Triple threat exit strategy

How would you like to have made +31,000 pips in the last eighteen months of trading? Well, then, we’d better get to work!

EQUITY CURVE, JANUARY 1/JUNE 30, 2012, GBP/JPY, AUD/USD, EUR/USD, GBP/USD


In sports, the triple threat player is one who excels at three major skills associated with a sport. In baseball that would be batting, throwing, and running. In football it is passing, kicking, and running. In trading we are going to suppose a triple threat to be the ability to capitalize on a quick move, an intermediate move, or a long move.

This simple system is for those who are able to watch the market or, at least, look in on it frequently. TRIPLE THREAT is more about the way you get out of a trade than the way you get in. In fact, you might choose to use these exits with virtually any entry system you favor.

For our purposes, I will be using entries from the DAYBREAK system. That is an entry method I am intimately familiar with (Duh!) and it is easy to understand and implement. We will make a change in the pairs we study. DAYBREAK uses gbp/jpy, aud/usd, eur/usd, and usd/jpy. We will continue with the first three here, but drop usd/jpy in favor of gbp/usd. You can visit the DAYBREAK thread at this site to learn more about this “set it and forget it” type method, but for now, all we need from DAYBREAK are the entry signals.

No indicators, no formulas. Watching four pairs, starting with three positions in each pair that gets triggered:

  1. Go long at yesterday’s high plus one pip.
  2. Go short at yesterday’s low minus one pip.
  3. Exception: For Monday trades, go long at yesterday’s high plus 11 pips or go short at yesterday’s low minus 11 pips.

We’re in. On to exits. With TRIPLE THREAT, any time we enter a trade, we are going to enter three positions and we will use the same initial stop-loss for all three, but a different profit exit strategy for each of the three. Size your positions consistent with your own money management approach, tolerance for risk, etc. What you will find is the three exits essentially make this into three trading systems and, in any given month, one may do better than the others.

We are currently evaluating initial stop-loss ideas. At this time, I suggest you experiment with a fixed sl around -25 to -35. Important: Only one of our positions will have a take-profit. Establishing initial stop loss is only applicable before we collect our first profit. When Position 1 hits its TP, we will start operating with Profitable Exits rules.

PROFITABLE EXITS

Aside from initial stop-loss, our profitable exits for TRIPLE THREAT will be looking at round numbers or, more precisely, quarters, halves, and wholes. A round number is simply where the price of our pair is at a number ending in 00. We will use the actual round numbers, but we will also use quarters and halves, i.e., numbers ending in 25, 50, and 75. There is information on this site and others about the presumed significance of round numbers and quarters theories and how to use them in your trading.

Our TRIPLE THREAT exits are, “Take Profit,” “Air,” and “Retrace.” They are so named because “Take Profit” will close a position at a predetermined quarter level that results in +25 or more; “Air” will create a new following stop each time an additional quarter is achieved (when there are at least +25 pips “air” between the current price and the last profitable quarter); “Retrace” works on the halves and, after achieving a new half, will get its stop moved to the former half. Conceptually, Take Profit will attempt to capture a quick profit and protect the other two positions, while Air and Retrace provide opportunities to capitalize on intermediate and large moves.

Keep in mind that we are entering three positions for each pair, same initial stop-loss for all, and only Position 1 gets a TP. Therefore, if an initial stop-loss is hit on a pair, our risk is on three positions, not just one. That’s why Positions 2 & 3 are open to achieve theoretically unlimited profit.

When our initial TP is hit, we will always move our other two positions to the nearest quarter that results in at least a small profit. It will happen frequently that we make a little profit on the first position, but the market reverses to take us out of the other two. No need to let these gains turn into losses.

Think about it as if Position 2 is governed by the quarters and Position 3 is governed by the halves, or Position 2 is being trailed by the nearest quarter and Position 3 is being trailed by the nearest half. Position 2 stops will always be resting at a 00, 25, 50, or 75 level (quarters). Position 3 stops will always be resting at a 00 or 50 level (halves). To help myself stay on top of my trades, I draw yellow horizontal lines at all 00 and 50 levels and red lines at all 25 and 75 levels. Since you may end up in multiple trades across four pairs, it is also a good idea to keep a notepad close by.

Observe that we may be waiting a while for these positions to fall back to close. They may be carried overnight, sometimes for days, and through the weekend. Though I prefer to be flat over the weekend, there are times when it seems appropriate to leave trades open. Often, while we are waiting for trades to close, a new day comes and we may have new entry signals. System testing assumes those signals were taken and more positions were added. But, that can be a judgment call on your part. If existing positions are safely in profit, i add the new positions. If not, I prefer to mitigate risk and just stay with what is already open.

