Liquidity Sweeps Tested: After a Stop Hunt, Price Reversed First Only 39% to 48% of the Time, and Fading It Lost on All 4 Charts
Key takeaways
- A liquidity sweep, or stop hunt, is a candle that pokes past a swing high or low and closes back inside. The story is that price “grabbed the stops” and will now reverse. I marked every one on EURUSD and gold, M5 and M15: 16,926 EURUSD bars and 13,736 gold bars.
- After a sweep, price went 1 ATR the reversal way first only 39% to 48% of the time. On all four charts it more often went 1 ATR further in the sweep direction first.
- Fading the sweep lost on all 4 charts, even before the spread. On EURUSD M5 it lost 2.63 pips a trade, worse than 95% of random-entry runs with the same exit.
- Trading the same wick the other way, as if the breakout were real, lost less than the fade on all 4 charts and made money on 2 of them. It still did not beat random entries.
- On gold more than half of all swing levels were taken by a wick, not a close. On this EURUSD feed only 9% were, because 91% of its M5 candles closed at their high or low.
The liquidity sweep is the second pillar of “smart money” trading, next to the fair value gap. Equal highs and swing lows hold stop orders. Big players push price through them, fill their orders against those stops, and then send price back the other way. You wait for the wick, see the close back inside, and enter the reversal.
It is a clean rule on open, high, low and close, so it can be counted. The question is simple: after a sweep, does price reverse more often than it would anyway?
What is a liquidity sweep?
| Term | Rule used in this test |
|---|---|
| Swing high | The highest high of 5 bars on each side. It is known only 5 bars later. |
| Swing low | The lowest low of 5 bars on each side, same delay. |
| Live level | A swing high or low that price has not traded through since. |
| Sweep | The first candle whose wick goes past a live level and which closes back inside it |
| Break | The first candle that goes past a live level and closes beyond it |
| Depth | How far the wick went past the level, in ATR(14) |
A sweep of a high is read as bearish, a sweep of a low as bullish. The same idea is called a stop hunt, a liquidity grab, a stop run or a “Turtle Soup” when the level is a 20-bar high.
How was it tested?
- Data: EURUSD M5 from June 3 to August 26, 2026, 16,926 bars; gold M5 from June 16 to August 26, 13,736 bars. M15 built from the same bars.
- Levels: swing highs and lows with 5 bars on each side, the same swings as in market structure tested. Every level counts once: the first candle that reaches it is either a sweep or a break.
- Reversal check: from the close of the sweep candle, which comes first within 48 bars: 1 ATR in the reversal direction or 1 ATR in the sweep direction. The same check on random bars as a control.
- Trades: fade the sweep (the textbook trade), follow the sweep (the opposite), follow a real break, and fade the sweep only on the EMA 200 side. Stop 1.5 ATR, target 3 ATR, 48 bars at most, one trade at a time.
- Cost and pass: 1 pip on EURUSD and $0.30 an ounce on gold, from spread and swap costs in MT4. A run passes if it makes money after cost, in both halves (split July 15), and beats the 95th percentile of 200 random-entry runs, as in algorithmic rule vs random entry.
How often does price sweep a level instead of breaking it?
| Swing 5 bars each side | EURUSD M5 | EURUSD M15 | Gold M5 | Gold M15 |
|---|---|---|---|---|
| Levels taken | 1,301 | 452 | 1,197 | 379 |
| Taken by a sweep | 112 | 109 | 677 | 199 |
| Share swept, not broken | 8.6% | 24.1% | 56.6% | 52.5% |
| Sweeps a trading day | 1.9 | 1.9 | 13.5 | 4.0 |
| Median wick past the level | 1.3 pips | 1.3 pips | $0.90 | $1.90 |
| Same, in ATR | 0.56 | 0.38 | 0.19 | 0.23 |
On gold a sweep is the normal way a level gets taken: more than half of all swing highs and lows were poked by a wick and closed back inside. That alone tells you a sweep is not a rare signal from big money. It is what gold candles do 13 times a day on M5.
The EURUSD numbers come with a warning. On this EURUSD feed 91% of M5 candles closed exactly at their high or low, against 7% on gold, so a wick that closes back inside is structurally rare. On M15, where three candles merge into one, the share rose to 24%. Before you use any wick pattern, look at how your own broker’s candles close.
