THE ANA LYSIS OF PROBABILISTIC CORREL ATIONS BETWEEN THE RATINGS OF THE LEADING INTERNATIONAL COMPANIES MOODY’S, FITCH AND STANDARD & POOR’S

Abstract

The article surveys leading international rating agencies (Moody’s, Fitch, and Standard&Poor’s) and analyses the correlation problem of their credit attractiveness rating scales. The research briefly reviews the existing methodological approaches to the solution of this problem. As a result, a new method based on the theory of binary relations and vector criterion selection functions is proposed. A strict order relation generated bythe probability of default is taken as a preferable relation. A specially introduced rating resulting from Pareto ranking of multidimensional parameters of current estimates of Moody’s, Fitch, and Standard&Poor’s is chosen as a benchmark. The mechanism of ranking a particular international agency is immaterial in this approach. It eliminates the dependence on politically-biased expert assessments. Naturally, such an extension may not lead to mutual unambiguity of scales and provides their probabilistic adequateness in return.