## _wp1688

## Source details

**Canonical URL:** [_wp1688](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp1688.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp1688.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp1688.pdf.json)

---

### Introduction and key findings
- Foreign banking claims grew until 2007 and have declined as a share of world GDP since then.
- The ratio of foreign banking claims to total banking assets of host countries declined from a peak of 24 percent in 2007 to 16 percent in 2013.
- Baseline regressions indicate that regulatory changes can account for about a half of the decline in the ratio of foreign lending to GDP between 2007 and 2013.
- Changes in home regulations related to banks’ international operations are identified as a key driver of the decline; regulatory changes in host countries played a less important role.
- Home country banks with higher capitalization in 2007 were associated with smaller declines in foreign banking claims following the crisis.
- The larger the pre-crisis growth of claims, the larger their decline since then.
- Higher initial home country engagement in the host countries is associated with a lower drop in lending.

### Data sources and scope
- Foreign banking claims: BIS, Consolidated Banking Statistics on an immediate risk basis (quarterly data until 2013Q3 averaged for each year).
- Portfolio investment assets: IMF, Coordinated Portfolio Investment Survey (equities and bonds).
- GDP: World Bank.
- Total banking assets: IMF, International Financial Statistics.
- Bilateral foreign claims (analysis): BIS consolidated banking statistics on an ultimate risk basis (available from 2005 onward).
- Time focus: international banking activities post-crisis, particularly after 2007.

### New regulatory survey data (IMF 2014 survey)
- Survey timing and coverage:
  - Conducted in the fall of 2014 for the April 2015 IMF Global Financial Stability Report preparation.
  - Confidential answers from supervisors in 40 advanced and emerging economies.
  - Survey asked 31 questions from home and host perspectives; questions classified into six categories for both home and host regulations.
  - Respondent jurisdictions listed in the source (40 jurisdictions).
- Coding and indexes:
  - Responses on changes between 2006 and 2014 coded as “tightened,” “loosened,” or “no change.”
  - For each category and country, index = 1, 0, or -1 when number of tightenings is greater than, equal to, or smaller than number of loosenings.
  - An overall index is the arithmetic mean of the six category scores.
- Use in analysis: only responses about changes between 2006 and 2014 are used.

### Regulatory changes documented by the survey
- Many countries tightened regulations on banks’ international operations between 2006 and 2014; only a few loosened them.
- In 2014, supervisors were more likely than before to limit banks’ activities (e.g., discretionary ring-fencing measures).
- Many resolution authorities obtained more powers over local branches of foreign banks.
- Some countries amended banking secrecy laws to enhance information sharing with foreign supervisors.
- A few countries loosened regulations regarding foreign banking presence (acquisition conditions) and activity (cross-border lending and borrowing).
- The share of countries that tightened regulations from 2006 to 2014 is higher in advanced economies than in emerging market economies.
- Euro area periphery countries tightened their home country regulations less than other advanced countries during this period.

### Methodology summary
- Cross-sectional regression model with dependent variable: growth rate of bilateral foreign banking claims from home country i to host country j between 2007 and 2013.
- Uses BIS consolidated data on an ultimate-risk basis for bilateral claims to capture risk transfers and bilateral exposures.
- Immediate-risk-basis data used in figures to show longer time series where risk-transfer issues are irrelevant.
- Key covariates:
  - Home-specific: one or two regulatory indexes, change in exchange rate vs U.S. dollar, U.S. dollar-denominated GDP growth, aggregate bank capital-to-total assets ratio in 2007.
  - Host-specific: one or two regulatory indexes, exchange rate change vs U.S. dollar, U.S. dollar-denominated GDP growth.
  - Bilateral: log distance, common language dummy, importance of claims from home in host in 2007, importance of claims in host from home perspective in 2007, pre-crisis growth rate of bilateral claims.
- Sample measurements: growth rates computed as change divided by average of the two periods; pre-crisis growth measured 2005–2007; post-crisis measured 2007–2013.

### Home countries — benchmark specifications and covariates
- Benchmark regression: growth in bilateral claims as dependent variable to separate home and host effects (formal specification given in source).
- Home-specific variables include regulatory change indexes and initial bank capital-to-total assets ratio in 2007.
- Notes on country groupings used in figures: AE = advanced economies; EM = emerging market economies; Euro area core and periphery countries listed in source.

