I am running TGARCH, EGARCH and PGARCH to test calendar effects using Stata but it is not producing out put for some returns , just giving iterations and at the end following message appear" flat log likelihood encountered, cannot find uphill direction".
Please help and recommend how to resolve it and why such problem occur, if anyone has expertise in this domain
This can simply mean that a constant equally well explains your data. But you also may tinker with max number of iterations or convergence criteria.
You might use Eviews software, it is more user friendly. There are available video tutorials in the youtube. Thanks....
Try to use another distribution or change from ML to QML, another trick is to reduce the number of ARCH effects till you make the model to converge ! Farah Naz I hope I am useful
Olumide Olaoye
yes that's true, I used to have the same problem, but as Md Bokhtiar Hasan stated sometimes it would be better to change the softwareBest,
Adelajda
- Badges
- Science topic
- Similar topics
- Phytochemistry
- Extracts