Triple threat exit strategy

pipwoof · August 5, 2012, 11:49pm

pip, i got my usual two bars from ibfx. the 22 and 23.
g/j, 22 bar hi 122.91, lo 122.73
g/j, 23 bar hi 123.17, lo 122.83
these make for a sunday hi of 123.17, which would be a long entry at 123.28
a sunday lo of 122.73, makes a lo entry at 122.62.
i don’t see anything above the hi and the short was entered on the 1:00 bar which went to a lo of 122.49.
let me look at the e/u, just a minute.

pipwoof · August 5, 2012, 11:55pm

pip,
e/u, sunday bar 22 hi 1.2395, lo 1.2382
bar 23 hi 1.2442, lo 1.2384
makes a long entry at 1.2452 and a short at 1.2371
neither has been triggered yet

Bijoymj · August 6, 2012, 12:01am

I am using +1/-1 pips on Friday bars for entry as my broker is not providing any candles on Sunday

pipwoof · August 6, 2012, 12:20am

pip, oanda just opened their new day bar and are showing me 9 hours of sunday data from their 15 thru 23.
g/j hi 123.16, lo 122.36, which makes a long entry at 123.27 and a short at 122.25.
e/u hi 1.2442, lo 1.2374, makes a long at 1.2453 and a short at 1.2363.
since oanda just started their 00 bar, nothing has been triggered yet in any pair.

pipwoof · August 6, 2012, 3:13am

following five pairs, we currently have four entries:
g/j sell 122.62, currently printing just below 122.50. it is monday, stops to 122.50, locks +10, +10, +10.
a/u, no entry
e/u, sell 1.2371, stop loss at 1.2400, risking -30 X 3 = -90.
g/u, sell 1.5617, stop loss at 1.5650, risking -35 X 3 = -105.
e/j, sell 97.12, stop loss at 97.50, risking -40 X 3 = -120.

edit 3:00 am:
g/j sell 122.62, trades below 122.25. tp +35, air and retrace stops stay at 122.50 locks +10, +10.
a/u sell 1.0543, stop-loss at 1.0575, risking -35 X 3 = -105.
e/u still open as above, risking -90.
g/u sell 1.5617, trades below 1.5600, since we now have a new open position with a/u, reduce risk here by moving stops to 1.2625, risking -10 X 3 = -30.
e/j sell 97.12, trades below 97.00, reduce risk here by moving stops to 97.25, risking -15 X 3 = -45.
summary:
closed trades +35
locks +20
at risk -270

edit 3:25 am:
g/j sell 122.62, trades below 122.25. tp +35, air and retrace stops stay at 122.50 locks +10, +10.
a/u sell 1.0543, stop-loss at 1.0575, risking -35 X 3 = -105.
e/u sell 1.2371, trades below 1.2350, reduce risk by moving stops to 1.2375, risking -5 X 3 = -15
g/u sell 1.5617, trades below 1.5575, tp +40, air and retrace stops to 1.5600, locks +15, +15
e/j sell 97.12, trades below 96.75, tp +35, air and retrace stops to 97.00, locks +10, +10
summary:
closed trades +105
locks +70
at risk -120

Read more: 301 Moved Permanently

dineshydv · August 6, 2012, 3:25am

Fellas,
following this thread silently since day1 and recently started automating this strategy. So below is the EA which places 3 trades(TP, AIR RETRACE) on break of previous’s day high/low by delta pips(parameter).

It opens 3 trades
Set SL on all 3 trades and TP only on TP trade.
Moves SL on AIR & RETRACE trades automatically.

