Zhu Shangshuai
This paper delves into the dynamics of the exchange rate by analyzing 2189 daily median price data of the USD-RMB exchange rate between January 5, 2015 and December 29, 2023. The study adopts the ARMA model to construct the mean equation of the exchange rate, and combines the GARCH family of models, including GARCH, TGARCH and EGARCH, to analyze the data in detail. The analysis results reveal the existence of characteristics such as clustering, asymmetry and leverage effect in the exchange rate of USD-RMB. In general, the GARCH(1,1) model has a more balanced performance, especially in the most important MAPE and MAE indicators, so it can be considered that the GARCH(1,1) model has a better forecasting effect.
RMB exchange rate; ARMA; GARCH family model; leverage effect; asymmetry