EAs in Market - page 6
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
MQL Improvement Proposal 03 :
MQL Improvement Proposal 03 :
Forum on trading, automated trading systems and testing trading strategies
...
Nillscash, 2023.07.26 00:54
Hi Friends,
I'm happy to know that I'm not alone worrying about the future of our Platform --MQL5.
Yesterday I sent, SOS message to Meta Quotes on this issue. One thing should be clear. Offering Experts, that one knows, only shows
good performance on DEMO account testing, but not good enough for trading on REAL account is a CRIME.
This have to be dealt with seriously. Then we are here to trade forex for one common reason, which is, to make few Dollars for our Families.
I suggested few new Rules for Sellers on MQL5.com platform in other to protect members, buyers of products offered on MQL5 using their
hard earned cash because they trust and love MQL5.
My suggestion was- that Sellers of Experts should present at least 4 to 6 months proof, showing the performance on Real account, and also the names of 2/3 Brokers they have traded with,
before they are permitted to sell their products on the platform.
* Also 3 months CASH-BACK guarantee for all clients who buy these products, when within 3 months the Expert fails to trade as promised.
I hope Meta Quote will take this matter serious and help make the platform, the best place to trade FOREX.
Wishing you all a good day and happy trading in the future.
Best Regards,
Nillscash
It was the post from one thread (I deleted the thread to keep everything in one place.
So, it is about the following:
good performance on DEMO account testing, but not good enough for trading on REAL account is a CRIME.
-----------------------
Hi Nillscash,
who told you that strategy testing results should be same with real trading for the EAs?
I did not find anything related to it in the Market Rules.
or you want for strategy tester report should be same as real trading on every EA from the Market, and because of that you are trolling here naming the sellers as a "CRIME"? ... without any proofs for every seller ...
I have one name for that - you are the trolling.
-----------------------
My reply on your post (just to collect everything in one place):
Forum on trading, automated trading systems and testing trading strategies
...
Sergey Golubev, 2023.07.26 06:57
Your proposal is not balanced one.Why?
1. Just imagine that some of the seller is not crime (you told that many of them are crime) ... so what will you do? nothing?
It is trolling in this case.
There were users who fought against the scam calling/naming someone as a scam ... but once they did mistake (they named the person as a scam but in reality - he is not) so they became a scam and they were removed from the forum because of that (because of trolling).
Do not you affraid to make a same mistake in that?
2. Responsibility of the customers who are buying the product from the Market.
Some customers turn their personal responsibility/risk to someone else (to MetaQuotes, to some seller, etc).
And it is completely illegal action (in general).
3. As to responsibility of the sellers - yes I agree. But this proposal should be made from the persons who are taking all those items (1 and 2) into consideration to be a balanced proposal.
Because in case the proposal is not balanced one so it is trolling.
-----------------------
The is the discussion thread about it - Scam EAs in Market - so we (on the forum) do not need many threads (just keep one thread related to the problem).
-----------------------
For information.
More (from Rules of Using the Market Service):
Hello,
In the Market appear recently Expert Advisors that look as Holy Grail by testing them in Strategy Tester. On real market they do not perform as good as on history. The issue is that they are based on RNN (neural network) models that fit to the history by teaching them on history data, but they cannot predict new data. So on the history they have learned to predict based on the model. This model is stored in the EA file entirely. To protect the clients MetaQuotes needs to detect these EAs and remove them from the market. Otherwise many clients will be scammed trusting the results in Strategy Tester but seeing bad results on live trading market.
Please care about this issue!
Regards,
Dmitry
Have you tried using free robots? :)
Hello,
In the Market appear recently Expert Advisors that look as Holy Grail by testing them in Strategy Tester. On real market they do not perform as good as on history. The issue is that they are based on RNN (neural network) models that fit to the history by teaching them on history data, but they cannot predict new data. So on the history they have learned to predict based on the model. This model is stored in the EA file entirely. To protect the clients MetaQuotes needs to detect these EAs and remove them from the market. Otherwise many clients will be scammed trusting the results in Strategy Tester but seeing bad results on live trading market.
