A Stress Test Approach to the Calibration of Borrower-Based Measures: A Case Study of the Czech Republic

This paper focuses on the calibration of borrower-based measures using a semi-structural modelling framework and defines two approaches to the setting of these measures. The first approach takes into account the magnitude of losses in the mortgage portfolio and the associated absorption potential of banks, while the second, preferred approach, considers both the benefits of regulation in terms of loss reduction and its costs manifested as foregone profits. This approach thus facilitates the optimization of the macroprudential strategy to minimize Type I error (no regulation) and Type II error (excessive regulation). The case of the Czech Republic serves as an illustrative example, demonstrating that borrower-based regulation appears unnecessary and costly during periods of low credit growth, specifically in the downward phase of the credit cycle. However, if any regulation is preferred with respect to other factors and circumstances that are not captured by the modelling framework, a purely loan-to-value regulation shows the best results in terms of cost-benefit analysis.