yesterday’s a/u trade still holding pos 2 & 3,
pos 2, air, stop-loss 1.0400, locked in +81
pos 3, retrace, stop-loss 1.0350, locked in +31
new entries today:
g/j sell 123.07
a/u buy 1.0373
e/u buy 1.2306
g/u buy 1.5668
a/u has broken 1.0425, so tp hit at 1.0400, +27
air stop 1.0400, retrace stop 1.0375
note: if you chose to give retrace more room, you would set stop at 1.0350, risking -23, i chose the conservative
after a low of 122.90, g/j began to reverse, stopped out at 123.25, -18, -18, -18
g/j flat
e/u stopped out at 1.2275, -31, -31, -31
e/u flat
a/u is holding two trades from yesterday with lock-ins +81 and +31
a/u has hit tp on today’s entry, +27
a/u is holding two trades from today with air stop at 1.0400, lock in +27, retrace stop 1.0375, lock in +2
g/u hit tp at 1.5700, +32
both air and retrace for g/u stop moved to 1.5675, lock in +7, +7
edit: g/u touched below 1.5675, remaining positions closed, +7, +7
g/u flat
summary:
we have closed losses today of -18, -18, -18, -31, -31, -31 = -147
profits closed today of +27, +32, +7, +7 = +73
locked in as of now, +81, +31, +27, +2 = +141
we still have 4 a/u positions open, all with locked in profits
if anyone is following these sequences, let me know if this is helping you to understand the method. thanks.
What usually happens with TSL strategies, we should have majority of trades hitting either SLs or avg. profits (hopefully more of those), especially when the pair is in a range. It’s when you get some rare fat tails, big winning trades, then you realize how profitable this can be.
I have started my EA to trade on same 4 pairs as Pipwoof. I got off quote errors on GBPJPY and so did not take those trade, broker issue on Demo, I started it late on 7/17 so it took EURUSD price from where it started. I made so far 7.9% profit in less than 1.25 days of trading, had I taken trades on eurusd and gbpjpy, I would loose 5%, still 3.9% in 1.25 days not bad a t all. I will have to check why my program did not take third trade on gbpusd as you would see only 2 trades. Summary we took 15 trades, 9 made money 6 lost, so we made 9.9% and gave back 6%.
at this time we are holding 4 long a/u positions.
2 from yesterday at 1.0319 and 2 from today at 1.0373
as thursday slowly winds down and we head into friday, i would make a judgment call
no additional positions
move stops on all 4 a/u to 1.0400. this locks in +81, +81, +27, +27 = +216
this is now a worse-case scenario with our current balance today -74
if we settle for +216, we are net +142
if a/u favors us with an additional move up between now and tomorrow, better. otherwise, we will be looking for a good place to take these out.
Im liking the idea so far. Y’all are doing a good job of making it easy to follow. Just curious, where do you stand overall? Are you going to post overall P/L on regular intervals.
I haven’t forward tested it yet as I have an ea running on the demo acct. I also trade live on Oanda. So, I’m curious how this goes.
Just a thought, do you take into consideration the gmt offset or just base entry off the candle your broker calls a “day”?
I know my ea bases on gmt time. So, I have to watch 0:00 gmt. I was just pondering if trading the 0:00 gmt candle would yield different results.
I like your system. Start to test now…
yesterday a/u h/l is 1.0443/1.0356 but you entered today is 1.3073. how?
i use ibfx data for all results. their day is defined by gmt. a/u h/l for 7-18 was 1.3072/1.0287. brokers will define “day” according to their platform time and different brokers will have different numbers, therefore, different results. this issue is also covered in the daybreak thread, posts 21, 30, and 113. i have also been profitable with oanda data, but have only checked the two brokers.
looking back is tedious, but the best way to see how your own broker comes out. check at least a whole month, all pairs, and see how close you get to the results posted here. good luck!
my objective with tracking a few specific days is to help those trying to understand the stops. since different platforms and judgment calls are parts of outcome, our p/l’s are going to be different. better that you get familiar with your own platform and see your personal results.
i’m not entirely clear on your platform time. if i understand, you are using oanda data (edt), but basing “day” on gmt. if so, your results should be very close to mine.
this is from post 21 on the daybreak thread: “That is, the definition of “day” will depend on your platform and may give you completely different highs and lows from what I am using. I am in Chicago, central time. Oanda is on EDT and their day for me is 11PM to 10PM. The IBFX day is on GMT and that is 7PM to 6PM for me.” see response to sheriff below, also. since i am using ibfx data for results, if you are basing on gmt, you should be very close. oanda and ibfx seem to differ only a pip here and there.
appreciate your interest. hope this helps.
