The exit of a trade: how? when? where?

[firewalker99]

Exercise

Ok, let's try something.

Here's a part of a chart from DAX (doesn't really matter which instrument) from some time ago.

The blue line indicates potential resistance. You can see price stalling between 10-11am before breaching the resistance and breaking upwards. Volume on the breakout confirms the move. Later on we have a pullback to what now is support and near 1400pm we move higher.

Suppose we are long from 6905 early in the day, in anticipation of a break higher. Where would we exit and why? Most people will probably exit on the breakout or at least scale out half of their position there. But how do you manage your trade now? A trailing stop will stop you out much too early and most likely you will want to buy (or add to your position) again on the pullback.

So for the sake of the exercise we are still holding longs and price is currently at 6925 on the right side of the chart. What would you do to manage your trade and do you have a specific target in mind? I'll post the second part of the chart later on.

 

[fibonelli]

I'm a newbie so I'll have a go!

Assumption: 3 lots.

1. Lack of volume on 6914 to 6927 upmove concerns me. Take profit on one lot @ 6925.
2. Next move looks like a retest of 6915 (alt: 6925-6930 congestion area)
3. Raise stop to 6910. Failure of support @ 6915. Sell remaining two lots.
4. Next swing target circa 6940. Being 6929-6903+6914. Sell one lot and raise stop to 6925.
5. Continue holding third lot unless stop loss is hit. The stop loss is set underneath the previous swing low.

The worse case is that you make 30 points. I would use the third lot to keep me in the successful position as long as possible.

 

[fxwinner22]

1. put the stop below the last valley until you get a new valley

2. when you get a new valley move the stop to below the new valley.

3. repeat 1 and 2 till the stop is hit

 

[laptop1]

The interesting thing about scaling out is that it doesn't necessarily mean your profits are bigger compared to NOT scaling out and taking all your profits at once. This depends on
- where do you scale out the last part of your position?
- what if after taking profit at the first target, price goes back to your entry point?
- what is the % of trades that reach your second target?

Situation 1
Consider for instance that 70% of your winning trades go back to breakeven after your first target was hit. The other 30% hit your second target. Suppose your first (fixed) target is 10 points away and (to make thing simple) your second is at a fixed 25 points.
Suppose we have a batch of 100 winning trades and we trade 2 lots:

- By scaling out:
70% hit target 1 then reverse back to your entry: 70 x 1 lot x 10 + 30 x 1 lot x 0 = 700 pt
30% hit target 2 after hitting target 1: 30 x 1 lot x 10 + 30 x 1 lot x 25 = 1050 pt
Total points = 1750pt

- Not scaling out:
We exit all of our position at target 2. We move our stop to breakeven after price has passed the "virtual" target 1 point:
70% go back to your entry: 70 x 1 lot x 0 = 0 pt
30% hit target 2 with both our lots: 30 x 2 lot x 25 = 1500 pt
Total points = 1500pt

Situation 2
Consider for instance that 50% of your winning trades go back to breakeven after your first target was hit. The other 50% hit your second target. Suppose your first (fixed) target is 10 points away and (to make thing simple) your second is at a fixed 20 points.
Suppose we have a batch of 100 winning trades and we trade 2 lots:

- By scaling out:
50% hit target 1 then reverse back to your entry: 50 x 1 lot x 10 + 50 x 1 lot x 0 = 500 pt
50% hit target 2 after hitting target 1: 50 x 1 lot x 10 + 50 x 1 lot x 20 = 1500 pt
Total points = 2000pt

- Not scaling out:
We exit all of our position at target 2. We move our stop to breakeven after price has passed the "virtual" target 1 point:
50% go back to your entry: 50 x 1 lot x 0 = 0 pt
50% hit target 2 with both our lots: 50 x 2 lot x 20 = 2000 pt
Total points = 2000pt

