Why are credit ratings coarse?

Abstract

An enduring puzzle is why credit ratings are coarse indicators of issuer credit quality, with a relatively small number of ratings categories being used to describe credit qualities that lie in a continuum. We develop a theoretical model to explain why ratings are coarse even though coarseness reduces welfare. We model the ratings-determination process as a cheap-talk game in which there is a divergence between the goals of issuing firms and those of investors, and the rating agency's objective gives positive weight to both goals. We show that the rating agency is willing to inflate ratings, but wishes to keep the rating inflation below a threshold bias. Consequently, it prefers an unbiased rating to a rating inflated by an amount exceeding the threshold bias. Coarse ratings arise as an equilibrium phenomenon due to the effects of these forces. The coarseness of ratings means that rating inflation, if it were to exist, would have to exceed the threshold bias, so ratings exhibit no bias/inflation in equilibrium, and legal liability to eliminate bias is unnecessary. Moreover, competition among rating agencies does not necessarily reduce ratings coarseness. We also examine welfare implications of regulatory initiatives involving ratings.