Code: Daybreak_v1.1.zip (19.3 KB)
NOTE: Not suitable on live accounts…

Below is the description of its parameters:

double Delta ( pips difference from previous day high/low. )
double Delta_Mon ( Monday’s pips difference from previous day high/low. )
double Lots=0.1;
int MagicNum=12345;
bool Broker_Sunday_Candle ( TRUE if broker has Sunday candle ) My demo broker thinkforex provides Sunday candle b’coz of their server time

bool UseBreakEven ( TRUE if breakeven trade after certain pips )
int BE_After_pip ( Break even trade after price reaches in profit )
int BE_lockinpip ( At Break even lock-in pips. This includes commision pips )

bool Friday_Close_All ( Close all positions at below time on Friday id SL/TP not hit )
int Friday_Close_Hour
int Friday_Close_Minute

// Below mainly for testing individuals trades
bool Trade_TP ( trade TP trade )
bool Trade_AIR ( trade AIR trade )
bool Trade_RETRACE ( trade RETRACE trade )

I backtested it for 2/3 month and didnt got good results. Could anyone test it out and reply back if i missed out anything…??

Godwin_Igili · August 6, 2012, 3:52am

dineshydv
Will test the EA nd post results. why the ex4? kindly let have mq4 too.
thanks for your work

Godwin_Igili · August 6, 2012, 4:26am

Rpptpor
Thank you for you endurance on the EA for this system.I am demoing it with the ex4 file. On compiling the mq4 file got the following errors, what could be wrong?
Compiling ‘Londonbreak-TT-OrderEntry-v1.01.mq4’…
Function “isNycDaylightSaving” is not referenced and will be removed from exp-file
Function “GetGMTDayHHMM” is not referenced and will be removed from exp-file
Function “GetNewYorkDayHHMM” is not referenced and will be removed from exp-file
Function “GetTokyoDayHHMM” is not referenced and will be removed from exp-file
0 error(s), 4 warning(s)

rpotor:

At the below link is the first beta version of the Londonbreak EA. It is still a work in progress. In this current iteration of the EA, the following rules are observed (numbering as per Bijoymj’s quoted rules above):

#1 rule is observed in the code
#2 rule is implicitly observed since the user decides which pairs he/she wants to run the EA on
#3 rule is observed with the following modification: GMT time does not observe daylight saving adjustments, however London time does. Therefore I understand this rule is meant to work like this: always enter trades between 09:00 London time (1 hour after the London open) and 17:00 London time, regardless if it is winter time or summer time
#4 rule is observed with the following modification: this “phase” of order placement is between 09:00 and 09:05 London time -> if price in this time period is above/below previous day Low/High, then orders are placed.
#5 rule is observed between 09:05 and 17:00 London time
#6 rule is not yet observed, this needs to be manually handled (but please be aware that if you cancel an order and then later the same order is triggered again as per rule #3 or rule #4, then the EA might place the order again (I don’t know if this is true, this needs to be tested, because I don’t know whether a deleted order keeps its magic number or not and whether it shows up in the order history or not)).
#7 rule is observed by setting an expiration on all the pending orders placed, the EA does not close orders on its own.
#8 rule is not yet observed, this needs to be manually handled, but be aware that in its current iteration, the EA will place an order even if you already have a running order on a currency pair that is in negative “profit”. You can cancel it, but then see the comments on rule #6 above.

The EA will still take previous day high/low data at the time of each individual broker’s forming of a new daily candle. So differences between results are still very much possible, mainly on Monday when high/lows might be wildly different between brokers due to the presence or absence of Saturday/Sunday candles.

Please be aware that due to the programming challenge posed by the fact that there are several brokers with wildly different GMT offsets and the EA needs to be able to determine the correct London time on any broker platform, regardless of the user’s local time, I had to use a so called function library to make the coding of this EA easier. This function library can be checked out here (it’s not coded by me): stevehopwoodforex.com • View topic - Broker-independent time-routine: LibGMT.h (actually to be able to see the code, you need to register on that forum, but it’s free, of course and there are very very useful things there).

