MULTIDIMENSIONAL REGRESSIVE MODELING AND FORECASTING OVERDUE CREDIT INDEBTEDNESS

The paper proposes a new approach to formation of bank reserve on possible credit losses. This methodology is considered as an alternative to the traditional approach of Making Provisions for Possible Losses on Loans, Loans and Similar Debts, acting on the territory of the Russian Federation. In the research the econometric models of multiple regression of credit indebtedness from macroeconomic indicators are constructed, dot and interval forecasts of values of credit indebtedness are calculated, and also the necessary reserve volume on a covering of the corresponding credit losses is calculated. At the end of the paper the efficiency of the proposed approach to the system of reserve allocation is estimated and the rationality of its application as a quantitative method of estimating credit risk in the system of modern bank risk management is defined.