We can walk through an example from May 17, 2012, GBP/JPY. You can follow along on this chart.


The high last May 16 was 128.49. The low was 127.65. Our resting sell stops for three positions will take us short at 127.64. Our initial stop-loss for all three positions is 127.99, risk -35 per position, -105 per trade, but it turns out we don’t even have to worry about that. Our TP for Position 1 is 127.25, potential +39. Let’s now track the activity bar by bar as the trade unfolds. Time references are to the hourly bar that is building.

  1. 7:00 Short orders are triggered for three positions at 127.64
  2. 8:00 Little movement, no action, no decisions to be made.
  3. 9:00 Price moves in our favor to 127.50, but still no action.
  4. 10:00 Price moves further in our favor through 127.25. Position 1 is cashed, +39. We move stops on Positions 2 & 3 to 127.50. We are now locking in 14 pips profit on each of two remaining positions.
  5. 11:00 Further positive as price breaks through 127.00. Position 2 is on the quarters, move Position 2 stop to 127.25. Position 3 is on the halves, leave the stop at 127.50.
  6. 12:00 Position 2 barely survives as prices reach up within two pips of our stop (127.23). But, it manages to hang on while prices fall through 126.75. Move Position 2 stop to 127.00. Leave Position 3 stop at 127.50.
  7. 13:00 Positions hold their own as the market prepares to give us one of those times when we remember why we like to trade.
  8. 14:00 Dramatic, and I mean serious, move in our favor. By the end of the hour, prices have penetrated 125.75 and we have replaced stops. Position 2 stop now rests at 126.00, Position 3 at 126.50.
  9. 15:00 No! No more! Stop tickling me, I can’t stand it! Prices fall through 125.50. Position 2 stop moved to 125.75 and Position 3 moved to 126.00.
  10. 16:00 Prices go south below 125.25. Position 2 stop to 125.50. Position 3 stop stays at 126.00.
  11. 16:00 Action on the 16 bar was such that Position 2 is taken out at 125.50, +214.
  12. 17:00 thru 3:00 of May 18 No news for our Position 3.
  13. 4:00 Price moves through 125.00. Position 3 stop is moved to 125.50.
  14. 5:00 to 8:00 Nada.
  15. 9:00 Price retraces to hit our Position 3 stop, taken out at 125.50, +214.
  16. Total pips +467. Boy, let’s do that again tomorrow!

Granted, May 17 was exceptional and chosen to illustrate the process of moving stops on the quarters and halves. We will certainly have our share of losers.

DISCRETION

We have worked with the ea to code the decision process, but most traders prefer to stay in control of their trades and may elect to make certain judgment calls against the rules. If you suppose yourself to be like Kipling’s admonition and can keep your head when all about you are losing theirs, then make a judgment call here and there. Or, if you’re like me and just admit human frailty, that is, right or wrong, sometimes I’m just gonna make a decision and pull the trigger, then discretion is certainly on your menu.

One discretionary area is whether to hold trades over the weekend. If you are going to trade the system, by all means hold overnight during the week. Your stops are set. You can rest as easily as any real trader can rest and you can always set alerts to come get you if progress takes you to a new level and requires resetting stops. Rules of the system will hold over the weekend also. Over the weekend can be a judgment call for the individual trader.

When i see a good runup in my favor that falls just short of that first tp level, followed by an immediate reversal, i may make a judgment call to close out and grab a few pips. Watch your price behavior and decide whether you’re willing to see that long from 26 trade touch 74, then fall all the way back to a loss.

Mondays and Fridays, all bets are off for me. I am like a cat on meth on Monday and Friday and I frequently override the rules. DAYBREAK and TRIPLE THREAT can make for some great trades on these days, but I am always on the alert to cut reverses quickly or grab a quick buck. For the results section, I adhered to the rules on other days, but on Mondays and Fridays I was outta control.

A major premise of the program is adding to winning positions when we have additional entry signals. There may be times when we just don’t want to increase our risk, so we pass. Having positions open, we may get a new signal breaking out from Thursday’s high or low, but it’s probably not a good idea to add more positions on Friday if we don’t like holding over the weekend.

Finally, when the end of the day is close to the high or low. I have observed that one day may make a significant move, but it isn’t destined to continue into the next day. If we are already in positions, we may want to just continue those into the next day. For new positions, we may want to require a few extra pips (10 or so) to trigger a new trade.

EDIT, AUGUST 16, 2012, RE: EA’S

At this time, we have at least two programmers working to help us build a working and accurate ea. Current progress is that we have a working ea that is not entirely accurate. These ea’s can help you backtest, but they will show trades that don’t fit the original rules or the improvements we are trying to make. DO NOT USE THESE TO TRADE A LIVE ACCOUNT! One of those working is rpotor who has made a webpage to house the ea’s. Next paragraph is a post from rpotor explaining the use of the page and providing a link.

