Corporate Credit Risk Modelling in the Supervisory Stress Test of the Magyar Nemzeti Bank

As a regulatory and decision-supporting tool, the stress test framework plays an important role in assessing the vulnerability of the domestic financial system and the individual institutions. Consequently, continuous development of the models used in parameter estimation is of crucial importance. This study aims to improve credit risk loss estimation, which is one of the most important components of the supervisory stress test framework, by making the estimation of corporate default and transition probability more accurate. The study is based on a client-level default database, which contains various actors in the Hungarian banking sector and covers an entire economic cycle (2007-2017). It is unique as it introduces a uniform stage classification rule for determining the transition probabilities which attempts to create harmony with domestic institutions' loan loss provision policies under IFRS 9. Based on the research findings, it can be concluded that - relying on a wide-ranging set of macroeconomic and client-level variables - it is possible to separate corporate debtors with adequate discriminatory power as well as to estimate point-in-time probability of default (PIT PD) and transition probabilities at the corporate level relevant in terms of the stress test, and thus to approximate the loan loss provisioning requirement arising in a stress situation. Of the factors capturing the cyclical nature of corporate default probability, the state of the labour market and the income position of the household sector were identified as the main determinants by the study.