[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]
RE: st: cointegration
> there is no command for cointegration as such. One defines
> an OLS regression and then tests the residuals to ensure
> they are white noise- if they are you're regression is a
> co-integrating relationship.
> On the other hand Johansen's method of testing for
> co-integration is better but I am not aware that STATA 6
> has the capability to test for co-integration using
> Johansen's method. In fact I'm pretty sure it has no
> Johansen 'command'.
> Use another package, EViews for example if you want to use
> Johansen's method otherwise you're stuck with OLS and
> testing residuals
A search using -findit- using keywords such
as cointegration or Johansen finds several
items, some of which predate Stata 6.
* For searches and help try: