My 2 cents: I think both scaling in and scaling out are inferior. If you're scaling in, it means you're timid about your entries and you need to work that: find better entries. If you're scaling out, it means you're timid about your exits and you need to work on that: find better exits. It's all about timing, folks: when to get in and when to get out. That's your edge. Money management comes in to wring the maximum advantage out of your edge, but it doesn't create your edge. Scaling in and scaling out are ad hoc money management gimmicks to try to overcome weak entries and weak exits. If that works for you, OK, but let's not pretend like it's better than having strong entries, strong exits and optimal position sizing. Let the flames begin.
I did. I'm not looking for certitude, I'm looking for maximum profits. The folks looking for certitude are those who are trying to maximize their win rate. Trust me, I'm not one of those people. I have no problem with shutting down a losing trade if I'm confident in the overall system that I've backtested the hell out of. And a main point of optimal position sizing is to minimize your potential loss, as long as you stick to the tested strategy. Losses are inevitable. I think scaling out is an attempt to maximize win rate rather than maximize profits. To each his own.
Scaling out on winners ensures lower exposure on the inevitable big winners. It would make more sense to scale out of the losers, because that would ensure lower exposure on the inevitable big losers. Better yet, get out the losers, and move on, getting back in the hunt for winners. Psychologically, it is harder to let winners ride than to cut losers short. No one wants to be in that 95% camp of the losers, so they cut short winners out of fear they'll turn into losers. Cutting losers short is not enough. One must have the courage to let winners ride. Few can do that; hence, few are winners. This, btw, is a cognitive bias referred to as loss aversion. Perhaps it should be called loss IMMERSION, since in trading terms it will lead in the long run to drowning in losses. In the end, they want to win less than they don't want to lose, and ironically, they end up losers anyway.
And because of the uncertainty associated with the markets and price movement, irrespective of how good you think your entries are, I think that single exits are a bit like playing a lottery with undue confidence. No matter how good of a marksman you are, if you want to hit a target under foggy conditions, then I think it's better to have a few smaller bullets rather than just one bigger bullet. Yes.
It's not undue if you use optimal position sizing. That's what it's there for. The problem is that scaling really (I mean REALLY) complicates your position sizing. Instead of a single return per trade, you're now juggling two or more returns per trade. Give me any sequence of single trade returns and I can give you an ongoing updated optimal-fraction sizing of your trading account for the next trade. It's a lot harder to do that with scaling out, and even more so if there's in-scaling in different proportions than the out-scaling.
I think I now understand where you're coming from. And I disagree. I believe you are placing far too much confidence in your mathematical specificity. Good luck with that. Optimal-fraction sizing? Are you a Ralph Vincer? If you think you can model trade size with such clarity and confidence and apply such specificity to future trades with back tested performance, then I think you are due for a surprise that has not visited upon you just yet. Remember the definition of a calculator: a device that allows you to take two seat-of-the-pants estimates, multiply them, and get accuracy to the 8th decimal point. Past performance is at best an educated guess at future performance, not that much better than an informed "seat-of-the-pants" estimate. Too much emphasis on the word "optimal" in this context is a joke. Just my opinion, of course.
Ralph Vince? He couldn't optimal position size his way through wet tissue paper, at least not based on the swill he's selling to the public. Tell me, what are you using for position sizing if you don't believe in trying to optimize it? I'm honestly curious.
A very small fraction of my account size. I might vary it slightly, depending on the quality of the setup, but that's mostly just to make myself feel important.