EViews.com

How I can write my GARCH (1.1) with the dummy variable?
I have the following data set. now I need to calculate GARCH (1.1) for twice,
one for date 03/10/2001 to 31/10/2001 with dummy 0 and
one for date 03/03/2001 to 31/03/2001 with dummy 1
The parameters value should be positive. Any one please help me, how I can do this in eviews 8? Thanks in advance.

Date Returns Dummy
03/10/2001 -0.002396294 0
04/10/2001 -0.02741479 0
06/10/2001 0.034219338 0
07/10/2001 -0.006785041 0
08/10/2001 -0.000753952 0
09/10/2001 -0.005050092 0
10/10/2001 -0.000497682 0
11/10/2001 -0.00760143 0
13/10/2001 -0.010963988 0
14/10/2001 0.005626831 0
15/10/2001 -0.006540975 0
16/10/2001 -0.007122104 0
17/10/2001 0.007598369 0
18/10/2001 -0.004446422 0
20/10/2001 -0.003892107 0
21/10/2001 -0.003987656 0
22/10/2001 0.000157636 0
23/10/2001 -0.006712036 0
24/10/2001 0.002604159 0
25/10/2001 0.007706424 0
27/10/2001 -0.002223391 0
28/10/2001 -0.002940081 0
29/10/2001 -0.00580369 0
30/10/2001 -0.004650034 0
31/10/2001 -0.001271651 0
03/03/2001 0.011669299 1
04/03/2001 0.001787059 1
05/03/2001 0.005022924 1
10/03/2001 0.009637503 1
11/03/2001 0.014444066 1
12/03/2001 -0.002406239 1
13/03/2001 -0.007744283 1
14/03/2001 -0.00123507 1
15/03/2001 0.009525963 1
18/03/2001 -0.001185963 1
19/03/2001 -0.008058648 1
20/03/2001 0.005204527 1
21/03/2001 -0.004414948 1
22/03/2001 -0.009820858 1
24/03/2001 -0.006261181 1
25/03/2001 -0.004073736 1
27/03/2001 0.002368852 1
28/03/2001 0.006408405 1
29/03/2001 0.001048676 1
31/03/2001 -0.00310402 1

Which bit are you having trouble with?

1) You have only 45 observations, which can be problematic for GARCH analysis.
2) There are large gaps in your data (i.e. jump from March to October).
3) Since you are using a dummy variable, you do not have to run the analysis twice.
4) Negative parameter values are not much of a problem, unless they are significant.

I guess you are trying to conduct an event study. If so, I suggest you to revise your approach and methodology.

Which bit are you having trouble with?

I do not know how I can run this in E-views. Can you please let me know what steps should I follow to run the GARCH (1.1) with dummy in eviews?

1) You have only 45 observations, which can be problematic for GARCH analysis.
2) There are large gaps in your data (i.e. jump from March to October).
3) Since you are using a dummy variable, you do not have to run the analysis twice.
4) Negative parameter values are not much of a problem, unless they are significant.

I guess you are trying to conduct an event study. If so, I suggest you to revise your approach and methodology.

Hi Trubador, thanks for your answer. Yes, I am doing an event study. I used over 100 stocks to get this sample average return. I like to compare march-April (pre-event) data with October-November data (post-event) to see how volatility changes. so I need to run the model twice with two different dummy variable. I can not change the approach as I am calculating the volatility and I like to see how GARCH (1.1) perform.

First I calculated the return for 100 stocks and then made an average which you see in the data. Can you please let me know what steps I should follow to calculate them?