Ever been stopped out at the exact low, right before the market reversed without you? Traders call that a liquidity sweep, or less politely, a stop hunt. It's one of the most searched — and most conspiracy-loaded — ideas in trading. This article explains what a sweep actually is, why it happens without anyone hunting you, how to tell a sweep from a breakout, and what the footprint shows when a sweep is real.
(Quick note: nothing here is financial advice. Sweeps are a pattern to understand, not a signal that guarantees a reversal.)
What a liquidity sweep is
A liquidity sweep is a three-beat move:
- Price approaches an obvious level — a prior high or low everyone can see.
- It spikes through it, often fast and wick-heavy, triggering the stop orders resting there.
- It fails to hold — closing back inside the old range, often followed by a strong move the other way.
The move takes the level but doesn't keep it. Same pattern, many names: stop hunt, stop run, liquidity grab, raid, swing failure pattern (SFP). Different vocabularies, one chart picture.
Why stops cluster at obvious levels
There's no mystery here. Nearly everyone is taught the same rulebook: stop below the last swing low, stop above the last swing high. Equal highs and equal lows look "confirmed," so even more stops stack behind them. Previous day high and low are drawn on every platform. Round numbers attract orders — one of the few things in this topic with actual academic research behind it.
Millions of traders, same rules, same levels. The stops don't cluster because anyone sees your order — they cluster because everyone thinks alike.
The honest mechanics — no villain required
Here's the part most stop-hunt content gets wrong. A stop order is just a conditional market order: when price touches it, it fires as a market order in the direction of the move. Sell stops below a low fire as market sells.
Now think like a fund that wants to buy 2,000 contracts. Lifting offers in a quiet market moves price against you while you fill. But just below an obvious low, a burst of triggered sell stops is dumping contracts at market — that's exactly the selling a large buyer wants to fill against. So large passive buyers rationally sit below obvious lows. The sweep is structural, not personal.
Worth saying clearly for futures traders: on CME there is one central order book, your broker doesn't set the price, and resting stops aren't visible to anyone. The "my broker spiked the feed to tag my stop" story belongs to unregulated CFD land — not exchange-traded futures.
And who caused any particular sweep? Honestly: unknowable. A big player probing the level, breakout traders failing and covering, or simple order flow running out of fuel — all three paint the same candle. The good news is the trade logic never requires knowing intent. Only recognizing failure-to-hold.
Sweep vs breakout — the real question
In hindsight, trivial. In the moment, they are the same first move. The difference is never the break — it's what happens after. Evidence traders stack, roughly in order of weight:
- Close back inside vs acceptance outside. A candle closing back within the old range points to a sweep. Consecutive closes and building value beyond the level point to a breakout.
- Speed of the reclaim. Sweeps tend to reject fast — a few candles. The longer price camps outside, the more it looks like acceptance.
- The retest. After a real breakout, the old level tends to hold when retested. After a sweep, it fails immediately.
- What follows. A meaningful sweep gets confirmed by displacement in the opposite direction — this is where CISD plugs in as the confirmation trigger.
None of this is certain in real time — some sweeps become breakouts on the second push. It's evidence stacking, not fortune telling.
Where sweeps happen most
Ranked by how often traders watch them: previous day high and low first — the most-watched intraday levels in futures. Then session highs and lows (the overnight range before the US open is the bread-and-butter example). Then equal highs and lows — the SMC favorite. Then prior week highs and lows, which matter more but happen less often. And round numbers.
In SMC language, these outer levels are external liquidity: price sweeps them, then often rotates back toward the fair value gaps and levels inside the range. That cycle has its own article — Internal & External Liquidity. And when two correlated markets disagree at the sweep — one makes the new low, the other refuses — that's SMT strengthening the read.
What the footprint shows on a real sweep
This is where order flow earns its keep, because most sweep content is taught with zero volume evidence. At a genuinely swept low, the footprint often shows:
- High volume, no progress. Big volume prints beyond the level, but price refuses to follow through — the breakout attempt is being absorbed by passive buyers. That pattern has its own article: Absorption vs Exhaustion.
- A delta flip. The sweep candle prints heavy negative delta, then the next candles flip positive as price reclaims the level — aggression changed sides.
- CVD divergence. Price makes a marginal new low, but cumulative delta makes a higher low — less real selling behind the push than the last one. Delta and CVD are covered in Delta & Cumulative Delta.
The honest caveat, as always: these raise probability, they don't confer certainty. Absorption can break on the second push.
Common mistakes with sweeps
- Calling every wick a sweep. After the fact, every rejection wick looks like a stop hunt. A real sweep needs a level that actually held liquidity — obvious, visible, meaningful timeframe.
- Expecting every sweep to reverse. Sweeps in the direction of the higher-timeframe trend often just keep going. Sweep ≠ automatic fade.
- Ignoring the bigger picture. Fading a 5-minute "sweep" while the daily chart is expanding toward that same level is fighting the tide.
- Entering on the sweep itself. No reclaim, no confirmation — that's catching knives with extra vocabulary.
- Using "stop hunt" as a coping word. If every loss is a stop hunt, no trade ever gets honestly reviewed. Sometimes the idea was simply wrong.
The stop-placement lesson isn't "trade without stops." It's that one tick beyond the obvious level is where everyone else's stop is too. Stops beyond the sweep-completion zone — sized accordingly — stand where the sweep has already failed.
What you need
Sweeps are visible on any candle chart. Seeing whether a sweep is being absorbed or accepted takes a footprint and delta — ATAS has both built in, with a free trial to practice on.
The bottom line
A liquidity sweep is price briefly taking an obvious level, triggering the stops there, and failing to hold — structural market behavior, not a personal attack. You can't know in the moment who swept or why, but you can read the response: reclaim speed, closes, the retest, and what the footprint says about absorption and delta. The pattern is real. The villain is optional.