Credit risk modelling within the euro area in the COVID‐19 period: Evidence from an ICAS framework

Abstract

This paper develops a logistic regression model in an in‐house credit assessment system (ICAS) framework for predicting corporate defaults in the Greek economy. We consider the impact of the COVID‐19 pandemic and the associated government financial support schemes, aiming to protect against financial vulnerabilities, on the probability of default of non‐financial firms, as well as the relevant sectoral and firm‐size effects. In developing the ICAS framework, we address methodological issues such as the predictive performance of statistical versus machine learning approaches and the imbalanced dataset problem, indicating ways to evaluate such models with strong predictive power. Our findings suggest that the effect of the financial support measures dominates the pandemic shocks, thus substantially reducing the probability of firms' default, while the size‐ and industry‐based models show that firms in the micro and services sectors benefited the most. Furthermore, using a random forest model, our findings highlight the trade‐off between the transparency of traditional statistical models and the predictive value of machine learning models.