Climate risk stress testing: A critical survey and classification

Abstract

We conceptually investigate Climate Risk Stress Testing (CRST) exercises to assess the impact of climate-related shocks on financial system stability. We distinguish between climate, economic, and financial modeling steps, and classify CRST exercises into six types of climate shocks and four different approaches (macro-financial, micro-financial, non-structural, and disaster risk). We identify several key limitations in current CRST approaches: (i) neglect of certain climate shock types (Green Swan and Minsky-type events); (ii) overreliance on macro models (with low sectoral and spatial granularity); (iii) incomplete modeling (lack of feedback effects); and (iv) limited scope (subset of causal channels and asset classes). We argue that these limitations may lead to significant underestimation of potential system-wide financial losses and offer suggestions for improving CRST approaches. They have also led CRST exercises to diverge from the traditional stress testing objective of capital adequacy.