投资者情绪与股市收益相关性的实证研究

发布时间：2018-03-09 19:47

本文选题：投资者情绪　切入点：股市收益　出处：《兰州大学》2017年硕士论文　论文类型：学位论文

【摘要】：近年来,随着行为金融学对多种金融异象的研究,其对经典的金融理论提出了一定程度上的质疑。中国股票市场起步晚、发展快,时常发生剧烈波动,投资者不理性、不成熟的特征也十分明显。本文针对中国当前股票市场的现状,展开了关于投资者情绪与股票市场收益联动性的实证研究。本文将投资者情绪定义为投资者自身对于金融市场上未来产品价格的带有一定程度偏差的投资者预期,并把现有的投资者情绪衡量指标,按照获取数据时的性质和方式的差异划分为三类:显性指标、隐性指标和情绪代理变量,并最终选定中国证券投资者保护基金有限责任公司发布的月度投资者信心指数作为考量指标。结果表明,投资者信心指数变化率曲线与上证指数收益率曲线及创业板指数收益率曲线之间表现出协同变化的趋势,而且相对于创业板指数收益率曲线,投资者信心指数变化率曲线与上证指数收益率曲线间的协同变化更加明显。接着,本文构建了滞后期为1期的向量自回归模型(VAR)来对投资者情绪与股市收益之间的动态关联展开深入分析。通过脉冲响应分析和方差分解后,笔者由格兰杰因果检验得知投资者信心指数变化率与上证指数收益率以及创业板指数收益率之间互为因果。然后运用Engle和Granger两步法建立误差修正模型,对结论再次进行验证。在确定投资者情绪与股市走向之间存在关联性的基础上,将时间跨度纳入考量,利用GARCH模型和TARCH模型实证检验得知股市收益的波动具有杠杆效应,利空消息会比等量的利好消息产生更大的波动。最后,针对价格波动的非对称性,本文分别采用TARCH、EGARCH和CARCH模型,研究了市场不同阶段中利好与利空消息的冲击效应。总体而言,笔者通过金融建模得出了本文的研究成果。但是,由于投资者结构的变动性和衡量指标的局限性,研究结论未必能完全准确的反映当前的市场状态。下一步的研究中,可结合投资者个体差异和市场运行趋势,纳入多维度因素来编制综合性的衡量指标,以便于量化投资者情绪的温度,为证券监管机构和市场参与者提供市场波动预判及风险预警的参考。
[Abstract]:In recent years, with the study of various financial anomalies in behavioral finance, the classical financial theory has been questioned to a certain extent. The stock market in China starts late, develops rapidly, often fluctuates violently, and investors are irrational. The characteristics of immaturity are also obvious. This paper aims at the current situation of stock market in China. An empirical study on the linkage between investor sentiment and stock market returns is carried out. In this paper, investor sentiment is defined as the investor's own expectations of the future product price in the financial market with a certain degree of deviation. And according to the difference of the nature and the way of obtaining the data, the existing investor sentiment measurement index can be divided into three categories: dominant index, implicit index and emotion proxy variable. Finally, the monthly investor confidence index issued by China Securities Investor Protection Fund Limited is selected as the index. The results show that. The variation rate curve of investor confidence index, the yield curve of Shanghai Stock Exchange index and the yield curve of growth enterprise board index show the trend of synergistic change, and compared with the yield curve of growth enterprise market index, the change rate curve of investor confidence index shows the trend of synergistic change. The synergistic change between the rate of change curve of investor confidence index and the yield curve of Shanghai stock index is more obvious. In this paper, a vector autoregressive model with a lag of 1 period is constructed to analyze the dynamic correlation between investor sentiment and stock market returns. After impulse response analysis and variance decomposition, By Granger causality test, the author finds out that the rate of change of investor confidence index is causality with the rate of return of Shanghai Stock Exchange Index and the rate of return of growth Enterprise Market Index. Then the error correction model is established by using Engle and Granger two-step method. The conclusion is verified again. On the basis of determining the correlation between investor sentiment and the trend of stock market, the time span is taken into account, and the empirical test of GARCH model and TARCH model shows that the volatility of stock market returns has leverage effect. Good news has more volatility than good news. Finally, aiming at the asymmetry of price volatility, this paper studies the impact effect of good and bad news in different stages of the market by using TARCHUE EGARCH and CARCH models, respectively. However, due to the volatility of investor structure and the limitation of measurement index, the conclusion of the study may not be able to reflect the current market state exactly. Combining individual investor differences and market operating trends, we can incorporate multi-dimensional factors to compile comprehensive metrics to quantify the temperature of investor sentiment. For securities regulators and market participants to provide market volatility prediction and risk early warning reference.
【学位授予单位】：兰州大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：F832.51

【参考文献】