A Study of Cross Sectional Stock Returns Using High-Dimensional SUR Model and Many Firm Level Characteristics

In this paper, we propose to study the cross sectional stock returns using the high-dimensional seemingly unrelated regression (SUR) model [1] with many common factors as well as observed firm level characteristics. The advantages of the proposed new method include: first, we consider a large number of firm level variables that could potentially be important in explaining the stock returns; second, we allow the heterogeneity in pricing of each asset; third, the cross sectional correlations of stocks are embedded in the estimation procedure to improve the estimation efficiency.