## _wp11278

## Source details

**Canonical URL:** [_wp11278](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp11278.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp11278.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp11278.pdf.json)

---

### I. Introduction — scope and hypotheses
- Episodes of rapid growth in bank credit to the private sector are frequent (Terrones, 2004; Barajas et al., 2007).
- Two hypotheses tested:
  - Rapid credit growth weakens banks.
  - Credit grows more rapidly in sounder or less sound banks (sign ambiguous).
- Credit boom definition:
  - Credit-to-GDP ratio grows faster than implied by a backward-looking, country-specific, cubic time trend.
  - Identified 90 credit booms between 1995 and 2005.
  - Around 23 percent of these booms ended in a systemic banking crisis (Laeven and Valencia, 2008) within two years of the end of the boom.
- Sample period: 1995–2005, split roughly into 1995–2000 and 2001–05.

### II. Main empirical findings and interpretation
- General findings:
  - Credit growth has a negative impact on bank soundness, significant only in 1995–2000.
  - In 1995–2000 growth is driven mainly by sounder banks; in 2001–05 the relationship weakens as less sound banks catch up in loan growth.
  - Weakened link between credit growth and bank soundness implies increased risks as loan creation becomes equally likely in less sound banks.
  - Results robust to alternative measures of bank soundness and model specifications.
- Threshold and boom effects:
  - Negative impact of credit growth on bank soundness is stronger during credit booms than during moderate-growth periods.
  - Feedback from credit growth to bank soundness is significantly negative over the whole sample period only during booms.
  - During booms, weak banks may expand as fast as sound banks; banks growing faster than the average bank in a country show a negative relationship between soundness and credit growth.
- Ownership effects:
  - Foreign-owned banks’ loan growth is not statistically related to their soundness level.
  - Credit growth significantly weakens domestic banks but not foreign banks.
  - Implication: cross-border cooperation in bank supervision is important because foreign banks’ accumulation of risks and/or reliance on parent resources could generate spillovers.

### III. Econometric approach — model and estimation
- Simultaneous equation framework with two-way causality between credit growth and bank soundness.
- Dependent variables:
  - Credit growth = annual percent change in total outstanding loans of individual banks.
  - Bank soundness = distance to default (DD).
- Representative equations include lagged dependent variables and controls for macroeconomic and bank-specific factors (e.g., GDP per capita, real GDP growth, real interest rates, real exchange rate depreciation, net interest margin, liquidity ratio, bank size, foreign/public ownership).
- Estimation method: three-stage least squares (3SLS).
  - Rationale: suitable for simultaneous equation panels with short time dimension and lagged dependent variables; more efficient than 2SLS and robust to arbitrary residual autocorrelation.
- Specification and robustness tests conducted (unit-root tests, Hausman, residual analyses).

### IV. Data and sample
- Data sources:
  - Bank financial ratios: BankScope (Bureau van Dijk), March 2007 update and historical merges.
  - Macroeconomic data: March 2007 version of IMF’s International Financial Statistics.
- Sample coverage: banks in 90 countries from 1995 to 2005.
- Key sample-processing rules:
  - Trimmed 5 percent of observations on tails of bank-level credit growth and distance to default.
  - Matched bank identifiers, avoided duplications (handling consolidated vs. unconsolidated data, merger treatments), and updated ownership data using multiple sources.
- Summary statistics (selected highlights):
  - Bank-level credit growth (percent change in total loans):
    - 1995-2005: Obs. 49422, Mean 7.42, Std. Dev. 139.51
    - 1995-2000: Obs. 25090, Mean 4.70, Std. Dev. 29.34
    - 2001-2005: Obs. 24332, Mean 10.22, Std. Dev. 196.54
  - Distance to default:
    - 1995-2005: Obs. 49422, Mean 18.53, Std. Dev. 22.29
    - 1995-2000: Obs. 25090, Mean 7.15, Std. Dev. 13.69
    - 2001-2005: Obs. 24332, Mean 30.26, Std. Dev. 23.33
  - Net interest margin (1995-2005): Obs. 47819, Mean 0.04, Std. Dev. 0.16
  - Liquidity ratio (1995-2005): Obs. 45092, Mean 0.23, Std. Dev. 0.18
  - Foreign ownership (1995-2005): Obs. 49420, Mean 12.30, Std. Dev. 31.75
  - Country-level credit growth (1995-2005): Obs. 985, Mean 20.86, Std. Dev. 359.23
  - Credit boom indicator (1995-2005): Obs. 985, Mean 0.13, Std. Dev. 0.34
- Country list: sample includes 90 countries (examples: USA Obs. 7,112; Germany Obs. 18,703; France Obs. 3,244; Bulgaria Obs. 211; Latvia Obs. 227; Romania Obs. 235; etc.).

### V. Policy-relevant implications and contextual notes
- Rapid credit growth raises prudential risks by:
  - Causing capacity constraints in loan origination and risk management.
  - Increasing systemic risk via greater interconnection and homogeneity of exposures.
- Policy implications drawn from empirical results:
  - Monitor speed of aggregate credit growth (boom detection) because booms amplify the negative effect of credit growth on bank soundness.
  - Supervisory attention should consider heterogeneity across banks: weaker banks may expand aggressively during booms, increasing systemic vulnerabilities.
  - Cross-border supervisory cooperation is crucial given foreign banks’ differential behavior and potential reliance on parent resources.
- Historical observation: risks identified in the findings materialized in countries such as Bulgaria, Estonia, Latvia, Lithuania, and Romania with the trigger of the global financial turmoil.

### Appendix I — Distance to Default (DD) measures (definition and caveats)
- DD ≡ (k + μ) / σ, where:
  - k = equity capital as percent of assets,
  - μ = average return as percent on assets,
  - σ = standard deviation of return on assets.
- Interpretation: DD measures number of standard deviations a return realization must fall to exhaust equity; higher DD implies lower probability of insolvency.
- Methodological choices:
  - Balance-sheet-based annual DD used due to limited stock market liquidity in many sample countries.
  - Standard deviation of returns calculated for the entire sample period.
- Caveats:
  - DD depends on average return; aggressive banks expanding portfolios may appear artificially sounder.
  - Normality assumption for returns may not hold; DD performs better where bank stocks are actively traded.
  - This balance-sheet-based DD is sometimes called z-score.