### Host countries — data, variables, and benchmark results
- Key variables and measurement:
  - Post-crisis claims growth (2007–2013), international operations regulatory index (2006–14 changes), capital regulatory index (difference Barth et al. 2006–2011 normalized to -1..1), exchange rate percent change vs U.S. dollar (2007–2013), GDP growth (2007–2013), bank capital ratio (home, 2007), importance measures (shares in 2007), pre-crisis claims growth (2005–2007).
  - Data notes: BIS ultimate risk basis; quarterly averaging over 2005Q2–2013Q3; observations with negative claims dropped.
- Main benchmark regression coefficients and significance (Table 3; columns (1) and (2); significance: *** p<0.01, ** p<0.05, * p<0.1; robust p-values in parentheses):
  - International operations regulatory index (home, change): -177.60*** (0.00) in (1); -190.54*** (0.00) in (2).
    - Text interpretation: tightening of one type of regulation (index increase by one-sixth) is associated, on average, with a 30 percentage-point drop in foreign claims between 2007 and 2013.
  - International operations regulatory index (host, change): -35.23** (0.04) in (1); -43.52** (0.02) in (2).
  - Capital regulatory index (home, change): -82.94*** (0.00) in (2).
  - Capital regulatory index (host, change): 7.93 (0.57) in (2) — not statistically significant.
  - Exchange rate (home, % appreciation): 1.09*** (0.00) in (1); 0.83** (0.03) in (2).
  - Exchange rate (host, % appreciation): 0.21 (0.41) in (1); 0.13 (0.67) in (2) — not statistically significant.
  - GDP (home, % change): 0.98*** (0.00) in (1); 1.05*** (0.00) in (2).
  - GDP (host, % change): 0.99*** (0.00) in (1); 0.96*** (0.00) in (2).
  - Bank capital to total assets (home, % in 2007): 4.85*** (0.00) in (1); 3.15** (0.04) in (2).
  - Host country's share of claims from home (% in 2007): 0.73 (0.10) in (1); 0.81* (0.09) in (2).
  - Home country's share of claims on host (% in 2007): 1.40*** (0.00) in (1); 0.90*** (0.01) in (2).
  - Claims (% change from 2005 to 2007): -0.17** (0.02) in (1); -0.14* (0.09) in (2).
  - Observations: 726 in (1); 597 in (2). R-squared: 0.17 in (1); 0.20 in (2).
- Key interpretations:
  - Tighter home-country regulations on international operations reduced international banking activities post-2008 rather than encouraging regulatory arbitrage.
  - Home-currency appreciation positively associated with growth of foreign claims.
  - GDP growth in home and host countries positively associated with increases in foreign claims; host GDP coefficient close to one.
  - Higher initial home-country bank capitalization in 2007 associated with higher subsequent growth in foreign claims.
  - Higher initial bilateral engagement associated with smaller declines in lending.
  - Higher pre-crisis growth of bilateral claims associated with larger subsequent contractions.

### Standardized regressions and contributions to claims/GDP growth
- Standardized-variable (one-standard-error change) results (Table 4):
  - International operations regulatory index (home, change): -17.17*** (0.00) in (1); -19.08*** (0.00) in (2).
  - International operations regulatory index (host, change): -7.09** (0.04) in (1); -8.91** (0.02) in (2).
  - Capital regulatory index (home, change): -21.78*** (0.00) in (2).
  - GDP (home, % change): 17.67*** (0.00) in (1); 19.85*** (0.00) in (2).
  - GDP (host, % change): 24.93*** (0.00) in (1); 24.32*** (0.00) in (2).
  - Bank capital to total assets (home, % in 2007): 11.09*** (0.00) in (1); 7.50** (0.04) in (2).
  - Observations: 726 in (1); 597 in (2). R-squared: 0.17 in (1); 0.20 in (2).
- Contributions to foreign claims/GDP growth (Table 5, Percentage Points):
  - Foreign claims/GDP (average growth): -43.7 in (1); -47.5 in (2).
  - Regulatory changes total contribution: -16.9 in (1); -27.7 in (2).
    - International operations regulatory index (home, change): -10.5 in (1); -12.7 in (2).
    - International operations regulatory index (host, change): -6.4 in (1); -8.4 in (2).
    - Capital regulatory index (home, change): -7.2 in (2).
    - Capital regulatory index (host, change): 0.5 in (2).
  - Other drivers (non-regulatory): -26.8 in (1); -19.7 in (2).
- Interpretation: Regulatory changes (especially home international-operations measures) materially contributed to the post-crisis decline in foreign banking claims; their overall effect is comparable to non-regulatory factors.

### Robustness checks
- Categories of robustness checks:
  1. Alternative regulatory indexes (Barth et al. (2013) indexes and averages).
  2. Instrumental variables addressing endogeneity.
  3. Excluding bilateral claims from euro area countries.
  4. Additional controls (banking crisis dummy, sovereign rating changes, real interest rates, financial openness, trade controls, share of government-owned banks).
- Selected alternative-measure results (Table 6):
  - International operations regulatory index (home, change) remains negative and significant across specifications:
    - -125.44*** (0.00) in (1); -133.37*** (0.00) in (2); -295.15*** (0.00) in (3); -286.72*** (0.00) in (4).
  - International operations regulatory index (host, change) remains negative and significant: -40.69** (0.04) in (1); -46.51** (0.02) in (2); -54.71** (0.02) in (3); -46.90** (0.03) in (4).
  - Official Supervisory Power (home, change): 35.60** (0.03) in (2) — positive and significant.
  - Financial Statement Transparency (home, change): -79.08*** (0.00) in (3).
  - Average of general financial regulations (home, change): -175.40*** (0.00) in (4).
  - Observations: 544 in (1); 597 in (2); 527 in (3); 477 in (4). R-squared range 0.155–0.210.
- Instrumental variables (Table 7 summary):
  - Instruments: pre-crisis levels of four general financial regulatory indexes in 2003 and 2006 for home and host.
  - IV estimates broadly confirm benchmark results for home regulatory changes:
    - International operations regulatory index (home, change): -157.99*** (0.00) in (1); -291.57*** (0.00) in (2).
    - International operations regulatory index (host, change): 31.95 (0.27) in (1); 28.92 (0.37) in (2) — not significant in IV.
    - Capital regulatory index (home, change): -113.71*** (0.00) in (2).
  - Observations: 690 in (1); 565 in (2). R-squared: 0.13 in (1); 0.14 in (2).
- Excluding euro area claims (Table 8):
  - International operations regulatory index (home, change): -261.51*** (0.00) in (1); -1,037.60*** (0.00) in (2).
  - Capital regulatory index (home, change): -76.57 (0.12) in (2) — not significant.
  - Observations: 328 in (1); 230 in (2). R-squared: 0.227 in (1); 0.245 in (2).
  - Interpretation: Main qualitative conclusions hold though some estimates sensitive to smaller sample.

