Corporate Distress and Bankruptcy Prediction–A Critical Review of Statistical Methods and Models

Bankruptcy prediction has interested and intrigued accountants and researchers alike since early 1930’s. The empirical and theoretical research conducted till date seek to find the best statistical method to develop distress / bankruptcy prediction model and also check the validity of the models developed across different industries, sectors and countries. This paper is an attempt to critically review the various models developed for bankruptcy prediction and the statistical methods adopted in such studies. It is observed that there is no consensus as to the best method and model for corporate distress and bankruptcy prediction. However certain techniques like multivariate analysis. Logistic regression and Artificial Neural Networks have found favour with researchers and academicians alike. Continuous attempts are on to discover newer techniques and methods to develop a robust bankruptcy prediction model.