## wpiea2021222-print-pdf

## Source details

**Canonical URL:** [wpiea2021222-print-pdf](https://www.imf.org/-/media/files/publications/wp/2021/english/wpiea2021222-print-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2021/english/wpiea2021222-print-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2021/english/wpiea2021222-print-pdf.pdf.json)

---

### Introduction and context
- Sample and period:
  - 29 economies: 19 AEs and 10 EMs.
  - Sample period: 1996:Q1 to 2020:Q4.
- Pre-COVID leverage:
  - Global nonfinancial sector debt increased from 138 percent of GDP to 152 percent of GDP over the decade leading up to the end of 2019.
  - Nonfinancial corporate sector debt reached 91 percent of GDP at the end of 2019.
  - Household debt reached 60 percent of GDP worldwide at the end of 2019.
- Pandemic-period changes:
  - Global nonfinancial corporate and household debt increased by 8 and 4 percentage points of GDP, respectively, between the end of 2019 and the fourth quarter of 2020.
- Definitions and measures:
  - Leverage measured as debt-to-GDP ratio (IIF data for debt; GDP scaling factor).
  - Financial conditions measured by IMF Financial Conditions Index (FCI); a decrease in FCI = loosening (lower price of risk).
  - Macroprudential measures drawn from IMF iMaPP database covering 17 policies in six categories.
- Key empirical questions:
  - Role of financial conditions in leverage buildups.
  - Macro-financial implications of high and rising leverage.
  - Role of macroprudential policies in arresting impacts of financial conditions and leverage.

### Drivers of leverage buildups (financial conditions evidence)
- Baseline empirical findings:
  - An easing in financial conditions by one unit is followed by an increase in nonfinancial corporate debt by 4 percentage points of GDP over three years (LS estimate).
  - IV estimate: increase in nonfinancial corporate debt of 3 percentage points of GDP over three years.
  - A one-unit easing of financial conditions implies an increase in household leverage by about 1½ percentage points of GDP over a three-year horizon (LS and IV).
- Controls and dynamics:
  - Output growth effect: increased income flows reduce need for external financing (significantly negative up to one year ahead for firms, up to two years for households).
  - Inflation: mixed; becomes positive and significant after one year for households (especially in AEs).
  - Policy rate hikes: associated with an initial increase in corporate leverage; reduce household leverage.
- Nonlinearity / credit boom amplification:
  - Credit boom definition: eight-quarter change in private nonfinancial sector debt-to-GDP in top three deciles AND country-specific FCI below median.
  - Under credit boom regime:
    - Easing financial conditions by one unit → increase in nonfinancial corporate debt by 10 (12) percentage points of GDP over three years in LS (IV) estimations.
  - No credit boom:
    - Easing by one unit → about 5 percent over three years (LS and IV).

### Macro-financial stability implications (Growth-at-Risk and tail risks)
- Intertemporal tradeoff from a one-standard-deviation easing in FCI:
  - Generates a stronger short-term boost but greater medium-term downside at the 10th percentile (GaR) than at the median.
  - 10 quarters after a one-standard-deviation increase in FCI, the adverse impact on GaR is ½ percentage point (more than three times that of the median).
- Leverage impacts on downside risk:
  - Nonfinancial corporate leverage: adverse impact on the 10th percentile peaks at about 1 percentage point after four quarters.
  - Household leverage: adverse impact takes longer to materialize, about 9 quarters.
  - For both sectors, impacts are larger at the 10th percentile than at the median.
- AE vs EM differences:
  - One-standard-deviation easing in FCI leads to significant negative impact in EMs after about 4 quarters; AEs show subdued left-tail impact in first 6 quarters, then significant negative impact.
  - Disruptive left-tail impact of increased household leverage is significant and persistent only in AEs.
  - Increases in nonfinancial corporate leverage have an impact on downside risk mainly in EMs.
- Regime amplifications:
  - Credit booms amplify the intertemporal tradeoff.
  - Positive output gap amplifies the tradeoff; tradeoff largely disappears when output gap is negative.
  - Combined regimes: credit boom with positive output gap shows oscillation in downside risk between +4 and -4 percentage points over different horizons; credit boom with negative output gap shows much milder tradeoff.
- Additional amplifiers:
  - High debt-service ratio regime (debt-service-to-income above country median and FCI looser than median) amplifies the tradeoff.
  - High FX share of debt (above country median and FCI looser than median) in predominantly EM subsample worsens medium-term downside risks for both corporate and household FX exposure regimes.

### Macroprudential policy: effects on leverage and downside risk
- Observed macroprudential actions:
  - Net tightening trend post-global financial crisis; close to 300 net tightening actions during 2016-18 across sample.
  - Bank capital and liquidity-related tightenings most prevalent.
  - FX-related measures played a larger role in EMs than in AEs.
- Effectiveness on leverage buildups:
  - Borrower-based measures (LTV, DSTI):
    - Tightening followed by reduction in household debt-to-GDP by up to 1 percentage point over a two-year horizon.
    - Borrower-based measures are the only tools with statistically significant decline in household leverage over two- and three-year horizons.
  - Banks’ liquidity requirements:
    - Net tightening associated with reduction in corporate leverage by up to 1 percentage point of GDP over a two-year horizon.
    - On average across sample, banks’ liquidity requirements deliver most consistent results to reduce corporate debt-to-GDP ratio.
  - FX-related measures in EMs:
    - Net tightening of FX-related measures for banks followed by a decline in nonfinancial corporate leverage by about 2½ percentage points over three years.
  - Caveats:
    - Estimates use net counts of tightening vs loosening actions (intensity of measures not quantified).
    - Reverse causality and leakages to NBFIs/cross-border funding could bias estimates toward zero or positive association; true impact likely stronger.
- Macroprudential mitigation of intertemporal tradeoff:
  - A macroprudential tightening measure is associated with a reduction in downside risk to future growth by about half a percentage point in the medium run.
  - When loosening financial conditions coincides with macroprudential tightening, the medium-term deterioration in downside risk is almost entirely mitigated (particularly after 9 quarters).
- Policy roles of macroprudential tools:
  - Leaning against the wind: tighten targeted measures to temper or reverse leverage buildups (household or firm-focused).
  - Mitigating intertemporal tradeoff: overall tightening reduces medium-term downside risk and counteracts risk from loose financial conditions when leverage is rising.
  - Importance of addressing leakages and expanding coverage to NBFIs as intermediation shifts away from banks.