### Appendix II — Specification tests, determinants, and robustness
- Unit-root tests for three-dimensional panel data reject unit roots at the 1 percent significance level.
- Baseline specification includes lagged dependent variables to capture persistence; specification analyses confirm adequacy.
- Determinants of credit growth (selected):
  - Real GDP growth: positive and statistically significant.
  - GDP per capita: negative coefficient (faster credit growth in poorer countries).
  - Sub-period differences:
    - 1995–2000: real depreciation and higher real interest rates associated with faster credit growth.
    - 2001–05: real appreciation and lower real interest rates linked to stronger credit growth.
  - Net interest margin: positive relationship with loan growth.
  - Public ownership share: significant negative coefficient.
- Determinants of bank soundness (selected):
  - Liquidity: more liquid banks are sounder.
  - GDP per capita: banks in richer countries are sounder.
  - Foreign ownership and size: negative and significant during 1995–2000 (larger, foreign banks take more risks in that sub-period).
  - Lagged distance to default: positive and significant, indicating persistence (coefficient close to unity but statistically different from 1).
- Robustness checks:
  - Results robust to time- and country-specific factors, alternative ownership measures, adding explicit financial development measures, alternative DD calculations, sub-period volatility, quadratic DD terms (insignificant), and alternative estimators (Arellano-Bond).

### Key regression findings — selected coefficients and fit (baseline, 1995-2005, Table 2)
- Selected coefficients (1995-2005):
  - Credit growth → Credit Growth equation: Coefficient 0.067**, [0.032]
  - Credit growth → Bank Soundness equation: Coefficient -0.017, [0.011]
  - Distance to default → Credit Growth equation: Coefficient 0.091***, [0.007]
  - Distance to default → Bank Soundness equation: Coefficient 0.951***, [0.002]
  - Real GDP growth → Credit Growth equation: Coefficient 0.688***, [0.062]
  - GDP per capita → Credit Growth equation: Coefficient -0.038***, [0.002]
  - Net interest margin → Credit Growth equation: Coefficient 31.43***, [4.448]
  - Real interest rate → Credit Growth equation: Coefficient -0.203***, [0.029]
  - Liquidity ratio → Credit Growth equation: Coefficient 1.014***, [0.333]
  - Size → Credit Growth equation: Coefficient -0.165***, [0.032]
  - Foreign ownership → Credit Growth equation: Coefficient -0.016***, [0.002]
- Fit statistics (1995-2005):
  - Observations: 28409 (columns shown)
  - R-squared reported for equations (examples): 0.39 (Credit Growth eq.), 0.85 (Bank Soundness eq.).

### Boom-specific regression highlights (Table 3, 1995-2005 boom results)
- Selected coefficients (boom periods):
  - Credit growth → Credit Growth (boom): 0.032**, [0.013]
  - Credit growth → Bank Soundness (boom): -0.605**, [0.271]
  - Distance to default → Bank Soundness (boom): 0.951***, [0.017]
  - Real GDP growth → Credit Growth (boom): 2.659***, [0.328]
  - Net interest margin → Credit Growth (boom): 51.42**, [24.67]
  - Real interest rate → Credit Growth (boom): -0.791***, [0.170]
  - Liquidity ratio → Credit Growth (boom): -2.403*, [1.333]
  - Size → Credit Growth (boom): -0.380***, [0.132]

### Rapid-bank and ownership heterogeneity (Tables 4 and 5 — selected)
- Rapid vs. slow growers:
  - Negative impact of credit growth on bank soundness is statistically significant only during credit booms and for banks expanding faster than the country-year average.
  - Slowly-growing banks during non-boom periods: statistically significant positive relationship between bank soundness and credit growth rates.
- Foreign vs. domestic banks (1995-2005):
  - Foreign banks:
    - Credit growth → Credit Growth: 0.019***, [0.005]
    - Distance to default → Bank Soundness: 0.961***, [0.007]
  - Domestic banks:
    - Credit growth → Credit Growth: 0.120***, [0.033]
    - Credit growth → Bank Soundness: -0.019, [0.013]
    - Distance to default → Bank Soundness: 0.950***, [0.003]
- Consistent regularity: distance to default is a strong positive predictor of bank soundness across subgroups.

*Source: _wp11278 - References.*

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

### _wp11278 - References

### I. Introduction — scope and hypotheses
- Episodes of rapid growth in bank credit to the private sector are frequent (Terrones, 2004; Barajas et al., 2007).
- Two-way relationship posited between credit growth and bank soundness: credit growth can weaken banks; bank soundness can influence credit growth (sign ambiguous).
- Paper tests two hypotheses:
  - Rapid credit growth weakens banks.
  - Credit grows more rapidly in sounder or less sound banks.
- Definition of credit booms: credit-to-GDP ratio grows faster than implied by a backward-looking, country-specific, cubic time trend.
  - Identified 90 credit booms between 1995 and 2005.
  - Around 23 percent of these booms ended in a systemic banking crisis (Laeven and Valencia, 2008) within two years of the end of the boom.
- Sample period limited to 1995–2005 for two reasons:
  - To derive policy implications available to policymakers around 2007 (real-time information up to the first signs of the global financial crisis).
  - Distribution of booms allows splitting the sample into roughly even sub-periods, 1995–2000 and 2001–05, with most booms in 2001–05.

### II. Main empirical findings and interpretation
- General findings:
  - Credit growth has a negative impact on bank soundness, but this effect is significant only in 1995–2000.
  - In 1995–2000, growth is driven mainly by sounder banks; in 2001–05 the relationship weakens, with less sound banks catching up in loan growth.
  - The weakening of the link between credit growth and bank soundness suggests increased risks as loan creation becomes equally likely in less sound banks.
  - These findings are robust to alternative measures of bank soundness and model specifications.
- Threshold and boom effects:
  - The negative impact of credit growth on bank soundness is stronger during credit booms than during moderate-growth periods.
  - During booms, weak banks may expand as fast as sound banks; banks that grow faster than the average bank in a country exhibit a negative relationship between soundness and credit growth.
  - Feedback from credit growth to bank soundness is significantly negative over the whole sample period only during booms.
  - Interpretation: during credit booms banks are more likely to dip into marginal borrower pools, contributing to deterioration in loan quality and explaining why credit booms are often associated with financial crises.
- Ownership effects:
  - Foreign-owned banks’ loan growth is not statistically related to their soundness level.
    - Possible interpretations: foreign banks are more aggressive to build market presence; or foreign banks have better access to wholesale funding and superior risk management via parent banks.
  - Credit growth significantly weakens domestic banks but not foreign banks.
  - Implication: cross-border cooperation in bank supervision is important because foreign banks’ accumulation of risks and/or reliance on parent resources could generate spillovers.

