What I would like to know is if trader3cnd even knows what Engle's fundamental ideas behind GARCH were... Well trader3cnd? I'll give you a hint, they're often referred to as "Stylized facts of Volatility". What's interesting is that Engle's original paper starts by introducing these stylized facts. Subsequently, via the course of proper albeit sophisticated modeling he *validates* his original hypothesis. What do you think was more important in creating that model? The fundamental concept, or, the resulting fit model? Mike
Yeah, what you were doing is optimization. You weren't basing your system on the parameters, but tweaking inputs based on an original hypothesis. This is acceptable as you realized.