
- Domestic holdings:
  - Domestic investor holdings = total debt minus foreign debt holdings.
  - Domestic central bank and domestic bank holdings obtained from IFS (gross claims of monetary authorities; “other depository corporations”) or national sources where IFS not available (Australia, Canada, Korea, New Zealand, Norway, Switzerland, U.K.).
  - Valuation treatment for domestic banks similar to foreign banks; no valuation adjustment applied on IFS statistics except for countries reporting all securities at market value (Austria, Denmark, United Kingdom).
- Domestic nonbank holdings:
  - Estimated as residual: total domestic holdings minus domestic banks and domestic central bank holdings.

### Robustness checks and adjustments
- Comparison with national data:
  - Compared estimates with national comprehensive holder data for 12 advanced economies. Despite differences in sector, instrument, and valuation, estimates, including domestic nonbanks, are in line with national data where national data are for general government at face value.
- Comparison with ECB statistics for euro area institutional investors:
  - For euro area countries, total nonbank holdings (domestic and foreign) compared with ECB’s Insurance Corporations and Pension Funds Statistics and Investment Fund Statistics.
  - Result: 70 percent of the domestic and foreign nonbank estimates for euro area countries can be explained by ECB data; remaining part likely reflects non-euro area holdings and domestic household and non-financial corporate holdings not in ECB data.
- Basic consistency checks and adjustments:
  - Checked that foreign central bank holdings are lower than total foreign holdings of debt securities; exceptions (Australia, Ireland) adjusted to comply with general trend.
  - Checked that foreign bank holdings are less than total foreign holdings minus foreign official holdings; exceptions (Australia, Ireland, Korea, Switzerland) adjusted by distributing the difference between foreign banks and foreign nonbanks in line with previous-quarter ratios.
  - Confirmed domestic nonbank estimates are positive.

### Limitations, caveats, and extensions
- Limitations due to data constraints:
  - COFER currency breakdown covers about half of foreign exchange reserves; assumption made that unallocated half has same composition as allocated half.
  - Assumption that foreign exchange holdings in a currency mainly reflect government debt securities of the issuing country.
  - SMP country decomposition and distribution across Eurosystem central banks not publicly disclosed; country-level SMP estimates are based on simplifying assumptions and may differ from actual figures.
  - Foreign bank estimates rely partially on BIS consolidated statistics to estimate share of claims on government; may overstate foreign bank holdings for countries with large international financial centers.
  - Foreign and domestic nonbanks are residuals and may include measurement errors; foreign nonbanks may include holdings of foreign banks resident in countries not reporting locational statistics (e.g., China, Russia).
- Investor transactions vs holdings:
  - Changes in holdings between periods reflect (i) transactions; (ii) price revaluations; (iii) exchange rate changes; (iv) statistical reclassifications.
  - Dataset compiled at face value or converted to face value mitigates price revaluation effects, but valuation adjustments are top-down approximations.
  - Exchange rate changes considered minor for the sample because most outstanding debt is denominated in local currency (80 percent or more as of end-2011).
  - Statistical reclassifications noted (e.g., Italy: inclusion of Cassa Depositi e Prestiti in domestic banks in IFS from 2006; changes in BIS reporting banks in 2008–09; Canadian adoption of IFRS in 2011).
- Country-of-origin extension:
  - Dataset can estimate country-of-origin of investors using CPIS under the assumption that country of origin of government debt investors aligns with country of origin of all external debt holders; more plausible when government debt is a high share of external debt (e.g., Greece, Japan with government debt in external debt >90 percent).
  - Such estimates reflect exposures on an immediate borrower basis and may not capture ultimate risk if holdings are custodial.

