学位论文独创性声明
本人郑重声明：本人所呈交的学位论文，是在导师的指导下，独立进行研究
所取得的成果。除文中已经注明引用的内容外，本论文不含任何其他个人或集体
已经发表或撰写的作品，也不包含为获得安徽财经大学或其他教育机构的学位或
证书所使用过的材料。对本文的研究做出重要贡献的个人和集体，均已在文中标
明并表示了谢意。
本声明的法律后果由本人承担。
论文作者（签名）：
2022 年 6 月 30 日
学位论文使用授权书
本论文作者完全了解学校关于保存、使用学位论文的管理办法及规定，即学
校有权保留并向国家有关部门或机构送交论文的复印件和电子版，允许论文被查
阅和借阅。本人授权安徽财经大学可以将本学位论文的全部或部分内容编入学校
有关数据库和授权学校研究生处与中国知网和万方数据签订收录协议及收录并
由作者本人享有、承担相应的权利和义务，也可以采用影印、缩印或扫描等复制
手段保存或汇编本学位论文。
注：保密学位论文，在解密后适用于本授权书。
作者签名：
2022 年 6 月 30 日

摘要
摘要
2019年底，“新冠肺炎”(COVID-19)疫情爆发，由于西方各国的消极抗疫态
度，新冠疫情蔓延之势愈来愈猛烈，逐渐成为了当今世界经济格局的首要不稳定
因素。世界各国股票市场股价暴涨暴跌发生的频率相较疫情之前大幅增加，而在
这背后是否是由于疫情带来的实体经济活动受限从而使得大量个人投资者携带
资金入市从而加剧了投资者情绪波动对股价崩盘风险的影响值得我们探究，相反,
股价的持续暴跌是否会使投资者对未来股市失去看涨预期,进而产生更加悲观的
判断?最终造成更为严重的股市崩盘乃至实体经济的衰退。因此,在新冠疫情视角
下研究投资者情绪与股价崩盘具有实际意义。
本文选取了A股市场股票收益率作为研究对象,选取了从2003年到2021年
共18年的时间序列数据,并以此构建了结构向量自回归模型( Structural Vector
Autoregressive Model, SVAR),并在结构向量自回归模型的基础上分别分析了新
冠疫情前和新冠疫情下股票崩盘风险与投资者情绪之间的交互动态影响。实证结
果表明投资者情绪对股价崩盘风险的影响不论在疫情前还是在疫情当下在时间
上产生的影响都不是瞬时发生的而是存在一段时间的滞后，在持续性方面，疫情
下，两者之间互相影响持续的时间更长，且投资者情绪对股价崩盘风险在股价下
跌区间会有更加显著的羊群效应的。研究表明最大的区别在于股价崩盘对投资者
情绪脉冲响应的第一期到第二期的时间点在新冠疫情下是逐渐下降的，而在新冠
疫情下，这两者的脉冲响应是逐渐上升的，说明在新冠疫情下投资者情绪对股价
崩盘风险的影响的衰退期会延后，造成的影响会更加深远。文章进一步通过对中
国资本市场羊群效应进行实证分析，实证结果表明羊群效应在中国股票市场上存
在非对称效应即在市场大幅度上涨时，羊群效应并不显著，而在市场出现大幅下
跌时，出现了显著的羊群效应，且实证结果表明，在疫情下，当股市大幅下跌时，
羊群效应在1%的水平上相较于疫情前5%水平上更加显著，表明非理性的羊群效
应是投资者情绪与股价崩盘间重要的中介渠道。最后，文章在实证结果的基础上
给出了对应的政策建议。
关键词：股价崩盘风险;投资者情绪;羊群效应;新冠疫情
1

Abstract
Abstract
At the end of 2019, the “COVID-19” epidemic broke out. Due to the negative
anti-epidemic attitudes of Western countries, the spread of the COVID-19 has become
more and more violent, and it has gradually become the primary destabilizing factor
in today’s world economic structure. The frequency of stock market surges and
plunges in stock markets around the world has increased significantly compared to
before the epidemic, and whether this is because the real economic activities brought
about by the epidemic have been restricted, which has caused a large number of
individual investors to carry funds in to market, thereby exacerbating investor
sentiment fluctuations. The impact of the risk of stock price collapse is worthy of our
investigation. On the contrary, will the continuous plunge of stock prices make
investors lose their bullish expectations for the future stock market, and thus produce
more pessimistic judgments? Eventually, it will cause a more serious stock market
crash and even the recession of the real economy. Therefore, it is of practical
significance to study investor sentiment and stock price collapse from the perspective
of the COVID-19 epidemic.
This paper selects the stock returns of the A-share market as the research object,
selects 18 years of time series data from 2003 to 2021, and constructs a Structural
Vector Autoregressive Model (SVAR) based on this, and uses it in the Based on the
structural vector autoregressive model, the interactive dynamic effects of stock crash
risk and investor sentiment under the COVID-19 epidemic and under the COVID-19
epidemic were analyzed respectively. Empirical results show that the impact of
investor sentiment on stock price crash risk, both before the epidemic and at the time
of the epidemic, is not instantaneous but lags for a period of time. The mutual
influence between them lasts longer, and investor sentiment will have a more
significant herd effect on the stock price crash risk in the falling stock price range.
The research shows that the biggest difference is that the time points from the first
period to the second period of the impulse response of stock price crash to investor
sentiment are gradually declining under the COVID-19 epidemic, while under the
COVID-19 epidemic, the impulse responses of the two are gradually rising. ,
I
。。。以下略