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]

st: Do i have a sort of nested model? If so, how can I estimate it using Stata? FE and RE.


From   Herman Haugland <[email protected]>
To   Statalist - Send questions <[email protected]>
Subject   st: Do i have a sort of nested model? If so, how can I estimate it using Stata? FE and RE.
Date   Thu, 22 Aug 2013 11:32:09 +0200

Dear Stata listers,

I have a problem, and I would appreciate your inputs.


I have a panel dataset, which is defined as follows:

N: 5 banks
T: 60 quarters

After performing several tests, and accounting for some economic
facts, I found that the model that best represents our data is one in
which we divide the 5 banks in 2 groups.  For that effects, I created
a categorical variable for separating the banks in groups, like this:

Group 1: Banks 1 and 4
Group 2: Banks 2, 3 and 5.

I then estimated the model using -regress- and -areg- (for fixed
effects).  I get very good results that make very much sense with the
theory and intuition.  However, given the fact that my panel is
defined as BANKS and each group is a sample of BANKS, I cannot account
for both effects at the same time.  I can in practical terms, but I
should not do it.  Because of this, I cannot simply run -xt- commands
to estimate my parameters, since, for example, running -xtreg dep ind
i.group, fe- would be accounting for the effects of banks and groups
at the same time.

In order to solve this "double-effects" problem, I can run -areg-
which allows me to specify the effect I want to include in the
-absorb()- option.  However, two problems remain:

1) I don't get the within R2 from the FE regression, just the overall.

2) I have not been able to find a way to calculate RE without using areg.


Therefore, the main questions are:

1) How can I consistently estimate such a model, in which the effect I
want to account for is not the panel id effect, but one of a subgroup
of that id, hopefully using some similar technique to -xtreg-.

2) How can I get the within R2 for my FE estimation?


Thank you for your consideration.




Med vennlig hilsen / Best regards,

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


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