### Analysis by subcategories — by type and by counterparty
- By type (cross-border vs local claims):
  - Cross-border claims: strong negative effects from home international-operations regulatory changes and home capital regulatory changes.
    - International operations regulatory index (home, change): -189.15*** and -251.54*** (Table 9 columns (1) and (2); p-value 0.00).
    - Capital regulatory index (home, change): -64.78*** and -79.98*** (Table 9 columns (1) and (2); p-value 0.00).
    - GDP (home, % change): 1.24*** and 1.25***; GDP (host, % change): 1.02*** and 0.92***.
    - Claims (% change from 2005 to 2007): -0.20** and -0.16*.
  - Local claims: limited or mixed impact from home international-operations regulatory changes.
    - International operations regulatory index (home, change): 26.07 and -133.89 (Table 9 columns (3) and (4); p-values 0.71 and 0.15).
    - Distance (log, km): -16.91*** and -19.09*** — large negative and significant.
    - Host country's share of claims from home (% in 2007): 0.77** and 0.48.
  - Interpretation: home-country regulations strongly reduce cross-border claims; local lending is less responsive, contributing to resilience of local lending.
- By counterparty (non-banks vs banks):
  - Non-banks:
    - International operations regulatory index (home, change): -178.57*** and -278.17*** (Table 10 columns (1) and (2); p-value 0.00).
    - GDP (home, % change): 1.53*** and 1.66***.
    - Distance (log, km): -7.25* and -9.99** in some specs.
  - Banks:
    - International operations regulatory index (host, change): -44.46** and -56.75** (Table 10 columns (3) and (4); p-values 0.04 and 0.02).
    - Home-country international-operations index not significant for claims on banks.
    - Host country's share of claims from home (% in 2007): 1.85*** and 1.69***.
  - Interpretation:
    - Home-country international-operations regulation reductions primarily affected lending to non-banks.
    - Host-country international-operations regulations are associated with reductions in foreign claims on domestic banks, consistent with limiting wholesale funding reliance.
    - Distance matters more for non-bank lending; host-country importance matters more for interbank claims.

### Aggregate conclusion, magnitudes, and policy implications
- Regulatory tightening explains a substantial portion of the post-crisis decline in international banking:
  - Empirical analysis suggests regulatory tightening can explain half of the decline in the foreign lending-to-GDP ratio between 2007 and 2013.
- Home-country regulatory changes are more influential than host-country changes in explaining the decline in foreign claims.
- Regulations targeted at banks’ international operations mattered more than general financial regulatory changes (capital regulation), especially for cross-border lending from headquarters.
- Shift in composition of foreign claims:
  - Cross-border lending fell more than local lending by affiliates; because local lending is less pro-cyclical, the compositional shift toward local lending could positively affect host-country financial stability.
- Policy-relevant takeaways:
  - Home-country regulatory tightening materially reduces cross-border banking claims and foreign lending-to-GDP ratios.
  - Host-country measures targeting international operations can reduce foreign claims on domestic banks.
  - Compositional shifts toward local affiliate lending may reduce pro-cyclicality in host countries and enhance financial stability.

*Source: IMF working paper content (survey and analysis on regulations and foreign banking claims, 2005–2014 data coverage).*

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

### _wp1688 - References .............................................................................................................

### Introduction and key findings
- Foreign banking claims grew until 2007 and have declined as a share of world GDP since then.
- The ratio of foreign banking claims to total banking assets of host countries declined from a peak of 24 percent in 2007 to 16 percent in 2013.
- Baseline regressions indicate that regulatory changes can account for about a half of the decline in the ratio of foreign lending to GDP between 2007 and 2013.
- Changes in home regulations related to banks’ international operations are identified as a key driver of the decline; regulatory changes in host countries played a less important role.
- Home country banks with higher capitalization in 2007 were associated with smaller declines in foreign banking claims following the crisis.
- The larger the pre-crisis growth of claims, the larger their decline since then.
- Higher initial home country engagement in the host countries is associated with a lower drop in lending.

### Data sources and scope
- Foreign banking claims are from BIS, Consolidated Banking Statistics on an immediate risk basis (quarterly data until 2013Q3 averaged for each year).
- Portfolio investment assets are from IMF, Coordinated Portfolio Investment Survey and include both equities and bonds.
- GDP is from the World Bank.
- Total banking assets are from IMF, International Financial Statistics.
- Consolidated banking statistics on an ultimate risk basis are used for bilateral foreign claims; these are available only from 2005 onward.
- The analysis focuses on international banking activities in the post-crisis period (particularly the period after 2007).

