Discussion of the Determination of Long-Term Credit Standing with Financial Ratios

Abstract

The general concern of Professor Horrigan's paper is said to be the investigation of the utility of accounting data in long-term credit administration, where the major problem is the determination of default risk on bonds. The specific question which he attempts to answer, however, is whether or not one year's financial data, viewed primarily in ratio form, can be used to predict corporate bond ratings. Because this question is rather limited in scope, at least relative to Horrigan's general concern, it seems appropriate to begin my comments by examining the logic which led him to choose it. On page 45, Professor Horrigan states that the ideal independent variable in study of default risk would be a measure which could be scaled on continuum ranging from certain repayment through certain default. He then comments, however, that such variable obviously does not exist. Few, if any, students of the bond market would seriously question this last statement. However, it does not follow that all of them would immediately agree that the best possible surrogate measure of default risk is bond ratings. (I realize, of course, that Professor Horrigan does not say ratings are the best possible surrogate, but it also seems reasonable to assume that he regards them as better surrogate than others which might just as easily have been used.) I would suggest, however, that an alternative and perhaps better measure of default risk might be bond yields themselves, or the risk premium on bonds, that is, the difference between bond yields and yields or long-term, default-free treasuries. After all, the market is assessing default risk too, and its assessments are considerably more continuous than those made by Moody's. I hasten to add that my remarks are not meant to imply that bond ratings are