Autocorrelation in Eviews

[kjb]

I recently ran a model with an AR(1) term in Eviews and in SAS. The coefficients were different. (1) How does Eviews estimate rho and (2) how does it transform data in the presence of an AR(1) term for reruns?

In addition, the predicted values from the data and the estimated coefficients are different from the forecasts generated by Eviews. How can I use the estimated coefficients and the data to arrive at the forecasts (and backcasts) generated by Eviews?

[EViews Gareth]

http://forums.eviews.com/viewtopic.php?f=7&t=465

[kjb]

QMS Gareth,
I followed your suggestions but still found differences in the Eviews generated forecasts (static and dynamic) versus the Excel generated one. If you send me an email address, I will send you the details. You have my email address. Alternatively, you can tell me where to send the details.

Here is why it is necessary for me to be able to replicate in Excel what Eviews generates as forecast. My company has SAS and I am trying to introduce them to Eviews. At present I am the only having Eviews. (I also have SAS). The forecasts generated from Eviews with AR(1) correction do not match what the estimated coefficients suggest when you do the fitting in Excel, unlike in SAS. If I do not convince them that the results from Eviews are not wrong, because that is what they currently think, I may not be allowed to use Eviews for any serious work. I like Eviews and I want to keep it and also promote the use of it as I did in my previous employment. Please help.
Thanks

[EViews Gareth]

You can email support@eviews.com, along with your EViews serial number.