### Key trends in the post-crisis period (2004–11)
- Research questions explored:
  - How did foreign demand for government debt and related refinancing risk change for advanced economies after the global financial crisis?
  - How did domestic banks’ demand for own government debt change and what are implications for bank-sovereign linkages and domestic financial stability?
  - Focused on two episodes: global financial crisis (2008–09) and euro area debt turmoil (starting 2010).
- Aggregate foreign ownership trends:
  - Aggregate foreign ownership of advanced economy sovereign debt increased from US$ 5 trillion to US$ 12 trillion between 2004 and 2011.
  - Increase driven by large purchases by foreign central banks and foreign nonbanks; foreign bank ownership remained relatively unchanged at around US$2 trillion.
- Share of foreign investors by group (sample period 2004–11, average shares reported):
  - Low-spread euro area countries: foreign share increased from 50 to 62 percent.
  - “Other advanced economies”: foreign share increased from 20 to 31 percent.
  - Traditional reserve countries: foreign share increased from 14 to 21 percent.
  - High-spread euro area countries: foreign share stopped rising at end-2009 and then declined from 60 percent to 35 percent during the euro area debt turmoil.
- Implications highlighted for subsequent analysis:
  - Which foreign investors (foreign central banks, foreign banks, foreign nonbanks) shifted portfolios and in which directions?
  - In countries where foreign demand dried up (notably high-spread euro area), did domestic investors (banks vs nonbanks) take up the slack, and what were consequences for sovereign-bank linkages and domestic financial stability?

*Source: Authors' calculations and supporting data as presented in the document.*

### Box 1. Country Groups

### Box 1. Country Groups

### Grouping criteria
- Four groups among advanced economies in the sample are constructed along the following dimensions: (i) exchange rate regime (independent or currency union); (ii) reserve currency status, and (iii) perceived credit risk in 2011.
- Group definitions use sovereign CDS spreads as of August 2011 and other country characteristics.

### High-spread euro area
- Definition: euro area countries identified by the IMF in the April 2012 GFSR as having sovereign CDS spreads of more than 200 basis points as of August 2011.
- Countries: Belgium, Greece, Ireland, Italy, Portugal, and Spain.
- Note: Although spreads for Belgium have come down in 2012, it is included in this group as our analysis is mainly about trends until end-2011.
- Aggregate share: Together, these countries account for nearly half of the euro area government debt market and 11 percent of the total advanced economy sovereign debt outstanding, at end-2011.

### Low-spread euro area
- Definition: the other euro area countries in the sample.
- Countries: Austria, Germany, France, Finland, Netherlands, and Slovenia.
- Credit-rating note: Except for Slovenia, all were rated AAA by at least two out of the three major credit rating agencies, as of end-

*Source: _wp12284 - Box 1. Country Groups*

### 2011. This group accounts for 14 percent of the

### _wp12284 - 2011. This group accounts for 14 percent of the

### Investor composition of advanced economy sovereign debt (end-2011)
- High-spread euro area, low-spread euro area, traditional reserve countries, and other advanced economies segmentation.
- Numerical shares:
  - This group accounts for 14 percent of the total advanced economy sovereign debt outstanding.
  - Traditional reserve countries (Japan, Switzerland, United Kingdom, and United States) represent 68 percent of total advanced economy sovereign debt outstanding.
  - Other advanced economies (Australia, Canada, Czech Republic, Denmark, New Zealand, Norway, Korea, and Sweden) represent 6 percent of total advanced economy sovereign debt outstanding.
  - Total advanced economy sovereign debt universe: Total: US$42 trillion.

### Foreign investor behavior and shifts, 2008–11
- Crisis-period patterns:
  - During the global financial crisis (2008–09), foreign investors increased holdings of virtually all advanced economy sovereign debt except Czech Republic, Japan, Korea, and Switzerland.
  - During the euro area debt turmoil (starting 2010), foreign investors differentiated markedly across sovereigns.
- Cumulative shifts mid-2010 to end-2011:
  - Foreign investors cumulatively reduced exposure to high-spread euro area sovereign debt by about US$ 400 billion.
  - Non-euro area countries, particularly “other advanced economies,” received large inflows as a percent of their total debt stock.
- Country-level example:
  - Foreign investors were net buyers of sovereign debt of Greece, Ireland, and Portugal up until 2010Q1, then sentiment changed quickly.

