Stop hunting (stop hunt, liquidity sweep): what Dow Jones futures data says
Stop hunting (stop hunt, liquidity sweep) is the claim that big players deliberately push price beyond an obvious high or low, trigger smaller traders’ stop orders and reverse the market. I tested it on 143 sessions of Dow Jones futures: after such levels were broken, the market reversed practically as often as after crossing a random line.
What it is
The concept spread on social media together with ICT and Smart Money Concepts. The typical story: retail stops sit above highs and below lows, “smart money” pushes price there, fills those orders as its own liquidity and reverses the market. Hence the slogans “sweep”, “inducement” and “they grabbed liquidity”.
The phenomenon itself is old and has had names for about a hundred years: spring and upthrust in the Wyckoff method (1910s–1930s), the false breakout in Edwards and Magee’s “Technical Analysis of Stock Trends” (1948), Linda Raschke and Laurence Connors’ “Turtle Soup” in “Street Smarts” (1995), excess and poor highs/lows in Market Profile (Dalton, “Mind Over Markets”, 1990). What is new is mostly the label.
Test on Dow Jones futures: 143 sessions from CME
I tested Dow Jones futures (MYM), every trade, regular session 9:30 AM–4:00 PM New York, from 21 April 2026 to 4 November 2026. The “liquidity” levels are the places where stops sit most obviously: the high and low of the previous session and the high and low of the overnight session. A break = the first trade 3 points beyond the level. Then I check what came first within 90 minutes: a reversal of X points (what stop hunting promises) or a continuation of X points. For comparison, the same test on random lines drawn 30–250 points from the open.
| Threshold X | Day and overnight highs/lows: reversal | Random lines: reversal | Difference |
|---|---|---|---|
| 50 pts | 47% (115 of 244) | 42% (143 of 341) | +5 pp |
| 100 pts | 53% (117 of 222) | 52% (162 of 309) | +1 pp |
| 150 pts | 57% (98 of 173) | 58% (135 of 231) | −1 pp |
After a break of an “obvious” high or low the market reversed practically as often as after crossing any line. On reversals, price overshot the level (median) by about 20–45 points, and at random lines by about 38 points, so no difference. Reversal is a normal feature of the market, not proof that someone was hunting.
This does not mean levels are meaningless. It means that a break and a return on their own decide nothing. The difference between a trap and the start of a move shows in what happened at the level: in volume, aggression and absorption. More in the order flow section and in the Dow Jones statistics.
How I count: the first break of a given level in the session, levels already broken at the open are skipped, outcome decided within 90 minutes. Historical statistics, not a recommendation.
How people use it
- entering after price returns below a broken high or above a broken low
- the previous day’s and overnight high and low as “liquidity pools”
- combining a sweep with fair value gaps and kill zones
- explaining every wick on the chart as “liquidity taken”
What the research and evidence say
I know of no independent research confirming an edge for the rule “trade against a broken high after stops are taken”. There is also a logical problem: when price comes back, people say “they took the stops”, and when it keeps going, “it was a real breakout”. A theory that explains both outcomes predicts nothing. That is why, instead of quoting opinions, I counted it on my own data (section above).
Weaknesses
- after the fact it fits every move: both the reversal and the continuation
- “liquidity” inferred from candle shape, with no order data
- shown on hand-picked examples, without counting the cases where the breakout kept going
- attributes intent (“someone was hunting”) to what may simply be a lack of participants on the other side
What order flow shows
Order flow replaces the question “was someone hunting?” with one that can be answered: what happened at the level. If a wave of market orders hit beyond the high (stops being filled), large limit orders absorbed it and price stopped moving, that is the picture of a failed breakout. If aggression beyond the level continues and trading moves to new prices, the market accepts the breakout. You see this in the footprint, delta and volume profile, not in the shape of the wick.
How to use it sensibly
- treat a “sweep” as a hypothesis about a specific level, not an explanation of every move
- count all breakouts, including those that kept going
- compare with random levels before you decide a place is special
- at an important level, check for absorption or acceptance in the order flow
Related
- Glossary: Stop run
- Glossary: Sweep
- Glossary: Liquidity
- Glossary: Absorption
- Glossary: Acceptance and rejection
- Atlas of biases: Cherry-picking
- Atlas of biases: The narrative fallacy
- Order flow: Absorption on a real market: 51,300
- Also under the microscope: Candle Range Theory (CRT)
- Also under the microscope: ICT (Inner Circle Trader)
- Also under the microscope: Smart Money Concepts (SMC)
- Also under the microscope: The Wyckoff method
Frequently asked questions
What is a stop hunt?
The claim that big players deliberately move price beyond an obvious high or low to trigger other traders’ stop orders, take that liquidity and reverse the market. The concept was popularised by ICT and Smart Money Concepts.
What is a liquidity sweep?
Another name for the same phenomenon: a quick break of a high or low after which price comes back. Earlier it was described as a false breakout, a spring and upthrust (Wyckoff) or excess in Market Profile.
Does stop hunting work?
In our test on 143 sessions of Dow Jones futures, after a break of the previous day’s or overnight high or low the market reversed by 100 points in 53% of cases, and at random lines in 52%. These levels did not produce clearly more frequent reversals.
Do big players hunt stops?
Stop orders beyond obvious extremes are real liquidity and the market often gets there. But a wick on a candle does not show who the aggressor was or whether anyone planned anything. You only see that in volume and order flow.
How do you tell a stop hunt from a real breakout?
Not by the shape of the candle. In a failed breakout, aggression beyond the level is absorbed by limit orders (absorption) and price comes back. In a real breakout, aggression continues and trading moves to new prices.
Educational content, not investment advice. No signals, entries, stops or position sizes.