Protecting Open Profits

[nizar]

When trading long-term trend following systems, as many on here already do, because of the wideness of the stop to allow for the big winners, the trade-off is often giving back alot of open profit at the end of the trend (open equity drawdown).

One way to perhaps overcome (?) this is to have a profit stop. Using tradeSim i can set a profit stop eg. exit after 20/50/100% profit is reached.

What I think may work better is a profit stop where only a percentage of the position was sold to lock in profit, but the remainder left to ride and will be exited when the trailing stop is hit.

So Exit X% of the position after Y% profit has been reached?

What do others think of this idea? I think its at least worth testing.

TradeSim manual makes it clear how to set Profit Targets, but how to set up to sell only an X% of the position is beyond me.

Any help here with the coding would be much appreciated.

 

[BBand]

Hi Nizar,
I like your post topics.

Protecting profits.
If we use the same logic to protect profits as we do to protect capital - how does this work out on Tradesim?
e.g. if we are prepared to risk say 2% of capital on our initial stop (protect capital), then maybe we should follow thro and use the same philosophy (2%) to protect profit, only give back 2% of profit?

Wide stops work well in strong trending markets - in range bound markets perhaps the above will be more pertinant?

I do not like giving back profits, but then again I am a discrectionary trader, I consider changing my exit plan if the market tells me to. Sometimes it works perfect, other times I come out too early (whats new). If I come out too early, I can always re-enter, no problem. (unless I think my money can be used more fruitfully elsewhere).

Good trading
Peter

 

[rub92me]

So Exit X% of the position after Y% profit has been reached?

What do others think of this idea? I think its at least worth testing.

Conceptually that will only yield better results if you immediately reinvest in a different position that has a better return than the one you had (tax and brokerage issues aside). So you'll get better diversification of your portfolio, but doubtful you'll be able to get better results. Definitely worth testing though!

 

[nizar]

Conceptually that will only yield better results if you immediately reinvest in a different position that has a better return than the one you had (tax and brokerage issues aside). So you'll get better diversification of your portfolio, but doubtful you'll be able to get better results. Definitely worth testing though!

Correct.

After giving it some more thought, this may not as well as I initially thought when posting.

Because in my system that Im developing now, i pyramid trades. So for champion stocks that keep running (ie.they make new highs and trigger another buy), ill effectively be selling cheaper, and buying it back at a dearer price.

Peter thanks for your thoughts, you may some good points.
ie. why arent we as prudent in protecting open profits as we are at protecting initial capital?

I think generally we need to give a trend room to develop so 2% may be too tight, but definately an area worth exploring.

Now... for the Tradesim code....

 

[BBand]

Hi Nizar,
"2% may be too tight" - depends on the underlying structure of the trade (trending or range trading)

Protecting open profits:
Other instances where it may prove financially beneficial to consider adjusting exits include:
If you are fortunate enough to be in a trade that goes parabolic - if you do not use some means of protecting profit - then you may give ALL of it back to the market when the parabolic trend ends!!
Parabolic trends tend to collapse explosively when they go vertical.

If we are in a trade that has a blow off top - again the least we should do is tighten our exit stop.

Pattern trades: - why exit at your target? there may be more profit for the taking. Tighten your stop when price nears your target and give it room to go higher (if price action is still strong as price approaches the target.)

Excessive moves- where price makes an excessive move in one period - tighten your stops (just in case)

Nizar - I hope I have not way laid your post (however it is titled "protecting profits"), your post was aimed at system traders. As I said before I am a discretional trader, and trade what the market is telling me - it works for me.

Good luck with your TradeSim code

Peter

 

[Shane Baker]

Hi Nizar

I personally subscribe to Perry Kaufman's idea that a system has to have a premise. it has to have a structured reasoning behind each component that you add in and and each part of the system ie entry/exit/money and risk management.

If you are considering a longer term, pyramiding trend following system then you might want to consider a few simple concepts.

