基于朴素贝叶斯算法的投资者情绪对于股票收益率的影响

The Impact of Investor Sentiment on Stock Returns Based on the Naive Bayes Algorithm

马科维茨的投资组合理论假设投资者是理性的,然而,最近几年股票市场出现了剧烈的波动,这种波动很难用宏观经济因素或公司自身情况来解释,行为金融学应运而生,投资者情绪在股市中的作用也备受关注。随着中国股市的发展,投资者的情绪很容易对股票市场产生影响,从而导致股票收益的波动,鉴于投资者情绪状态无法直接观测到,对其如何进行度量一直是一个难题。本文在传统金融学的基础上引入了行为金融理论,考虑了市场因素和公司基本面因素,并加入了投资者情绪因子,用以解释股票收益率的波动。在实证方面,研究使用了沪深300指数成分股中300个企业的数据,并使用朴素贝叶斯分类算法构建了投资者情绪指标。通过个股技术指标来衡量投资者情绪,填补了沪深300成分股中个股情绪面板数据的空白。同时,还构建了多元线性模型进行回归,分析了投资者情绪对股票收益率的影响。结果表明,随着投资者情绪得分的提高,股票的回报率也会随之上升,投资者情绪的变化在一定的程度上会对股票市场的投资收益产生影响,更清晰体现出股市中投资者情绪对股票风险的影响机理,这对个人和机构投资者在选股时具有重要意义。

Harry Markowitz’s portfolio theory assumes that investors are rational;however, the stock market has experienced dramatic volatility in recent years, which is difficult to explain in terms of macroeconomic factors or a company’s own circumstances. Behavioral finance has emerged, and the role of investor sentiment in the stock market has received much attention. With the development of China’s stock market, investors’ emotions can easily have an impact on the stock market, which leads to the fluctuation of stock returns, and given that investors’ emotional state cannot be directly observed, it has been a difficult problem to measure how it is measured. This paper introduces behavioral finance theory based on traditional finance, considers market factors and firm fundamentals, and adds an investor sentiment factor to explain the volatility of stock returns. In terms of empirical evidence, the study uses data from 300 companies in the constituent stocks of the CSI 300 index and constructs an investor sentiment indicator using a simple Bayesian classification algorithm. By measuring investor sentiment through technical indicators of individual stocks, it fills the gap of panel data of individual stock sentiment in CSI 300 constituent stocks. A multivariate linear model is also constructed for regression to analyze the impact of investor sentiment on stock returns. The results show that as the score of investor sentiment increases, the return of stocks will also increase, and the change of investor sentiment will have an impact on the investment return of the stock market to a certain extent, and these conclusions reflect more clearly the mechanism of the impact of investor sentiment on stock risk in the stock market, which is of great significance to individual and institutional investors in stock selection.