Credit Risk Theoretical Model on the Base of DCC-GARCH in Time-Varying Parameters Framework

The research paper is devoted to developing a mathematical approach for dealing with time-varying parameters in rolling window logit models credit risk assessment. Forecasting coefficients yields better model accuracy than trivial of using computed past statistics the next time period. In this paper, new method scoring proposed, which aimed at computing default probability borrower. It was empirically shown that continuously changing economic environment factors’ influence on target variable also changing. Therefore, forecasting financial result simply applying obtained by accumulated over periods. develops theoretical approach, incorporating combination ARIMA class model, DCC-GARCH and state–space more accurate, only model. Rigorous simulation testing provided confirm efficiency proposed method.