The Role of Policy on Commodity Markets: From the Perspective of Narrative News Based on NLP

ABSTRACT Using a cutting‐edge natural language processing method, this study constructs the China narrative policy index (CNPI) to capture policy content, drawing on 10 authoritative government reports and official media sources and incorporating expert input to enhance thematic precision and relevance. On this basis, the study employs a ripple‐spreading network model to examine CNPI's impact on global commodity futures markets, revealing China's position in policy‐driven spillovers across both temporal and spatial dimensions. Through case analyses of the supply‐side reform and four sub‐policy themes, we validate CNPI's effectiveness in capturing policy content and highlight China's influence as a global policy transmitter.