## _wp12292

## Source details

**Canonical URL:** [_wp12292](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp12292.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp12292.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp12292.pdf.json)

---

### Introduction and research objective
- The 2007–09 crisis originated in developed economies and produced a global recession with exceptional declines in output, employment, and trade.
- Recovery defined as the state of the economy at the end of 2010 versus the global crisis episode (2008–09).
- Main goal: characterize firm performance after the crisis (the recovery period) and identify factors affecting vulnerability and resilience at the firm, sector, and country levels.
- Macroeconomic context quantitative facts:
  - World output in 2009 (year over year) decreased by 0.5 percent.
  - Collapse in advanced economies in 2009 was around 3.5 percent.
  - World trade volume declined by almost 12 percent.

### Empirical strategy and key variables
- Estimation approach:
  - Ordinary least-squares (OLS) cross-sectional regression to explain percentage change of real sales in 2010 with respect to 2009.
  - Dependent variable: mkji,Growth,, = percentage change of real sales in 2010 with respect to 2009.
  - Control for depth of decline: Growth_09 (mkji,Growth,09) = percentage change in real sales in 2009 with respect to 2008.
  - All explanatory variables measured as of 2007 (unless otherwise indicated).
  - Outlier trimming: observations above the 97.5th percentile and below the 2.5th percentile for each variable by country at the firm level are discarded.
- Regression specification (variables preserved as in source):
  - mkjimkjimkjimkjmkji TangLevSizeGrowth ,,,3,,,2,,,1,,,, 
  - mkjimkjimkji GrowthStdebt ,,,,,,5,,,4 09 
- Firm-level explanatory variables (definitions preserved):
  - mkji,Size,, = natural log of total assets (in millions of U.S. dollars).
  - mkji,Lev,, = leverage, the ratio of total liabilities to total assets.
  - mkji,Tang,, = tangibility, the ratio of fixed assets (gross property, plant, and equipment less accumulated reserves for depreciation, depletion, and amortization) to total assets.
  - mkji,Stdebt,, = ratio of short-term debt and the portion of long-term debt payable within one year to total debt.
- Theoretical rationale:
  - High leverage can impair access to credit and amplify downturns (financial accelerator literature).
  - Tangible assets facilitate collateralized lending and access to external finance during crises.
  - High short-term debt creates refinancing risk and worsens performance during stressed credit conditions.
  - Size effects ambiguous: small firms may grow faster but have higher exit risk; larger firms may benefit from reputation, collateral, and accounting transparency.

### Data description and sample
- Sample scope and exclusions:
  - Database contains data for 6,581 publicly listed manufacturing firms in 48 countries.
  - Coverage: 21 advanced economies and 27 emerging economies.
  - Sample restricted to manufacturing sector (SIC codes 200 to 399).
  - Financial firms excluded.
- Data source and characteristics:
  - Data based on annual reports from Worldscope (Thomson Financial).
  - Balanced panel: all firms have observations for all variables.
  - Data set covers mostly large, publicly listed firms.
  - SIC industry grouping used for sectoral analysis.

### Descriptive statistics and key sample characteristics (exact figures)
- Median real sales growth among sample firms:
  - Full sample (2010): 4.36 percent.
  - Crisis year (2009): –5.5 percent.
  - 2007 median: around 8 percent.
- By degree of development:
  - Advanced economies: median growth in 2010 was around –3 percent; decline in 2009 was about –8 percent.
  - Emerging economies: growth rates were 9.5 percent in 2010 and –3.8 percent in 2009.
- Leverage and growth in 2010 (median real sales growth by leverage):
  - Low leverage firms (below median leverage): 6.3 percent.
  - High leverage firms: 1.9 percent.
  - Emerging economies: low leverage 11.8 percent; high leverage 6.4 percent.
  - Advanced economies: low leverage –2.1 percent; high leverage –3.8 percent.
- Share of firms with negative growth:
  - 2009: more than 60 percent of sample firms had negative growth.
    - Advanced economies (2009): 66 percent negative.
    - Emerging economies (2009): 55 percent negative.
  - 2010: around 37 percent negative overall.
    - Advanced economies (end-2010): 46 percent negative.
    - Emerging economies (end-2010): 35 percent negative.
- Pre-crisis (2007) medians for sample firm characteristics:
  - Firm size (natural logarithm of total assets, in U.S. dollars): advanced economies almost 6; emerging economies 4.71.
  - Leverage (total liabilities over total assets): advanced economies 0.52; emerging economies 0.50.
  - Tangibility (fixed assets over total assets): emerging economies 0.34; advanced economies 0.26.
  - Short-term debt ratio (short-term debt over total debt): emerging economies 0.7; advanced economies 0.5.
- Summary statistics (Table 3, all values measured in 2007 unless noted):
  - Full Sample (Obs = 6,581): Growth mean 17.69; Median 4.36; Std. Dev 27.04; p25 -8.85; p75 20.05.
  - Growth09 mean -2.12; Median -5.57; Std. Dev 54.15; p25 -17.76; p75 6.75.
  - Size mean 15.15; Median 5.04; Std. Dev 1.67; p25 4.00; p75 6.16.
  - Lev mean 0.51; Median 0.51; Std. Dev 0.18; p25 0.38; p75 0.63.
  - Tang mean 0.33; Median 0.31; Std. Dev 0.16; p25 0.19; p75 0.44.
  - Stdebt mean 0.61; Median 0.63; Std. Dev 0.30; p25 0.36; p75 0.90.
  - Emerging Economies (Obs = 4,198): Growth mean 12.95; Median 9.50; Std. Dev 28.58; p25 -4.73; p75 26.78.
  - Advanced Economies (Obs = 2,383): Growth mean -1.57; Median -2.98; Std. Dev 21.11; p25 -13.76; p75 7.87.
  - Sector Tradesens mean 2.01; Median 1.21; Std. Dev 1.08; p25 0.43; p75 1.43.
  - Country Deprec (REER change 2007–2010, in percent) Obs 48; Mean -3.65; Median -2.74; Std. Dev 10.67; p25 -9.54; p75 1.97.