### III. Econometric approach — model and estimation
- Simultaneous equation framework models credit growth and bank soundness as functions of lagged values of each other and macroeconomic and bank-specific factors; allows two-way causality.
- Dependent variables:
  - Credit growth: annual percent change in total outstanding loans of individual banks.
  - Bank soundness: distance to default (see Appendix I for discussion and shortcomings).
- Baseline specification includes lagged dependent variables and controls for macroeconomic and bank-specific variables.
- Representative equations (as presented):
  - Equation 1: Credit Growth — Credit growth modeled as function of lagged distance to default, lagged credit growth, GDP per capita, real GDP growth, real interest rates, real exchange rate depreciation, net interest margin, liquidity ratio, bank size, foreign ownership, public ownership, and other controls.
  - Equation 2: Bank Soundness — Distance to default modeled as function of lagged credit growth, lagged distance to default, GDP per capita, real GDP growth, liquidity, bank size, foreign ownership, and other controls.
- Estimation method: three-stage least squares (3SLS)
  - Rationale: 3SLS convenient for simultaneous equation panels with short time dimension and lagged dependent variables.
  - Advantages cited:
    - Applies to simultaneous equation setting (unlike Arellano-Bond single-equation dynamic panel).
    - More efficient than 2SLS by accounting for cross-equation correlation.
    - Leaves auto-covariance matrix unrestricted, producing estimates robust to arbitrary residual autocorrelation.
  - Tests conducted to examine covariance structure and specification problems.

### IV. Data and sample
- Data sources:
  - Bank financial ratios from BankScope (Bureau van Dijk).
  - Macroeconomic data from IMF’s International Financial Statistics.
- Sample coverage: banks in 90 countries from 1995 to 2005.
- Summary statistics highlights:
  - Banks lent at higher rates on average during 2001–05 than 1995–2000 (Table 1a).
  - Banks in high-income, non-OECD group grew faster on average than banks in other groups in both periods.
  - Distance to default increased on average in all country groups, with more significant improvement in the high-income, non-OECD group; variation in distance to default also increased, especially in the middle-income group.
  - Variation in bank credit growth increased only in high-income, OECD group; remained somewhat unchanged in other groups.
  - Aggregate credit growth higher in the upper middle income group; incidence of credit booms almost as likely in high-income, non-OECD group (Table 1b).
  - Interpretation: credit booms are not only an emerging markets phenomenon; relationship between credit growth and bank soundness may depend more on whether a credit boom is occurring than on country income level.
- Appendices referenced for methods and data:
  - Appendix I: Distance to Default Measures.
  - Appendix II: Data Sources and Definitions.
  - Appendix Table 1: List of Countries in the Sample.

### V. Policy-relevant implications and contextual notes
- Rapid credit growth can raise prudential risks at micro and macro levels by:
  - Causing capacity constraints in loan origination and risk management (Berger and Udell, 2004).
  - Increasing systemic risk via greater interconnection and homogeneity of exposures (e.g., mortgage credit boom in the U.S. linked to house-price shock exposure; Dell’Ariccia, Igan, and Laeven, 2008).
- Key implications drawn from empirical results:
  - Monitoring speed of aggregate credit growth (boom detection) is critical because boom periods amplify the negative effect of credit growth on bank soundness.
  - Supervisory attention should consider heterogeneity across banks: weaker banks may expand aggressively during booms, raising systemic vulnerabilities.
  - Cross-border supervisory cooperation is important given foreign banks’ differential behavior and potential reliance on parent resources, which can create cross-border spillovers.
- Historical note:
  - Some risks identified in the findings materialized in countries such as Bulgaria, Estonia, Latvia, Lithuania, and Romania with the trigger of the global financial turmoil.

_Italic: Content based only on the supplied excerpt from _wp11278 - References._

### Appendix II.

### _wp11278 - Appendix II.

### Specification tests and model adequacy
- Feasible unit root tests for three-dimensional panel data (Kónya and Ohashi, 2005) reject unit roots at the 1 percent significance level.
- The Hausman specification test, based on a model excluding lagged dependent variables, is inconclusive; residual analysis points to non-stationarity due to failure to capture persistence.
- Baseline specification includes lagged dependent variables to capture persistence; this is confirmed as adequate by the specification analyses.
- Significance levels reported in the tables are based on robust standard errors; significance levels are similar when standard errors are clustered by country.

### Determinants of credit growth
- Higher real GDP growth has a statistically significant positive impact on credit growth.
- GDP per capita: coefficient on GDP per capita is negative in all periods, indicating catching-up (faster credit growth in poorer countries with less financial depth).
- Exchange rate and real interest rate effects differ by sub-period:
  - 1995–2000: real depreciation has a positive impact on credit growth; higher real interest rates are associated with faster credit growth.
  - 2001–05: real appreciation is linked to stronger credit growth; lower real interest rates are found to boost credit growth.
- Net interest margin: positive relationship with loan growth (more profitable banks can grow faster).
- Share of bank capital owned by the state: significant negative coefficient, implying financial sector reforms increased private sector access to credit.

### Determinants of bank soundness
- Significant determinants of bank soundness include:
  - Liquidity: more liquid banks are sounder.
  - GDP per capita: banks in richer countries are sounder.
  - Foreign ownership variable and size: coefficients are negative and statistically significant during 1995–2000, suggesting larger, foreign banks take more risks in that sub-period.
  - Lagged distance to default: coefficient is positive and statistically significant, indicating persistence (banks sound and stable in the past are likely to remain so). Although the coefficient on lagged distance to default is rather close to unity, statistical tests confirm that it is different from 1.

### Two-way relationship between credit growth and bank soundness
- Credit growth had a negative impact on bank soundness in both sub-periods, but the impact is statistically significant only in 1995–2000 (Table 2, Column 6 vs. Column 4).
- Sounder banks tend to grow faster throughout the period:
  - 1990s: sounder banks expanded credit faster but were also weakened by fast pace of growth.
  - 2001–05: pace of credit growth is less dependent on bank soundness; weakened banks continued to expand credit almost as fast as sound banks.
- Policy-relevant implication: rapid expansion by weak banks could ultimately undermine banking system soundness if weak banks outgrow problems by venturing into high-risk/high-return activities.
- Robustness checks:
  - Results robust to controlling for time- and country-specific factors; alternative measures of foreign and public ownership; adding explicit measure of financial development.
  - Robust to alternative calculations of distance to default; using sub-period volatility for returns renders a statistically significant negative coefficient on credit growth in bank soundness for 2001–05.
  - Including a quadratic term of distance to default produces an insignificant quadratic term.
  - Results broadly robust to estimating the equations separately using the Arellano-Bond method, although short time dimension precludes sub-sample analysis with that method.
  - Estimating baseline excluding lagged dependent variables does not significantly alter main parameters.
  - Replacing lagged variables with contemporaneous values (faster feedback) does not alter results.

