The course
Optimise a strategy on a stretch of history and of course it fits — you chose the settings that fit it. Walk-forward moves the goalposts: tune on the left block, then score only on the block the optimiser never saw, and step both forward. What comes out is a record assembled entirely from data the strategy had not met.
Fold 1 of 4 — the optimiser only ever sees the left block. Every result that counts comes from the block it had not seen.
No coding background — you learn to review code, not write it. No capital — demo accounts throughout.
Each is specified, generated with Claude, broken, fixed, validated and audited on camera — errors left in.
Module 08 · 9 lessons · 90 min
A XAUUSD trend-pullback system. The full pipeline on familiar ground: handles, buffers, sizing and the Strategy Tester.
Module 10 · 9 lessons · 95 min
Structure-based. Break of structure, change of character, order blocks, fair value gaps and liquidity sweeps — subjective concepts turned into unambiguous rules.
Module 12 · 10 lessons · 105 min
Non-directional. Position maths, basket accounting, margin and compounding risk — built with hard caps, then stress tested to destruction.
Module 13 · 10 lessons · 100 min
Non-directional and hedged. Basket mechanics, the exit ladder, true margin cost, and the four bounds that turn an unbounded system into one you can size.
Grid and recovery systems are taught with their failure modes, not their marketing. You should finish those modules more cautious, not less.
Every project reuses the same risk layer and the same 20-point audit checklist, so you finish with a library — not four disconnected files.
How often you win, and how much you stake. Move either one. The interesting part is that they are not interchangeable — past a certain quality of edge the size stops mattering, and below a certain quality no size helps at all.
At 2:1, that is +0.14R a trade in expectation
Of account equity, on every trade
Positive expectancy, and still a meaningful chance of not surviving to see it pay.
A teaching model, and the simplifications matter: 2,000 simulated runs of 200 trades, a fixed payoff of 2:1, every trade independent of the last, and no spread, slippage or commission. Real systems have none of that — strike rates drift, losses arrive in clusters, and costs are real, all of which push these figures the wrong way rather than the right one. Module 02 runs it on your own measured expectancy instead of a dial. Nothing here is a prediction, a result, or advice.
18 modules · 125 lessons · 20h 15m of video. Lessons run 4–15 minutes, except live build sessions where continuity matters more than brevity.
3 lessons · 20 min
Nine things, not a certificate.
Plus a written risk plan, made before it was needed.
One payment. Lifetime access to the material and every source file.
Self-paced, with an optional live cohort track.
$999
Attendees of the free live webinar receive a 50% enrolment discount.
The free live webinar walks through putting an Expert Advisor together in MQL5, start to finish, with questions answered as they come up.
Join the free webinar