Macroeconomic determinants of corporate CDS spreads: an empirical study

Abstract

Several theoretical studies suggest the importance of the macroeconomy for credit default swap (CDS) pricing. However, only few studies empirically investigate the effect of macroeconomic variables on CDS spreads. The previous analysis is further limited to only one or two macroeconomic variables. This provides a motivation for this PhD thesis to conduct an extended analysis on how macroeconomic variables, capturing various dimensions of the economy, affect CDS spreads. In addition, in contrast to the previous literature, the analysis here examines not only the effect of macroeconomic fundamentals but also macroeconomic uncertainty.
The thesis comprises three empirical chapters, all employing U.S. CDS data between March 2009 and December 2016. Chapter 3 employs the time series framework to explore how macroeconomic variables affect spreads of investment-grade and high-yield CDX, which are traded CDS indices. Chapter 4 accounts for the firm heterogeneity by incorporating firmspecific variables and explores how macroeconomic level and macroeconomic volatility affect single-name CDS spreads within the panel data analysis framework. Chapter 5 adopts a different angle by studying how macroeconomic news announcements affect CDS spreads. Chapter 3 finds that CDX spreads decrease with total nonfarm payroll growth but increase with industrial production growth volatility. A larger share of explained variation in CDX spreads is accounted for by macroeconomic level variables, while macroeconomic volatility is responsible for a smaller but, nevertheless, sizable share of explained variation. High-yield CDX spreads are more sensitive than investment-grade CDX spreads to macroeconomic variables. Chapter 4 further finds that single-name CDS spreads increase with leverage, industrial production growth volatility and 3-month Treasury Bill rate volatility but decrease with total nonfarm payroll growth. Firm-specific variables account for more than 90% of explained variation in single-name CDS spreads, with macroeconomic variables explaining a considerably smaller remaining share of variation. Similar to CDX spreads, single-name CDS spreads of high and low credit quality differ in their sensitivity to macroeconomic variables. Furthermore, Chapter 5 finds that unexpected announcements in total nonfarm payroll, advanced retail sales, and the ISM manufacturing index reduce CDX spreads. Unexpected announcements in total nonfarm payroll have the most profound effect. The analysis in all empirical chapters highlight the importance of nonfarm payroll, an employment indicator which has not been previously investigated in the CDS pricing literature, indicating its relevance for future theoretical and empirical CDS pricing models.