持续结构突变的理论影响与检验方法研究

A Research on the Theoretical Consequences and Testing Procedure of Shift in Persistence

研究目标:研究持续结构突变的理论影响与检验方法,检验国际能源价格的持续结构突变。研究方法:理论证明与蒙特卡洛模拟研究相结合。研究发现:当序列存在持续结构突变,且非平稳区间较窄时,ADF单位根检验倾向于支持序列的平稳性,但采用该序列进行回归分析会导致伪回归。研究创新:证明了存在持续结构突变时ADF与t检验统计量的渐近性质,研究了允许存在未知均值突变的持续结构突变的检验方法。研究价值:给出了持续结构突变的计量影响以及更加稳健的检验方法。

Research Objectives: Investigating the economteric impacts of shifts in persistence and its robust testing procedure,testing the shift in persistence of the international energy prices. Research Methods: Theoretical derivation and Monte Carlo experiments. Research Findings: When the time series undergoes a shift in persistence and the nonstationary part is relatively small,ADF unit root test tends to favor the stationarity of the data. However,the linear regression conducted based on data with shifts in persistence suffers from the spurious regression problem. Research Innovations: We derive the asymptotic distributions of the ADF and t test statistics,and provide testing procedure allowing for structural breaks in the mean process. Research Value: We illustrate the econometric consequences of shift in persistence and more robust testing procedure.