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 at the end of May, and its replacement, statalist.org is already up and running.


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

st: RE: Serial Autocorrelation in Panel Data


From   "Schaffer, Mark E" <M.E.Schaffer@hw.ac.uk>
To   <statalist@hsphsun2.harvard.edu>
Subject   st: RE: Serial Autocorrelation in Panel Data
Date   Mon, 2 Jul 2012 16:21:58 +0100

addyplavo,

> -----Original Message-----
> From: owner-statalist@hsphsun2.harvard.edu 
> [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of addyplavo
> Sent: 02 July 2012 09:32
> To: statalist@hsphsun2.harvard.edu
> Subject: st: Serial Autocorrelation in Panel Data
> 
> Hello everyone,
> I have an unbalanced panel and -xtserial- indicates that the 
> data suffer from autocorrelation. Now I am looking for an 
> answer of how to deal with serial correlation in panel data, 
> as the Newey-West adjustment only seems to work with time 
> series.

But if you have a large-T panel, the kernel-robust approach (Newey-West
et al.) is feasible.

> I have red a previous thread

Can you provide a link to the thread, please?

> that stated I should use 
> the -xtreg, cluster- command and -, mle-. However I am not 
> entirely clear on the exact implementation:
> I used -xtset ID month- to form the panel. Does this mean my 
> cluster variable is ID?

That depends on how you call -xtreg-.  -xtreg, cluster- is not valid
syntax.  You will probably say -xtreg <variables>, vce(cluster ID)-, in
which case the answer is obvious.

> (I observe company returns of 1000 
> companies accross 10 countries, so maybe I should also 
> cluster by country?).

Big problem here.  The consistency of the cluster-robust covariance
estimator is asymptotic in the number of clusters, and 10 is not very
far on the way to infinity.  (Second time today I've used that phrase in
this context, and the umpteenth time in total.  Such a common
pitfall....)

You're probably better off using country dummies (or -areg-, which
amounts to the same thing).  Or, if you have a large-T panel, -xtivreg2-
or -xtscc- with a kernel-robust covariance estimator.

> Furthermore, why should I also use the 
> -mle- specification with my -xtreg- command?

Good question.  I don't understand either why that would be recommended.
In fact, -xtreg- with the -mle- option does not allow cluster-robust
VCEs.

HTH,
Mark

> 
> I am very new to stata and not a statistic pro, thus I am 
> thankful for any help here!
> 
> 
> --
> View this message in context: 
> http://statalist.1588530.n2.nabble.com/Serial-Autocorrelation-
> in-Panel-Data-tp7580214.html
> Sent from the Statalist mailing list archive at Nabble.com.
> *
> *   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/
> 


-- 
Heriot-Watt University is the Sunday Times
Scottish University of the Year 2011-2012

Heriot-Watt University is a Scottish charity
registered under charity number SC000278.


*
*   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