st: GARCH with dummy variables

 From Katia Bobulova To statalist@hsphsun2.harvard.edu Subject st: GARCH with dummy variables Date Tue, 12 May 2009 17:25:40 +0100

```Dear All,

I am trying to estimate a GARCH model to capture price volatility of
securities and I am using daily return.
Now, I want to see if there is a particular behaviour in a specific
period of time and I use a dummy variable (d_D1).

I formulated my GARCH in this way but I am not sure if I have used the
right procedure:

arch return ,het(d_D1) arch(1) garch(1)

ARCH family regression -- multiplicative heteroskedasticity

Sample:  2 - 5309                               Number of obs      =      5308
Wald chi2(.)       =         .
Log likelihood =  14948.95                      Prob > chi2        =         .

------------------------------------------------------------------------------
|                 OPG
return |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
return       |
_cons |  -.0000342   .0001505    -0.23   0.820    -.0003291    .0002608
-------------+----------------------------------------------------------------
HET          |
d_D1 |  -.3115308   .0415504    -7.50   0.000     -.392968   -.2300936
_cons |  -10.24161   .0779456  -131.39   0.000    -10.39438   -10.08884
-------------+----------------------------------------------------------------
ARCH         |
arch |
L1. |   .2284142   .0095525    23.91   0.000     .2096916    .2471368
garch |
L1. |   .6932624    .010922    63.47   0.000     .6718557    .7146691
------------------------------------------------------------------------------

Another thing that I would like to know is how to perform in Stata10
the Ljiung-Box test to decide the order of the GARCH.