Will share my first EA for help with backtesting

Hi guys,

After quite a number of years (learning) in forex, I have now started to develop my own EA (FXDT), with help of a skilled developer.
My preferred strategy revolves mostly around repulse indicator (to indicate trend) and HAOS (to indicate dips and tops for entry & exit).
So far I am quite profitable with it.

On the above basis I have three sub-strategies:
1. let's call it 'low intensity';
SL can be set but I prefer not to use it or set it far away, and close trades either in profit on signal, on TP, or by conservatively averaging them out.
Recovery trades only open after at minimum x pips away. Can take a long time but will do the job.
Naturally small initial lot sizes and a hard stop can be set on % drawdown to close all trades.
You won't get rich overnight but not lose sleep over it either while still over time accumulate your balance.
(forward test: FXDT HAOS v06 | Myfxbook)

2. 'Medium intensity'
Stop-Loss set at around 50 - 100 pips, but mostly closing trades on signal.
No restriction for recovery trades to kick in on nr of pips away from previous open, but recovery trades are triggered by signal.

3. 'High(er) intensity'
Again Stop-Loss set at around 50 - 100 pips but mostly closing trades on signal.
Even more relaxed condition for opening recovery trades, mostly depending on close of previous bar.

For 'low intensity', I have my first EA version coming out early next week.
It comes with a lot of options like draw-down control, partial closing, position sizing options (initial & recovery trades), limit max number of pairs traded simultaneously, prevent opening of trades on high spread, current drawdown etc.
So I aim for an EA that proves profitable over the long term and lets you sleep well at night.

It would be great if there are people out there interested in helping me with high quality backtesting and finding optimal settings for currency pairs, in exchange for a year long free use of the EA.

If interested please let me know. I will also post here some test reports but don't have neither enough time nor good enough data quality for top quality testing..

BR,

Barry

I have Tick Data Suite and can help you out with 99% back testing.

Great!
I'll reach out to you probably mid next week when I hope to have the first version ready

@mrdfx please have a go at attached EA. I added one setfile I'm now trying out. But as said needs quite a bit of quality testing and optimization.

Ok I'll test your set file and see what else I can come up with and share the results good or bad. Give me a day or 2.

Ok, I tested your EA, and I have some feedback.

Firstly, the good news is that it does appear to be profitable on initial testing; however, your EA is very slow. I started with a one-year-back test using your set file and I noticed it was taking a long time, so to speed things up, I did a three-month-back test, thinking it would be quicker. Well, I was wrong. It took 21 hours to do a 3-month back test. I'm doing another test now with some different settings I chose, and I'm about 10 hours in and not even close to half way yet.

A well-written EA should take around 5–10 minutes max to do a 1-year-back test, not hours.

You need to analyze your code to find out what is causing the bottleneck and fix it. I would suggest doing a debug using profiling to see what part of the code needs fixing.

Unfortunately, until this is fixed, I can't do more testing as it takes too much time to do so.

Allright thanks for this useful feedback!

No problem, when you figure out whats causing the bottleneck and get it fixed I can test it more.

Have you had any progress with your EA? Have you managed to speed it up?

Hi, yes I believe I did, hope this EA loads OK for you? intended for pairs where the rate is 1.xxx or 0.xxx and on M15 timeframe. I would set position sizing not below 1.000.000 and max 1 pair traded at a time.

Variant for 1 minute chart

This version cuts loss earlier so safer

This version is more profitable.

On my end it works fine, but ex4 file may be missing some resource files, checking for bugs now.

Code optimised by the developer. On my end this works fine. For 1 min timeframe, pairs with rate 1.xxx or 0.xxx

Hi BarryD, I gave this new version a test on EU on M1 default settings. The results were not profitable and the EA still tests extremely slow. It took over 8 hours to do a one year back test. Are you using resource files with iCustom? Doing this can make an EA function very slowly, if so you can make it more efficient so that it doesn't run so slow.

Besides this your previous initial version from post #4 functioned better and was profitable.

Thanks again for your feedback. I have since improved the code further and published the ea and signal here
Buy the 'Atreyu' Trading Robot (Expert Advisor) for MetaTrader 4 in MetaTrader Market.