What a liquidity sweep is
In trading, liquidity refers to the orders available to trade against. Many orders cluster at predictable places: just above obvious highs (stop-losses of short sellers and buy-stop breakout entries) and just below obvious lows (stop-losses of buyers and sell-stop breakout entries).
A liquidity sweep is a move through one of those obvious levels that quickly reverses back inside the previous range. The orders beyond the level get triggered, but the move lacks follow-through. Traders also call this a stop hunt, a stop run or a false breakout.
Whether the sweep is “engineered” by large players, as Smart Money Concepts suggests, or simply the natural result of many orders clustering in one place, the behavior is real and visible on the chart — and gold produces plenty of it.
Where liquidity tends to sit on XAUUSD
- Previous day’s high and low — widely watched by intraday gold traders.
- The Asian-session range — frequently tested at the London open.
- Equal highs and equal lows — two or more swing points at roughly the same price look like “strong” levels and attract stops.
- Round numbers — gold often trades through whole numbers before reacting.
- Obvious trendlines — many traders place stops just beyond them.
How sweeps behave on gold (XAUUSD)
| Liquidity pool | When it’s often tested | Gold-specific note |
|---|---|---|
| Asian range high/low | London open, about 02:00–05:00 New York time | The thin Asian session builds a tight range that London often probes first. |
| Previous-day high/low | London or New York AM | Widely watched, so sweeps here tend to be sharp. |
| Equal highs/lows | Any session | Two clean touches attract many stops. The cleaner they look, the more traders notice them. |
| Round numbers (4,100, 4,150, 4,200) | Any session | Gold often trades a few dollars through a round number before reacting. |
| Previous-week high/low | Early in the week or on big data days | Rarer, larger sweeps that can set up multi-day moves. |
| Pre-news highs/lows | Around 8:30 New York data | Spikes in both directions are common. Spreads widen, and slippage can be large. |
Two habits help on gold. Check ATR 14 to judge whether a sweep is a small probe or a large spike. And remember that a stronger US dollar or rising real yields can turn a “sweep of the lows” into a genuine breakdown. Macro flow beats patterns when the two disagree.
Sweep or breakout? The key distinction
Every sweep starts out looking like a breakout. The difference only appears afterwards:
| Behavior | Likely sweep | Likely breakout |
|---|---|---|
| Close after the move | Back inside the prior range | Beyond the level |
| Follow-through | Little or none | Continues or consolidates beyond the level |
| Retest | Level becomes resistance/support against the move | Level holds from the other side |
| Structure | Short-term structure shifts back | Structure continues in breakout direction |
This is why the most important rule in sweep trading is wait for the rejection. Trying to fade a move while it’s still pushing through a level is simply betting against a breakout.
Variations and settings: from beginner to pro
| Choice | Beginner approach | What experienced traders refine |
|---|---|---|
| What counts as a sweep | Any wick through the level | A trade through the level followed by a candle close back inside on a defined timeframe |
| Confirmation | Close back inside only | Close back inside plus a market structure shift with displacement |
| Entry | Market order on the rejection candle | Pullback into the order block or fair value gap left by the shift |
| Sweep size | Ignored | Measured against ATR. Very large spikes can be breakouts that briefly pause. |
| Which levels | Any high or low | Only pre-marked pools: Asian range, previous day, equal highs/lows, round numbers |
| Extra filters | None | Time windows, inducement and SMT divergence, higher-timeframe draw on liquidity |
SMT divergence compares gold with a related market, such as silver or the US dollar index. If gold sweeps its low but silver does not make a new low, some traders read that as a sign of a failed move. It’s a useful filter to test, not a confirmation.
