本文聚焦于我国上证A股市场,以2013年1月至2023年12月的数据为基础,对CAPM模型以及Fama-French三因子模型在A股市场的适用状况展开实证研究。研究发现,在CAPM模型中市场风险溢价系数显著,但其仅能解释个股超额收益变动的13.08%;而同时,...本文聚焦于我国上证A股市场,以2013年1月至2023年12月的数据为基础,对CAPM模型以及Fama-French三因子模型在A股市场的适用状况展开实证研究。研究发现,在CAPM模型中市场风险溢价系数显著,但其仅能解释个股超额收益变动的13.08%;而同时,在三因子模型中三个指标,分别是市场风险溢价、规模与价值因子,它们的系数均显著,能解释65.4%的个股超额收益变动。结果表明,虽CAPM模型有一定理论基础,但Fama-French三因子模型因因子结构丰富和拟合优度高,在解释资产收益率方面更优。不过二者在金融研究中均重要,未来可结合其他因素或改进模型结构提升解释和预测能力,为投资决策提供依据。This paper focuses on China’s Shanghai A-share market and conducts an empirical study on the applicability of the CAPM model and the Fama-French three-factor model in the A-share market, based on data from January 2013 to December 2023. The research finds that the market risk premium coefficient of the CAPM model is significant, but it can only explain 13.08% of the variation in individual stock excess returns. In contrast, for the Fama-French three-factor model, the coefficients of market risk premium, size, and value factors are all significant and can explain 65.4% of the variation in individual stock excess returns. The results show that although the CAPM model has a certain theoretical basis, the Fama-French three-factor model is better at explaining the asset return rate due to its rich factor structure and high goodness of fit. However, both models are important in financial research. In the future, other factors can be combined or the model structure can be improved to enhance the ability of explanation and prediction and provide a basis for investment decisions.展开更多
多年来,学者们基于市场和公司基本面,建立了成熟的多因子定价模型,并通过市场数据验证。然而,市场并非完全有效,非理性现象时常出现,挑战了有效市场理论。行为金融理论逐渐重要,新的定价因子不断被探索并纳入模型,分析其对股票收益和资...多年来,学者们基于市场和公司基本面,建立了成熟的多因子定价模型,并通过市场数据验证。然而,市场并非完全有效,非理性现象时常出现,挑战了有效市场理论。行为金融理论逐渐重要,新的定价因子不断被探索并纳入模型,分析其对股票收益和资产定价的影响。我国股市发展较晚,政策尚不完善,散户较多,非理性因素可能对市场影响更大。国内研究发现,三因子模型和五因子模型对于我国来说,三因子的适配度会更高,因此在实证部分,本文选取了经典的Fama-French三因子模型,然后以隔夜收益率、个股换手率增长额、指令不平衡这3个指标构建投资者情绪指标,使用主成分分析法构造投资者情绪因子SENT。将情绪因子添加到三因子模型中,构建出一个新的四因子模型,然后对两个模型进行对比分析。Over the years, scholars have built mature multi-factor pricing models based on market and company fundamentals, which have been validated by market data. However, markets are not entirely efficient, and irrationalities often arise that challenge the theory of efficient markets. As behavioral finance theories become increasingly important, new pricing factors are being explored and incorporated into models to analyze their impact on stock returns and asset pricing. China’s stock market has developed late, the policy is not perfect, there are many retail investors, and irrational factors may have a greater impact on the market. Domestic research has found that compared with the five-factor model, the three-factor model is more suitable for China. Therefore, in the empirical part, this paper selects the classic Fama-French three-factor model, and then constructs the investor sentiment index with the three indicators of overnight return, the growth of individual stock turnover rate and the imbalance of instructions, and uses the principal component analysis method to construct the investor sentiment factor SENT. A new four-factor model was constructed by adding the sentiment factor to the three-factor model, and then the two models were compared and analyzed.展开更多
文摘本文聚焦于我国上证A股市场,以2013年1月至2023年12月的数据为基础,对CAPM模型以及Fama-French三因子模型在A股市场的适用状况展开实证研究。研究发现,在CAPM模型中市场风险溢价系数显著,但其仅能解释个股超额收益变动的13.08%;而同时,在三因子模型中三个指标,分别是市场风险溢价、规模与价值因子,它们的系数均显著,能解释65.4%的个股超额收益变动。结果表明,虽CAPM模型有一定理论基础,但Fama-French三因子模型因因子结构丰富和拟合优度高,在解释资产收益率方面更优。不过二者在金融研究中均重要,未来可结合其他因素或改进模型结构提升解释和预测能力,为投资决策提供依据。This paper focuses on China’s Shanghai A-share market and conducts an empirical study on the applicability of the CAPM model and the Fama-French three-factor model in the A-share market, based on data from January 2013 to December 2023. The research finds that the market risk premium coefficient of the CAPM model is significant, but it can only explain 13.08% of the variation in individual stock excess returns. In contrast, for the Fama-French three-factor model, the coefficients of market risk premium, size, and value factors are all significant and can explain 65.4% of the variation in individual stock excess returns. The results show that although the CAPM model has a certain theoretical basis, the Fama-French three-factor model is better at explaining the asset return rate due to its rich factor structure and high goodness of fit. However, both models are important in financial research. In the future, other factors can be combined or the model structure can be improved to enhance the ability of explanation and prediction and provide a basis for investment decisions.
文摘多年来,学者们基于市场和公司基本面,建立了成熟的多因子定价模型,并通过市场数据验证。然而,市场并非完全有效,非理性现象时常出现,挑战了有效市场理论。行为金融理论逐渐重要,新的定价因子不断被探索并纳入模型,分析其对股票收益和资产定价的影响。我国股市发展较晚,政策尚不完善,散户较多,非理性因素可能对市场影响更大。国内研究发现,三因子模型和五因子模型对于我国来说,三因子的适配度会更高,因此在实证部分,本文选取了经典的Fama-French三因子模型,然后以隔夜收益率、个股换手率增长额、指令不平衡这3个指标构建投资者情绪指标,使用主成分分析法构造投资者情绪因子SENT。将情绪因子添加到三因子模型中,构建出一个新的四因子模型,然后对两个模型进行对比分析。Over the years, scholars have built mature multi-factor pricing models based on market and company fundamentals, which have been validated by market data. However, markets are not entirely efficient, and irrationalities often arise that challenge the theory of efficient markets. As behavioral finance theories become increasingly important, new pricing factors are being explored and incorporated into models to analyze their impact on stock returns and asset pricing. China’s stock market has developed late, the policy is not perfect, there are many retail investors, and irrational factors may have a greater impact on the market. Domestic research has found that compared with the five-factor model, the three-factor model is more suitable for China. Therefore, in the empirical part, this paper selects the classic Fama-French three-factor model, and then constructs the investor sentiment index with the three indicators of overnight return, the growth of individual stock turnover rate and the imbalance of instructions, and uses the principal component analysis method to construct the investor sentiment factor SENT. A new four-factor model was constructed by adding the sentiment factor to the three-factor model, and then the two models were compared and analyzed.