### Main empirical results (baseline regressions and magnitudes)
- Baseline sample: 48 countries (Observations = 6,581).
- Leverage:
  - Statistically significant negative effect of pre-crisis (2007) leverage on 2010 corporate performance.
  - An increase of 10 percentage points in a sample firm’s leverage decreases its real sales growth, on average, by almost 0.7 percentage points.
  - Regression coefficients (Table 4) include β2: -5.90***; -6.23***; -6.59***; -6.76*** across specifications.
- Tangibility:
  - Statistically significant positive effect of pre-crisis tangibility on 2010 firm growth.
  - An increase of 10 percentage points in firm’s tangibility is associated with an increase in real sales growth, on average, by around 0.8 percentage points.
  - Regression coefficients (Table 4) include β3: 9.12***; 7.76***; 7.60***.
- Short-term debt:
  - Statistically significant negative effect of pre-crisis short-term debt (over total debt) on 2010 corporate performance.
  - An increase of 10 percentage points in a firm short-term debt ratio decreases its real sales growth, on average, by over 0.3 percentage points.
  - Regression coefficients (Table 4) include β4: -3.12**; -3.18** and β5: -0.02***.
- Rebound and firm size:
  - Small but statistically significant rebound effect (firms’ growth in 2009).
  - Firm size average effect statistically non significant in full sample (Table 4).

### Subsample comparisons: emerging versus advanced economies
- Leverage:
  - Negative effect present and slightly larger in emerging economies subsample.
  - Emerging β2: -6.29**; -6.82***; -6.89***; -7.11*** (Table 5).
  - Advanced β2: -5.41**; -5.51**; -6.64**; -6.47** (Table 6).
- Tangibility:
  - Positive and statistically significant among emerging economies (~1.1 percentage points; Table 5).
  - Not statistically different from zero in advanced economies subsample (Table 6).
- Short-term debt:
  - Negative and statistically significant effect in advanced economies of around 0.6 percentage points for each 10 percentage point increase in the short-term debt ratio (Table 6).
  - Effect not significantly different from zero for sample firms in emerging economies (Table 5).
- Firm size:
  - No statistically significant effect in emerging economies.
  - Statistically significant negative effect in advanced economies.
- Rebound effect:
  - Stronger for sample firms in advanced economies.

### Nonlinear effects of leverage (threshold analysis)
- Methodology:
  - Hansen (2000) threshold regression framework with leverage as threshold variable.
  - Grid search over the 0.30–0.95 leverage interval (median leverage is 0.51).
  - Threshold value identified for full sample: 0.483.
  - Bootstrap p-values for inference (100, 300, 500, and 1,000 replications).
- Threshold results:
  - For leverage beyond 0.48 (full sample), an increase of leverage of ten percent is associated with a statistically significant decrease in real sales of 0.1 percentage points.
  - Table 7 (Full Sample): η: -8.17***; γ: 0.48**; β3: 7.43***; β4: -3.16**; β5: -0.02***.
  - Robustness: no statistically significant effect of short-term debt once controlling for nonlinearity in leverage.

### Exchange rate, trade-sensitivity interactions, and robustness
- Exchange rate channel motivation:
  - Liability effect: foreign-currency indebtedness raises domestic-currency value of liabilities upon depreciation.
  - Competitiveness effect: depreciation can improve export competitiveness.
- Aggregate observation (48 countries):
  - Fixed exchange rate regimes: grew in real terms on average –3.5 percent in 2009 and around 2 percent in 2010.
  - Flexible exchange rate regimes: grew in real terms on average –0.6 percent in 2009 and 5.4 percent in 2010.
- Trade-sensitivity interaction:
  - Adds interaction of sector-level trade sensitivity with percentage change in REER during 2007–2010 (REER increase interpreted as depreciation).
  - Trade sensitivity constructed by regressing change in global exports at 2-digit SIC level (2000-06) on change in global GDP; coefficients used as sector-level trade sensitivity.
- Findings (Table 8):
  - Positive and statistically significant interaction between real depreciation and trade sensitivity for full sample and both subsamples.
  - Magnitudes:
    - Full sample: a 1 percent depreciation in the REER generates an increase of 0.5 percent in real sales growth for firms in sectors more open to trade (median growth referenced as 4.36 percent).
    - Advanced economies: effect around 0.8 percent per 1 percent depreciation.
  - The competitiveness effect dominates the liability effect in these exercises.
- Additional robustness:
  - Using 2006 instead of 2007 as pre-crisis year yields consistent results.
  - Results consistent with alternative measures of leverage (total debt over total assets) and performance (EBITDA).
  - Tests allowing for nonlinearity of other explanatory variables found non statistically significant results.

