Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down on April 23, and its replacement, statalist.org is already up and running.


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

Re: st: Granger causality test in Stata10


From   Muhammad Anees <anees@aneconomist.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: Granger causality test in Stata10
Date   Tue, 9 Aug 2011 09:24:35 +0500

Our discussion follow the question, how to test the Granger causality
after VEC. Really, I have never tried that. I usually practice it
after VAR which using versions of vargranger. Your querry was about
testing it after VEC. I hope the discussion has led you to guess like
a practioners. I would appreciate it if you just read the Enders:
Applied Econometric Time Series, 2nd Edition. It exposes well how to
follow time series analysis. Also read the time series manual of stata
to check how to do things.

Best wishes

Anees

On Mon, Aug 8, 2011 at 2:59 PM, Merijn Groenenboom
<m.groenenboom@students.uu.nl> wrote:
> Dear Anees,
>
> Thanks again for your help. I really appreciate it that you keep
> helping me. It is very important for me since I am writing  my
> masterthesis and I completed the whole thesis, except for this last
> granger causality test. Furthermore the deadline is almost there, so I
> am a little bit stressed at the moment. So thanks a lot for all the
> help!
>
> I followed all your tips, but I still did not succeed in completing
> the test. I added all the new variales.
> However, I do not think the problem was with the lags. Even when I
> only use the following command after the vec model:
> test (d.lexr=0)
> I get an error with: regressor d.lexr not found.
>
> Are you sure you can do granger causality test with a VEC model in
> Stata? I have the feeling that it is not possible to use the command
> function test after creating a VEC model.
>
> Thanks in advance!
>
> MerijnG
>
> 2011/8/8 Muhammad Anees <anees@aneconomist.com>:
>> Sorry to mention, use the previous command with the newly created
>> lag1xvar and lag2xvar etc.
>>
>> On Sun, Aug 7, 2011 at 9:19 PM, Merijn Groenenboom
>> <m.groenenboom@students.uu.nl> wrote:
>>> Thanks again Anees! I feel I am very close now...
>>>
>>> My vec regression now is:
>>>
>>> vec d.lstp d.lexr l.d.lexr  l2.d.lexr d.lsint l.d.lsint l2.d.lsint
>>> d.lip l.d.lip l2.d.lip d.lur l.d.lur l2.d.lur d.lyie l.d.lyie
>>> l2.d.lyie d.loil l.d.loil l2.d.loil d.lcon l.d.lcon l2.d.lcon d.lgold
>>> l.d.lgold l2.d.lgold d.ltb l.d.ltb l.2.d.ltb d.linf l.d.linf
>>> l.2.d.linf
>>>
>>> After that I want to test for example lexr, as you recommanded:
>>>
>>> test (d.lexr=0) (l.d.lexr=0) (l2.d.lexr=0)
>>>
>>> However then Stata10 sais:
>>> d.lexr not found
>>>
>>> How is this possible?
>>>
>>> Thanks in advance for your reply!
>>>
>>> MerijnG
>>>
>>> 2011/8/6 Muhammad Anees <anees@aneconomist.com>:
>>>> For example you want to include lags into the system, your command should appear
>>>>
>>>> vec yvar d.xvar1 l.d.xvar l2.d.xvar d.xvar2 l.d.xvar2 l2.d.xvar and so on
>>>> test granger causality using
>>>>
>>>> test (d.xvar1=0) (l.d.xvar=0) (l2.d.xvar=0) and so on for other set of xvars.
>>>>
>>>>
>>>> On Fri, Aug 5, 2011 at 3:21 PM, Merijn Groenenboom
>>>> <m.groenenboom@students.uu.nl> wrote:
>>>>> Thanks for the help Muhammad!
>>>>>
>>>>> Only one more question about the command with the lags:
>>>>>
>>>>> Lets say my y variable= d.lstp and my x variables are d.lexr and d.lgold.
>>>>> I use a lag of 3 for both variables.
>>>>> What should I type in Stata?
>>>>> vec d.lstp d.lexr.......
>>>>> How can i include lags in the middle of the regression? What command
>>>>> should i use?
>>>>>
>>>>> Thanks in advance for the help!
>>>>>
>>>>> 2011/8/5 Muhammad Anees <anees@aneconomist.com>:
>>>>>> Granger causality is tested using F or Modiffied Wald Chi2
>>>>>> statistitic. You can try the -test- command on the respective
>>>>>> variables after your regression command.
>>>>>>
>>>>>> vec yvar xvar1with lags xvar2withlags xvar3withlags
>>>>>> test xvar1withlags
>>>>>>
>>>>>> will give you the causality statistic for xvar1 and its included lag.
>>>>>> This way you have to keep mind the routine of hypothesis for Granger
>>>>>> causality
>>>>>>
>>>>>> Hope It explains the answer
>>>>>>
>>>>>> Anees
>>>>>>
>>>>>> On Thu, Aug 4, 2011 at 1:07 PM, Merijn Groenenboom
>>>>>> <m.groenenboom@students.uu.nl> wrote:
>>>>>>> Dear Statalisters,
>>>>>>>
>>>>>>> I am writing a research about the effect of economic indicators on stock prices.
>>>>>>> Granger causality test is one of the tests I want to use.
>>>>>>> I started with the following command in Stata10:
>>>>>>>
>>>>>>> quietly var d.lstp d.lexr d.lsint d.lip d.lur d.lyie d.lm1 d.loil d.lcon
>>>>>>> d.lgold d.ltb d.linf, lags(1/2)
>>>>>>> vargranger
>>>>>>>
>>>>>>> However the results showed spurious results, since all variables are
>>>>>>> only stationary at first differences and cointegration exists
>>>>>>> I should use a vec model instead of a var model.
>>>>>>> So I used the following command:
>>>>>>> vec d.lstp d.lexr d.lsint d.lip d.lur d.lyie d.lm1 d.loil d.lcon
>>>>>>> d.lgold d.ltb d.linf, lags(1/2)
>>>>>>>
>>>>>>> However I am not sure about the next step. How does Stata show me the
>>>>>>> results of granger causality test now?
>>>>>>> Or did I already make a mistake before?
>>>>>>>
>>>>>>> Thank you.
>>>>>>>
>>>>>>> Merijn G
>>>>>>> *
>>>>>>> *   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/
>>>>>
>>>>
>>>>
>>>>
>>>> --
>>>>
>>>>
>>>> Regards
>>>>
>>>> Anees
>>>>
>>>> *
>>>> *   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/
>>>
>>
>>
>>
>> --
>>
>>
>> Regards
>>
>> Anees
>>
>> *
>> *   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/
>



-- 


Regards

Anees

*
*   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 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index