Hi Tom,
You make some very good points. Perhaps I am way off base trying to compete with indexes. Certainly, if one does comparisons against buy&hold, really valuable findings will result. My trials with ISIP using evolutionary tuning suggested that, while AIM out of the box (AIMOOTB?) did better than b&h, a bit of tuning did even better. Jack J's numbers suggest the same.
Let me resubmit my notion of a competition: one for best under AIMOOTB, and one with tuning -- perhaps that can fall into sub categories (Vealies, IW, tuning) and (anything goes -- including additional attributes, as I did in my experiments). My talk is intended to explore the latter, but I think some kind of competition would be fun as well.
Thinking that a bit further, it might be useful to gang up and design a data presentation form -- something along the lines of those published in S&C where you show all the factors (e.g. #trades, #buys, #sells, draw downs, etc etc etc). Sure, we don't do all the things conventional traders do, but we look to achieve the same bragging rights.
Say what?
Jack |