### Credit booms and threshold effects
- Definition of credit boom (following Gourinchas, Valdes, and Landerretche, 2001):
  - Trend: country-specific, backward-looking, cubic time trend estimated over period starting at 1980 and ending at t-1.
  - A credit boom is identified if either:
    - (i) deviation from trend > 1.5 times historical country-specific standard deviation AND annual growth rate of credit-to-GDP ratio exceeds 10 percent; or
    - (ii) annual growth rate of credit-to-GDP ratio exceeds 20 percent.
  - Boom start: earliest year when either:
    - (i) credit-to-GDP ratio exceeds trend by more than three-fourths of its historical standard deviation AND annual growth rate exceeds 5 percent; or
    - (ii) annual growth rate exceeds 10 percent.
  - Boom end: when either:
    - (i) growth of credit-to-GDP ratio turns negative; or
    - (ii) credit-to-GDP ratio falls within three-fourths of one standard deviation from its trend AND annual growth rate is lower than 20 percent.
- Key findings from boom/non-boom split (Table 3 comparisons):
  - Negative impact of credit growth on bank soundness is significant over whole sample only during booms (Table 3, Column 2 vs. Column 8).
  - In sub-periods, feedback from credit growth to distance to default is much larger during credit booms than in non-booms (Table 3, Column 4 vs. Column 10).
  - Sounder banks expand credit faster than less sound banks more in non-boom periods than in booms.
  - In 2001–05 booms, coefficient on bank soundness in credit growth equation is negative (statistically insignificant), suggesting weak banks grow as fast as sound banks during booms.
- Interpretation:
  - Credit booms associated with decline in lending standards at system level due to relative shift of credit from sound banks to less sound banks, rather than uniform easing of individual bank standards.
  - Example linkage: U.S. subprime mortgage market loan originations shifted to non-traditional lenders with less experience and greater funding risk exposure (Dell’Ariccia, Igan, and Laeven, 2008).

### Heterogeneity by bank growth rates
- Splitting sample into rapidly-growing banks vs. slowly-growing banks (Table 4):
  - Negative impact of credit growth on bank soundness is statistically significant only during credit booms and for banks expanding faster than the average bank in the country in a given year (Table 4, Column 2 vs. Columns 4, 6, and 8).
  - Slowly-growing banks during non-boom periods: statistically significant positive relationship between bank soundness and credit growth rates.
- Mechanism suggested:
  - Weaker banks “gamble to survive” by extending loan portfolios via riskier loans during rapid credit expansion to maintain market share; this helps explain why financial crises are often preceded by credit booms.

### Bank ownership and cross-border considerations
- Regressions separately for foreign and domestic banks (Table 5):
  - Foreign-owned banks: loan growth is not statistically related to their soundness level (Table 5, Column 3 vs. Column 9), suggesting more aggressive expansion as new-comers.
  - Credit growth significantly weakens domestic banks but not foreign banks (Table 5, Column 4 vs. Column 10).
- Interpretation:
  - Foreign banks may take on more risks via rapid expansion but possess skills and resources (better access to wholesale funding and superior risk management via parents) to manage these risks.
  - Risks taken by foreign banks may be aligned with the soundness level of their parents (Igan and Tamirisa, 2008).
  - Cross-border cooperation in bank supervision remains crucial as foreign banks’ accumulation of risks and reliance on parent resources could generate spillovers.
- Data limitation noted: foreign bank organization mode (branch vs. subsidiary) is not widely available in the data set.

### Concluding remarks (summary of implications)
- Rapid credit growth during the last decade weakened banks and became less dependent on bank soundness; effect most pronounced for banks growing particularly fast and during overall credit booms.
- Foreign banks appeared willing to take on greater risks than domestic banks, indicating potential cross-border spillovers.
- Principal mechanism linking credit booms to financial instability: rapid growth of weak banks during booms shifts credit provision toward banks with less sound fundamentals, which begets crises.
- Empirical illustration: countries with higher growth rates, such as Latvia and Bulgaria, were caught in a more vulnerable state during the recent financial crisis.

### Appendix I — Distance to Default (DD) measures
- Definition used in this paper:
  - DD ≡ (k + μ) / σ, where:
    - k is equity capital as percent of assets,
    - μ is average return as percent on assets,
    - σ is the standard deviation of return on assets (proxy for return volatility).
  - DD measures number of standard deviations a return realization must fall to exhaust equity, assuming normally distributed returns. Higher DD implies lower probability of insolvency.
- Methodological choices:
  - While option-pricing, market-data-based DD measures exist (Black and Scholes, 1973; Merton, 1974), they require liquid stock markets.
  - Given many sample countries may have illiquid bank stock markets, this paper uses a simpler annual DD based on balance sheet and income statement data:
    - Annual data on equity capital (valued at end-year market prices) and return on assets.
    - Standard deviation of returns is calculated for the entire sample period to obtain a sufficiently long-term view on risks for a given bank.
- Caveats and shortcomings of DD measure:
  - Because DD depends on average return to assets, aggressive banks expanding portfolios may appear sounder than they actually are.
  - DD likely performs better where bank stocks are actively traded and market equity values reflect investor perceptions.
  - Normality assumption for returns may not hold, as bank returns may be skewed.
- Note: This balance-sheet-based measure is sometimes called z-score to differentiate it from option-price-based distance to default.

*Source: _wp11278 - Appendix II.*

### Appendix II. Data Sources and Definitions

### Appendix II. Data Sources and Definitions

### Macroeconomic and bank-level data sources
- Macroeconomic data: taken from the March 2007 version of the IMF’s International Financial Statistics.
- Bank-level data: downloaded from the March 2007 version of BankScope and cleaned by:
  - carefully matching bank identities and deleting duplicate entries;
  - deleting entries with possible measurement errors.
- Bank ownership data: complemented from various sources, such as Euromoney and banks’ websites.

### Matching bank identifiers
- BankScope uses a unique identifier for each bank; the identifier:
  - remains unchanged when the bank’s name changes;
  - sometimes remains unchanged when the bank is merged with or acquired by another bank;
  - is reassigned only if a merger or acquisition intrinsically changes the bank.
- Data for 2002–05 were downloaded using the March 2007 update of BankScope and merged with the historical data set provided by Ugo Panizza using unique identifiers and cross-checking based on the 2002 data.

