Cantor, Richard and Frank Packer, "Multiple Ratings and Credit Standards: Differences of Opinion in the Credit Rating Industry", Federal Reserve Bank of New York, Research Paper #9527, (December 1995).

Abstract: We test whether the tendency of third rating agencies to assign higher ratings than Moody's and Standard & Poors results from more lenient standards or simple selection bias. More lenient standards might result from incentives to satisfy issuers who are, in fact, the purchasers of the ratings. Selection bias might be important because issuers that expect a low rating from a third agency are unlikely to request one. Our analysis of a broad sample of corporate bond ratings at year-end 1993 reveals that, although sample selection bias appears important, it explains less than half the observed difference in average ratings.

We also investigate why bond issuers seek ratings in addition to those of Moody's and Standard & Poors. Contrary to expectations, we do not find that the probability of obtaining a third rating is related to levels of ex ante uncertainty over firm default probabilities. In particular, a firm's decision to obtain a third rating appears unrelated to its Moody's and Standard & Poors ratings or the amount of disagreement between them. Instead, the most important determinants of the decision are a firm's age and size.

Our results should be of interest to investors and financial market regulators who generally use the ratings of different agencies as if they correspond to similar levels of default risk. In addition, our findings raise a number of questions about the certification role of rating agencies and about the strength of the rating agencies' incentive to maintain a reputation for high quality (accurate) ratings.