We study climate-related central bank communication using a novel dataset containing 35,487 speeches delivered by 131 central banks from 1986 to 2023. We employ natural language processing techniques to identify and trace the evolution of key climate-related narratives centred around (i) green finance, and (ii) climate-related financial risks. We find that central bank public communication strategies are primarily driven by underlying institutional factors, rather than exposure to climate-related risks. We then study the impact of climate-related communication on financial market dynamics through both a portfolio and a firm-level analysis. We find that equity returns of « green » firms outperform those of « dirty » firms when central banks engage more frequently and intensely with climate-related topics.
Keywords: Central banking; Climate change; Low-carbon transition; Central bank communication; Climate-related risks; Green finance; Text analysis; Topic modelling; Asset pricing.
JEL codes: C38, E44, E58, Q54, Z13.
