Most Gold EAs you see on the Market lead with one number: a huge profit factor or a "500% in 3 months" equity screenshot. I want to explain why I do the opposite — and why I think it actually sells better in the long run.
The number nobody wants to show
Every backtest has two sides: what it made, and what it risked to make it. The second half is the drawdown — the deepest the account fell before it recovered. It is the single most important number for deciding whether you can actually live with a system, and it is the number most sellers quietly leave out of the first screenshot.
I publish it on every listing. When one of my modules shows a 40% equity drawdown in the tester, I put that 40% right next to the profit — same size, same screenshot. Not because it looks good, but because a strategy you don't understand the risk of is a strategy you will panic out of at the worst possible moment.
Real ticks or it didn't happen
There is a second place backtests lie: the modelling quality. MetaTrader 5 can run a test on "Open prices only" or "Mathematical calculations" in seconds — and those modes routinely produce numbers that evaporate the moment you go live, because they never simulated the spread, the wicks, or the exact fill sequence inside each bar.
Everything I publish is tested on 100% real ticks. It is slower (minutes, not seconds, for a multi-year run) and the numbers come out lower and uglier than the shortcut modes. That is the point. A backtest is only useful if it was hard on the strategy.
Why four modules instead of one
Gold doesn't trade one way. It grinds sideways for weeks, then breaks out, then whips around the news. A single logic that's tuned for one of those regimes gets destroyed by the others.
So I built my Gold system, Sentinel Gold, as four independent modules, each solving a different piece of the market: a patient, progressive core for the quiet phases; an aggressive engine that only wakes up in a clear trend regime; a fast momentum sniper that takes single shots with a hard stop loss; and a slow M15 swing hunter that trades rarely and holds with a hard stop.
Each module is its own Expert Advisor and can run completely on its own. That matters for a reason most module-systems ignore: if you buy the momentum module, you should get the momentum module — not the whole system hidden behind a switch. I compile each product so it physically contains only its own logic.
What pre-calibrated really means
I lock the parameters. The only input a buyer sees is an on/off switch. This is unpopular with the "let me optimise it myself" crowd, and I understand why — but optimisation on the buyer's side is how good systems get curve-fit into bad ones. The calibration is the product. If I did my job in the tester, there is nothing left for you to tune.
The honest part
None of this makes a system a money printer. Past results — simulated or live — never guarantee the future. Trading Gold carries real risk of loss, drawdowns happen, and quiet or losing weeks are normal by design in some of these modules. Anyone who tells you otherwise is selling you the screenshot, not the strategy.
What I can promise is the process: one market (XAUUSD), real-tick testing, the drawdown shown next to the profit, locked parameters, and a person who answers every message. If that's the kind of thing you're looking for, the Sentinel Gold family — the four modules and the full combined system — is on my MQL5 profile. Always start on a demo account.
— Servet Coban