Situation 3
Consider for instance that 40% of your winning trades go back to breakeven after your first target was hit. The other 60% hit your second target. Suppose your first (fixed) target is 10 points away and (to make thing simple) your second is at a fixed 20 points.
Suppose we have a batch of 100 winning trades and we trade 2 lots:

- By scaling out:
40% hit target 1 then reverse back to your entry: 40 x 1 lot x 10 + 40 x 1 lot x 0 = 400 pt
60% hit target 2 after hitting target 1: 60 x 1 lot x 10 + 60 x 1 lot x 20= 1800 pt
Total points = 2200pt

- Not scaling out:
We exit all of our position at target 2. We move our stop to breakeven after price has passed the "virtual" target 1 point:
40% go back to your entry: 40 x 1 lot x 0 = 0 pt
60% hit target 2 with both our lots: 60 x 2 lot x 20= 2400 pt
Total points = 2400pt

Conclusion
I realize these are a lot of numbers... But depending on how much trades you have that travel significantly far enough scaling out is not necessarily to best way to maximize on profits as is shown by situation 3.

These examples are hence very dependent on two parameters
- the win% ratio of your system
- the offset between T1 (= target 1) and T2 (=target 2)

This thread isn't meant to go into the numbers in depth, I'm sure there are statistical methods to analyze which method yields the best results under which circumstances... Nevertheless I hope these examples will be helpful for those who are trying to improve on their strategy.

I try to make this short and sweet ... In some cases based on fix targets you could be correct over long term trading, I looked in to this myself, However, if you don't depend 100% on fixed targets, then the results are going to be different from what you have said in the above post.

Personally, if I trade out of a position, I will scale out 3 times, if the trade goes to plan. However, if the market is talking to me, indicting strength or weakness based on pure price action time & sales. I may exit all my position before my targets or stops are hit. So based on that, the results are always going to be diffrent and hard to back test.

 

[zigglewigler]

I day trade the emini Russell 2k, I'm a range trader. I even consider trend days as ranging, except narrow as hell, I'll accept a small loss or scratch then take the continuation move, this way I can catch the trend reversals.

My aim is to be 'always in' the market, hence the exit points are simply swing highs and swing lows.

I look for the turning points in ranges within overbought/oversold areas, look for strong candle formations and click.

I can't trade without Bol Bands, the whole s/r line stuff I ignore, though Cama pivots are very helpfull.

So that's the idea, I let the market show me when it's exhausted either way, take the turn for a decent run or take hopefully a scratch and get back on trend.

 

[firewalker99]

I day trade the emini Russell 2k, I'm a range trader. I even consider trend days as ranging, except narrow as hell, I'll accept a small loss or scratch then take the continuation move, this way I can catch the trend reversals.

My aim is to be 'always in' the market, hence the exit points are simply swing highs and swing lows.

I look for the turning points in ranges within overbought/oversold areas, look for strong candle formations and click.

I can't trade without Bol Bands, the whole s/r line stuff I ignore, though Cama pivots are very helpfull.

So that's the idea, I let the market show me when it's exhausted either way, take the turn for a decent run or take hopefully a scratch and get back on trend.

If I may ask, how many trades do you on average take per day and on what timeframe? You must be a very confident trader, wanting to be "always in" the market. I think most people would loosen up on their rules and begin to act on feelings rather than strategy. But whatever suits your personality. Thanks for the post!

 

[firewalker99]

I try to make this short and sweet ... In some cases based on fix targets you could be correct over long term trading, I looked in to this myself, However, if you don't depend 100% on fixed targets, then the results are going to be different from what you have said in the above post.

Personally, if I trade out of a position, I will scale out 3 times, if the trade goes to plan. However, if the market is talking to me, indicting strength or weakness based on pure price action time & sales. I may exit all my position before my targets or stops are hit. So based on that, the results are always going to be diffrent and hard to back test.

Thanks for the reaction. Interesting to hear you have a quite discretionary style of exiting a trade then. Indeed, difficult to backtest probably because most likely you won't get the same feeling that the market is talking to you while reading an old static chart compared to looking at the bars or candles form in real time.