### Policy-relevant conclusions and implications
- Summary findings:
  - Firm-specific characteristics—size, leverage, tangibility, and short-term debt ratio—had significant impacts on post-crisis firm performance.
  - Leverage exhibits a strong nonlinear negative effect on firm performance; above a threshold (around 0.48) higher leverage is associated with further declines in real sales.
  - Macroeconomic policy matters: countries allowing exchange rate depreciation experienced faster recovery among firms in trade-dependent sectors, as competitiveness effects dominated liability effects in the sample.
- Policy considerations derived from results:
  - Monitor and address high pre-crisis corporate leverage, especially when leverage approaches or exceeds identified thresholds.
  - Consider the composition of corporate debt (short-term versus long-term), given negative associations of short-term debt with post-crisis performance, particularly in advanced economies.
  - Exchange rate policy and flexibility can influence sectoral recoveries; depreciation may aid firms in trade-sensitive sectors through competitiveness gains, but currency-composition of firm liabilities should be monitored to manage liability effects.
  - Strengthen financial-market and regulatory frameworks in emerging economies to reduce persistent excessive leverage that may reflect market imperfections and regulatory weaknesses.

*Source: Excerpts from _wp12292 (IMF working paper PDF).*

### 1. Country and Region Coverage .........................................................................................

### _wp12292 - 1. Country and Region Coverage .........................................................................................

### Introduction
- Context and research objective
  - The 2007–09 crisis originated in developed economies and produced a global recession with exceptional declines in output, employment, and trade.
  - The paper defines the recovery as the state of the economy at the end of 2010 versus the global crisis episode (2008–09).
  - Main goal: characterize firm performance after the crisis (the recovery period) and identify factors affecting vulnerability and resilience at the firm, sector, and country levels.
- High-level empirical findings reported in the text
  - Higher levels of pre-crisis leverage and short-term debt have negative effects on firms’ real sales growth in the recovery.
  - Tangibility (ratio of fixed assets to total assets) has positive effects on recovery growth.
  - The negative effect of leverage is nonlinear and particularly strong for firms with high pre-crisis leverage.
  - Firms in countries that allowed the exchange rate to depreciate during the crisis experienced faster recovery in sectors highly dependent on trade.
- Macroeconomic context quantitative facts (as stated)
  - The world’s output in 2009 (year over year) decreased by 0.5 percent (in advanced economies, the collapse was around 3.5 percent).
  - The world trade volume declined by almost 12 percent.

### Empirical strategy and key variables
- Estimation approach
  - Ordinary least-squares (OLS) cross-sectional regression model estimated to explain percentage change of real sales in 2010 with respect to 2009.
  - Dependent variable: mkji,Growth,, = percentage change of real sales in 2010 with respect to 2009 (subindices mkji,, stand for firm, sector, country and region).
  - Control for depth of decline: Growth_09 (mkji,Growth,09) = percentage change in real sales in 2009 with respect to 2008.
  - All explanatory variables are measured as of 2007 (unless otherwise indicated) to minimize endogeneity.
  - Outlier treatment: values above the 97.5th percentile and below the 2.5th percentile for each variable by country at the firm level are discarded.
- Regression specification (variables preserved as in source)
  - mkjimkjimkjimkjmkji TangLevSizeGrowth ,,,3,,,2,,,1,,,, 
  - mkjimkjimkji GrowthStdebt ,,,,,,5,,,4 09 
- Firm-level explanatory variables (definitions preserved)
  - mkji,Size,, = natural log of total assets (in millions of U.S. dollars).
  - mkji,Lev,, = leverage, the ratio of total liabilities to total assets.
  - mkji,Tang,, = tangibility, the ratio of fixed assets (gross property, plant, and equipment less accumulated reserves for depreciation, depletion, and amortization) to total assets.
  - mkji,Stdebt,, = ratio of short-term debt and the portion of long-term debt payable within one year to total debt.
- Theoretical rationale summarized
  - High leverage can impair access to credit during adverse shocks and amplify downturns (financial accelerator literature).
  - Tangible assets facilitate collateralized lending and can improve access to external finance during crises.
  - High short-term debt (maturity mismatch) can create refinancing risk and worsen performance during stressed credit conditions.
  - Size effects are ambiguous: small firms may grow faster but have higher exit risk; larger firms may have advantages via reputation, collateral, and accounting transparency.

### Data description and sample
- Sample scope and exclusions
  - Database compiled contains data for 6,581 publicly listed manufacturing firms in 48 countries.
  - Coverage: 21 advanced economies and 27 emerging economies.
  - Sample restricted to manufacturing sector (Standard Industrial Classification [SIC] codes 200 to 399).
  - Financial firms (e.g., banks and insurance companies) are excluded.
- Data source and characteristics
  - Data are based on annual reports from Worldscope (Thomson Financial).
  - The data set is balanced (all firms have observations for all variables).
  - The data set covers mostly large firms that are publicly listed.
  - SIC industry grouping used to enable sectoral analysis of firm performance.

### Descriptive statistics and data treatment
- Heterogeneity in sample
  - Despite restriction to publicly traded manufacturing firms, there is substantial heterogeneity in firm growth before, during, and after the crisis.
- Outlier trimming (reiterated)
  - Observations above the 97.5th percentile and below the 2.5th percentile for each variable by country at the firm level are discarded.

