quantifying influencing factors

[churchman]

I want to quantify systematic risk factors, which influences the market spread of credit derivatives like CDS, CDOs and the credit indices (iTraxx, CDX).
So first i'am using a linear regression method to examine the single-name CDS for the factors, which may possibly determine the spread (maybe S&P500, FamaFrench, risk less yield, etc.). This gives me an answer to the question, what factors really have the power to effect something.
But now is my question, how can i quantify the variable, for which the R^2 indicates significance.

 

[churchman]

nobody an idea?

 

[churchman]

What would you think is to do, when the topic of the study is called: "Quantification of systematic factors that influence the pricing of credit derivatives - a comparative empirical study"

So, I considered to choose the causal chain CDS - CDS Index - CDO (tranched) in order to show that the systematic risk is most probably increasing because of the impossibility of diversification. Only the idiosyncratic (specific) risk of the CDS can be diversified by pooling in an index or additionally by tranching. Now I have the plan to check the sensitivity of the CDS at my assumed risk factors. For this purpose I will use a univariate or multivariate regression of changes in CDS Spreads. But I don't really know what is about the quantification.