Analysts’ ESG attention and stock pricing efficiency: evidence from machine learning and text analysis
利用深度学习构建ESG词典，研究分析师对ESG的关注能否提升股票定价效率，发现环境和社会因素影响更大，且通过提高投资者关注和股票流动性发挥作用。
Purpose Corporate ESG performance has attracted widespread attention from various sectors of society. This paper aims to investigate whether analysts’ ESG attention can convey additional information to the market and consequently influence stock pricing efficiency. Design/methodology/approach Using A-share listed companies from 2014 to 2021 as the research subjects, this paper employs a deep learning algorithm, word2vec, to construct an ESG dictionary. Text analysis is then applied to create an analysts’ ESG attention index, delving into its impact on stock pricing efficiency. Findings Empirical research reveals that: (1) Analysts' ESG attention effectively enhances stock pricing efficiency, with a more significant impact from analysts’ attention to environmental (E) and social (S) factors compared to governance (G); (2) Further analysis indicates that this effect becomes more pronounced when there is higher disparity in corporate ESG ratings, greater marketization in the province where the company is located, and a higher institutional ownership percentage and (3) The mechanism by which analysts' ESG attention influences stock pricing efficiency is through an elevation in investor attention and stock liquidity. Additionally, it is observed that analysts prioritize ESG information to enhance their reputation and business capabilities. Originality/value From the perspective of ESG rating divergence, this paper innovatively uses analyst reports to construct ESG attention indicators and analyzes their impact on the efficiency of stock pricing.