Provide a two-stage model for predicting corporate bankruptcy using data envelopment analysis

Predicting the bankruptcy of companies is one of the most basic activities in auditing risk and uncertainty of companies. Therefore, designing bankruptcy prediction models is essential for many decision-making processes. The purpose of this study is to provide a two-stage model for predicting corporate bankruptcy using data envelopment analysis.

To conduct this research, first data envelopment analysis was used using market-oriented and accounting-based financial data as well as non-financial data to measure the scores of management efficiency and stock market efficiency of companies and then scores were used. The obtained performance has been used to predict the bankruptcy of companies and 9 models of bankruptcy prediction have been proposed. Also, the proposed models were performed using logistic regression of conditional fixed effects and the best model was selected using the ROC curve. The study period of this research is 13 years (1385-1397) and the number of samples of this research includes 184 companies and 2392 year-company observations.

The results showed that the proposed two-stage model has a very good predictive power. The results also showed that management efficiency scores are directly related to bankruptcy. In other words, companies with lower management efficiency scores were in a bad position in terms of bankruptcy risk. Also, accounting-based data had better estimates of bankrupt companies.

The two-stage model presented in this study can be used with high confidence to predict bankrupt companies.