Corporate Policy Information Disclosure and Analysts’ Earnings Forecast Bias

Policy arrangements exert profound impacts on corporate earnings performance and long-term development, and policy information serves as a vital reference for analysts when conducting earnings forecasts. Using a sample of A-share listed firms on the Shanghai and Shenzhen Stock Exchanges from 2007 to 2023, this paper constructs a corporate policy information disclosure index based on the BERT large language model, and investigates the impact of corporate policy information disclosure on analysts’ earnings forecast bias as well as its underlying mechanisms. The results show that corporate policy information disclosure significantly reduces analysts’ earnings forecast bias, and this conclusion remains robust after a battery of robustness tests and endogeneity treatments. Mechanism tests indicate that expanding the supply of incremental information and reducing earnings uncertainty are two core channels through which the above inhibitory effect works. Heterogeneity analyses reveal that the mitigating effect of policy information disclosure on forecast bias is more pronounced for firms where analysts face higher barriers to obtaining proprietary information, firms with higher business complexity, and firms with negative annual report tone. In addition, policy information disclosure by peer firms also curbs analysts’ earnings forecast bias, implying a significant industry spillover effect. This study provides empirical evidence for revealing the value effect of corporate policy information disclosure.