TL;DR
This paper uses a Markov regime-switching model to show that the factors influencing the credit risk of emerging markets (as measured by Credit Default Swap spreads) change depending on whether the market is in a "good" or "bad" state. Specifically, local and regional factors dominate in good times, while global factors, particularly US stock market returns, become more influential during market downturns.
What Problem Does This Paper Address?
Sovereign credit risk, a crucial indicator of a country's ability to repay its debts, is often reflected in the pricing of Credit Default Swaps (CDS). However, understanding what drives changes in this risk, especially for emerging markets, is complex. Previous research has identified various factors (local, regional, global) influencing CDS spreads, but it's unclear if these influences remain constant or shift with market conditions. This paper addresses this gap by investigating how the determinants of emerging market sovereign CDS spread changes vary across different market regimes (good vs. bad), aiming to provide a more nuanced understanding for risk management.
Methods
The core methodology employed is a Markov regime-switching regression model. This statistical technique allows for the estimation of relationships that change over time, assuming the underlying process can switch between a finite number of unobservable "states" or "regimes." In this study, a two-state model is used, representing a "good" market state and a "bad" market state. The model estimates separate coefficients for each regime, allowing the influence of explanatory variables to differ. The authors also conduct standard Ordinary Least Squares (OLS) regressions as a benchmark and a cross-sectional analysis to explore how country-specific characteristics moderate these regime-switching effects.
Datasets
The study utilizes weekly data for 11 representative emerging market countries from May 2001 to December 2012. The countries are drawn from Asia, Europe, Latin America, and the Middle East/Africa. The primary data source for CDS spreads is Markit Financial Information Services. Explanatory variables are sourced from Bloomberg, S&P, and the Federal Reserve.
Key Insights and Ideas
- Market Regimes Matter: The most significant insight is that the drivers of sovereign credit risk are not static. The "good" and "bad" market states exhibit distinct patterns of influence from various factors.
- Global Factors Dominate in Bad Times: During market downturns ("bad" regime), global factors, particularly the US stock market return (RS&P500), become the dominant influence on emerging market CDS spreads. This highlights investor behavior like loss aversion and a "flight-to-quality" phenomenon.
- Local/Regional Factors Lead in Good Times: In more stable periods ("good" regime), local financial variables (stock returns, exchange rates) and regional CDS spreads play a more significant role in determining a country's credit risk.
- Country Characteristics Influence Sensitivity: The study reveals that a country's openness, economic size, and integration into the global economy influence how susceptible its CDS spreads are to different factors and regimes. For instance, more open and smaller economies tend to be more affected by regional and global shocks.
- US Stock Market as a Key Global Indicator: The US stock market return emerges as a particularly strong indicator of global sentiment and risk, especially during market stress.
Key Results
- Regime-Specific Coefficients: The Markov regime-switching model shows statistically significant differences in the coefficients of explanatory variables between the good and bad regimes for most countries.
- Dominance of RS&P500 in Bad Regime: The negative impact of US stock market returns on emerging market CDS spreads is significantly stronger in the bad regime compared to the good regime, often by a factor of 2 to 5, and in some cases, much more (e.g., China).
- Local Factors' Influence: Local stock market returns generally have a negative impact on CDS spreads (tightening spreads in good market conditions), and this effect is more pronounced in the good regime for many countries.
- Regional Interconnectedness: Regional CDS spread changes consistently show a positive and significant effect across both regimes, underscoring the strong inter-coupling of countries within geographic proximity.
- Cross-Sectional Findings: Open countries are more influenced by regional factors in the bad state, while smaller economies are more susceptible to both regional and global shocks. Rating changes are more influential in the good state and for closed economies.
Limitations and Future Work
- Two-State Simplification: The study uses a two-state model, which might oversimplify the complex dynamics of market regimes. Future research could explore models with more than two states.
- Data Granularity: While weekly data is used, daily data might offer finer insights into short-term dynamics, though it presents challenges for regime-switching analysis.
- Explanatory Power: While the model captures regime-switching behavior, the paper acknowledges that not all variation in CDS spreads is explained. Further exploration of additional variables or more complex interactions could be beneficial.
- Country-Specific Nuances: While the cross-sectional analysis provides valuable insights, future work could delve deeper into the specific institutional, political, and structural characteristics of individual countries that drive these differences.
- Policy Implications: The authors suggest that their findings are crucial for emerging market investors in formulating country-specific credit risk management strategies. Further research could explore the practical implementation of these insights into dynamic trading and hedging strategies.
Summary updated