股吧和交易:股吧中的信息内容研究  

Stock BBS and Trades:The Information Content of Stock BBS

作　　者：熊熊[1,2] 罗春春 张烨 

机构地区：[1]天津大学管理与经济学部,天津300072 [2]中国社会计算研究中心,天津300072

出　　处：《系统科学与数学》2017年第12期2359-2374,共16页Journal of Systems Science and Mathematical Sciences

基　　金：国家自然科学基金(71532009;71320107003;71271145);天津市教委社会科学重大项目(2014ZD13);天津市人才发展特殊支持计划高层次创新创业团队项目资助课题

摘　　要：利用中国最大的股吧2011年1月至2014年6月的数据,研究了中国股吧与个股股票市场的关系.采用朴素贝叶斯的文本情绪分类技术,将帖子按投资意见分为"买入"、"中性"、"卖出",构造了情绪看涨指数、意见一致指数、发帖量等股吧特征变量,利用FamaMacBeth截面回归方法,研究股吧特征变量与市场特征变量包括收益率、成交量、波动性的关系,结果显示情绪看涨指数与收益率、发帖量与成交量、投资者意见分歧与波动性之间存在双向的预测作用,说明股吧包含了未反应在当前股票市场价格的信息.利用股票收益率和帖子情绪构造了衡量帖子投资意见的质量,研究股吧的信息传播机制,结果发现高质量的投资意见能够通过股吧帖子阅读量得以传播识别,但不能通过帖子的评论量被投资者识别.另外,有评论和无评论的帖子质量没有显著区别,进一步说明投资者对待不同质量帖子的关注度没有差异.The increasing popularity of ＂big data＂ research and development on finance is the common interesting point both in industrial and academic circles. The mining and analyses of online message contained therein has been generally considered one of the most important approaches on information efficiency of stock market. Online message board has become a vibrant online platform for exchanging stock- related information. This study explores the relationship between online message board and stock market, taking advantage of data derived from the biggest stock BBS in China from January 2011, to June 2014. Using text classification algorithm based on Na＇/ve Bayes, we analyze roughly 8.8 million stock-related messages on a daily basis. We find a bidirectional prediction association between stock BBS sentiment and stock returns, message volume and trading volume, as well as disagreement and volatility by Fama-Macbeth cross-section regressions. We empirically explore the mechanism behind the efficient aggregation of information in stock BBS. Specifically, we investigate the association between the quality of investment advice and the level of readers and comments. Our results demonstrate that high quality of investment advice has more readers, but not comments, which amplifies their share of voice. And there is no difference in the quality of investment advice between the comments and non-comments, which further suggest that greater weight is not given to high quality pieces of investment advice.

关 键 词：股吧 收益率 预测 帖子质量 传播机制 

分 类 号：F832.51[经济管理—金融学]