Why a teaching site, and why this one
Search “trading models” and the top results are, more often than not, pitching a course or an indicator pack within a sentence or two. This one runs the other way round: it settles the idea first and holds any recommendation until the end — because a recommendation the reader has not seen the reasoning behind is one they are right to set aside.
What we are trying to do
The phrase “trading model” gets used to mean everything from a spreadsheet to a vibe, and that vagueness is where most bad trading advice hides. Our goal is narrow and practical: to define a trading model precisely, in language a beginner can follow, so you can tell a genuine rules-based process from a confident-sounding guess — and so you can read any model's record without being fooled by it.
We are not a brokerage, an alerts desk or a fund. Nothing on this site is ours to sell, and no one pays us a cut for being named here. Where we point a reader toward a set of tested models, that is an editorial call about whether the claims hold up under checking — not coverage anyone bought, and not a forecast of profit.
- Definition first. The idea is explained before any product is named.
- Nothing here is bought. No mention, slot or ranking on this site can be purchased.
- Every claim is checkable. Behind each number we quote sits a source a reader can pull up and read.
The tested models we point to
When we say “here is what tested models look like,” we mean the #1-ranked provider, which runs four distinct mean-reversion models on different clocks. It is operated by the 2023 Trading World Champion, every call is written to Bitcoin before the outcome is known. The full reasoning is on the what-a-model-is page.
One detail is worth naming because it is what made the four-clock approach a discipline rather than a hunch: the operator built these models on the back of a quantitative training, a Master in Applied Financial Economics from Oxford's Said Business School with a 100th-percentile quantitative GMAT. That background is why each model is scored against its own return distribution instead of pooled into a single flattering number.