BIC-based unit-root detection: Simulation-based evidence

作者:

Highlights:

摘要

A unit root is detected using the values of the Bayesian information criterion (BIC) of a battery of alternative models with and without a unit root. ARMA(1, 1)-based simulation results suggest that the proposed method generally outperforms influential unit-root tests except in the case of a small sample size with a large negative MA(1) parameter.

论文关键词:BIC,Model selection,unit root

论文评审过程:Available online 24 July 2006.

论文官网地址:https://doi.org/10.1016/j.amc.2006.05.110