Anyone Want to With Some EA Edits?
AUDJPY short only
Mirror test long only.
Observations I had made on this;
1 - It does poorly in ranges/choppy markets.
2 -It does poorly trading aganst a strong trend.
3 - It does well trading in line with a trend.
4 -It does very well trading as a trend correction strategy.
Possible ways to improve;
1 - ADX filter would cut out the mass of these losing trades.
2 - A triple SMA indicating strong trend with rules to not trade against would cut out losers here.
3 - ADX coupled with SMAs (poss 200) could optimise it trading in these conditions.
4 - Adding fibs retacement levels and trading more aggressively in prime conditions may work well.
Being able to manually add in major SR levels that it would stop tading at (or even look for reversals from) would beneft this aso since as it is, it just constantly opens new trades when it closes a batch at a pre-defined dollar profit. Having it set to stop opening trades on major levels and wait for breakout/retrace could make it very much more efficient.
https://www.mql5.com/en/code/8014
- www.mql5.com
Thanks for the link!
Tested it with dax data from Jan, 1st 2017 till now. I have more data, but the EA does so many trades, that it takes forever...
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
You agree to website policy and terms of use
I found a free EA in the codebase. The default settings were not good, they were strange and confusing actually. I made some adjustments to them and have managed to turn out various profitable back tests on it.
See one of them above.
The interesting thing about this is this was a long only strategy on EURUSD and I am currently running the exact same sim but running sell only. It is not finished yet but looks promising.
I cant add the code because it says the post is too long (even in a seperate post), this is the link to the EA. https://www.mql5.com/en/code/8014
It is using a reverse margingale strategy and the settings I have are as follows;
20 pip grid.
14 pips stop
$130 profit target.
Step 1 - 0.1 lots
Step 2 - 0.1 lots
Step 3 - 0.4 lots
Step 4 0.6 lots
Step 5 0.9 lots
(In the visual sims I have never seen it pass step 5).
The really interesting thing here is as I said before, its producing uptending results over multiple years regardless of the direction it is trading. Even in the tests that did not do so well (over short periods of time) it tended to track about breakeven.
The settings are extremely basic and can be far better, I think.
Things I want to test;
1 - This version can only run one direction. I'd like to see what it looks like simulatiously running both directions (I think this will work).
2 - The strategy is "sticky" in direction. I'd like to see how it does when using a trending finding method. SMAs/EMAs with and perhaps an ADX to make direction less arbitary.
3 - The current settings only allow for a dollar profit target. I'd like to see how a compounding aspect would affect results.
4 - Perhaps a way to optimize grid spacing using ATR.
The EA also needs some common sense safefails and cut offs to be used in the live market I think. I am sure someone will know more about this than me but off the top of my head, max risk cut off and something to stop it placing orders when the spreads are stupid would be beneficial.
Little footnote on timframes, I have ran it on 4 hour for long tests and 15 mins for short visual tests. I do not think this matters much since the trades are based on set pips rather than patterns, so all charts are probably the same.
I have been manualy trading for over 5 years and know a fair bit about that but no almost nothing regarding coding EAs. I am willing to do the testing and generate the ideas on how to improve the strategy if someone can help with the implimenting of the coding for that.
I cant add the code because it says the post is too long (even in a seperate post), this is the link to the EA. https://www.mql5.com/en/code/8014