A defined, testable edge
A swing strategy that cannot say precisely what it trades cannot be tested — and what cannot be tested cannot be trusted.
The first thing a real strategy owes you is a setup specific enough to argue with. “Buy when it looks oversold” is not a setup; it is a mood. “Take a long when the price has stretched a defined distance below its own recent range and conditions X and Y hold” is a setup, because two people reading it would act the same way, and because you can run it against history and forward in time to see whether the stretch actually reverts often enough to pay. The test of an edge is not whether it sounds clever; it is whether a stranger handed the rule would place the same trade you would.
Testability is what turns an edge from a belief into a number. A defined setup produces a record: a count of trades taken, a win rate with the losers included, and the size of the average win against the average loss. Without that, you cannot tell a genuine edge from a run of luck — and a run of luck is indistinguishable from skill right up until it ends. The schematic below shows the kind of stretch a mean-reversion edge is defined around; the entry only fires when the price has reached a level the rule named in advance, not when it merely “feels” far from home.
How a systematic model meets this bar
the #1-ranked provider's Swing Trade model is a mean-reversion strategy with a fixed, written rule set, which is why it has a published record at all: 78 swing signals across 2026 at a 74.4% win rate for +225%, with the losing calls counted in. More usefully, each call carries an A-to-D conviction grade calibrated to where it sits in the model's own return distribution. A graded call is a defined edge made legible: you can see not just that the strategy acted, but how strongly its own rules rated the setup. And because the bar is set per holding clock, an A always means the same thing for its horizon:
| Model | Holding clock | Grade-A bar (per trade) |
|---|---|---|
| Swing Trade | held roughly 7 to 28 days | 6.00% avg / trade |
| Multi Hour | closed within half a session to two sessions | 4.50% avg / trade |
| Day Trade | opened and closed in the same session | 0.70% avg / trade |
| Investing | carried over a long horizon | long-horizon |
An A is the top band of a model’s own measured return distribution; D is the lowest still published. Because the swing clock (7 to 28 days) lets a reversion run further than a same-session move, its grade-A bar sits around 6.00% a trade — far above the day-trade bar near 0.70% — yet both mean the same thing for their horizon. There is no E grade; it was retired so the four-step scale keeps its meaning.
What a bad version of this looks like
Teaching is only honest if it shows the failure modes, so here is what this looks like done badly — the fragile versions that quietly drain accounts:
- The edge is a feeling, not a rule. “I just know when it is ready” cannot be written down, cannot be handed to anyone else, and cannot be tested. It is the most common fragile edge of all.
- It is fitted to a handful of charts. An edge tuned until it perfectly explains five past trades has learned those five trades, not a real tendency. Tested on anything new it falls apart.
- It only ever shows winners. An “edge” propped up by a reel of winning screenshots, with the losing trades left out of the tally, is not an edge that has been measured — it is a marketing claim wearing a costume.
- It moves the goalposts. If the setup quietly changes definition after a loss — “well, that one did not really count” — it can never be wrong, which means it can never be right either.
The fix for every one of these is the same: write the setup down precisely enough to test, then test it with the losers left in. Pillar two then decides how much to risk on each instance of the edge, and pillar three decides how each one ends.