Triple threat exit strategy

sheriff, you have the entries and initial sl right, but appear to still be in pursuit on the exits. let’s see if we can catch this guy. here’s where i think you start to go wrong, “Placing t/p for all 3 positions with 3 different t/p.” tp applies to only one of our three positions. we call that one “take profit.” the other two are managed by trailing stops based on points on the chart called quarters, i.e, prices ending in 00, 25, 50, or 75, and we call position 2 “air” and position 3 “retrace.”

it may help us to separate our thinking into two parts: part one is when we intially set up and part two is after our first profit. when setting up, we will want our buy/sell stops and our sl the same on all three positions. if trades are triggered, three things can happen: 1. they don’t move and we just patiently wait until one of the other things happens, 2. they move against us and sl is hit for all three positions, 3. they move in our favor enough so that position 1 hits tp. remember, tp applies only to position 1.

now that position 1 has taken a profit, we shift our thinking into part two. we just took a profit and still have two positions open, air and retrace. this is when we engage our quarters theory and move stops on both remaining positions to the nearest quarter that would give us either a small loss or a win. don’t risk a sizeable loss here. i always favor an sl that holds on to a few pips, but if you want to risk a few to see if you can get further, your call. for example, if you had long entries at 125.20 and price rose to 125.50, you would tp on position 1 +30. that would be the tp point you determined when you initially set up, i.e., nearest quarter that give you at least +25. now, you need to move stops on your other two positions. your choices are to set stops at 125.00 where you would now be risking -20 on each of your two remaining positions or set them at 125.25where you would have less wiggle room but would lock in +5 on each position.

from here, once again, three things can happen: 1. movement stall, be patient, 2. prices reverse and take out our positions at small loss or win, whatever decision we made, 3. additional movement in our favor. if there is additional movement in our favor such that it starts crossing quarter lines, we then start trailing position 2 at every quarter behind it and position 3 at every half behind it. example, our long entries from 125.20 see prices rise to 126.00. we now set sl for position 2 at 125.75, which is the nearest quarter behind it, and we set position 3 sl at 125.50, which is the nearest half behind it.

people tell me this is easy once the light bulb goes on, but you may have to read post 1 and this resonse several times. if you can download and attach rpotor’s ea, it will help. i am using v3.21. when a new day begins, attach the ea and it will establish all entry stops, long and short. i think it may also move the air and retrace stops. can’t say for sure because i don’t think we’ve had enough profit anywhere for me to see stops move. lol. hopes this helps. good luck.

Hey again! Darn the market not triggering much today. Anywho, I went and downloaded rpotor’s EA, since manually setting all those long/short entries is tedious. I went and put the file into my “experts” folder, and it appears in the expert advisers list when I open IBFX, but I am not sure if I am “doing it right” so to speak. What are the steps that need to be taken once I have it all installed and in my expert advisers list? (the icon is greyed out, if that matters)

Thanks all! Here’s to hoping tomorrow brings prosperity!

mine is not greyed out. i am using v3.21. click on the link. open. it should open in metaeditor. compile. then, file, save as, and take it where you want it. let me know if this works for you.

see post 281 on page 29 for a link to 3.21.

my bad. paid a little more attention and it seems to be working for me. thanks for the contribution. i will be testing and see if there is some reconciliation for the difference in outcomes. hope this doesn’t work out to depend solely on “discretion.”

once you have it installed, wait for the opening of a new day, then attach to a chart. it should place all entry stops and initial stop losses properly.

It is all installed I believe and is not greyed out. So I go to the chart, expert advisors, strategy tester, then start, yes?

edit: Of course, this would only be a test to make sure the system is working, and I will be waiting until 7 until I put in the real entries.

also, in post 346, dines has provided an experimental ea. i have just started to figure it out and try to use it on demo. it works differently from rpotor’s. this one does not place the entry stops where you can see them in pending orders. rather, when conditions are met, it opens the orders. it also moves stops as prices move and seems to be accurate there. more testing needed.

Thanks! I may check it out as well.

Cheers.

