Dummy variables and explanatory variables in ARMA/GARCH models
3 views (last 30 days)
Show older comments
Two questions regarding what appears to be significant limitations in the garch functions in the econometrics toolbox:
Is it possible to introduce dummy variables in the variance (GARCH) equation (garchset and garchfit), as might be needed to ascertain whether volatility is lower in summer months than in the rest of the year?
Is it possible to introduce explanatory variables in the GARCH variance equation. This is done in the mean (ARMAX) eqn through the input 'Regress' in garchset, but I have not seen how it can be done for the variance eqn as well.
These types of questions have been posted on Matlab Central for 5 years now, but no answers.
Thank you.
Jeff
0 Comments
Answers (1)
Oleg Komarov
on 20 Jun 2011
Huge limitation and not only this one: the answer is NO to both.
Look here for an unanswered question i posted some time ago: http://www.mathworks.com/matlabcentral/answers/3996-egarch-1-1-with-an-additional-term-in-the-conditional-variance-specification
Another thing you can't do is select some specific AR and AM coefficient (say AR_3 AR_5 etc.. instead of AR1to3 or AR1to5)
0 Comments
See Also
Categories
Find more on Conditional Variance Models in Help Center and File Exchange
Products
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!