Research on Multistage Dynamic Trading Model Based on Gray Model and Auto-Regressive Integrated Moving Average Model

Abstract

Quantitative portfolio investment mainly depends on historical data analysis and market trend prediction to make appropriate decisions, which is an important mean to reduce risks and increase returns. Based on summarizing the existing traditional single forecasting models and multiobjective dynamic programming models, this paper puts forward a new quantitative portfolio model to improve the accuracy of asset price forecasting results and the appropriateness of investment trading strategies, to better realize the maximization of investment returns. This model analyzes and forecasts daily price data by establishing a combination forecasting model of the gray GM (1,1) model and the ARIMA time series model and establishes a multiobjective dynamic programming model with moving average convergence divergence (MACD) and Sharpe ratio indicators as risk constraints to formulate appropriate investment trading strategies. The results show that by solving the quantitative portfolio trading model established in this paper and analyzing the sensitivity and robustness of the model, the price of gold and Bitcoin, two volatile assets, can be accurately predicted, and the best investment portfolio trading strategy can be effectively worked out on the premise of considering the risk level.