Please care about this issue!
Regards,
Dmitry
Hi everyone,
I've seen quite a few messages lately accusing certain EAs of being scams — especially those that use artificial intelligence or machine learning. I get why it can be frustrating, but I think there’s some confusion that needs to be cleared up.
An Expert Advisor (EA) is just a tool. It executes trades based on rules. These rules can be simple (like “if RSI < 30, then buy”) or come from a trained algorithm. But at the end of the day, the EA does what it’s told: it follows logic, that’s all.
The real issue isn’t the tool — it’s what people expect from it. Many believe that buying an EA, renting a VPS, and funding an account is all it takes to start making money. But no — trading doesn’t work like that.
Backtests don’t guarantee anything
We often see EAs with insane backtests: profit factor of 10, 20, even 100+. That’s not impressive — that’s suspicious. When a profit factor is too high, it usually means the strategy has been over-optimized for past data. It’s been “trained” to win on history. But history tells you nothing about what’s coming next.
A great reminder from someone in the industry:
So if you’re evaluating an EA, don’t just look at the pretty backtest curve. Dig deeper: look at the drawdown, MAE/MFE, number of trades, trade duration, consistency across years or timeframes… that’s where the truth lies.
MetaQuotes does check EAs, but…
Before an EA is published, MetaQuotes performs a technical validation: they make sure it doesn’t crash, that it follows trading rules, that it works with different brokers, and so on. But that doesn’t mean the EA is profitable — just that the code is clean.
Let’s be clear: no backtest can predict the future. I’ve seen EAs perform well for months, then crash overnight because market conditions changed. That’s part of the game.
Start with risk management
Most people skip this part. They want to talk about gains, not losses. But if you don’t put risk at the center of your strategy, profits won’t last. It’s that simple.
In summary
An EA isn’t magic. It’s a tool. In the right hands, it can be very useful. In the wrong hands, it can be dangerous. Don’t get hypnotized by dream backtests. Look for logic, consistency, and common sense.
And most importantly: educate yourself. Understand what a profit factor means. Learn how to read a report. Learn how to evaluate a strategy. That’s how you stop being just a customer… and start becoming a real trader.
An Expert Advisor (EA) is just a tool.
I see the issue more as a whole spectrum of quality. On the extremely poor end of the spectrum, there are the EA's (and Signals) that run wild profits for a bit only to ultimately blow up. And sometimes these things are removed from publication only to reappear under a different name. Obviously, these are total Scammy McScammers. In the middle of the spectrum, there are (alleged) scalping, HFT, arbitrage, and machine learning EA's. The reason that so many traders go for those EA's is a lack of trading capital, which most traders unfortunately have. Well, you can't really get into MFT swing trading with a 200 USD account because you can't afford the stops (nor the swaps in the case of forex/CFD's). Making things worse, this is where the individual account nuances come into play--differing price feeds, broker-dealers, execution processes, trading machines, connection latencies, swaps, commissions, etc. The worst kept secret in trading is that you have to move up to a higher timeframe/chart structure in order to be able to absorb market noise. This is simply not an option with a small account. For example, look at a 100 pip Renko chart of the GBPJPY. Unless you're trading with an absolute bucket shop, the aforementioned differences will be less impactful. Unsurprisingly, the MFT and positional EA's are on the excellent end of the spectrum... but only for those who can afford to run them.
By the way, EA source code is not visible in the Market. Who's to say whether a "machine learning" EA is actually machine learning at all? What's next... Rampant honesty on Wall Street?