Three sweep setups traders use
1. Asian range sweep at the London open
| Element | Rule |
|---|---|
| Context | Asian range marked (about 20:00–00:00 New York time). Higher-timeframe bias known. |
| Trigger | In the London kill zone, price trades through one side of the range and closes back inside on M15. |
| Entry | Pullback after a short-term structure shift, often into the FVG it leaves. |
| Stop (invalidation) | Beyond the sweep extreme plus a buffer. |
| Targets | The opposite side of the Asian range, then the previous-day level. |
| Management | Partial profit at the range midpoint or opposite edge. Exit if the London window ends without progress. |
2. Previous-day high or low sweep in New York AM
| Element | Rule |
|---|---|
| Context | New York AM, about 07:00–10:00 New York time. Previous-day high and low marked. Calendar checked. |
| Trigger | Price runs the previous-day level, ideally after the 8:30 data has printed, then closes back inside on M15 or M5. |
| Entry | Retest of the swept level from the other side, or the order block left by the reversal. |
| Stop (invalidation) | Beyond the sweep wick. |
| Targets | The New York or London session low/high, then the opposite previous-day level. |
| Management | Avoid entries in the minutes around the release. Reduce size on high-impact days. |
3. Equal highs or lows with structure shift
| Element | Rule |
|---|---|
| Context | Two or more swing points at nearly the same price, visible on H1 or M15. |
| Trigger | A sweep of the equal level, then displacement that breaks the most recent opposing swing. |
| Entry | The order block or FVG created by the displacement. |
| Stop (invalidation) | Beyond the sweep extreme. |
| Targets | Opposing liquidity: the range low/high or the next equal level. |
| Management | If price returns beyond the sweep extreme, accept the breakout and stand aside. |
Step-by-step: analyzing a potential sweep
- Mark obvious liquidity in advance. Before the session, mark the previous-day high/low, the Asian range and any equal highs or lows.
- Define context. What is the higher-timeframe trend? Sweeps against the higher-timeframe trend into a major level are generally more interesting than sweeps in the middle of a range.
- Watch for the run. Price trades through the level.
- Require rejection. A close back inside the range on your execution timeframe — and, for a stricter rule, a break of the short-term swing in the opposite direction (market structure shift).
- Define the plan. Entry area (often a pullback after the rejection), invalidation beyond the sweep extreme, and an objective such as the opposite side of the range.
Entry conditions and invalidation (educational example)
On a hypothetical H1 chart, gold has formed equal highs near 3,825.4 with a range low near 3,814.6. Early in the New York session price spikes to 3,830.2, then the candle closes back at 3,822.4, inside the range. A trader using a sweep model might wait for a pullback toward the swept level, considering a short with invalidation above 3,830.2 plus a buffer (for example 3,833), and the range low around 3,814.6 as an initial objective. A sustained close back above the equal highs would show the move was a breakout, not a sweep.
Risk management
- Invalidation beyond the extreme. The sweep wick defines where the idea fails. A stop inside the wick will often be taken by a second probe.
- Second sweeps happen. Gold sometimes sweeps a level twice. Decide in advance whether you’ll re-enter, and with what risk.
- News distortion. Sweeps around US data are often fast and wide, with slippage. Many traders avoid entries in the minutes around releases.
- Size from the stop. Sweep wicks on gold can be large, so the logical stop can be wide. Reduce size accordingly — see XAUUSD risk management.
Worked example (hypothetical)
Account: $25,000. Risk per trade: 0.5% ($125). Contract: 1 lot = 100 oz, so 0.01 lot = $1 per $1 move.
Context. The H4 trend is up. The Asian range runs from 4,054.6 to 4,063.0. The previous-day high is 4,078.0.
Setup. At about 03:00 New York time (08:00 GMT in winter), gold drops to 4,049.8, below the Asian low. The M15 candle closes back inside at 4,055.2. Two candles later price breaks the last M15 lower high at 4,060.5 and leaves a small bullish FVG.
Plan. - Entry: buy limit at 4,056.5, inside the FVG. - Stop: 4,048.3, which is $1.50 below the sweep low. - Stop distance: 4,056.5 − 4,048.3 = $8.20. - Risk per 0.01 lot: $8.20. - Position size: $125 ÷ $8.20 = 15.2, rounded down to 0.15 lot. Actual risk: 15 × $8.20 = $123. - Target 1: Asian high at 4,063.0 ($6.50, about 0.8R). - Target 2: previous-day high at 4,078.0 ($21.50, about 2.6R).
Management. Close 0.05 lot at Target 1 ($32.50) and hold 0.10 lot for Target 2 ($215 if reached). Note that Target 1 is below 1R. Some traders would skip this trade unless they believe Target 2 is realistic. That judgment belongs in your written plan, not in the moment.
