The Study Of RMB Exchange Rate Prediction Based On Mixed Frequency Data

The exchange rate is an important issue in the field of international finance.Research on the exchange rate of the RMB has always been a research hotspot at home and abroad.The prediction of the RMB exchange rate has the function of regulating market expectations and preventing financial risks for the supervisory institution.For enterprises and individuals,it helps to optimize the portfolio strategy and avoid market risks.The article first sorts out the relevant literature on macro and micro influencing factors of RMB exchange rates,forecasting methods of RMB exchange rate prices,and the study of RMB exchange rate volatility.Then,this paper makes further application research on the prediction of China's exchange rate market based on the mixed data sampling models proposed by Ghysels et al.(2004).The main work and conclusions are:For the high-frequency(daily)forecast of exchange rate,the mixed-frequency data sampling model(MIDAS)is used to directly model the mixed-frequency data of high-frequency data(minute data of different frequencies),which more fully uses high-frequency data information of China's exchange rate market.The results show that the prediction effect of Beta Non-Zero-MIDAS model is obviously better than exchange rate long-term maturity structure model,ARIMA model and GARCH model,but there is no significant difference compared with the SVM model,which proves the superiority of the mixed data sampling model over the traditional model.For the low-frequency(monthly)forecast of the exchange rate,this paper uses the multi-mixed data sampling model(M-MIDAS)to forecast the RMB exchange rate and analyzes the effects of the foreign exchange reserves,terms of trade,the price ratio of non-traded goods and trading goods,the Fed's benchmark interest rate,and high-frequency financial variables on exchange rate price,respectively.The prediction accuracy of the M-MIDAS model is compared with the theoretical model of the same frequency and the traditional modeling method(ARIMA,BEER,ARDL),and the prediction effect is significantly improved,which further enriches the research on China's exchange rate prediction.For the study of the volatility of the RMB exchange rate,this paper analyzes the impact of China's macroeconomic economic prosperity and uncertainty of economic policies on the exchange rate fluctuation of the RMB by constructing the GARCH-MIDAS model,and forecasts the exchange rate volatility.The empirical results show that the mixed-frequency volatility model has good applicability in China's exchange rate market,and can monitor the impact of macroeconomic factors on long-term components of exchange rate fluctuations.This paper uses the research pattern of sampling of mixed data to further deepen the understanding of exchange rate price movements and volatility behavior.This is favorable for the regulatory authorities to grasp the changing trend of exchange rates and the characteristics of exchange rate fluctuations.