“Below is a link to an absolutely plain, no-frills webpage where from now on you can find the current version of the EAs coded by me for this trading system. By putting this webpage live, I’ve deleted all previous versions of the EA from my website, so links posted previously in this topic to the various old versions of the EA will now come up with an error message and only the current version can be downloaded. Hopefully this will prevent newcomers from downloading an old and potentially buggy version. I put instructions on how to find that webpage in my forum signature, so hopefully anyone with eyes should be able to get to the website even if he/she first encounters a link to a previous EA version which is not there anymore. On this webpage I also included dineshydv’s EA for the sake of completeness, if this is not okay for any reason, please tell me and I’ll remove it. I also reproduced the latest directory of topics there as collected by pipwoof.”

See here: http://www.rpotor.com/forex/TTE/index.html

Read more: http://forums.babypips.com/free-forex-trading-systems/45633-triple-threat-exit-strategy-45.html#ixzz23i1b5YSL

So, let’s start talking about results. For eighteen months, starting in January 2011 through June 2012, I have laboriously looked at each hourly bar for the GBP/JPY, EUR/USD, AUD/USD, and GBP/USD to see what might have happened in real-time had we been using the TRIPLE THREAT exits with the DAYBREAK entries. I used IBFX data, F12, a pencil, F12, a tablet, and F12. Oh, yeah, and F12. Please let it sink in that this was all done by hand.

I made discretionary calls sometimes, mostly on Mondays and Fridays. Are there errors in here? Did I miss a few things? Did I add it up to +49 when it was actually +39? Absolutely! Furthermore, the hourly bars also have limitations. We are looking at a whole hour. We can’t see for sure if some jerky behavior within the hour may have made a different outcome, better or worse. But, I do believe we have some achievable results here. Again, we are looking for generalizations. Caveat emptor! Satisfy yourself that this might be for you. Check my work.

Just so we know what we’re looking at, left to right, these sheets tell us the month, the pair, and the losses and wins for each of the three exit strategies, with a total at the end. I haven’t a clue why the Aussie did as well as the Yen and don’t know if it will continue.



A few comments here. We can see that I am showing the losses for each of the triple-threat exits as always the same each month. I held that because they all share the same initial stop. It may change slightly among the exit strategies if we make judgment calls. Usually, if one gets stopped out, they all do, and at the same point. We also see that TP rarely does better than the other two. Well, could we just eliminate the TP altogether and trade larger positions with the other two exits? We could. Don’t lose sight of the fact, however, that trading is also a psychological game. For me, having the target get reached, putting some money in the bank, and still having open positions is psychologically rewarding and I will continue that way.

Interesting strategy Pipwoof. Curious on couple things. 1st, what dates are you backtesting these 18 months? 2nd, do you always follow the trade on the H1 as you have in the picture? 3rd, What do you consider to be the “end of day extreme?” Is it GMT or end of a certain session or something else?

thanks for the post,

-GarrisonArmstrong

Can you tell the closing time of your daily candle in GMT ?

garrison, per the post above, “For eighteen months, starting in January 2011 through June 2012.” I watch the one-hour bar.

the issue of “day” is going to depend on your broker, as each will define the day according to their platform time. here is some information from the DAYBREAK thread, post 145:

"… i use my ibfx charts for updates to the thread. ibfx usually shows me only one or two hourly bars on sunday. the method requires +/- 11 pips for a monday trade. on ibfx, gbp/jpy sunday high was 123.42 +11 = 123.53 long entry. sunday high for eur/usd was 1.2279 + 11 takes us long at 1.2290.

oanda is on a different time and usually shows me nine or ten hourly bars on sunday. oanda shows the gbp/jpy sunday high for their “day” as 123.53, which would put us long on their platform at 123.64. oanda high sunday for eur/usd was 1.2296 + 11 = 1.2307. i have found both ibfx and oanda to be profitable, but oanda less than ibfx.

each trading platform will create a discreet data set called “day.” i have assumed that breakouts from any day may be profitable, but i haven’t checked more than the two brokers. not familiar with fxcm, so would be interested in what you determine. june was a particularly difficult month with gbp/jpy taking some hard losses and ending the month, i think, about -150 for the portfolio. you might walk through that month by hand and see what you get." DAYBRAK thread, posts 21, 30, and 113 also discuss the issue of differing platforms. Hope this helps, thanks.

