EViews offers sevreal distributions for errors in the ARCH/GARCH estimates.
How to I choose among Normal, t, and Generalized error? and What is the reason for selecting the distribution?
After choosing it, how can i test whether the choice is right?
thanks a lot.
Search found 3 matches
- Sat Sep 10, 2011 11:41 pm
- Forum: Econometric Discussions
- Topic: How to select the distribution of error in GARCH?
- Replies: 0
- Views: 1582
- Tue Aug 16, 2011 3:03 pm
- Forum: Estimation
- Topic: How to wirte the output of ARMA(p,q)?
- Replies: 3
- Views: 3113
Re: How to wirte the output of ARMA(p,q)?
Either of the first two, depending on what you mean by resid in: Yt=b1*Yt-1+b2*Yt-2+C1*resid(-1)+C2*resid(-2)+resid If I were a betting man, reading a strict interpretation of that equation, I'd say you want: y y(-1) y(-2) ma(1) ma(2) Thanks a lot. But I am not sure whether MA(1) MA(2) are able to ...
- Tue Aug 16, 2011 2:50 pm
- Forum: Estimation
- Topic: How to wirte the output of ARMA(p,q)?
- Replies: 3
- Views: 3113
How to wirte the output of ARMA(p,q)?
I wanna run the following regression:
Yt=b1*Yt-1+b2*Yt-2+C1*resid(-1)+C2*resid(-2)+resid.
What should I write in EViews under LS method?
Should it be ...
Y AR(1) AR(2) MA(1) MA(2) ?
OR
Y Y(-1) Y(-2) MA(1) MA(2)?
OR
Ohters?
coz we cannot use Y Y(-1) Y(-2) RESID(-1) RESID(-2).
Thanks a lot.
Yt=b1*Yt-1+b2*Yt-2+C1*resid(-1)+C2*resid(-2)+resid.
What should I write in EViews under LS method?
Should it be ...
Y AR(1) AR(2) MA(1) MA(2) ?
OR
Y Y(-1) Y(-2) MA(1) MA(2)?
OR
Ohters?
coz we cannot use Y Y(-1) Y(-2) RESID(-1) RESID(-2).
Thanks a lot.