An Early Warning System of Bank Failures: Identification of Key Indicators

Unfavorable socio-economic conditions and the policy of the Bank of Russia of improving and strengthening the banking sector over the past few years have resulted in a significant reduction in the number of credit institutions. The past events demonstrated the presence of a significant number of credit institutions with financial problems in the banking system, in the process of maintaining or liquidating which large public expenditures were incurred. During the crisis, significant losses of financial assets are mainly were incurred by creditors and investors, and also significant budgetary funds used to maintain financial stability of financial organizations or liquidation of banks. In particular, the Deposit Insurance Agency bears significant insurance liability when the arbitration court institutes bankruptcy proceedings against a bank deprived of a banking license. In connection with this, the development of early warning systems for banks' bankruptcy is beginning to take on special importance recently because this makes it possible to identify the potential problems of the bank at the initial stage of their formation. The use of the systems which allow identifying financially unstable banks before the occurrence of a case of their financial insolvency (bankruptcy) will cause taking the effective measures to improve the financial condition of the bank. The purpose of this paper is to develop probabilistic statistical models for predicting the bank’s bankruptcy using the system of advancing indicators of the financial condition. Due to differences in the specifics and characteristics of large and small banks, we assume the influence of a different set of financial indicators on the probability of their bankruptcy. Therefore, the model will be built on two different samples: large and small banks.