"Scaling out" is inferior behavior

[DHOHHI]

Quote from Buy1Sell2:

This one's pretty far off topic, but someone else is welcome to respond to it.

Not off topic at all ... you said "Certainly a good trader can determine optimal exit points" and "Yes, I am able to do that".

Consider that we trade in penny or sub-penny increments these days. Thus, the optimal exit is nearly impossible to compute ... even if you were running it on a Cray. Variables are just that and at times unpredictable. So, to suggest that a human can find such optimal points. Sure, someone can back test with some mechanical solution and compute optimal solutions. But that's after the fact ... with historical data. To be able to look into the future and calculate optimality just isn't possible.

 

[Buy1Sell2]

Quote from traderNik:

For me?

I am not smart enough to define the optimal exit point before I place a trade. In my method, I need to observe price action after the trade is on and then determine the optimal course (add, remain flat or reduce, all decisions underpinned by total exposure and correlation rules). This is how I trade. I am not even smart enough to determine the optimal exit point after I enter a trade, especially if the trade is going my way, although good old S & R seems to work pretty well for this.

This is how I trade in general. I typically do not use profit targets , rather I let trades run until they reverse, then I exit the full position. The discussion about optimal exit comes up because I mentioned that some people who use profit targets could benefit from better analysis of the "normal" moves in a market to place their profit targets. I don't use profit targets generally (loosely I may), because I want as much of the full move per commission that I can obtain. You and I trade somewhat alike. I use trailing stops below reaction lows/highs and I stay in. Many times I am trading the weekly chart and adding positions on spikes etc until the chart tells me no.

(For me?)-- Quite possibly since you are in the group of people known as traders and everyone is faced with the same issues.

 

[Buy1Sell2]

Quote from DHOHHI:

Not off topic at all ... you said "Certainly a good trader can determine optimal exit points" and "Yes, I am able to do that".

Consider that we trade in penny or sub-penny increments these days. Thus, the optimal exit is nearly impossible to compute ... even if you were running it on a Cray. Variables are just that and at times unpredictable. So, to suggest that a human can find such optimal points. Sure, someone can back test with some mechanical solution and compute optimal solutions. But that's after the fact ... with historical data. To be able to look into the future and calculate optimality just isn't possible.

What would perhaps make your original post somewhat on topic would be the part about defining optimal. Earlier in the thread , I mentioned that I can define the area to get out. This is what is meant here. (Someone else began using the word optimal). My main point with this thread, is that the strategy of scaling out is subpar/inferior. No one knows the exact tick of the optimal exit, only the area.

 

[kiwi_trader]

Buysell,

You miss that there may be two "optimal areas." I used to trade the euro on 5min charts and there was a "this thrust" probable range and a "major thrust" probable range.

So, every thrust, I set half my position to exit on reaching the first target. I snugged and trailed stops for the other half. On retracements I would add another half unless the stop on the "free" part was so high that I could add still more "without risk"

This was a very workable approach to scaling out and back in. To evaluate the efficacy of such an approach I suggest treating it as two different trade setups --- evaluate each to determine whether it is worth pursuing. One has a rr in the 2:1 range with slippage and commission and a 70% win rate. The other is more like 4:1 and 45%. Trading the combination is both easy and profitable.

 

[OddTrader]

Quote from illiquid:

Thunder, you beat me to it.

There's always a possibility of a wide gulf between real-time vs theoretically "optimal" and "correct", etc. It's like asking a chartist to look at a chart, and tell you what he (obviously) "would have" done, then trying to replicate the results.

But really it comes down to how much faith one has in the past in the first place, and how the past is used in one's trading. "Backtests" are inferior criteria to trade off of -- now that would be an interesting thread.

"Leap of Faith? - How much you've attained (so far) vs backtesting profits?"
http://www.elitetrader.com/vb/showthread.php?s=&threadid=79206

 

[GTS]

Quote from Thunderdog:
1) It may be sub-optimal after the fact, but a priori? Perhaps I should consider my scaling out as a form of diversification within a single position. Diversification is, by definition, a sub-optimal strategy if you know full well in advance where specific markets are going to go. Few of us do. I am not among them. Therefore, a priori, I consider scaling out as prudent, and therefore optimal in an environment of uncertainty.

