Subject: | |
From: | |
Reply To: | |
Date: | Fri, 25 Jul 2008 12:59:24 -0400 |
Content-Type: | text/plain |
Parts/Attachments: |
|
|
Hello all,
I have a question about implementing a variant of a regime switching
regression. The preferred model has three regimes and has endogenous
determinants of the regime. In other words, I?m modeling something
like this
Y_t = b_1*X_1 + e_1
=b_2*X_2 + e_2
= b_3*X_3 +e_3
Pr(R=r)=b_4*X_4 + e_4, with r taking on one of three values, which
indicates the current (unobservable) regime. The current regime
determines which of the three parameters is having an effect on our
dependent variable.
What solutions are available for coming up with parameter estimates?
Stata has an implementation for an endogenous two regime approach, and
I?ve heard that GAUSS has some packages which are similar but I?ve
been unable to pin down if they are capable of handling this
particular problem. R does have some regime switching packages,
though they are strictly time-series related. As far as I can tell,
these aren?t flexible enough to model the rhs parts of the equation
(since what we have isn?t lagged values of the dep var). Does anyone
know of a Stata package which can handle the three regime approach?
Or in any other program?
Any help on this problem would be appreciated.
Thanks in advance,
Bill Mac Millan
Dept. of Political Science
University of Michigan
**********************************************************
Political Methodology E-Mail List
Editors: Melanie Goodrich, <[log in to unmask]>
Delia Bailey, <[log in to unmask]>
**********************************************************
Send messages to [log in to unmask]
To join the list, cancel your subscription, or modify
your subscription settings visit:
http://polmeth.wustl.edu/polmeth.php
**********************************************************
|
|
|