Forecast And Research For The RMB Exchange Rate Basing On Time Series Analysis

In 2008 after the financial crisis, the RMB held on US dollars once again, only fluctuating within a small range. However, the United States has strongly urged revaluation of the RMB. Whether it will appreciate? And how much space of appreciation? Many people give different views. This article will give explanation and forecast of the RMB/dollar exchange rate base on time series analysis.Time series analysis is an important method to the dynamic data analysis. In the process, a mathematical model will be used to give description by way of the information from the random series. Some models turn out to be successful, such as ARMA(p,q), GARCH (1,1), ARCH-M, EGARCH, and so on. In the empirical researching, a reasonable explanation for the volatilities is the most important.In the paper, Eviews5.0 and SAS statistical softwares were used, with the RMB exchange rate data as the research object. It turns out that AR (2)-ARCH (1) and GARCH (1,1) are more suitble to the rate series than other models. However theσand asymmetric information give a little influence to the series. It proved successfully, we got the Y245=682.819849,only 0.0000144 than the real value 682.81, and residuals is -0.009849, especially the GARCH(1,1) model, it has a less standard variance than 0.05. It makes a good sense to the application. However, the ARCH(1) model is more radical than the GARCH(1,1) at the forecast, especially the exchange rate series with violent fluctuations.