Je vois le problème plutôt comme une question de qualité. À l'extrémité la plus médiocre du spectre, on trouve les EA (et les signaux) qui génèrent des profits exorbitants pendant un certain temps, avant de finalement exploser. Et parfois, ces produits sont retirés des publications pour réapparaître sous un nom différent. Il s'agit évidemment de véritables arnaqueurs. Au milieu du spectre, on trouve les EA (présumés) de scalping, de HFT, d'arbitrage et d'apprentissage automatique. Si tant de traders optent pour ces EA, c'est par manque de capital de trading, ce dont la plupart des traders disposent malheureusement. En effet, il est impossible de se lancer dans le swing trading MFT avec un compte de 200 USD, car on ne peut pas se permettre les stops (ni les swaps dans le cas du Forex/CFD). Pire encore, c'est là qu'entrent en jeu les nuances propres à chaque compte : flux de prix, courtiers, processus d'exécution, machines de trading, latences de connexion, swaps, commissions, etc. Le secret le moins bien gardé du trading est qu'il faut adopter une structure graphique/période de temps plus élevée pour absorber les bruits de marché. Ce n'est tout simplement pas envisageable avec un petit compte. Prenons par exemple un graphique Renko à 100 pips de la paire GBP/JPY. À moins de trader avec une plateforme de trading de premier ordre, les différences mentionnées ci-dessus auront moins d'impact. Sans surprise, les MFT et les EA positionnels sont d'excellents outils… mais seulement pour ceux qui peuvent se les permettre.
Au fait, le code source d'EA n'est pas visible sur le Market. Qui peut dire si un EA « machine learning » est réellement de l'apprentissage automatique ? Et après ? Une honnêteté débridée à Wall Street ?
I get the concerns people are raising here — I share some of them — but I think the real issue is more nuanced than just “EAs are scams.” The problem isn’t the tool itself or the tech behind it (AI or not), it’s how it’s used… and more often, how it’s abused to sell hype.
1. EAs built to be deliberately opaque
We’re seeing more and more EAs on the Market with zero transparency. No logic you can follow, no adjustable settings. Everything is hardcoded — indicators, conditions, values. The user has no control, no understanding, nothing to tweak. These EAs are basically fixed sets disguised as unique strategies.
2. Signals used as marketing bait
A lot of people don’t realize how easy it is to fake a “winning” EA. Here’s how it’s done:
That signal looks impressive, but it’s just statistical luck. Then it gets used as social proof to push sales.
3. Gaming the Market algorithm
Another common move: as soon as the EA is published, you’ll see dozens of demo downloads, quick purchases, and glowing reviews — all within hours.
Let’s be honest: that’s not organic. It’s coordinated.
This pushes the EA into the MQL5 Top 10, giving it a ton of exposure. Real users see the popularity, assume it must be legit, and buy in. It’s artificial hype — same playbook we’ve seen in NFTs, crypto, and other speculative bubbles.
4. The illusion of support
After launch, the seller acts like they’re engaged and responsive. They answer questions, post “updates” — which are usually just minor tweaks to indicators or internal values — and keep users hopeful for a while.
Then, when things go south: radio silence.
Two months later, the same EA reappears under a new name. Same game.
5. A shady but profitable business model
Let’s say an EA sells for $1,500 and gets 200 sales. That’s $300,000 — in a month. Way more than you’d make actually trading the strategy with discipline.
What’s worse: the signal they use often runs on a $100 or $1,000 account, while the product is pitched as something that can “transform your financial future.”
Instead of showing how to grow a $100K account, they show how to survive on $100. It’s more accessible, more marketable — but way less honest.
Bottom line
Yes, there are solid EAs out there — with real strategy, smart risk management, and transparency.
But let’s not ignore the “Market mafia”: a small group of sellers who know exactly how to game the system — using signals, fake hype, reviews, and multiple accounts — to manufacture the social proof they need to sell anything.
They’re not developers. They’re marketers.
And as long as the Market stays wide open and loosely moderated, this stuff will keep happening.
I’m not writing this to discourage people. Quite the opposite — I want to help users stay sharp and ask the right questions.
A nice backtest means nothing. A $100 signal proves nothing. And a $1,500 EA isn’t necessarily better than one that costs $100.
Learn. Test. Analyze.
Very nice indeed, thanks. This is basics for market economy - when strategy doesnt work well, it becomes product to sell instead.
Critical thinking remains the only real protection for buyers.