Can somebody help me with how I can determine what deterministic regressors(trend, drift and intercept) I shall include in the Dickey Fuller test. A Stata help file outlines four cases:

1 Random walk without drift α(Intercept)=0,θ(trend)=0 2 Random walk without drift α=0, 3 Random walk with drift θ=0 4 Random walk with drift or without drift No deterministic regressor

I have estimated all of them but how do I know which result to take. So far I cannot reject the Null for all of them except if a drift term is included.

  • $\begingroup$ This is a general and still persisting problem in econometrics/statistics. In general you - as an applied person - "must" have some knowledge about the trending behavior of your data. Though there are some ways out of this situation when it come to trends... see this for example. $\endgroup$ – Math-fun Jul 19 '16 at 14:30
  • $\begingroup$ This is probably a duplicate as it is a frequently encountered issue; you might try looking it up. Probably you will find a good answer there, too. $\endgroup$ – Richard Hardy Jul 22 '16 at 18:45

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.