A Study Of Sino-US Exchange Rate Based On Time Series

Exchange rate is the ratio or price of one country's currency to another country's currency.Exchange rate changes have a direct regulatory effect on a country's import and export trade.Since 2014,the RMB has continued to depreciate substantially,falling 1641 basis points,with a depreciation of 2.71%.This is the longest and largest depreciation since the informal link between the RMB and the US dollar in 1994.In view of the large fluctuation of RMB exchange rate since 2014,this paper will study the fluctuation and development trend of RMB under the influence of international economic environment since 2014.Through the study of time series,we can understand that many models in time series can be used to study the fluctuation of financial series.For example,the AR model,MA model and ARMA model of stationary time series,the ARMA model,ARCH model and GARCH model of non-stationary time series.On the other hand,the wavelet transform proposed by J.Morlet,an engineer engaged in petroleum signal processing in France,uses the localization analysis of time and frequency to refine the signal step by step through the scaling translation operation,and finally achieves the time subdivision at the high frequency and the frequency subdivision at the low frequency,so as to better fit the fluctuations of each frequency,and gives effective prediction results.The empirical research in this paper mainly fits and predicts the fluctuation of 988 RMB exchange rate data from August 1,2014 to May 18,2018.It is mainly used in ARMA,GARCH model and ARIMA combined with wavelet analysis model.The model is established by R and Python software and forecasted.