Economic Policy Uncertainty And RMB Exchange Rate Volatility

Since the reform of the RMB exchange rate system in July 2005,there are significant fluctuations in the RMB exchange rate.Traditionally,the GARCH model has been widely used to predict exchange rate volatility,but it only utilizes historical yield information of the RMB exchange rate and ignores other information(such as macroeconomic information).Later,some scholars introduce the mixed frequency data sampling method and propose the GARCH-MIDAS model,which can effectively introduce low-frequency macro variables(such as macroeconomic information)and improve the prediction accuracy of RMB exchange rate volatility to a certain extent.However,the GARCH-MIDAS model still ignores some empirical factual features in the conditional distribution of returns in empirical studies,such as time-varying higher moments(skewness and kurtosis).In addition,in recent years,changes in economic policy uncertainty in various countries have seriously affect financial markets,especially exchange rate markets.Therefore,comprehensive consideration of time-varying higher moments and economic policy uncertainty is crucial for accurately modeling and predicting the volatility of the RMB exchange rate.Specifically,this article first analyzes the mechanism between economic policy uncertainty and RMB exchange rate volatility,and then proposes the research hypotheses of this article;Secondly,this article constructs the GARCH-MIDAS-SK model,and introduces the economic policy uncertainty(EPU)index to construct the GARCH-MIDAS-EPU model and GARCH-MIDAS-SK-EPU model under the GARCH-MIDAS model and GARCH-MIDAS-SK-EPU model frameworks,respectively;Then,this paper selects daily USD/CNY exchange rate return data and monthly global economic policy uncertainty index(GEPU),America economic policy uncertainty index(AEPU)and China economic policy uncertainty index(CEPU)data for empirical analysis through in sample fitting,out-of-sample forecasting,MCS test and robustness test,respectively.And we explore the long-term impact of EPU on RMB exchange rate volatility and its predictive effect;Subsequently,this article conducts economic value analysis by calculating Va R to prevent market risks.Finally,we conclude the research of this article and propose targeted recommendations based on the research findings.The empirical study shows that both GEPU,AEPU and CEPU have a significant positive impact on the long-term exchange rate volatility between the USD/CNY;Compared with various competition models(including GARCH,GARCH-MIDAS,GARCH-MIDAS-EPU and GARCH-MIDAS-SK models),the GARCH-MIDAS-SK-EPU model with time-varying higher moments and EPU can produce more accurate forecasting effect of RMB exchange rate volatility out-ofsample.In addition,the superior prediction ability of GARCH-MIDAS-SK-EPU model is robust to different out-of-sample prediction windows,different countries’ EPUs,bilateral EPUs,new economic policy uncertainty indexes,different time-variant skewness and kurtosis and different out-of-sample intervals.The research results emphasize the value of introducing time-varying higher moments and EPU to predict the RMB exchange rate volatility.