Page 1 of 1
OLS Estimation and Level of Data
Posted: Wed Nov 26, 2014 7:22 am
by alim
Dear TomDoan
Trust you are well and helping people like me. I need your quick help. Suppose, I have four time series data. They are all non-stationary at level but 1st difference they are stationary and they are co-integrated at level then, what level data we use for OLS estimation ( at level or 1st difference).
I would be highly glad if you kindly reply as early asible.
Thanking you
with regards
Ali
Re: OLS Estimation and Level of Data
Posted: Wed Nov 26, 2014 8:05 am
by TomDoan
This is such a common question (this is probably the 4th almost identical forum thread) that we address it in the version 9
User's Guide:
Before you run unit root tests on your data, you might want to ask yourself why you
are doing it. There have been many papers published that have included unit root
tests even though there was nothing in the paper that actually depended upon the
result of the tests. Worse than pointless unit root tests, we have had more than a few
users who are almost paralyzed by the (incorrect) idea that you can’t run regressions
when some of the variables are “I(1)”. It’s true that static regressions (ones without
lags) on series involving unit roots have a good chance of being “spurious” (Granger
and Newbold, 1974), but regressions that have proper handling of dynamics (lags)
such as vector autoregressions (Chapter 7) and ARDL’s (page UG–63) are generally fine.
Let your regression be driven by the economics, not by the unit root tests.
Re: OLS Estimation and Level of Data
Posted: Thu Nov 27, 2014 3:17 am
by alim
Dear TomDoan
Thank you for your posted mail. I like to know my case if I want to estimate the model by using OLS then, what level of data I will use. You know that VAR or VECM do not produce result as OLS produces. That the reason I need your advice what level data will I use for estimation?
I am looking your best help in this regard.
Thanking you
with kinds
Ali
Re: OLS Estimation and Level of Data
Posted: Thu Nov 27, 2014 8:19 am
by TomDoan
You seem to be missing the point. Please read that quote carefully. You're asking me about a model where all you've described is the I-ness of the data, nothing about the content. What does the economics tell you to do?