### How did foreign demand change by investor type
- Foreign central banks:
  - Had negligible exposure to the high-spread euro area before the turmoil (less than 7 percent).
  - Outflows of about US$ 60 billion during 2010Q2–2011Q4 (mainly from Ireland and Spain).
  - Increased holdings of low-spread euro area government debt by about US$ 240 billion during 2010Q2–2011Q4.
  - Share of “other advanced economies” in foreign central bank advanced-economy sovereign debt holdings rose from 2 to 6 percent since 2007.
- Foreign banks:
  - Began cutting exposure to high-spread euro area countries from early 2010.
  - Timeline: began reducing exposure to Greece in 2010Q1; Portugal and Italy in 2010Q2; Ireland and Spain in 2010Q3.
  - Reduction in exposure to high-spread euro area countries of around US$ 170 billion during 2010Q2–2011Q4.
  - Increased exposure to low-spread euro area countries (especially Germany) by around US$ 50 billion.
  - As of end-2011, estimated half of sovereign exposure of foreign banks was to the United States and Germany alone.
- Foreign nonbanks:
  - Pulled out a cumulative US$ 160 billion during 2010Q2–2011Q4, most in late 2011.
  - Inflows into low-spread euro area countries of US$ 110 billion partly offset outflows.
  - Net sellers of euro government debt during 2010Q2–2011Q4; net buyers of debt of traditional reserve countries and, to a lesser extent, other advanced economies.

### Investor reactions to sovereign credit ratings and implications
- Sovereign downgrades (multi-notch) for high-spread euro area countries appear associated with accelerated foreign sales once ratings approached the lower end of AA.
- Despite reductions to high-spread countries, increased exposure to other euro area countries—particularly Germany—more than offset sales, helping explain relative euro resilience during the turmoil.

### Domestic investor response and sovereign-bank linkages
- Filling the gap from foreign outflows:
  - Falls in foreign demand were filled mostly by domestic banks in countries with severe foreign outflows (Greece, Ireland, and Portugal).
  - Domestic nonbanks also provided additional demand in some countries (e.g., Belgium retail bond issuance in November 2011).
- Domestic bank holdings of own government debt:
  - Government debt held by domestic banks in the euro area increased from 15 percent to about 20 percent of GDP.
  - Home bias for euro area banks (own sovereign as percent of overall euro area government holdings):
    - End-2007: 57 percent.
    - End-2011: 69 percent.
  - In some high-spread euro area countries (Greece, Italy, Spain), the ratio of own government debt to overall euro government debt holdings is now close to 100 percent.
- Drivers of increased bank demand for own-government debt (non-exhaustive list):
  - The global recession reducing private sector loan demand.
  - Bank deleveraging and incentives from zero percent regulatory risk weights on own sovereign debt.
  - Basel III and other regulatory changes increasing demand for high-quality liquid assets.
  - Central bank operations (e.g., ECB LTROs) and changes in collateral criteria; contrasting effects from quantitative easing in Japan, U.K., and U.S.
- Risks:
  - Stronger sovereign-bank interdependence and potential negative feedback loops affecting domestic financial stability.
  - Possible crowding out of bank credit to the private sector, with adverse long-term growth implications.

### Sovereign investor-base risk indicators and funding-shock framework
- Objective:
  - Develop three risk indicators to capture vulnerability to sudden investor outflows: sovereign funding shock scenarios, investor base risk index, and foreign investor position index.
- Sovereign Funding Shock Scenarios (FSS) — methodology and assumptions:
  - Trigger: foreign private investors (foreign banks and foreign nonbanks) pull out of a country’s sovereign debt market over a one year horizon.
  - Domestic institutional investors (insurance companies, pension funds, investment funds) assumed able to reallocate one percent of their assets under management to domestic sovereign debt.
  - The remainder of the gap is filled entirely by domestic banks.
  - Domestic central bank is assumed not to directly buy government debt during the stress episode.
  - The three euro area countries under the EU/IMF program are excluded from the scenarios.
  - Projected overall fiscal deficit and projected supply of debt over the next year remain the same in all scenarios.
- Three shock scenarios (in increasing severity):
  - Scenario 1 (100 percent rollover): Foreign private investors stop financing their share of the government deficit but roll over all debt holdings coming due over the next year.
  - Scenario 2 (0 percent rollover): Foreign private investors stop financing their share of the government deficit and do not roll over the debt maturing over the next year.
  - Scenario 3 (0 percent rollover plus 30 percent sale): Foreign private investors stop financing their share of the government deficit; do not roll over the debt maturing over the next year; and sell a portion (30 percent) of their remaining exposure.
    - Note: 30 percent is based on recent experience of distressed euro area countries (Greece, Ireland, and Portugal), where foreign private investors reduced their holdings by 30 percent, on average, over a one year period after their peak level.

*Source: Authors' calculations and figures in the provided content.*

### Box 2: Methodology of Sovereign Funding Shock Scenarios

### Box 2: Methodology of Sovereign Funding Shock Scenarios

### Methodology overview
- The sovereign funding shock scenarios (FSS) rely on three investor decision parameters:
  - α: foreign private investors' contribution to funding of the overall fiscal deficit over the next year.
  - β: foreign private investors' rollover of short-term government debt by residual maturity (i.e., debt maturing over the next year).
  - γ: foreign private investors' sale of long-term government debt holding by residual maturity.
- α determines the extent foreign private investors contribute to net financing needs; α and β together determine contribution to gross financing needs; γ determines liquidation beyond gross financing needs.
- Fiscal funding needs are calculated using projected general government deficit from the WEO and the stock of short-term government securities (remaining maturity basis) from BIS Securities Statistics.
- Foreign share of fiscal funding needs and rollover needs is based on the foreign share of the total debt stock as of end-2011.

### Calculation approach and assumptions
- Decomposition: Gross financing needs (GFN) are decomposed into fiscal deficit (D) and maturing debt (M), and further by investor type into FD/OD (fiscal deficit financed by foreign private/other investors) and FM/OM (maturing debt held by foreign private/other investors).
- Investor share assumption: foreign private investors’ share of financing fiscal deficit and short-term debt is proportional to their share in the total debt stock (FS/TS/OS notation used).
- Behavioral assumptions: foreign private investors contribute to α percent of the annual fiscal funding needs, roll over β percent of their share of maturing debt, and sell γ percent of their remaining stock.
- Domestic absorption assumptions:
  - Debt taken up by domestic banks is derived from foreign outflows net of domestic nonbank absorption.
  - Debt taken up by domestic nonbanks is assumed to be one percent of the financial assets of domestic institutional investors (assets under management data from OECD Institutional Investors Database).
- Resulting domestic bank holding at time t (DBS) computed from prior domestic bank holdings plus the additional debt taken up.

### Scenario parameters (specific scenarios used)
- Scenario 1: α = 0, β = 100 percent, γ = 0.
- Scenario 2: α = 0, β = 0, γ = 0.
- Scenario 3: α = 0, β = 0, γ = 30 percent.