### Main results and interpretation (as reported)
- Firm-level determinants of recovery growth
  - Higher pre-crisis leverage negatively affects firms’ real sales growth in the recovery period.
  - Higher pre-crisis short-term debt ratio negatively affects firms’ real sales growth in the recovery period.
  - Higher tangibility (fixed assets to total assets) positively affects firms’ real sales growth in the recovery period.
  - The negative effect of leverage is nonlinear—particularly pronounced for firms with high pre-crisis leverage.
- Country- and sector-level interactions
  - The macroeconomic framework matters for firm growth: firms in countries that allowed the exchange rate to depreciate during the crisis had faster recovery in sectors highly dependent on trade.
- Consistency with literature
  - Results are consistent with theoretical and empirical literature on collateral, leverage, credit constraints, and firm growth dynamics.

*Source: Excerpts from _wp12292 - 1. Country and Region Coverage (IMF working paper PDF).*

### 4.36 percent, substantially above the one during the crisis (–5.5 percent in 2009) but still

### _wp12292 - 4.36 percent, substantially above the one during the crisis (–5.5 percent in 2009) but still

### Key patterns in corporate performance and sample characteristics
- Median real sales growth among sample firms:
  - Full sample (2010): 4.36 percent.
  - Crisis year (2009): –5.5 percent (crisis median).
  - 2007 median: around 8 percent.
- By degree of development:
  - Advanced economies: median growth in 2010 was around –3 percent; decline in 2009 was about –8 percent.
  - Emerging economies: growth rates were 9.5 percent in 2010 and –3.8 percent in 2009.
- Distributional observations:
  - Corporate performance in advanced economies clustered around the mean in 2007; emerging economies presented a wider distribution around the mean.
  - In 2010, tails of the distribution for advanced economies were fatter (particularly the left tail).
- Leverage and growth in 2010 (median real sales growth by leverage):
  - Low leverage firms (below median leverage): 6.3 percent.
  - High leverage firms: 1.9 percent.
  - Emerging economies: low leverage 11.8 percent; high leverage 6.4 percent.
  - Advanced economies: low leverage –2.1 percent; high leverage –3.8 percent.
- Share of firms with negative growth (yearly percentage change in real sales):
  - 2009: more than 60 percent of sample firms had negative growth.
    - Advanced economies (2009): 66 percent negative.
    - Emerging economies (2009): 55 percent negative.
  - 2010: around 37 percent negative overall.
    - Advanced economies (end-2010): 46 percent negative.
    - Emerging economies (end-2010): 35 percent negative.
- Pre-crisis (2007) medians for sample firm characteristics:
  - Firm size (natural logarithm of total assets, in U.S. dollars): advanced economies almost 6; emerging economies 4.71.
  - Leverage (total liabilities over total assets): advanced economies 0.52; emerging economies 0.50.
  - Tangibility (fixed assets over total assets): emerging economies 0.34; advanced economies 0.26.
  - Short-term debt ratio (short-term debt over total debt): emerging economies 0.7; advanced economies 0.5.

### Empirical findings (baseline regressions)
- Main baseline sample: 48 countries (Tables referenced: Table 4 baseline; Tables 5 and 6 for subsamples).
- Leverage:
  - Statistically significant and economically meaningful negative effect of pre-crisis (2007) leverage on 2010 corporate performance.
  - An increase of 10 percentage points in a sample firm’s leverage decreases its real sales growth, on average, by almost 0.7 percentage points.
  - Median growth among sample firms referenced as around 4 percent.
- Tangibility:
  - Statistically significant and economically meaningful positive effect of pre-crisis tangibility ratio on 2010 firm growth.
  - An increase of 10 percentage points in firm’s tangibility is associated with an increase in real sales growth, on average, by around 0.8 percentage points.
- Short-term debt:
  - Statistically significant and economically meaningful negative effect of pre-crisis short-term debt (over total debt) on 2010 corporate performance.
  - An increase of 10 percentage points in a firm short-term debt ratio decreases its real sales growth, on average, by over 0.3 percentage points.
- Rebound and firm size:
  - Small but statistically significant rebound effect (firms’ growth in 2009).
  - Average statistically non significant effect of firm size in the full sample (Table 4).

- Subsample comparisons (emerging versus advanced economies):
  - Leverage:
    - Negative effect present and slightly larger in emerging economies subsample (Tables 5 and 6).
  - Tangibility:
    - Positive and statistically significant among emerging economies (~1.1 percentage points; Table 5).
    - Not statistically different from zero in advanced economies subsample (Table 6).
  - Short-term debt:
    - Negative and statistically significant effect in advanced economies of around 0.6 percentage points for each 10 percentage point increase in the short-term debt ratio (Table 6).
    - Effect not significantly different from zero for sample firms in emerging economies (Table 5).
  - Firm size:
    - No statistically significant effect on performance for sample firms in emerging economies.
    - Statistically significant negative effect of firm size on performance for sample firms in advanced economies.
  - Rebound effect:
    - Stronger for sample firms in advanced economies.

### Nonlinear effects of leverage (threshold analysis)
- Motivation: theories imply leverage can raise corporate risk, funding costs, reduce investment, amplify shocks; threshold effects may exist where high leverage has different impacts than low leverage.
- Methodology:
  - Hansen (2000) threshold regression framework applied.
  - Threshold regression model used where leverage serves as its own threshold variable.
  - Grid search over the 0.30–0.95 leverage interval (median leverage is 0.51).
  - Threshold value identified for full sample: 0.483.
  - Bootstrap p-values used for inference (robustness with 100, 300, 500, and 1,000 bootstrap replications).
- Threshold results (Table 7):
  - For leverage beyond 0.48 (full sample), an increase of leverage of ten percent is associated with a statistically significant decrease in real sales of 0.1 percentage points.
  - Robustness note: no statistically significant effect of short-term debt once controlling for the nonlinearity in leverage.

