A liquidity sweep is price pushing just past an obvious level, far enough to trigger the stop-losses sitting there, then reversing hard the other way. Some people call it a stop hunt. Same event, two names. I used to think it was personal — like the market waited for my stop specifically. It wasn’t. It was waiting for everyone’s stop at once, because that cluster of orders is the easiest liquidity on the board.
I don’t trade with RSI, MACD, or volume profile anymore. I deleted all of it. What’s left is a blank chart, a few horizontal lines drawn where price has stopped before, and a habit of watching how a candle closes instead of where it pokes. That habit is the entire difference between getting swept out of a good level and holding through one.
Why a liquidity sweep happens in the first place
Every trader watching an obvious support or resistance level puts a stop just past it. Long positions get stops a few cents below support. Short positions get stops a few cents above resistance. Breakout traders set buy orders just above resistance, ready to enter the moment it “breaks.” All of that sitting order flow is liquidity — real money, resting at a predictable address.
A level with a lot of eyes on it isn’t dangerous because of the level itself. It’s dangerous because of what’s parked past it. Larger players need volume to fill big positions without moving price against themselves the whole way. A cluster of stops just beyond a well-known level is exactly that volume — a pool that can be traded into, filled against, and then left behind once price reverses. The sweep isn’t malice. It’s the shortest path to the order flow that happens to be sitting there.
This is why sweeps cluster at levels that look obvious on a chart. A level nobody’s watching doesn’t have stops stacked behind it. A level everybody’s watching does. The more textbook the support looks, the more bait it becomes.
How to tell a liquidity sweep from a real breakdown
Here’s the mechanical difference, and it doesn’t require anything beyond the candle itself. A sweep pushes past the level, then reverses fast and closes back inside the prior range — often within the same candle, sometimes within one or two afterward. A real breakdown holds below the level and closes there cleanly, without snapping back the way a sweep does.
Look at the shape. A sweep candle usually has a long wick poking through the level and a small body that closes back near where it opened — sellers hit the bid hard the second price dipped below support, and buyers took it right back. A real breakdown candle closes near its low, with a body that runs most of the candle’s length. One shows rejection. The other shows conviction. You’re reading which one happened by waiting for the candle to finish, not by reacting to the wick crossing the line.
Time matters too. A genuine breakdown tends to keep making lower lows over the next several candles. A sweep usually snaps back within minutes to a few hours and rarely revisits the extreme again. If price is back above the old support inside the same session it dipped below, that’s not a breakdown anymore. That’s a stop hunt that already did its job.
Why indicator traders get faked out by a sweep more than price-action traders do
Most indicators confirm a break after it happens, not while it’s happening. An RSI reading crossing oversold, a moving average cross, a volume spike alert — these all fire once price has already moved through the level, which is exactly the moment a sweep is running out of steam and starting to reverse. The indicator trader gets their “confirmed breakdown” signal right as the real move is turning against them. They short the bottom of the sweep because the tool told them to, one candle after the market already decided to go the other way.
A trader watching raw price action isn’t waiting on a lagging calculation. They’re watching whether the candle closes past the level or snaps back into it. That’s the same information the indicator is trying to summarize, just without the delay a formula built on past bars always carries. The read isn’t better because it’s simpler. It’s better because it isn’t late.
The sweep I held through instead of panicking
SPY had support at 548 that had held twice over the prior three weeks — the kind of level enough people were watching that I half expected someone to test it a third time. On a Wednesday afternoon it did, dropping to an intraday low of 546.80 before the candle closed back at 548.90, above the level again. I had a long position on with a stop resting at 546.50, and for about four minutes I watched price sit under my level and had to fight the urge to close the trade early and take a smaller loss before it “got worse.”
I didn’t touch it. The candle’s body was thin, the wick beneath it was long, and by the time it closed, price had already reclaimed 548. That’s the anatomy of a sweep, not a breakdown — the stops below 548 got run, filled, and left behind, and price didn’t have anywhere to go but back up once that supply was gone. SPY closed the week at 554.20. My stop never got touched. The only thing that would have hurt me was reacting to the low of the candle instead of waiting for its close.
The breakdown that kept going
A different setup, a few weeks later, taught me the other side of this. NVDA had support at 128 — tested twice, held both times, another level with plenty of stops stacked underneath it by the third approach. It came down to 128 on a Monday, and this time the candle closed at 126.40. No snapback. No reclaim the next session either — Tuesday’s candle opened below 128 and closed at 124.90, making a clean lower low with a body that ran most of the candle instead of a thin one tucked under a long wick.
I was flat going in, watching from the sideline, and I stayed flat until the second candle confirmed what the first one suggested — this wasn’t a sweep-and-reverse, it was a level actually giving way. NVDA kept falling to 119 over the following eight sessions. Same setup on paper as the SPY trade — price dipping below an obvious level with stops behind it — but the close behavior told a different story from the first candle on, and the second candle removed any doubt. Reading the closes instead of the wicks is what separated these two trades, not the tickers, not the sector, not luck.
How to actually trade around a sweep
The rule I use is short enough to fit on an index card. Don’t react to price touching a level. React to how the candle closes relative to that level, and if you’re unsure, wait one more candle for confirmation before doing anything.
If price wicks past a level and closes back inside the prior range, that’s a sweep — the stops got taken, the move is probably exhausted, and the higher-probability trade is often in the opposite direction of the wick, back toward where price came from. If price closes past the level and the next candle holds there instead of snapping back, that’s a real break, and the sweep read was wrong. The two look identical for the first thirty seconds after the wick prints. They only separate once the candle closes, which is the entire reason waiting for the close beats reacting to the touch.
This won’t catch every sweep before it happens, and it won’t spare you every loss. Some sweeps run for two candles before reversing instead of one, and by the time you’re sure, you’ve given back part of the move. That’s the cost of not guessing. It’s smaller than the cost of getting stopped out on every single stop hunt because you treated the wick as the verdict instead of the opening argument.
I trade with a blank chart and a few lines in Static, the free daily chart room run by Draw Lines Make Money. If this way of reading a liquidity sweep makes sense to you, you can sit in and watch how it’s done live:
Join the free Static chart room →
Disclosure: that’s an affiliate link — I may earn a commission if you join a paid tier later, at no extra cost to you. The free room is free.