 

[barjon]

Pyramids

mornin', fire

You may be interested in Kaufman's (Trading Systems & Methods) conclusions relating to scaling in and out.

He found that all forms of pyramiding increase both profits and risk but show consistently higher returns than holding an unchanged position. Of all the various methods the reflecting pyramid did best. Here it is:

Good trading

jon

 

[firewalker99]

Scaling out in 3 phases

Yes, I will agree with Windowsill if you got a good entry point, best to place max on and scale out using something like,

Profit Target 1, 25% of your postion

Profit Target 2, 50% of your postion

Profit Target 3, 25% of your position, and let it run as far as you can go.

Laptop I hope you don't mind me copying this post from your journal, I think it's an good departure point for some further discussion on scaling out.

So to sum things up, we scale out 25% at Target 1 (T1), 50% at T2 and the remaining 25% at T3.

For me, I'm not clear if this is the optimized strategy. There are several scenarios that I can think of in which you will be leaving profits on the table. Right, any trader will probably leave profits on the table because squeezing out every last tick isn't our job, but let's focus for a second here on the scaling out in three phases:

Assume T1 = 10 points away, T2 = 15 points and T3 = 30 points (we want to hang on to this 3rd and final lot as long as possible)

1) We have a situation where a trade that goes to T1, but immediately reverses back to your entry. Personally, I move my stops to breakeven after my first target is hit so I basically enjoy "a free trade". I am curious which tactics other traders employ after they have scaled out a certain part of their position.

2) We can have a trade that goes to T1, retraces a part of that move (but not whole the way back to your entry point). It then continues on it's way to T2. Immediately after the market turns and reverses back to your entry. The important thing will be to move your stops accordingly. Would your stop be at the original entry point, at target 1, or trailing price closely, or something else?

3) In another situation, the best we can think of but also the one that will occur the least, the trade flies to T1 and T2 and never looks back before hitting T3.

Now consider what would the best strategy for each of the scenario's:

Scenario 1 We have exited 25% at T1, but the best thing would've been to exit your position completely at that point. Obviously we couldn't tell in advance that the market was going to reverse straight away. And if we exited our position completely but the market went to T2 we would be feeling pretty sorry we took are profits so soon.

Scenario 2 We have exited 25% at T1 and 50% at T2. Now the market reverses back to your entry point. The best thing would've been to exit your position completely at T2. But again, if the market traveled to T3, we would be feeling sorry we didn't leave at least a small portion of our position in the trade.

Scenario 3 All of our targets are hit, but we only exited 25% at T3 (the best price level). If we exited everything at that point, this trade would've been the big winner we've been looking for and probably make up for the losses of other trades. For everybody who is in the process of testing a system, it will be important to determine what % of trades travel as far as T3.

All of this should spark some thoughts, although I realize we are getting into numbers a lot it should be clear that it's important to have a good idea about the probabilities of your trades reaching a certain target. Based on the likeliness of reaching a target level, the approriate scaling out method should be applied.

 

[firewalker99]

Scaling out: a case-study

3 scenarios, 5 different styles... 1 conclusion

If anybody is interested in the spreadsheet file instead of the PDF, send me a PM.

 

[firewalker99]

mornin', fire

You may be interested in Kaufman's (Trading Systems & Methods) conclusions relating to scaling in and out.

He found that all forms of pyramiding increase both profits and risk but show consistently higher returns than holding an unchanged position. Of all the various methods the reflecting pyramid did best. Here it is:

Good trading

jon

Pyramiding, that is definitely worth looking into for longer term trading imo. I always found that pyramiding doesn't improve my results (rather the contrary), because of my (rather agressive) entry style which means I often have a nice and early entry. Adding to the position later on means more risk while the trade has already moved. Nevertheless, it's an interesting path and thanks for the suggestion!

 

[lurkerlurker]

Firewalker, thanks for the case study. I found it to be very interesting. When I trade full lots of the YM I'll only be trading one contract at a time, but if I am spreadbetting for awhile (at £2pp) or I eventually move up to 2 YM lots this will be very helpful. Thanks.