### Key simulated results and interpretation
- Assessment metrics:
  - Additional sovereign debt domestic banks would need to take up under scenarios, expressed as percent of bank assets.
  - Resulting level of sovereign debt held by domestic banks, expressed as percent of bank assets.
- Findings (aggregate and country examples):
  - Countries most sensitive (additional amount of debt to be absorbed > 5 percentage points of assets under severest shock — 0 percent rollover plus 30 percent sale): Belgium, Czech Republic, Italy, Slovenia.
  - Countries highly resilient: Australia, Denmark, Sweden, Switzerland (due to lower share of foreign private investors and larger banking sectors).
  - Countries where domestic banking system would end up with more than 15 percent of bank assets in own sovereign debt under severest shock: Italy, Slovenia.
  - Countries that already have high bank exposure to own sovereign debt and thus are vulnerable: Japan, Czech Republic.
- General cross-country vulnerabilities:
  - Greater sensitivity where investor base has a higher share of foreign private investors, banking sector assets are smaller, and government debt maturity is shorter.

### Limitations of the FSS
- Results may underestimate shocks if foreign investors also withdraw from non-sovereign debt markets (domestic bank securities or corporate debt); positive correlation could increase total impact.
- Assumes foreign private investors’ share of short-term debt is proportional to share in total debt; if foreign investors hold a disproportionately larger share of short-term debt (e.g., Japan), outflows may be larger; if disproportionately smaller (e.g., Slovenia), outflows may be lower.
- Does not address different maturity structures and rollover needs between foreign banks and nonbanks; banks generally hold shorter duration debt than nonbanks. EBA stress-test maturity information can partly address this, but EBA data have limited coverage and low frequency.

---

### Investor Base Risk Index (IRI) — methodology
- Purpose: reflect likelihood of sudden outflows by different investor types; index ranges 0 to 100 (higher = more prone to sudden outflow). Complementary to FSS: FSS assesses vulnerability under hypothetical outflow; IRI captures likelihood of event materializing.
- Construction steps:
  1. Historical correlations: calculate historical correlation between log changes in investor holdings and log changes in 10-year sovereign bond yields (Bloomberg), estimated over the post-crisis period (2008Q3–2011Q4). Data on investor holdings from authors’ dataset.
     - Aggregate finding: domestic investors and foreign central banks tend to increase holdings when sovereign yields rise; foreign private investors tend to reduce holdings.
  2. Risk scores per investor: apply linear transformation to correlation coefficients so highest positive correlation → risk score 0 and highest negative correlation → risk score 100.
     - Correlations and corresponding risk scores shown in source Table 6:
       - Domestic central bank: correlation 0.148 → Risk Score 0
       - Domestic bank: correlation 0.056 → Risk Score 26
       - Domestic nonbank: correlation 0.093 → Risk Score 16
       - Foreign central bank (excluding SMP holdings): correlation 0.097 → Risk Score 14
       - Foreign bank: correlation -0.162 → Risk Score 87
       - Foreign nonbank: correlation -0.207 → Risk Score 100
  3. Aggregate IRI: weighted average of investor risk scores using each investor’s share in total debt stock as weights. Examples:
     - Country fully held by domestic central bank → IRI = 0.
     - Country fully held by foreign nonbanks → IRI = 100.
- Simpler alternative: foreign private share alone (score domestic investors and foreign central banks as 0; foreign banks and nonbanks as 100) produces broadly similar results as sensitivity analysis.

### IRI findings and applications
- Euro area countries that later received EU/IMF support in 2010–11 exhibited high investor base risk as early as 2010; all euro area countries show a high IRI reflecting high financial integration and high share of foreign private investors.
- Greece, Portugal, Ireland identified with especially risky investor bases in 2010; Spain and Italy show moderately high scores due to high share of domestic banks/nonbanks; their risk scores decline after SMP purchases began in 2011Q3.
- Austria and Finland have high IRI despite lower levels of debt.
- Low-risk investor base countries: Australia, Japan, Switzerland, U.S.
- Stylized framework: combining IRI (demand-side) with supply-side indicators (e.g., debt-to-GDP ratios) classifies countries into four quadrants:
  - Quadrant I: High debt, low investor base risk (e.g., Germany, Japan, U.K., U.S. as of end-2009).
  - Quadrant II: High debt, high investor base risk (e.g., high-spread euro area countries, Austria, France as of end-2009).
  - Quadrant III: Low debt, low investor base risk (e.g., Switzerland).
  - Quadrant IV: Low debt, high investor base risk (e.g., Finland, Netherlands, New Zealand, Slovenia as of end-2009).
- Limitations of IRI:
  - Relies on historical correlations during 2008Q3–2011Q4; correlations may change over time.
  - Based on simple correlation coefficients; could be augmented by more formal empirical analysis.
  - A very low IRI may indicate excessive reliance on domestic investors (especially domestic banks), which can be a separate source of vulnerability.

