在线情绪对股票波动的影响及预测  

The influence and forecast of online emotion on stock volatility

作　　者：徐雨迪 Xu Yudi(Nanjing Audit University,Nanjing,Jiangsu,211815)

机构地区：[1]南京审计大学商学院,江苏南京211815

出　　处：《市场周刊》2022年第7期169-173,190,共6页Market Weekly

摘　　要：文章以沪深证券交易所的股票作为研究对象,以控制周一效应的HAR_M模型为基线模型,并将通过社交媒体数据构建的在线情绪代理引入建立新的波动率预测模型,通过比较HAR_M类模型预测精度的差异研究在线情绪对股票波动额外的预测能力。研究结果表明:来源于新闻、股吧和搜索引擎的在线情绪均包含对股票波动的预测信息,且包含的预测信息存在差异;三种在线情绪在对股票波动预测上有互补价值,三种在线情绪一起引入预测模型时,模型的预测性能最好。This paper takes the stocks of Shanghai and Shenzhen Stock Exchange as the research object,takes the HAR_M model which controls the Monday effect as the baseline model,and introduces the online sentiment proxy constructed by social media data to establish a new volatility prediction model,and studies the additional prediction ability of online sentiment to stock volatility by comparing the differences in prediction accuracy of HAR_M models.The results show that the online sentiments from news,stock bar and search engines all contain the prediction information of stock volatility,and the prediction information is different,and the three online sentiments have complementary value in the prediction of stock volatility.When the three online sentiments are introduced into the prediction model,the prediction performance of the model is the best.

关 键 词：股票波动 在线情绪 周一效应 预测价值 

分 类 号：F832.51[经济管理—金融学]