摘要
根据我国股市的经验事实与实证分析,发现不同于欧美股市的短期动量效应与长期反转效应,中国股市在短期与长期皆存在显著的反转效应特征;而依据市场设计理论,价格机制不可能脱离文化、习俗、制度等市场环境单独发挥作用。在此事实基础上,本文引入自谦归因偏向修正了Daniel等人(1998)模型中不符合中国投资者行为特征的假设,构建了符合中国股市的模型。仿真实验表明,在过度自信与自谦归因偏向假设下,经过短暂的过度反应后,股票收益自相关函数在短期与长期皆为负值,这与经验事实、实证研究和市场设计理论是相符的;此外,股票收益短期自相关函数绝对值,即短期反转效应水平,与投资者受自谦归因偏向影响程度呈正向关系。
According to the empirical evidence and researches of Chinese stock market,different from the short-term momentum and long-term reversal effects in European and American stock markets,Chinese stock market only has significant reversal effects in both short-term and long-term.Based on the market design theory,any price mechanism cannot play a role independently from the market environment,including culture,custom,system and so on.In this paper,we introduce the self-effacing attribution bias to modify the hypothesis in Daniel,Hirshleifer and Subrahmanyam's model(1998)that does not accord with behavior characteristics of Chinese investors,then construct a new model that conforms to the Chinese stock market.The simulation results show that under the hypothesis of overconfidence and self-effacing attribution bias,the autocorrelation function of stock returns will be negative in both short-term and long-term after a short overreaction,which consistent with empirical facts and market design theory.In addition,the absolute value of short-term autocorrelation function of stock returns,that is,the level of short-term reversal effects,has a positive relationship with the degree to which investors are affected by self-effacing attribution bias.
作者
贺京同
贺坤
赵子沐
郑为夷
He Jingtong;He Kun;Zhao Zimu;Zheng Weiyi(School of Economics,Nankai University,Tianjin 300071,China;Industrial and Commercial Bank of China,Tianjin Branch,Tianjin 300074,China)
出处
《南开经济研究》
CSSCI
北大核心
2021年第1期85-101,224,共18页
Nankai Economic Studies
基金
教育部人文社会科学研究规划基金项目“发挥有效投资对优化供给结构的关键作用研究”(19YJA790025)
中国特色社会主义经济建设协同创新中心以及国家社科基金重大项目“经济稳定增长前提下优化投资与消费的动态关系研究”(12&ZD088)的资助。
关键词
行为金融
反转效应
市场设计
自谦归因
DHS模型
Behavioral Finance
Reversal Effects
Market Design
Self-effacing Attribution
DHS Model