 

[laptop1]

Laptop I hope you don't mind me copying this post from your journal, I think it's an good departure point for some further discussion on scaling out.

So to sum things up, we scale out 25% at Target 1 (T1), 50% at T2 and the remaining 25% at T3.

For me, I'm not clear if this is the optimized strategy. There are several scenarios that I can think of in which you will be leaving profits on the table. Right, any trader will probably leave profits on the table because squeezing out every last tick isn't our job, but let's focus for a second here on the scaling out in three phases:

Assume T1 = 10 points away, T2 = 15 points and T3 = 30 points (we want to hang on to this 3rd and final lot as long as possible)

1) We have a situation where a trade that goes to T1, but immediately reverses back to your entry. Personally, I move my stops to breakeven after my first target is hit so I basically enjoy "a free trade". I am curious which tactics other traders employ after they have scaled out a certain part of their position.

2) We can have a trade that goes to T1, retraces a part of that move (but not whole the way back to your entry point). It then continues on it's way to T2. Immediately after the market turns and reverses back to your entry. The important thing will be to move your stops accordingly. Would your stop be at the original entry point, at target 1, or trailing price closely, or something else?

3) In another situation, the best we can think of but also the one that will occur the least, the trade flies to T1 and T2 and never looks back before hitting T3.

Now consider what would the best strategy for each of the scenario's:

Scenario 1 We have exited 25% at T1, but the best thing would've been to exit your position completely at that point. Obviously we couldn't tell in advance that the market was going to reverse straight away. And if we exited our position completely but the market went to T2 we would be feeling pretty sorry we took are profits so soon.

Scenario 2 We have exited 25% at T1 and 50% at T2. Now the market reverses back to your entry point. The best thing would've been to exit your position completely at T2. But again, if the market traveled to T3, we would be feeling sorry we didn't leave at least a small portion of our position in the trade.

Scenario 3 All of our targets are hit, but we only exited 25% at T3 (the best price level). If we exited everything at that point, this trade would've been the big winner we've been looking for and probably make up for the losses of other trades. For everybody who is in the process of testing a system, it will be important to determine what % of trades travel as far as T3.

All of this should spark some thoughts, although I realize we are getting into numbers a lot it should be clear that it's important to have a good idea about the probabilities of your trades reaching a certain target. Based on the likeliness of reaching a target level, the approriate scaling out method should be applied.

Few!! firewalker99, Well, like I said in my last post I don't use fix targets as such. I set T1,T2 ,T3 and a Stop
at the start of every trade, but don't trade it mechanically.I am more a discretionary trader. However, if it goes to plan It would work something like this

Enter trade! Stop 12 ticks!...First target hit at 5 pips move stop to 6 ticks, second target hit move stop to break-even, the final part of my position I trail my stop till stopped out.

As for moving stop to break-even after first target is hit, yes good to trade like this, but you will get lots of break-even trades. This is why I use 25%, 50% 25%.… I like hearing the cash register ka-ching. I don't think there is no right way or wrong way as long as you making money. If you have a sound method, whether it be moving your stop to break-even at first target or scaling out. The money is made my having a good money management plan in place to move up you lot size in line with your account.

 

[zigglewigler]

If I may ask, how many trades do you on average take per day and on what timeframe? You must be a very confident trader, wanting to be "always in" the market. I think most people would loosen up on their rules and begin to act on feelings rather than strategy. But whatever suits your personality. Thanks for the post!

As I am a lazy b'stard, and don't want to get up at 6am LA time, I only trade the afternoon session, so around 9am to 1pm and I'm done. The advantage apart from sleep being that the market has established some form of character by then, the afternoon session being a reaction or response to the morning half.

But still can do 6 trades in that time, using 1 & 3 minute candlestick charts.

