"Scaling out" is inferior behavior

[Moneyball]

Quote from Buy1Sell2:

It depends on how often it runs to 125 or 150. If it runs there often enough, then one of those should really be the 100% target and the optimal target has been calculated wrong from the beginning. Optimal targets for setups can be calculated by anyone using backtesting etc. As you know, I use trailing stops with full position on.

Exactly- in this hypothetical world you're living in, you can change what ever figures you want after the fact, but in reality, we can't trade in hindsight. No one knows what the extent of a move is going to be until the moves are over. Scaling allows you to hedge your bets against miscalculation of the optimal target (among many other things). The optimal target is always calculated wrong in reality, unless of course you always pick the exact bottoms and tops.

Backtesting? Please, tell me that's not the main way you set your optimal targets .

 

[Buy1Sell2]

Quote from Moneyball:

Backtesting? Please, tell me that's not how you set your optimal targets .

I use trailing stops with no scaling as I have mentioned many times before. And yes, backtesting is certainly one of the factors involved in determining optimal targets. I do trade in the real world, by the by.

 

[smilingsynic]

Quote from volente_00:

but the fact remains is that no one knows exactly what the market what will do so scaling gives you the opportunity to exit half at what you may think will be optimal target and let the trade run versus missing not selling any at the optimal point and letting it come back to your trailing stop and giving up extra profit.

If no one knows exactly what the market is going to do, then why scale out? Why not sit tight until a trailing stop takes you out, thereby affirming change of trend?

 

[smilingsynic]

Scaling out winners before the target price only makes sense if one also scales out of losers before the stop.

If one scales out of winners and does not scale out of losers, then one is skewing the original risk/reward ratio, turning a 2/1 into something less.

If I scalped stocks for nickels, dimes, and quarters, I might feel differently, I'll admit, since I know little about scalping stocks.

 

[Moneyball]

Quote from Buy1Sell2:

I use trailing stops with no scaling as I have mentioned many times before. And yes, backtesting is certainly one of the factors involved in determining optimal targets. I do trade in the real world, by the by.

This isn't going to get very far if you continue to just respond to very small parts of posts while ignoring the bulk of them. I know what you do (at least what you say you do), that wasn't the question. FYI, it doesn't make the way you do it the best or only way to trade.

Why not quote and respond to the entire post? Do you really think you can precisely determine the optimal entry and exit points in advance, using backtesting or any other methods? No one can with any consistency.

Smiling, I do use trailing stops, just not always for the entire position at once. Having your trailing stop get hit doesn't necessarily affirm a change of trend, it only does so according to your set parameters for that particular trade. Since no one knows exactly how the market is going to behave at any moment in time, you can't say for certain that a certain level or price absolutely changes the trend, not to mention it all depends on the time frame you're talking about. Have you ever seen a huge spike in the opposite direction of the general trend in the market, only to see it resume it's longer term direction shortly after? Trailing stops aren't fool proof either, nor do they tell you "exactly what the market is going to do".

Also, most people that I know that scale out of winners also scale out of losers, at least to some extent. Yet another reason why B1S2's example is purely hypothetical and optimized to fit his argument (similar to optimizing backtesting), most people that scale don't do so in the increments or at the levels that he gave. Thanks for at least acknowledging that it may make sense for some traders to scale, unlike your friend here .

 

[jagmot]

Quote from Buy1Sell2:


Four ES Contracts 50% win ratio versus Four ES Contracts 50% win ratio scaling out at half target.

9 pt target 3 pt initial stop loss

1st example with 20 trades
10 winners for 9 X (4 conracts) = 360 pts ($18000)
10 loser for 3 X (4 contracts) = 120 pts (-$6000)
Net profit $12000


2nd example with 20 trades
10 winners for 9 X(2 Contracts)=180 pts ($9000)
10 winners for 4.5 X(2 Contracts)=90 pts ($4500)
10 Losers for 3 X(4 Contracta) =120 pts (-$6000)
Net profit $7500

Money can be made scaling out, but it is inferior behavior.

The numbers can easily be manipulated. Here is an example

2nd example with 20 trades

10 winners for 9 x (2 contracts) = 180 pts ($9000)
16 winners for 4.5 x (2 contracts) = 144 pts ($7200)
4 losers for 3 x (2 contracts) = 24 pts (-$1200)
5 losers for 3 x (4 contracts) = 60 pts (-$3000)
Net profit is $12000

Since you are scaling out that win rate is 60% vs 50% on the scale out trades. I think that is a fair and reasonable assumption.

Both examples are now equally well off with $12,000 in profits.
Would you like me to create another example that the scaler is better off?

Either way, you can see my earlier post about the subject. I belive that scaling out is inferior once you have backtested your system with thousands of actual trades over many different market cycles. In the meantime, the system that I discuss on this board I scale out because it has only been around for 1 year and doesn't have enough actual trade data.

Another system that I've traded for over 7 years does not scale out. I have thousands of trades over many different market cycles and know the optimal point to exiting my trade.

 

[jagmot]

Quote from jagmot:

The numbers can easily be manipulated. Here is an example

2nd example with 20 trades

10 winners for 9 x (2 contracts) = 180 pts ($9000)
16 winners for 4.5 x (2 contracts) = 144 pts ($7200)
4 losers for 3 x (2 contracts) = 24 pts (-$1200)
5 losers for 3 x (4 contracts) = 60 pts (-$3000)
Net profit is $12000

Since you are scaling out that win rate is 60% vs 50% on the scale out trades. I think that is a fair and reasonable assumption.

I'm surprised no one has commented on my math error. Didn't realize it until I went for a walk to clear my head and thought about it more.

 

[romik]

Quote from jagmot:

I'm surprised no one has commented on my math error. Didn't realize it until I went for a walk to clear my head and thought about it more.

That's OK, I mailed you a calculator

 

[hilojack]

I'm not going to read every post in this thread, but the bottom line is this: size players scale out, pikers do not. It is nearly impossible to pick absolute tops or bottoms, size players use areas of accumulation and dispersion, not trying to hit exact figures. Also to consider is the timeframe that one is trading in, typical length of trade, and overall methodology. To each their own.

With a name like buy1sell2, its pretty obvious there is nothing to scale out. Come back and repost when u change your name to buy3sell6 and let us know what you find.

 

[Joab]

Quote from Buy1Sell2:

(borrowed line from George Bush). No sense being a weak hand.

I actually started to agree with your post till you quoted the knucklehead.

I have scaled out 1/2 most of my career but I'm now thinking it may not have been the smartest way to trade.

I'm actually going to experiment with keeping all till my 2nd targets and see if that works better.

I don't have the balls to scale in