### New regulatory survey data (IMF 2014 survey)
- Survey conducted in the fall of 2014 for the preparation of the April 2015 IMF’s Global Financial Stability Report.
- Answers were confidentially collected from supervisors in 40 advanced and emerging economies.
- Survey asked 31 questions from the perspective of home and host countries (see Table 1 structure in source).
- Questions were classified into six categories for both home country and host country regulations.
- The survey asked about regulations in place in 2014 and about changes between 2006 and 2014; only the latter responses are used in this analysis.
- For most questions about regulatory changes, answers were restricted to three options and coded to “tightened,” “loosened,” and “no change.”
- Index variables: for each category and country, index = 1, 0, or -1 when the number of tightenings is, respectively, greater than, equal to, or smaller than the number of loosenings. An overall index is the arithmetic mean of the six category scores.
- Survey respondent jurisdictions: Argentina, Australia, Austria, Belgium, Brazil, Canada, Chile, China, Croatia, Denmark, Finland, France, Germany, Greece, Hong Kong SAR, Hungary, Indonesia, Ireland, Italy, Japan, Luxembourg, Malaysia, Mexico, the Netherlands, Norway, the Philippines, Portugal, Romania, Russia, Saudi Arabia, Singapore, Slovak Republic, South Africa, Spain, Sweden, Switzerland, Thailand, Turkey, United Kingdom, and United States.

### Regulatory changes documented by the survey
- Many countries tightened regulations on banks’ international operations between 2006 and 2014; only a few loosened them.
- In 2014, many supervisors were more likely than before to limit banks’ activities (for example, by imposing ring-fencing measures in a discretionary way).
- Many resolution authorities obtained more powers over local branches of foreign banks.
- Some countries amended banking secrecy laws to enhance information sharing with foreign supervisors.
- A few countries loosened regulations regarding foreign banking presence (for example, acquisition conditions) and activity (for example, cross-border lending and borrowing).
- The share of countries that tightened regulations on banks’ international operations from 2006 to 2014 is higher in advanced economies than in emerging market economies.
- Euro area periphery countries tightened their home country regulations less than other advanced countries during this period, despite facing serious banking stress.

### Methodology summary
- A cross-sectional regression model is estimated with the growth rate of bilateral foreign banking claims as the dependent variable.
- The paper uses consolidated BIS data on an ultimate-risk basis for bilateral claims to better capture risk transfers and bilateral exposures.
- Immediate-risk-basis data are used in Figures 1 and 2 to show a longer time series; risk-transfer issues are irrelevant for those aggregate figures.

### Subcategory and robustness findings
- Analysis by subcategories of claims suggests:
  - Cross-border lending from banks’ headquarters has been more affected by regulations on banks’ international operations than local lending from banks’ affiliates.
  - The resilience of local lending relative to cross-border lending may be largely attributable to the limited response of local lending to changes in regulations on banks’ international operations.
  - Host country regulations played a significant role in reducing international bank borrowing.
- Robustness checks include use of different BIS data bases (immediate-risk vs. ultimate-risk) and various regression specifications.

### Contribution to literature and implications
- The study adds to literature on determinants of foreign banks’ presence by emphasizing the role of regulatory changes, a factor largely ignored in many prior studies.
- Results suggest that post-crisis tighter home-country regulations on international operations are an important factor explaining the reduction in foreign bank lending and the decline in foreign banking claims as a share of GDP and host-country banking assets.

*Source: IMF working paper content (survey and analysis on regulations and foreign banking claims, 2005–2014 data coverage).*

### 1. Home  Countries

### 1. Home Countries

### Benchmark specifications
- The benchmark regression model uses the growth in bilateral claims as the dependent variable to separate effects of home and host variables.
- The baseline specification takes the following form as presented in the source:
  ݏ݈݉݅ܽܿ∆
  ௜௝
  ܍ܕܗܐ൅઺∙ߙൌ
  ௜
  ܜܛܗܐ൅઻∙
  ௝
  ܔ܉ܚ܍ܜ܉ܔܑ܊൅઼∙
  ௜௝
  ߝ൅
  ௜௝
  ,   (1) 
- The dependent variable ݏ݈݉݅ܽܿ∆
  ௜௝
  denotes the growth rate of claims from home country i to host country j between 2007 and 2013.

### Model covariates and data features
- Home-specific variables (܍ܕܗܐ
  ௜
 ):
  - Include one or two indexes of regulatory changes.
  - Include the change in the exchange rate against the U.S. dollar.
  - Include the U.S. dollar-denominated GDP growth rate.
  - Additionally include the aggregate bank capital-to-total assets ratio in 2007 as an indicator of initial banking sector health.
- Host-specific variables (ܜܛܗܐ
  ௝
 ):
  - Include one or two indexes of regulatory changes.
  - Include the change in the exchange rate against the U.S. dollar.
  - Include the U.S. dollar-denominated GDP growth rate.
- Bilateral variables (ܔ܉ܚ܍ܜ܉ܔܑ܊
  ௜௝
  ):
  - Log distance between home and host countries.
  - A common language dummy.
  - A variable capturing the importance of the claims from the home country in the host country in 2007.
  - A variable capturing the importance of the claims in a given host country from the home country perspective in 2007.
  - The pre-crisis growth rate of bilateral claims.
- Note: Classens and van Horen (2014a) include a regressor measuring the importance of the claims from the home country in the host country, but do not include the variable measuring the importance of the claims in a given host country from the home country perspective.

### Sample period and country groupings
- Growth rate of claims measured between 2007 and 2013.
- Regulatory tightening/loosening captured over the 2006 and 2014 window in figure references.
- Country groupings referenced in figures and notes:
  - AE = advanced economies.
  - EM = emerging market economies.
  - Euro area core countries: Austria, Belgium, Finland, France, Germany, Luxembourg, and the Netherlands.
  - Euro area periphery countries: Greece, Ireland, Italy, Portugal, and Spain.