### System evidence and robustness
- PVAR results corroborate main findings:
  - One-standard-deviation loosening in financial conditions leads to:
    - Increase in household leverage of about 0.2 percentage point after 4 years.
    - Increase in nonfinancial corporate leverage of about 0.4 percentage point after 4 years.
  - Loosening in financial conditions leads to an increase in real GDP by about 0.2 to 0.3 percent in the 4 years after the shock (boost frontloaded in first year).
  - Shock to household leverage leads to a fall in real GDP of about 0.2 percent after 4 years.
  - Shock to nonfinancial corporate leverage reduces GDP by about 0.1 percent within first 4 years.
- Robustness and alternative measures:
  - Results robust to different estimation methods (local projections, PVAR), specifications, and variable definitions.
  - Alternative leverage measures used in robustness (debt-service-to-income, foreign currency share) show consistent amplifications.
  - Domestic financial conditions generally more informative on future leverage than international conditions.

### Conclusions and policy implications
- Three major empirical results:
  - Loosening financial conditions is associated with stronger buildup in nonfinancial corporate and household leverage, particularly over the medium term.
  - Looser financial conditions lead to an intertemporal tradeoff: reduced downside risk near-term but accentuated downside risk in the medium term; tradeoff amplified during credit booms, positive output gaps, high debt-service, and high FX debt shares.
  - Selected macroprudential policies can slow leverage buildups (borrower-based tools for households; liquidity and FX-related measures for corporate leverage in EMs) and mitigate the intertemporal tradeoff, though with implementation lags.
- Policy recommendations and implications:
  - Near-term accommodative policy may be appropriate where recovery fragile, but continued extraordinary easing after recovery risks adding to elevated leverage vulnerabilities and moral hazard expectations.
  - Swift implementation of macroprudential tightening depending on country-specific circumstances (pace of recovery, postcrisis vulnerabilities, available tools) will be needed as soon as macroeconomic conditions permit.
  - Policymakers should limit leakages and expand the macroprudential perimeter, particularly to cover NBFIs, as intermediation migrates away from banks.
  - Use of targeted borrower-based measures and bank liquidity/FX exposure measures can be effective complements to monetary policy to contain leverage buildups and reduce medium-term downside risk.

*Italic source: Extracted from wpiea2021222-print-pdf (29-economy analysis, 1996:Q1–2020:Q4) contained in the supplied PDF content.*

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

### wpiea2021222-print-pdf - References  ............................................................................................................................

### Introduction and context
- Sample and period:
  - 29 economies: 19 AEs and 10 EMs.
  - Sample period: 1996:Q1 to 2020:Q4.
- Pre-COVID leverage:
  - Global nonfinancial sector debt increased from 138 percent of GDP to 152 percent of GDP over the decade leading up to the end of 2019.
  - Nonfinancial corporate sector debt reached 91 percent of GDP at the end of 2019.
  - Household debt reached 60 percent of GDP worldwide at the end of 2019.
- Pandemic-period changes:
  - Global nonfinancial corporate and household debt increased by 8 and 4 percentage points of GDP, respectively, between the end of 2019 and the fourth quarter of 2020.
- Definitions and measures:
  - Leverage measured as debt-to-GDP ratio (IIF data for debt; GDP scaling factor).
  - Financial conditions measured by IMF Financial Conditions Index (FCI); a decrease in FCI = loosening (lower price of risk).
  - Macroprudential measures drawn from IMF iMaPP database covering 17 policies in six categories.
- Key empirical questions:
  - Role of financial conditions in leverage buildups.
  - Macro-financial implications of high and rising leverage.
  - Role of macroprudential policies in arresting impacts of financial conditions and leverage.

### Drivers of leverage buildups (financial conditions evidence)
- Baseline empirical finding:
  - An easing in financial conditions by one unit is followed by an increase in nonfinancial corporate debt by 4 percentage points of GDP over three years (LS estimate).
  - IV estimate: increase in nonfinancial corporate debt of 3 percentage points of GDP over three years.
  - A one-unit easing of financial conditions implies an increase in household leverage by about 1½ percentage points of GDP over a three-year horizon (LS and IV).
- Controls and dynamics:
  - Output growth effect: increased income flows reduce need for external financing (significantly negative up to one year ahead for firms, up to two years for households).
  - Inflation: mixed; becomes positive and significant after one year for households (especially in AEs).
  - Policy rate hikes: associated with an initial increase in corporate leverage; reduce household leverage.
- Nonlinearity / credit boom amplification:
  - Definition: credit boom = eight-quarter change in private nonfinancial sector debt-to-GDP in top three deciles AND country-specific FCI below median.
  - Under credit boom regime:
    - Easing financial conditions by one unit → increase in nonfinancial corporate debt by 10 (12) percentage points of GDP over three years in LS (IV) estimations.
  - No credit boom: easing by one unit → about 5 percent over three years (LS and IV).

### Macro-financial stability implications (Growth-at-Risk and tail risks)
- Intertemporal tradeoff from a one-standard-deviation easing in FCI (loosening financial conditions):
  - Generates a stronger short-term boost but greater medium-term downside at the 10th percentile (GaR) than at the median.
  - 10 quarters after a one-standard-deviation increase in FCI, the adverse impact on GaR is ½ percentage point (more than three times that of the median).
- Leverage impacts on downside risk:
  - Nonfinancial corporate leverage: adverse impact on the 10th percentile peaks at about 1 percentage point after four quarters.
  - Household leverage: adverse impact takes longer to materialize, about 9 quarters.
  - For both sectors, impacts are larger at the 10th percentile than at the median.
- AE vs EM differences:
  - One-standard-deviation easing in FCI leads to significant negative impact in EMs after about 4 quarters; AEs show subdued left-tail impact in first 6 quarters, then significant negative impact.
  - Disruptive left-tail impact of increased household leverage is significant and persistent only in AEs.
  - Increases in nonfinancial corporate leverage have an impact on downside risk mainly in EMs.
- Regime amplifications:
  - Credit booms amplify the intertemporal tradeoff.
  - Positive output gap amplifies the tradeoff; tradeoff largely disappears when output gap is negative.
  - Combined regimes: credit boom with positive output gap shows oscillation in downside risk between +4 and -4 percentage points over different horizons; credit boom with negative output gap shows much milder tradeoff.
- Additional amplifiers:
  - High debt-service ratio regime (debt-service-to-income above country median and FCI looser than median) amplifies the tradeoff.
  - High FX share of debt (above country median and FCI looser than median) in predominantly EM subsample worsens medium-term downside risks for both corporate and household FX exposure regimes.