### Robustness tests: trade sensitivity and real depreciation effects
- Exchange rate channels considered: liability effect (foreign-currency indebtedness raises domestic-currency value of liabilities upon depreciation) and competitiveness effect (depreciation can improve export competitiveness).
- Aggregate observation for sample of 48 countries:
  - Countries with fixed exchange rate regimes: grew in real terms on average –3.5 percent in 2009 and around 2 percent in 2010.
  - Countries with flexible exchange rate regimes: grew in real terms on average –0.6 percent in 2009 and 5.4 percent in 2010.
- Trade-sensitivity interaction specification:
  - Regression adds interaction of sector-level trade sensitivity measure with percentage change in the real effective exchange rate (REER) during 2007–2010.
  - Trade sensitivity constructed by regressing change in global exports at 2-digit SIC level (2000-06) on change in global GDP (equation (5)); coefficients used as sector-level trade sensitivity.
  - REER change: an increase interpreted as depreciation.
- Findings (Table 8):
  - Consistent positive and statistically significant effect of the interaction between real depreciation and the trade sensitivity coefficient for the whole sample and both subsamples.
  - Magnitudes:
    - For the full sample: a 1 percent depreciation in the REER generates an increase of 0.5 percent in real sales growth for firms in sectors more open to trade (median growth referenced as 4.36 percent).
    - Advanced economies: effect around 0.8 percent per 1 percent depreciation.
  - The effect does not differ for highly leveraged firms in this exercise.
  - Interpretation: in this exercise, the competitiveness effect overpowers the liability effect.
- Additional robustness:
  - Using 2006 instead of 2007 as pre-crisis year yields consistent results (results not shown).
  - Results consistent with alternative measures of leverage (total debt over total assets) and performance (EBITDA).
  - Tests allowing for nonlinearity of other explanatory variables found non statistically significant results.

### Conclusions and policy-relevant implications
- Summary findings:
  - Firm-specific characteristics—size, leverage, tangibility, and short-term debt ratio—had significant impacts on post-crisis firm performance.
  - Leverage exhibits a strong nonlinear negative effect on firm performance; above a threshold (around 0.48 leverage) higher leverage is associated with further declines in real sales.
  - Macroeconomic policy matters: countries allowing exchange rate depreciation experienced faster recovery among firms in trade-dependent sectors, as competitiveness effects dominated liability effects in the sample exercises.
- Implicit policy considerations (derived from empirical results in the study):
  - Monitor and address high pre-crisis corporate leverage, especially when leverage approaches or exceeds identified thresholds.
  - Consider the composition of corporate debt (short-term versus long-term) given negative associations of short-term debt with post-crisis performance, particularly in advanced economies.
  - Exchange rate policy and flexibility can influence sectoral recoveries; depreciation may aid firms in trade-sensitive sectors through competitiveness gains, but currency-composition of firm liabilities should be monitored to manage liability effects.
  - Strengthen financial-market and regulatory frameworks in emerging economies to reduce persistent excessive leverage that may reflect market imperfections and regulatory weaknesses.

*Source: IMF working paper content unit _wp12292 - 4.36 percent, substantially above the one during the crisis (–5.5 percent in 2009) but still*