### Avoiding duplications
- BankScope includes both consolidated and unconsolidated balance sheet data; when both are available for the same bank, a different identifier is assigned to each type of data.
- At mergers, pre-merger banks and merged entities may coexist in the data set; the following procedure was applied to avoid duplicated observations:
  - Using the “rank” variable in BankScope, non-ranked banks were dropped to avoid duplications.
  - If a non-ranked bank had assets greater than the country average, its history of mergers and acquisitions was examined.
  - Premerger banks were re-ranked to ensure inclusion in the data set; post-merger banks were de-ranked to exclude them from the premerger period.
  - When both consolidated and unconsolidated balance sheets existed, the unconsolidated data were preserved to identify individual banks.
  - If unconsolidated data were unavailable, consolidated data were used to avoid dropping banks from the sample.

### Excluding outliers
- To reduce influence of measurement errors and misreporting, 5 percent of the observations on the tails of the distributions of the two main variables (bank-level credit growth and distance to default) were dropped.

### Coding ownership
- BankScope provides only current-year shares held by foreign and public investors and does not provide historical ownership information.
- Historical ownership data up to 2002 were available from Micco, Panizza and Yañez (2004).
- For 2003, 2004, and 2005, ownership information was updated using:
  - the most recent ownership information from BankScope data on NMS banks;
  - banks’ websites; and
  - BankScope data on parent banks.

### Observations (country — Observations — Region — World Bank classification)
- Algeria — 32 — North Africa — Upper middle income
- Argentina — 891 — Latin America — Upper middle income
- Australia — 299 — Pacific — High income: OECD
- Austria — 1,546 — Western Europe — High income: OECD
- Bangladesh — 268 — South Asia — Low income
- Belgium — 697 — Western Europe — High income: OECD
- Bolivia — 137 — Latin America — Lower middle income
- Brazil — 1,422 — Latin America — Upper middle income
- Bulgaria — 211 — Eastern Europe — Upper middle income
- Canada — 507 — North America — High income: OECD
- Chile — 290 — Latin America — Upper middle income
- China — 342 — East Asia — Lower middle income
- Colombia — 308 — Latin America — Upper middle income
- Costa Rica — 234 — Latin America — Upper middle income
- Croatia — 344 — Eastern Europe — High income: non-OECD
- Cyprus — 154 — Eastern Europe — High income: non-OECD
- Czech Republic — 249 — Eastern Europe — High income: OECD
- Denmark — 1,021 — Western Europe — High income: OECD
- Dominican Republic — 225 — Caribbean — Upper middle income
- Ecuador — 250 — Latin America — Lower middle income
- Egypt — 326 — North Africa — Lower middle income
- El Salvador — 152 — Latin America — Lower middle income
- Estonia — 47 — Eastern Europe — High income: non-OECD
- Finland — 104 — Western Europe — High income: OECD
- France — 3,244 — Western Europe — High income: OECD
- Germany — 18,703 — Western Europe — High income: OECD
- Greece — 180 — Western Europe — High income: OECD
- Guatemala — 338 — Latin America — Lower middle income
- Honduras — 178 — Latin America — Lower middle income
- Hong Kong — 453 — East Asia — High income: non-OECD
- Hungary — 270 — Eastern Europe — High income: OECD
- Iceland — 76 — Western Europe — High income: OECD
- India — 751 — South Asia — Lower middle income
- Indonesia — 631 — East Asia — Lower middle income
- Ireland — 372 — Western Europe — High income: OECD
- Israel — 197 — Middle East — High income: non-OECD
- Italy — 6,021 — Western Europe — High income: OECD
- Jamaica — 108 — Caribbean — Upper middle income
- Japan — 5,076 — East Asia — High income: OECD
- Jordan — 130 — Middle East — Lower middle income
- Kazakhstan — 167 — Central Asia — Upper middle income
- Kenya — 315 — Sub-Saharan Africa — Low income
- Korea — 197 — East Asia — High income: OECD
- Kuwait — 83 — Middle East — High income: non-OECD
- Latvia — 227 — Eastern Europe — Upper middle income
- Lebanon — 568 — Middle East — Upper middle income
- Lithuania — 107 — Eastern Europe — Upper middle income
- Luxembourg — 1,173 — Western Europe — High income: OECD
- Malaysia — 377 — East Asia — Upper middle income
- Malta — 68 — Western Europe — High income: non-OECD
- Mauritius — 74 — Sub-Saharan Africa — Upper middle income
- Mexico — 421 — Latin America — Upper middle income
- Morocco — 118 — North Africa — Lower middle income
- Nepal — 92 — South Asia — Low income
- Netherlands — 502 — Western Europe — High income: OECD
- New Zealand — 86 — Pacific — High income: OECD
- Nigeria — 391 — Sub-Saharan Africa — Lower middle income
- Norway — 442 — Western Europe — High income: OECD
- Oman — 91 — Middle East — High income: non-OECD
- Pakistan — 229 — South Asia — Lower middle income
- Panama — 623 — Latin America — Upper middle income
- Papua New Guinea — 39 — Pacific — Lower middle income
- Paraguay — 200 — Latin America — Lower middle income
- Peru — 182 — Latin America — Upper middle income
- Philippines — 343 — East Asia — Lower middle income
- Poland — 442 — Eastern Europe — Upper middle income
- Portugal — 335 — Western Europe — High income: OECD
- Romania — 235 — Eastern Europe — Upper middle income
- Russia — 977 — Eastern Europe — Upper middle income
- Saudi Arabia — 104 — Middle East — High income: non-OECD
- Senegal — (not listed in source) — (not listed in source) — (not listed in source)
- Singapore — 152 — East Asia — High income: non-OECD
- Slovak Republic — 177 — Eastern Europe — High income: OECD
- Slovenia — 199 — Eastern Europe — High income: non-OECD
- South Africa — 309 — Sub-Saharan Africa — Upper middle income
- Spain — 1,641 — Western Europe — High income: OECD
- Sri Lanka — 91 — South Asia — Lower middle income
- Sweden — 632 — Western Europe — High income: OECD
- Switzerland — 2,196 — Western Europe — High income: OECD
- Thailand — 180 — East Asia — Lower middle income
- Tunisia — 139 — North Africa — Lower middle income
- Turkey — 386 — Eastern Europe — Upper middle income
- Ukraine — 278 — Eastern Europe — Lower middle income
- United Arab Emirates — 197 — Middle East — High income: non-OECD
- United Kingdom — 2,048 — Western Europe — High income: OECD
- USA — 7,112 — North America — High income: OECD
- Uruguay — 283 — Latin America — Upper middle income
- Venezuela — 400 — Latin America — Upper middle income
- Vietnam — 182 — East Asia — Low income
- Zambia — 90 — Sub-Saharan Africa — Low income
- Zimbabwe — 105 — Sub-Saharan Africa — Low income