### Macroprudential policy: effects on leverage and downside risk
- Macroprudential actions observed:
  - Net tightening trend post-global financial crisis; close to 300 net tightening actions during 2016-18 across sample.
  - Bank capital and liquidity-related tightenings most prevalent.
  - FX-related measures played a larger role in EMs than in AEs.
- Effectiveness on leverage buildups (regression specification analogous to equation (3)):
  - Borrower-based measures (LTV, DSTI):
    - Tightening followed by reduction in household debt-to-GDP by up to 1 percentage point over a two-year horizon.
    - Borrower-based measures are the only tools with statistically significant decline in household leverage over two- and three-year horizons.
  - Banks’ liquidity requirements:
    - Net tightening associated with reduction in corporate leverage by up to 1 percentage point of GDP over a two-year horizon.
    - On average across sample, banks’ liquidity requirements deliver most consistent results to reduce corporate debt-to-GDP ratio.
  - FX-related measures in EMs:
    - Net tightening of FX-related measures for banks followed by a decline in nonfinancial corporate leverage by about 2½ percentage points over three years.
  - Caveats:
    - Estimates use net counts of tightening vs loosening actions (intensity of measures not quantified).
    - Reverse causality and leakages to NBFIs/cross-border funding could bias estimates toward zero or positive association; true impact likely stronger.
- Macroprudential mitigation of intertemporal tradeoff (GaR framework, equations (4a)-(4b)):
  - A macroprudential tightening measure is associated with a reduction in downside risk to future growth by about half a percentage point in the medium run.
  - When loosening financial conditions coincides with macroprudential tightening, the medium-term deterioration in downside risk is almost entirely mitigated (particularly after 9 quarters).
- Policy roles of macroprudential tools:
  - Leaning against the wind: tighten targeted measures to temper or reverse leverage buildups (household or firm-focused).
  - Mitigating intertemporal tradeoff: overall tightening reduces medium-term downside risk and counteracts risk from loose financial conditions when leverage is rising.
  - Importance of addressing leakages and expanding coverage to NBFIs as intermediation shifts away from banks.

### System evidence and robustness
- PVAR (system) results corroborate main findings:
  - One-standard-deviation loosening in financial conditions leads to:
    - Increase in household leverage of about 0.2 percentage point after 4 years.
    - Increase in nonfinancial corporate leverage of about 0.4 percentage point after 4 years.
  - Loosening in financial conditions leads to an increase in real GDP by about 0.2 to 0.3 percent in the 4 years after the shock (boost frontloaded in first year).
  - Shock to household leverage leads to a fall in real GDP of about 0.2 percent after 4 years.
  - Shock to nonfinancial corporate leverage reduces GDP by about 0.1 percent within first 4 years.
- Robustness and alternative measures:
  - Results robust to different estimation methods (local projections, PVAR), specifications, and variable definitions.
  - Alternative leverage measures used in robustness (debt-service-to-income, foreign currency share) show consistent amplifications.
  - Domestic financial conditions generally more informative on future leverage than international conditions.

### Conclusions and policy implications
- Three major empirical results:
  - Loosening financial conditions is associated with stronger buildup in nonfinancial corporate and household leverage, particularly over the medium term.
  - Looser financial conditions lead to an intertemporal tradeoff: reduced downside risk near-term but accentuated downside risk in the medium term; tradeoff amplified during credit booms, positive output gaps, high debt-service, and high FX debt shares.
  - Selected macroprudential policies can slow leverage buildups (borrower-based tools for households; liquidity and FX-related measures for corporate leverage in EMs) and mitigate the intertemporal tradeoff, though with implementation lags.
- Policy recommendations and implications:
  - Near-term accommodative policy may be appropriate where recovery fragile, but continued extraordinary easing after recovery risks adding to elevated leverage vulnerabilities and moral hazard expectations.
  - Swift implementation of macroprudential tightening depending on country-specific circumstances (pace of recovery, postcrisis vulnerabilities, available tools) will be needed as soon as macroeconomic conditions permit.
  - Policymakers should limit leakages and expand the macroprudential perimeter, particularly to cover NBFIs, as intermediation migrates away from banks.
  - Use of targeted borrower-based measures and bank liquidity/FX exposure measures can be effective complements to monetary policy to contain leverage buildups and reduce medium-term downside risk.

*Italic source: Extracted from wpiea2021222-print-pdf (29-economy analysis, 1996:Q1–2020:Q4) contained in the supplied PDF content.*

