Comparison of Indicators for Determining the Thresholds of Banks\' Financial Crisis in EWS Based on Business Cycles (in Persian)

The purpose of this paper is to design a prediction system for thresholds of the bankruptcy of banks based on the business cycle and examine the effects of different approaches in defining the bankruptcy threshold in predicting bankruptcy time of Iranian banks using the Kaplan-Meier and Cox Proportional-Hazards Models. So, the data of listed banks in Tehran Stock Exchange were used from 1385-1395. The results indicate that Iranian banks are affected by five leading variables, which bank supervisors can identify risky banks using these indicators. These variables are operating profit to operating costs ratio, the ratio of total net income of interest and operating income to the total assets, the income of banking services to total income ratio, the ratio of administrative and general expenses to total expenses and bank’s size. Although, results indicate that base on AIC, Z-score approach is the best criteria for defining and identifying bankschr('39') bankruptcy thresholds in comparison with capital adequacy ratio and non-performing loans ratio. Based on this approach, at overall, banks will be on the thresholds of bankruptcy for up to 7 years and this will be reduced to 5 years in recession and during the boom, bankschr('39') bankruptcy thresholds will increase to 8 years.