---

### Foreign Investor Position Index (FIPI) — methodology and findings
- Purpose: estimate position of foreign investors relative to outstanding debt portfolio; focuses on foreign investor holdings excluding foreign official loans and SMP holdings of foreign central banks.
- Benchmarking:
  - Different benchmark portfolios for euro area and non-euro area countries, constructed as total debt outstanding of all countries within the group with weights equal to each country’s share in group total debt.
  - FIPI = country’s share in foreign investor holdings minus country’s share in benchmark portfolio.
- Findings (end-2011):
  - Euro area: foreign holdings overweight Germany and France; underweight Italy and Spain.
  - Non-euro area: foreign holdings overweight U.K. and Australia; underweight Korea and Switzerland.
  - Outliers: U.S. significantly overweight (U.S.: 2555 basis points in figure note); Japan significantly underweight (Japan: -2822 basis points).
  - "Other advanced economies" generally overweight except Korea.
- Limitations:
  - Index varies with benchmark choice; alternative benchmarks (e.g., by credit rating) can give different perspectives.
  - Does not differentiate loans vs. securities exposures; could be extended to marketable securities only.

---

### Policy implications and conclusions (selected)
- Dataset scope: methodology tracks US$ 42 trillion of advanced economy sovereign debt held by six investor classes—domestic central banks, domestic banks, domestic nonbanks, foreign official sector, foreign banks, foreign nonbanks. Dataset covers 24 countries, 2004 to 2011, quarterly.
- Main cross-cutting findings: rising share of foreign investors in sovereign debt markets and rising exposure of banks to own government debt after the global financial crisis.
- Policy recommendations (selected):
  - Public debt managers should monitor investor relations and government refinancing risk through scenario analysis; retain flexible issuance strategies and focus on investor asset-allocation strategies. Reference: principles in the "Stockholm Principles".
  - Financial sector supervisors should emphasize macro-financial risks from bank holdings of sovereign debt; consider strengthening banking system backstops, higher loss-absorbing capital, regular systemic stress tests with credible sovereign risk scenarios, and coordination with public debt managers.
  - Complement standard debt sustainability analyses with investor base risk indicators to capture sovereign risk from market and investor dynamics.
  - Reduce data gaps on investor holdings of sovereign debt; improve granularity of holder information and foreign investor classifications. Support ongoing initiatives such as the 2009 G20 Data Gaps Initiative and the BIS-ECB-IMF Handbook on Securities Statistics (Part 2); ECB Centralized Securities Database initiative for euro area.
- Further research areas noted:
  - Empirically examine relationship between level/volatility of sovereign yields and investor holdings.
  - Analyze how investor base diversity affects sovereign debt market liquidity.
  - Factor investor base dynamics into unwind of crisis-related monetary and financial policies.
  - Assess impacts of new financial regulations on investor holdings and sovereign debt markets.

*Source: Box 2: Methodology of Sovereign Funding Shock Scenarios, from the supplied IMF content unit.*

### Annex Figure 1. Advanced Economies: Holders of Government Debt, 2004–11

### Annex Figure 1. Advanced Economies: Holders of Government Debt, 2004–11

### Key definitions and measurement notes
- Government debt indicates general government gross debt on a consolidated basis, which excludes intergovernmental holdings.
- Domestic banks are depository corporations residing in the country (IFS definition).
- Foreign banks are BIS reporting banks residing outside the country.
- Foreign official includes foreign central bank holdings as foreign exchange reserves, SMP holdings of foreign central banks, and foreign official loans.
- Foreign nonbanks and domestic nonbanks are imputed from external and total debt.
- Charts present components in percent; total in percent of GDP (axes labeled "Percent of GDP" and "Percent of total").
- Time coverage indicated across panels: 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011 (displayed along horizontal axes).

