The RATS Software Forum

Thank you Tom,

Is there a test to test for contegration and structural break at the same time?

elhampa wrote:Thank you Tom,

Is there a test to test for contegration and structural break at the same time?

None that I'm aware of or at least not in the way that I think you mean it. Enders-Siklos is a test for cointegration with a specific form of break, which is in the loadings, not the cointegrating vector.

Thank you Tom,

How about the Gregory and Hanseon test "Residual-based tests for cointegration in models with regime shifts" Journal of Econometrics, 1996, vol. 70, issue 1, 99-126?
This test seems test for cointegration and break together. Please correct me if I am wrong.

And do you have any idea whether there is a more updated of this type?

Cheers
Elham

No. That's a test for cointegration allowing for breaks (in the cointegrating vector). There is never any attempt to test whether there actually is a break.

Sorry Tom,

I am a confused.
Do you mean the Gregory and Hansen is test for break in a cointegrated system?

Regards
Elham

No. Please read the description:

https://estima.com/ratshelp/gregoryhansenprocedure.html

It is assumed that there is a break at an unknown location in the (possible) cointegrating vector and it tests for cointegration under those assumptions. It does not test for a break. What you're looking for basically can't exist (logically). If the null is no cointegration (i.e. under root residuals) and no break, then the alternative is either cointegration or breaks but not necessarily both.