Looking for help to build EA based on ICT strategies
-
Where can i find help to create few EA based on ICT strategies? are there any freelancers out there or anyone else interested in join forces to implement ICT strategies like judas swing, silver bullet, MMXM… I have good requirements details for strategies that i have manually backtested and proven promising results.
-
What's the amount of data of the manual backtest?
-
@emcoin If you make a start, then ask about 1 point here, I will help you create.
-
@bnjmnstsma i have got two strategies with 40 trades each which is from 18 months backtesting. Then i have got two more strategies very promising but requires more data that is why i though about automation.
-
Start simple, just doing 1st condition, back test and adjust. You can then add a shared link, and say what is not working, once working, start on condition 2.
-
I have started to code directly in mql4 using ChatGPT. The first condition i am working on is OnTick identifying the high and low price between 00:00 to 03:59. I think to implement it is required to implement like a state machine to: 1) identify the level of interest; 2) liquidity grab on those levels; 3) a break of structure and back to the level if the BOS happened above/below level; 4) set entry in the most extreme FVG. Not easy since i dont have much coding experience. Do you guys can code directly in C++ mql4 directly?
-
The above is one of the profitable strategies a judas swing with filters identified to increase the winrate.
-
Only FX on here mate.
-
@emcoin I recently started learning about automation with new capital fx. I'd really recommend starting follow their course and community, because this is very complicated.
Especially ICT/SMC concepts.
What are the rules in the strategy? Maybe I can help you also. Btw, 40 trades on data is way way too small. But as you said, automation can make this dataset bigger in a quite little time
-
@bnjmnstsma for starters the best two strategies I have got working after 6 months and I got funded it is a filtered version of the Judas Swing for NAS100. The key is to have the data so i can continue build on more filters. For example removing news days, removing Fridays and limiting the execution time window to certain timeframes (9:40 to 11:15 NY time) i was able to increase win-rate from 40% to 60%. Where is the catch? there are only 2-3 trades a month and sometimes either i dont execute well or i am not in front of the screen. Also, I find journalling and keep it clean tedious and it is preventing me to further iterate in the filters i am working on like adding bias to the trade to see if win-rate can be increased.
This course looks great new capital fx but $150 a month. I am thinking to start using ChatGPT code + try out a freelancer in mql4 to build a version of basic judas swing...
-
hi @jstap @bnjmnstsma I have spent the afternoon writing the full spec for the EA. I prefer to be written in C++ than FXDreema since the code maintenance will be easier as the code grows (thanks to ChatGPT changes will be easier to make). If any of you interested in help, please share your email and i can forward you the spec. let me know,
-
C++ means a lot more work, even with chatGPT, infact this will get some code wrong, then you will have to correct. If done from FX, you can be helped with blocks.