### Investor categories shown (panel legend wording preserved)
- Foreign official
- Foreign nonbank
- Foreign bank
- Domestic central bank
- Domestic nonbank
- Domestic bank
- Total debt (rhs)

### Country-level holdings (panel structure and axis values visible)
- Panels for individual advanced economies include Australia, Austria, Belgium, Canada, Czech Republic, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Japan, Korea, Netherlands, New Zealand, Norway, Portugal, Slovenia, Spain, Sweden, Switzerland, United Kingdom, United States.
- Axis tick values and ranges shown across country panels (examples preserved exactly as labeled in figures):
  - 0, 5, 10, 15, 20, 25, 30, 35, 40, 45, 50
  - 0, 10, 20, 30, 40, 50, 60, 70, 80, 90, 100
  - 0, 20, 40, 60, 80, 100, 120, 140, 160, 180
  - 90, 95, 100, 105, 110, 115, 120, 125
  - 0, 200 (appearing on some country panels)
- Percent-of-GDP and percent-of-total scales are used concurrently (dual axes), with "Total debt (rhs)" indicating total debt plotted on the right-hand side scale.

### Aggregate group panels
- Aggregate panels include: All Countries, G7, High-Spread Euro Area, Low-Spread Euro Area, Traditional Reserve Countries, Other Advanced Economies.
- Axis tick values and ranges visible for aggregates:
  - 0, 20, 40, 60, 80, 100, 120, 140
  - 0, 5, 10, 15, 20, 25, 30, 35, 40, 45, 50
  - 0, 200 (and multi-year axis labels 2004–2011)
- Aggregates use the same investor-category decomposition and percent/GDP presentation.

### Annex Figure 2. Advanced Economies: Cumulative Net Purchases of Government Debt by Investor Type, 2008–11
- Time coverage explicitly 2008–11 (horizontal axis labeled 2004 2005 2006 2007 2008 2009 2010 2011 in some panels; title specifies 2008–11 for cumulative net purchases).
- Panels mirror country and aggregate group structure used in Annex Figure 1.
- Investor-category legend preserved (as in Figure 1) and includes:
  - Domestic central bank
  - Domestic bank
  - Domestic nonbank
  - Foreign central bank
  - Foreign bank
  - Foreign nonbank
  - Foreign official loans
  - SMP_external
- Axis tick values and ranges shown across panels for cumulative net purchases (examples preserved exactly as labeled in figures):
  - -50, 0, 50
  - -40, -20, 0, 20, 40, 60, 80
  - -30, -25, -20, -15, -10, -5, 0, 5, 10, 15
  - -10, -5, 0, 5, 10, 15, 20, 25, 30, 35, 40
  - -4, -3, -2, -1, 0, 1, 2
  - -2, 0, 2, 4, 6, 8, 10, 12, 14
  - Billion euros; Billion U.S. dollars; Trillion U.S. dollars (units shown for aggregate panels)
  - Percent of 2011 GDP (unit label appearing on many country panels)

### Visual findings implied by figure layout (preserved qualitative structure without inferring magnitudes beyond axes)
- Figures decompose sovereign debt holdings by investor type across 2004–11 and show changes in investor composition.
- Annex Figure 2 presents cumulative net purchases by investor type for 2008–11, with panels indicating both positive and negative net flows across investor categories and countries.
- SMP_external and foreign central bank holdings are explicitly separated in the investor legend.

*Source: Authors' calculations. Note text and legend reproduced from original figures.*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp12284.pdf_