User-defined VAR?

[donihue]

It is not (yet) possible to estimate a BVAR in EViews, although it can be done in other programmes (eg, RATS or Dynare). It is possible to import the parameters into the Eviews "system" object and thence obtain a "model", but it would be very helpful if one could import the parameters estimated in these outside programmes in such a way as to use the very nice functions which EViews has to handle VARs. This does not yet seem to be possible in EViews7; is there any plan to permit this in the future?
Regards
Donihue

[EViews Gareth]

I'm not quite sure I follow what you're asking here.

[startz]

In VARs, and other places it would be nice to be able to choose a value of a coefficient rather than have it estimated, and then still use the provided post-estimation tools. (At least those that don't require standard errors and the like.) For example, it would be nice to be able to find impulse responses.

[donihue]

Yes, thank you Startz: that is exactly what I was asking...
Regards
Donihue

[donihue]

Coming back to this exchange: we now have the excellent BVAR add-ins, for which those of us working in this field are extremely grateful.

However, to quote Startz, it is still the case that:

In VARs, and other places it would be nice to be able to choose a value of a coefficient rather than have it estimated, and then still use the provided post-estimation tools. (At least those that don't require standard errors and the like.) For example, it would be nice to be able to find impulse responses.

Are there any plans to provide this functionality?

Regards
Donihue

[EViews Gareth]

It is on our todo list