Application of Seasonal ARIMA Model in Forecasting the Exchange Rate: Taking RMB to USD as an Example

Forecasting exchange rate is very profitable and crucial. Realizing the forecasting of the exchange rate means that personal investors can do arbitrage; For the policy makers, it means that they can optimize their fiscal policy and monetary policy. The article has used seasonal ARIMA model in terms of forecasting the RMB to USD exchange rate, and captures the seasonal factors of the RMB to USD exchange rate which is a time series better. First, the author deals with the data by using first-order seasonal difference；then,the author applies the auto ARIMA model in Rstudio to fit the model with the data, receiving the ARIMA(4,2,1)(0,1,0)(365). After that, the author forecasts the RMB to USD exchange rate in the next 90 days. Finally, the author does the residual test, and discusses the autocorrelation,fitting degree, and whether or not it is overfitted of the model. The conclusion of the writer is that the RMB to USD exchange rate will fall in the next 90 days.