### References

### _wp12292 - References

### Key bibliographic citations
- Agénor, P., and P. Montiel, 2008, Development Macroeconomics, Third Edition, Princeton University Press.
- Baggs, J., Beaulieu, E. and Fung, L.,2010, “Are Service Firms Affected By Exchange Rate Movements?” Review of Income and Wealth, 56: S156–S176.
- Berger, A and E. B. di Patti, 2003,“Capital Structure and Firm Performance: A New Approach to Testing Agency Theory and an Application to the Banking Industry,” FEDS Working Paper No. 2002–54.
- Berkmen, P., G. Gelos, R. Rennhack, and J. Walsh, 2009, “The Global Financial Crisis: Explaining Cross-country Differences in the Output Impact.” IMF Working Paper WP/09/280.
- Bernanke B., M. Gertler and S. G. Gilchrist, 1994, "The Financial Accelerator and the Flight to Quality," Finance and Economics Discussion Series 94–18.
- Bernanke, B. S., Gertler, M., & Gilchrist, S., 1999, "The Financial Accelerator in a Quantitative Business Cycle Framework," Handbook of Macroeconomics, chapter 21, pages 1341–1393.
- Blanchard, O., H. Faruqee, and M. Das. 2010. “The Initial Impact of the Crisis on Emerging Market Countries.” Brookings Papers on Economic Activity.
- Bleakley H., and K. Cowan, 2009, “Maturity Mismatch and Financial Crises: Evidence from Emerging Market Corporations,” Journal of Development Economics.
- Braun, M., 2003,“Financial Contractibility and Asset Hardness,” UCLA, mimeo.
- Braun, M., and B. Larrain, 2005, “Finance and the Business Cycle: International, Inter-Industry Evidence,” Journal of Finance, 60(3), pp. 1097–1128.
- Chang, R. and A. Velasco, 1999, “Illiquidity and Crises in Emerging Markets: Theory and Policy,” NBER Macroeconomics Annual, 1999.
- Chor, Davin and Kalina Manova, 2011, “Off the Cliff and Back: Credit Conditions and International Trade During the Global Financial Crisis,” Journal of International Economics (forthcoming). NBER Working Paper 16174.
- Claessens, S., and Laeven, L., 2003, “Financial Development, Property Rights, and Growth,” Journal of Finance 58, pp. 2401–2436.
- Claessens, S., H. Tong, and S.-J. Wei, 2012, “From the Financial Crisis to the Real Economy: Using Firm-level Data to Identify Transmission Channels,” Journal of International Economics, Available online 8 March 2012.
- Cook, D., 2004, “Monetary Policy in Emerging Markets: Can Liability Dollarization Explain Contractionary Devaluation?” Journal of Monetary Economics 51 (September) pp.1155–81.
- Cooley, T. F. and V. Quadrini, 1999, “Financial Markets and Firm Dynamics,” NYU Working Paper No. S-MF-99–09.
- Cetorelli, N., and L. S. Goldberg, 2009, “Global Banks and International Shock Transmission: Evidence from the Crisis,” Federal Reserve Bank of New York Working Paper.
- Coricelli, F., Driffield, N. L., Pal, S., and Roland, I., 2009, “Excess Leverage and Productivity Growth in Emerging Economies: Is There a Threshold Effect?” IZA Discussion Paper No. 4834.
- Driffield, N.L., V. Mahambare and S. Pal, 2007, “How Does Ownership Structure Affects Capital Structure and Firm Value? Recent Evidence from East Asia,” Economics of Transition 15(3).
- Fischer, E. O., Heinkel, R., Zechner, J., 1989, “Dynamic Capital Structure Choice: Theory and Tests,” Journal of Finance, 44, pp. 19–40.
- Furman, J. and J.E. Stiglitz, 1998, “Economic Crises: Evidence and Insights from East Asia," Brookings Papers on Economic Activity 1998:2.
- Gertler, M., 1988, ‘‘Financial Structure and Aggregate Economic Activity: An Overview.’’ J. Money, Credit and Banking 20, pp.559–88.
- Gertler, M., and S. Gilchrist, 1994, “Monetary Policy, Business Cycles, and the Behavior of Small Manufacturing Firms,” The Quarterly Journal of Economics, 109(2), pp. 309-340.
- Goldstein, R., N. Ju and H. Leland, 2001, “An EBIT based Model of Dynamic Capital Structure”, Journal of Business, 74(4).
- Hansen, B. E., 2000, “Sample Splitting and Threshold Estimation,” Econometrica, Vol. 68, pp. 575–603.
- Huynh K., and R. Petrunia, 2010, “Age Effects, Leverage and Firm Growth,” Journal of Economic Dynamics & Control, 34, pp. 1003–1013.
- International Monetary Fund, World Economic Outlook (WEO), Rebalancing Growth, April 2010.
- International Monetary Fund, World Economic Outlook (WEO), Tensions from the Two-Speed Recovery: Unemployment, Commodities, and Capital Flows, April 2011.
- Kannan P., 2010, “Credit Conditions and Recoveries from Recessions Associated with Financial Crises,” IMF Working Paper 10/83.
- Kamil H., and K. Rai, 2010, “The Global Credit Crunch and Foreign Banks Lending to Emerging Markets: Why Did Latin America Fare Better?” IMF Working Paper 10/102.
- Kiyotaki, N., and Moore, J., 1995, ‘Credit Cycles, Discussion Paper No. TE/95/285; Working Paper no. 5083. NBER, April 1995.
- Kiyotaki, N. and Moore, J., 1997, "Credit Cycles,” Journal of Political Economy 105 (2): pp. 211–248.
- Laeven L., and F. Valencia, 2011, “The Real Effects of Financial Sector Interventions During Crises,” IMF Working Paper 11/45.
- Lane, P. R., and G. M. Milesi-Ferreti, 2007, “The External Wealth of Nations Mark II,” Journal of International Economics 73 No. 2 (November), pp. 223–250.
- Lang, L.E., Ofek, E., Stulz, R., 1996, “Leverage, Investment and Firm Growth,” Journal of Financial Economics, No. 40, pp. 3–29.
- McConnell, J. J., and H. Servaes, 1995, “Equity Ownership and the Two Faces of Debt,” Journal of Financial Economics 39, pp. 131–157.
- Merton, R. C., 1974, "On the Pricing of Corporate Debt: The Risk Structure of Interest Rates", Journal of Finance, Vol. 29, No. 2, pp. 449–470.
- Oliner, S. D., and G. D. Rudebusch, 1995, “Is There a Bank Lending Channel for Monetary Policy?,” Economic Review, pp. 1–20.
- Radelet, S. and J. D. Sachs, 1998, “The East Asian Financial Crisis: Diagnosis, Remedies, Prospects," Brookings Papers on Economic Activity (1): pp. 1–74.
- Rajan, R., and L. Zingales, 1995, “What Do We Know About Capital Structure: Some Evidence From International Data,” Journal of Finance, December 1995, 50 (5), pp. 1421–60.
- Rajan, R., and L. Zingales, 1998, “Financial Dependence and Growth,” American Economic Review, Vol. 88(3), pp. 559–86.
- Sachs, J., A. Tornell, and A. Velasco, 1996, “The Collapse of the Mexican Peso: What Have We Learned?" Economic Policy, 46:13–63, April.
- Strebulaev, I., 2007, “Do Tests of Capital Structure Theory Mean What They Say?” Journal of Finance 62, pp. 1747–1788.