GREAT…i suggest you backtest it to compare results with that of your’s. I ran on EU & EJ and its hardly BE.
Let me know hpw your backtesting goes with this EA…

Hi rpotor, since a few days I run you EA vers. 1.01 and it opens an endless numbr of sell stop and buy stop orders, i.e. today it entered at 16:00 Manila Time every minute 6 positions (360 Pos. by 17:00) is my setting wrong or what happens. But the chart looks well, just only 2 order lines and each one TP and on SL.
Since I am newbie with the use of EA’s I am confused.
woris

rpotor’s EA cant be backtested. Test with one available at this post [link].

i am testing g/u july only data with both dinesh and rpotor ea’s. dinesh shows +448. using the breakeven feature, it seems to have a problem. i can’t get a screenshot, but starting with line 30 on july 4 the ea begins modifying order #7 sl, going back and forth between 1.5653 and 1.5700 and continuing with this through line 413. starts again with line 440, sl for order #13 going first 1.5495 then 1.5425 and continues this through line 3,583. in all, the one-month test goes over 29,000 lines with this repetition for various orders. have not had the opportunity to verify trade by trade, but shows +542 for the month.

rpotor’s also seems to be working well, using 355 lines and showing +494.

I havent tested breakeven feature thoroughly so let me test it out and will do the necessary changes. I have tested it without BE feature.


after making some corrections to both sides, here is a comparison of rpotor’s ea v3.21 backtest with what i got by hand. data is g/u for july only.

though both are profitable, there are about +300 pips difference in the two. to account for the difference:

  1. judgment calls were sometimes right, sometimes wrong, but gave me an overall advantage of about +100. some of my judgment calls are when we see a run in our favor, almost to tp, followed by a reversal. i believe the ea could replicate at least part of that with a be stop after a select profit accumulation, maybe somewhere +15 to +25. need to make it variable so we can test it. dinesh has this feature in his ea and is working to correct a bug. if that gets solved, i will do a test using his be function and see if the ea can incorporate some of the judgment calls. other judgment calls were made on mondays and fridays where the market just didn’t move in a direction and i pulled the plug. i don’t see how an ea can accommodate for that so will hope that mechanized versions remain profitable without that benefit.
  2. another difference between the ea and pipwoof was that the ea would take both trades on a day when both entry points were reached. these double entries cost the ea -180, but it also traded on july 4, which i did not, and it made +190, so these differences are a wash.
  3. the ea would carry trades into the next day and into weekends. while i normally also carry into next weekday, i do try to be flat by eod friday. the ea showed an advantage of about +80 for keeping the trades open.
  4. a major difference occurs in sl’s. with the ea showing a short trade with sl 1.5575, it exits on a high of 1.5474. again, short trade showing sl 1.5700, it exits on a high of 1.5697. visually, i would not have exited either of these trades and both worked out to be more profitable than the ea. no doubt, this has to do with spread and the only ways to adjust would be either for me to count these as stopped out when the spread is considered or for the ea to include a spread factor before hitting sl. this would be harder to do visually, so would prefer that the ea be adjusted to account for spread, i.e., for g/u my ibfx spread probably averages 3.


dinesh, including a few days for comparison. your ea also takes both sides of the trade on days when both are hit.

seems to be some problem in adjusting stops. stops for pos 1 do not follow the same rules as for pos 2 and 3. initial stops for position 1, tp, should not move. trade on 7-2 should show a -40 for pos 1. air trade (pos 2) on 7-6 should not be modified. pos 2 and 3 are modified only after pos 1 takes a profit. in this case, having entered at 1.5497, pos 1 takes profit at 1.5525 +30. at this point, positions 2 and 3 sl’s would move to 1.5500.

Trading live on Oanda Fxtrade based on gmt, trading eur/usd, gbp/jpy, aud/usd, results to date are as follows:

TT: -1310 pips

TT w/ base x2: -1499

TT w/ RT x2: -1984

ITT: +1073

ITT w/ base x2: +1211

ITT w/ RT x2: +991

After watching a magnificent end to july with TT, it showed a lot of promise. This month, however, I have watched the markets range and TT give signals at almost the exact opposite times that it should have. It will sell at demand zones and buy supply. Both are never good. I will keep tracking it with trades on a demo account but I stopped trading it live. It has proven a great confirmation tool for timing entry if you do the inverse. It has given some and taken others as all systems do. On the g/j, it is down 327 pips whereas ITT is only +82. So, I am drawing the conclusion that some currency pairs are better suited to a different strategy altogether. That may be just as important for this system to work as following your rules for SL and entry. Maybe it is just a draw down period for the system. It’s too early to call it. Especially when we have seen it give wonderful results. I hope you all are not in it live and preserving your capital until the fruits are ripe for the picking.


