Do riskier firms pay out higher returns? Evidence from US CDS spreads and stock returns in 2020

The relation between the CDS spread and the stock return has caused a lot of debate in the literature and it is still now a discussed topic. Some papers suggest the existence of a positive relationship between the two, meaning that an high measure of CDS spread of a ?rm is associated with an high return on the same company on the day after. Others advise a negative cor- relation implying that an higher value of CDS spread tends to be associated with lower values of the sock return. In the latter case, the literature calls this negative relationship as "distress risk puzzle". I approach this debate by studying the sign of the relationship among the year 2020 by using daily CDS spread. The latter is a particular year since the spread out of Covid-19, that causes many stops leads to a downturn in the US economy but not only. In section 2, I brie y review the literature on the relationship about the CDS spread, citing how it is constructed, the in uence of the public and private information on it and the empirical relationship between the CDS spread and the stock return. Section 3 describes the data set and presents the descrip- tive analysis of the CDS, by studying how the data set is composed and by studying its term structure and comparing the behaviour of the slope with some important economic variables. I study, graphically, the relationship between the CDS spread and the main US indexes, such as Nyse, Nasdaq and Dow Jones, and also with the GDP growth rate. In the last case, since the GDP data is only given quarterly, if it is possible to obtain daily information about it from the CDS spread. The results suggest that the CDS spreads and the GDP growth rate are correlated in a positive way con?rming that it is possible to extract information about the GDP from the term structure of the CDS spread. Lastly, Section 4 describes the results of the empirical analysis. Firstly, I construct portfolios on daily, weekly and monthly basis by sorting stock re- turns CDS spreads, from the lowest to the highest. As maturity of the CDS contract, I use tenor 1, 5 and 10, since they are the most frequent traded in the market. I ?nd no clear pattern, since by increasing the CDS spreads, the stock returns do not increase and neither decrease. In the sample considered, the pattern between the CDS spread and stock return looks like a random noise. This can be caused by the random noise present in daily CDS spreads or can be caused by the particular year that I considered. After that, I conduct an ols regression in ordered to estimate coe?cients of linear regression equation which describes the relationship between a in- dependent variable, the CDS spreads at time t, and a dependent variable, stock returns at time t+1. I perform a study on the number and percentage of positive or negative and signi?cant or not alphas and betas, in order to understand whether the relationship, in the sample that I consider, is at most positive or negative. Moreover, I conduct a cross section and a cross rating analysis with the aim to verify the relationship in each sector and rating. Section 4.3 describes the results of the implementation of the bi factorial CAPM. The latter is a CAPM increased with CDS spreads, the dependent variable is always the stock return at time t+1 and the the independent vari- ables are the the CDS spread at time t and the market return at time t+1.