CyberVest EA review: a backtest with the answers built in

This CyberVest EA review starts with a backtest that wins 85.1 % of its trades. Stop loss and take profit are the
same size, 5 USD each. On that setup a coin flip wins about half the time. CyberVest wins 85 %, every month, for ten
months straight.

Then I moved the dates back one week and it fell apart.

Tested: CyberVest EA v3.30, the Market demo, 2 November 2026.

What does it promise?

$225, a 4.2 rating from 6 reviews, gold on M5 with an "AI signal filter" and a news filter. There's no live signal. So
again the backtest is all you get, and on Pepperstone real ticks it looks like this:

906 trades, profit factor 5.68, 10 out of 10 months green. Even the rough months for gold.

What made me suspicious?

Same author as Quantum iGold, Yassine Mouhssine. Same inputs, and
the same lock: it won't place a single trade before 2023, no matter what data you feed it. I switched off the AI filter,
the news filter and the funded mode all at once. Not one trade changed.

How did this CyberVest EA review catch it?

I copied real gold prices into a custom symbol and kept every price exactly as it was. The only change: every date moved
back 7 days. Same weekdays, same hours.

Run (0.01 lot)	Trades	Win rate	Profit
Real dates	1,463	92.1 %	+6,168
Dates −7 days, all trades	2,399	46.9 %	−585
Dates −7 days, new trades only	1,400	16.6 %	−4,513

Look at the last row. With equal stop and target, 16.6 % winners is way below a coin flip. Those are trades the EA
normally skips, on the real dates, because its backtest knows they lose. The skip list is tied to the calendar, not to
price. Move the calendar and the losing trades come back:

Shifting 14 days gave the same picture.

What happens on a live account?

Live is the future, and the list has no dates from the future. So you get the plain momentum signal without it. Even
inside the test period thats only about 54 % winners and PF around 1.2. A
close rule I ran on 2020 to 2026 won 46 to 52 % and lost money.

One buyer left a 1 star review: "Not actual as testing data." Another saw the lot go up after two losses. That's the
"Recovery" input, x2 per loss up to 20.48 lots. On the trades without the skip list, it took 1,000 USD down by 850 in 3
to 6 weeks.

Why will the next version look perfect too?

Expect new versions to keep coming. Each one can add the weeks since the last build to its list, so the backtest stays
perfect right up to the release date. Whatever happens after that date isn't in the list, and that's where your account
trades.

Verdict

Don't buy it. The 85 % comes from a list of dates, and the strategy underneath doesnt have an edge.

Want to check yourself? Run the demo on a custom symbol with your broker's gold, every date moved back 7 days. If the new
trades win far under half with equal SL and TP, you know.

The two kinds of problem in this series: MQL5 EAs I wouldn't buy.