文献类型：期刊文献

中文题名：情绪传染对股价联动的影响——来自股吧大数据的证据

英文题名：Research on the Impact of Sentiment Contagion on Stock Price Co-movement:Evidence from Big Data of Stock Forum

作者：易苗苗[1];任飞[1];何令杰[1];熊熊[2]

机构：[1]华东理工大学商学院,上海200237;[2]天津大学管理与经济学部,天津300072

年份：2024

卷号：37

期号：4

起止页码：147

中文期刊名：管理科学

外文期刊名：Journal of Management Science

收录：;北大核心:【北大核心2023】；CSSCI:【CSSCI2023_2024】；

基金：国家自然科学基金(71871094,72141304,71790594)。

语种：中文

中文关键词：投资者情绪;情绪传染;股价联动;股吧;大数据

外文关键词：investor sentiment;sentiment contagion;stock price co-movement;stock forum;big data

摘要：互联网社交平台已然成为投资者信息交互的重要媒介,以信息为载体的情绪表达作为投资者制定决策的重要依据,在股票定价过程中发挥着重要作用。已有研究证明了投资者情绪对单只股票价格的影响,却很少讨论投资者间的情绪传染是否会对股价的关联关系产生影响,也鲜有研究对投资者间的情绪传染进行直接测量。采用2018年7月至2021年7月上证A股东方财富股吧用户发帖和回帖数据,通过大数据挖掘、处理技术以及自然语言分析技术计算出投资者情绪,基于活跃在不同股吧中的投资者帖子数据和相关性分析方法,设计两两股吧间的投资者情绪传染指标,讨论情绪传染对股价联动的影响,以及这一影响在不同情绪倾向、不同热度帖子和同行业与非同行业股票样本中是否存在差异,并进一步讨论剔除市场因素后,投资者情绪传染对股价超额收益联动的影响,以及滞后期的情绪传染对股价联动的预测作用。研究结果表明,投资者情绪传染对当期股价联动具有显著的正向影响,经过一系列稳健性和内生性检验后结果依然成立。异质性分析表明,当积极帖子占比越高、帖子热度越高和两只股票属于同一行业时,投资者情绪传染对股价联动的影响更大。进一步研究表明,投资者情绪传染对个股间超额收益率联动有显著正向影响,此外,滞后一期的情绪传染对股价联动具有预测作用。研究结果丰富了基于情绪因素的股价联动成因研究,将视角从情绪对单只股票的影响延伸到股价联动的研究上,为深入理解资产定价过程和分析股市风险传染机制提供新的思路。情绪传染对股价联动的预测作用为投资者和监管机构提供了情绪应用的参考,投资者可以通过观察和分析情绪传染制定投资策略实现风险规避或投资收益,监管机构可以通过监测投资者的情绪水平和传染速度及时发现市场异常波动,采取加大对散播虚假消息等行为的惩罚力度、督促上市公司信息披露等政策措施维护市场稳定。
Social media platforms have become important medium for investors to exchange information.The sentiment contained in the information is a key factor for investors to make investment decisions,which plays an important role in stock pricing.Existing studies provide evidences of the impact of investor sentiment on the price of a single stock.However,there is by far limited research that directly measures the sentiment contagion between investors,let alone studies of its influence on the correlation of stock prices.Using data on posts and replies on the East Money Stock Forum for A-share stocks listed on SSE from July 2018 to July2021,we first measure investor sentiment based on big data mining and natural language processing algorithms.And we construct an indicator of investor sentiment contagion between two stocks by using the correlation analysis method based on the measure of stocks′investor sentiment.Then we explore the impact of investor sentiment contagion on the stock price co-movement and examine whether this impact would vary across samples with different sentiment tendencies,different levels of post popularity and different industries.Furthermore,investigate the impact of investor sentiment contagion on the excess co-movement and the prediction effect of investor sentiment contagion on the stock price co-movement.The research results show that investor sentiment contagion has a significant positive effect on the stock price co-movement in the current period,and this finding remains hold after we conduct a series of robustness and endogeneity tests.The heterogeneity analysis shows that the impact of sentiment contagion on the co-movement is more pronounced when there are higher proportion of positive posts,greater popularity of posts,or belonging to the same industry.This study further indicates that the influence of investor sentiment contagion on the excess co-movement is positive significant,and lagged investor sentiment contagion can well predict the stock price co-movement.This study supplements the research that explores the driving forces of stock price co-movement in terms of investor sentiment and extends these studies by studying the impact of investor sentiment on stock price co-movement rather than on a single stock,offering new evidence for a deeper understanding of asset pricing and a thorough analysis of the risk contagion mechanism in the stock market.The prediction effect of investor sentiment contagion on stock price co-movement provides references for investors and regulators to properly utilize sentiment analysis.That is,investors could set up trading strategies based on the observation and analysis of the spread of sentiment contagion on stock forums to avoid risks or to make better investment returns.And regulators could monitor the levels of investor sentiment and the spread of sentiment contagion to detect market fluctuations in a timely manner,and take measures such as heavier penalties for spreading fake information and urging the corresponding firms to release relevant information to ensure stock market stability.