ARDL in Eview 9 and 10.

[NipNip]

In Eviews 9, after estimating an ARDL model, it is easy to distinguish between the short term coeficients and the long term coeficients.
In eviews 10, however, I have problems with the short term dynamics. I mean the short term coeficients are the one in the "ECM regression" or the ones in the "Conditional error correction regression" (The ones that are not differenciated).

Also, are "Unrestricted error correction model" and " conditional error correction model" synonymous?

Can anyone please clarify please.

[EViews Mirza]

Please refer to our three part blog post on ARDL estimation. Virtually all of your questions will be answered there if you take the time to study the material carefully.

Part 1: http://blog.eviews.com/2017/04/autoregr ... -ardl.html
Part 2: http://blog.eviews.com/2017/05/autoregr ... rdl_8.html
Part 3: http://blog.eviews.com/2017/05/autoregr ... -ardl.html

Also, I strongly urge you to use EViews 10 since it is indeed up to date and improves on the features of previous version of EViews.

[NipNip]

So. ARDL short/long run coeficients and bound test before Eviews 10 are wrong?

[igor]

So, you need to update your Eviews 9 Currently, long- and short run coefficients and the bounds test are the same in Eviews 9.5 and 10, although the latter has some additional features.

[NipNip]

LOL. I used its results over Stata on my thesis thinking that they were better.

[NipNip]

Did it in Excel! :D