非财务信息的市场反应与预测功能——基于“京东商城”消费者大数据的经验证据

The Market Reaction and Prediction Function of Non-financial Information:Based on the Empirical Analysis of JD.com

本文运用Python数据搜索技术,以2014—2018年间消费者发表的在线评论信息为研究对象,实证检验消费者在线评论的市场反应和预测功能。研究结果显示,消费者关注度对股票收益率的影响取决于消费者评论情感,当消费者评论倾向于积极情感时,消费者关注度与股票收益率之间显著正相关,反之则反,由此可见存在“消费者关注度的情绪效应”。进一步研究通过有效甄别并排除水军言论,防止“噪声交易”导致股价明显偏离基本面,并抑制股价在长期出现反转的可能性。异质性分析表明,消费者评论的情绪效应在追加评论及竞争激烈的行业更为显著,且消极情绪下要比积极情绪下效用更强。本文从新的视角推进了非财务信息预判能力的研究,为投资者利用非财务信息提高决策质量提供重要参考,具有一定的现实意义与启示。

Using Python data search technology,this paper takes the online comment information published by consumers from 2014 to 2018 from JD.com to predict the market reaction.It also reveals that:the effect of consumer attention on stock returns depends on the emotional tendencies of consumer,when the attention of consumers tends to be positive,there is a significant positive correlation between consumer attention and stock return,and vice versa.It implies that there exists“emotional effect of consumer attention”.Moreover,the research removes paid posters comments to prevent“the noise trading”from causing stock prices to deviate significantly from fundamentals,and curbs the possibility of stock price reversal in the long run.Heterogeneity tests show that the“emotional effect”of consumer attention has a greater impact on additional comments and fierce competition groups.All in all,the study which promoting the research on the prediction ability of non-financial information from a new perspective that provides an important reference for investors to improve the quality of decision-making and has certain practical significance and enlightenment.