Quantitative Strategy Based on Decision Tree and Bollinger Band

Because stock investors often invest blindly and emotionally, and the quantitative investment strategy is more rational and reliable, so the purpose of this paper is to establish a suitable quantitative strategy for investors. We selected six indicators as stock price forecasting indicators, constructed the model by using CART algorithm, Mallat algorithm and Brin Belt index.The test results show that the earnings gap of the multi-short portfolio of the modified model is significantly increased, reaching the level of 3%. Combining with the timing strategy of Brin Belt, the total earnings can exceed the average earnings of technology stocks by 3.3%.