### REFERENCES

### REFERENCES

### Cited works
- Adrian, Tobias, Federico Grinberg, Nellie Liang, and Sheheryar Malik, forthcoming, “The Term Structure of Growth-at-Risk,” American Economic Journal: Macroeconomics, forthcoming.  
- Adrian, Tobias, and Nellie Liang, 2018, “Monetary Policy, Financial Conditions, and Financial Stability,” International Journal of Central Banking, Vol. 14 (January), pp. 73-131.  
- Adrian, Tobias, Nina Boyarchenko, and Domenico Giannone, 2019a, “Vulnerable Growth,” American Economic Review, Vol. 109 (April), pp. 1263-89.  
- Adrian, Tobias, Dong He, Nellie Liang, and Fabio Natalucci, 2019b, “A Monitoring Framework for Global Financial Stability,” IMF Staff Discussion Note 19/06 (Washington, D.C.: International Monetary Fund).  
- Aiyar, Shekhar, Charles W Calomiris, and Tomasz Wieladek, 2014, “Does Macro-prudential Regulation Leak? Evidence from a UK Policy Experiment,” Journal of Money, Credit and Banking, Vol. 46 (January), pp. 181–214.  
- Alam, Zohair, Adrian Alter, Jesse Eiseman, Gaston Gelos, Heedon Kang, Machiko Narita, Erlend Nier, and Naixi Wang, 2019, “Digging Deeper—Evidence on the Effects of Macroprudential Policies from a New Database.” IMF Working Paper 19/66 (Washington D.C.: International Monetary Fund).  
- Alter, Adrian, and Selim Ali Elekdag, 2020, “Emerging Market Corporate Leverage and Global Financial Conditions,” Journal of Corporate Finance, Vol. 62 (June), pp. 1-18.  
- Araujo, Juliana, Manasa Patnam, Adina Popescu, Fabian Valencia, and Weijia Yao, 2020, “Effects of Macroprudential Policy: Evidence from Over 6,000 Estimates,” IMF Working Paper 20/67 (Washington D.C.: International Monetary Fund).  
- Auerbach, Alan and Yuriy Gorodnichenko, 2012, “Measuring the Output Responses to Fiscal Policy,” American Economic Journal: Economic Policy, Vol. 4 (May), pp. 1-27.  
- Bank for International Settlements, 2014, “Debt and the Financial Cycle: Domestic and Global,” Chapter IV in the BIS Annual Report (Basel: Bank for International Settlements).  
- Bernanke, Ben S., and Alan S. Blinder, 1992, “The Federal Funds Rate and the Channels of Monetary Transmission,” American Economic Review, Vol. 82 (September), pp. 901-21.  
- Bernanke, Ben S., Mark Gertler, and Simon Gilchrist,1999, “The Financial Accelerator in a Quantitative Business Cycle Framework,” Chapter 21 in John B. Taylor and Michael Woodford (eds), Handbook of Macroeconomics (Amsterdam: Elsevier Science B.V.).  
- Brandao-Marques, Luis, Qianying Chen, Claudio Raddatz, Jérôme Vandenbussche, and Peichu Xie, 2019, “The Riskiness of Credit Allocation and Financial Stability,” IMF Working Paper 19/207 (Washington, D.C.: International Monetary Fund).  
- Brandao-Marques, Luis, Gaston Gelos, Machiko Narita, and Erlend Nier, 2020, “Leaning Against the Wind: A Cost-Benefit Analysis for an Integrated Policy Framework,” IMF Working Paper 20/123 (Washington D.C.: International Monetary Fund).  
- Brunnermeier, Markus K., and Yuliy Sannikov, 2014, “A Macroeconomic Model with a Financial Sector,” American Economic Review, Vol. 104 (February), pp. 379-421.  
- Caballero, Ricardo J., and Arvind Krishnamurthy, 2003, “Excessive Dollar Debt: Financial Development and Underinsurance,” Journal of Finance, Vol. 58 (March), pp. 867-93.  
- Calvo, Guillermo A., Alejandro Izquierdo, and Luis-Fernando Mejia, 2008, “Systemic Sudden Stops: The Relevance of Balance-Sheet Effects and Financial Integration,” Working Paper 14026 (National Bureau of Economic Research).  
- Cerutti, Eugenio, Stijn Claessens, and Luc Laeven, 2017, “The Use and Effectiveness of Macroprudential Policies: New Evidence,” Journal of Financial Stability, Vol. 28 (February), pp. 203–24.  
- Christiano, Lawrence J., Martin Eichenbaum, and Charles Evans, 1996, “The Effects of Monetary Policy Shocks: Evidence from the Flow of Funds,” Review of Economics and Statistics, Vol. 78 (February), pp. 16-34.  
- Christiano, Lawrence J., Mathias Trabandt and Karl Walentin, 2011, “DSGE Models for Monetary Policy Analysis,” in Benjamin M. Friedman, and Michael Woodford (eds), Handbook of Monetary Economics, Vol. 3A (Amsterdam: Elsevier Science B.V.).  
- Claessens, Stijn, Giulio Cornelli, Leonardo Gambacorta, Francesco Manaresi, and Yasushi Shiina, 2021, “Do Macroprudential Policies Affect Nonbank Financial Intermediation?” BIS Working Paper 927 (Basel: Bank for International Settlements).  
- Drehmann, Mathias, and Mikael Juselius, 2014, “Evaluating Early Warning Indicators of Banking Crises: Satisfying Policy Requirements,” International Journal of Forecasting, Vol. 30 (July-September), pp. 759-80.  
- Driscoll, John C., and Aart C. Kraay, 1998, “Consistent Covariance Matrix Estimation with Spatially Dependent Panel Data,” Review of Economics and Statistics, Vol. 80 (November), pp. 549-60.  
- Du, Wenxin, and Jesse Schreger, 2016, “Local Currency Sovereign Risk,” Journal of Finance, Vol. 71 (February), pp. 1027-70.  
- Duprey, Thibault, and Alexander Ueberfeldt, 2020, “Managing GDP Tail Risk,” Bank of Canada Staff Working Paper 2020-03 (Ottawa: Bank of Canada).  
- Eberhardt, Markus, and Andrea Presbitero, 2018, “Commodity Price Movements and Banking Crises.” IMF Working Paper 18/153 (Washington D.C.: International Monetary Fund).  
- ESRB, 2014, “Flagship Report on Macro-prudential Policy in the Banking Sector,” European System of Financial Supervision, European Systemic Risk Board.  
- Forbes, Kirstin, 2020, “The International Aspects of Macroprudential Policies,” MIT Sloan School Working Paper 6167-20.  
- Furceri, Davide, and Ricardo M. Sousa, 2011, “Does Government Spending Crowd Out Private Consumption and Investment?” World Economics, Vol. 12 (October), pp. 153-170.  
- Gertler, Mark, and Simon Gilchrist, 2018, “What Happened: Financial Factors in the Great Recession,” Journal of Economic Perspectives, Vol. 32 (Summer), pp. 3-30.  
