This Quantum iGold review comes down to one test. On real gold ticks the EA won 96.8 % of 1,141 trades, profit factor
36.5. I raised every gold price by 0.02 USD, two points, and ran it again. Profit factor 0.99.

A real strategy doesn't care about two points. This one does.

Tested: Quantum iGold MT5 v4.2, the Market demo, 2 November 2026.

What does it promise?

$125, around a hundred reviews at 4.6 stars, "LSTM + Transformer" neural network on gold. No live signal, buyers asked and
got "soon". So the backtest is all you have to go on, and it looks like this:

Quantum iGold backtest, Pepperstone real ticks Jan 2025 to Oct 2026, 0.01 lot: almost a straight line

What made me suspicious?

The losing trades were tiny, about 30 cents typical. Like the EA knew which trades were going bad. And it never trades
before 3 January 2023, even when I gave it data from 2020.

How did I catch it?

I copied real gold prices into a custom symbol and nudged them a little. The EA cant tell the difference between that and
real gold.

Data Win rate Profit factor
Real prices 94 % 60
Every price +0.02 USD 82 % 0.99
Every date moved back 7 days 35 % 0.86

The EA carries past prices inside it. When the chart matches, it trades from the answers. Shift the prices or dates and
there's nothing to match, so you see what it really does:

Same EA with every gold price raised by 0.02 USD, 2025 to Sep 2026: balance drifts back to break-even

What happens on a live account?

Live is always the future, and there's no stored data for the future. So you get the second chart. It takes +1 USD wins
and about −4.5 USD losses, which needs 82 % winners to break even. It wins 82 %. After spread and commission, thats a
slow loss.

The "Loss Recovery" option doubles the lot after every loss, up to 20.48 lots. That's a martingale, whatever the page
says. And switching the AI settings on or off didn't change a single trade.

Why will the next version look perfect too?

Because the author keeps updating it. Quantum iGold had 14 versions in four months. Each new build carries price data up
to its build date, so it fills in the weeks the old version was already trading live, and the backtest is a straight line
again. The days after the build are never in it. Those are the days your account trades.

Verdict

Don't buy it. The backtest is answers it already had, and the strategy underneath doesn't make money.

Want to check yourself? Make a custom symbol from your broker's gold with every price +0.02 and run the demo on both. If
the result falls apart, you know.

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