Bankruptcy prediction by using efficiency as a predictor variable based on Logit Panel data

Investigating the factors affecting bankruptcy is one of the most important issues in financial literature. Investigating the factors affecting bankruptcy can provides useful information to government, central bank, banks, financial institutions, creditors, capital market investors. In this paper we discussed the differences between two concepts of bankruptcy and insolvency, then calculated efficiency variable through data envelopment analysis (DEA) and we have fitted it as an explanatory variable alongside of available variables in logit panel data for 36 manufacturing companies during the period of 1388-93. The results show that efficiency model have been calculated by constant return to scale, variable return to scale, collective pattern and Anderson- Peterson methods, and also efficiency variable are significant with constant return to scale. Finally, logit regression model is fitted twice, without specification of efficiency variable and further with efficiency variable. We find that in all patterns which fitted by efficiency variable, accuracy of model prediction significantly increases.