Read more: 301 Moved Permanently
Pipwoof, Why did we replace UJ with GU (I do like the change)? More profitable I surmise…
If we are trading daybreak only rules, should we also make the currency pair change over there, as well? The triple threat also has a max SL of 50 pips…change the daybreak SL, as well?
How do you feel about adding any other additional pairs?
I tracked 14 currency pairs with the original strategy for Friday…nearly +400 pips…the only loser was NU and that was only 9 pips. So the market pretty much moved in unison for a positive gain. In your experience, does the market also move pretty much in unison [U][/U][I]against[/I] you at times?
Did I mention this is fun? :59:
Whoops. I probably should have posted this over at Daybreak, so next time I will.
bk, i made the original choice of pairs for daybreak on a combination of eyeballing their movement, raw results, and some influence from my futures trading. in retrospect, we could have looked better with some changes.
the four i chose looked inviting on what i call a casual observation of the daily bars. in other words, as i walked through the dailys, i would mutter to myself, “winner, winner, loser, winner, loser,” just looking at where they ended today relative to the highs and lows of yesterday. the influence from my futures trading is that you always add a clinker to your portfolio. eperience in futures, myself and others, showed that a portfolio system would often yield great tested results on, say, 3 or 4 contracts, and less than stellar on 2 or 3 others. but, sure as hell, it would be those at the bottom that performed better when you went live, sometimes saving your butt from the others. honestly, i think i’m ready to give up that habit for forex.
now, raw results. i defined raw results as what happened with a spreadsheet just looking at going with the break and holding until close. no stops, no eliminating sundays, nada. break to close. it didn’t surprise me that the gbp/jpy, wild little party, did well. but, there were some other findings that might suggest a different portfolio from that reported on daybreak.
raw results were approximately:
gbp/jpy, +18,500
aud/usd, +7,400
usd/jpy, +7,350
eur/usd, +5,100
gbp/usd, +1,850
usd/cad, +10,200
usd/chf, +600
eur/jpy, +10,700
from these findings, it appears that usd/chf wouldn’t have paid the rent and, surprisingly, the eur/usd looks weak. better choices may have been usd/cad or eur/jpy. the gbp/usd was chosen for triple threat because it appeared to reach tp a significant number of times. however, the raw results indicate it does not have the staying power during the day to reach the larger moves. again, usd/cad or eur/jpy would probably work better.
providing some backtesting by hand for triple threat was tedious. to get a year and a half results on a pair took about two long days of steady work. i haven’t done that for usd/cad or eur/jpy, but it certainly looks like they might be valid substitutions or additions to the portfolio.
in the end, i think it would be good for anyone considering trading either daybreak or db with tt to do some hand evaluation of their own. for one thing, it will familiarize them with the way the pairs move and what to expect when they are in live trades. they will, and should, trust their own findings more than mine and can choose for themselves which pairs they expect the most from.
while i’m at it, i will address trading difficulty here. the only reason i haven’t gone live with db with tt is the demand on my life. to trade this way, i would enter at 7 pm on ibfx. during the night, alerts would wake me to move stops, often several times. get up from a sound sleep, wake up, pay attention, make the changes, try to get back to sleep, doze off during the day. i’ve done that routine before and it does wear on you. just db is simpler, but, of course, doesn’t offer the improved results using tt.
yes, it does seem that these pairs often move in unison, one way or the other. one thing that seems to often be a saving grace is when one pair just touches up to go long, then falls like a rock. it does happen that another pair may follow a similar move, but stop just short of entering that initial long trade. thereby, compensating for the loser as they both move south together, one making pips, the other losing.
finally, regarding stops. remember that db is a break to close method and my studies were limited to the use of a spreadsheet. using daily data and a smaller stop, i can’t know how often the trade was stopped out before moving on to a more successful close. for example, a trade opens long at 120.50 and closes at 121.00, showing on the spreadsheet as +50. however, when we look back at a shorter time frame, we see the trade was stopped out long before it got to close. it may have required a sl of 75, 100, 150, to actually allow it to proceed to close. the large emergency stops are there to give the trade every opportunity to proceed to close. a process which shows a profit.
yes, this is fun. dogs woke me up at 5:30 and wife stayed in bed. i can play on the 'puter till she’s up. i was a silent member of bp and other forums for many years. i thought about what i was in for when i rejoined and decided to post a method. but, long ago i promised a mentor i would help someone else along the way… i can just hope this actually ends up being help.