1. Do I need to enter at the very start of a trend? How do I define a trend?
2. How do I define when a trend is over? Is that my final exit signal?
3. Do I hang onto all positions or am I willing to sacrifice one or two positions and allow the rest to remain until the trend ends. This may seem odd at first but if you use say an initial stop of 10% for each trade and you pyramid every time the share has moved 10% then on a significant retracement of say 22% you would lose two open positions. However if those two are just part of a large set of twelve positions ala a trend like CTX then you have not lost a lot compared to the profit of remaining in the trend until it ends.
4. How am I going to manage the risk with all these open positions. Obviously small risk amounts are a good idea but if i have twelve positions open of 0.5% risk each...I still have 6% risk in that share.
5. Money management. If I am considering riding a trend...but I wish to implement a scaling out strategy what is the likely outcome on profits? On drawdown? Looking at profits as an example, if I buy three parcels and sell at 2R, 4R and 8R, the last being the end of trend by my predetermined definition, I would make 14R total profit. Good stuff till I realise that if I held all three positions to 8R profit, I would have made 24R profit...ouch!

Just some starting thoughts for you to consider on your long and arduous journey

Cheers

Shane

 

[theasxgorilla]

If you are fortunate enough to be in a trade that goes parabolic - if you do not use some means of protecting profit - then you may give ALL of it back to the market when the parabolic trend ends!!
Parabolic trends tend to collapse explosively when they go vertical.

Nick Radge posted a comment on my blog about this topic. He describes that his system often manages to exit parabolic moves on the left-hand side ie. on the way up. Simple logic says that this may not be letting the trend play out to it's fullest extent, but parabolic is parabolic ie. a curve accelerating toward vertical and it seems that the very smallest minority of trends reach vertical and continue. By exiting on the way up you can side-step open equity drawdowns. Food for thought.

ASX.G

 

[bingk6]

Nick Radge posted a comment on my blog about this topic.
ASX.G

Gorilla,

Would really appreciate a link.

 

[weird]

Nobody knows how long a trend will last ... longer term trading systems for stocks means participating in a trend as long as possible, and when trading 'long' the upside is infinite .

I would be hestitate about using a profit stop in longer term trading .... a few reasons ...

1. Trade frequency can be low.

2. Once in a position, longer term trading tries to filter noise out as much as possible, which includes the shorter-term wave patterns (does not matter which theory you are subscribing to but there are shorter swings within the longer swings).

3. Longer term trend trading is not swing trading

Shorter term swing trading, which may use exits encompassing profit targets and also reversal patterns is perhaps better left to those more concentrating on those time frames where there is a higher trade frequency to make up for exiting earlier, and then re-enter the trend.

Reversal patterns could be further looked into, however these may, but not always, indicate a relatively smaller downward movement to an overall stronger longer term trend.

 

[tech/a]

Nobody knows how long a trend will last ... longer term trading systems for stocks means participating in a trend as long as possible, and when trading 'long' the upside is infinite .

Along my lines of thinking. What people seem to have not considered is that if your trading a system extensive Montecarlo testing will have given you a maximum high and low parameter for your system.Forward testing In sample etc etc.hat its saying is that if you dont play with it then you can expect x% return plus any other objectives you place in your "Wish list"
Twiddle and tweek it and your likely to fall outside of those parameters.Even worse kill it all together.

But thenI use 2 eye ball filters as my testing tells me I can take ANY trade triggered in my system and expect results between X and Y and forward tested 5 yrs so it does. But the eyeballing filters dont mess with the entry stop or exit criteria.So far blind luck or good eyeballing the results are in the upper end of test results.

a curve accelerating toward vertical and it seems that the very smallest minority of trends reach vertical and continue. By exiting on the way up you can side-step open equity drawdowns.

Has happened just reciently where my view and that of many others whom I respect was that this is an important topping.I took the decision to sell the whole portfolio---not in the system.
Human failing?
Well I would rather lose a small potential profit (As I will re enter if enough evidence comes to light) than a sizable open profit.
True had the open profit been less than it was I would probably not exited.
Whats right?---Im sure the opinion will be varied.

