Working Paper Series no. 1048. We construct a firm-bank level panel of French single-establishment micro and SME firms from 2010 to 2023 to investigate the differential impact of chronic versus acute heat events on banks’ loan supply. We show that both chronic temperature increases and acute heat days significantly reduce loan growth, primarily through medium- and long-term lending, but that banks respond differently to these two types of climate risk. To account for demand-side effects, we investigate sectoral dynamics and uncover pronounced maturity and sectoral heterogeneity. While the trade, transport, accommodations, leisure,manufacturing and mining sectors are affected by both types of shocks, sectors such as real estate, construction, services, and ICT are more strongly impacted by acute heat shocks. Our results are robust to alternative exposure measures and to the inclusion of credit ratings, highlighting that temperature exposure constitutes an additional source of risk not yet captured by standard bank credit assessment models.
Impact of temperature risk on credit growth to French firms
Non-Technical Summary
Physical climate risk is becoming increasingly salient for financial intermediation because it already affects firms’ operating conditions, asset values, and borrowing capacity. Among its many dimensions, temperature is particularly important. It captures both gradual warming and the increasing incidence of extreme heat, two manifestations of climate change that may affect firms through distinct channels and over different horizons. Yet relatively little is known about how banks adjust credit supply in response to these temperature-related risks, especially at a granular level.
This paper examines how banks’ lending to French single-establishment micro, small, and medium-sized enterprises (SMEs), measured using the Banque de France Fichier Central des Risques, responds to two distinct dimensions of temperature exposure: chronic warming and acute heat events. To do so, we combine station-level temperature observations from Météo- France with geocoded firm information to construct establishment level measures of local temperature exposure, providing a substantially finer geographic resolution than that used in the existing literature.
We document three main results. First, both chronic warming and acute heat exposure significantly reduce loan growth at high levels of temperature exposure, with effects concentrated in medium- and long-term credit. As illustrated in the Figure below, however, the dynamics differ markedly across the two dimensions of risk. Acute heat shocks generate sharp but transitory contractions in lending, persisting for up to five years at the 90th percentile of the exposure distribution. In contrast, chronic warming is associated with more persistent lending adjustments, lasting up to eight years at the same percentile, consistent with a longer-horizon reassessment of firm risk. Because chronic and acute temperature shocks are, on average, unevenly distributed across regions, the results suggest that these effects may also translate into geographically differentiated acess to credit. Second, the aggregate response masks substantial heterogeneity across industries. Trade and Transport, Accommodation and Leisure, as well as Manufacturing and Mining are vulnerable to both chronic warming and acute heat stress, whereas sectors such as Real Estate and Construction, Utilities, and Services and ICT respond more strongly to acute heat events. These patterns are consistent with heterogeneous transmission channels linking temperature exposure to firm performance and, ultimately, to bank lending. They also suggest that the observed contraction in credit largely reflects a supply-side adjustment. In this respect, our findings complement recent evidence that extreme temperatures reduce SME access to credit in other settings as well (Aguilar-Gomez et al., 2024). Third, the response to temperature exposure is not uniformly monotone. In some sectors, heightened temperature risk leads banks to shorten maturities rather than withdraw credit altogether, implying a shift toward short-term lending that partly offsets the contraction in medium- and long-term credit.
Overall, our results show that temperature-related climate risk already influences bank lending in economically meaningful ways. The effects differ across chronic and acute exposures, across sectors, and across loan maturities, and they persist over multi-year horizons. For policymakers and supervisors, these findings underscore the importance of climate-risk frameworks that are granular, sector-specific, and attentive to the maturity structure of lending. More broadly, they highlight the need to incorporate forward-looking climate-risk assessments into credit evaluation and portfolio monitoring practices.
Keywords: Bank Loans, Climate Change, Environmental Risks, Credit Register, Acute and Chronic Heat Stress, Firms
Codes JEL : C23, D22, G21, L25, Q54
Download the full publication
Updated on the 18th of June 2026