Abstract: This paper investigates whether supervisory climate stress tests change how euro-area banks price and allocate credit according to borrowers’ transition risk. We combine loan-level data from the Eurosystem’s AnaCredit database with a firm-level measure of greenhouse-gas (GHG) intensity and compare banks subject to the ECB’s 2021-22 climate stress test with non-participating banks. A difference-in-differences-in-slopes design estimates whether the sensitivity of new-loan spreads to carbon intensity changed after the exercise. The average shift in this spread-intensity relationship is small and statistically insignificant. However, decomposing the aggregate effect reveals economically meaningful and offsetting adjustments. Participating banks reallocate new credit away from more carbon-intensive incumbent borrowers, while applying more carbon-sensitive pricing to relationships that enter or exit their portfolios; for new borrowers, a one-standard-deviation increase in carbon intensity is associated with about 10 basis points higher spreads. By contrast, repricing within continuing bank-firm relationships remains limited. The effects are strongest among the most carbon-intensive firms, and participation in the more demanding bottom-up module does not generate a clearly additional response. Overall, climate stress tests appear to operate mainly through portfolio recomposition and extensive-margin pricing, consistent with an information-and-supervisory-scrutiny channel rather than a broad repricing of existing loans.
