Issues with using predict on double EGARCH specification #77
Description
Hi,
I am wondering how exactly the predict function works with a doubleEGARCH student-t distribution specification for nahead=1L. I am asking because your original paper references a closed form formula, but I don't find one for double EGARCH in Haas (2004). Moreoever, I am getting an issue whereby every once in a while my rolling double EGARCH model produces infinite one-step ahead volatility forecasts so I am trying to debug. Any help is appreciated.