### Figures and descriptive statistics (as presented)
- Figure title: Share of Countries that Tightened Regulations on International Banking Operations between 2006 and 2014, by Region.
- Note: Countries that tightened regulations are defined as countries with a positive index of changes in regulations on banks' international operations.
- Axis ticks and percentage scales shown in figures include: 0, 10, 20, 30, 40, 50, 60, 70, 80 (percent); and in another figure: 0, 20, 40, 60, 80, 100 (percent).
- Regional labels displayed in figures: AE excluding euro area; Euro area (core); Euro area (periphery); EM.

*Source: _wp1688 - 1. Home  Countries*

### 2. Host Countries

### 2. Host Countries

### Data, variables, and measurement
- Post-crisis claims growth: The growth rate of bilateral claims from 2007 to 2013, calculated by dividing the change in claims by the average of claims in 2007 and 2013. Source: BIS, CBS.
- International operations regulatory index: An index constructed from answers to survey questions about regulation changes for 2006–14. Source: IMF.
- Capital regulatory index: Difference between Barth et al.'s (2013) indexes in 2006 and 2011, normalized so that the index can take a value from -1 to 1. Source: Barth et al. (2013).
- Exchange rate: The percent change in the exchange rate against the U.S. dollar between 2007 and 2013. Source: IMF, IFS.
- GDP: The growth rate from 2007 to 2013, calculated with the same method used for post-crisis claims growth. Source: IMF, WEO.
- Bank capital ratio: The bank capital-to-total assets ratio of the home country in 2007. Source: World Bank, GFDD.
- Importance measures:  
  - Importance of host in the claims from home: The ratio of bilateral claims from a home country to a host country to total claims from the home country to all host countries in 2007. Source: BIS, CBS.  
  - Importance of home in the claims on host: The ratio of bilateral claims from a home country to a host country to total claims from all home countries to the host country in 2007. Source: BIS, CBS.
- Pre-crisis claims growth: Growth rate of bilateral claims between 2005 and 2007 (calculated as for post-crisis claims growth).
- Data source notes: BIS consolidated banking statistics on an ultimate risk basis; statistical breaks identified and corrected following Cerutti (2013); series constructed by averaging quarterly data over 2005Q2-2013Q3; growth rate computed dividing change by the average of the two periods. Observations with negative claims values are dropped.

### Benchmark results: effects of regulatory changes and controls
- Main finding: Regulations on banks’ international operations in home and host countries have significant negative effects on foreign lending.
- Key coefficients and significance from Table 3 (columns shown as (1) and (2); robust p-values in parentheses; significance: *** p<0.01, ** p<0.05, * p<0.1):
  - International operations regulatory index (home, change): -177.60*** (0.00) in (1); -190.54*** (0.00) in (2).
    - Interpretation in text: tightening of one type of regulations (which would increase the index by one-sixth) in the home country is, on average, associated with a 30 percentage-point drop in foreign claims between 2007 and 2013.
  - International operations regulatory index (host, change): -35.23** (0.04) in (1); -43.52** (0.02) in (2).
  - Capital regulatory index (home, change): -82.94*** (0.00) in (2).
  - Capital regulatory index (host, change): 7.93 (0.57) in (2) — not statistically significant.
  - Exchange rate (home, % appreciation): 1.09*** (0.00) in (1); 0.83** (0.03) in (2).
  - Exchange rate (host, % appreciation): 0.21 (0.41) in (1); 0.13 (0.67) in (2) — not statistically significant.
  - GDP (home, % change): 0.98*** (0.00) in (1); 1.05*** (0.00) in (2).
  - GDP (host, % change): 0.99*** (0.00) in (1); 0.96*** (0.00) in (2).
  - Bank capital to total assets (home, % in 2007): 4.85*** (0.00) in (1); 3.15** (0.04) in (2).
  - Host country's share of claims from home (% in 2007): 0.73 (0.10) in (1); 0.81* (0.09) in (2).
  - Home country's share of claims on host (% in 2007): 1.40*** (0.00) in (1); 0.90*** (0.01) in (2).
  - Claims (% change from 2005 to 2007): -0.17** (0.02) in (1); -0.14* (0.09) in (2).
  - Observations: 726 in (1); 597 in (2). R-squared: 0.17 in (1); 0.20 in (2).
- Interpretation highlights:
  - Tighter regulations (higher index values) reduced international banking activities post-2008 crisis rather than encouraging regulatory arbitrage.
  - Appreciation of the home currency is positively associated with growth of foreign claims (possible valuation effect or cheaper foreign assets).
  - GDP growth in both home and host countries is associated with increases in foreign claims; coefficient on host GDP is very close to one (one-to-one relationship).
  - Higher bank-capital-to-total-assets ratio in the home country in 2007 is associated with higher subsequent growth in foreign claims.
  - Bilateral relationships matter: higher initial home engagement with a host is associated with a smaller decline in lending.
  - Pre-crisis high growth bilateral claims experienced larger subsequent contractions (negative coefficient on pre-crisis claims growth).

