This is a system that uses only one indicator Swing_zz to generate signals. I’ll use the pics to show how to locate entry points. This will work on any pair and any time frame.
I use a TP of 10+spread and a SL of 20-spread on a 15min chart, if I entered all the trades this year on the Euro I’d be up about 170 pips and over 15% on margin used for the year if all trades were the same size, even better if you use a better money management strategy.
The settings for the Swing_ZZ indicator are minbars 4, and you could set the colors to whatever works for you I just reverse the defaults and increase the size.
The signals are easy to locate, ignore the last dot it repaints look at the two dots before the last dot. The upper dot will be your buy entry and the lower dot will be your sell entry, set your limit orders and wait for the next dot to form. I also attached Swing_zz with e-mail alert so you could set up your orders and leave the computer and have an e-mail sent to your phone when the next dot forms.
I think so I get confused with buy/sell stop, on my platform its a limit order with TP and SL, I set up the orders and set the swing_zz to alert me when a new dot is generated, then go check and make any adjustments that are needed after every dot is generated.
I went back a bit and it was a really rough in December but it still broke even with 22 winners and 11 losers, but December is always rough most traders take most of December off, 12/14 was a really bad day 5 losses in a row during NY session.
I would love to forward test this system with an EA, I wish knew how to program, I could repair or a computer and build one from the ground up but I’m lost when it comes to programming. Any programming help would be appreciated If anyone thinks this system is worth the time to program it into an EA.
Ok I have written a VERY basic EA. There is no money management, it simply places 1 standard lot for each trade.
Back testing is very slow, this is because of the swing_zz indicator. As it is not a ‘built-in’ indicator for MT4, we use the iCustom function which slows down back testing dramatically. It took about 15 mins to back test this year so far on my computer, which is fairly new.
My initial tests of the system do not agree with jaws810’s. I made it 29 wins and 15 losses with this EA. Don’t know if that’s a problem with my backtesting data, a bug in the EA or simply because that’s actually the correct results.
One problem with the system is that it has a poor R:R ratio, so one loss wipes out the profit from two wins. And as I had 15 losses, I ended up around break even. Looking at the back test, this maybe improved by adjusting the TP. It seemed a lot of times that the trade hit TP, but could have gone much further. Maybe this could be improved with a trailing stop, I don’t know.
Anyway, give the EA a go and let me know what you think. I’m happy to make changes if needed.
For the EA to work you must have copied the swing_ZZ_with_alert indicator to your experts/indicators directory, and you must not re-name it.
Nice work, it works fine, I posted the results I got from my test for this year, I think that when I did my initial manual backtest I was going off the charts and might have missed some trades due to the spread, different brokers, and model quality. I got 32 winners and 15 losers with an 80% model quality, not bad but you’re right it needs some adjustments and more testing but the tests take forever. Here’s an EA that was posted in another thread that uses swing_zz and backtests go fast, it might help speed up the backtests.
I’m running a backtest with a 20tp and 30sl from March of last year to today, I’ll post the results whenever it finishes. I’d also like to try 1:1 RR ratio and see how it does with a better modeling quality. With money management 32 winners and 15 losers isn’t that bad.
MM idea, a step progression: start with however many lots and after every win increase however many lots then go back down to 1 lot after a loss so you make a lot of money on strings of winning trades. Lets say you start with 1 lot you could increase 1 minilot after every win, then go back down to 1 lot after a loss.
I have 2 backtests running VERRY SLOOOOOW on the Euro one at 20tp/20sl and one at 20tp/30sl, and I’m also forward testing on E/G, Euro, cable, E/J and U/J. I’ll post backtest results when they get done, sometime tomorrow morning, and I’ll post results of the forward tests at the end of every day.
I have been trying a couple things with your trade indicator that might help. If you take 2 trades at once with one using a 20sl and 10tp and the second one with a 20sl but a much larger tp. When the one trade with the 10tp hits the target the second on gets the stop moved to break even and it is allowed to run. This allows you to capture larger runs as well as smaller ones. Not sure if this can be programed for a backtest or not. Thanks for sharing the indicator.
Ok thanks for posting the other EA. I’ve not tried it, but it made me realise where I can make mine faster.
Try this one, you’ll find back testing time drastically quicker.
I will also try Steve4K3’s idea, adding another order with a bigger TP and move SL to BE when the first order takes profit. I’ll post the EA later today.
I’m not ready to release the new EA yet, but initial tests look quite promising.
Using Steve4k3’s idea with two orders seems to work well. I had the initial order, plus another order with SL at 20 and TP at 30 and moving the second order’s SL to BE when order 1 hits TP.
With this setup this gave a 100% return on investment for 2012 for EURUSD (not including December, because my data is crap, no other reason). 2010 gave similar results.
2011 was not as good, so I’m currently I’m working on a money management strategy that will minimise losses when times don’t suit the system as well. Hopefully I’ll get it done this afternoon, but if not, realistically it’ll be Monday now.
I’m running V1.1 on a demo account and it’s opening multiple trades with one signal, it has to invalidate the signal after it triggers a trade, it didn’t seem to have that problem in the backtest at least I didn’t notice it, it might be a server error, I keep watching it.
I have the backtest results, they are all Eur/USD from 3-2012 till today, the first one had the best results with a 20TP/20SL
Here’s a graph from the 20/20 backtest from 3-2012 to today, I don’t know how to post a pic so you have to click on it to open to a full screen, I’ve tried everything to get them big enough but when you enlarge them the distort and are unreadable.
I put all the charts I’ve posted into a zip file you should be able to see them fine in windows photo viewer.Test results.zip (741 KB)