Does price reverse after a liquidity sweep?
| Which 1 ATR came first, 48 bars | EURUSD M5 | EURUSD M15 | Gold M5 | Gold M15 |
|---|---|---|---|---|
| Reversal side first | 45.5% | 39.4% | 47.9% | 47.7% |
| Sweep side first | 51.8% | 56.0% | 51.0% | 51.8% |
| Random bar, random side first | 51.8% | 47.7% | 48.7% | 45.7% |
| Move 12 bars later, reversal side, ATR | -0.67 | +0.10 | -0.15 | -0.40 |
No. On every chart the price more often kept going the sweep way for another ATR before it reversed by one. EURUSD M15 was the clearest: 56% continuation against 39% reversal. A random bar with a coin-flip direction gave 46% to 52%, so the sweep was, if anything, a slightly worse reversal point than a random candle.
A deeper wick did not help. Sweeps that went at least 0.5 ATR past the level reversed first 45% of the time on EURUSD M5 and 49% on gold M5, about the same as all sweeps.
Can you trade the liquidity sweep?
| Profit per trade after cost | EURUSD M5 | EURUSD M15 | Gold M5 | Gold M15 |
|---|---|---|---|---|
| Fade the sweep | -2.63 (102) | -2.54 (85) | -0.50 (409) | -3.42 (122) |
| Fade, EMA 200 side only | -3.68 (30) | -4.40 (35) | -0.29 (160) | -0.08 (45) |
| Follow the sweep | -0.88 (103) | +0.45 (83) | +0.22 (407) | -0.04 (131) |
| Follow a real break | -1.23 (741) | -1.21 (216) | -0.35 (358) | -1.71 (121) |
| Random entries, median | -0.98 | -1.19 | -0.61 | -0.12 |
| Random entries, 95th percentile | -0.10 | +0.52 | +0.38 | +3.08 |
Pips on EURUSD, dollars an ounce on gold, trades in brackets. The random rows match the fade’s trade count; each run was checked against its own random range, and the breakout, with more trades, had a narrower one (95th percentile -0.65 pips on EURUSD M5).
The textbook fade lost on all four charts, and not because of the spread: before cost it still lost 1.63 pips, 1.54 pips, $0.20 and $3.12 a trade. On EURUSD M5 and gold M15 it did worse than 95% of random-entry runs. The win rate was 22% to 34% with a target twice as far as the stop.
Turning the same signal around was the best of the four trades on every chart. On EURUSD M15 it made 0.45 pips a trade and made money in both halves, but random entries with the same number of trades reached 0.56 at the 95th percentile, so it did not pass. 0 of 16 runs passed. The head-to-head with a real breakout is in liquidity sweep vs breakout.
What about the London sweep of the Asian range?
A popular version waits for London, 07:00 to 10:00 UTC, to poke the Asian session high or low and close back inside. On these charts it happened 8 to 18 times in total, too few to judge. The result was -0.80 pips and +1.03 pips a trade on EURUSD M5 and M15, +$0.95 and -$5.17 on gold M5 and M15. How the hour of the day changes results in general is in session time filter tested.
What would I do with liquidity sweeps?
- Do not read a wick as a reversal signal on its own. On all four charts the next 1 ATR went the sweep way more often than back.
- If anything, respect the sweep direction. The poke past the level was the better hint, and the fade lost even before cost.
- Know your candles. A feed where most candles close at an extreme hides most sweeps. Gold showed them 13 times a day on M5.
- Treat swing levels as places, not signals. How often swing highs and lows held as support and resistance against a random control is in support and resistance indicators tested, and the gap version of the same idea in fair value gaps tested.
How to mark swing levels in MT4 and what each swing size and wick depth did is in liquidity sweep settings for MT4.
Frequently asked questions
What is a liquidity sweep in forex?
A candle that trades past a swing high or low, where stop orders sit, and closes back inside. It is also called a stop hunt or liquidity grab.
Does price reverse after a liquidity sweep?
Not in this test. Price went 1 ATR the reversal way first only 39% to 48% of the time on EURUSD and gold, M5 and M15.
Is trading the stop hunt reversal profitable?
No. Fading the sweep lost on all four charts, even before the spread, and on two charts it did worse than 95% of random entries.
How many liquidity sweeps happen a day?
With 5-bar swings, about 13.5 a trading day on gold M5 and 4 on gold M15. EURUSD showed about 2 a day on this feed.
Is a liquidity sweep the same as a false breakout?
Yes, on one candle. A sweep is a breakout that fails by the close of the same candle. A false breakout can also fail a few candles later.
Want levels drawn by a rule, not by eye?
It is my own enhanced DeMARK Trend Line indicator for MetaTrader 4 and 5. Non repaint, clean, and free.
Forex educator and indicator developer. I build and trade my own MetaTrader tools, and share the ones that genuinely help.