### Standardized regressions and contributions to claims/GDP growth
- Standardized-variable results (Table 4) compare sensitivities to a one-standard-error change:
  - International operations regulatory index (home, change): -17.17*** (0.00) in (1); -19.08*** (0.00) in (2).
  - International operations regulatory index (host, change): -7.09** (0.04) in (1); -8.91** (0.02) in (2).
  - Capital regulatory index (home, change): -21.78*** (0.00) in (2).
  - GDP (home, % change): 17.67*** (0.00) in (1); 19.85*** (0.00) in (2).
  - GDP (host, % change): 24.93*** (0.00) in (1); 24.32*** (0.00) in (2).
  - Bank capital to total assets (home, % in 2007): 11.09*** (0.00) in (1); 7.50** (0.04) in (2).
  - Observations: 726 in (1); 597 in (2). R-squared: 0.17 in (1); 0.20 in (2).
- Contributions to foreign claims/GDP growth (Table 5, Percentage Points):
  - Foreign claims/GDP (average growth): -43.7 in (1); -47.5 in (2).
  - Regulatory changes total contribution: -16.9 in (1); -27.7 in (2).
    - International operations regulatory index (home, change): -10.5 in (1); -12.7 in (2).
    - International operations regulatory index (host, change): -6.4 in (1); -8.4 in (2).
    - Capital regulatory index (home, change): -7.2 in (2).
    - Capital regulatory index (host, change): 0.5 in (2).
  - Other drivers (non-regulatory): -26.8 in (1); -19.7 in (2).
- Interpretation: The overall effect of regulatory changes on foreign banking claims is comparable to that of non-regulatory factors. Regulatory changes targeted at international operations have larger effects than capital requirements. Effects of home-country regulatory changes are larger than host-country changes.

### Robustness checks
- Categories of robustness checks:
  1. Alternative financial regulatory indexes (Barth et al. (2013) indexes: Official Supervisory Power; Financial Statement Transparency; Overall Restrictions; and an average of the four indexes).
  2. Instrumental variables to address endogeneity of regulatory variables.
  3. Excluding bilateral claims from euro area countries.
  4. Additional control variables (discussed in Appendix; not detailed here).
- A. Alternative measures of regulatory changes (Table 6):
  - International operations regulatory index (home, change) remains negative and significant across specifications:
    - -125.44*** (0.00) in (1); -133.37*** (0.00) in (2); -295.15*** (0.00) in (3); -286.72*** (0.00) in (4).
  - International operations regulatory index (host, change) remains negative and significant: -40.69** (0.04) in (1); -46.51** (0.02) in (2); -54.71** (0.02) in (3); -46.90** (0.03) in (4).
  - Specific alternative index results (selected):
    - Official Supervisory Power (home, change): 35.60** (0.03) in (2) — positive and significant.
    - Financial Statement Transparency (home, change): -79.08*** (0.00) in (3).
    - Average of general financial regulations (home, change): -175.40*** (0.00) in (4).
  - Observations: 544 in (1); 597 in (2); 527 in (3); 477 in (4). R-squared ranges 0.155–0.210.
  - Interpretation: International operations regulatory indexes for home and host remain significant negatives across alternative regulatory measures. Capital regulatory index appears to contain more information than other three indexes (higher R-squared in baseline).
- B. Instrumental variables (Table 7 summary):
  - Instruments: pre-crisis levels of the four general financial regulatory indexes in 2003 and 2006 in both home and host countries.
  - IV estimates broadly similar to benchmark regressions:
    - International operations regulatory index (home, change): -157.99*** (0.00) in (1); -291.57*** (0.00) in (2).
    - International operations regulatory index (host, change): 31.95 (0.27) in (1); 28.92 (0.37) in (2) — not significant in IV results shown.
    - Capital regulatory index (home, change): -113.71*** (0.00) in (2).
  - Observations: 690 in (1); 565 in (2). R-squared: 0.13 in (1); 0.14 in (2).
  - Interpretation: IV results confirm negative and statistically significant effects of home regulatory changes; baseline results do not appear to seriously overestimate impacts due to endogeneity.
- C. Excluding claims from euro area countries (Table 8):
  - Results continue to show significantly negative coefficient on international operations regulatory index in home countries:
    - International operations regulatory index (home, change): -261.51*** (0.00) in (1); -1,037.60*** (0.00) in (2).
  - Some coefficients vary substantially (sample size reduction concerns):
    - Capital regulatory index (home, change): -76.57 (0.12) in (2) — not significant.
  - Observations drop to 328 in (1); 230 in (2). R-squared: 0.227 in (1); 0.245 in (2).
  - Interpretation: Main qualitative conclusions hold when excluding euro area observations, though some parameter estimates are sensitive to the reduced sample.

### Analysis by subcategories
- Aggregate patterns motivating disaggregation:
  - Cross-border claims fell since 2007 and by 2013 were much below peak levels, while local claims recovered (Figure 5, Panel 1).
  - By counterparty sector (Figure 5, Panel 2): foreign claims on (non-affiliated) banks continued to decline while claims on non-banks stabilized.
- Note: More detailed analysis by subcategories is pursued, but smaller numbers of observations make robustness harder to confirm (analysis initiated; detailed results not included in supplied excerpt).

*Source: _wp1688 - 2. Host Countries (excerpt).*

### 1. By type 2. By counterparty

### _wp1688 - 1. By type 2. By counterparty

### Key empirical findings: by type of claims
- Cross-border claims:
  - International operations regulatory index (home, change): -189.15*** and -251.54*** (columns (1) and (2) in Table 9) — large, negative, and statistically significant (Robust p-value 0.00).
  - Capital regulatory index (home, change): -64.78*** and -79.98*** (columns (1) and (2) in Table 9) — large, negative, and statistically significant (Robust p-value 0.00).
  - GDP (home, % change): 1.24*** and 1.25*** (columns (1) and (2) in Table 9) — positive and statistically significant (Robust p-value 0.00).
  - GDP (host, % change): 1.02*** and 0.92*** — positive and statistically significant (Robust p-value 0.00).
  - Distance (log, km): -5.02 and -3.68 (not statistically significant in Table 9 cross-border specifications).
  - Claims (% change from 2005 to 2007): -0.20** and -0.16* — negative and statistically significant at conventional levels.
- Local claims:
  - International operations regulatory index (home, change): 26.07 and -133.89 (columns (3) and (4) in Table 9) — not statistically significant (p-values 0.71 and 0.15).
  - Distance (log, km): -16.91*** and -19.09*** — large, negative, and statistically significant (Robust p-value 0.00).
  - Host country's share of claims from home (% in 2007): 0.77** and 0.48 — positive for local claims, significant in one specification (p-value 0.05).
  - Home country's share of claims on host (% in 2007): 0.84*** and 0.72** — positive and statistically significant.