Pipwoof,
Its perfectly valid to take into account the larger timeframes like daily bars for trade exits, however in that case stops will be based on daily close too. As you say "finally, regarding stops. remember that db is a break to close method and my studies were limited to the use of a spreadsheet. using daily data and a smaller stop, i can’t know how often the trade was stopped out before moving on to a more successful close. for example, a trade opens long at 120.50 and closes at 121.00, showing on the spreadsheet as +50. however, when we look back at a shorter time frame, we see the trade was stopped out long before it got to close. it may have required a sl of 75, 100, 150, to actually allow it to proceed to close. the large emergency stops are there to give the trade every opportunity to proceed to close. a process which shows a profit. "
Read more: 301 Moved Permanently
Did you calculate your stops based on daily closes as well. Let me know which timeframe you used to calculate/eye ball your exits? I can add that to may EA along with emergency exits, say 150 pips, or daily ATR etc. Because my back tests do not match your results and may be there is something that I did not code properly, after updates I intend to share my EA with those who request it.
Spreadsheet, original studies.zip (277 KB)
farruk, i appreciate your continued efforts to make a working ea. much of what bobkat and i were covering in the posts above pertained to the original daybreak thread. that thread does not use the triple threat exits, but is designed as a break to close method for those who are unable to watch the markets and move stops as is required by the triple threat exit method. the wide stops in daybreak are designed to permit the trade to go to close as often as possible. using the triple threat exits is a completely different horse. if you want to see, attached is the first spreadsheet i used for the daybreak thread. as i covered in post 1, the limitations of the spreadsheet are: 1. uses +/- 1 for mondays as well as for all other days, 2. takes both trades on days when two are triggered, and 3. daily bars are unable to tell if a certain stop was hit prior to close. to obtain results for this triple threat thread, i used hourly bars and logged trade results by hand.
for point of reference, here are the gbp/usd trades i have logged by hand so far this month:
7-2, sell 1.5660, -15, -15, -15
7-5, sell 1.5571, +45, +45, +95
7-6, sell 1.5497, be, be, be
7-9, buy 1.5497, -5, -5, -5
7-10, buy 1.5534, -35, -35, -35
7-11, buy 1.5548, +25, be, be
7-12, sell 1.5483, +30, +10, +10
7-13, buy 1.5516, +30, +30, +30
7-16, sell 1.5565, +40, +15, +15
7-17, buy 1.5654, -5, -5, -5
7-19, buy 1.5668, -20, -20, -20
and here are the eur/usd so far this month:
7/2, sell 1.2638, -15, -15, -15
7-5, sell 1.2506, +30, +105, +230
7-9, buy 1.2290, -15, -15, -15
7-10, buy 1.2324, -25, -25, -25
7-11, sell 1.2233, +30, +30, +5
7-12, sell 1.2211, +35, +10, +10
7-13, sell 1.2165, -10, -10, -10
7-16, sell 1.2233, +30, +30, +30
7-17, buy 1.2289, -40, -40, -40
7-19, buy 1.2306, -30, -30, -30
7-20, sell 1.2227, +25, +25, +25
it would help if someone else would take a look at these trades and see if they can come up with the same or similar results, some judgment calls notwithstanding. just take a look and determine if you agree with my thought process on these trades. i also believe when we have an ea that starts getting close to replicating these results, we will be on the right track.
a nice start for the week as all 4 pairs trigger short trades:
gbp/jpy, short 122.25, tp hit +25, air out at 121.50 +75, retrace open with stop at 122.00 +25 locked in
aud/usd, short 1.0323, tp hit +45, both air and retrace stops at 1.0300 +20 and +20 locked in (monday decision)
eur/usd, short 1.2104, after hitting a low of 1.2081, began to reverse and all 3 closed at be
gbp/usd, short 1.5588, tp hit +35, air and retrace stops set at 1.5550 +35 and +35 locked in
closed trades: +180
locked in: +135
edit: gbp/usd prints below 1.5500, air stop moved to 1.5525 locks in +60, retrace stop stays at 1.5550
total locked in now +165
edit: aud/usd prints below 1.0250, air stop moved to 1.0275 locks in +45, retrace stop stays at 1.0300
total locked in now +190
edit: gbp/usd touches, and i do mean just touches, 1.5525. air closed +60, retrace stop still at 1.5550
closed trades: +240
open trades: gbp/jpy short 122.25, retrace open with stop at 122.00 locks in +25
aud/usd short 1.0323, both air and retrace still open with air stop 1.0275 locks in +45, retrace stop at 1.0300 locks in +20
gbp/usd short 1.5588, retrace open with stop at 1.5550, locks in +35
total locked in: +125
edit 11:00 am cst: because it is monday and the market has kinda gone into that sleepy afternoon drift around, i would make the following judgement calls:
gbp/jpy, move stop to 121.75, locks in +50
aud/usd, leave air stop at 1.0275, move retrace stop to same place, locks in +45 and +45
gbp/usd, move retrace stop to 1.5525, locks in +60
now locked in total: +200
pipwoof, since GJ is the runaway winner over time, would it make sense to trade triple threat with just this pair? I realize that the basket approach would render more consistency, but GJ raw data results are +18,500 vs +11,700 when using the top 4 performers. The triple threat 4 pairs only average +8,213.
