摘要
提出一种衡量股价因子有效性的三维度检验方法,认为有效因子应同时具有超额收益解释力、解释能力显著性及经济学逻辑性.基于2016年—2019年我国沪深300指数成分股相关数据,通过评分法、回归法、经验法的复合有效因子检验法筛选出各截面期有效因子,并建立基于有效因子的多因子模型.结论认为:每股收益、行业市盈率等在内的质量、估值类因子能够有效解释沪深300指数成分股的超额收益;所构建的多因子模型在回测期间表现远超基准收益率;多因子模型在我国资本市场仍能够获得超额收益,具有较强的普适性.
Traditional factor selections when forming a multi-factor model are often based on single-dimensional tests such as correlation coefficient method (IC,IR value,etc.),and therefore lead to poor excess return prediction ability and model failure. On this basis,the paper proposes a multi-dimensional test method to measure the effectiveness of stock factors,and suggests that effective factors should have the following characteristics:explanatory capability,explanatory significance and economic-based logic. Based on the relevant data of China s CSI 300 index from 2016 to 2019,effective factors are tested and selected through the composite effective factor test method of scoring method,regression method and empirical method,and a multi-factor model is built based on the effective factors. Results suggest that:Quality and valuation factors such as EPS,Industry P/E ratio can effectively explain and predict future excess return of CSI 300 s constituent stocks;The multi-factor model outperforms CSI 300 during back-test period,indicating its capability and universality in China stock market.
作者
赵子铭
赵文声
ZHAO Zi-ming;ZHAO Wen-sheng(School of Mathematical Science,South China Normal University,Guangzhou Guangdong 510000,China;Road and Bridge Planning Research Center of Guangdong Province,Guangzhou Guangdong 510000,China)
出处
《广州航海学院学报》
2019年第3期74-78,共5页
Journal of Guangzhou Maritime University
关键词
多因子模型
有效因子
评分法
回归法
multi-factor model
effective factors
scoring method
regression method
