Having run panel unit root tests, we cannot reject the null of individual unit root (Im, Pesaran), suggesting a non-stationary series.

Question 1 At this point, how do we know if we have a stochastic or deterministic trend or both?

Sub-question: Do panel unit root tests on intercepts only (1) and intercepts and trend (2) correspond to tests on stochastic (the former) or deterministic component (the latter)?

Question 2 How can we make the time series stationary?

To be specific, the panel unit root tests unanimously not reject the null of unit root, for both on the intercept (1) and intercept and trend (2). Can we conclude from these tests whether there is stochastic or deterministic trend (from the latter) in our data?

  • $\begingroup$ Just a comment about your test for stationarity: You could complement the Im-Pesaran-Shin test with the Hadri-Lagrange test which has the null hypothesis that all time series are stationary (with or without linear trend), and the alternative hypothesis that at least some of the panels have unit root. That way you look for evidence in both directions. $\endgroup$ – Alexis Mar 14 '18 at 18:10

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.