5 / 7 · 7 min
Liquidity: where other people's stops are sitting
Under every visible low there is a cluster of stop orders. That is verifiable mechanics, not a conspiracy — and the difference between those two explanations decides how you will trade.
Everyone who bought puts a stop below the nearest low, because that is where their reasoning stops being valid. So under every noticeable low a cluster of sell-stops accumulates. When price reaches it they all fire at once and turn into market sells: a sharp jab down, often followed by an equally sharp return. The selling happened not because somebody decided to sell but because they were forced to.
Why a reversal often follows a sweep
The key word is forced. An ordinary sale means somebody judged the price to be high. A triggered stop means nothing of the sort: the person did not judge the price, a rule did it for them. Such selling ends instantly — once the cluster is exhausted there is nobody left to press. Hence the familiar picture: a long wick down and a return into the old range within a candle or two.
Where clusters sit most often
- Below a significant low and above a significant high — the most obvious place, and therefore the most crowded.
- Beyond a round number: people set targets and stops on round figures more often than on any others.
- Beyond the previous day's high and low — every trading application computes them.
- Beyond EQUAL highs or lows: two attempts at the same price leave two layers of stops in one place.
Equal highs and what lies beyond them
Price reached 2,486 and 2,487 twice and pulled back both times. Those are equal highs: whoever sold there put stops slightly above — say around 2,490. A day later price spikes sharply to 2,498 and within ten minutes is back below 2,486. What happened: the stop cluster fired, the shorts closed with market buys, and after them there was nobody left to buy. An entry on such an event is taken NOT on the spike — there you buy alongside other people's stops at the worst price — but after the return below 2,486, with a stop beyond 2,500. The spike alone is not a signal: what makes it one is the fast return.
Believing stops are taken DELIBERATELY
"The market maker went for your stops" is the familiar explanation, and nothing can test it. Worse, it is harmful: it supplies a picture of the world in which somebody specific is playing against you, so you can take offence, seek revenge and "outplay" them. Meanwhile the mechanics work WITHOUT any intent: a cluster of orders attracts price simply because there is somebody there for a large participant to trade with. He does not need to punish you, he needs size — and he goes where size is.
Calling every wick a liquidity sweep
Half the candles on a chart have wicks. Call each of them a sweep and everything is explained while nothing is predicted. Three conditions worth demanding: the spike must go BEYOND a noticeable level, not a random local peak; the return must be fast — one or two candles of your scale; and volume on the spike must be above normal, otherwise there was nothing there to sweep. Without those conditions you have an ordinary swing with a fancy name.
Putting your own stop exactly where everyone else does
The direct practical consequence of the whole lesson. If you understand that a cluster sits below the low, do not add your own stop to it: it will be taken out along with the rest and price will return without you. It is wiser to move the stop further — beyond the zone, not beyond the round number — and reduce position size so the money risk stays the same. This is exactly the trade-off from the lesson on levels: a wide stop costs a smaller size but survives the noise.
Forced closing leaves a trace visible in numbers: open interest. If during a jab price moved and open interest FELL SHARPLY, positions were being closed rather than opened — that is, it was a sweep, not new entry. If price rises and open interest rises with it, new money is coming in, which is an entirely different event. The same candle means different things depending on what open interest was doing.
Open interestTwenty spikes and what came after
Find twenty cases where price went beyond a noticeable low or high and came back within two candles. For each, record: was volume on the spike above normal, what did open interest do, and where did price go over the next ten candles. Then split the cases into those where open interest fell and those where it rose, and compare how many reversed. That is the only way to learn whether this sign works for YOU — rather than believing it works for the author of yet another course.
Price pierced the low and did not come back but kept falling. Is that also a liquidity sweep?
No, and confusing the two is expensive. A sweep is when price went for a cluster and returned: the selling was forced and it ended. If price left and stayed there, then it was not only triggered stops selling — a genuine reason to go lower appeared. The first minutes look identical from outside, and only the return distinguishes them. That is exactly why the entry is taken AFTER the return and not on the spike: the return is the thing that tells one from the other.
Why do equal highs attract price more strongly than a single one?
Because stops accumulate there twice. After the first high, stops were placed by whoever sold at it; after the second, by whoever sold at the second — and they landed at PRACTICALLY the same price. That gives a double layer in one place, and for a large participant who needs size it is more profitable to go there than anywhere else. No intent is required for this — it is enough that there is somebody there to trade with.