Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

st: RE: Remarks and questions on the VEC function

From   DE SOUZA Eric <>
To   "" <>
Subject   st: RE: Remarks and questions on the VEC function
Date   Sat, 30 Apr 2011 16:05:15 +0200

The second issue has been discussed at length yesterday and the day before.

Eric de Souza
College of Europe
Brugge (Bruges), Belgium

-----Original Message-----
From: [] On Behalf Of Steve06
Sent: 30 April 2011 00:45
Subject: st: Remarks and questions on the VEC function

i've tried Stata's VEC command to estimate a vector error correction model with one co-integrating equation.
Here's what I found:
- all reported significance statistics are Z-statistics, while in published empirical work you find t-statistics. I did not find a switch to change it but would find it a good idea to have.
- if you specify a constant, you have it in both the main model and in the cointegrating equation. If you specify no constant, it's neither in the main model nor in the coint. equation, which is annoying, since you might want to have a constant in the coint. equation but not in the main model.

Any suggestions how to overcome these limitations? Are there alternative VEC functions by the Stata community? (like Xtabond2 vs Xtabond)

View this message in context:
Sent from the Statalist mailing list archive at
*   For searches and help try:

*   For searches and help try:

© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index