Is Scaling Out a Fallacy?

[Deringer]

Scaling in and out is a form of disciplinary money management.

You certainly don't want to put all your eggs in the same basket at the same time when you're choosing some top or bottom because, unless you have absolutely certain information, you could always be wrong. If you're wrong, you would be carrying an immediate load of risk, but if you're right, you still get to ride it and gradually add or subtract to the position. If you're a good trader, you shouldn't need to jump all in or out drastically anyway because you should be catching a general trend that's functional. Accepting a bunch of singles rather than trying to hit one giant homerun seems like a lot safer strategy, or at least that's what I personally would prefer. It just contributes to a lot more consistency.

 

[fader]

i agree with one of the above posts - scaling out is as much of an integral part of a strategy as entry, so asking if it's a fallacy is akin to asking if entry is a fallacy...

you need to analyze if scaling out makes your strategy more profitable or not and then decide if it works for your strategy

personally, i have considered scaling in/out in developing my strategies, however, i have found that fixed size works better when you look at the numbers

also, psychologically i prefer fixed size vs fixed size entry + scaling out (i am not considering scaled entry here) because if i get stopped out, then it's always on full size but the profits may be only on partial size if i don't get any runners...

 

[kiwi_trader]

fader,

That "I lose the whole lot but make on only a portion of the bet" is a van tharp argument (ie not as good as vt thinks it is imnsho). Intuitively it feels right but as is often the case in trading our intuition may lead us to a conclusion that is statistically incorrect. Treat each scaling in/out as a separate trading strategy; evaluate them separately; then evaluate them as a portfolio. Thats the only way to see the impact on ones ideal performance.

(Not saying you didnt do this - just objecting to the vt argument which may or may NOT be true, depending on the situation and whether its eod vs day trading and how close you are to moving the market)

Then if you take into account the reduced psychological pressure involved in such a strategy you may find that it also helps move your real performance closer to your ideal performance.

If you havent read it yet you might look at acrary's thread --- there's a man who understands money management and trading objectives.

 

[fader]

thx kiwi - i have gone through acrary's, ericp's and a few others' material some time ago, excellent material, very useful indeed.

 

[prt_systems]

Quote from cashonly:

I was considering some of my strategies this evening and was wondering about the wisdom or lack thereof of scaling out of positions.
.....
Cash

I have an algorithm that I use to decide how to scale out and how to scale in based upon the character of the particular market. Of course the idea is to catch more of the move - my current system does better than a linear scale out/in and that was the goal. If you could perfectly (or near perfectly) identify moves then you would not need to scale in/out. Let me know if you find that system

 

[FredBloggs]

Quote from cashonly:

I was considering some of my strategies this evening and was wondering about the wisdom or lack thereof of scaling out of positions.

Let's take the simplest example. When it reaches a certain target profit, we take half the position off the table, looking to take the other half off later.

On the one hand, if the trading strategy was solid, why not just wait until you get to the point where you'd close the second half and have that much more profit?

On the other hand, if you need to unload half, maybe you should be closing the whole position.

From a dollars and cents standpoint, is this really the best approach. The more I think about it, the more it seems like a psycological panacea. We feel good that we took a profit, and then if we make even more of a profit on the 2nd half, we feel better (maybe not as good as if we had not scaled out). Or if the stock starts retracing, and we close at a lesser profit or even a loss, we feel good that we got some profit out of the trade rather than a much lower profit or even a loss.

But, it would seem that one or the other is more optimal, and I question whether I should really be scaling out or if I should try to nail down which is the truly better approach.

Any thoughts?

Cash

good post.

yes it is a psychological crutch as others have pointed out.

but lets not forget that a healthy psychological make up is key to success.

so, inother words, what ever makes you happy

 

[jason_l]

so far, nearly every system I have backtested did better w/o scaling (entry or exit).. I realize one can't prove a negative, and that it may simply be the strategies I have migrated to don't do as well with scaling.. but over the past few years this has been my experience.
It seems just setting a stop where I would first scale out, then letting it all "ride", has had better results for me. Again though, this is may very well be dependent on ones trading methods/strategies.
I can't help but be drawn to the concept though, but it just doesn't seem to work for my methods.

 

[darkhorse]

I've thought about this from a longer term perspective (not daytrading).

In addition to the psychological element, which is subjective, I think scaling can have value by smoothing out the volatility of returns over time, while still accounting for the uneven way in which the market distributes profit.

Swing and position traders might have a handful of trades that make their year, standing out dramatically in relation to the other 90 percent of trades.

If you press hard every time waiting for the rare big score, you will give up a lot of short run profit in reversals / setbacks etc. and potentially get yourself mired in extended drawdown.

On the other hand, if you always take profits at a predetermined point, you automatically disqualify yourself for those occasional huge wins--and it becomes much harder to exploit the trends that literally run for months (like VLO or FNM or RGLD etc). In my opinion the market does not distribute potential profits equally--like earthquakes and forest fires, trends tend to follow a power law. Rigidly predetermined exits ignore the potentially exploitable effects of that power law.

Scaling thus serves a few useful non-psychological purposes for the swing / position trader:

By taking partial profits at a predetermined exit point, you smooth your equity curve and better avoid long droughts.

By maintaining a partial position, you stay in place to exploit the sweetest trends, and retain the option of re-adding to your existing profitable position if the timing is right.

Furthermore, if you use a 'pig at the trough' approach to your available capital, scaling out frees up money for new positions without requiring you to give up on profitable current trends.

You simply give yourself more options.

 

[cashonly]

Quote from jason_l:

so far, nearly every system I have backtested did better w/o scaling (entry or exit).. I realize one can't prove a negative, and that it may simply be the strategies I have migrated to don't do as well with scaling.. but over the past few years this has been my experience.
It seems just setting a stop where I would first scale out, then letting it all "ride", has had better results for me. Again though, this is may very well be dependent on ones trading methods/strategies.
I can't help but be drawn to the concept though, but it just doesn't seem to work for my methods.

Yeah, that's what I would *think* would be the case.

Once you set that stop, I take it you trail it as the market moves in your direction?

Cash

 

[prt_systems]

Quote from jason_l:

so far, nearly every system I have backtested did better w/o scaling (entry or exit).. I realize one can't prove a negative, and that it may simply be the strategies I have migrated to don't do as well with scaling.. but over the past few years this has been my experience.
It seems just setting a stop where I would first scale out, then letting it all "ride", has had better results for me. Again though, this is may very well be dependent on ones trading methods/strategies.
I can't help but be drawn to the concept though, but it just doesn't seem to work for my methods.

There are a number of situations in which scaling in or out is quite beneficial and beats all or none entries and exits. I use this technique when appropriate and not, as some others suggest, as a crutch. The strategy is part of a fully automated trading system. It is used only when appropriate. Under other conditions single entry and exit points are used.