Now, to make distribution of this EA easier, I had to compile the EA code together with the above mentioned function library. What this means is that the extension of the EA is now ex4. You still place it into your experts folder, like a “regular” EA, but you are no longer able to just open it with a Notepad and check out the code for yourself. This means that if I’m a baaaad, baaaad person I could theoretically program a virus or some other nastiness into the code and you will have no way of verifying that the code “hidden” inside the ex4 file indeed does what I’m saying it does. So this means that you should only download and use this EA if you personally decide that I’m trustworthy enough for you and you believe that I don’t want to do harm to you or your computer. This is the internet, folks and there might be wolves hiding in a sheep’s clothing, so to speak, so this is a judgement call each of you should make before downloading. I can’t give you any assurances, beside my previous contributions to this forum topic and the fact that I take the words of the Dalai Lama and buddhist teachings in general very seriously, therefore I would never intentionally do harm to another human being. Obviously this can also be a lie, since this is the internet and you should take everything with a grain of salt.

Anyway with the above warning, the link to the very first beta version of the Londonbreak EA is below, if you have balls of steel you might try testing it next week, I definitely will and post my results here. The rule still applies that this is to be used only on a demo account, you should not take this “out” to a live account. I’ll try and upgrade the EA to observe the remaining two rules (see above), but I feel like I seriously need to take a break from coding for a while, because seriously, sometimes I’m dreaming about the code, these days. :-))

http://www.rpotor.com/forex/TTE/Londonbreak-TT-OrderEntry-v1.01.ex4

Updating this post: I found the solution for the above mentioned trust issues. Actually the answer is right in the code of the function library I referenced above. By the rules of the author of the function library I’m required to share with you guys the original code as well. Please find these below:

http://www.rpotor.com/forex/TTE/Londonbreak-TT-OrderEntry-v1.01.mq4
http://www.rpotor.com/forex/TTE/LibGMT.mqh

Now, what this means for you?

If you just decide to trust me, you can download the Londonbreak-TT-OrderEntry-v1.01.ex4 file above, install it like a normal EA and be happy about it.

If you don’t trust me, you can download the Londonbreak-TT-OrderEntry-v1.01.mq4 file (please note the differences in extensions) and also the LibGMT.mqh file. With these two files you can compile the EA for yourself with the MetaEditor that comes with all MT4 installations (LibGMT.mqh needs to be in experts/include). If you’re tech savvy enough you know what I’m saying, if not, ask here and I’ll make a detailed post of how to compile an EA for yourself using the mq4 code.

Update2: Ummm, just one more thing that I forgot. Okay, this is seriously the last update to this post guys, and I’m off from here for the day When you put this EA on a chart, please make sure that in the window that pops up when you place the EA, you should have “Allow DLL imports” enabled. Why? Because the LibGMT function library I talked about above needs to “ask” your Windows operating system this question: “Hey Windows, could you please tell me what the exact local time of this computer is?” Now, the EA can only do this if you enable “Allow DLL imports”. If you don’t let it do this, Windows will be rude and will not communicate at all with the EA. Then the EA in turn will become a sad panda and refuse to do anything with an error message.

pipwoof · August 6, 2012, 4:44am

edit 3:25 am:
g/j sell 122.62, trades below 122.25. tp +35, air and retrace stops stay at 122.50 locks +10, +10.
a/u sell 1.0543, stop-loss at 1.0575, risking -35 X 3 = -105.
e/u sell 1.2371, trades below 1.2350, reduce risk by moving stops to 1.2375, risking -5 X 3 = -15
g/u sell 1.5617, trades below 1.5575, tp +40, air and retrace stops to 1.5600, locks +15, +15
e/j sell 97.12, trades below 96.75, tp +35, air and retrace stops to 97.00, locks +10, +10
summary:
closed trades +105
locks +70
at risk -120

okay, by now everyone knows how i feel about mondays and fridays. so, i have a few of those outstanding judgment calls, then need to run off for two or three hours sleep.

