摘要
宏观经济环境的变化会传导作用于外汇市场,然而宏观数据多为月度低频数据,汇率为日度高频数据,这使得自变量与因变量之间的时间间隔不一。鉴于此,选用GARCH-MIDAS与GJR-GARCH两种混频时间序列模型对其展开研究,选择Shibor, M2与工业增加值的水平值作为宏观低频输入变量,通过滑动窗口测得的22日CNY/USD汇率方差作为真实波动率的代理变量。结果表明:低频成分的解释变量会对高频汇率的条件方差产生影响,利率以及利率与其它因子的双因子组合具有相对较好的拟合性,且这些因子对长期成分非对称性影响是不显著的。
Changes in the macroeconomic environment can have in impact on the foreign exchange market.However,the macro data are mostly monthly low-frequency data,and the exchange rate is daily high-frequency data,which makes the time interval between the ind ependent variable and the dependent variable different.In view of this,this paper selects GARCH-MIDAS and GJR-GARCH two mixed frequency time series models to study it,selects Shibor,M2 and the level of industrial added value as comprehensive input variables,and measu res the 22-day CNY through a sliding window.The variance of the USD exchange rate as a proxy for true volatility.The results show that the explanatory variables of the low-frequency components will have an impact on the cond itional variance of the high-frequency exchange rate,the interest rate a nd the two-factor combination of interest rates and other factors have a relatively good fit,and these factors have no significant effect on long-term compositional asymmetry.
作者
赵耀
张英辉
ZHAO Yao;ZHANG Ying-hui(Xingzhi College,Xi'an University of Financeand Economics,Xi'an,Shaanxi 710026,China)
出处
《计算机仿真》
2024年第7期383-389,共7页
Computer Simulation
基金
陕西省教育厅专项科学研究计划(19JK0330)。
