Default Probability Modelling for Russian Banks

The aim of current research is to build a default probability model for Russian banks using CAMELS methodology and additional regressors including institutional parameters (participation in deposit insurance system and location of head office of the bank) and macroeconomic variables (GDP growth and inflation). Importance and relevance of this research can be attributed to the increasing number of bankruptcies after the appointment of Ms. Nabiullina as the head of the Central Bank and consecutive policy of liquidation of insolvent and financially unstable banks. Due to this fact, the time span under consideration is from 3rd quarter 2013 till 3rd quarter 2018. Using logit model the author built a model based on the whole sample which showed that institutional parameters contribute more to the accuracy of the “default state” prediction than macroeconomic variables. The model also revealed negative correlation between capital adequacy, share of liquid assets, dummy standing for participation in the deposit insurance system and default probability, positive correlation between location of the bank’s head office in Moscow, inflation and default probability. Moreover, U-shaped relationship was present between default probability and profitability of the bank. The next step was the division of all the banks in the sample into 4 groups according to the size of their assets and capital sufficiency level (S/L, S/M, S/H, B) which was accomplished in order to construct a separate model specification for each group. The optimal model judging from accuracy and type I error levels was the one for big banks (B): 77% and 42% accordingly. It should be noted that some of the models for other groups of banks had higher accuracy (80% for group S/H) or lower type I error (27% for group S/M). In conclusion, it was proved that 2 quarter lag for regressors is the optimal one because increasing the lag to 1 year leads to lower accuracy of default prediction while reducing the lag won’t allow banks to react to worsening of their financial position in time in order to prevent potential bankruptcy.