Airline Stocks’ Return Prediction with Modified Fama-French 3-Factor Model
Download as PDF
DOI: 10.23977/FMESS2022.040
Author(s)
Shuwen Chen, Yijia Xue, Jianan Yang, Lingxuan Zhu
Corresponding Author
Shuwen Chen
ABSTRACT
Will the industry factor perform better than the market factor in the Fama-French 3-factor Model for airline stocks in the pandemic? The covid-19 pandemic has brought about the extreme downside of the airline industry. This paper reconstructs the Fama-French 3-factor model by replacing the market risk premium with the industry ETF risk premium to investigate whether the industry factor improves the Fama-French 3-factor model. The JETS ETF is chosen because it is the only airline industry ETF. Three sample airline stocks with different scales of capital size are selected to test the performance of FFM and the modified FFM. Besides directly comparing the two factor models, we combine time series models and factor models to test the future stock price prediction ability of the two models without future factor data. The comparison result has been analyzed from several quantitative levels: correlation and covariance, adjusted R squared, F-statistics’ p value, root mean square errors (RMSEs), and correct ratio. The result shows that the industry factor performs much better than the market factor in a pure factor model. However, when combining FFM or the modified FFM with time series models, the performance of FFM and the modified FFM is quite similar and hard to compare.
KEYWORDS
modified Fama-French 3-factor model, time series prediction, industry factor, machine learning approaches