Interpretation:
- Home country regulations on banks’ international operations strongly reduce cross-border claims but have limited or mixed impact on local claims.
- Local lending is less responsive to home-country international-operations regulations, consistent with greater resilience of local lending.
- Physical distance is negatively associated with growth of local lending, suggesting decreased exposures to far-away countries and potential post-crisis regionalization of banking activity.

### Key empirical findings: by counterparty (non-banks vs banks)
- Non-banks:
  - International operations regulatory index (home, change): -178.57*** and -278.17*** (columns (1) and (2) in Table 10) — large, negative, statistically significant (Robust p-value 0.00).
  - International operations regulatory index (host, change): 20.46 and 26.67 (not statistically significant).
  - GDP (home, % change): 1.53*** and 1.66*** — positive and statistically significant.
  - Distance (log, km): -7.25* and -9.99** — negative and statistically significant in some specifications.
  - Host country's share of claims from home (% in 2007): -0.54 and -0.60 (not significant).
- Banks:
  - International operations regulatory index (home, change): 40.70 and 36.02 (columns (3) and (4) in Table 10) — not statistically significant.
  - International operations regulatory index (host, change): -44.46** and -56.75** — negative and statistically significant (Robust p-values 0.04 and 0.02).
  - GDP (host, % change): 1.38*** and 1.25*** — positive and statistically significant.
  - Host country's share of claims from home (% in 2007): 1.85*** and 1.69*** — positive and statistically significant.

Interpretation:
- Regulations on banks’ international operations in home countries have had a pronounced negative effect on lending to non-banks.
- Host-country regulations targeting international operations are negatively associated with claims on banks (foreign claims on banks), suggesting host regulators sought to limit banks’ reliance on wholesale funding.
- Distance and host-country importance affect non-bank and bank claims differently: distance matters more for non-bank lending (operational costs), while host-country importance matters for interbank claims (sunk costs and existing bilateral bank relationships).

### Aggregate conclusion and policy-relevant magnitudes
- Regulatory tightening explains a substantial portion of the post-crisis decline in international banking:
  - The empirical analysis suggests that regulatory tightening can explain half of the decline in the foreign lending-to-GDP ratio between 2007 and 2013.
- Regulatory changes in home countries appear more influential than those in host countries in explaining the decline in foreign claims.
- Regulations targeted specifically at banks’ international operations have mattered more than general financial regulatory changes (capital regulation), especially for cross-border lending emanating from headquarters.
- Shift in composition of foreign claims:
  - Cross-border lending fell more than local lending by affiliates; to the extent that local lending is less pro-cyclical, the compositional shift toward local lending could positively affect financial stability in host countries.

### Robustness checks: controls and sensitivity
- Main results remain robust to addition of multiple controls, with significant coefficients preserved:
  - Banking crisis dummy (Laeven and Valencia, 2012): banking crisis dummy is negative (significant in one specification) but the significance of regulatory coefficients persists (Table A.1).
  - Sovereign rating index (number of notches the sovereign rating improved between 2007 and 2013): inclusion does not change main results; rating index is insignificant (Table A.2).
  - Real interest rates (short-term and long-term, home and host): main results robust; host real interest rates are negatively associated with changes in foreign claims (Table A.3). Some counterintuitive signs noted for real short-term rates in home countries (possible carry-trade unwinding or endogeneity).
  - Financial openness (Chinn-Ito index): increase in host country financial openness is associated with stronger growth of foreign claims; main results unchanged (Table A.4).
  - Bilateral trade variables (imports growth; trade growth): adding trade controls does not reveal a strong relationship between foreign banking claims and these trade variables (Table A.5).
  - Share of government-owned banks (2006): higher home-country government-owned bank share has a positive effect on growth rate of claims in one specification (significant at the 10 percent level); regulatory index results continue to hold (Table A.6).

### Policy implications
- Home-country regulatory tightening, especially regulations aimed at banks’ international operations, can materially reduce cross-border banking claims and foreign lending-to-GDP ratios.
- Host-country measures targeting international operations can reduce foreign claims on domestic banks, consistent with efforts to limit wholesale-funding reliance.
- A compositional shift from cross-border to local lending by affiliates may reduce pro-cyclicality of foreign banking flows and enhance host-country financial stability.

