If I am holding a strong hand in a market, I do not fold part of the hand, I hold my max position to reap the full benefit. In addition, I get the whole position in the market as early as possible.
scaling out is just an attempt to increase sample size in order to smooth out the equity curve. In it's own scaling out does not provide an edge.
Here is a post from another thread that supports much of what I have said is the reason for scaling-- http://elitetrader.com/vb/showthread.php?threadid=64965&perpage=6&pagenumber=10854
this argument is truly a joke. b1s2 knows he is insane. Everyone who has backtested multiple systems knows if you can blend systems you can have much better results. While the scaled part of the trade might have a better smoother curve (because you will have a higher percent right) ... you do not even have to argue the scale can do better. Even if the second system is not as good as the first ---- blending systems can lead to a better, more profitable and less risky program. you have to be a novice to argue against scaling.
We're well past the proof stage of this thread. All in , All out has already been established as superior over the long haul. Please address the points currently being discussed.==Ishmael
Such a concept could never be established. That has got to be a blatant lie or foolish ignorance. Given real life real time trading only a non trader could try and make such a silly assertion. But even given your assertion... you tell me your given trade amount and your percent right on your trades. If you get it get it right more than 50% of the time without hitting your stop I will tell you to trade larger and scale out one tick into profit once you recover costs? Boom you lose. (In reality you would hold out for half your normal trade or 1 atr or something but my example proves you are incorrect. Or at the very least it proves you could never prove your point.) If your trades are losers more than 50% of the time, I guarantee there will be scale outs that make more money for you. Boom you lose. Do you trade with real money? Do you know your winning percentages. Have you consider stats such max adverse excursion and max profitable excursion. If you did look at such stats you would know how absurd your claim is.
The opposite is the case. This thread has shown that scaling out is optimal in many situations. Furthermore, there has been not one iota of proof offered for the OP's assertion, aside from the OP's statement that it is so. Apparently the OP believes that if he says 'Scaling out is inferior behaviour', this constitutes a rigorous proof. The OP is a failed daytrader who has made the following statements on these boards 1. "Daytrading is inferior behaviour". 2. "I can identify the extent of a price move before the move occurs (the 'optimal exit point'). I can do this through backtesting". 3. "I know the best entry and exit protocols for every conceivable trading technique". 4. "In addition to scaling out being sub-optimal, scaling in is also sub-optimal" It would take a very long post to enumerate the errors in thinking and conception contained in the OP's posts. Flexibility and adaptability are the most important traits in a trader. Blanket statements such as the ones made by the OP should be a warning to beginners. This type of thinking will eventually lead to ruin. The OP's assertion is 100% based on the mistaken notion that the optimal exit point can be determined before a trade is placed. This is the central tenet of the thread. As this is not possible, scaling out will be optimal in those situations in which price action shows the trader that the conditions which obtained when the trade was placed have changed.
Unless you're operating on a very fast timeframe, why on earth would you not scale in and out as price action evolves? You trade on a huge timeframe; how can you know whether the move will go 50pts and last two months, or go 500pts and last a year?
This is only relevant when time is considered. Unless one is operating on a timeframe where things are moving too quickly for good decision-making, whether you make five decisions per week or five per year is totally irrelevant. You might as well claim that it's more "optimal" to win five times out of ten than fifty times out of a hundred, when there's clearly no difference.
The assumption here seems to be that 'scaling out' means exiting fixed portions of the position at fixed, arbitrary points, such as 'breakeven,' '$1000 profit' etc. I can certainly believe that kind of strategy is suboptimal, since the market doesn't move according to arbitrary PnL reference points.