Measuring climate risk with ChatGPT: the influence of firms market value within supply chains in China

Abstract

ABSTRACT This study examines the influence of climate risk on upstream and downstream firms within supply chains using Chat Generative Pre-trained Transformer (ChatGPT). Climate risk not only affects corporate asset pricing but also propagates through interconnected supply chains, amplifying its impact on firms’ financial performance and market value. The Chat Generative Pre-trained Transformer 4th generation optimised mini (ChatGPT-4o-mini) model measures firms’ climate risk levels, and the consistency and validity of this metric are verified against multiple external benchmarks. Empirical analysis based on two datasets reveals that corporate climate risks reduce the market value of upstream and downstream partners, with the magnitude of this impact varying by ownership structure, size, market power, and geographical proximity. This study provides empirical evidence and theoretical support for quantifying firm-level climate risk, mitigating climate risk shocks in supply chains, and constructing secure and resilient supply chain networks.