All this talk about Expert Advisors (EA) has got me doing some serious thinking since last week. I have tried a few free EAs in the past but their performances were underwhelming and were not even close to the results I produce manually.
I believe in doing things elegantly. Elegantly in this instance meaning the most optimal/efficient in terms of costs, effort, time, simplicity etc. Take for example earning money…I have concluded that trading forex is one of the most elegant ways of earning money, which is why I have persevered with it for so long. Now if I could somehow cause an EA to be created which could duplicate the way I trade and thus produce similar results wouldnt that be great (pronounced elegant)?
The thing is the way I trade (see the attached screenshot) I honestly dont know how that could be captured to build an EA. I know for sure it cant be backtested. Just the mere thought of attempting to put my strategy down on paper is giving me a headache.
But just imagine though if I could let such an EA go on a $2k account and say …GO!!!
Maybe one of these days I will just take a week examining it and see what I can come up with. What do you guys think?
I think given enough time, anything can be done. Problem is, unless you’re a programmer or know someone who will take on the challenge as a favor, it could get expensive. There are some good companies out there who do these things, might be worth reaching out with your idea just to see what their thoughts are and IF it’s feasible. Definitely would be a good investment if you could turn it into an EA though.
I appreciate the suggestions.
I with my slow self, however, have been doing some preliminary thinking, which means I may have hurt myself, lol.
I will use my elegant concept and get Chatgpt and another of the Artificial Intelligence guys to build me an EA based on what I can tell them about my strategy. So the issue of cost is not a factor for now. I will test them and see how it goes. Like I said…should be interesting.
You have to start somewhere. After that, it’s a lot of testing, tweaking, and back and forth.
I fully support you asking ChatGPT for a solution. This link may help you start off in an elegant way. I have used this method for 37 use cases to date, have never spent more than 2 hours on any one of them, and am keen to see the breadth of what can be achieved by participants. The subject matter is referred to as prompt engineering.
- YouTubeI watched the first 2 minutes of it. Interesting. I will have to get back to it later today when I have some free time.
ChatGPT is very annoying at this stage when working on MQL projects (even simple, small ones). That’s what I’ve discovered so far.
A very good alternative (in terms of time and money spent) for transforming your ideas into code that actually works is to use an Online Expert Advisor Builder. The subscription costs a few bucks per month and once your understand how it works, you will be able to create almost any automation you want. While it may take some time to become familiar with the interface, it is significantly less than the time required to learn the MQL programming language. Take this into consideration, at least until you’ll afford to pay thousands of dollars to a real MQL programmer.
Yes I came across this but decided to hold off on it until I had some experience under my belt with the AI ones first. I think I saw where that one was giving a 15 day free trial so I wouldnt want t waste any of those free days, lol.
Hi 
Judging by the screenshot, you might be looking for a Scalper EA.
But the first thing I ask someone who wants an EA created (mostly experienced traders): What all did you consider and how did you get to the point where you open a trade?
I sort out 99% of the requests because there is too much “human interaction”.
Yes, I describe my strategy as scalping so the EA would naturally be a scalping EA.
I have a question though. What do I do in a scenario where I pay someone to create such an EA for me, (which they do and it works perfectly as intended) but then I find out they are using it themselves or are selling it to other people?
That right there is the main reason why I hadnt given much thought to the idea of creating an EA for my strategy before now.
@THE_GOATE I do not wish in any way to discourage you in this very worthy pursuit, I have been through this movie myself. I had an EA developed by the very reputable developers 4XDEV, I will not go into the full story here, as I’ve posted it on another thread, suffice it to say - it did not work, probably for the reasons point by @Ravenash which he indeed pointed out to me. I am not very techy but it seems, also, that EAs work only on ‘candle close’.
Al Brooks was asked once why he didn’t use an EA and he said that to get a bot of the calibre used by institutions would cost in the order of half a million dollars with quants being involved.
Things are moving fast though with AI, so I wish you all the best in your ventures.
You should have a non disclosure agreement to protect your ideas, then again a bit of tweaking would perhaps nullify any agreement.
Good luck.
@Johnny1974 I appreciate the input.
I dont know if it makes a difference but so far my instructions that will be given to construct the EA stand at 13.
I worked with 4xDEV as well to create an EA. They were good and it works as intended. However, it’s a pretty simple strategy.
Any reputable company will have you sign a non-disclosure agreement before the work starts. They are trying to make a name for themselves and ripping off their customers certainly isn’t the way to do it.
As for my EA, as I said it works great but I stopped using it because I was not comfortable with the large drawdowns the strategy was capable of producing.
Yes, that’s interesting. My pro trader friend Hal spent his time developing strategies suitable for EAs which he uses on his Hedge fund. However, he has an EA for up to five differing FX market conditions.
IMO, I suggest that’s why a single EA would only work well for the tested strategy in the market condition that suits it. How Hal decides which one to use is determined each day, at Wall St by the state or the art servers as used also by Goldman Sachs. Yes, he is a millionaire - great down to earth guy, though.
Which most likely confirms that the big boys have the edge over any retail trader’s EA, however good their strategy.
@THE_GOATE It’s like with all programming work: If you are not the programmer’s employer, the work belongs to the programmer and not to you. Even if you had the idea. Basically, you buy the permission to use the EA. Of course you can change this with an agreement, but then the EA becomes massively more expensive! But as @Johnny1974 said: You only have to change the behavior or the code a little bit and it is a different EA and not subject to the agreement anymore. So if you want to be really sure, you have to program it yourself ;). But most people don’t care about the idea, because they just want to have an EA to trade.
@Johnny1974 Correct! The EA waits for a signal and that is mostly candle open/close and not in the middle of a candle.
“mostly candle open/close and not in the middle of a candle”
I’m not sure that this is correct as you can use Bid & Ask when it meets a particular level as a trigger…
(i.e Ask < ima 20 - or Ask at 23.8% of the previous candle - that kind of thing)
Coding is time consuming !
Believe me, it is correct. I’m doing this for a long time. What do you mean here “(i.e Ask < ima 20 - or Ask at 23.8% of the previous candle - that kind of thing)” is the rule of the EA. When the rule is positiv, the EA waits for the next tick from the market to set/open the trade. And the only ticks you get from the market are Open/Close.
And coding is not time consuming.But backtesting is!