### Country and sample coverage (exact figures)
- Emerging (28); Developed (20).
- Americas (5), Asia (13), Europe (7), Middle East and Africa (3) listed in one grouping; separate developed grouping lists Americas (1), Asia (3), Europe (16), Middle East and Africa.
- Table 2: Number of listed manufacturing firms by country for 2010. Total = 6,581 firms.
  - Examples of country counts as listed:
    - ARGENTINA 25
    - AUSTRALIA 118
    - BRAZIL 66
    - CANADA 114
    - CHINA 781
    - INDIA 906
    - JAPAN 1,152
    - KOREA (SOUTH) 297
    - TAIWAN 433
    - UNITED KINGDOM 98
    - VIETNAM 128
  - (Complete country counts presented in Table 2 sum to TOTAL 6,581.)

### Summary statistics (Table 3, all values measured in 2007 unless noted)
- Full Sample (Obs = 6,581)
  - Growth (Percentage change in real sales in 2010): Mean 17.69; Median 4.36; Std. Dev 27.04; p25 -8.85; p75 20.05
  - Growth09 (Percentage change in real sales in 2009): Mean -2.12; Median -5.57; Std. Dev 54.15; p25 -17.76; p75 6.75
  - Size (Natural logarithm of total assets in US$): Mean 15.15; Median 5.04; Std. Dev 1.67; p25 4.00; p75 6.16
  - Lev (Total liabilities over total assets): Mean 0.51; Median 0.51; Std. Dev 0.18; p25 0.38; p75 0.63
  - Tang (Fixed assets over total assets): Mean 0.33; Median 0.31; Std. Dev 0.16; p25 0.19; p75 0.44
  - Stdebt (Short-term debt over total debt): Mean 0.61; Median 0.63; Std. Dev 0.30; p25 0.36; p75 0.90
- Emerging Economies (Obs = 4,198)
  - Growth: Mean 12.95; Median 9.50; Std. Dev 28.58; p25 -4.73; p75 26.78
  - Growth09: Mean 0.17; Median -3.80; Std. Dev 63.47; p25 -17.58; p75 10.62
  - Size: Mean 14.71; Median 4.64; Std. Dev 1.47; p25 3.68; p75 5.63
  - Lev: Mean 0.50; Median 0.50; Std. Dev 0.19; p25 0.37; p75 0.63
  - Tang: Mean 0.35; Median 0.34; Std. Dev 0.17; p25 0.21; p75 0.48
  - Stdebt: Mean 0.65; Median 0.70; Std. Dev 0.29; p25 0.41; p75 0.95
- Advanced Economies (Obs = 2,383)
  - Growth: Mean -1.57; Median -2.98; Std. Dev 21.11; p25 -13.76; p75 7.87
  - Growth09: Mean -6.19; Median -8.22; Std. Dev 31.25; p25 -17.91; p75 1.41
  - Size: Mean 35.93; Median 5.79; Std. Dev 1.73; p25 4.79; p75 7.05
  - Lev: Mean 0.52; Median 0.52; Std. Dev 0.17; p25 0.39; p75 0.64
  - Tang: Mean 0.27; Median 0.26; Std. Dev 0.15; p25 0.17; p75 0.38
  - Stdebt: Mean 0.51; Median 0.49; Std. Dev 0.29; p25 0.27; p75 0.74
- Sector level
  - Tradesens (Trade sensitivity coefficient): Mean 2.01; Median 1.21; Std. Dev 1.08; p25 0.43; p75 1.43
- Country level
  - Deprec (Real depreciation between 2007 and 2010, in percent): Obs 48; Mean -3.65; Median -2.74; Std. Dev 10.67; p25 -9.54; p75 1.97

### Baseline regression results (Tables 4–6: dependent variable = annual growth in real sales in 2010 vs 2009; all explanatory variables dated 2007 except Growth09)
- Table 4. All Countries (Observations = 6,581; Adjusted R-squared across specifications: 0.228, 0.229, 0.232, 0.232, 0.234)
  - Estimated coefficients (specification numbers correspond to columns (1)–(5)):
    - β1: 0.03; 0.08; -0.00; -0.10; -0.13 (standard errors in parentheses: (0.241), (0.241), (0.231), (0.234), (0.233))
    - β2: -5.90***; -6.23***; -6.59***; -6.76*** (standard errors: (1.858), (1.869), (1.897), (1.886))
    - β3: 9.12***; 7.76***; 7.60*** (standard errors: (2.790), (2.706), (2.711))
    - β4: -3.12**; -3.18** (standard errors: (1.460), (1.461))
    - β5: -0.02*** (standard error: (0.007))
  - Country, sector, and region fixed effects included; clustering at country-industry pair level.
- Table 5. Emerging Economies (Observations = 4,198; Adjusted R-squared across specifications: 0.175, 0.176, 0.179, 0.179, 0.180)
  - β1: 0.29; 0.34; 0.19; 0.15; 0.13 (std. errors: (0.417), (0.416), (0.396), (0.396), (0.394))
  - β2: -6.29**; -6.82***; -6.89***; -7.11*** (std. errors: (2.507), (2.538), (2.548), (2.536))
  - β3: 11.00***; 10.30***; 10.23*** (std. errors: (3.608), (3.410), (3.417))
  - β4: -1.39; -1.41 (std. errors: (1.946), (1.950))
  - β5: -0.01** (std. error: (0.006))
- Table 6. Advanced Economies (Observations = 2,383; Adjusted R-squared across specifications: 0.192, 0.194, 0.194, 0.198, 0.207)
  - β1: -0.32; -0.29; -0.31; -0.46*; -0.62*** (std. errors: (0.225), (0.225), (0.223), (0.239), (0.239))
  - β2: -5.41**; -5.51**; -6.64**; -6.47** (std. errors: (2.467), (2.448), (2.599), (2.640))
  - β3: 5.70*; 4.21; 3.31 (std. errors: (3.385), (3.628), (3.611))
  - β4: -5.04**; -5.34** (std. errors: (2.199), (2.202))
  - β5: -0.07** (std. error: (0.028))

