An Empirical Research of Asset Pricing Models in the US Market under COVID-19

Abstract

Asset pricing theory has been a key area of research in finance. Since the in-troduction of CAPM, asset pricing models have been very widely used. Re-searchers have discovered many phenomena that cannot be explained by the CAPM model, so they have made many refinements to the asset pricing model and developed various improved versions that can better explain stock returns. The Fama and French three-factor model (FFTFM) and the five-factor model (FFFFM) are two of the most classic and most used asset pricing models. However, as time goes by and financial markets continue to evolve, many scholars have questioned whether traditional asset pricing models are still valid. Starting from the theoretical explanation of CAPM, FFTFM, and FFFFM factor selection, this paper uses the stock market data from 2000 to March 2023 to divide stocks into six portfolios based on their size and book value and conduct regressions according to the above three models respectively to evaluate and compare the recent performance of these models. In addition, this paper also performs separate regressions using data from the novel coronavirus epidemic period in particular to verify the robustness of the model to different market environments. According to the results of this paper, the FFFFM model has the best explanatory power. The role of the factors in the model on stock returns varies with portfolio selec-tion, i.e., firm size and book value, and the market environment also has an impact on the significance of the factors, leaving room for improvement in the model to cope with particular markets, but overall, FFFFM is still effec-tive in explaining present market stock prices.