Quantitative Investment with Machine Learning in US Equity Market

Quantitative investment attempts to use computer algorithms to predict the price of securities and make automatic trading, in order to gain excess return on the stocks. This paper introduces a strategy based on machine learning algorithms and technical indicators. The model uses several popular technical indicators as inputs and predicts the movement of stock price after a certain short period. Then, a portfolio is constructed using the prediction result. The strategy buys the stocks whose returns exceed the predetermined threshold and sells (shorts) the stocks whose return are below the negative threshold. The empirical results show that the annual return is above 40%, which is far higher than the S&P500 index(2.14%). Considering the risk-adjusted return, the machine learning strategy is better than the S&P500 index. The Sharpe Ratio is higher than that of the S&P500.