摘要
在非对称的门限自回归模型下,由于传统单位根检验式的误设,会导致单位根检验势下降。本文通过一系列的Monte-Carlo模拟表明:非对称性对ADF和PP检验的检验势会产生较大影响,而对其他四种常用的单位根检验势产生的影响较小,也就是说,在非对称的门限自回归下,非对称性对退势单位根检验势产生的影响较小。模拟中也发现:NP单位根检验对TAR模型和持久性都具有稳健性。
It is widely known that test power is reduced owing to the specification error of test equation while in testing the unit root hypothesis against a threshold autoregressive model（TAR） as the alternative. By means of a Monte Carlo study, the paper investigates the performance of six popular unit root tests against a TAR model. Our simulation results show that the asymmetry of TAR makes a great impact on test power of ADF and PP tests, and however, makes a small impact on test power of the other unit root tests. That is to say, unit root tests based on detreuding to data process are more powerful for TAR models. In addition, simulation results reveal that NP test has robust property while against a TAR model.
出处
《统计研究》
CSSCI
北大核心
2010年第2期98-106,共9页
Statistical Research
关键词
门限自回归
单位根检验
检验势
MC模拟
Threshold Autoregressive Model
Unit root test
Test power
Monte-Carlo Simulation
