I am studying cointegration theory in time series using online resources.

As per https://www.econometrics-with-r.org/16-2-ooiatdfglsurt.html

DF-GLS test is implemented in the package urca can be used to test for cointegration, but the null hypothesis of this test is to check for unit root in zt=xt- \theta*yt.

There are other R functions that provide the cointegrating vector (e.g., VECM function in tsDyn package, ca.jo function in urca package), but I could not find a way to do any hypothesis testing on the estimated cointegrating vector.

To be specific, how do I test if this cointegrating vector (1, \theta) is equivalent to (1, -1) ?

I found a related question here, but not specifically on hypothesis testing.

Test for cointegration between two time series using Engle–Granger two-step method

Please advise.

  • 1
    $\begingroup$ hi: when you do engle granger, ultimately ( at the last step ), you run a regression. a test of beta = -1 in that regression is the test of the cointegrating vector being (1, -1). $\endgroup$
    – mlofton
    Jul 23, 2020 at 14:20


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.