Liquidity, Sweeps and Stop Runs: A Descriptive Look

Before you read: Reading the DOM Honestly: Spoofing, Icebergs and Hidden Size

What you will learn

• What liquidity and resting liquidity mean in an order book.
• What a liquidity sweep and a stop run describe, and why the words are applied after the fact.
• Why claims about “hunting” stops cannot be checked from a chart.

What is liquidity in an order book?

If you wanted to buy 50 contracts right now, how much of the book would you use?

Liquidity is how much can trade near the current price without moving it far. Resting liquidity is the set of limit orders waiting in the book at each price. A deep book has many contracts resting near the best bid and ask. A thin book has few, so a modest market order can trade through several prices.

Where do stop orders tend to cluster?

If many people use the same chart, where might they place their stops?

A basic stop order becomes a market order once price reaches its trigger. Stop-limit and other stop variants behave differently, and handling can differ by venue and broker. Many traders look at the same obvious levels: a prior day high or low, a round number, the overnight high and low. Stops are often described as clustering just beyond such levels. Those orders are not visible in the book before they trigger, so a cluster is an assumption, not an observation.

What is a sweep?

What does it look like on a chart when price trades a short way through a level and then returns?

A sweep describes a fast move that trades through a level, fills orders resting beyond it, and then trades back. On a bar chart, it often leaves a long wick beyond the level. On a bid/ask volume chart, it may show heavy aggressive volume at the extreme of the bar. When the triggered orders are thought to be stops, the move is often called a stop run.

The same picture has other readings. Price may have traded through a level because new sellers arrived, because a large order was being worked, or because the book was thin for a moment. A move is usually only called a sweep after price has returned. If price had kept going, the same first leg would be called a breakout.

Sweep diagram

Illustrative diagram constructed for teaching. Not real market data.

The diagram shows a fictional market trading briefly above a prior high on a volume spike and returning below it. One dashed path continues lower, labeled “later called a sweep”, and another continues higher, labeled “later called a breakout”. The first leg is the same in both. All values are made up.

Is someone “hunting” stops?

What evidence would you need to show that one participant moved price to trigger others’ stops?

Some commentary describes sweeps as deliberate “stop hunts” by large participants. That claim needs to know who traded and why, which a chart, the DOM or a bid/ask volume chart do not show. Deliberately moving price to trigger other people’s orders can be market manipulation, which regulators prohibit. A descriptive reading stays with what is visible: where price traded, how fast, on how much volume, and what happened next.

What this does not tell you about sweeps

• Where stops actually rest. They are not visible before they trigger.
• Who caused a move or why. Charts do not identify participants or intent.
• Whether the next move through a level will return or continue. The label depends on what follows.
• Nothing here suggests buying, selling, or where to place an order.

Method in Practice: trading the return, not the label

Fictional example. John, Harbor Index futures, the prices, times, costs and results are all invented for teaching. This shows how a written plan can be built and tested. It is not a recommendation, and nothing here says how often this idea works.

Try it on a real chart. Test this idea yourself on past sessions with bar replay. Keep your own count, including the days with no trade.

The idea. The lesson says a move is only called a sweep after price comes back, and no chart shows who caused it. So John's plan never uses the word. It asks for something he can check as it happens: price trades below the prior day low, then closes back above it. The no-trade replay is the day the same first leg was later called a breakdown.

John's plan

Setup

prior day low 100.00, marked on the pre-session map

Enter if

price trades at least 0.10 below 100.00, then a 5-minute bar closes back above 100.00. A buy stop goes one tick above that bar's high.

Wrong if

price trades one tick below the day's low (stop order)

Manage

take half at 100.70, the rest at 101.10 or by 14:30

Risk

$150 maximum, costs included. Size depends on the distance to the stop.

Three outcomes

• ✅ Worked: price came back and kept rising → +$120

• ❌ Failed: price came back, then fell through the low → −$120, inside the limit

• ⏸ No trade: price went through the low and never closed back above it → $0

Show the full replays

Item	Plan
Context	Prior day low 100.00
Entry condition	Price trades 99.90 or lower, then a 5-minute bar closes above 100.00. Buy stop one tick above that bar's high
Wrong if	A stop order one tick below the day's low
Management	Take profit on half at 100.70 (limit) and move the stop for the rest to the entry price. Rest at 101.10 (limit). Time limit 14:30 CT
Risk cap and size	$150. Cost per contract = (entry stop − exit stop + 2 ticks of slippage) × $100 + $5. Contracts = $150 ÷ that, rounded down
Not trading if	Scheduled major release within the hour, or no setup by 12:00 CT

Replay A: worked

Time (CT)	What happened	John
09:00–09:10	Trades to 99.80	Watching
09:10–09:15	Closes 100.10, high 100.15	Buy stop at 100.20, stop below at 99.75. Cost per contract $60, so 2 contracts
09:18	Trades 100.20	Filled 2 at 100.25
09:18–10:50	Trades 100.70 at 10:50	Sells 1 at 100.70 (limit). Stop for the other moves to 100.25
10:50–13:15	Never back to 100.25, trades 101.10 at 13:15	Sells the last 1 at 101.10 (limit)

Result after costs: +0.45 point = +$45, +0.85 point = +$85, commissions −$10. Net +$120. Journal: "Below the low, back above, and through my buy stop. Both exits were limits."

Replay B: failed

Time (CT)	What happened	John
09:30–09:40	Trades to 99.80	Watching
09:40–09:45	Closes 100.08, high 100.15	Buy stop at 100.20, stop below at 99.75, 2 contracts
09:47	Trades 100.20	Filled 2 at 100.25
09:47–10:05	Turns down, trades 99.75	Stop triggers, filled 99.70
10:05–15:00	Falls to 99.30	Done for the idea

Result after costs: 2 × −0.55 point = −$110, commissions −$10. Net −$120. Journal: "It came back above the low, then went through it. The full cost was the one I planned."

Replay C: no trade

Time (CT)	What happened	John
09:05–09:15	Trades to 99.80	Watching
09:15–12:00	Keeps falling to 99.20. No 5-minute close above 100.00	No setup. 12:00 cutoff

Result: $0. Journal: "Same first leg as the other two days. This one would later be called a breakdown, not a sweep. My rule only needed to know whether it closed back above."

Test it yourself

• Write the minimum distance below the level (here 0.10) and the bar length before testing.
• Replay every trade below a marked low, including the ones that kept going. Those are the days the first leg looked the same.
• Work out the size from the distance to the stop each time, using the position size calculator.

Where it breaks

• You can't see stops before they trigger. "Stops below the low" is an assumption.
• A return above the level can fail (the failed replay).
• The same first leg can become a breakdown (the no-trade replay).

Calculators: position size · risk/reward · expectancy

Reflection question: Which words in your plan can you check while it's happening, and which can only be applied afterwards?

Check yourself

1. Why is a stop cluster an assumption rather than something you can see in the book? Hint: when does a stop order enter the book as a market order?
2. Price trades two ticks above the prior day high and then falls. What would the same first leg be called if price had kept rising? Hint: what label depends on what follows?
3. What information would you need to support the claim that a move was a deliberate stop hunt? Hint: who traded, and why?

Last reviewed November 8, 2026. Spotted a mistake? Report an error.