Stress-testing on Credit Risk of Commercial Bank

The paper discusses the results of research on credit risk of the commercial organization Alfa-Bank. Within the framework of the study we approximate credit risk by non-performing loans (hereafter, NPL) using dataset of macro- and micro- variables over the period 2001-2015. The variables include RUB to USD exchange rate, consumer price index, Brent oil price, unemployment, real wage rate, interest rates spread, the volume of overdue loans and the volume of the loan portfolio. Based on developed VAR models, we carry out credit risk stress testing, NPL back-test and NPL forecast for the individual and corporate loan portfolios with respect to three hypothetical scenarios. Also, we analyze the impulse-response functions and the variance decompositions. Those make possible to determine the time profile of the effect of shocks on the future NPL values. The result suggests that the main factors influencing the individual NPL are spread, unemployment and the volume of the individual loan portfolio. For the corporate NPL main factors are, RUB to USD exchange rate and the volume of the corporate loan portfolio. It was revealed that the loan portfolio growth shocks reduce the overdue loans share for both individual and corporate portfolios. Spread shocks increase individual portfolio arrears with a quarter lag and unemployment shocks – without any lags. With regards to the factors affecting the corporate portfolio arrears RUB to USD exchange rate shocks lead to an NPL increase without lags, meanwhile Brent oil price shocks - reduce with a four quarter lag. The result can be used to carry out early warning credit risk management policy in this bank.