### Nonlinear effects and interaction results (Tables 7–8)
- Table 7. Nonlinear Effects of Leverage (Observations and Adjusted R-squared)
  - Full Sample: Observations 6,581; Adjusted R-squared 0.235
    - β1: -0.09; β2: 6.54* (std. error (3.410)); β3: 7.43*** (std. error (2.722)); β4: -3.16** (std. error (1.449)); β5: -0.02*** (std. error (0.007)); η: -8.17*** (std. error (1.908)); γ: 0.48**; Observations 6,581.
  - Emerging Economies: Observations 4,198; Adjusted R-squared 0.183
    - β1: 0.20; β2: 9.98** (std. error (4.158)); β3: 10.05*** (std. error (3.444)); β4: -1.23 (std. error (1.924)); β5: -0.01** (std. error (0.006)); η: -10.69*** (std. error (2.449)); γ: 0.48**; Observations 4,198.
  - Advanced Economies: Observations 2,383; Adjusted R-squared 0.211
    - β1: -0.57**; β2: 13.08* (std. error (6.792)); β3: 3.62 (std. error (3.556)); β4: -5.42** (std. error (2.196)); β5: -0.07** (std. error (0.028)); η: -13.64*** (std. error (4.601)); γ: 0.37*; Observations 2,383.
  - Robust standard errors in parentheses. Significance: *** p<0.01, ** p<0.05, * p<0.1.
- Table 8. Depreciation and Trade Effects (Full Sample and subsamples; Observations and Adjusted R-squared reported)
  - Full Sample (Observations 6,581; Adjusted R-squared 0.243)
    - β1: -0.09; β2: 6.71** (std. error (3.410)); β3: 7.08*** (std. error (2.721)); β4: -3.16** (std. error (1.449)); β5: -0.02*** (std. error (0.007)); η: -8.21*** (std. error (2.078)); φ: 0.49*** (std. error (0.114)); Ѳ: 0.00 (std. error (0.050)); γ: 0.48**.
  - Emerging Economies (Observations 4,198; Adjusted R-squared 0.185)
    - β1: 0.20; β2: 9.86** (std. error (4.158)); β3: 9.79*** (std. error (3.449)); β4: -1.23 (std. error (1.924)); β5: -0.01** (std. error (0.006)); η: -10.26*** (std. error (2.523)); φ: 0.30** (std. error (0.141)); Ѳ: 0.05 (std. error (0.087)); γ: 0.48**.
  - Advanced Economies (Observations 2,383; Adjusted R-squared 0.235)
    - β1: -0.57**; β2: -0.51 (std. error (6.792)); β3: 2.41 (std. error (3.662)); β4: -5.42** (std. error (2.196)); β5: -0.07** (std. error (0.028)); η: -3.58 (std. error (3.530)); φ: 0.82*** (std. error (0.103)); Ѳ: -0.02 (std. error (0.047)); γ: 0.37*.

### Figures and distributions (high-level descriptors with exact labels)
- Figure 1. Density Distribution of Corporate Performance: 2007, 2009, and 2010 — panels for All Countries, Emerging Economies, Advanced Economies; kernel = epanechnikov with bandwidths 2.6571, 3.2454, and 2.5894 respectively.
- Figure 2. Density Distribution of Corporate Performance by Levels of Leverage: 2010 — panels for All Countries, Emerging Economies, Advanced Economies. High and low leverage refer to levels above and below the median leverage for each sample; kernel = epanechnikov with bandwidths 3.3710, 3.9805, and 3.4490 respectively.
- Figure 3. Corporate Performance: 2007–2010 — median percentage growth in real sales (solid lines, left scale) and percentage of firms with negative growth (bars, right scale) for All Countries, Emerging Economies, Advanced Economies.

### Appendix: Data sources and variable definitions (exact)
- Principal data source: Worldscope (Thomson Financial).
- Inflation and REER data: IMF’s International Financial Statistics.
- Deflation: Variables in real terms are deflated using the CPI of the country where the firm is located.
- Variable definitions:
  - Sales: gross sales and other operating revenue less discounts, returns, and allowances (net sales or revenues, WS 01001).
  - Leverage: ratio of all short- and long-term obligations (total liabilities, WS 03351) to total assets (WS 02999).
  - Tangibility: ratio of fixed assets to total assets, with fixed assets defined as gross property, plant, and equipment less accumulated reserves for depreciation, depletion, and amortization (WS 02501).
  - Short-term debt: ratio of short-term debt and the portion of long-term debt payable within one year (WS 03051) to total debt (WS 03255).

*Source: _wp12292 - References (PDF).*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp12292.pdf_