Business Cycle and Stability of Bankruptcy Prediction Models

Abstract

This study examines the effects of business cycles on the stability of bankruptcy prediction models in Iran economic environment. In this study, business cycles are identified by Hodrick-Prescott Filter and bankruptcy prediction is examined by Logit and Discriminate Analysis models. The sample consists of 118 firms listed in Tehran Stock Exchange for the years 2002 to 2011 and the hypothesis is tested by comparing the efficiency and stability of the fitted Logit and Discriminate Analysis models in the business cycles. Empirical findings of the study suggest that Business cycle has effects on the efficiency and stability of the models. In other words, the efficiency and structure of the models are different in each phase of economic cycle. On the other hand, Designing the models and comparing their efficiency indicate that the logistic regression was more efficient than discriminate analysis in bankruptcy prediction.

Keywords

• Business cycle
• Bankruptcy Prediction Models
• Hodrick-Prescott Filter
• Logit and Discriminate Analysis models