Sign up with our partner broker and get free XAU/USD signals. JOIN US
LIVE MT5 BUILD 5830 LAST REV 2026-05-23
SESSION -- UTC --:--
SESSIONS
SYDNEY22:00–07:00 UTC
TOKYO00:00–09:00 UTC
LONDON07:00–16:00 UTC
NEW YORK12:00–21:00 UTC

EA scam red flags: how to spot a fake before you pay

The MQL5 Market has thousands of EAs. Maybe 5% are honest products with real (if modest) edges. The rest range from naive failures to outright fraud. The patterns are repetitive enough that you can filter most of the junk in under five minutes per listing.

PUBLISHED 2026-05-24 READING TIME 12 MIN MT5 BUILD 5830 CATEGORY TOOLS
Bottom line: If an EA seller cannot show a live MyFXBook-verified or FXBlue-verified track record covering at least 12 months, the equity curve does not exist. Backtest images, screenshots, and "client testimonials" prove nothing. Walk away.

1. The grand promise red flags

Start with the marketing language. Specific claims are tells.

ClaimWhat it really means
"99% win rate"Almost certainly a martingale or grid system that never closes losing trades. Wins are tiny; losses are catastrophic when they finally come.
"Risk-free strategy"Either a scam or a misunderstanding of what risk-free means. Every position-taking strategy has risk.
"Guaranteed monthly returns"Securities fraud in most jurisdictions. Nobody can guarantee market returns.
"AI-powered" or "uses machine learning"Sometimes true, usually meaningless marketing. Most "AI" EAs are linear regression with extra colour.
"Used by professional banks"Almost never true. Banks build proprietary execution systems, not MT5 EAs.
"Holy grail" or "secret formula"If it existed, the seller would deploy it privately rather than sell it for $200 on a marketplace.
"Limited copies available, price going up soon"Standard urgency manipulation. There is no scarcity of digital files.

2. The track record question

This is the single most important filter. Honest EA sellers publish their EA running on a verified third-party tracker. The two standard ones:

  • MyFXBook. Connects directly to a broker account read-only and publishes live equity curve, drawdown, and trade history. Hard to fake.
  • FXBlue. Similar concept. Slightly less common but equally credible.

What honest looks like:

  • Account is verified (a checkmark and the broker name are shown).
  • The track record covers at least 12 months. Three months is short enough to lie with a martingale before blowing up.
  • The equity curve has visible drawdowns. Curves that go straight up have been faked or used capital top-ups during losing periods to mask drawdown percentage.
  • Trade history is downloadable. You can inspect individual trades.
  • The account uses normal leverage (1:100 or less, ideally lower for the strategy claim).

What dishonest looks like:

  • Screenshots of MT5 terminal showing big profits. Trivial to edit; meaningless.
  • "Backtest results" with a Sharpe ratio of 10 and zero drawdown.
  • MyFXBook with the "broker verified" badge but the connection is in custom or manual mode (the user types trades in, so it can be fictional).
  • Live account that started two weeks ago. Selling on the first short streak of wins.
  • Multiple accounts shown, all 100% profitable on the website but the seller cannot tell you which one their EA actually runs on.

3. The martingale and grid signature

Most "no-loss" or "guaranteed profit" EAs use one of two techniques: martingale (doubling lot size after losses) or grid (placing multiple positions at price intervals). Both work for a while; both eventually destroy accounts.

Tells in the trade history:

  • Many trades closed at small profit, occasional big drawdown.
  • Position size grows during losing sequences.
  • Multiple open positions at different prices in the same direction.
  • The EA never closes a losing trade in the red, only by net average crossing into profit.
  • Equity curve grinds upward then has a vertical drop. The vertical drop is when the martingale finally ran out of margin.

Martingale EAs can look fantastic for 6 to 18 months before a single bad week wipes everything.

4. The vendor evasiveness test

Email the seller before purchase. Ask:

  1. What is the maximum historical drawdown on the live account?
  2. How does the EA behave during high-impact news (NFP, FOMC, CPI)?
  3. What broker, account type, and leverage was the live track record run on?
  4. Does the EA use martingale, grid, averaging, or hedging?
  5. Can you provide a video of the EA opening and closing trades in real time?

Honest sellers give direct answers. Scammers respond with marketing copy, deflect, or stop replying.

