统计与决策2025，Vol.41Issue(18)：11-17,7.DOI:10.13546/j.cnki.tjyjc.2025.18.002
非对称平滑转移误差修正模型的协整检验研究
Cointegration Test on Asymmetric Smooth Transfer Error Correction Model
摘要
Abstract
The asymmetric smooth transfer error correction model(AST-ECM)is popular with empirical scholars because of its ability to describe the characteristics of asymmetric nonlinear continuous adjustment in economic reality.However,in its coin-tegration test,both the traditional linear EG cointegration test and nonlinear cointegration test statistics constructed based on Tay-lor expansion have problems of low test efficacy or serious distortion of the test level.Therefore,based on the introduction of the AST-ECM,this paper relies on the unidentified parameter space method to construct the nonlinear cointegration test statistics for the AST-ECM,deriving the asymptotic distribution of the statistics,and simulating its critical value,and furthermore uses the Monte Carlo simulation method to make comparisons on the test level and efficacy between this statistics,the traditional EG statis-tics and various statistics based on Taylor expansion,so as to test its limited sample properties.The research results go as below:(1)The asymptotic distribution of the proposed nonlinear cointegration test statistic is not a standard distribution.When the null hypothesis holds,it converges to a complex stochastic function;when the alternative hypothesis holds,it tends to positive infinity.(2)The Monte Carlo simulation results show that compared with the traditional EG statistic and various statistics based on Taylor expansion,the proposed nonlinear cointegration test statistic has a lower distortion degree of test level and relatively better test ef-ficacy,and its statistical properties are more excellent.关键词
AST-ECM模型/参数空间/检验水平/检验功效Key words
AST-ECM model/parameter space/test level/test efficacy分类
管理科学引用本文复制引用
吴建銮,霍宇辰,游万海,南士敬..非对称平滑转移误差修正模型的协整检验研究[J].统计与决策,2025,41(18):11-17,7.基金项目
陕西省社会科学基金项目(2022D014) （2022D014）
陕西省教育厅专项科研计划项目(21JK0302) （21JK0302）