*Source: _wp1688 - 1. By type 2. By counterparty (IMF Working Paper PDF content provided).*

### REFERENCES

### _wp1688 - REFERENCES

### References

- Barth, James R., Gerard Caprio, and Ross Levine. 2013. “Bank Regulation and Supervision in 180 Countries from 1999 to 2011.” Working Paper 18733, National Bureau of Economic Research, Massachusetts.  
- Bernanke, Ben S. 2005. “The Global Saving Glut and the U.S. Current Account Deficit.” Remarks at the Sandridge Lecture, Virginia Association of Economists, Richmond, Virginia, March 10.  
- Bouvatier, Vincent and Anne-Laure Delatte. 2015. “Waves of International Banking Integration: A Tale of regional Differences.” European Economic Review, 80: 354–373.  
- Bremus, Franziska, and Marcel Fratzscher. 2015. “Drivers of Structural Change in Cross-Border Banking since the Global Financial Crisis.” Journal of International Money and Finance: 52(C), 32–59.  
- Cerutti, Eugenio. 2013. “Banks’ Foreign Credit Exposures and Borrowers’ Rollover Risks: Measurement, Evolution and Determinants.” Working Paper 13/9, International Monetary Fund, Washington.  
- Cerutti, Eugenio and Stijn Claessens. 2014. “The Great Cross-Border Bank Deleveraging: Supply Side Characteristics.” Working Paper 14/180, International Monetary Fund, Washington.  
- Cetorelli, Nicolas, and Linda S. Goldberg. 2011. “Global Banks and International Shock Transmission: Evidence from the Crisis.” IMF Economic Review 59 (1): 41–76.  
- Claessens, Stijn, and Neeltje Van Horen. 2014a. “The Impact of the Global Financial Crisis on Banking Globalization.” Working Paper 14/197, International Monetary Fund, Washington.  
- Claessens, Stijn, and Neeltje Van Horen. 2014b. “Location Decisions of Foreign Banks and Competitor Remoteness,” Journal of Money Credit and Banking 46(1): 145–170.  
- Coeurdacier, Nicolas. 2009. “Do trade costs in goods market lead to home bias in equities?” Journal of International Economics 77(1): 86–100.  
- Cull, Robert, and María Soledad Martínez Pería. 2010. “Foreign Bank Participation in Developing Countries. What Do We Know about the Drivers and Consequences of this Phenomenon.” Policy Research Working Paper 5398, World Bank, Washington.  
- De Haas, Ralph, and Neeltje Van Horen. 2012. “International Shock Transmission after the Lehman Brothers Collapse: Evidence from Syndicated Lending.” American Economic Review: Papers & Proceedings 102(3): 231–237.  
- Demirgüç-Kunt, Asli, and Enrica Detragiache. 2002. “Does Deposit Insurance Increase Banking System Stability? An Empirical Investigation.” Journal of Monetary Economics 49 (7): 1373–1406.  
- Goldberg, Lawrence G. and Denise Johnson. 1990. “The determinants of US banking activity abroad” Journal of International Money and Finance 9 (2): 123–137.  
- Hawkins, John and Dubravko Mihaljek. 2001. “The Banking Industry in the Emerging Market Economies: Competition, Consolidation and Systemic Stability: an Overview.” BIS Papers Chapters (4): 1–44.  
- Houston, Joel, Chen Lin, and Yue Ma. 2012. “Regulatory Arbitrage and International Bank Flows.” Journal of Finance 67 (5): 1845–895.  
- Ichiue, Hibiki and Kentaro Koyama. 2011. “Regime switches in exchange rate volatility and uncovered interest parity.” Journal of International Money and Finance: 30(7), 1436–1450.  
- International Monetary Fund (IMF). 2014. “Global Liquidity—Issues for Surveillance.” IMF Policy Paper, International Monetary Fund, Washington.  
- ———. 2015. “International Banking After the Crisis: Increasingly local and Safer?” Global Financial Stability Report, Chapter 2, April: 55–91.  
- Kapan, Tümer, and Camelia Minoiu. 2013. “Balance Sheet Strength and Bank Lending during the Global Financial Crisis.” Working Paper 13/102, International Monetary Fund, Washington.  
- Laeven, Luc, and Fabian Valencia. 2013. “Systemic Banking Crises Database.” IMF Economic Review 61 (2): 225–70.  
- Lane, Philip R., and Gian Maria Milesi-Ferretti. 2008. “International Investment Patterns.” The Review of Economics and Statistics 90 (3): 538–549.  
- McGuire, Patrick and Philip Wooldridge. 2005. “The BIS Consolidated Banking Statistics: Structure, Uses and Recent Enhancements.” BIS Quarterly Review, September: 73–86.  
- Obstfeld, Maurice, and Kenneth Rogoff. 2001. “The Six Major Puzzles in International Macroeconomics: Is There a Common Cause?" NBER Chapters, in: NBER Macroeconomics Annual 2000 (15): 339–412.  
- Ongena, Steven, Alexander Popov, and Gregory F. Udell. 2013. “’When the Cat’s Away the Mice Will Play’: Does Regulation at Home Affect Bank Risk-taking Abroad?” Journal of Financial Economics 108 (3): 727–50.  
- Reinhardt, Dennis, and Steven J. Riddiough. 2014. “The Two Faces of Cross-Border Banking Flows: An Investigation into the Links between Global Risk, Arms-length Funding and Internal Capital Markets.” Bank of England Working Paper 498.  
- Rose, Andrew K. 2004. “ Do We Really Know that the WTO Increases Trade?” American Economic Review 94 (1): 98–114.  
- Van Rijckeghem, Caroline, and Beatrice Weder di Mauro. 2014. “Deglobalization of Banking: The World is Getting Smaller.” Discussion Paper 10139, Center for Economic Policy and Research, London.  
- White, Halbert. 1980. “A Heteroskedasticity-Consistent Covariance Matrix Estimator and a Direct Test for Heteroskedasticity.” Econometrica 48 (4): 817–38.

*Source: _wp1688 - REFERENCES*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp1688.pdf_