Cointegration & error correction model

[Apple Tan]

Hi,

Using Stata 13.

I have posted a similar question at

http://stats.stackexchange.com/quest...rrection-model


But anyway, after finding out that (x, y) are cointegrated, do I run the usual OLS or use a error correction model? If the latter, how do I get a error correction model and run a regression on it?

I believe some of you do know the answer, please help.

Thank you

[Sebastian Kripfganz]

See the Stata command vec for vector error-correction models.

In the context of a single-equation model, an ARDL approach might be what you are looking for:
ARDL in Stata