Do climate stress tests change lending?
Stress testing appears on 40 recent papers in the finance journals, and climate stress testing on 12. Central banks in Europe and Asia have run climate scenarios on their banks. Did banks that fared badly in the scenarios cut lending to exposed borrowers, or only write longer reports?
What prompted this
Since October 2025, 57 papers carry a stress-testing keyword (40 in finance journals), and 17 of them also carry a climate keyword; "climate stress testing" appears on 12 papers. Climate risk itself appears on 242 papers, so the stress-test exercises are a small corner of a large literature.
Questions
- After the European Central Bank's 2022 climate stress test, did banks with larger projected losses reduce loans to carbon-intensive or flood-exposed borrowers?
- Did borrowers simply switch to banks that were not tested, or to bond markets?
- Do the scenarios' long horizons and lack of capital consequences mean that banks treat them as reporting exercises?
Data that could help
Published stress-test results by bank, loan-level data where available (AnaCredit, syndicated loans), and borrowers' emissions and locations.
Next step
Compare the syndicated lending of banks with high and low projected climate losses to carbon-intensive borrowers before and after the 2022 exercise.