Will code in MT5 for free.. EAs only, can't be f*cked with Indicators
I am thinking when trade is halfway to TP (150 fib level), we move to BE+5 pips so that way there is no way we lose from that trade. I feel like if not for news today, I would not have hit my TP target and we would have the same issue as yesterday where we got faked out and then it would've hit SL. {image} Running Total - 1/2 Trades Won - 50% Accuracy
Got your SL idea coded... doesn't perform too well. Standard shit below, just tested for 2018. $1000 account, 15-45 pips window limit, 1 trade per day max, 0.01 lots. Severely hinders profit for a bit of a minor trade-off on DD:
Will code in MT5 for free.. EAs only, can't be f*cked with Indicators
- #1,323
- Edited 4:15pm Oct 26, 2018 3:39pm | Edited 4:15pm
- | Joined Jan 2018 | Status: Trader | 70 Posts
{quote} Thanks for considering my post..ive been sort of applying your logic manually after I get an open trade. Another thought I had was if the candle breaking closed too far (and so limited trades) that an EA could still open at any later candle that retraced back to within the range set for a valid entry. It would have to close just break back into the valid entry.
Please see below. Same settings as above, and includes a 10% maximum breakout with or without a 'First candle only' rule. (useFirstClose true means only the first breakout can be used, false means later ones can if they abide by the max rule). So as described, if subsequent candles close in our eligible gap, it will take trades later on. If anything, this amplifies losses; probably because our breakout has shot straight through the buy line for example, lost momentum, and we may be catching the retracement each time.
Didn't like the limited testing in this with just 2018, so ran it back to 2016:
Seems it doesn't impact us too much, but definitely doesn't seem to have any positive benefits.
Haven't programmed any tester logic in, so the useFirstClose is only applicable when useMaxPercent is true... hence why when its false, both results are the same.
EDIT: This post above was changed quite heavily, as I found I still had IvanYo's SL setting applied from the previous tests.
Will code in MT5 for free.. EAs only, can't be f*cked with Indicators
{quote} let say you opened 0.1 lot on moday and hit SL, on tuesday you opened 0.2 lot to cover monday lost.. but the problem is when you had losing strike.. ut from my observation, maximum consecutive lost is no more than 3 days in a row...
Programmed a basic Martingale management into the system.. Doesn't operate too smoothly though, as we could lose 5 trades which are 45 pip windows, then our win could be a 15 pip window or an end-of-day close for 0.5pips or something. Could probably program around this though, and essentially ratio our Martingale based on pips rather than trades.
Backtested to 2016, and there are a few 5-6 loss-streaks which hurt a bit. Had to up the account to $10,000 as the $1,000 wasn't beefy enough to take the loss-streaks, even at 0.01lots as standard.
Will code in MT5 for free.. EAs only, can't be f*cked with Indicators
{quote} Evening dimash, Programmed a basic Martingale management into the system.. Doesn't operate too smoothly though, as we could lose 5 trades which are 45 pip windows, then our win could be a 15 pip window or an end-of-day close for 0.5pips or something. Could probably program around this though, and essentially ratio our Martingale based on pips rather than trades. Backtested to 2016, and there are a few 5-6 loss-streaks which hurt a bit. Had to up the account to $10,000 as the $1,000 wasn't beefy enough to take the loss-streaks, even at 0.01lots...
You’re a great coder, I admire your work. 
Grind until you no longer have to introduce yourself. 4:30am club
- #1,331
- Edited 5:48am Oct 27, 2018 1:26am | Edited 5:48am
- | Joined Jan 2018 | Status: Trader | 70 Posts
My apologies, doing this a lot.. I still had multiple trades turned on, as its my default...
If you're not trading multiple trades per day (my default allows this), the max is 4. Worst place for this is Oct 5 2016 to Oct 20 2016, where we have 2 4x loss streaks back to back with 1 win inbetween, and the win is a small window range.
If you are trading multiple, this same window gives a 6 loss streak, whilst Apr 20 2016 is also a shitty day giving 5 in 1 day.
Results without multiple trades per day: (note that due to only 4 consecutive losses, I could drop to a $1000 account)
If you're not trading multiple trades per day (my default allows this), the max is 4. Worst place for this is Oct 5 2016 to Oct 20 2016, where we have 2 4x loss streaks back to back with 1 win inbetween, and the win is a small window range.
If you are trading multiple, this same window gives a 6 loss streak, whilst Apr 20 2016 is also a shitty day giving 5 in 1 day.