i had originally reported e/u for july at +400. that’s a big oops and i’ll have to go back and see what i did wrong. meantime, here is a more accurate representation and a comparison with what rpotor’s ea came up with.

btw, i wanted to see what happened when i changed the ea results to mimic the original rules more closely. this is what i changed:

  1. eliminated the second trade if two were taken in a day
  2. showed all positions closed by friday afternoon
  3. didn’t sl unless loss point PLUS spread was hit
  4. didn’t trade a holiday or the day before

the only thing i could not make allowance for were the judgment calls. the ea results were still +825. you can see that mine were +945, a difference of +120 attributable to judgment. imo, it is worth making those calls, but without them the strictly mechanical approach does well, at least in this case.


this is a case where the judgment calls played a bigger role than with some of the other pairs. i made six judgment calls on this pair in july. but, though the results are not good when the machine has a mind of its own, when we bring it back into line with the system, it does a lot better and the judgment calls become less significant. just by taking only one trade per day, avoiding holidays, and closing out on friday, the ea goes to +630 and that leaves about +200 improvement using judgment.

these ea’s are great. they help me to spot mistakes made by hand and give me a new look at the method. i will appreciate it so much if we can just bring them into line with the original rules. thanks for your help.

I wanted to thank you PipWoof as you continue to write on this thread and help us perfect an EA that can come close to your results. I’m unfortunately a trader that doesn’t posses the capacity to code and therefore am subject to the generosity of the other traders. Also, I would like to continue to trade with FinFx and I’m not available at London’s open for manual trading so its been very important to me and many others to follow each detail, of each post, to make this strategy work.

To address some things you’ve recently mentioned, I too have tried to figure if there is a BE threshold that is less than the TP mark. I think 15 pips is most likely the best figure but the market really hasn’t moved enough recently to give a proper sample. Secondly, I’ve experimented with 10 pips T-F instead of 1 pip as the entry point on the daily extreme. Again, not sure if it makes a long-term difference but I’m willing to try.

Thirdly, I demoed the London EA on Eur/Usd and Gbp/Usd and the Daybreak 1.1 on the jpy pairs and the aussie. Interesting idea that the Euro based pairs would perform better with a London based time frame and the the Asian session pairs doing better with their own “Asian” time frame. Unfortunately, my results haven’t outperformed this week to say one way or another. It probably doesn’t make a difference in the long run.

I share this with you mostly because very few posts have come in this week and I wanted you to know that you’re not alone in your continued effort to perfect an already excellent strategy. I’m sure you have a good reason, but has I move forward I wonder if it has occurred to perhaps set tighter SL’s or TP/BE around closer round numbers? It just seems to me that if it doesn’t reach that 25+ TP within the first few hours of the day it’s destined to reverse and hit a sometimes astronomical SL at -49 x3. I personally would prefer a few break even trades to avoid the gutting days, but I’ve been known to be too squeamish in my risk appetite. I would appreciate any thoughts you have in that direction.

Again, thanks for all your willingness to share.

-Armstrong

armstrong, thanks for your support and good wishes. yes, it does seem that the quieter the markets get, the quieter my thread gets. lol. np, though, i have said that if it deserves to die a natural death, it will. it is normal for traders to be attracted to the excitement of success. any craps table will show you that. but, i think they’re still watching, maybe hoping.

as you, i lack programming capability, so my hope is that someone will be available to help us out with what seem to be a few minor tweaks to get these ea’s closer to the rules. it does seem that, even eliminating the judgment calls, a totally mechanized system still comes out.

i am certainly considering some be stop and share your observation that these positions seem to either take off pretty quickly in our chosen direction or go the other way. i really don’t see much where it fails to meet at least the first tp level because the stop was too tight. that would suggest a tighter initial stop and we could certainly be more comfortable with 3 X 25 than 3 X 40 or 49. unless someone comes up with some ea help, i will start looking at that manually in the next few days.