- Gilchrist, Simon, David López-Salido, and Egon Zakrajšek, 2015, “Monetary Policy and Real Borrowing Costs at the Zero Lower Bound,” American Economic Journal: Macroeconomics, Vol. 7 (January), pp. 77-109.  
- Gilchrist, Simon, Michael Siemer, and Egon Zakrajsek, 2018, “The Real Effects of Credit Booms and Busts: A County-Level Analysis,” Meeting Papers 99, Society for Economic Dynamics.  
- Gilchrist, Simon, and Egon Zakrajšek, 2012, “Credit Spreads and Business Cycle Fluctuations,” American Economic Review, Vol. 102 (June), pp. 1692-720.  
- Gross, Marco, Thierry Tressel, Xiaodan Ding, and Eugen Tereanu, forthcoming, “Structural Risk Drivers of Household Default in a Micro-Macro Simulation Model for 21 EU Countries and the US,” IMF Working Papers forthcoming (Washington D.C.: International Monetary Fund).  
- Hanson, Samuel G., Jeremy C. Stein, Adi Sunderman, and Eric Zwick, 2020, “Business Credit Programs in the Pandemic Era,” BPEA Article, September (Washington D.C.: Brookings Institution).  
- Institute of International Finance, 2020, “COVID-19 Exacerbates Household Debt Burdens,” IIF Weekly Insight, March 26.  
- International Monetary Fund, 2014, “Staff Guidance Note on Macroprudential Policy” (Washington D.C.: International Monetary Fund).  
- International Monetary Fund, 2017a, “Household Debt and Financial Stability,” Global Financial Stability Report, (October), Chapter 2, pp. 53-89 (Washington D.C.: International Monetary Fund).  
- International Monetary Fund, 2017b, “Is Growth At Risk?” Global Financial Stability Report, (October), Chapter 1, pp. 1-52 (Washington D.C.: International Monetary Fund).  
- International Monetary Fund, 2018a, “Online Annex 1.1 Technical Note,” Global Financial Stability Report, (April), Chapter 1, pp. 57-92 (Washington D.C.: International Monetary Fund).  
- International Monetary Fund, 2018b, “The Riskiness of Credit Allocation: A Source of Financial Vulnerability?” Global Financial Stability Report, (April), Chapter 2, pp. 57-92 (Washington D.C.: International Monetary Fund).  
- International Monetary Fund, 2020a, “Risky Credit Markets,” Global Financial Stability Report,” (April), Chapter 2, pp. 29-46.  
- International Monetary Fund, 2020b, “Liquidity Strains Cushioned by a Powerful Set of Policies,” Global Financial Stability Report, (October), Chapter 3, pp. 53-69.  
- International Monetary Fund, 2020c, “Fiscal Policies to Address the COVID-19 Pandemic,” Fiscal Monitor, (October), Chapter 1, pp. 1-29.  
- International Monetary Fund, 2021a, “Tailoring Fiscal Responses,” Fiscal Monitor (April), Chapter 1, pp. 1-26.  
- International Monetary Fund, 2021b, “An Asynchronous and Divergent Recovery May Put Financial Stability at Risk, Global Financial Stability Report, (April), Chapter 1, pp. 1-33.  
- International Monetary Fund, 2021c, “Global Prospects and Policies,” World Economic Outlook, (April), pp. 1-41.  
- Jordà, Òscar, 2005, “Estimation and Inference of Impulse Responses by Local Projections,” American Economic Review, Vol. 95 (March), pp. 161-82.  
- Jordà, Òscar, Martin Kornejew, Moritz Schularick, and Alan M. Taylor, 2020, “Zombies at Large? Corporate Debt Overhang and the Macroeconomy,” NBER Working Paper 28197 (National Bureau of Economic Research).  
- Jordà, Òscar, Moritz Schularick, and Alan M. Taylor, 2016, “Macrofinancial History and The New Business Cycle Facts.” NBER Macroeconomic Annuals, Vol. 31, pp. 213-63.  
- Kalemli-Ozcan, Sebnem, Luc Laeven, and David Moreno, 2020, “Debt Overhang, Rollover Risk, and Corporate Investment: Evidence from the European Crisis,” NBER Working Paper 24555 (National Bureau of Economic Research).  
- Kirti, Divya, 2018, “Lending Standards and Output Growth,” IMF Working Paper 18/023 (Washington D.C.: International Monetary Fund).  
- Kiyotaki, Nobuhiro, and John Moore, 1997, “Credit Cycles,” Journal of Political Economy, Vol. 105 (April), pp. 211-48.  
- Krugman, Paul, 1999, “Balance Sheets, the Transfer Problem, and Financial Crises,” International Tax and Public Finance, Vol. 6 (November), pp. 459-72.  
- Machado, José A.F., and J.M.C. Santos Silva, 2019, “Quantiles Via Moments,” Journal of Econometrics, Vol. 213 (November), pp. 145-73.  
- Mian, Atif, Amir Sufi, and Emil Verner, 2017, “Household Debt and Business Cycles Worldwide,” Quarterly Journal of Economics, Vol. 132 (November), pp. 1755-817.  
- Opler, Tim C., Lee Pinkowitz, Rene Stultz, and Rohan Williamson, 1999, “The Determinants and Implications of Corporate Cash Holdings,” Journal of Financial Economics, Vol. 52 (April), pp. 3-46.  
- Ostry, Jonathan, Atish Ghosh, Marcos Chamon, and Mahvash Qureshi, 2012, “Tools for Managing Financial-Stability Risks from Capital Inflows,” Journal of International Economics, Vol. 88 (November), pp. 407–21.  
- Peydró, José-Luis, Francesc Rodriguez-Tous, Jagdish Tripathy, and Arzu Uluc, 2020, “Macroprudential Policy, Mortgage Cycles and Distributional Effects: Evidence from the UK,” Staff Working Paper 866 (London: Bank of England).  
- Rancière, Romain, Aaron Tornell, and Athanasios Vamvakidis, 2010, “Currency Mismatch, Systemic Risk and Growth in Emerging Europe,” Economic Policy, Vol. 25 (October), pp. 597-658.  
- Rancière, Romain, and Aaron Tornell, 2016, “Financial Liberalization, Debt Mismatch, Allocative Efficiency, and Growth,” American Economic Journal: Macroeconomics, Vol. 8 (April), pp.1-44.  
- Schularick, Moritz, and Alan M. Taylor, 2012, “Credit Booms Gone Bust: Monetary Policy, Leverage Cycles, and Financial Crises, 1870-2008,” American Economic Review, Vol. 102 (April), pp. 1029-61.  
- Verner, Emil, 2019, “Private Debt Booms and the Real Economy: Do the Benefits Outweigh the Costs?” Working Paper, MIT, Sloan School of Management.  
- Verner, Emil, and Győző Gyöngyösi, 2020, “Household Debt Revaluation and the Real Economy: Evidence from a Foreign Currency Debt Crisis,” American Economic Review, Vol. 110 (September), pp. 2667-702.