Read more: 301 Moved Permanently

Read more: 301 Moved Permanently

bijoymj, i am using ibfx data. they are on gmt. they close the bar for the day at 23:03 and start a new day at 00.

just in case anyone is having difficulty following, i really want you to understand this strategy and be able to use it if you determine it to be valuable to your trading. to help, i will post at least a few days results here. any questions, send 'em over.

today, july 12, 2012
the chart is for the gbp/usd today and an aid to help you follow the steps.


gbp/jpy long @123.98, stopped out @ 123.50, -48 loss X 3 = -144.

aud/usd short @ 1.0177. pos 1 (TP) takes profit @ 1.0150, +25. pos 2 (AIR) achieves one air level, then reverses to be taken out at 1.0125, +50. pos 3 (RETRACE) achieves one level, then reverses to be taken out at 1.0150, +25.

eur/usd short @ 1.2211. pos 1 takes profit @ 1.2175, +35. Pos 2 & 3 reverse to 1.2200, +10 and +10.

gbp/usd short @ 1.5483. pos 1 takes profit @ 1.5450, +30. pos 2 get air, then reverses to be taken out @ 1.5475, +10. pos 3 still open with a stop @ 1.5450, locked in profit, +30.

to summarize, today we had -144 in losses, +195 wins, net +50 and we are holding one position open with a locked in profit of +30.

edit, 7-13, gbp/usd moves back up to take us out of our last open position @ 1.5450, +30. so, entries yesterday have all finished and we are flat again. after this final close we are -144, +225, net +81. today, we are getting additional entry signals on some pairs.

nice!! I’ll be watching

Are these real trade results or back tested results please? Thanks.

already testing it :smiley:

backtested.

i appreciate your participation.

Pipswoof…l think this is an important question could you answer please.
Thanks in advance.

asked and answered in post #11 above. not sure that this should be causing any confusion. here is a statement from post 2, where i discuss results: “For eighteen months, starting in January 2011 through June 2012, I have laboriously looked at each hourly bar for the GBP/JPY, EUR/USD, AUD/USD, and GBP/USD to see what might have happened in real-time had we been using the TRIPLE THREAT exits with the DAYBREAK entries.”

Read more: 301 Moved Permanently

Sorry, l missed that post…thanks for clarifying!

np. most of the time anything i post will be strictly experimental and in the testing phase. i apologize if i didn’t make that clear on this one. if i ever post regarding a live account, i will include pictures of that account.

imo, before trading live with any method, there are at least two important things you should do: 1. look back at what would have happened. that’s what i mean when i say trading is an open book test. 2. subject the method to what can sometimes be harsh scrutiny from a trading community like babypips. i may think i have the mamas pajamas for a trading system, but some “joined yesterday” may spot a hole you could drive a truck through.

history (backtesting) can’t tell us what will happen, it can only tell us what did happen. but, we at least have an indication. the input of other traders can identify weak spots and suggest changes. i have posted this exit strategy and have another thread, DAYBREAK, where i have had lots of competent input. this process is leading me to a level of confidence, or lack thereof, which will result in discarding the method, revising it, or adding it to what i am already doing.

i do appreciate your interest and hope you will get something beneficial out of this.

it is friday and you’ll get a good idea what i mean by discretionary trading on mondays and fridays.
today, we had no signals from g/j or a/u.

e/u went short @ 1.2165. this trade moved a few pips in the right direction, then headed back up toward the most recent quarter. when it got there, i gave it no quarter and cut it down at 1.2175, -10 X 3 = -30. had you wanted to give it an opportunity to go your way, your stop would have been 1.2200, risking -35 X 3 = -105.

g/u went long @1.5516, hit pos 1 target @ 1.5550, +34.
it is now just past 2 pm central time and ibfx will close the door on the day in less than an hour. don’t want to hold these two open positions over the weekend. looking for price to touch 1.5575 (currently at 1.5571) and will take both out there, otherwise, will close sometime between now and 15" of closing.

ah, 1.5575 it is at 2:09 central time. final two positions exited, flat for the weekend. +59 on each position = +118. with our tp of +34 and our stop-loss of -30, we are up today +122.

had we chosen to allow that e/u trade more room, we would have -105, +34, +59, and +59 = +47. not bad.

Hello pipwoof

Are these tests from real trading/demo or from historical test data please? Thanks.

from post 11 above, “backtested.”

from post 14 above, not sure that this should be causing any confusion. here is a statement from post 2, where i discuss results: “For eighteen months, starting in January 2011 through June 2012, I have laboriously looked at each hourly bar for the GBP/JPY, EUR/USD, AUD/USD, and GBP/USD to see what might have happened in real-time had we been using the TRIPLE THREAT exits with the DAYBREAK entries.”

Read more: 301 Moved Permanently