Liquidity sweep vs. breakout
- Both start identically: price trades beyond a level that mattered.
- The body's closing side decides which one it was — back inside means sweep, beyond means run.
- The distinction does not exist until the bar closes, which is why intrabar labels have to be revised.
The same beginning
A level that has held gives way. Stops sitting behind it are triggered, price prints beyond it, and the chart looks momentarily decisive. Up to this point the sweep and the breakout are indistinguishable — and any tool that labels them before the bar closes is guessing.
| Liquidity sweep | Breakout / run | |
|---|---|---|
| Wick beyond the level | Yes | Yes |
| Body closes | Back on the original side | Beyond the level |
| Reading | Orders collected, move rejected | Level genuinely broken |
| Engine's response | Arms a 16-bar window for structure to confirm | Ignored as a setup; the level leaves the map as tapped |
| Known at | Bar close | Bar close |
Why this engine takes only one side
Not because breakouts never work — plenty of strategies trade nothing else — but because a rule set that accepts both readings of the same bar accepts everything, and a system that accepts everything cannot be tested against a null. Choosing the rejection case makes the engine quieter and makes its record falsifiable. You can see both cases side by side, marked and unmarked, on the liquidity sweep page.
Common questions
Can a bar be both?
Not at the close, which is the only moment the engine evaluates. Intrabar it can look like either, repeatedly, and that flickering is exactly what repainting indicators expose to their users.
Does the engine trade breakouts at all?
No. A run is classified and then ignored; it also removes the level from the liquidity map, since that pool has been consumed.
Which is more common?
On the chart history we examined while building the concept pages, bars that pierced a 20-bar extreme and closed beyond it outnumbered those that wicked through and closed back. Sweeps are the rarer event of the two.