Open AccessCredit Risk Modelling of Mortgage Loans in the Supervisory Stress Test of the Magyar Nemzeti Bank
Author(s) -
András Viktor Szabó
Publication year - 2022
Publication title -
financial and economic review
Language(s) - English
Resource type - Journals
eISSN - 2415-928X
pISSN - 2415-9271
DOI - 10.33893/fer.21.1.56
Subject(s) - credit risk , macro , business , database transaction , equity (law) , stress test , business cycle , econometrics , actuarial science , economics , finance , database , computer science , macroeconomics , political science , law , programming language
The study aims to develop a model that can estimate potential credit risk losses for housing and home equity loans using both macro and micro data, can be applied uniformly to all banks and takes into account the new accounting standards (IFRS 9). The model is based on a deal-level database for several Hungarian credit institutions, covering an entire business cycle (2004–2018). It uses economic indicators that strengthen risk sensitivity while also including transaction characteristics that mitigate procyclicality. Modelling in a two-step process allows risk groups to be created during forecasting in accordance with various credit characteristics. The results show that the evolution of employment has a stronger effect on riskier groups which potentially have only ad-hoc employment, while net wealth was not even among the explanatory variables for the group containing the best debtors, who presumably rely more on stable earned income.