What a load of crap. Yes, of course there is only one optimum entry and exit. And there is no way of knowing whether you have hit them except in hindsight. So, you hedge your bet so to speak by scaling in and out. The only time this is not more advantageous, is when you are lucky enough to hit the optimum points.
I think he is trading computerized, backtested systems; that's why he made the statement. If a system is based on historical market action statistics, then, based on this past history there is an optimal entry and exit. If you trade in such a manner, then scaling might be wrong, depending on how you did your statistics. For those who adapt to the market on the fly, scaling is the right thing to do, if the entrys/exits are based on market conditions, not points. If you scale out with preset point values, you give up the advantage of adapting to each trade (otherwise known as letting it run). I mean, if I am short I won't close it up until I see some sort of support. Etc.
...why from your perspective is my assertion a load of crap? I am coming myself from the perspective of unambiguously defining the rules of a trading system, backtesting them with parametric variations, and finding the optimum historical entries and exits. I have yet to see a significantly profitable and robust system which had multiple optima. This means that given the choice of competing systems, you should always bet the full account on the system which has the highest expectation.
All in, all out for an active trader.. though for a fund manager or a long term investor (weekly/monthly+) with huge pockets, scaling and adding is the way to go.. that's the thing with books, they are written by traders in different situations and there will always be conflicting answers.
good stuff. i think there are many variables that we're all mentioning here -- specifically, the type of trading being done. i am currently always in the market, either long or short, and part of my objective is to capture entire trends. therefore, all in and all out only works for me when i nail the beginning and end. that happened often, but what would happen even more often was i would reverse, get stopped out, reverse, get stopped out, ad infinitum until my p/l went from way up to way down. so i decided to have more of a lagging entry (exit) method when the trend changed. by doing this though, i created many situations where i would get some great moves, but then the confirmation would not come until my profit on that trade was almost gone. so then i started taking profits at important s/r levels along the way. 1/3 at first s/r level (initial thrust), 1/3 at second s/r level (decent move), and leave the remaining 1/3 until i reverse at confirmation. i'm tellin ya, for those of you having this problem or something like it, scaling out does wonders for your longetivity throughout the day. of course, for me, there are some deep-rooted psychological problems i have identified and finally have admitted to myself. i would rather have lots of small winners than have big winners and small losers. i know this is not optimal in theory, but it turns (somehow?) to be optimal for my equity curve, because i have better execution throughout the day. go figure. this game really is all about knowing yourself. thanks for your input guys.....
funky - great thread - I like the support the folks are giving you and wish to add my own. If you think about it, scaling in and out of a trending position mimics the market itself. If this method helps you keep your finger straight on the market's pulse without getting so panicked yourself about letting a profitable trade become a loser that you can't hear the patient's heartbeat over your own, then JUST DO IT. I totally agree with you - this game is all about knowing yourself. Calling out the market isn't the whole game - for some its the easiest part, its how you execute what you see/think. Peace,