单位根检验的模型设定问题研究

On model specification and testing for unit root

统计量tαρ^和tτρ^会受到"噪声参数"α和δ的影响,DF在联合假设基础上给出的不变统计量仅能检验部分假设,二者的不一致可能导致通常检验程序的错误推断。依据模型设定正确和真实过程完全未知两种情况下的单位根检验,得出前者需要统计量tρ^和Φ配合使用,而后者需要遵循一定的检验程序,而DJSR程序是其中比较稳健的选择。

The statistics tρ^α and tρ^τ may be affected by ＂nuisance parameters＂ α and δ , so DF formulate composite null hypothesis to derive the invariant statistics, but which can only test a part of the null hypothesis. The underlying inconsistency will induce wrong inference made by usual test procedure. Ih this paper, tests for unit root under two different conditions are discussed. To make effective inference, statistic tρ^ together with Φ is suggested when model specification is right, while some test procedures such as robust DJSR procedure should be followed when the form of the data - generating process is completely unknown.