 

[howardbandy]

Greetings --

May I make two comments?

First, about stops. Parabolic trailing stops generally work very well. Chandelier trailing stops generally work well. Fixed percentage trailing stops do not work as well. Maximum loss stops work very poorly.

Second, about taking partial profits. Taking partial profits, say at break-even, then letting the remainder of the position run until the normal trade exit will dramatically reduce the overall profitability of a trading system and increase its drawdowns.

Thanks,
Howard

 

[BBand]

Hi Howard,
What do you use as a Parabolic trailing stop? I tend to use the same %age that I used as my initial stop ($value that I am prepared to give back to the market), this is the only application where I use this.

If the last bar does not have a large spread - then I would use a close below the low of that bar - provided it was less than the above method

My experience has been that ATR type trailing stops do not work too well for me.with regard to parabolic trends

Cheers
Peter

In general, how much we give back to the market on exit to a large extent depends on what timeframe we trade.

 

[nizar]

In response to Shane's post.

1. Do I need to enter at the very start of a trend? How do I define a trend?

No, entering at the very start is not a neccessity. Neither is riding it to the exact top. The middle 70-80% of the trend is enough (probably more than enough).

Definition of a trend, higher highs, lower lows. You cant really see the start of a trend until it develops. Half of breakouts (at best) result in trends.

2. How do I define when a trend is over? Is that my final exit signal?

Well you really cant. At best you can have a trailing ATR exit or a moving average exit, but even if it hits these i guess it could still rebound to the moon.

And thats why we test. To see which stop works well MOST of the time, over a large, statistical significant number of trades. Then we can focus on those and forget about the exceptions.

3. Do I hang onto all positions or am I willing to sacrifice one or two positions and allow the rest to remain until the trend ends. This may seem odd at first but if you use say an initial stop of 10% for each trade and you pyramid every time the share has moved 10% then on a significant retracement of say 22% you would lose two open positions. However if those two are just part of a large set of twelve positions ala a trend like CTX then you have not lost a lot compared to the profit of remaining in the trend until it ends.

I think its okay to have a common exit when pyramiding.

4. How am I going to manage the risk with all these open positions. Obviously small risk amounts are a good idea but if i have twelve positions open of 0.5% risk each...I still have 6% risk in that share.

Well thats true but if it triggered 12 entries then the stock is probably going in the right direction.

5. Money management. If I am considering riding a trend...but I wish to implement a scaling out strategy what is the likely outcome on profits? On drawdown? Looking at profits as an example, if I buy three parcels and sell at 2R, 4R and 8R, the last being the end of trend by my predetermined definition, I would make 14R total profit. Good stuff till I realise that if I held all three positions to 8R profit, I would have made 24R profit...ouch!

You're totally right with all your points.

This last point especially is what i was thinking about when i thought it wasnt such a good idea.
Profit stops probably wont improve the system to any great extent. Your basically chopping your winners too early.

Shane,
Iv been reading some of your posts on the reefcap forum from back in 2002-2003 and you knew your stuff even from then!
Just a question, on your weekly systems, what sort of open equity drawdowns were getting during testing?
And how about during real-life testing?
Is there any way we can minimise this while keeping the system intact?

I suspect not.
But ill be keen to hear your thoughts.
Thanks for your continous contributions.

 

[nizar]

Greetings --

May I make two comments?

First, about stops. Parabolic trailing stops generally work very well. Chandelier trailing stops generally work well. Fixed percentage trailing stops do not work as well. Maximum loss stops work very poorly.

Second, about taking partial profits. Taking partial profits, say at break-even, then letting the remainder of the position run until the normal trade exit will dramatically reduce the overall profitability of a trading system and increase its drawdowns.

Thanks,
Howard

Similar to my findings on the initial stop.

With the weekly system im developing, an initial stop does not improve the results. Iv tried several types. None of them increase profitability or decrease drawdowns, which is my definition of improvement.

A chandelier trailing stop is the same thing as an ATR stop?
If so, then yes iv found that works nicely for me as well.