Please correct me if I am wrong with this conclusion, but there appear to be three main advantages trading just GJ:
Thanks again for supplying these 2 strategies and all of the data for us to make decsions on how we can individually best apply these concepts to our own trading styles. :35:
bk, i think you make a good point, especially when i consider how often i want to wake up to alerts in the middle of the night! much easier to deal with one going off instead of four at all different times. i suppose some of it gets down to how much we want to do. for the top four pair, just using daybreak, raw data, we get almost +47,000. of course, we would suppose all we have to do to get that with gbp/jpy is increase lot size. there would be the assumption that gbp/jpy would continue to perform as it has in the past. but, i suppose that’s an assumption with any backtested method.
there is also the assumption that, should gbp/jpy fail, the others would not carry the portfolio. since we have observed that these pairs usually move in unison, i doubt we could count on one or more to behave differently enough to save us in a drawdown situation, unlike silver and soybeans, heh.
one more thought. on the results i got by hand for a year and a half using triple threat, gbp/jpy got just over +10,000, but aud/usd was close behind with around +9,700. i think one or both of us need to relook at gby/jpy, aud/usd, and eur/jpy (a pair i have not looked at in detail at all) for that time period and satisfy ourselves that we think we know what we’re about to do. i must say, though, overall it sounds very appealing to me and may be what i need to be willing to go live and interrupt my night at least minimally. thanks for the idea.
pw, thank you for your response…you offer more insight and I appreciate it.
I do like spreading the risk a bit, such as using your idea of both GJ and AU. The profitability is still high, and we have 4 different currencies involved.
I have a question regarding what period of time most closely mirrors today’s performance of the currency pairs we are tracking.
What concerns me a bit about the data is the very long period of time the data was collected. I believe that the US economic meltdown 4 years ago may have indelibly altered the financial landscape and forex market. What may have amounted to relative financial stability prior to the meltdown no longer exists?
Is it possible to truncate the data and find the results over the past 4 years (new era economics)? This may not be the solution either as the bailouts and stimulus package have provided interventions that under normal times would not have occurred.
once again, turns out you are absolutely right. really don’t have any way to look at db using tt, but could look at the spreadsheet data for db. the method remained profitable from 1-10 thru 6-12, but much less so than 1-07 thru 12-09. this time, i used the modified data, i.e., excluding sunday trades and adding failsafe stops.
1-07/12-09: gbp/jpy +22,200, aud/usd +5,000, usd/jpy +8,000, eur/usd +4,500, gbp/usd +7,100, eur/jpy +9,600
1-10/6-12: gbp/jpy +24,200, aud/usd +9,100, usd/jpy +8,600, eur/usd +7,538, gbp/usd +7,000, eur/jpy +11,784
these numbers are the equity growth from one point to another. in other words, gbp/jpy made +22,200 in the first segment and added only +2,000 in the second. aud/usd did well and the euro not bad. but usd/jpy only added +600 and gbp/usd actually slipped a little. if we look at the equity curve on post 1 of db, we can see the flattening of the curve in the past three or four years, but to separate the data like this is revealing.
i may do some spot checking prior to 2011 using tt, but remember that my hand studies for db with tt are the most recent year and a half. it could be that we had more dramatic moves and tendencies to trend in the first segment and we are seeing extraordinary choppiness in recent years. but, db with tt could capitalize on shorter moves and still catch the possibly less frequent big moves. from my mentor, “all systems fail and not because you just bought 'em. they fail because markets change.”
i’ve about decided i’m willing to get up a time or two at night for the yields shown for gbp/jpy in post 2 here. thoughts?
final stops placed as described in post 55 have been hit. total for today: +440
