Predicting corporate bankruptcy: where we stand?

Abstract

PurposeThe incidence of important bankruptcy cases has led to a growing interest in corporate bankruptcy prediction models since the 1960s. Several past reviews of this literature are now either out‐of‐date or too narrowly focused. They do not provide a complete comparison of the many different approaches towards bankruptcy prediction and have also failed to provide a solution to the problem of model choice in empirical application. Seeks to address this issue.Design/methodology/approachThrough an extensive literature review, this study provides a comprehensive analysis of the methodologies and empirical findings from these models in their applications across ten different countries.FindingsThe predictive accuracies of different models seem to be generally comparable, although artificially intelligent expert system models perform marginally better than statistical and theoretical models. Individually, the use of multiple discriminant analysis (MDA) and logit models dominates the research. Given that financial ratios have been dominant in most research to date, it may be worthwhile increasing the variety of explanatory variables to include corporate governance structures and management practices while developing the research model. Similarly, evidence from past research suggests that small sample size, in such studies, should not impede future research but it may lead researchers away from methodologies where large samples are critically necessary.Originality/valueIt is hoped that this study will be the most comprehensive to‐date review of the literature in the field. The study also provides a unique ranking system, the first ever of its kind, to solve the problem of model choice in empirical application of bankruptcy prediction models.