Bookmark and Share

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


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

Re: st: RE: Cointegration analysis including exogenous variables


From   Syed Basher <[email protected]>
To   [email protected]
Subject   Re: st: RE: Cointegration analysis including exogenous variables
Date   Wed, 9 Mar 2011 02:15:30 -0800 (PST)

Thank you Eric for the reference, luckily I have a copy of the book. Seems like 
I am fortunate for not having an intervention dummy in my VAR model, my dummy 
variable does not cumulate to trend, so asymptotic distribution is less likely 
to be affected, but not necessarily the finite sample distributions. Interesting 
read, thanks again.




----- Original Message ----
From: DE SOUZA Eric <[email protected]>
To: "[email protected]" <[email protected]>
Sent: Wed, March 9, 2011 12:38:49 PM
Subject: RE: st: RE: Cointegration analysis including exogenous variables

If you have access to Juselius' textbook on cointegration, see Chapter 8, 
sections 3 and 4 for a brief discussion of how dummy variables can affect the 
distribution of the rank test statistic. The chapter and section references are 
from the manuscript which is in my computer. The book is in the our library.


Eric de Souza
College of Europe
Brugge (Bruges), Belgium
http://www.coleurope.eu


-----Original Message-----
From: [email protected] 
[mailto:[email protected]] On Behalf Of Syed Basher
Sent: 09 March 2011 10:25
To: [email protected]
Subject: Re: st: RE: Cointegration analysis including exogenous variables

Dear Beat,

I still didn't find a convincing answer to Eric's question. I recently estimated 
a VECM model in Eviews where I used a binary dummy variable as an exogenous 
regressor, without really knowing how that affect the rank statistics.

Syed



----- Original Message ----
From: Beat Hintermann <[email protected]>
To: [email protected]
Sent: Wed, March 9, 2011 12:00:26 PM
Subject: Re: st: RE: Cointegration analysis including exogenous variables

Dear Syed

thank you very much for your offer to estimate it in Eviews. I might take you up 
on it, but only if I don't manage to estimate it myself. I have access to Eviews 
in our finance lab, I've just never used it. But now that I know that Eviews 
allows for exogenous variables, I might give it a shot. If I can't get it done I 
might give you another shout.

Thanks again,

Beat


On 08.03.2011 17:24, Syed Basher wrote:
> This may not be relevant for Stata users, but the problem you 
> described is
easy
> to handle in other packages. For example, Eviews has an option to 
> include exogenous variables in cointegrating model. If you do not mind 
> sharing your data, I can attempt to estimate it for you in Eviews.
>
>  Syed Basher
> Doha, Qatar.
>
>
>
>
> ----- Original Message ----
> From: Beat Hintermann<[email protected]>
> To: [email protected]
> Sent: Tue, March 8, 2011 6:53:13 PM
> Subject: Re: st: RE: Cointegration analysis including exogenous 
> variables
>
> i get your point. but i'm working with monthly weather measures, not 
> long-term climate trends. i think it's safe to say that monthly 
> heating-degree days and precipitation events are exogenous to today's 
> fuel and electricity prices.
>
> exogenous or not, they are I(0) and therefore not part of the 
> cointegrating relationship. how should I take them into the model? any 
> idea?
>
>
>
> On 08.03.2011 16:35, Nick Cox wrote:
>> Treating the weather as "truly exogenous" is not the best current 
>> science, although I imagine you don't want to build climate change into your 
>>model.
>>
>> Nick
>> [email protected]
>>
>> Beat Hintermann
>>
>> I would like to estimate the cointegrating relationship between 3 
>> I(1) variables, but in the presence of exogenous I(0) variables.
>> Specifically, I want to estimate the long-term relationship between 
>> electricity, gas and coal prices, but taking into account exogenous 
>> weather shocks (like heating degree days or precipitation, stuff that 
>> is truly exogenous).
>>
>>    From what I gathered in the Statalist archive, it is not advisable 
>>(and  perhaps even impossible) to include exogenous and/or I(0) 
>>variables into  the vec model framework. But since the weather 
>>definitely influences  electricity and fuel prices, I don't see how 
>>ignoring this information  will give me the best result.
>>
>>
>> *
>> *  For searches and help try:
>> *  http://www.stata.com/help.cgi?search
>> *  http://www.stata.com/support/statalist/faq
>> *  http://www.ats.ucla.edu/stat/stata/
>

--
Beat Hintermann
Assistant Professor
University of Basel, Faculty of Business and Economics (WWZ) Peter Merian-Weg 6, 
4002 Basel, Switzerland Tel. +41 61 267 3339
------------------------------------------

*
*  For searches and help try:
*  http://www.stata.com/help.cgi?search
*  http://www.stata.com/support/statalist/faq
*  http://www.ats.ucla.edu/stat/stata/



      

*
*   For searches and help try:
*  http://www.stata.com/help.cgi?search
*  http://www.stata.com/support/statalist/faq
*  http://www.ats.ucla.edu/stat/stata/

*
*   For searches and help try:
*  http://www.stata.com/help.cgi?search
*  http://www.stata.com/support/statalist/faq
*  http://www.ats.ucla.edu/stat/stata/



      
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/


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