### Figures and captions (selected)
- Figure 2. Leverage as an Amplifier of Shocks  
  - Source: IMF staff, based on Adrian and others 2019b.  
  - Note: “Nonlinearities” indicates that the impact of financial conditions on macro-financial stability may be amplified in the presence of elevated financial vulnerabilities, such as a high level of leverage.  

- FIGURES — Figure 1. Private nonfinancial Sector Leverage, by Country Group  
  - Panels and measures:  
    - 1. Nonfinancial Corporate Leverage: Debt-to-GDP Ratio, 2001:Q1–20:Q4 (Percent)  
    - 2. Household Leverage: Debt-to-GDP Ratio, 2001:Q1–20:Q4 (Percent)  
    - 3. Nonfinancial Corporations: Decomposition of the Debt-to-GDP Ratio Increase from 2019:Q4 to 2020:Q4 (Percentage points)  
    - 4. Household Sector: Decomposition of the Debt-to-GDP Ratio Increase from 2019:Q4 to 2020:Q4 (Percentage points)  
  - Sources: Institute of International Finance; and authors’ calculations.  
  - Notes: The figure includes 27 advanced economies (AEs) and 25 emerging markets (EMs). Leverage is measured as the ratio of debt to GDP. Global, AE, and EM leverage are measured as the ratio of aggregate debt to aggregate GDP across different country groups. Nonfinancial corporate debt figures are nonconsolidated.

- Figure 3. Leverage, Financial Conditions, and Output Growth  
  - Panels:  
    - 1. Nonfinancial Corporations: Within-Country Correlation between Financial Conditions and a Future Change in Leverage  
    - 2. Households: Within-Country Correlation between Financial Conditions and a Future Change in Leverage  
    - 3. Nonfinancial Corporations: Within-Country Correlation between a Change in Leverage and Future Output Growth  
    - 4. Households: Within-Country Correlation between a Change in Leverage and Future Output Growth  
  - Sources: Institute of International Finance; and authors’ calculations.  
  - Notes: The sample includes 19 advanced economies (AE) and 10 emerging markets (EMs) over the period 1996:Q1 – 2020:Q4. Panels 1 and 2 show the correlation of the negative of the Financial Conditions Index (FCI) (that is, higher values indicate looser financial conditions) and changes in the debt-to-GDP ratio over the subsequent 12 quarters. Panels 3 and 4 show the correlation of a change in the nonfinancial corporate sector and household debt-to-GDP ratio, respectively, over eight quarters and a change in output growth over the subsequent 12 quarters.

- Figure 4. Financial Conditions Easing and Nonfinancial Sector Leverage Buildups: All Countries  
  - Panels 1–6 present impacts on cumulative h-quarter-ahead changes in debt-to-GDP ratios over horizons of 1–12 quarters.  
  - Source: Authors’ calculations.  
  - Notes (selected):  
    - (a) Panel regressions with country-fixed effects are estimated for 1996:Q1-20:Q4 for Panels 1-2 (1997:Q1-20:Q4 for Panels 3-6).  
    - (b) The shock is scaled as a one-unit decrease in the Financial Conditions Index (FCI) to reflect the effect of financial condition easing.  
    - (c) Regressors in Panels 1 and 2 (regression 1a) comprise one-period-lagged variables of the core covariates (FCI, output growth, inflation, and policy rate change) as well as lagged one-quarter changes in the sector-specific debt-to-GDP ratio; Panels 3-6 (regression 1b) additionally include credit-boom and no-credit boom dummies interacting with these regressors.  
    - (d) IV regressions include instruments: two lags of commodity price inflation, policy rate change, 2-year government bond yield change, and 10-year government bond yield change purged by 2-year bond yield change in addition to the regressors in Panels 1 and 2; and these variables interacting with credit-boom and no-credit-boom dummies in Panels 3-4.  
    - (e) The credit-boom dummy equals one when, for a given country, the country-specific FCI is below the median of its distribution and the eight-quarter change in the credit-to-GDP ratio is in the top three deciles of its distribution.

- Figure 5. Impact of Financial Conditions, Household Leverage, and Nonfinancial Corporate Leverage on Future GDP Growth (10th percentile and median, percentage points)  
  - Panels:  
    - 1. FCI  
    - 2. Household Leverage  
    - 3. Nonfinancial Corporate Leverage  
  - Sources: International Institute of Finance; and authors’ calculations.  
  - Notes: The panels show the 10th percentile (red) and median (blue) impact of a one-standard-deviation easing in the FCI, one-percentage-point increase in household leverage, one-percentage-point increase in nonfinancial corporate leverage, respectively, on future year-on-year real GDP growth. Control variables include year-on-year real GDP growth, inflation, policy rate change (all one-period lagged), and country fixed effects. Dashed lines indicate 90 percent confidence intervals based on standard errors computed from bootstrapping.

