Impact of the Industries, Behavioral Information, and Covid-19 Measures on the Non-financial Corporates Default Rates

In this paper, the author examines an impact of behavioral, macro-economic, and Covid-19 measures on the default rates on a client level for sub-portfolios of clients from six industries with the highest number of clients in the Czech banking system. As the model is based upon individual client data over the period from 2005Q2 until 2021Q2 (plus 12-month observation period for default identification), the generalized linear model with logit link function is employed. The detailed data allow to consider, beyond the commonly used macro-economic variables, the information about past client repayment behavior, size of their exposures, and the information about using client specific and banking related Covid-19 support measures (postponements of repayments and state guarantees). The results confirm macro-economic determinates (especially unemployment and inflation measures) of defaults rates available by the others authors’ studies. The results also identify the strong statistically significant differences among sub-portfolios, defined by client’s previous repayment history (leading default indicator for probability of default within the next 12 months), industry (for six industries with the highest number of observations), and assesses the effect of Covid-19 measures.