公司特征是否和收益率成非线性关系?--基于中国股市的实证研究

Are Company Characteristics and Stock Return in a Non-linear Relationship?——Based on the Empirical Research of Chinese Stock Market

资产定价领域已成为金融学研究的热点。与传统因子模型不同,本文提出了加权非线性因子模型,将公司特征的未知函数作为因子载荷,时变的因子收益作为权重,研究股票超额收益率与公司特征之间是否存在非线性关系,以及模型对我国股票市场的适用性和解释力度。选取1995年7月—2020年6月我国A股上市公司的财务数据,考虑公司规模特征、价值特征和动量特征,采用核函数回归方法同时估计出因子载荷和因子收益,结果发现三个特征都是显著的,并且动量特征与收益率成非线性关系。此外,与美国股票市场相比,本文模型更加适用于中国股票市场,解释力度更强;与Fama-French三因子模型相比,本文估计误差更小。本文提出的模型在基本面分析和因子投资策略领域中具有借鉴意义。

Asset pricing has been overwhelmingly researched in finance.A weighted nonlinear factor model is established based on existing research results,with unknown functions of firm characteristics as factor loadings and timevarying factor returns as weights,to investigate whether there is a nonlinear relationship between stock excess returns and firm characteristics,and the applicability and explanatory strength of the model to China's stock market.The financial data of China's A-share listed companies from July 1995 to June 2020 are selected,and the characteristic functions and factor returns are estimated simultaneously by using kernel function regression method considering firm size characteristics,value characteristics and momentum characteristics,and it is found that all three characteristics are significant and the momentum characteristics are nonlinearly related to the returns.In addition,the model in this paper is more applicable to the Chinese stock market and has stronger explanatory power compared to the US stock market;compared with the Fama-French three-factor model,the estimation error in this paper is smaller.The model proposed in this paper has implications in the field of fundamental analysis and factor investment strategies.