Hello,
My first question is theoretical: 1) I am estimating EGARCH models via Gaussian and Student-t QMLE. I want to conclude, which one is more suitable for my model. Is it correct to use AIC and BIC for model comparison or are there other metrics/ways to answer that?
2)As long as I estimate QMLE and then I use the estat ic command, will that yield the AIC and BIC metrics or should I compute them separately, since STATA calculates the pseudo log-likelihood of the model?
Thank you in advance,
My first question is theoretical: 1) I am estimating EGARCH models via Gaussian and Student-t QMLE. I want to conclude, which one is more suitable for my model. Is it correct to use AIC and BIC for model comparison or are there other metrics/ways to answer that?
2)As long as I estimate QMLE and then I use the estat ic command, will that yield the AIC and BIC metrics or should I compute them separately, since STATA calculates the pseudo log-likelihood of the model?
Thank you in advance,
