统计与数据科学学院知识库

自1990年沪深交易所成立，至今三十多年以来，我国股票市场迅速发展，规模不断扩大，取得了重大的成就。但是与已有几百年发展历史的西方市场相比，我国股市起步较晚，并且市场中的投资者大多数都是个人投资者，投资知识和技能不足，在参与交易活动时极易受情绪支配。同时，个人投资者相比机构投资者，接收信息的渠道有限，速度较低，这容易导致部分投资者存在盲目跟风，听从小道消息进行非理性交易的现象。此外，面对市场波动，投资者的情绪趋同，易产生羊群效应，羊群效应的存在使得股票市场的波动更加剧烈。我国股票市场暴涨暴跌起伏不定，很多异象传统的金融学理论并不能得到很好的解释，而投资者情绪理论可以弥补其不足。因此，有必要引入投资者情绪，为研究股票收益的波动以及金融异象提供一个新视角。本文主要研究内容如下：

本文首先借鉴国内外学者对投资者情绪综合指标的构建，客观指标方面选取了封闭式基金折价率、市场换手率、新股发行数量、上市首日收益、新增投资者开户数，主观指标方面选取了消费者信心指数，基于主成分分析方法构建投资者情绪指数。并且进一步通过投资者情绪指数与上证指数的走势以及两者之间的相关性来验证其构建指数的有效性。

接着将投资者情绪指数引入五因子资产定价模型，构建出包含投资者情绪的五因子资产定价模型。然后，针对整个A股股票市场和不同行业，构建分位数回归模型，在不同股票超额收益率水平下，测度了投资者情绪对股票超额收益率影响的差异性。实证结果表明：第一，就投资者情绪对股票超额收益率影响的正负而言，投资者情绪对低、中超额收益率的影响是负向的，对高超额收益率的影响是正向的。第二，就投资者情绪对股票超额收益率的影响程度而言，投资者情绪对低、高超额收益率的影响相对较大。第三，针对不同行业而言，在低超额收益率水平下，不同行业超额收益率受投资者情绪的影响差异不大，在高超额收益率水平下，投资者情绪对金融、综合和工业三个行业的股票超额收益影响较大。

最后通过脉冲响应分析不同区制下投资者情绪与股票超额收益率两者之间的联动关系。结果显示：运用马尔可夫区制转换模型可以将我国股票市场划分为三个区制：迅速下跌期、低位徘徊期以及快速上涨期；三区制状态下，投资者情绪与股票超额收益两者之间的冲击响应路径基本相似，但冲击响应程度不同：投资者情绪对股票超额收益率的正向冲击均产生显著的正向效应，股票超额收益率对投资者情绪的正向冲击均产生负向效应；投资者情绪对股票超额收益率冲击的响应程度在低位徘徊区制下最小，迅速下跌期和快速上涨期两者相差不大，响应程度较高；股票超额收益率对投资者情绪冲击的响应程度在快速上涨期最大，低位徘徊期次之，迅速下跌期最小。

总之，本文构建的投资者情绪指数符合我国股票市场的现实股市的波动情况，可以用来分析不同行业下投资者情绪对股票投资者情绪超额收益率影响的差异性；同时，对于投资者情绪与股票超额收益之间的联动分析，有助于监管部门在股票市场运行的不同阶段制定出符合当期特征的监管措施，促进我国股票市场的健康发展。

Since the establishment of Shanghai and Shenzhen Stock Exchanges in 1990, the Chinese stock market has developed rapidly, expanded continuously in scale and made great achievements. But compared with western markets with hundreds of years of development history, Chinese stock market started late, and most of the investors in the market are individual investors, investment knowledge and skills are not enough, so that they can be easily controlled by emotions. At the same time, compared with institutional investors, individual investors have limited channels to receive information at a lower speed, which leads to some investors blindly follow the trend and follow the grapevine to make irrational transactions. In addition, in the face of market fluctuations, investors' emotions converge, which is easy to produce the herd effect, which makes the stock market volatility more intense. Chinese stock market booms and tumbles, many anomalies can not be well explained by the traditional finance theory, but the investor sentiment theory can make up for its shortcomings. Therefore, it is necessary to introduce investor sentiment to provide a new perspective for studying the volatility of stock returns and financial anomalies. The main research contents of this paper are as follows:

This paper first draws on the construction of comprehensive indicators of investor sentiment by domestic and foreign scholars. In terms of objective indicators, it selects closed-end fund discount rate, market turnover rate, number of new shares issued, first-day earnings, and number of new investor accounts. In terms of subjective indicators, it selects consumer confidence index and constructs investor sentiment index based on principal component analysis method. And further through the trend of investor sentiment index and Shanghai Composite Index and the correlation between them to verify the effectiveness of the construction of the index.

Then the investor sentiment index is introduced into the five-factor asset pricing model to build a five-factor asset pricing model including investor sentiment. Then, aiming at the whole A-share stock market and different industries, the quantile regression model is built to measure the difference of the impact of investor sentiment on stock excess return under different levels of stock excess return. The empirical results show that: First, in terms of the positive and negative impact of investor sentiment on stock excess return, investor sentiment has a negative impact on low and medium excess return, and a positive impact on high excess return. Second, in terms of the impact of investor sentiment on stock excess returns, investor sentiment has a relatively large impact on low and high excess returns. Third, for different industries, under the low excess return rate, the impact of investor sentiment on the excess return rate of different industries is not significant, while under the high excess return rate, investor sentiment has a great impact on the excess return of stocks of finance, comprehensive and industrial industries.

Finally, impulse response is used to analyze the linkage relationship between investor sentiment and stock excess return under different regional systems.The results show that the Chinese stock market can be divided into three regional systems by using the Markov zone conversion model: rapid decline period, low lingering period and rapid rise period. Under the three-zone system, the impact response paths between investor sentiment and stock excess return are basically similar, but the impact response degrees are different: investor sentiment has a significant positive effect on the positive impact of stock excess return, and stock excess return has a negative effect on the positive impact of investor sentiment. The response degree of investor sentiment to the shock of stock excess return is the smallest under the low lingering zone system, and there is no difference between the rapid decline period and the rapid rise period, and the response degree is high. The response degree of stock excess return to investor sentiment impact is greatest in the period of rapid rise, followed by the period of low hover, and minimum in the period of rapid fall.

In conclusion, the investor sentiment index constructed in this paper conforms to the reality of the stock market fluctuations, and can be used to analyze the differences of the influence of investor sentiment on the excess return of stock investor sentiment in different industries. At the same time, for the linkage analysis between investor sentiment and stock excess returns, it will help the supervision departments to formulate supervision measures in the different stages of the stock market operation, which will promote the healthy development of the stock market in our country.