The Measurement Of RMB Exchange Rate Risk Against USD Based On GARCH-VaR Model

In order to make the RMB middle exchange rate quotation mechanism more perfect,in August 2015,the central bank announced the adjustment of the RMB middle exchange rate formation mechanism,the market makers need to refers to the exchange rate between the banks,and then provide exchange rate to the Chinese foreign exchange trading center.This reform makes the quotation mechanis,m of the exchange rate between RMB and USD further marketization.Meanwhile,in the foreign exchange market,the RMB exchange rate against the USD also appeared in a sustained volatility.As there is a very closed economic relationship between China and the United States,in such a market environment,China’s various economic entities are facing the rising exchange rate risk.As a core of the exchange rate risk management process,the importance of exchange rate risk measurement is becoming more and more prominent in the present.This paper selects the middle exchange rate of the RMB against USD between August 11,2015 and March 21,2017,and the data is logarithmized to became the RMB exchange rate Yield series.We used GARCH,EGARCH and TAGRCH model,after the test,we obtain that the best model is EGARCH(1,1)model.After calculating the variance of the yield sequence using the EGARCH(1,1)model,we calculated the VaR of the yield sequence based on the variance obtained.Finally,we test the VaR value by observing and the Kupiec failure rate test,the test result shows that the yield series of the RMB exchange rate is more consistent with the t distribution than the normal distribution,and the yield series has the characteristics of smooth,no autocorrelation and heteroskedasticity,and is suitable for the GARCH family model.The VaR calculated on the basis of the EGARCH(1,1)model has a good coverage for the actual yield fluctuation,and only 15 failures have occurred during the selected data period,and the Kupiec failure rate test has been passed.We believe that it is an accurate and practical method to measure and forecast exchange rate risk through the GRACH-VaR model.