Abstract
With the further development of futures of PTA, the need of futures prices for long-term predictions is more and more imperative .Futures of PTA forecast model is also an economic issue to investors. Therefore, analysis a number of factors of influence, select the main factors, establish multi regression model with Stepwise Regression, test Goodness-of-fit and significant of regression model, and then make it to be strong forecast results. Moreover use time series to forecasting model. In the time series, it is important to determine the stability of the sample data. Use ADF test to determine stability. Granger causality test reuse, select the main effects of factors, The establish vector autoregressive model. Thus allows the prediction model has a contrast effect. Through the establishment of these two models, we could make a regression equation to predict futures of PTA in a short period of time. Investors can also play a role in the reference.
Keyword: Futures of PTA ;multi regression ;Time Series;Prediction model