I hope I didn’t claim that the trader can find the absolute optimal exit point (if I did that was a mistake) I agree that can only be found after the fact.

I was trying to allude to the fact that the person that scales out has already chosen two different exit points (mid-point and final point) and one of them is clearly superior to the other and that that trader should learn over time which one is better and then going forward exit the full position at that point (which ever one turns out to be more profitable overall).

I was not trying to imply that this superior exit point is in fact the "optimal" exit point - it is simply the best choice given the 2 possible exit points that a given trader had already established. So I am not asking the scaling out trader to change exit points or hold-out for bigger moves then he otherwise would.

I understand your point about diversification (in general) but do you really think diversification is appropriate when talking about a trades within a single strategy? E.g. you don't think past trades using that strategy are a sufficient basis for reaching an educated conclusion about which exit point tends to be better, negating the need to diversify between the two via scaling out?

 

[Thunderdog]

Quote from GTS:

...you don't think past trades using that strategy are a sufficient basis for reaching an educated conclusion about which exit point tends to be better, negating the need to diversify between the two via scaling out?

I think past trades can provide an indication or an idea, but not a straight answer. Therefore, I'm disinclined to bet entirely on a single number. And make no mistake, it is betting.

 

[illiquid]

Quote from GTS:

I understand your point about diversification (in general) but do you really think diversification is appropriate when talking about a trades within a single strategy? E.g. you don't think past trades using that strategy are a sufficient basis for reaching an educated conclusion about which exit point tends to be better, negating the need to diversify between the two via scaling out?

I don't think diversification (in general) is the main purpose behind scaling out, at least not for me. I think the argument for and against scaling revolves mainly around what methodology and rationale exists behind the trade and the trader. Some traders have an "ironclad" strategy that they stick to no matter what; others admit to not even knowing where their exit point will be after they enter, incorporating price action and other variables on the fly. It's the latter camp which operates more comfortably in fluid or "grey zones" that seem to condone partial or staggered exits, while the former will stick to predetermined setups and expectancy, determined by long-term averages of past trades. How much weight you give the past on influencing how you feel about the current trade seems to be the single overriding factor.

I really don't understand the justification for the claim of "inferior" -- just read through the Schwager interviews again and you'll see that more condone scaling and staggering entries and exits than those that don't, across all time frames, markets, methodology. I don't have the Lipschutz interview in front of me, but he gives his rationale for stepping into and out of trades that is pretty much the opposite of what b1s2 assumes scaling is all about.

 

[Cutten]

I think all traders would agree that one's position size shoud reflect two things - the amount of risk one is prepared to tolerate, and the opportunity presented by the market in question. Let's assume we have decided our level of risk tolerance - the only thing then left to do is to evaluate the current risk/reward available in the trade.

If the trade is presenting a great risk reward, you should have a larger position. If the trade is good but not great, you would have an average size position. If the trade is acceptable but has some flaws (e.g. higher than normal uncertainty), then you would have a small position.

So, if we accept that position size should vary according to opportunity, "scaling out" not only makes sense, but is a requirement, in certain situations. If you have a position that started out great, but has now degraded somewhat, then you should reduce your position size. You may have started out long 20 lots, but now the risk has increased, the volatility is expanding, so the appropriate position size may only be 10 lots or even 5 lots. The expectation is still positive, but the risk is higher and/or the probability of success is lower and/or the reward has decreased.

Anyone who says position sizes should remain fixed is basically saying that all market opportunities have identical risk/reward profiles, and furthermore, that the risk/reward of a position does not change over its lifetime. I certainly don't agree with that. Look at any market and volatility fluctuates over time. If volatility increases, then other things being equal, you should reduce your position.

Thus - scaling out makes perfect sense in some cases.

 

[DHOHHI]

Quote from Buy1Sell2:

What would perhaps make your original post somewhat on topic would be the part about defining optimal. Earlier in the thread , I mentioned that I can define the area to get out. This is what is meant here. (Someone else began using the word optimal). My main point with this thread, is that the strategy of scaling out is subpar/inferior. No one knows the exact tick of the optimal exit, only the area.

Actually one can make the case that there may be many optimal exits ... much depends on the time frame of the trade. A day trade would have a different optimal exit than a position trade where the expected time in the trade is a week for example. And so on ...