As for needing to be confident, my risk is clear, based on the strategy I use, I know exactly where my exit is and take it if I made the wrong call. Then jump back on the continuation. That's the benefit of range trading, though risky trying to catch tops and bottoms, easy to see where the exit is, from a few ticks to a point and a bit generally.

I'd much rather be in, with a known small risk, that wishing I was in and watching the sod peel away into the distance not sure whether to jump on the momentum train because it already looks over. The next reversal is then my current profit exit and next trade entry.

But yes, you have to be on the ball, but I can't bear looking at the screens and asking whether the arbitrary (my opinion) lines of support and resistance, will in fact hold or break. If they break, by how much to be certain it will continue? If they've held by a degree just after you jumped in, where will you get out? You exit, only to find that after a second or third push, the market breaks through and leaves you sitting there like an ar*e.

I've put alot of work over the years into making my form of range trading work, simply because it suits me. And yes, the psychology aspect of a trader has to be satisfied if the strategy is to work

 

[firewalker99]

Few!! firewalker99, Well, like I said in my last post I don't use fix targets as such. I set T1,T2 ,T3 and a Stop
at the start of every trade, but don't trade it mechanically.I am more a discretionary trader. However, if it goes to plan It would work something like this

As for moving stop to break-even after first target is hit, yes good to trade like this, but you will get lots of break-even trades. This is why I use 25%, 50% 25%.… I like hearing the cash register ka-ching. I don't think there is no right way or wrong way as long as you making money. If you have a sound method, whether it be moving your stop to break-even at first target or scaling out. The money is made my having a good money management plan in place to move up you lot size in line with your account.

I know you don't actually use those fixed targets, but your post in your journal was a good departure point for some numerical analysis

It's true however that trades can return back to your entry-point after reaching your first target. But at least you won't let a winning position turn into a loser then. You can always choose to re-enter your trade if the price action favours that.

Thanks you for your input laptop1.

 

[firewalker99]

I'm a newbie so I'll have a go!

Assumption: 3 lots.

1. Lack of volume on 6914 to 6927 upmove concerns me. Take profit on one lot @ 6925.
2. Next move looks like a retest of 6915 (alt: 6925-6930 congestion area)
3. Raise stop to 6910. Failure of support @ 6915. Sell remaining two lots.
4. Next swing target circa 6940. Being 6929-6903+6914. Sell one lot and raise stop to 6925.
5. Continue holding third lot unless stop loss is hit. The stop loss is set underneath the previous swing low.

The worse case is that you make 30 points. I would use the third lot to keep me in the successful position as long as possible.

Thanks for taking part in the exercise. Here are my comments on your excellent observations:

1. Notice how volume drops off and almost comes to a standstill before the actual breakout. You say the lack of volume concerns you, but it can be a signal of anticipation. At 1055 (the wide spread up bar that goes up to 6925) the volume is entering the bars. However price already broke resistance by then. This might be a slight difference, but keep notice how volume trails price.

2. Indeed a retest of 6915 was likely.

3. A sensible stop at 6910. But why do you say failure of support? I don't see 6915 failing?

4. Interesting method to determine the next target. So you add the length of the first swing to the pullback and estimate a second target. I don't know how reliable this is, but it's definitely better than just gambling on the exit. Price actually traveled to 6950 before retracing the move to 6933 (see attached chart). Notice the exuberant volume on the upbar on 1430. That's due to the release of economic figures at that time.

5. Setting your stop underneath the previous swing low is a safe bet. However in most cases it will stop you out when price has already reversed and leaving some points on the table.

All in all, a very decent analysis. Are you sure you are a newbie?

 

[firewalker99]

1. put the stop below the last valley until you get a new valley

2. when you get a new valley move the stop to below the new valley.

3. repeat 1 and 2 till the stop is hit

This works fine when there are clear moves and swings. That's however not always the case. In fact, you are quite lucky if you have a stairstep kind of trend moving day with swings higher and higher highs (or lower and lower lows). Also, when do you move your stop under the "new valley"? According to that, I believe you would move your stop underneath the blue circled area (say 6915) and be stopped out later on. I am right or misinterpreting your comment?

