Adding Restrictions/Equation Identity

For technical support, tips and tricks, suggestions, or any other information regarding the EViews model object.

Moderators: EViews Gareth, EViews Moderator

fmramos
Posts: 86
Joined: Fri Sep 19, 2008 7:30 am
Location: Sao Paulo / Brazil
Contact:

Adding Restrictions/Equation Identity

Postby fmramos » Thu Feb 19, 2009 6:03 am

Hi! I'm a Brazilian user of EV.

I have a model with some equations and identities. The format is something like:

eq1 = f (X1, X2, X3)+e1
eq2 = f(Y1, Y2, Y3)+e2


I'd like to add the restriction/identity:

eq1=eq2

And, of course, cov(e1,e2)<>0.

But I can't add 'eq1=...' twice (recognized as endogenous by the model specification). The problem is analogous to the demand and supply problem.

Someone has any idea how can I proceed?

Tks in advance!

Fabio

startz
Non-normality and collinearity are NOT problems!
Posts: 3491
Joined: Wed Sep 17, 2008 2:25 pm

Re: Adding Restrictions/Equation Identity

Postby startz » Tue Mar 24, 2009 2:04 pm

fmramos wrote:Hi! I'm a Brazilian user of EV.

I have a model with some equations and identities. The format is something like:

eq1 = f (X1, X2, X3)+e1
eq2 = f(Y1, Y2, Y3)+e2


I'd like to add the restriction/identity:

eq1=eq2

And, of course, cov(e1,e2)<>0.

But I can't add 'eq1=...' twice (recognized as endogenous by the model specification). The problem is analogous to the demand and supply problem.

Someone has any idea how can I proceed?

Tks in advance!

Fabio


you might try creating a series

Code: Select all

series z = 0

and then adding as an identity in the model

Code: Select all

z=eq1-eq2


But I haven't tried it. :)

fmramos
Posts: 86
Joined: Fri Sep 19, 2008 7:30 am
Location: Sao Paulo / Brazil
Contact:

Re: Adding Restrictions/Equation Identity

Postby fmramos » Tue Mar 24, 2009 5:32 pm

Yeah. Tks for the idea. I confess that i already thought about it. But it works only 'a little'. The model'll forecast a value to the next steps different than zero (but close to)... So, the two sides will still be different. But I think that's the better until now...

This solution is useful, but in fact is a deterministic restriction 'converted' into a stochastic one, I think.

Tks again anyway!,

Fabio


Return to “Models”

Who is online

Users browsing this forum: No registered users and 1 guest