Using credit history data to monitor financial stability of the Belarusian economy

Abstract

Data on credit histories may be very helpful for analysis of vulnerability of private sector to monitor financial stability in central banks. In National Bank of Belarus, there are credit register including vast database of credit histories of most companies and individuals on daily basis. These data may be used for estimating some useful risk measures such as Probabilities of Default (PD), Loss Given Default (LGD), etc. To construct aggregated measures from daily data special algorithms are developed. The event of default is defined according to Basel methodology. Some issues related to estimation of the risk measures are solved through analysis of the data distribution. The constructed measures can be further used for analytical reporting and statistical models estimation using supervised learning techniques