g/j sell 122.62, trades below 122.00, tp +35. seeing stall and some reversal, moving air and retrace stops to 122.25, locks +35, +35.
a/u sell 1.0543, don’t like what a/u just did and took it out at 1.0550, -10, -10, -10.
e/u sell 1.2371, trades below 1.2350, reduce risk by moving stops to 1.2375, risking -5 X 3 = -15
g/u sell 1.5617, trades below 1.5575, tp +40, move air and retrace stops to 1.5575, locks +40, +40
e/j sell 97.12, trades below 96.75, tp +35, air and retrace stops to 97.00, locks +10, +10
summary:
closed trades +110, -30 = +80
locks +170
at risk -15

g/u reversed to 1.5575. may not see it on a one-hour, but it did.
summary:
closed trades +110, -30, +40, +40 = +160
locks +90
at risk -15

Read more: 301 Moved Permanently

rpotor · August 6, 2012, 5:10am

This is normal. I get the same warnings at compile. What it means is that these four functions from the LibGMT.mqh function library are not used in the Londonbreak EA, therefore the compiler will not include them in the compiled ex4 file (probably to reduce filesize or something like that). This is not an error, the compiled EA should work fine.

Godwin_Igili · August 6, 2012, 5:20am

Thanks, I have installed the EA. But no trade had been taken yet.Has yours taken any trade?

rpotor · August 6, 2012, 5:25am

Started testing the Londonbreak EA and I found a bug in it. Calculating the correct start and end times for order entries were not correct, hotfixed version below.

http://www.rpotor.com/forex/TTE/Londonbreak-TT-OrderEntry-v1.03.ex4 (for “plug&play” installation)
http://www.rpotor.com/forex/TTE/Londonbreak-TT-OrderEntry-v1.03.mq4 (for manual compile, if you want to do this, then see the last 5 paragraph of this post)

rpotor · August 6, 2012, 5:26am

Nope, only pending orders are set yet, none of them triggered.

Godwin_Igili · August 6, 2012, 5:35am

No pending Orders sets in mine! Did not set Pending Orders on my Alpari UK. But loaded to Litefore platform and pending Orders are set

pipwoof · August 6, 2012, 7:29am

final edit:
g/j sell 122.62, trades below 122.00, tp +35. seeing stall and some reversal, moving air and retrace stops to 122.25, locks +35, +35, stopped.
a/u sell 1.0543, don’t like what a/u did and took it out at 1.0550, -10, -10, -10.
e/u sell 1.2371, trades below 1.2350, reduce risk by moving stops to 1.2375, risking -5 X 3 = -15, stopped.
g/u sell 1.5617, trades below 1.5575, tp +40, move air and retrace stops to 1.5575, locks +40, +40, stopped.
e/j sell 97.12, trades below 96.75, tp +35, air and retrace stops to 97.00, locks +10, +10, stopped.

closed trades +35,+35, +35, -10, -10, -10, -5, -5, -5, +40, +40, +40, +35, +10, +10 = +235.

rpotor · August 6, 2012, 7:40am

Ummm, I don’t know. Mine set the orders correctly. Do you have the Daybreak-TT EA also running? Because if the Daybreak-EA set orders for today already and let’s say you cancelled them, then the Londonbreak EA will not enter orders because it assumes it’s “done for the day”. In the next version of the Londonbreak, I’ll change its magic number, so it will not be possible for the two EAs to confuse each other even if they’re running on the same pair side-by-side. Also, by design the EA only runs once every minute. If you put it on a chart you need to wait at least one minute before it starts to work. Also, it only enters pending orders if the current Bid/Ask price is higher/lower then the entry price it needs to set. Also there is the issue of stop level. For each broker there is a minimum difference by which a pending order should differ from the current market price. Let’s say our current Ask price is 99.27 and we want to set a pending buy stop at 99.29 Now if our broker’s stop level is let’s say 3 pips, then the order will not be set, because the minimum buy entry price we can set would be 99.27 + 3 pips = 99.30

As you see, there might be a couple issues that could prevent a successful order placement, however, if you let the EA running, it should continually try to place the “missing” orders and once all stars align it should be successful. On my platform (also Alpari UK) it didn’t have any problems placing the orders, so please test it on your end for a few days and let’s see if this problem persists.