- Figure 6. Impact of Financial Conditions, Household Leverage, and Nonfinancial Corporate Leverage on the 10th Percentile of Future GDP Growth: by Country Groups (AEs vs. EMs)  
  - Panels:  
    - 1. FCI  
    - 2. Household Leverage  
    - 3. Nonfinancial Corporate Leverage  
  - Sources: International Institute of Finance; and authors’ calculations.  
  - Notes: Panels show impact on the 10th percentile of the future growth distribution of a one-standard-deviation easing in the FCI, one-percentage-point increase in household leverage, one-percentage-point increase in nonfinancial corporate leverage, respectively, on future year-on-year real GDP growth in advanced economies (red) and emerging market economies (blue). Control variables include year-on-year real GDP growth, inflation, policy rate change (all one-period lagged), and country fixed effects. Dashed lines indicate 90 percent confidence intervals based on standard errors computed from bootstrapping.

- Figure 7. Amplification Effects of Credit Booms and Output Gaps: Impact of Loosening Financial Conditions on the 10th Percentile of Output Growth under Different Regimes  
  - Panels:  
    - 1. Credit Boom  
    - 2. Output Gap  
    - 3. Frequency of Output Gaps and Credit Booms  
    - 4. Concurring Effect of Credit Boom and Output Gap  
  - Sources: International Institute of Finance; Bank for International Settlements; and authors’ calculations.  
  - Notes: Panels show the 10th percentile impact of a one-standard-deviation easing in the FCI on future year-on-year real GDP growth under different regimes. The regime in Panel 1 is credit boom in which the 8-quarter change in private nonfinancial sector leverage is higher than the top 3rd decile and financial conditions are looser than median in a country. The regime in Panel 2 is defined as the episodes where the quarterly output gap is positive or negative. Panel 3 shows the frequency of credit boom regimes and the concurring credit boom and positive (or negative) output gap regimes across 29 sample countries during 1997-2020. Panel 4 shows the impacts on output growth under the regimes defined by interactions between output gaps and credit boom. Control variables include year-on-year real GDP growth, inflation, and policy rate change (all one-period lagged), and country fixed effects. Dashed lines indicate 90 percent confidence intervals based on standard errors computed from bootstrapping.

- Figure 8. Macroprudential Policy Actions, by Category

*Source: wpiea2021222-print-pdf - REFERENCES*

### 1. Composition of Macroprudential Policy

### 1. Composition of Macroprudential Policy

### Composition and Net Tightening (1990–2018; 2011–2018)
- Sample: 19 advanced economies (AEs) and 10 emerging markets (EMs).
- Panel 1 (1990–2018 five-year subperiods): shows number of times each category of macroprudential measures was tightened (positive) or loosened (negative). Solid line indicates net tightening (tightened minus loosened).
- Panel 2 (2011–18): reports net tightening over the 2011–18 period by category and by country group.
- Note: FX = foreign exchange; L = loosened; T = tightened.

### Key empirical takeaways
- The database used: IMF, Integrated Macroprudential Policy database (as referenced in figures and notes).
- Net tightening is presented across 17 types of macroprudential measures (used in later outcome analysis).

### Policy implication (composition)
- Macroprudential policy composition varies by country group and measure type; analyses below quantify the effects of tightening/loosening on sectoral leverage and downside growth risks.

---

### Macroprudential Tightening and Change in Leverage (Figure 9)
- Regression setup: association between a one-unit net tightening in a category of macroprudential measures and subsequent change in household or nonfinancial corporate debt-to-GDP ratios over horizons of 1–16 quarters.
- Controls: one-quarter change in sector-specific debt-to-GDP ratio, GDP growth, FCI, inflation, policy rate change (all one-period lagged), and a dummy for the global financial crisis.
- Confidence intervals: dashed lines = 90 percent confidence intervals.

Key results described:
- Panel 1: Response of change in household leverage to borrower-based measures (percentage points).
- Panel 2: Response of change in nonfinancial corporate leverage to liquidity-related measures (percentage points).
- Panel 3: Response of change in nonfinancial corporate leverage to foreign currency exposure measures in Emerging Market Economies (percentage points).

---

### Macroprudential Measures and Downside Risks to Future Growth (Figure 10)
- Dependent variable: year-over-year growth of real GDP over horizons of 1–16 quarters.
- Panel 1: effects of a net tightening event across the 17 types of macroprudential measures.
- Panel 2: effect of a one-unit loosening in the FCI; compares regimes with vs. without recent macroprudential tightening (the “macroprudential tightening” regime contains all quarters with net macroprudential tightening in the past year).
- Controls: change in private nonfinancial sector leverage over past eight quarters, year-over-year GDP growth, and inflation.
- Estimation: quantile local projections at the 10th percentile of the growth distributions with country fixed effects.
- Confidence intervals: dashed lines = 90 percent.

Key qualitative insight:
- Macroprudential tightening regimes are evaluated for their ability to mitigate downside risks to future GDP growth (10th percentile outcomes) following loosenings of financial conditions.

---

### PVAR-Estimated Responses to Loosening Financial Conditions and to Increasing Leverage (Figure 11)
- Panels 1–3: responses to a one-standard-deviation loosening in financial conditions.
  - Panel 1: Household leverage responses (percentage points, quarters).
  - Panel 2: Nonfinancial corporate leverage responses (percentage points, quarters).
  - Panel 3: Real GDP responses to loosening (percent, quarters).
- Panels 4–5: responses to a one-standard-deviation increase in private nonfinancial sector leverage.
  - Panel 4: Real GDP response to increase in household leverage (percent, quarters).
  - Panel 5: Real GDP response to increase in nonfinancial corporate leverage (percent, quarters).
- Bands: dotted lines show +/- two-standard-deviation-error bands.

Key qualitative insight:
- Surprise loosening in financial conditions raises private sector leverage (household and corporate) and has measurable dynamic effects on real GDP; increases in leverage themselves feed back to GDP.

---

### Amplification Effects of High Debt Service or FX Share of Debt (Figure 12)
- Outcome: impact on the 10th percentile of year-on-year real GDP growth of a one-standard-deviation easing in FCI under different regimes.
- Regime comparisons:
  - Panel 1: Above-median private sector debt-service ratio + Loose FCI versus Other.
  - Panel 2: Above-median nonfinancial corporate debt FX share + Loose FCI versus Other.
  - Panel 3: Above-median household debt FX share + Loose FCI versus Other.
- Controls: year-on-year real GDP growth, inflation, and policy rate change (all one-period lagged), and country fixed effects.
- Confidence intervals: dashed lines indicate 90 percent confidence intervals based on bootstrap standard errors.

