Determining the maximum loss of loan portfolios I.

The Vasicek model with one financial product and with one independent variable

Abstract

Credit losses arise from a lack of knowledge of the future state of influencing factors, and these factors are classified in the literature as idiosyncratic causes, expressing internal characteristics, and external, so-called macro causes. Even in the presence of favorable external conditions, the default rate of loans can be estimated at 2-4\%, for which one can be prepared. However, adverse macro conditions may exist in the external environment, the occurrence of which causes an increase in the rate of failures. Therefore, it is important to analyse what macro factors, and to what extent, cause the change in the rate of failures. The use of the Vasicek model is well known in the financial sphere: the Vasicek model aims to determine the probability of default of a single type of credit product and condenses external, so-called macro effects into a single explanatory variable. This study concludes that the Vasicek model's estimation of the probability of failure is very sensitive to input parameters, which are already numerous and often difficult to measure. It is also important to note that the model assumes relationships that are difficult to prove. While retaining the values of the model, this study basically launches a family of models extended in two directions, the basic goal of which was to propose a useful methodology that is easy to understand (e.g. summarized in a figure) for a bank's senior management or supervisory body, where inputs are as clear and measurable as possible. We intend to continue this study, in which we simultaneously determine the magnitude of maximum losses for several credit products, when several external (macro) factors can be used as explanatory variables.