【作者】 张萌；

【导师】 张永杰；

【作者基本信息】 天津大学 ， 金融硕士（专业学位）， 2019， 硕士

【摘要】 Bali等研究美国1926年至2005年股票市场数据发现了一个市场异象,即最大日收益率效应(MAX效应),当月最大日收益率高(低)的股票,下个月的预期收益率更低(高)。本文主要研究MAX效应、MAX效应与投资者情绪和套利限制的关系。首先,本文研究2003年至2018年中国股票市场A股的数据,将股票根据上个月最大日收益率(MAX)和上个月五日平均最大日收益率(MAX5)排序分组以及对应的组合平均加权的收益率,以验证MAX效应在我国股票市场的存在性。然后从双因素二重分组的投资组合角度检验了在控制了公司规模,账面市值比,动量和短期反转后最大日收益和股票未来收益的关系。最后使用公司层面的Fama-Mac Beth回归来检验横截面上MAX和未来收益的关系。在研究MAX效应与投资者情绪关系中,本文根据中国投资者情绪指数(ISI)为基础,剔除宏观因素构造剔除了宏观因素的中国投资者情绪指数(EMISI)来衡量中国投资者情绪,根据剔除了宏观因素的中国投资者情绪指数(EMISI)对样本期间进行区分高情绪和低情绪区间,检验不同情绪区间下MAX投资组合的投资收益情况,以验证中国市场MAX效应和投资者情绪有关。在研究MAX效应与套利限制的关系中,本文主要参照Gu等在研究中国市场特质波动率效应与套利限制的关系中对套利限制的描述来构建套利限制指数以刻画每个股票不同的套利限制。然后采用双重分组的研究方法,分为低套利限制,高套利限制两组,以得到套利限制对MAX效应的影响。经过研究发现,在中国A股市场存在明显的MAX效应,存在MAX差额月收益为1.1%(t=4.07)。MAX效应与投资者情绪关系密切,在高投资者情绪期间,MAX差额月收益明显好于无条件MAX分组,高情绪时期MAX差额月收益为1.43%(t=3.68),而低情绪时期MAX差额月收益为0.78%(t=2.04),MAX差额月收益明显差于无条件MAX分组。套利限制因素影响MAX效应,在高套利限制的投资组合中,MAX差额月收益更为显著,为1.61%(t=5.73),而在低套利限制的投资组合中,MAX差额月收益较小,为0.87%(t=2.89)。更多还原

【Abstract】 Bali et al.studied the stock market data of the United States from 1926 to 2005 and found a market anomaly,that is,the Maximum daily yield effect(MAX effect).The stocks with the highest(lowest)daily yield in the current month and the expected return in the next month are lower(higher).This paper mainly studies the relationship MAX effect,between MAX effect,investor sentiment and arbitrage limit.the Maximum daily yield effect(MAX effect).The stocks with the highest(lowest)daily yield in the current month and the expected return in the next month are lower(higher).First of all,this paper studies the data of A-share in China’s stock market from 2003 to 2018.According to the Maximum daily yield(MAX)of last month and the average Maximum five-daily yield(MAX5)of last month,ten portfolios and the corresponding weighted average portfolio return are grouped to verify the existence of MAX effect in China’s stock market.Then,we test the relationship between the Maximum daily return and stock future return after controlling the company size,book to market ratio,momentum and short-term reversal from the perspective of dual factor and dual grouping portfolio.Finally,Fama-Macbeth regression at the company level is used to test the relationship between MAX and future return on the cross section.In the study of the relationship between MAX effect and investor sentiment.This paper uses the Chinese investor sentiment index(ISI)as basics,and constructs the excluding macro Chinese investor sentiment index(EMISI)to measure Chinese investor sentiment.According to the excluding macro Chinese investor sentiment index(EMISI),the sample period is divided into high sentiment and low sentiment terms,and the investment of MAX portfolio under different sentiment intervals is tested Earnings to verify that China’s market MAX effect is related to investor sentiment.In the study of the relationship between MAX effect and arbitrage restriction,this paper mainly refers to the description of arbitrage restriction by Gu et al.in the study of the relationship between China’s idiosyncratic volatility anomalies and arbitrage limit to construct an arbitrage restriction index to describe the different arbitrage restrictions of each stock.Then,using the double grouping method,we divided the two groups into low arbitrage limit and high arbitrage limit,in order to get the influence of arbitrage limit on MAX effect.It is found that there is an obvious MAX effect in China’s A-share market,and there is a MAX difference with a monthly return of 1.1%(t = 4.07).MAX effect is closely related to investor sentiment.In the period of high investor sentiment,the monthly return of MAX difference is significantly better than that of unconditional MAX group.In the period of high investor sentiment,the monthly return of MAX difference is 1.43%(t=3.68),while in the period of low investor sentiment,the monthly return of MAX difference is 0.78%(t=2.04).The monthly return of MAX difference is significantly worse than that of unconditional MAX group.In the portfolio with high arbitrage limit,the monthly return of MAX difference is more significant,which is 1.61%(t = 5.73),while in the portfolio with low arbitrage limit,the monthly return of MAX difference is smaller,which is 0.87%(t = 2.89).更多还原