In the paper, the econometrics and machine learning field in asset pricing are summarized. A complete review and comparison of techniques including Ordinary Least Squares, Elastic Net, Random Forest and Neural Networks for stock return prediction is performed. All approaches are tested using a predictive out-of-sample R2 to observe their behavior under an unstudied sample. Furthermore a portfolio using long-short strategy is constructed. Results show that machine learning algorithm are raising hope in the asset field but remain inaccurate for the moment.