I understand this limitation. I am trying to simulate with MT4 then when i see it is having a consistent edge. I am planning to migrate to java code with API to Dukascopy.
Here's what I think: 1) Retail bucket shops will not be suitable. And it will be hard for you to get a more direct access to the market without size; 2) The data you're testing on needs to be the data you will be trading. Most likely the data you're testing on now is irrelevant. You basically want to run an forex HFT program. Far from trivial. Get away from MT4 and retail forex brokers and their data if you want to get anywhere.
Testing for the most PRIMITIVE protype. LOL.. The final version for testing will be much different than this one in every aspect. However, it shows some elements that encourage continuing developing this monster. Around 40000 trades and from 10 to 50k.
I agree 100% about the DD. However, there are many features still not been functioning properly to mitigate this DD. I am learning to code now so i could speed the development process.
Hi McGene, I'm doing something the same these days. Yes, there a re a potential in such systems. Let's share info and see what will happen next By the way, does your screen looks like mine
Hi Osama, Welcome to ET. Sharing ideas is always welcomed. That is why we are here in ET. However, my screen does not look the same as yours. Yours is more rich . good luck my buddy. I am not sure yet i have a potential in my system or not. I have done the financial modeling but in reality, to trade such systems need a lot of tweaking. By the way, could you share the backtesting results
Now: PF is 4.82 instead of 1.6 DD is decreased about 50% than before. 85-90 % winning ratios instead of 70%. to be continued... N.B. This EA is not for selling or for signal renting. It is absolutely for in-house research and development purposes only. I do not/will not manage others' money also. Thanks