Key qualitative insight:
- The negative impact of loosening financial conditions on downside growth (10th percentile) is amplified when private sector debt-service ratios are high or when household or corporate debt FX shares are above-median.

---

### Estimated Drivers of Future Changes in Leverage (Tables 1–3; selected numeric findings preserved)

- Sample and estimation:
  - Panel regressions with country-fixed effects estimated for 19 AEs and 10 EMs for 1996:Q1–2020:Q4.
  - Horizons reported: h = 2, 4, 6, 8, 10, 12 quarters ahead.
  - Regressors: one-period-lagged FCI, output growth, inflation, policy rate change, and lagged one-quarter change in sector-specific debt-to-GDP ratio. IV regressions include additional instruments for policy rate change.
  - Standard errors: Driscoll and Kraay (1998) or robust SEs as indicated. Significance: *, **, *** indicate 0.1, 0.05, 0.01 levels respectively.

- Table 1 (Nonfinancial Corporate Leverage; Panel 1 least squares):
  - FCI coefficients (horizons h=2 to h=12): 0.348; 1.104***; 1.989***; 3.006***; 3.677***; 3.840***.
  - Output Growth coefficients: -0.060**; -0.109**; -0.064; -0.044; -0.039; -0.018.
  - Inflation coefficients: -0.026***; -0.052***; -0.056***; -0.072***; -0.082***; -0.102***.
  - Policy Rate Change coefficients: 0.084***; 0.111***; 0.101**; 0.075; 0.050; 0.053.
  - Observations decline with horizon: 2,816; 2,758; 2,700; 2,642; 2,584; 2,526.

- Table 1 (Nonfinancial Corporate Leverage; Panel 2 IV):
  - FCI coefficients (h=2..12): 0.097; 0.578*; 1.013**; 2.060***; 2.747***; 3.043***.
  - Output Growth: -0.054***; -0.113**; -0.026; -0.000; -0.000; 0.030.
  - Inflation: 0.041; 0.016; 0.006; -0.048; -0.046; -0.105.
  - Policy Rate Change: -0.100; 0.311; 1.581***; 1.720**; 1.669***; 1.214*.
  - Observations: 2,277; 2,219; 2,161; 2,103; 2,045; 1,987.

- Table 2 (Household Leverage; Panel 1 least squares):
  - FCI coefficients (h=2..12): 0.065; 0.244; 0.464*; 0.769***; 1.085***; 1.265***.
  - Output Growth: -0.040***; -0.076***; -0.068***; -0.055**; -0.039; -0.023.
  - Inflation: -0.006**; -0.007; -0.005; -0.003; -0.001; -0.001.
  - Policy Rate Change: 0.007; -0.009; -0.001; 0.005; 0.000; 0.003.
  - Observations: 2,816; 2,758; 2,700; 2,642; 2,584; 2,526.

- Table 2 (Household Leverage; Panel 2 IV):
  - FCI coefficients (h=2..12): 0.054; 0.259; 0.456; 0.741*; 1.221**; 1.492**.
  - Output Growth: -0.042***; -0.108***; -0.104***; -0.100***; -0.081**; -0.052*.
  - Inflation: -0.005; 0.026; 0.091**; 0.110*; 0.137**; 0.129.
  - Policy Rate Change: -0.435***; -0.700***; -0.841***; -0.504; -0.909**; -0.984**.
  - Observations: 2,277; 2,219; 2,161; 2,103; 2,045; 1,987.

- Table 3 (Impact on Future GDP Growth)
  - Panel 1 (10th percentile impact; percentage points) — FCI coefficients (h=2..12): -1.121***; -1.250***; -0.269; 0.467**; 0.593***; 0.609***.
  - Panel 1 — Change in HH Leverage coefficients: -0.295; -0.543; -0.644; -0.718; -1.078**; -1.263**.
  - Panel 1 — Change in NFC Leverage coefficients: -0.319**; -0.768***; -0.681***; -0.424*; -0.184; 0.003.
  - Panel 2 (Median impact) — FCI coefficients (h=2..12): -0.935***; -0.928***; -0.189; 0.130; 0.097; 0.120.
  - Panel 2 — Change in HH Leverage: 0.104; 0.061; -0.002; -0.144; -0.389; -0.553**.
  - Panel 2 — Change in NFC Leverage: -0.246***; -0.491***; -0.496***; -0.373***; -0.262***; -0.191***.
  - Observations (both panels): 2,257 for all horizons reported.

- Table 4: linear regression results document association between one-unit net macroprudential tightening and subsequent change in debt-to-GDP (household and NFC) at horizons of 8 and 12 quarters, controlling for one-quarter change in sector debt, GDP growth, FCI, inflation, policy rate change, and global financial crisis dummy. Standard errors: Driscoll-Kraay.

---

### Robustness and Appendix analyses (Figures A1–A2, Table A1)
- Additional country-specific controls and alternative FCI measures examined to support robustness.
- AE-specific findings:
  - Increase in unemployment lowers leverage for corporates and more strongly for households (Figure A1, Panel 7).
  - Government balance has a positive effect on leverage (firms in near/medium term; households after a lag) (Figures A1 and A2, Panel 5).
  - Aging speed (future-40-quarter change in elderly ratio) increases firm leverage somewhat but not household leverage (Figures A1 and A2, Panel 6).
- Alternative measures of financial conditions (Table A1):
  - Domestic financial conditions generally outweigh alternative measures in explaining dynamics of private nonfinancial sector leverage.
  - Purged domestic FCI (purged of domestic macro factors) gives qualitatively similar results to original.
  - Global measures of financial conditions (as in IMF (2018a)) are not significantly related to household leverage when included alone or with domestic measures.
  - VIX is significantly related to nonfinancial corporate leverage but not household leverage.

---

*Source: International Institute of Finance; and authors’ calculations.*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2021/english/wpiea2021222-print-pdf.pdf_