Grey Model And ARMA-GARCH Model Combination Forecasting Based On Wavelet Analysis

As an important part of the financial market,the stock market is a barometer of the economy.Predicting the stock price accurately not only can help managers supervise the market effectively and lead the market to run smoothly,but also can make the investor evade the risk as much as possible.The stock market is a complex and changeable nonlinear system,which has high uncertainty.It is influenced by many factors such as the political and economic situation at home and abroad,the company itself,and the market psychology.It's ineffective to use a single model for forecasting.Therefore,in this paper,based on wavelet analysis,a combined forecasting model that integrates grey model and ARMA-GARCH model for predicting stock price series was proposed.The main work and results are as follows:1.Introduce the gray prediction model and ARMA-GARCH model and the details of model principles and basic processes are included.2.Due to the complexity of the stock market,it's difficult to get ideal results.Therefore,based on wavelet analysis,a combined forecasting model that integrates grey model and ARMA-GARCH model for predicting stock price series was proposed.Stock price series is firstly decomposed by discrete wavelet transform.Use the improved GM(1,1)model capture the approximation(low-frequency)coefficients,use the ARMA-GARCH model capture the detail(low-frequency)coefficients.And then we get the final result by simulating each part.3.We select two sets of data with large fluctuations and small fluctuations respectively and use the above models to predict stock price.At last,this paper use RSME,MAE,and MAPE to evaluate the effect of predicting and experimental results indicate that the combined model based on wavelet analysis has better accuracy combined with those single models.