Results without multiple trades per day: (note that due to only 4 consecutive losses, I could drop to a $1000 account)
Will code in MT5 for free.. EAs only, can't be f*cked with Indicators
- #1,334
- Oct 28, 2018 3:40am Oct 28, 2018 3:40am
Worst place for this is Oct 5 2016 to Oct 20 2016, where we have 2 4x loss streaks back to back with 1 win inbetween, and the win is a small window range.
. Like to read this thread and see what is happening. Maybe I can share some of my questions and ideas with you all.Regarding the different box sizes: it may be better if the risk is the same every day. So if the box is e.g. 45 pip wide you trade 1 Lot and if it is 15 pip you trade 3 Lots. Martingale then wold multipliy it as before.
How do you handle if the breakout is to far away? E.g. the box is 20 pip wide and the candle closes 15 pip above. Then you have a risk of 35 pip for 5 pip TP. Am I right? And what if the breakout is 25 pip above the box?
Regarding the backtest: have you double checked your data? As you know the European, US and more use DST/no DST during the year while the Japanese do not. That might influence the time for the Asian Session Box in your history.
Have a nice weekend
!
{quote} First of all - great job. Appreciate it
{quote}How do you handle if the breakout is to far away? E.g. the box is 20 pip wide and the candle closes 15 pip above. Then you have a risk of 35 pip for 5 pip TP. Am I right? And what if the breakout is 25 pip above the box?
{quote}Regarding the backtest: have you double checked your data? As you know the European, US and more use DST/no DST during the year while the Japanese do not. That might influence the time for the Asian Session Box in your history. Have a nice weekend
Will code in MT5 for free.. EAs only, can't be f*cked with Indicators
Just sharing for those who care - I managed to get a hold of an MT4 demo account and did some back-testing since October 1st. The results are 13 wins and 7 losses, which isn't quite the 80% accuracy that is claimed in the thread title but not sure if the data has outliers or whether the small sample size has to do with it. Regardless, it's still a profitable strategy if you manage your money well. I will wait for the last two days of trading to see finish off the October test and see what % accuracy we end up with. {image}
Differences:
03/10/18 - No trade as box is > 45 pips
08/10/18 - Win
09/10/18 - Win
10/10/18 - Loss
I won't show the rest for screenspace purposes. Your results aren't far off though. With standard settings I have 12 wins, 7 losses (63%).
Will code in MT5 for free.. EAs only, can't be f*cked with Indicators
Hey guys, I've noticed an issue with the 3 Tier London Breakout V.3.2b indicator and Forex.com. Wondering if anyone else is experiencing this as well?
For some reason on Sunday evenings (EST), the indicator in MT4 with Forex.com gets all weird and can't create a box properly. Yet on Tuesday thru Friday it works perfectly.
The box for me in Forex.com starts at 22:00 to 2:00, but this is irrelevant since I can't get the box to form/generate/show no matter what time I enter. This only happens on Sunday evenings (EST) for some reason.
Interestingly I don't have any issues with MT4 when it's connected to OANDA, so I use a OANDA demo as a reference for the entry buy and entry sell and apply them to Forex.com.
Overall I find this breakout strategy to be good. It's definitely NOT the holy grail strategy that will generate you millions based entirely on an EA, unless you throw a ton of money at it and are in it for years.
Like FXMD said, it's an 80% winning strategy, but ONLY VISUALLY, and this is obviously quite different in real life when you factor in candle height, box height, opening on cross, opening on close, etc., etc., etc. But nevertheless, thank you FXMD for the foundations of this strategy. There are some similar strategies on YouTube based on the breakout model (London breakout), but I find that FXMD's settings work the best for the EURJPY.
For some reason on Sunday evenings (EST), the indicator in MT4 with Forex.com gets all weird and can't create a box properly. Yet on Tuesday thru Friday it works perfectly.
The box for me in Forex.com starts at 22:00 to 2:00, but this is irrelevant since I can't get the box to form/generate/show no matter what time I enter. This only happens on Sunday evenings (EST) for some reason.
Interestingly I don't have any issues with MT4 when it's connected to OANDA, so I use a OANDA demo as a reference for the entry buy and entry sell and apply them to Forex.com.
Overall I find this breakout strategy to be good. It's definitely NOT the holy grail strategy that will generate you millions based entirely on an EA, unless you throw a ton of money at it and are in it for years.
Like FXMD said, it's an 80% winning strategy, but ONLY VISUALLY, and this is obviously quite different in real life when you factor in candle height, box height, opening on cross, opening on close, etc., etc., etc. But nevertheless, thank you FXMD for the foundations of this strategy. There are some similar strategies on YouTube based on the breakout model (London breakout), but I find that FXMD's settings work the best for the EURJPY.