A Pattern for Dynamic Prediction of Financial Distress by Using Survival Analysis

In spite of extensive research on predicting financial distress and bankruptcy there seems to be no acceptable pattern for predicting based on known financial theories. Therefore. More research in this field will lead to better understanding of financial crisis. In this research, we tried to find a prediction pattern using the Cox model. Using a company-year sample list of 19536 companies during the time period between 1384 and1395, this study investigates dynamic probability predictions for Iranian firms. This estimated pattern uses a combination of accounting data and stock market information. The pattern provides predictions of survival probability using the cox model with time-varying variables. Forecast accuracy is evaluated using receiver operating characteristics curve and the Brier score. It was found that the dynamic model has acceptable predictive power.