Providing a range of values to coef in a state-space model

For technical questions regarding estimation of single equations, systems, VARs, Factor analysis and State Space Models in EViews. General econometric questions and advice should go in the Econometric Discussions forum.

Moderators: EViews Gareth, EViews Moderator

maragloria
Posts: 105
Joined: Tue Jun 29, 2010 7:09 am

Providing a range of values to coef in a state-space model

Postby maragloria » Fri May 24, 2013 1:45 pm

Hi there,

I am estimating a state-space model of output gap for the US. I'm reproducing the results from a BIS working paper (attached).

Here is a simple model to help clarify my question :

Code: Select all

gdpln_us = ystar + c(1)*(gdpln_us(-1) - ystarlag) + 1600*e0 @state ystar = 2*ystar(-1) - ystarlag(-1) + e0 @state ystarlag = ystar(-1) @ename e0 @evar var(e0) = exp(c(20))
The authors restrict the coefficient c(1) to be between 0 and 0.95 and to have a mean of 0.8. How can I accomplish that with Eviews 7?

Thanks!

Mara
Attachments
borio_disyatat_juselius_outputgap_BIS_feb13.pdf
(353.03 KiB) Downloaded 470 times

startz
Non-normality and collinearity are NOT problems!
Posts: 3798
Joined: Wed Sep 17, 2008 2:25 pm

Re: Providing a range of values to coef in a state-space mod

Postby startz » Fri May 24, 2013 2:16 pm

I believe the paper you attached is doing a Bayesian estimation, which EViews doesn't do.

maragloria
Posts: 105
Joined: Tue Jun 29, 2010 7:09 am

Re: Providing a range of values to coef in a state-space mod

Postby maragloria » Fri May 24, 2013 4:59 pm

Yes, he is doing Bayesian estimation. Too bad for me :(

I estimated the model with Eviews and c(1) is very close to 1. I guess that is why he's restricting it, to avoid the unit root. Thanks for you reply anyway.


Return to “Estimation”

Who is online

Users browsing this forum: No registered users and 2 guests