Smart Portfolio Trading Strategy Based on Random Forest Model with Limited Information

Quantitative trading refers to using advanced mathematical models instead of people's subjective judgment and the use of computer technology and big data to formulate trading strategies. Quantitative trading dramatically reduces people's irrational investment strategies due to emotional fluctuations. With the continuous development of global financial markets, quantitative trading has been welcomed by more and more investors. This paper takes gold and bitcoin as the research object, takes machine learning as the method, and aims to maximize the return of their portfolio. Before December 26, 2016, the daily stock price changes are selected as the training set, and the daily stock price changes from December 26, 2016, to September 10, 2021, are selected as the test set. Establish the rate of the return prediction model and backtest system, select the trading decision through a reasonable investment portfolio and buying and selling time points, and conduct sensitivity analysis. This paper provides an investment strategy on bitcoin and gold, which has the advantages of high yield, strong stability, and high application value.