An Empirical Research On The Selection Of RMB Exchange Rate Volatility Models

Based on the relevant daily data of exchange rate(RMB/dollar) after July21,2005,this paper did an empirical research on the RMB exchange rate by use of time series analysis method, Through the establishment of ARIMA model, GARCH family models (including the GARCH model, TGARCH model and the EGARCH model) and the promoted models (AR-GARCH family models and the GARCH-M family models),the paper fitted and projected the fluctuate status of currency exchange rate of RMB after a descriptive analysis. And then compared above models from both static and dynamic aspects.It is found that,the TGARCH-M model is the best model;and the RMB exchange rate fluctuations showed obvious asymmetry; the variance equations of GARCH model and the EGARCH model are not stable, it can not be applied to research on fluctuations of the RMB exchange rate.Furthermore, the risk mechanism does not play a role in foreign currency market of China,and the exchange rate fluctuations is very sensitive to government policy.