摘要
本文通过采用不同的线性和非线性一元时间序列模型对人民币实际汇率行为进行研究。研究结果表明,非线性的自我激励阈值自回归模型和平滑过渡自回归模型对人民币实际汇率历史数据有很好的拟合效果,且人民币实际汇率具有显著的非线性动态行为特征。
Several univariable time series models are used to study the behavior of RMB's real exchange rate in this paper.According to the empirical study,we find two nonlinear models(the self-exciting threshold autoregressive model and the smoothing transition autoregressive model)have better performance than the linear autoregressive model on describing the behavior of RMB's real exchange rate,which implies its obvious nonlinear dynamic properties.
出处
《数量经济技术经济研究》
CSSCI
北大核心
2007年第2期11-18,共8页
Journal of Quantitative & Technological Economics
关键词
时间序列模型
人民币实际汇率
线性自回归模型
自我激励阈值自
回归模型
平滑过渡自回归模型
Time Series Models
RMB's Real Exchange Rate
Linear Autoregressive Model
Self-exciting Threshold Autoregressive Model
Smoothing Transition Autoregressive Model