Well, I don't know about other people, but I prefer not to have to guess if, or for how long, a price move will continue. That is why I use "quasi-predetermined" D ) scaled exits. It's either that or eye of newt.
Certainly the more confidence that people develop in their trading beyond merely careful, cautious optimism is a result of the "eye of newt" factor in one form or another. That's my opinion. As an aside, that's a rather interesting name you have in view of the topic of discussion.
I use the name "gnome" in the sense of the subterranian dwarf in folklore who guards the treasure....
Sorry, just having some fun here. Obviously, with our reference to "eye of newt," we are implying an element of witchcraft and sorcery (wizardry?) in connection with some trading methods and their perceived performance reliability. Since a gnome is a mythical creature...oh, never mind. In any event, I assure you that there was no offense intended, particularly since we are essentially in agreement here.
thanks. I'll give it a try anyway. My goal isn't to forecast the timeseries. Rather, I am just looking for a small but real statistical advantage within a short time window. anyway, I have used numerical recipes quite a lot before. I'm still trying to get used to developing in this non-scientific windows environment -- I find myself having to spend a lot of time on stupid winapi/MFC issues rather than the numerical algorithms themselves....
oh one more thing. are you saying that all patterns in the stock market are not real? if you admit that there are good (human) traders out there, then that must mean there are some real trends that they are picking out. Otherwise we must conclude that all good traders are merely lucky.
Your right, momentum strategies (trend following) profits are made on 20% of their trades. The focus on these types of strategies is to start by minimize your losses during directionless phases of the market which occur 80% of the time.