The Prediction of Corporate Failure: Testing the Australian Experience

Abstract

The purpose of this paper is to address some of the methodological issues which have evolved from the literature on corporate failures, and to report the results of an empirical investigation on the usefulness of financial models for the prediction of corporate failures. The experimental design is based on the 21 listed public companies that failed in Australia during the period 1963 to 1977, inclusive, and which met with minimum data requirements. It examines the performance of linear versus quadratic classification rules; temporal versus atemporal models; equal versus unequal priors of failure, variable dimension reduction, and a validation test proposed by Lachenbruch (1967). The results of the study suggest that it is difficult to identify a unique model to predict corporate failures, without specifying the utility preference of the user. Utility preference in this context refers to the minimization of either Type I, Type II, or the overall error rate of a failure model.