2005年-世界发展银行全球_Global_Monetary_Conditions_versus_Country-Specific_Factors_in_the_Determination_of_Emerging_Market_Debt_Spreads_31页_394kb

Summary of "Global Monetary Conditions versus Country-Specific Factors in the Determination of Emerging Market Debt Spreads"

Core Content

This paper investigates the relationship between global monetary conditions, particularly US interest rate policy, and emerging market (EM) debt spreads over US Treasury securities. It challenges existing literature by emphasizing the role of market expectations and non-linearities in the link between US interest rates and EM spreads.

Main Points

• Influence of US Interest Rates: US interest rate policy significantly affects EM debt spreads. The study shows that the effect is non-linear and depends on the country's debt level and solvency.

• Market Expectations: The market anticipates changes in US monetary policy by observing macroeconomic indicators such as non-farm payrolls and retail sales. These expectations shape the perception of default risk and thus the spreads.

• Non-Linearity in Default Probability: The probability of default is not linearly related to US interest rates. A moderate level of debt may not be affected much by a rise in US rates, but a country near the solvency borderline may see a sharp increase in default risk.

• Multiple Equilibria: The model suggests that shifts in expectations can lead to multiple equilibria, where a rise in US rates may push a country into a higher probability of default, especially if the curve shifts up enough.

• Crisis vs. Non-Crisis Periods: The paper divides the sample into crisis and non-crisis periods. It finds that during non-crisis times, EM spreads are more sensitive to global monetary conditions, while during crises, they are more influenced by country-specific factors.

• Contagion Effects: Contagion from other countries in crisis is significant during non-crisis periods but not during crisis periods. This suggests that during crises, the spread is more a function of internal factors than external ones.

• Empirical Methodology: The study uses panel regressions with both "pull" (country-specific) and "push" (global) factors. It estimates a long-run relationship and short-run dynamics using an error correction model (ECM).

• Estimation Techniques: The Pooled Mean Group (PMG) estimator is used to capture the long-run relationship across countries, while the short-run dynamics are modeled with lagged residuals.

• Key Variables:

  • US Interest Rates: US short-term and long-term rates are included, with the short-term rate showing a non-linear effect when interacting with the country's debt/GNI ratio.
  • Corporate Bond Spreads: The spread between high and low risk US corporate bonds is a significant "push" factor, especially in non-crisis periods.
  • Country-Specific Factors: Trade openness, debt/GNI, reserves/debt, and short-term debt ratio are important "pull" factors.
  • Contagion Dummy: Reflects the number of other countries in crisis and has a positive effect on spreads in non-crisis periods but not in crisis periods.

Key Findings

• Non-Crisis Periods:

  • US short-term interest rate has a non-linear effect, especially when interacting with debt/GNI.
  • The spread between high and low risk corporate bonds is a strong explanatory variable.
  • Contagion effects are significant, indicating that EM spreads can be influenced by external crises.

• Crisis Periods:

  • US interest rate variables (including short-term and long-term rates) have less explanatory power.
  • The effect of global monetary conditions is less important compared to internal creditworthiness.
  • Contagion has no significant effect during crisis periods.

• Simulation Results:

  • A 200 bps increase in US short-term interest rates leads to EM spreads increasing from 6 bps to 65 bps, depending on the country's debt/GNI ratio.
  • This highlights the sensitivity of EM spreads to US monetary policy in countries with higher debt levels.

Conclusion

The paper concludes that EM debt spreads are influenced by both global monetary conditions and country-specific creditworthiness, with the relationship being non-linear and context-dependent. The inclusion of market expectations and the distinction between crisis and non-crisis periods enhances the understanding of how global interest rate changes affect EM spreads. The use of more recent data and a more sophisticated econometric approach helps in capturing these dynamics more accurately.

Key Variables and Their Effects

Variable	Non-Crisis Periods	Crisis Periods
US short-term rate (linear)	Not significant	Not significant
US long-term rate	Negative but insignificant	Negative and insignificant
Hi-low corporate spread	Strong positive effect	Strong positive effect
Trade openness	Strong negative effect	Strong negative effect
Debt/GNI	Positive effect	Positive effect
Reserves/Debt	Negative effect	Negative effect
Short-term debt ratio	Negative effect	Negative effect
Contagion dummy	Positive effect	No significant effect
US Tbill rate × Debt/GNI	Significant positive effect	No significant effect

Data and Methodology

• Time Period: 1991M1 to 2004M6.
• Countries: 17 major EM countries, including Argentina, Brazil, Mexico, Russia, and others.
• Data Sources: EMBI+ spreads from JP Morgan, secondary market data.
• Model: Panel regression with long-run and short-run components, using an error correction model.
• Estimation Technique: Pooled Mean Group (PMG) estimator for long-run relationships and short-run dynamics.

Implications

• The study underscores the importance of understanding the non-linear and expectation-based nature of the relationship between US interest rates and EM spreads.
• It highlights the role of contagion and the need to differentiate between crisis and non-crisis periods in modeling EM spreads.
• The findings suggest that global monetary policy is a key determinant of EM spreads, especially in countries with moderate debt levels.