Liquidity Sweeps: How to Spot a Stop Hunt Before It Traps You

You put the stop three ticks under yesterday low. Price drifts down, tags it, and thirty seconds later runs forty points without you. It feels like somebody is reading your screen. Nobody is. What happened is simpler and far more useful to understand: somebody needed to buy size and went to the one place where sellers were guaranteed, which is exactly where your stop was resting.

That is a liquidity sweep. It can be recognized, it can be traded, and above all it can stop being personal.

What a liquidity sweep is, and why it is not a conspiracy

Futures trade through continuous auction: for you to buy, somebody has to sell at that exact price and moment. If a large participant needs 800 NQ contracts, a quiet tape crossing twenty contracts a minute is useless. They need a place with sellers in size.

Where are sellers in size? Right below an obvious low. Two things sit stacked there:

• Stops from everyone who is long, which convert into market sell orders the instant they trigger.
• Breakout sell orders from everyone waiting to short if that low gives way.

Both do the same thing: they sell. And both fire automatically if price touches the level. For a large buyer that is a grocery store with the doors open. Push price a few ticks below the low, absorb all of that supply, and price snaps back because the selling pressure exhausted itself in twenty seconds.

Nobody is watching your account. What exists is a map of where the orders are, and that map is public, because it is made of the same obvious levels you can see. That is the good news. If you can see them, you can anticipate where price will go looking for the other side.

"Stop hunt" is a bad name

It implies personal intent. Liquidity seeking is more accurate. Price is not coming for you, it is coming for volume. Swapping that phrase in your head changes how you trade: you stop being outraged and start asking where the next pocket of orders sits.

Where the stops actually pile up

Not every level qualifies. Liquidity stacks where a lot of people see the same thing. Here is the list ranked by reliability, with the window in which the sweep usually happens, in Eastern time.

Level	Where orders stack	Reliability	Usual window
Yesterday high and low	2 to 10 ticks beyond	High	9:30 to 11:00 a.m.
Opening range extremes	3 to 8 ticks beyond	High	9:40 to 10:30 a.m.
Overnight high and low	2 to 6 ticks beyond	Medium high	9:30 to 10:15 a.m.
Intraday double top or bottom	Just past the second touch	High	All session
Low of a 30 minute balance	4 to 10 ticks below	Medium	After 11:00 a.m.
Round numbers (21,000 NQ, 6,000 ES)	Around the figure	Medium	All session
Weekly high and low	5 to 20 ticks beyond	High but slow	Monday and Friday

Notice the pattern: the most reliable levels are the oldest ones and the ones most traders have drawn. Everybody sees yesterday low. A swing low from eleven minutes ago on a 512 tick chart is visible to you and four other people. Liquidity accumulates where there is visual consensus.

For more on the levels that keep getting respected without anyone quite knowing why, see previous day high, low and close.

What a sweep looks like on the chart

A textbook sweep has four parts, and you need all four:

1. Slow approach. Price drifts toward the level without urgency, usually with small bars and declining volume. That is the market waiting.
2. Fast penetration. One or two bars break the level with speed and heavy volume. Usually the largest bar of the last twenty.
3. No continuation. This is the key. The break does not extend. Price hangs there for two, three, five bars going nowhere.
4. Return. Price comes back through the level and does it decisively, not crawling.

Step 3 is what separates a sweep from a real breakout. A good breakout extends: it breaks and keeps going in the next three bars. A sweep breaks and stalls. If five bars after the break price has not advanced past the break point, what most likely happened is that somebody collected orders and that was the whole event.

A long wick proves nothing

You hear constantly that a candle with a long wick under the low is a sweep. Sometimes. But a long wick only says price was there and came back. On one minute charts at the open you get long wicks every third bar and most of them mean nothing. The wick is a hint, not a confirmation. Confirmation comes from volume and delta.

Confirming it in the order flow

This is where order flow stops being decoration and does actual work. Three things to check when price breaks the level:

Volume on the break. A sweep produces a volume spike. If price loses yesterday low on the same volume as the previous bars, nothing was swept, it is simply going down. The spike means a lot of orders executed at once, which is exactly what a cascade of stops does.

Delta against price. If price prints a new low but cumulative delta does not, buyers are absorbing the selling. That is the classic divergence and it shows up in almost every real sweep. Detail in cumulative delta and divergences.

Absorption at the level. On a footprint chart you see a cell with heavy sell volume at one price and price refusing to drop. Somebody is eating all of that supply with resting limit orders. It is the most reliable of the three, and we cover it separately in absorption in order flow.

With all three lining up, the odds move a lot. With one of them, you are guessing with decoration.

How to trade it: entry, stop and target

This trade has one huge virtue: the stop is small and obvious. If the sweep is real, price should not revisit the wick extreme.

Entry. Do not enter while price is still outside the level. Wait for it to come back inside and for a bar to close inside. In MNQ that usually costs you 8 to 15 points of worse fill, and in exchange it removes about half your losers.

Stop. A few ticks beyond the sweep wick extreme. In MNQ that is typically 20 to 35 points, or $40 to $70 per contract. In MES, 6 to 10 points, $30 to $50. If the wick extreme sits 60 points away, the sweep was too deep. Pass.

Target. The natural first target is the liquidity on the other side. If you went long after a sweep of yesterday low, the logical destination is the session high or yesterday high. That usually pays 2R to 4R, and that ratio is why the trade is worth taking even though it fails often.

When it is not a sweep and it costs you money

This is the part most videos skip, and it is where the money goes.

On strong trend days everything looks like a sweep

On a day when the Nasdaq drops 400 points, every broken low looks like a sweep and none of them are. Price breaks, pauses ten seconds because a few buyers show up, and keeps falling. Fade each of those "sweeps" and you take six losers in a row.

Simple filter: if price has broken three consecutive lows in the last hour without reclaiming any of them, you are not in a sweep market. You are in a trend. Stop looking for reversals.

In dead hours

At 3 a.m., a twenty point move that breaks a low and returns is not an institutional sweep. There were four orders in the book. The pattern looks identical and it means something else, and the continuation never arrives. The volume map by hour is in futures market hours.

The second sweep of the same level

The first run at yesterday low usually works. The third time price visits that same level in the same session, there is no liquidity left. It got taken. A swept level is a spent level.

Sources and further reading

• barras volumétricas de NinjaTrader · cómo ver la agresión compradora o vendedora en la mecha del barrido
• documentación del SuperDOM de NinjaTrader · dónde se observa la profundidad y la retirada de órdenes
• curso de CME Group sobre profundidad de mercado · base objetiva de por qué el precio busca zonas de liquidez