*Appendix II. Data Sources and Definitions — source document.*

### Appendix Table. List of Countries in the Sample

### _wp11278 - Appendix Table. List of Countries in the Sample

### Bank-level variables — Summary statistics (selected highlights)
- Sample coverage: Obs. counts vary by variable and period; overall bank-level sample spans 1995-2005.
- Credit growth (percent change in total loans):
  - 1995-2005: Obs. 49422, Mean 7.42, Std. Dev. 139.51
  - 1995-2000: Obs. 25090, Mean 4.70, Std. Dev. 29.34
  - 2001-2005: Obs. 24332, Mean 10.22, Std. Dev. 196.54
- Distance to default (number of standard deviations by which ROA should fall to deplete equity):
  - 1995-2005: Obs. 49422, Mean 18.53, Std. Dev. 22.29
  - 1995-2000: Obs. 25090, Mean 7.15, Std. Dev. 13.69
  - 2001-2005: Obs. 24332, Mean 30.26, Std. Dev. 23.33
- Net interest margin (interest income less interest expense divided by average earning assets):
  - 1995-2005: Obs. 47819, Mean 0.04, Std. Dev. 0.16
  - 1995-2000: Obs. 23828, Mean 0.04, Std. Dev. 0.04
  - 2001-2005: Obs. 23991, Mean 0.03, Std. Dev. 0.22
- Cost-to-income ratio:
  - 1995-2005: Obs. 48639, Mean 66.94, Std. Dev. 38.90
  - 1995-2000: Obs. 24629, Mean 65.50, Std. Dev. 40.12
  - 2001-2005: Obs. 24010, Mean 68.43, Std. Dev. 37.56
- Liquidity ratio:
  - 1995-2005: Obs. 45092, Mean 0.23, Std. Dev. 0.18
  - 1995-2000: Obs. 22691, Mean 0.23, Std. Dev. 0.18
  - 2001-2005: Obs. 22401, Mean 0.23, Std. Dev. 0.18
- Size (logarithm of total assets):
  - 1995-2005: Obs. 49422, Mean 6.77, Std. Dev. 1.83
  - 1995-2000: Obs. 25090, Mean 6.69, Std. Dev. 1.82
  - 2001-2005: Obs. 24332, Mean 6.87, Std. Dev. 1.84
- Foreign ownership (percent of equity held by foreign entities):
  - 1995-2005: Obs. 49420, Mean 12.30, Std. Dev. 31.75
  - 1995-2000: Obs. 25090, Mean 12.25, Std. Dev. 31.62
  - 2001-2005: Obs. 24330, Mean 12.36, Std. Dev. 31.88
- Public ownership (percent of equity held by public entities):
  - 1995-2005: Obs. 49420, Mean 3.69, Std. Dev. 17.70
  - 1995-2000: Obs. 25090, Mean 4.24, Std. Dev. 18.98
  - 2001-2005: Obs. 24330, Mean 3.13, Std. Dev. 16.24

Notes on bank-level data:
- Bank-level variables calculated from BankScope.
- Data trimmed by dropping 5 percent of observations on tails of distributions for bank-level credit growth and distance to default.
- Definitions: Credit growth = percent change in total loans; Distance to default = number of standard deviations by which return on assets should fall to completely deplete equity; Size = logarithm of total assets.

### Country-level variables — Summary statistics (selected highlights)
- Sample coverage: Country-level panels and subsets reported for 1995-2005, 1995-2000, and 2001-2005.
- Credit growth (country-level):
  - 1995-2005: Obs. 985, Mean 20.86, Std. Dev. 359.23
  - 1995-2000: Obs. 540, Mean 30.25, Std. Dev. 484.92
  - 2001-2005: Obs. 445, Mean 9.46, Std. Dev. 17.29
- Credit boom (indicator; definition below):
  - 1995-2005: Obs. 985, Mean 0.13, Std. Dev. 0.34
  - 1995-2000: Obs. 540, Mean 0.13, Std. Dev. 0.34
  - 2001-2005: Obs. 445, Mean 0.13, Std. Dev. 0.34
- Real GDP growth:
  - 1995-2005: Obs. 990, Mean 3.90, Std. Dev. 6.19
  - 1995-2000: Obs. 540, Mean 3.37, Std. Dev. 6.78
  - 2001-2005: Obs. 450, Mean 4.54, Std. Dev. 5.32
- Real interest rate:
  - 1995-2005: Obs. 690, Mean 11.46, Std. Dev. 14.07
  - 1995-2000: Obs. 336, Mean 14.29, Std. Dev. 14.82
  - 2001-2005: Obs. 354, Mean 8.77, Std. Dev. 12.77
- Real depreciation:
  - 1995-2005: Obs. 607, Mean 0.85, Std. Dev. 14.44
  - 1995-2000: Obs. 276, Mean 3.63, Std. Dev. 11.64
  - 2001-2005: Obs. 331, Mean -1.46, Std. Dev. 16.07
- GDP per capita:
  - 1995-2005: Obs. 990, Mean 109.21, Std. Dev. 126.21
  - 1995-2000: Obs. 540, Mean 106.17, Std. Dev. 121.78
  - 2001-2005: Obs. 450, Mean 112.85, Std. Dev. 131.37
- Credit-to-GDP ratio:
  - 1995-2005: Obs. 985, Mean 56.98, Std. Dev. 42.16
  - 1995-2000: Obs. 540, Mean 52.78, Std. Dev. 40.43
  - 2001-2005: Obs. 445, Mean 62.08, Std. Dev. 43.68

Notes on country-level data:
- Country-level variables calculated using IMF International Financial Statistics.
- Credit boom definition: indicator = 1 if deviation of credit-to-GDP ratio from backward-looking, country-specific, cubic time trend exceeds 1.5 times the historical standard deviation AND the growth rate of credit-to-GDP ratio is greater than 10 percent OR the growth rate of credit-to-GDP ratio is greater than 20 percent.
- Five country groups based on World Bank classification.

