1

Looking to estimate a VARX(p,q) type VECM in R if possible.

I'd like to estimate a VECM with p lags (lags relative to the level, not diff of the vars) on the endogenous variables and q lags on the exogenous components. Any ideas? At the moment I'm using the ca.jo() function and adding the contemporaneous and lagged exogenous variables to the dumvar= option. I do not feel confident about this approach.

Can I do this in R? Should I switch software?

1 Answers1

4

If your variables are exogenous and their lags are exogenous, then there is no problem in treating them as dummy variables. So you can add them to dumvar option. My answer to similar question applies.

Community
  • 1
mpiktas
  • 33,140
  • 5
  • 82
  • 138
  • There is no need to merge -- this is an example of "good duplicate" –  Jan 24 '11 at 15:22