A text-mining analysis of climate risk disclosure and corporate energy efficiency in Chinese listed firms

This study examines the association between climate risk disclosure and corporate energy efficiency using Chinese A-share data from 2012 to 2024. A firm-level disclosure index is constructed via text mining based on the TCFD framework. Two-way fixed effects models show a positive association, with a one-standard-deviation increase in disclosure corresponding to approximately a 3.6% increase in energy efficiency, robust to endogeneity tests. Transition risk disclosure has a stronger association with energy efficiency than physical risk; the difference widens further when standardized by each dimension’s dispersion. Mediation analysis indicates green innovation and eased financing constraints are partial channels whose separately estimated indirect effects sum to approximately 20% of the total effect, though the direct effect dominates. Heterogeneity analysis reveals stronger effects in state-owned, heavily polluting, and carbon pilot region firms, with insignificant effects outside pilot regions; the association also strengthens significantly after the launch of the national carbon market in 2021. The study contributes to the literature by shifting the focus from capital market reactions to real operational outcomes, distinguishing between physical and transition risk disclosure, and highlighting the role of institutional context in determining whether climate disclosure produces substantive effects on energy performance.