PipWright · August 6, 2012, 7:46am

Thanks Pipwoof.
My bar is within a few pips of yours.
G/j hi:123.18 lo:122.18
E/u hi: 1.2443 lo: 1.2377

I wouldn’t suggest using friday and Sunday data to set levels for Monday. Base it of of Sunday only and you’ll get better results.

rpotor · August 6, 2012, 8:03am

Pending trades for today as set by the Londonbreak EA (+10 pips entry delta because of Monday). No trades triggered yet.

audusd high/low 1.05691 / 1.04447
audusd buy 1.05801
audusd sell 1.04337

eurjpy high/low 97.383 / 95.013
eurjpy buy 97.493
eurjpy sell 94.903

eurusd high/low 1.23918 / 1.21659
eurusd buy 1.24028
eurusd sell 1.21549

gbpusd high/low 1.56559 / 1.55038
gbpusd buy 1.56669
gbpusd sell 1.54928

gbpjpy high/low 123.03 / 121.065
gbpjpy buy 123.140
gbpjpy sell 120.955

Seems these Mondays are waaaaaay off between pipwoof’s numbers and Alpari UK (and by extension I guess all brokers where there are no weekend candles). For example pipwoof reports above a g/j sell entry at 122.62, currently in some good number of pips profit, while as you can see Friday’s low for g/j on Alpari UK was 121.065, so we will only enter sell at 120.95 (11 pips below Friday’s low) or at best 121.05 (1 pips below Friday’s low). So most likely we won’t even see any entries on g/j today while pipwoof pockets some nice profits… I’m starting to envy those on “weekend-candle” brokers.

pipwoof · August 6, 2012, 1:02pm

no promotion intended here, but ibfx offers free non-expiring demo accounts. everyone here who is seriously considering the approach should download a demo. it will facilitate our communication as you can be looking at exactly what i’m talking about. if you wanted, it would enable you to base your signals off the ibfx data, but use another platform for your entries.

another free download is available here: Vitrite - Download. this is a utility that enables you to make any selected window transparent. you can even control the degree of transparency on a scale of 1-9 with ctrl/shift and your choice of numbers. reverse the transparency by ctrl/shift 9. example, keep the ibfx platform as non-transparent. place your own platform on top with a transparency around 5-6, be sure to line them up so you know you’re looking at the same bars. you can now see both charts and yours is on top, available for orders.

doing this, would you be able to use rpotor’s order entry ea? i don’t know, but i trust he is willing to help us with that. perhaps he can devise a way to use the ibfx entry points with any platform. since i am not yet enamored with the london open entry proposal, i have not familiarized myself with his new ea’s and am still learning what to do with v3.21. certainly learned last night to light up the ea on daily bars and, whatever you do, don’t cancel the orders unless you’re ready to redo them without help from the ea. lol! not even deleting, recompiling, and reattaching keeps it from saying, “nope, been there, done that. not doing it again.” missed the first half-hour of the movie.

now, count on it when i say this thought is among the most spurious and unfounded. here it comes. is it even remotely possible that gmt 00 represents some kind of psychological beginning of the day and, like round numbers or fibs, might have some subconscious attachments? if so, does that offer any advantage at all in staying with that beginning? whoooooeeeeoooo, spooky. it may be a long shot considering that probably the only place in the world currently on gmt 00 is Reykjavik, Iceland. on the other hand, we build all our time zones and world clocks around gmt +/-. let me know when you have had time to think about this and when you have those downloads taken care of. let’s get on the same page.

pipwoof · August 6, 2012, 1:54pm

this was one of those days that favored entry at gmt 00 rather than the london open. there will be days that go absolutely the other way. today favored cut and run when our positions were in profit. there will be days that watch us run, then move in the original direction like phelps on the straight. we just have to decide which choices probably have the advantage.

you can see above how gmt 00 entries and judgment came out. for london open bases, g/j would have caught the first tp had you entered at 7, but stopped out if you waited 'til 8. same for g/u. a/u and e/u would have been stopped either way. e/j at 7 is currently negative, but still open, 8 entry would have been stopped. did i mention that tomorrow may favor the london open idea?