Path: news.mathworks.com!not-for-mail
From: "Sebastian Arslanogullari" <sebastian.arslanogullari@robur.se>
Newsgroups: comp.soft-sys.matlab
Subject: Re: Estimation of a vector error correction model
Date: Wed, 4 Nov 2009 09:45:02 +0000 (UTC)
Organization: Robur AB
Lines: 16
Message-ID: <hcrieu$rn1$1@fred.mathworks.com>
References: <hcqf13$l7$1@fred.mathworks.com>
Reply-To: "Sebastian Arslanogullari" <sebastian.arslanogullari@robur.se>
NNTP-Posting-Host: webapp-05-blr.mathworks.com
Content-Type: text/plain; charset="ISO-8859-1"
Content-Transfer-Encoding: 8bit
X-Trace: fred.mathworks.com 1257327902 28385 172.30.248.35 (4 Nov 2009 09:45:02 GMT)
X-Complaints-To: news@mathworks.com
NNTP-Posting-Date: Wed, 4 Nov 2009 09:45:02 +0000 (UTC)
X-Newsreader: MATLAB Central Newsreader 69570
Xref: news.mathworks.com comp.soft-sys.matlab:582303


Hi Chris
I've worked quite a lot with the var_bvar library in the spatial econometrics toolbox in the past. My experience is that the functions work quite well and are sufficient if you want to estimate simple VAR/Error correction models. 

If you want to estimate slightly more complicated models you will find that you will have to write your own versions of e.g. the ecm function. Some examples include models where you want to estimate the error correction variables yourself instead of using the Johansens ML method and/or if you want to use additional exogenous variables. If this is the case, I suggest that you use the ecm as a starting point and change it according to you needs.

So in summary, I think that starting out with ecm is smarter than writing your own code from scratch.
Regards
Sebastian Arslanogullari
"Chris" <AzizC0rn@bluestring.com> wrote in message <hcqf13$l7$1@fred.mathworks.com>...
> Dear Matlab users,
> 
> I was wondering if anyone has estimated a Vector Error Correction model (e.g. between two cointegrated variables) based on the code of spatial econometrics of Johansen's cointegration test (http://www.spatial-econometrics.com/coint/johansen.m)
> 
> Any suggestions on how to implement/start with will be greatly appreciated,
> 
> Thank you