st: Panel SUR with Unbalanced Observations

[Chris Parker <cparker.phd2007@london.edu>]

Hi statalist,

I have quarterly data on 24 brokers for two exchanges over a span of
5.5 years. I am trying to explain the allocation of the brokers orders
by a number of other variables. The regression I would like to run is
as an SUR with two equations, one for each exchange. However I run
into two problems. First one exchange opened after another so the
equations do not have the same number of observations. I attempted to
solve this problem using advice from Allen McDowell:
http://www.stata-journal.com/sjpdf.html?articlenum=st0079 This is
where I enter the second problem which is that I have multiple
brokers. The final step before running xtgee requires a tsset which I
cannot do because I have multiple observations per time period
corresponding to the different brokers. I think I need to wide my data
so that in each exchange-time observation I have all of the
independent and dependent variables. However I am not sure if this is
what needs to be done. Any help understanding this process would be
greatly appreciated.

Additionally, I noticed a message from Kit Baum in previous posts that
I think is what I need.
http://www.stata.com/statalist/archive/2007-11/msg00346.html Kit, do
you still have this stata code available? I would really appreciate a
copy if possible.

Does anyone else have an idea on how to run this regression?

Chris

Chris Parker
________________________________

PhD Candidate | Management Science & Operations
London Business School | Regent's Park | London NW1 4SA | United Kingdom
Direct line +44 (0)20 7000 8816 | Email cparker.phd2007@london.edu

*
*   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/