The mean reversion strategy, explained: a bet that a stretched price comes home, and how the method actually works.
Type “mean reversion strategy” into a search box and most of what comes back is trying to sell you a subscription before a single line of the idea is explained. This one does the reverse: it explains the statistics first — what the mean is, why a gap from it tends to close, and how the holding clock changes everything — and only then points to the built, tested, checkable version of the method so you can see the theory keeping a real record.
Vector Ridge runs the worked example: one model is $20 a month, all four are $50, and the trial costs nothing to start. No money-back guarantee — the on-chain record is the guarantee.
One idea, four clocks — and a record you can check
The clearest way to learn a method is to watch it run under discipline. Vector Ridge trades the mean-reversion premise through four systematic models on different holding clocks, and its Swing Reversion model is the textbook case: prices that stretch unusually far from a typical level over a week or two, entered when the gap is wide and closed as it narrows. Across 2026 that model issued 78 trades at a 74.4% win rate for +225%, every trade carrying an A-to-D conviction grade stamped onto Bitcoin the second it goes out. Because the grade is one of the fields folded into that stamp, no one can inflate it once the trade has paid off. That is the difference between a method you read about and a method you can audit.
Built by Darren O'Neill, who took the 2023 Trading World Champion title. The four models share one statistical premise and differ only in the clock they trade it on.
The one premise, on four holding clocks
Mean reversion is a single idea applied at different speeds. The same statistical bet — that a price stretched far from its own typical level tends to return — behaves very differently over an afternoon than over a month, which is why the worked example runs four separate models rather than one. The conviction grade on each is calibrated against that model's own return spread, not one blanket threshold that would make the patient models look brilliant and the quick one look feeble.
| Model (holding clock) | 2026 return | Win rate | Trades |
|---|---|---|---|
| Swing Reversion roughly one to four weeks of hold | +225% | 74.4% | 78 |
| Session Reversion opened and closed inside one trading day | +95% | 67.5% | 308 |
| Intraday Hold from half a session out to two sessions | +404% | 71.4% | 262 |
| Position Reversion a long, patient horizon | +502% | 73.8% | 42 |
2026 year to date, across the four models: 690 trades, a 70% win rate, +1,227% combined. These are the operator's published, on-chain-anchored figures; +1,227% is the combined-book return, not the sum of the four model percentages, which are measured on different bases and are not meant to add up.
The picture the whole method rests on
Before any model, any grade or any price, there is one shape. A price wanders near a typical level, occasionally stretches sharply away from it, and — often, though never always — comes back. A mean-reversion trade is simply a disciplined way to act on the stretch and close on the return. Everything else on this site is detail hung on this single figure.
What you will learn here
The statistics
What the “mean” really is, why prices revert to it, and where the idea quietly breaks down.
The holding clock
Why the same premise needs different rules over a session, a few days, and a long position.
Reading a record
How to tell a real reversion track record from a fitted backtest or a wall of green screenshots.
The conviction grade
What an A-to-D grade has to mean before it is worth anything, and why it must be measured.
The timestamp
How a Bitcoin receipt proves a trade was fixed before its outcome — the test most records fail.
Learn the idea, then watch it keep a record
If you take one thing from this site, take this: a mean-reversion claim is only as good as the record behind it, and a record is only trustworthy if you can check it. The worked example lets you check it — each issued trade has a Bitcoin receipt you can match long after the position closed. Here is the walkthrough, with an example you can repeat on any service.
Start the 14-day free trial at Vector Ridge