2
$\begingroup$

I currently apply a VAR models with the vars package in R. There are the options to select "both", "none", "trend", or "constant".

I searched a little but I didn't find conclusive evidence to know what "const" exactly means.

enter image description here

I found this plot and I'm curious, whether "const" means "cycle", also sometimes described as "drift"?

"trend" should be clear as a time series that is trending up- or downwards is trending. I assume most economic time series to be either constant+trend or constant?

$\endgroup$
  • $\begingroup$ I was going through my old answers and noticed this one was not accepted. Do you perhaps need further clarification? $\endgroup$ – Richard Hardy Feb 24 '17 at 14:12
1
$\begingroup$

I found this plot and I'm curious, whether constant means "cycle", also sometimes described as "drift"?

type=const would include an intercept in the model equations.
I do not see how that could mean "cycle".
It could be interpreted as the drift term if the data supplied to the VAR function is a multivariate time series in first differences (rather than levels).

I assume most economic time series to be either constant+trend or constant?

It is hard to generalize, but I guess you are right (if you mean a VAR with a constant or constant+trend, not just constant or constant+trend). But I am sure it is possible to find some examples where other options make more sense.

$\endgroup$

Your Answer

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

Not the answer you're looking for? Browse other questions tagged or ask your own question.