Credit Risk: An Empirical Study of the Credit Default Swap Spread

Rikke Oien Carlsen & Louise Benche

Student thesis: Master thesis

Abstract

This thesis investigates the determinants of the credit default swap spread by a multiple linear regression model, using ordinary least squares method. The thesis opens with an outline of credit risk and the credit derivatives market. Furthermore, the characteristics of a CDS contract and the credit default swap spread are covered. On the basis of economic theory on credit risk and previous empirical researches, a set of determinants is identified. The thesis is highly inspired by the theory behind Merton’s structural credit risk model, and the findings in three empirical studies by Collin-Dufresne et al. (2001), Benkert (2004), and Ericsson et al. (2009). A model, which is consistent with the economic theories and at the same time both statistically and economically significant, is drawn up to investigate the linear relationship between the identified determinants and the credit default swap spread. The analysis is based on daily data from January 2005 to December 2016 on 79 American companies collected from the Markit CDX NA IG index. Both macroeconomic and firm-specific factors prove to be of importance, when determining the credit default swap spread. The three theoretical factors proposed by Merton (1974): firm leverage, volatility, and the risk free rate level were all found to have a significant influence on the CDS spread. CDS liquidity, GDP growth, and the firm-specific factors: credit rating and the price/book value also turned out to be statistically significant. However, the empirical analysis proposes that equity volatility, especially option-implied volatility, is the most central determinant of the credit default swap spread. Since the dressed up model only seemed to explain a little less than half of the variation in the CDS spread, the model was divided into shorter periods, representing different economic stages, and also run on the individual years. It was found that the explanatory power of the model and the variables’ behaviour vary, depending on the respective periods The leverage seemed to have a decreasing significant impact on the CDS spread over the years while at the same time, both of the volatility measures had an increasing impact on the CDS spread. By running the regression model on the individual years, it was found that the model, in general, is better at explaining the variation in the CDS spread in single years. The model and hence the identified determinants, proved to be much better at explaining the variation in the years prior to the financial crisis.

EducationsMSc in Finance and Accounting, (Graduate Programme) Final ThesisMSc in Finance and Investments, (Graduate Programme) Final Thesis
LanguageEnglish
Publication date2017
Number of pages138