Empirical Analysis of Credit Default Swaps

Lasse Frederik Samuelsen

Student thesis: Master thesis

Abstract

This paper examines variables explanatory power in the movements of Credit Default Swap (CDS) spreads in the years of 2006 to 2016. By collecting daily spreads from 112 entities in the CDX.NA.IG index and 58 entities in the CDX.NA.HY index, multiple regressions is conducted to examine the relationship between the CDS spreads and the variables. Furthermore, the relationship between stock returns and CDS spreads are examined by cointegration and causality tests. This paper finds that there is no cointegration between the CDS Spreads in CDX.NA.IG and the stock returns of these entities nor between the CDX.NA.HY and the stock returns of these entities. Also, the result of the causality tests is that the causality runs from stock returns to CDS spreads i.e. Stock returns causes the CDS spreads to move. The variables included in the multiple regression is a mix of structural variables from the traditional Merton model, the Merton jump diffusion model, the Altman z-score and a handful of variables suggested by previous literature. This paper finds, that 20 percent of the movements in the CDX.NA.HY index can be explained by the these variables and 10% percent of the movements in the CDX.NA.IG can be explained by these variables. The robustness of these variables is tested by splitting up the time periods in three periods. A before, during and after crisis time period. The variables that are most robust to the three time periods is the variables suggested by the traditional structural Merton theory.

EducationsMSc in Finance and Investments, (Graduate Programme) Final Thesis
LanguageEnglish
Publication date2017
Number of pages112