### Key regression findings — Baseline (Table 2: 1995-2005, 1995-2000, 2001-2005)
- Estimation method: 3SLS in a two-equation system; dependent variables: (1) annual percentage change in outstanding loans; (2) distance to default. All RHS variables (except foreign and public ownership) lagged by one year. Robust standard errors in brackets; significance: ***, **, * denote 1 percent, 5 percent, 10 percent.
- Selected coefficients (1995-2005, Table 2 columns (1)–(6)):
  - Credit growth → Credit Growth equation: Coefficient 0.067**, [0.032]
  - Credit growth → Bank Soundness equation: Coefficient -0.017, [0.011]
  - Distance to default → Credit Growth equation: Coefficient 0.091***, [0.007]
  - Distance to default → Bank Soundness equation: Coefficient 0.951***, [0.002]
  - Real GDP growth → Credit Growth equation: Coefficient 0.688***, [0.062]
  - GDP per capita → Credit Growth equation: Coefficient -0.038***, [0.002]
  - Net interest margin → Credit Growth equation: Coefficient 31.43***, [4.448]
  - Real interest rate → Credit Growth equation: Coefficient -0.203***, [0.029]
  - Real depreciation → Credit Growth equation: Coefficient -0.069***, [0.016]
  - Public ownership → Credit Growth equation: Coefficient -0.062***, [0.011]
  - Liquidity ratio → Credit Growth equation: Coefficient 1.014***, [0.333]
  - Size → Credit Growth equation: Coefficient -0.165***, [0.032]
  - Foreign ownership → Credit Growth equation: Coefficient -0.016***, [0.002]
- Fit statistics (1995-2005):
  - Observations: 28409 (columns shown)
  - R-squared: ranges reported, e.g., 0.39 (Credit Growth eq. column 1), 0.85 (Bank Soundness eq. column 2)

### Boom effect (Table 3 — regressions during credit booms vs. non-boom periods)
- Columns (1)–(6) show regression results during credit booms; columns (7)–(12) show results during moderate aggregate credit growth.
- Selected coefficients (1995-2005 boom results, Table 3 columns (1)–(2)):
  - Credit growth → Credit Growth equation (boom): 0.032**, [0.013]
  - Credit growth → Bank Soundness equation (boom): -0.605**, [0.271]
  - Distance to default → Bank Soundness equation (boom): 0.951***, [0.017]
  - Real GDP growth → Credit Growth equation (boom): 2.659***, [0.328]
  - GDP per capita → Credit Growth equation (boom): -0.082***, [0.012]
  - Net interest margin → Credit Growth equation (boom): 51.42**, [24.67]
  - Real interest rate → Credit Growth equation (boom): -0.791***, [0.170]
  - Liquidity ratio → Credit Growth equation (boom): -2.403*, [1.333]
  - Size → Credit Growth equation (boom): -0.380***, [0.132]
  - Foreign ownership → Credit Growth equation (boom): 0.001, [0.008]
- Interpretation evident from coefficients:
  - During booms, Credit growth has a positive and sometimes statistically significant association with subsequent credit growth, while the direct coefficients on Bank Soundness vary and can be negative (e.g., -0.605**).
  - Distance to default strongly predicts bank soundness (positive, highly significant).

### Rapid-bank effect (Table 4 — rapid vs. slow growers; boom vs. non-boom)
- Columns distinguish rapidly-growing banks (loan growth above country-year average) and slowly-growing banks (below average); results reported separately for boom and non-boom periods.
- Selected coefficients (1995-2005, Table 4):
  - Rapid banks, 1995-2005 overall (columns (1)–(2)):
    - Credit growth → Credit Growth: -0.141, [0.110]
    - Distance to default → Credit Growth: 0.113**, [0.057]
    - Distance to default → Bank Soundness: 0.961***, [0.007]
  - Slow banks, 1995-2005 overall (columns for slow banks in Table 3/4):
    - Credit growth → Credit Growth (slow): 0.087***, [0.023] (boom subset)
    - Credit growth → Bank Soundness (slow): -0.012**, [0.005] (boom subset)
- Liquidity ratio, size, and ownership coefficients vary by subgroup; distance to default consistently predicts bank soundness positively and significantly across subgroups.

### Ownership effect (Table 5 — foreign vs. domestic banks)
- Banks classified as foreign (equity share of foreigners > 50 percent) vs. domestic (domestic investors > 50 percent); results reported for 1995-2005, 1995-2000, 2001-2005.
- Selected coefficients (1995-2005):
  - Foreign banks (columns (1)–(2)):
    - Credit growth → Credit Growth: 0.019***, [0.005]
    - Distance to default → Credit Growth: 0.760, [0.526]
    - Distance to default → Bank Soundness: 0.961***, [0.007]
  - Domestic banks (columns (7)–(8)):
    - Credit growth → Credit Growth: 0.120***, [0.033]
    - Credit growth → Bank Soundness: -0.019, [0.013]
    - Distance to default → Bank Soundness: 0.950***, [0.003]
- Other notable regularities:
  - Net interest margin, real interest rate, real depreciation, liquidity ratio, size, and ownership variables show heterogeneous impacts across foreign vs. domestic banks and across subperiods.
  - Distance to default remains a strong positive predictor of bank soundness across ownership groups.

### Methodological and definitional notes (from source)
- Regressions estimated by 3SLS in a two-equation system. Dependent variables:
  - Equation 1: annual percentage change in outstanding loans.
  - Equation 2: distance to default.
- All RHS variables except foreign and public ownership are lagged by one year.
- Robust standard errors reported in brackets; significance levels: *** 1 percent, ** 5 percent, * 10 percent.
- Data sources: BankScope; IMF International Financial Statistics.
- Data cover 90 countries in total; five country groups based on World Bank classification.

*Source: _wp11278 - Appendix Table. List of Countries in the Sample*

### References

### _wp11278 - References

### Empirical and Econometric Methods
- Arellano, Manuel, 1990, “Testing for Autocorrelation in Dynamic Random Effects Models,” Review of Economic Studies, Vol. 57, No. 1, pp. 127–34.
- Arellano, Manuel, and Stephen Roy Bond, 1991, “Some Tests of Specification for Panel Data: Monte Carlo Evidence and an Application to Employment Equations,” Review of Economic Studies, Vol. 58, No. 2, pp. 277–97.
- Hall, S.G., 1987, “A Forward “Looking Model of the Exchange Rate,” Journal of Applied Econometrics, Vol. 2, No. 1, pp. 47–60.
- Sab, Randa, and Stephen C. Smith, 2002, “Human Capital Convergence: A Joint Estimation Approach,” IMF Staff Papers, Vol. 49, No 2, pp. 200–11.

### Panel Data, Dynamic Models, and Specification Testing
- Arellano, Manuel, 1990, “Testing for Autocorrelation in Dynamic Random Effects Models,” Review of Economic Studies, Vol. 57, No. 1, pp. 127–34.
- Arellano, Manuel, and Stephen Roy Bond, 1991, “Some Tests of Specification for Panel Data: Monte Carlo Evidence and an Application to Employment Equations,” Review of Economic Studies, Vol. 58, No. 2, pp. 277–97.

