各位大侠：
本人有一疑问， 想请教各位。
我用Johansen and Juselius (1990) 检验得出变量之间存在协整（cointegration），但是用Gregory and Hansen (1996) 检验无法拒绝原假设。
也就是说检验无法拒绝变量之间不存协整关系的假设。该怎么解释这种情况呢?

Dear Friends on the forum,

I have something boggling my mind. When I use the Johansen and Juselius (1990) (JJ) test, it indicates there is a least one cointegration  among the variables under investigation, however, when I use the Gregory and Hansen (1996) (GH test)cointegration with regime shift, None of the proposed test such as ADF*, Za and Zt tests able to reject the null hypothesis of no cointegration. This results makes me very uncomfortable.  By right, the GH test should reconfirm the results of JJ test. I am stuck here right now. I have no idea what to next?

I hope friends on the forum may advise on this issue. thank you very much.