Thank you statalist,
Yes I am using xtabond2 with the following command line: xtabond2 CR L.CR L(0/1).(MCY NE G) yr*, gmm(L3.(L.CR L2.MCY L.NE)) ivstyle(L(0/2).G yr*) two robust. Unfortunately, I have not come accross any literature that explains on how I can use either the lag() or collapse options.
Grant.
-----Original Message-----
From: [email protected]
To: [email protected]
Date: Fri, 14 Dec 2007 18:59:51 +0100
Subject: Re: st: Problematic Result with two-step System GMM.
You may find a reply if you provide exactly the command-line you used and the output from which you guessed that you have too many instruments. From your e-mail, I guess you are using -xtabond2- and you need to work with lag() option to reduce the number of instruments, or collapse option. Read carefully the help file and author's working paper referenced in there.
Nicola
At 02.33 11/12/2007 -0500, Grant Peter Kabango wrote:
>I am using two-step System GMM for my estimation, with a panel which has 20 industries observed over 35 years (i.e.T=35 and N=20). However, the results show that I have too many instruments. I have been trying to collapse the data but I do not know how to go about it. Pleasehelp.
>
>Further, my estimated coefficients carry opposite signs in levels and in the first lag, for each variable. Is this normal or there is something I could be doing wrongly.
>
>Please Help
>
>Grant Kabango
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
