Integration of physical climate risks into banks’ credit risk and capital assessment: a case study on the impact of flooding on real estate portfolios

What the paper says

Purpose This study addresses the challenges banking institutions face in integrating physical climate risks into credit risk parameters—the probability of default (PD) and loss given default (LGD)—and calculating their internal capital requirements. Design/methodology/approach We employ an extension of the Merton (1974) model to integrate the impact of physical climate risks into credit risk assessments and banks' internal capital requirements. We focus on structural physical damage affecting real estate exposure and collaterals due to river flooding in medium-term scenarios. We integrate losses from such damages into the credit risk parameters used to determine a bank's capital requirements, particularly PD and LGD. Findings The physical risk stemming from flooding impacts PD and LGD, affecting banks' internal capital requirements. However, this impact is not systematic when the damage ratio remains below the expected return on immovable assets, which is based on historical observations. Specifically, the impact becomes material when firms exhibit existing financial vulnerability in terms of indebtedness and collateral value. Practical implications Banks can use the proposed model to determine the additional credit risk and internal capital requirements stemming from climate events, particularly river flooding. Originality/value This study offers banks a methodology to integrate physical climate risks into credit risk assessments and internal capital requirements within the internal capital adequacy assessment process (ICAAP) framework. This integration is important for ensuring that banks can absorb additional credit risk due to physical losses impacting their counterparties and/or received collateral without jeopardizing their solvency, while also complying with present regulatory requirements and anticipating future ones. Moreover, it complements the existing literature by proposing a novel methodology and extending existing foundational credit risk models, enabling the aforementioned integration without relying on historical methods.