基于投资者情绪的股票价格及成交量预测研究
STOCK MARKET PREDICTION BASED ON INVESTOR SENTIMENT
基于新浪微博2014年数据,运用情感分析技术构建情绪指数, 考察了投资者情绪与股票市场指数的相关性及其预测能力, 并通过加入百度指数平台获取的搜索指数, 考察了两种指数共同作用的影响. 研究结果表明, 搜索指数与创业板指收盘价及交易量均呈正相关, 情绪指数与创业板指收盘价及交易量均呈负相关: 情绪指数对股票价格有预测作用, 仅短期影响显著, 对股票交易量预测作用不显著;搜索指数对交易量的预测有显著作用, 但无法提升股价的预测精准度: 当两种指数同时作用时, 股价预测精准度得到较大的提升, 交易量未能达到预期的改善. 研究证明网络信息中蕴含的有效信息可以有助于预测包括股价、成交量在内的市场变量, 有助于投资者更好的利用网络中的有效信息进行金融决策行为.
To determine whether investor sentiment is correlated with stock market indicators or even predictive of financial indicators, this paper built sentiment index by sentiment analysis techniques based on Weibo posts in 2014. Then, adds search index to investigate comprehensive predictive power over financial indicators. The result shows Sentiment Index was negatively correlated with financial indicators, while Search Index was positively correlated with them; Sentiment index is predictive of the stock price, but is not a significant predictor of stock trading volume; On the contrary, search index is a significant predictor of trading volume, but cannot raise the accuracy of stock price prediction; When the two indices simultaneously work, the stock price prediction accuracy is greatly upgraded, although trading volume failed to achieve the desired improvement. Studies have shown that effective information within network can help predict financial indicators, which helps investors make better use of effective information to make financial decision-making.
/
〈 | 〉 |