世界发展银行-The-Impact-of-Oil-Shocks-on-Sovereign-Default-Risk_45页_1mb

Summary of "The Impact of Oil Shocks on Sovereign Default Risk"

Core Content

This paper investigates the impact of daily oil supply and demand shocks on sovereign credit default swap (CDS) spreads for G10 countries and major oil-exporting countries. The study employs quantile regression and two-state Markov-switching models to analyze the non-linear and state-dependent effects of oil shocks on sovereign default risk.

Main Findings

• Oil demand shocks have a uniformly negative impact on CDS spreads for both G10 and oil-exporting countries.
• Oil supply shocks have differential impacts:
  • They increase CDS spreads for G10 countries.
  • They reduce CDS spreads for oil-exporting countries.
• Quantile regression analysis reveals:
  • Demand shocks affect CDS spreads across the entire conditional distribution.
  • Supply shocks mainly influence the upper quantiles of spread changes.
• Markov-switching model confirms:
  • The effect of oil shocks on CDS spreads is non-linear.
  • The impact varies depending on the state of the economy.
  • Supply shocks have a significant impact only in high volatility states.
  • Demand shocks have a significant impact in low global economic activity states.
• The impact of oil shocks on CDS spreads also depends on the country's oil dependence:
  • For G10 countries, oil-import dependence is associated with stronger CDS spread changes.
  • For oil-exporting countries, heavy reliance on oil revenue amplifies the impact of supply shocks.

Key Contributions

1. Robust evidence is provided that both supply and demand shocks significantly affect CDS spreads for G10 and oil-exporting countries.
2. The study shows how oil shocks influence CDS spreads across the entire distribution using quantile regression.
3. The state-dependent nature of the relationship between oil shocks and CDS spreads is confirmed through Markov-switching models.
4. The differential impact of oil shocks on oil-importing and oil-exporting countries is emphasized, highlighting the need to identify the source of oil price changes and acknowledge heterogeneity in responses.

Methodology

• Regression Model (4) is used to test the first hypothesis, incorporating:
  • Lagged CDS spread changes.
  • Domestic control variables (e.g., stock returns, FX rates).
  • Global control variables (e.g., TED spread, VIX, S&P 500, bond yields).
  • Lagged oil demand and supply shocks.
• Quantile regression (6) is applied to assess the distributional impact of oil shocks on CDS spreads.
• Markov-switching model (7) is used to examine the state-dependent effects of oil shocks.

Data and Variables

• Sample Countries:
  • G10 countries: United States, United Kingdom, France, Germany, Italy, Japan, Belgium, Netherlands, Sweden, Switzerland.
  • Oil-exporting countries: Russia, Iraq, UAE, Kazakhstan.
• Data Sources:
  • Daily oil prices from the NYMEX - Light Sweet Crude Oil contract.
  • World Integrated Oil and Gas Producer Index from Thomson Reuters.
  • CBOE volatility index (VIX).
  • Control variables include:
    • Domestic: stock market returns, FX rates.
    • Global: TED spread, VIX, S&P 500, Euro Stoxx 50 Volatility index, bond yields, and Treasury yields.
• Global Economic Activity is measured using the OECD and major country industrial production index.
• Oil Dependency is measured using:
  • Energy import to GDP ratio for G10 countries.
  • Oil rent to GDP ratio for oil-exporting countries.

Empirical Results

• G10 countries:
  • Demand shocks affect CDS spreads across all quantiles.
  • Supply shocks affect CDS spreads only in extreme quantiles.
• Oil-exporting countries:
  • Demand shocks affect CDS spreads only in low economic activity states.
  • Supply shocks significantly impact CDS spreads only in high volatility states.
• CDS spreads are highly sensitive to changes in sovereign solvency, as illustrated by the case of Greece during the 2015 financial crisis.
• Correlation between CDS spreads is higher within regions, with Germany and France showing a strong correlation (0.79), and Netherlands and Belgium showing a moderate correlation (0.69).
• Control variables show moderate to low correlation with each other and with oil shocks.

Conclusion

The paper concludes that oil shocks significantly influence sovereign default risk as measured by CDS spreads, with differential impacts based on the type of shock, country's oil dependence, and economic state. The findings highlight the importance of understanding the non-linear and state-dependent nature of oil shocks in assessing sovereign credit risk, and suggest that policy responses should be tailored to the specific characteristics of oil-importing and oil-exporting economies.