"Scaling out" is inferior behavior

[romik]

I still think scaling out MAY be superior at times. Another example:

Going for 10 points using 2 lots, 1 lot scaled out after +2, 1 lot scaled +10 or close at entry for b/e and non-scale out goes for 10 points and also closes b/e if price fails to run and returns to entry.

1st Case:

Scale out:

1 lot x 2pts x $50 x 5 times = $500
1 lot x 0pts x 4 times = $0
1 lot x 10 pts x $50 x 1 time = $500
Total: $1000


Non-scale out:

2 lots x 10 pts x $50 x 1 time = $1000
2 lots x 0pts x 4 times = $0
Total: $1000

EVEN OUTCOME

2nd Case:

Scale out

1 lot x 2pts x $50 x 10 times = $1000
1 lot x 0pts x 9 times = $0
1 lot x 10pts x $50 x 1 time = $500
Total: $1500

Non-Scale out:

2 lots x 10pts x $50 x 1 time = $1000
2 lots x 0pts x 9 times = $0
Total: $1000

SCALE OUT OUTPERFORMS

3rd Case:

Scale out

1 lot x 2pts x $50 x 5 times = $500
1 lot x 0pts x 3 times = $0
1 lot x 10 pts x $50 x 2 time = $1000
Total: $1500

Non-Scale out:

2 lots x 10pts x $50 x 2 times =$2000
2 lots x 0pts x 3 times=0
Total: $2000

NON-SCALE OUT OUTPERFORMS

Conclusion:

Everything is dependent upon different variables including randomness of future price action. Sometimes scale outperforms, sometimes non-scale out outperforms and at times they would be even. Therefore, it is impossible to be certain which would outperform, perhaps just by making hypothetical examples like the ones above.

 

[romik]

There are many other scenarios which can support either strategy, especially when stops come into play.

 

[romik]

'Scaling in' can be superior to 'all in' too, depending on strategy used. Example would be starting with 1 contract and ONLY adding to winning trades another 1 contract whenever price increases (if long) by 1 point, after +8 points you would have 8 contracts on with different gains opposed to the 10 contracts when 'all in'. Clearly 'all in' is ahead here. But here is the catch. When you are all in your stop would be -2 points x10 contracts, whereas 'scale in' strategy would be down -1.5 points x 2 contracts. That way you greatly reduce risk when wrong, yet still beat 'all in' strategy on both sides - risk management and total gains generated.

 

[Buy1Sell2]

'Scaling in' can be superior to 'all in' too, depending on strategy used. Example would be starting with 1 contract and ONLY adding to winning trades another 1 contract whenever price increases (if long) by 1 point, after +8 points you would have 8 contracts on with different gains opposed to the 10 contracts when 'all in'. Clearly 'all in' is ahead here. But here is the catch. When you are all in your stop would be -2 points x10 contracts, whereas 'scale in' strategy would be down -1.5 points x 2 contracts. That way you greatly reduce risk when wrong, yet still beat 'all in' strategy on both sides - risk management and total gains generated.

Doesn't make sense---- If adding contracts, they could all still go against you and have larger losses per contract.

 

[Handle123]

I think it one concentrates on lowering losing percentages, then more prudent to average down and this way you can risk half as much and target 50% more on this position. Whereas adding as market becomes profitable is increasing risk if based on charting support/resistance and targeting less, never made any sense to me

 

[Buy1Sell2]

I think it one concentrates on lowering losing percentages, then more prudent to average down and this way you can risk half as much and target 50% more on this position. Whereas adding as market becomes profitable is increasing risk if based on charting support/resistance and targeting less, never made any sense to me

Both are sub-par. Both will having you missing the full move when you are right.

 

[Handle123]

Both are sub-par. Both will having you missing the full move when you are right.

Well I been averaging down past three years, I do not recommend others to do so as losing trades are multiplied, but you do enough back testing, numbers show it is well worth it. BUT you have to be able to trade like a robot, like no mistakes and don't bet the farm. And no way I am going to miss the full move as I have entered where I am supposed to enter on original entries, but also enter at -1,-2,-3,-5. Really comes down to how much back testing one does and how far back one does and not a couple months or years. Only time I have a target is very long term trading of 5/10k as I might get couple of them when markets swing to go other way to make new highs/lows, might as well as take something out, but day trading risk to reward is often inverse and getting out early makes little sense unless larger target is so large enough to make numbers right for all the times it doesn't work when smaller target is best.

 

[K-Pia]

In these circumstances it is best to not scale in/out (futures):

You are crossing the market in a liquid contract and your size is so small to have no market impact.
You are providing liquidity far from the current market price.


Scaling can have the following benefits:

Smooths PnL which allows for trading larger size in general
Reduces market impact


One could make the argument that if you are not trading enough size to move the market then you are trading the wrong product. I have found that for providing liquidity on a variety of products that scaling in using 3 clips is optimal.

Doesn't it add the benefit to make someone convex ?
You lose little against adversity and win a maximum while fortunate.

 

[Xela]

Authors like Van Tharp, Tushar Chande and others certainly seem anxious to explain that scaling out without having scaled in is, overall, sub-optimal trade-management (and it's not difficult to see why, arithmetically). I prefer to add to winning positions, not that that's got anything to do with the price of bouillabaisse.

I've never tried (or fancied) averaging down, myself. I need my trades to move in my favor pretty quickly, really, otherwise my reason for entry is probably invalid and I don't want to be in them. But I see that other methods might suit people with very different trading styles from mine.

Apparently there's 160 pages of this thread - I won't pretend to have read all of it, before replying. So I'm probably only repeating comments made earlier on (though I'll bet nobody else specifically mentioned bouillabaisse).

 

[murray t turtle]

Both are sub-par. Both will having you missing the full move when you are right.

%%
Most likely commenting on adding+ scale out.
That can be true. Dr Tharp tends to be right....
But as a practical matter, a good trend has much more than one excellent entry + exit point.
As a practical matter, no one knows the exact bottom except in hindsight; bottoms takes more than one candle, [except in hindsight; no replies needed for the elite that says he always picks the bottom-LOL]

BUT that is one area- dont be real slow about some exits.............., on an old, old trend LOL