Liquidity sweep: why price goes hunting for your stop

Updated on September 15, 2026

Your stops are someone else’s raw material

Start from a constraint everyone forgets: to buy, someone has to sell. On a small size that is no problem at all. On a large size the seller does not spontaneously exist, and buying at market would push price against you before you had even finished positioning.

So the question becomes: where do you reliably find a pile of sellers at one precise place? Answer: below a low. Because that is where buyers keep their stops, and a buyer’s stop is a sell order. Price only has to drop there for all of them to fire at once and supply the missing other side.

There is nothing personal in any of this, and that matters if you want to stop telling yourself stories. Nobody is targeting your position specifically. What is targeted is a concentration of orders that your stop happens to be part of, because you placed it in the same obvious spot as everyone else.

Where liquidity sits, and why it is predictable

Above highs sits buy-side liquidity: the stops of sellers and the buy orders waiting on a break. Below lows sits sell-side liquidity: the stops of buyers and the sell orders waiting. The symmetry is exact.

Some places hold far more than others, and those are the ones worth recognising. Two highs at the same price, or two lows at the same price, are the strongest magnets on the chart: once a level has held twice, everyone piles their stops just behind it, to the same fraction of a point. Then come yesterday’s high and low, the extremes of the current session, and round numbers.

The practical consequence is counter-intuitive. A level that looks solid because it has held several times is not only a line of defence: it is also, and mainly, a pool of orders that everyone has identified. Its apparent solidity is precisely what makes it a target.

Telling a sweep from a real break

This is the only question that really matters, and it is settled on two criteria.

A sweep goes clearly through the level and then closes back on the side it came from. The chart keeps a long wick and a candle body that did not follow. The move back is usually fast, as if price had no intention of staying up there.

A real break, by contrast, closes beyond the level and, above all, stays there. It is that second point that settles it: after a genuine break, price accepts the new prices and consolidates beyond the level instead of diving straight back.

One honest caveat remains: while the candle is forming, the two cases are rigorously identical on screen. You do not know whether you are watching a sweep or a break before the close, and sometimes not even then. Any method claiming to tell them apart live, every time, is describing a chart from the past.

The trap of fading it too early

Once the mechanism clicks, the temptation is immediate: see a wick poke below a low and buy right away, telling yourself it was a sweep and price will come back up.

Except a wick below a low, taken on its own, says nothing at all. It is compatible with a sweep and a bounce, and just as compatible with the start of a fall that keeps going. The sequence serious readings wait for has a second step: the sweep and then a change of character, meaning a break of structure in the new direction. The sweep provides the fuel, the CHoCH confirms somebody used it.

Without that second step you are not buying a liquidity grab, you are buying a decline in progress.

What it does not tell you

A sweep gives no timing. A level loaded with orders can sit untouched for weeks, and nothing obliges the market to go and take it on any given day.

There is also a memory bias worth guarding against. You remember vividly the times your stop was hit just before price left without you, far less vividly the times it was hit and the decline simply continued. The first is painful and tells a story, the second is unremarkable. That asymmetry makes the phenomenon feel more frequent than it is.

Finally, the most tempting conclusion is also the worst: widening your stops so you stop getting taken out. A wider stop is a larger risk at constant size, and the only correct way to compensate is to reduce the size, which means recomputing the position. That is exactly the subject of our page on risk per trade.

How AlphaGPT detects them

Liquidity zones are identified by computation: significant highs and lows, equal levels, extremes of previous sessions, across several timeframes at once. A sweep is then recognised by its geometry, a move through the level followed by a close on the opposite side, rather than by the impression it gives.

It serves as context and never as a trigger on its own: a recent liquidity grab increases the weight of a zone, it does not justify an entry without the rest of the reading. The full vocabulary is in our Smart Money Concepts glossary.