CMStatistics 2016: Start Registration
View Submission - CFE
Title: A joint model of firm failure and credit ratings Authors:  Laura Vana - WU Wirtschaftsuniversitaet Wien (Austria) [presenting]
Rainer Hirk - Vienna University of Business and Economics (Austria)
Kurt Hornik - WU Wirtschaftsuniversitaet Wien (Austria)
Abstract: Credit risk modeling including the measurement of credit quality has been intensively investigated by academics and practitioners over the past decades. We contribute by developing a framework for jointly modeling firm failures (e.g. bankruptcies) and ordinal credit ratings as outcomes. This model, unlike prior work, simultaneously incorporates failures and credit ratings and allows inference about the quantitative relationships between these outcomes by simultaneously making use of both sources of information. In addition the model does not require a balanced data set in the outcome variables, i.e. missing values in the outcomes are possible. We hypothesize that 1) the model alleviates the `low failure portfolio' problem and outperforms univariate models of failure or credit rating models in terms of prediction accuracy; 2) useful insights into rating heterogeneity are gained; 3) information about the systematic rating patterns of the credit rating agencies is uncovered. The joint model will be formulated such that it takes the ordinal nature of the credit ratings into account and can therefore be incorporated in the class of multivariate ordinal models. Failure, firm-level and stock price data for publicly traded North American companies as well as issuer credit ratings from the big three rating agencies (S\& P, Moody's and Fitch) are collected and analyzed to illustrate the proposed framework.