### Banking, Credit Growth, and Lending Behavior
- Aydin, Burcu, 2006, “Banking Structure and Credit Growth in Central and Eastern European Countries,” IMF Working Paper No. 08/215 (Washington: International Monetary Fund).
- Berger, Allen N., and Gregory F. Udell, 2004, “The Institutional Memory Hypothesis and the Procyclicality of Bank Lending Behavior,” Journal of Financial Intermediation, Vol. 13, No. 4, pp. 458-95.
- Cottarelli, C., G. Dell’Ariccia, and I. Vladkova-Hollar, 2005, “Early Birds, Late Risers, and Sleeping Beauties: Bank Credit Growth to the Private Sector in Central and Eastern Europe and in the Balkans,” Journal of Banking and Finance, Vol. 29 (January), pp. 83–104.
- de Haas, Ralph, and Iman van Lelyveld, 2005, “Foreign Banks and Credit Stability in Central and Eastern Europe,” Journal of Banking and Finance, Vol. 30 (July), pp. 152-92.
- Dell’Ariccia, Giovanni, and Robert Marquez, 2006, “Lending Booms and Lending Standards,” Journal of Finance, Vol. 61, No. 5, pp. 2511-46.
- Dell’Ariccia, Giovanni, Deniz Igan, and Luc Laeven, 2008, “Credit Booms and Lending Standards: Evidence from the Subprime Mortgage Market,” IMF Working Paper No. 08/106 (Washington: International Monetary Fund).
- Igan, Deniz, and Natalia Tamirisa, 2008, “Are Weak Banks Leading Credit Booms? Evidence from Emerging Europe,” Comparative Economic Studies, Vol. 50, pp. 599-619.
- Neir, Erlend, and Lea Zicchino, 2006, “Bank Weakness, Loan Supply and Monetary Policy,” IMF Seminar Series No 2006-11 (Washington: International Monetary Fund).

### Credit Booms, Crises, and Systemic Risk
- Barajas, Adolfo, Giovanni Dell’Ariccia, and Andrei Levchenko, 2009, “Credit Booms: the Good, the Bad, and the Ugly,” unpublished (Washington: International Monetary Fund).
- Dell’Ariccia, Giovanni, Deniz Igan, and Luc Laeven, 2008, “Credit Booms and Lending Standards: Evidence from the Subprime Mortgage Market,” IMF Working Paper No. 08/106 (Washington: International Monetary Fund).
- Dell’Ariccia, Giovanni, Enrica Detragiache, and Raghuram Rajan, 2005, “The Real Effect of Banking Crises,” IMF Working Paper No. 05/63 (Washington: International Monetary Fund).
- Gourinchas, P.O., R. Valdes, and O. Landerretche, 2001, “Lending Booms: Latin America and the World,” Economia, Spring Issue, pp. 47–99.
- Laeven, Luc, and Fabian Valencia, 2008, “Systemic Banking Crises: A New Database,” IMF Working Paper No. 08/224 (Washington: International Monetary Fund).
- Mendoza, Enrique G. and Marco Terrones, 2008, “An Anatomy of Credit Booms: Evidence from Macro Aggregates and Micro Data,” IMF Working Paper No. 08/226 (Washington: International Monetary Fund).
- Maechler, Andrea, Srobona Mitra, and DeLisle Worrell, 2010, “Decomposing Financial Risks and Vulnerabilities in Emerging Europe,” IMF Staff Papers, Vol. 57, No. 1, pp. 25-60.

### Regional Studies: Central and Eastern Europe; Emerging Europe; Latin America
- Aydin, Burcu, 2006, “Banking Structure and Credit Growth in Central and Eastern European Countries,” IMF Working Paper No. 08/215 (Washington: International Monetary Fund).
- Cottarelli, C., G. Dell’Ariccia, and I. Vladkova-Hollar, 2005, “Early Birds, Late Risers, and Sleeping Beauties: Bank Credit Growth to the Private Sector in Central and Eastern Europe and in the Balkans,” Journal of Banking and Finance, Vol. 29 (January), pp. 83–104.
- Égert, Balázs, Peter Backé, and Tina Zumer, 2006, “Credit Growth in Central and Eastern Europe: New (Over) Shooting Stars?” unpublished (Vienna: Oesterreichische Nationalbank).
- Igan, Deniz, and Natalia Tamirisa, 2008, “Are Weak Banks Leading Credit Booms? Evidence from Emerging Europe,” Comparative Economic Studies, Vol. 50, pp. 599-619.
- Gourinchas, P.O., R. Valdes, and O. Landerretche, 2001, “Lending Booms: Latin America and the World,” Economia, Spring Issue, pp. 47–99.
- Micco, Alejandro, Ugo Panizza, and Monica Yañez, 2004, “Bank Ownership and Performance,” IADB Working Paper No. 518 (Washington: Inter-American Development Bank).

### Financial Stability, Supervision, and Prudential Policies
- Danmarks National bank, 2004, “Market-Based Risk Measures for Banks,” in Financial Stability 2004 (Copenhagen: Danmarks National Bank).
- Hilbers, Paul, and others, 2005, “Assessing and Managing Rapid Credit Growth and the Role of Supervisory and Prudential Policies,” IMF Working Paper No. 05/151 (Washington: International Monetary Fund).
- De Nicoló, Gianni, and others, 2005, “European Financial Integration, Stability and Supervision,” in Euro Area Policies; Selected Issues, by Anthony Annett and others, IMF Country Report No. 05/266 (Washington: International Monetary Fund).
- Schadler, S., and others, 2004, Adopting the Euro in Central Europe: Challenges of the Next Step in European Integration, IMF Occasional Paper No. 234 (Washington: International Monetary Fund).

### Asset Pricing and Credit Risk Models
- Black, F., and M. Sholes, 1973, “The Pricing of Options and Corporate Liabilities,” Journal of Political Economy, Vol. 81, No. 3, pp. 637–54.
- Merton, R. C., 1974, “On the Pricing of Corporate Debt: The Risk Structure of Interest Rates,” Journal of Finance, Vol. 29, No. 2, pp. 449–70.

### Miscellaneous Studies and Conference Papers
- Coricelli, Fabrizio, and Igor Masten, 2004, “Growth and Volatility in Transition Countries: The Role of Credit,” a paper presented at the IMF conference in honor of Guillermo Calvo, Washington, April 15–16, 2004.
- Kónya, István, and Hiroshi Ohashi, 2005, “International Consumption Patterns among High-Income Countries: Evidence from the OECD Data,” CIRJE Discussion Papers, CIRJE-F-327 Tokyo: Center for International Research on the Japanese Economy.
- Micco, Alejandro, Ugo Panizza, and Monica Yañez, 2004, “Bank Ownership and Performance,” IADB Working Paper No. 518 (Washington: Inter-American Development Bank).

*References as listed in the source PDF.*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp11278.pdf_