Hello,
I am conducting a volatility study based on DCC-GARCH methodology. I built 3 empirical models based on standard GARCH, EGARCH and GJR-GARCH. I defined the goodness fit indices but as the image below, the post-invasion period models confuses me.
How can I choose the best fit model?
Tihana Škrinjarić
Hi,
I cannot decide which model has the best suit among Post-invasion models.
You can rely on the information criteria, e.g. AIC or SBC. The lower, the better.
Bayes is the SBC and it is well-matched to the big sample size.
Nadiya Gandalialikhani In my paper, my total observation is 2464 which I used for the last modes (DCC(1,1)-GARCH, DCC(1,1)-EGARCH, DCC(1,1)-GJR-GARCH). The other dataset, I divided by events dates, for example, pre-pandemic has 1720 observations, post-pandemic has 571 observations, and post-invasion has 228 observations.
- Badges
- Science topic