摘要
本文从因子择时视角重新探讨了中国股票市场的动量效应。为此,本文基于因子历史收益率构建了全市场因子动量策略。该策略不仅驱动了传统的动量,而且对市场异象有显著的解释能力。此外,本文从风险补偿、行为金融和投资者行为视角出发,探究因子动量的收益来源。实证结果表明,因子动量既不是对系统性风险的补偿,也不能被投资者过度自信或反应不足解释。投资者对于错误定价的缓慢修正有助于解释因子动量效应,策略在投资者情绪持续性较强、流动性较低时期及在信息不对称和卖空约束较高的股票中表现更好。本文进一步从构建窗口、指标构建等角度验证了策略的稳健性。最后,本文通过最大夏普比率检验证明因子动量为传统因子模型提供了增量信息。
This paper re-explores the momentum effect of the Chinese stock market from the per-spective of factor timing.This paper constructs a novel factor momentum strategy based on the historical performance of factor returns.Factor momentum not only subsumes momentum but also helps explain anomalies.Furthermore,this paper investigates the sources of factor momentum re-turns from three perspectives of risk compensation,behavioral finance theory,and investor be-havior.Empirical results indicate that factor momentum is neither a compensation for traditional systematic risk nor fully explained by investor overconfidence or underreaction.Instead,the slow correction of mispricing by investors contributes to the factor momentum effect.It produces stron-ger returns during periods of strong investor sentiment persistence and low market liquidity,as well as among stocks with higher aggregate information asymmetry and short-sale constraints.This paper further explores the robustness of the strategy from the perspectives of window building,in-dicator construction,and so on.Finally,the maximum Sharpe ratio test confirms that factor mo-mentum provides incremental information beyond traditional factor models.
作者
姜富伟
王月洁
刘向丽
马甜
Jiang Fuwei;Wang Yuejie;Liu Xiangli;Ma Tian(School of Economics,Xiamen University;School of Finance,Central University of Finance and Economics;School of Economics,Minzu University of China)
出处
《经济科学》
北大核心
2026年第2期36-61,共26页
Economic Science
基金
国家自然科学基金项目“财务基本面信息与金融风险预测:机器学习与经济理论”(项目编号:72072193)
国家自然科学基金项目“海量异构金融数据协同建模与机器学习”(项目编号:72342019)的支持
国家青年科学基金项目“基于完善数据的可解释深度学习模型及多市场定价研究”(项目编号:72303271)的支持。
