| Summary: | We investigate the relationship between firm specific and macro variables on credit default spreads.
We collect weekly CDS spreads for North American companies, as well as firm specific and macro
variables, from July 2011 to July 2016. A panel based model to conduct both fixed effects and OLS
regressions on the levels and differences datasets. The variables used in our regressions are in line
with the theoretical and additional variables mentioned by Ericsson and Colin Dufresne which
consist of leverage, risk free rate, equity volatility, Vix index, and slope. We collect weekly CDS
spreads for North American companies, as well as firm specific and macro variables, from July 2011
to July 2016.The explanatory power of our variables in the OLS regression is 22% and extends to 35%
in our robustness analysis for our levels dataset. While the R square for our differences dataset in
the OLS regression is 13.8% and extends to 14.5% in the robustness analysis. Our results suggest that
theoretical determinants (leverage, equity volatility & risk free rate) have rather limited explanatory
power, and that additional determinants contain useful information about the CDS spreads.
|
|---|