As part of my bachelor's thesis, I am running the following regression on daily bank excess returns:
(r-rf)=Beta * Market excess return + Beta * Level(5Y)+ Beta * Credit Risk + error
Level(5Y) is the daily return on a portfolio of 5-year maturity treasury bonds and captures interest rate risk. Credit risk is obtained by taking the residual of the daily returns of a portfolio of BBB bonds when regressing them on the interest rate factor, so they are orthogonal.
While all other betas turn out reasonable, the credit risk beta is -0.59 (significant) in 2016-2019 and I do not understand how it can be so negative. BBB bonds yield positive returns when yields decrease, which means decreased credit risk - and my results show that bank returns decrease following this. Shouldn't this be good for bank equity returns? I was thinking if they may be using credit derivatives but found that this has decreased a lot since 2008. Does anyone have an explanation to this credit risk result? It would be so greatly appreciated.