基于GARCH模型的外汇风险度量中VaR的测度研究

Research of VaR in Foreign Exchange Risk Measurement Based on GARCH Model

近年来,中国外汇储备的规模不断扩大,如何控制和防范风险成为当务之急。通过利用残差项服从t分布的GARCH模型,对人民币/美元汇率风险进行测度,计算和预测了2005-2014年人民币/美元收益率的VaR值。研究结果表明:基于该方法估计得到的风险值是可靠的,精度较高,可以体现其收益率的分布状态和波动性的影响作用,为中国外汇风险度量提供了有力的参考依据。在此基础上,呼吁商业银行加强汇率风险管理意识并建立完善的汇率风险管理体系。

In recent years,since Chinese foreign exchange reserves have risen continuously,how to control and prevent foreign exchange risk becomes imperative. This paper employs GARCH model with t- distribution residual errors to estimate RMB / USD exchange risk,calculates and forecasts the value at risk of return of RMB / USD from 2005 to2014. The study indicates that the estimation of VaR by GARCH model is reliable with high precision and can both manifest the distribution condition of the returns ratio and portray its volatility,which can provide some references for the measurement of Chinese foreign exchange risk. Therefore,this paper suggests banks should strengthen consciousness of exchange risk management with complete system.