An Analysis of the Determinants of S&P Ratings Assigned to Canadian Firms : Application of a Multinomial Logit

This study identifies the factors affecting the ratings assigned to Canadian firms by the rating agency Standard and Poor's, and proposes a multinomial logit model to predict ratings. We first propose a two-class model which classifies ratings as either investment or speculative grade. We then extend this model to account for all the rating levels used by the rating agency. The out-of-sample forecasts respectively lead to concordance rates of 86% and 60% in the two-class and extended models. Our credit rating models thus capture the main determinants affecting the ratings changes to some extent