Rating Alignment, Rating Shopping and Reputation of Credit Rating Agencies: Evidence from the Subprime Crisis

Abstract

This paper compares conflict of interest incentives and reputational concerns of credit rating agencies (CRAs) in the context of the subprime crisis. We argue that, during up-market periods, ratings levels are affected by both a strong tendency for alignment across CRAs and ratings “shopping” by issuers, while, during periods of economic slowdown, these distortions disappear since CRAs are then more concerned about their long-run reputation. We test our hypotheses by analyzing the gap between Moodys and S&Ps ratings on US residential, subprime mortgage-backed securities before and after the 2007 crisis. Overall, our results show a clear reduction in ratings alignment. Moreover, we find strong evidence that harsher downgrades came from S&P, which had higher ratings before the crisis, and that the gap reduction is strongly correlated with the rating gap before the crisis. We interpret this as evidence that CRAs try to “reverse the gap”, to reduce the impact on their (relative) reputation. Finally, we find that harsher downgrades tend to occur for securities not jointly rated and that the relation between downgrades and initial rating is significantly different across the two agencies, this being consistent with the rating shopping hypothesis.

Keywords: credit rating agencies, subprime, conflicts of interest, rating alignment, rating shopping

JEL Classification: G01, G14, G24, G38