Credit-Rating Shopping, Selection and the Equilibrium Structure of Ratings

Abstract

An important feature of the microstructure of credit ratings is the ability of a security issuer to choose which ratings to purchase. Such choices can reflect explicit or even implicit shopping for favorable credit reviews and induce effects in the structure of ratings. When there is considerable heterogeneity in views the issuer selects the ratings that are the most positive. Because the correlation among models is least when heterogeneity is greatest, we examine the influence of the correlation on the extent of ratings shopping and bias. Selectivity is central to the decision to solicit (purchase) a credit rating or for the credit rating agency to provide an unsolicited rating based upon coarser information. Our analysis highlights the interaction between the decision about whether to rely on unsolicited ratings and the potential for ratings shopping, illustrating the interaction between different types of potential conflicts of interest in the credit rating process. We use selection to provide an equilibrium interpretation for notching (formulaic haircutting of a credit rating agency's rating) by a rival. We point to the potential winner's curse that is implicit when the most favorable rating is selected, raising the issue of how rating agencies, regulatory authorities and investors interpret solicited ratings. We conclude by presenting theoretical results and numerical solutions to illustrate qualitative aspects of rating shopping. For example, we show that higher cost of obtaining indicative ratings and regulatory mandates to charge fees for obtaining indicative ratings reduce the extent to which these are obtained, increasing the extent of ratings bias. The higher cost of ratings also reduces the probability of notching, i.e., the likelihood that an unpublished rating is below an existing published one.