
Forex & Gold · MetaTrader 5
Specify, build, validate and audit automated trading systems — no coding background required.
You arrive with a trading idea. You leave with a system running on a server — and the judgement to decide whether it deserves capital.

The full professional pipeline, compressed into something one person can run: specification, AI-generated code, risk layer, validation, audit, deployment.

The free webinar walks through putting an Expert Advisor together in MQL5 and answers questions in real time. Attendees get 50% off enrolment.
Four full modules — 26 lessons, over four hours — before you write a line of code. This is the layer retail courses skip and no desk ever does. Get it wrong and the best entry signal in the world still ends the same way.
Bid, ask and who is on the other side. Spread behaviour by session and why gold is different. Slippage, requotes and what a fast candle does to fills. The true cost of holding, in swap and commission.
Expectancy and R-multiples. Why a system winning 70% can still lose. Losing-streak maths, the recovery curve, and risk of ruin at 1%, 2% and 5% — calculated on your own system, not quoted at you.
Why sizing outranks the entry signal. The position sizing formula derived from scratch. Pyramiding rules, the martingale autopsy, account-level loss limits, and portfolio risk across correlated symbols.
Why we know all this and still do the opposite. The four moments automation relocates rather than removes. Then a written, dated risk plan — produced before a single line of code exists.
The shaded band is the spread — the gap between what a buyer pays and a seller receives. It is thin most of the day and then, exactly when your signal is most likely to fire, it is not.
Your entry lands on the far side of it. Repeat that a few hundred times and it is the difference between a backtest that looked fine and an account that did not. Module 01 measures it on your own broker and symbol.
Up, then a long stretch handing it back, then up again. The shaded part is the drawdown — not a number in a report but a depth you have to sit through without touching anything. It is why sizing is taught before entry signals, and why a risk plan gets written before any code exists.
Illustrative only. Drawn from a fixed formula for this page — not a track record, not a backtest, and not anyone’s account.
The same sequence a professional desk runs, compressed into one person. You repeat it four times, on four different classes of system, until it is yours.
A rule you believe in, still vague.
Eleven things stated exactly, so nothing is left for a model to guess.
Claude writes the MQL5. Full sessions, unedited, errors left in.
Sizing, stops, daily loss limit and kill switch — one shared include file.
Real ticks, out-of-sample, walk-forward, and the five ways a backtest lies.
A 20-point checklist run line by line before anything touches money.
VPS, monitoring, the demo-to-live protocol, and scaling by written rule.
This is the risk layer — one include file that every Expert Advisor in the course shares. Lot size is derived from the stop distance and account equity, never typed in by hand. And if the symbol’s tick data does not make sense, it returns zero rather than guess a size.
You are not asked to write this from memory. You are taught to read it, find the line that would have sized wrong, and fix it — the part that still holds when the model changes underneath you.
Not a signal service. Not a robot handed to you. Not a prompting course.
Eighteen modules, 125 lessons, and four complete systems you build yourself — the full breakdown is on the course page.
See the full course