A Study on the Influencing Factors of Stock Returns Based on PCA -LASSO Model
Abstract
This paper analyzes the influence factors to explain the cross-section of expected returns. Our paper proposes the PCA-LASSO model and its reduced model, comparing them with the results of traditional linear regression model and LASSO regression model. The results show that the proposed PCA-LASSO model and its reduced model are superior to the existing two methods in the prediction of the stock returns of the industry. For practical applications, the research method proposed in this paper has wide applicability and the research conclusion can provide important reference for the securities investors.
Keywords
PCA-LASSO model, Stock returns, Influencing factors, Out of sample prediction
DOI
10.12783/dtcse/cmsam2017/16370
10.12783/dtcse/cmsam2017/16370
Refbacks
- There are currently no refbacks.