State space estimation

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

Posts: 32
Joined: Mon Oct 05, 2009 4:28 am

State space estimation

Postby wolly77 » Tue May 06, 2014 1:05 am

Hi, could somebody please help me clarify this point? Thank you in advance for the help.
I am estimating a state space model of this kind:

@SIGNAL D(log(prod)) = SV1 + c(1)*Gap + [VAR=EXP(C(2))]
@STATE SV1 = c(3)*SV1(-1) + [VAR=EXP(C(4))]

c(1) is a fixed regression coefficient. How is it estimated c(1) in State Space framework? Is it obtained via Maximum likelihood (ML) or OLS?
I think that only unobservable components and variances are estimated via ML, because if I run an OLS of this kind:

D(log(prod)) = sv1 + c(1)*Gap,

c(1) in OLS and in state space are equal. Thank you.

Non-normality and collinearity are NOT problems!
Posts: 3573
Joined: Wed Sep 17, 2008 2:25 pm

Re: State space estimation

Postby startz » Tue May 06, 2014 6:03 am

Maximum likelihood.

But, of course, in some circumstances OLS and MLE are the same thing.

Return to “Estimation”

Who is online

Users browsing this forum: No registered users and 9 guests