5. Technical red flags in the listing

Even before contacting the seller, the MQL5 Market listing itself contains tells:

TellWhat it suggests
Seller has 47 EAs listed, all marketed as different "systems"Asset flipper, churning low-effort products. None are genuinely tested.
EA name contains "Pro", "X", "Ultimate", "Elite", or "AI"Marketing-driven; no signal about content. Not a hard disqualifier, but raises the bar.
Listing was created 2 weeks ago, no reviewsYou are the beta tester. Buy at your own risk.
Reviews are all 5-star with identical phrasing or one-word praiseSuspicious. Real reviews vary in tone, mention specific outcomes, and include criticism.
Backtest screenshots in the listing show 99% win rate or zero losing weeksCurve-fit or martingale. See section 3.
"Risk-disclaimer" in tiny text says "results not guaranteed, do not represent live performance"The seller knows the displayed results are misleading. The disclaimer is legal cover.
Recommended deposit is $10,000+ for "stable performance"The strategy probably needs the buffer to survive its drawdowns. Smaller accounts blow up.

6. The signal seller variant

The same playbook exists for trade signal services. The mechanic is different (the seller sends you trades to copy rather than an automated EA), but the lies are identical:

  • Fake screenshots of profits.
  • Cherry-picked winners shown, losers hidden.
  • "Lifetime access" pricing implies trust the buyer cannot verify.
  • VIP groups behind paywalls so claims cannot be independently checked.
  • "Free signals" used to recruit subscribers to paid tiers; the free signals are deliberately good-looking, the paid ones are average.

The same verification rule applies: ask for a MyFXBook or FXBlue track record. Refusal to provide one is the answer.

7. The "prop firm" angle

A newer variant: EAs marketed specifically to pass prop firm challenges (FTMO, Topstep FX, etc.). Many of these are tuned to take large risks during the evaluation phase, pass the targets in a few days, then collapse during the funded phase.

The prop firm itself collects the evaluation fee, the EA seller collects the EA fee, and the trader takes the loss. Both businesses are economically incentivised to sell evaluations and EAs respectively without caring about funded-phase performance.

8. The free EA Trojan horse

Some scams are not even sold. Free EAs uploaded to MQL5 Code Base or random forums sometimes contain:

  • Credential exfiltration. The EA logs your MT5 account number, password, and broker info to a remote server.
  • Forced broker referral. The EA refuses to run unless you create an account through a specific affiliate link, generating commission for the developer.
  • Hidden backdoor for the developer. The EA accepts external commands that can place trades against your wishes.

If the EA is provided only as compiled .ex5, you cannot audit it. If you must use a third-party EA, prefer source-available .mq5 from established authors.

9. Refund policy

MQL5 Market lets buyers return purchases under certain conditions. The official policy allows refunds within a limited window for products that do not function. It does not allow refunds because the EA lost money. This is reasonable from the marketplace's perspective and an important caveat from the buyer's.

Subscription-based signal services and EA rentals usually do not refund. Once you have paid for a month, that month is gone.

10. The honest path

Rather than buying EAs, consider:

  • Write your own EA implementing a strategy you understand and have tested manually first. Even if the result is unprofitable, you have learned something.
  • Copy trade verified track records on Darwinex, ZuluTrade, or similar copy-trading platforms where strategy providers cannot fake performance. You still need to vet providers carefully.
  • Pay for education rather than automation. A solid understanding of risk management, market microstructure, and a tested manual approach is worth more than any EA.

FAQ

Is the MQL5 Market itself a scam?

No. MetaQuotes runs a legitimate marketplace with review and refund systems. The scams are the sellers; the marketplace cannot vet whether a strategy is profitable.

Are any paid EAs worth buying?

A small number, yes. The qualifying criteria: verified live track record over 18+ months, transparent strategy description, modest claimed returns (10-30% annual), reasonable drawdown (under 30%), responsive seller who answers technical questions. Most EAs fail at least three of these.

What is a realistic annual return for a good EA?

Honest professional systems target 15 to 40% per year with controlled drawdown. Anything claiming 100%+ monthly is either martingale (will blow up) or fiction.

Are propietary firm EA challenges worth the EA fee?

Generally no for retail buyers. The math of evaluation pass rates, EA fees, and prop fees rarely works out positively. Most buyers churn through multiple challenges paying fees with no funded account at the end.

What about signal services that say "verified by third party"?

Check who the third party is. MyFXBook and FXBlue are real. Random websites that look official but are run by the same seller are not third parties. If the verification badge is not clickable to an external account page, it does not count.