The relationship between firm characteristic variables and stock returns: An empirical study based on principal component analysis

摘要

A large number of empirical studies show that the firm characteristic variables, including the firm size, book to market ratio (B/M), momentum and etc., they have a certain correlation with future stock returns. However, the information contained in many stock feature variables is too complicated, there is a linear correlation between them, leading to overlapping problems of the information. So in the study of stock returns, how to avoid information overlapping is worth studying. Based on this starting point, this paper uses principal component analysis method to extract new principal component variables from many variables that reflect the characteristics of stocks, and then use the new variables combined with Fama-Macbethe regression to predict future stock returns. It is of great theoretical and practical significance for the investors to make full use of the real-time information of the stock market, to effectively create the investment strategy and master the law of the securities market.