 

[barjon]

Pyramiding, that is definitely worth looking into for longer term trading imo. I always found that pyramiding doesn't improve my results (rather the contrary), because of my (rather agressive) entry style which means I often have a nice and early entry. Adding to the position later on means more risk while the trade has already moved. Nevertheless, it's an interesting path and thanks for the suggestion!

fire

In drawing your attention to the reflecting pyramid I was thinking as much about the exit strategy of scaling out as the scaling in. However, I'm not sure that adding to a position later on "means more risk" if you follow the rule that no unsuccessful secondary position should offset the entire profits of the prior position.

Anyway, a way to reduce the risk is to exit positions once specific targets are reached rather than relying wholly on judgement of the market action. For example, analysis of your successful trades will establish a reasonable average of the profit expectation of the move (not the trade) - for argument's sake let's say that comes out at 30 points.

You'd be looking to scale out at +20, +25 and +30. That can either be done by a straight exit once those targets are reached or, in case the thing keeps going like a train, by moving up a tight stop on a proportion of the overall position as those targets are reached.

good trading

jon

 

[fibonelli]

Thanks for taking part in the exercise. Here are my comments on your excellent observations:

1. Notice how volume drops off and almost comes to a standstill before the actual breakout. You say the lack of volume concerns you, but it can be a signal of anticipation. At 1055 (the wide spread up bar that goes up to 6925) the volume is entering the bars. However price already broke resistance by then. This might be a slight difference, but keep notice how volume trails price.

2. Indeed a retest of 6915 was likely.

3. A sensible stop at 6910. But why do you say failure of support? I don't see 6915 failing?

4. Interesting method to determine the next target. So you add the length of the first swing to the pullback and estimate a second target. I don't know how reliable this is, but it's definitely better than just gambling on the exit. Price actually traveled to 6950 before retracing the move to 6933 (see attached chart). Notice the exuberant volume on the upbar on 1430. That's due to the release of economic figures at that time.

5. Setting your stop underneath the previous swing low is a safe bet. However in most cases it will stop you out when price has already reversed and leaving some points on the table.

All in all, a very decent analysis. Are you sure you are a newbie?

Hi Firewalker,

1. I would expect volume to be higher in the primary direction. My interpretation of the chart was that there was a discrepancy between price and volume at the re-test of 6925.

2. It looked logical.

3. Ooops, I didn't explain myself clearly. I actually meant in the event of 6915 failing (not that I was forecasting it would), I would liquidate all remaining lots at 6910.

4. It's a very simple method of assuming a swing pattern.

5. I agree. The stop is there to take you out on the failure of the swing pattern (my reason for the trade). Probably needs to be set a bit more below the previous swing low to avoid stop hunting antics.

I am a newbie and learning each day.

Thank you for putting up the example.

Fibonelli

 

[firewalker99]

Hi Firewalker,
3. Ooops, I didn't explain myself clearly. I actually meant in the event of 6915 failing (not that I was forecasting it would), I would liquidate all remaining lots at 6910.

5. I agree. The stop is there to take you out on the failure of the swing pattern (my reason for the trade). Probably needs to be set a bit more below the previous swing low to avoid stop hunting antics.

Thank you for putting up the example.

You're welcome. You've already explained your reasoning, which sounds very logical so I have little more to add. About the stop placement and stop hunting actics I'd say the ideal stop size should show up from testing. Probably different for each instrument and depending on the volatility of the market. But 6910 would've been a fair bet in this situation.

I've attached the full chart of the day to see what happened next. Most traders would probably be out of their longs:
1) after price swings back to 6935 (at 1700) and doesn't make a higher low
2) or when price fails to make a higher high (around 1730pm) at 6950 and drops lower

Interesting action near the close however. A run-up of about 40 points in just a bit more than 1 hour. The possibility of a re-entry later on existed although some would probably argue this is a H&S-pattern and the probabilities of more upside were limited.