IMF-债务风险（英）-2025_59页_4mb

Debt-at-Risk Summary

Core Content

This working paper introduces a novel framework called "Debt-at-Risk" to analyze the risks surrounding the public debt outlook. The framework uses a quantile panel regression approach to assess how current macro-financial and political conditions influence the full distribution of future debt outcomes, not just the average. The paper highlights the importance of understanding both downside and upside risks in debt projections and provides empirical evidence on the magnitude and sources of these risks.

Main Points

• Debt-at-Risk Definition: Debt-at-Risk is defined as the 95th percentile of the future debt-to-GDP distribution, representing the upper tail risk. It is used to quantify how high public debt could rise in a severe adverse scenario.

• Key Findings:

  • In a severely adverse scenario, global public debt could be approximately 20 percentage points higher than current projections.
  • Financial stress and tighter financial conditions have asymmetric and nonlinear effects on the debt distribution, particularly affecting the right tail (upside risks).
  • Higher initial debt levels amplify the impact of economic and financial conditions on debt-at-risk.
  • Political uncertainty and social unrest also contribute to rising debt risks, especially in the short term.
  • Economic variables (such as GDP growth and primary balance) have long-lasting effects on the debt distribution.

• Methodology:

  • The framework uses a location-scale model to estimate the predictive distribution of debt-to-GDP.
  • It incorporates country fixed effects to control for non-time-varying characteristics.
  • Quantile regression is used to estimate the conditional distribution of future debt.
  • A skewed t-distribution is applied to model the density function of the predicted quantiles.
  • Pooled density functions are constructed by combining individual conditional densities using country-specific weights derived from out-of-sample predictive accuracy.

• Empirical Results:

  • A one-standard deviation increase in the financial conditions index is associated with a 3 percentage points of GDP increase in three-year-ahead debt-at-risk.
  • Debt-at-risk is a strong predictor of fiscal crises, outperforming other economic indicators.
  • The framework is applied to a sample of 175 economies, with 90 economies accounting for over 90% of global sovereign debt.
  • Global debt-at-risk for 2027 is estimated at 117% of GDP, significantly higher than the IMF's baseline projection.

Key Variables

• Financial Factors:

  • Financial Conditions Index
  • Financial Stress Index
  • Sovereign spreads (difference between 10-year government bond yields and U.S. Treasury yields)

• Political Factors:

  • World Uncertainty Index
  • Reported Social Unrest Index

• Economic Factors:

  • GDP growth
  • Inflation
  • Primary balance
  • Initial debt-to-GDP ratio

Extensions and Contributions

1. Fiscal Crisis Prediction:

   • Debt-at-Risk is shown to be a strong early warning indicator of fiscal crises.
   • It outperforms other variables in Bayesian Model Averaging and Random Forest models.

2. Global Sample Expansion:

   • The model is expanded to cover approximately 175 economies, including emerging market and low-income countries.
   • This allows for a more comprehensive assessment of debt risks across different income groups and economic structures.

3. Non-linearity and Heterogeneity:

   • The paper identifies non-linear relationships between conditioning variables and future debt.
   • Initial debt levels and income status (advanced vs. emerging) are found to be key sources of heterogeneity in debt-at-risk.

4. Methodological Advancement:

   • The framework provides a flexible and robust method for modeling the full distribution of debt risks.
   • It allows for country-specific analysis and flexible incorporation of new conditioning variables in future research.

Conclusion

The paper emphasizes the importance of quantifying the full distribution of debt risks, not just the mean. It provides a new analytical tool for policymakers to assess the potential severity of public debt outcomes in adverse scenarios. The Debt-at-Risk framework has been shown to be effective in predicting fiscal crises and is robust across different econometric methods. It also contributes to the broader literature on sovereign debt vulnerabilities, early warning systems, and macro-fiscal policy analysis.