Imposing non-negative constraints for TGARCH/GJR-GARCH(1,1)
Posted: Mon Apr 24, 2017 11:04 am
How do I go about doing this. The estimation on Eviews when I change TARCH to 1 produces a negative Alpha 1 coefficient. The literature says it should be >0. How do I get around this. Need an answer asap if possible. Thanks.
To be clear this is the RESID(-1)^2 coefficient I am talking about.
To be clear this is the RESID(-1)^2 coefficient I am talking about.