Invalidation. A move back below 4,049.8 means the sweep reading is wrong. These numbers illustrate the method, not a forecast, and many trades like this lose.
Combining sweeps with other concepts
- Market structure: a structure shift after the sweep is the most common confirmation.
- Order blocks and fair value gaps: give you a defined entry area after the rejection.
- Power of 3 (AMD): the manipulation phase of the daily template is, in practice, a sweep.
- Kill zones and the Silver Bullet: sweeps inside these windows are the ones ICT traders study most.
- Inducement and SMT divergence: helps separate a decoy level from the main liquidity pool.
- Asian range breakout and breakout and retest: the other side of the same coin. Knowing both helps you avoid fading every real breakout.
- Premium, discount and OTE: a refinement for where to enter after the shift.
Best timeframes and sessions
| Timeframe | Typical use on gold |
|---|---|
| Daily / H4 | Mark previous-day, previous-week and major swing liquidity. |
| H1 / M15 | Mark Asian range and equal highs/lows; judge the close back inside. |
| M5 / M1 | Time entries after the shift. Sweeps on these timeframes alone are very common and less meaningful. |
Sessions (New York time, GMT when the US is on standard time): - London kill zone, about 02:00–05:00 (07:00–10:00 GMT): the classic Asian-range sweep window. - New York AM, about 07:00–10:00 (12:00–15:00 GMT): previous-day and London-session sweeps, often around 8:30 data. - Silver Bullet windows: 03:00–04:00, 10:00–11:00 and 14:00–15:00. - Asian session, about 20:00–00:00: mainly for building the range you’ll watch later. - Avoid: the 17:00–18:00 daily break and the thin market just before the Friday 17:00 close.
Check your local times with the gold market hours tool.
Common mistakes and limitations
- Calling every break a sweep. Many breakouts are genuine. If you fade them all, you’ll be on the wrong side of every trend day.
- Entering before the rejection. “It’s going to sweep and reverse” is a prediction, not a setup.
- Labelling after the fact. Sweeps are easy to spot in hindsight. Mark the liquidity before the session to test yourself honestly.
- Ignoring the trend. A sweep of a low in a strong downtrend often leads only to a brief bounce.
- Stops inside the wick. Fix: place invalidation beyond the sweep extreme and reduce size to fit.
- Trading sweeps during the release. Fix: let the 8:30 candle finish and spreads normalize before judging the close.
- Taking trades with no room to target. Fix: check the distance to the opposite liquidity pool before entry. If the first target is well under 1R, have a clear reason to take the trade.
Limitations and honest caveats
- “Stop hunting” is a narrative. A retail chart can’t show who triggered the orders beyond a level or why. Clustered orders and fading momentum explain the same picture.
- Liquidity locations are estimates. Nobody outside the broker or exchange knows exactly where stops sit. Obvious highs and lows are educated guesses.
- Gold is decentralized. Spot XAUUSD trades over the counter, and each broker’s feed is slightly different. A “sweep” on one chart may be a near miss on another.
- Confirmation costs price. Waiting for a close back inside and a structure shift reduces false entries but often means a wider stop or a missed move.
- No verified universal success rate exists. The only useful figure is the one from your own rule-based testing, collected before you risk real money.
The limitation of any liquidity-based model is that the concept describes a common pattern but can’t tell you which instance will follow through. Clear rules, consistent invalidation and honest journaling are what turn the idea into a testable approach.
Fading a move through a level is one of the fastest ways to lose money on a trend day. Never enter a sweep trade before the rejection, keep risk small and fixed, and accept that some of your “sweeps” will turn into breakouts.
Pro tips
- Mark liquidity before the session starts. Screenshot your chart at, say, 01:30 New York time. It’s the only honest way to test yourself.
- Judge the close, not the wick. A wick through a level means little until the candle closes. Pick your confirmation timeframe and stick to it.
- Prefer sweeps into higher-timeframe levels. A sweep of the Asian low that also tags an H4 order block or demand zone has more context than one in mid-range.
- Watch the second probe. Gold often retests the swept level. A second failure with a higher low (or lower high) can be a cleaner entry than the first.
- Respect displacement. A sweep followed by slow, overlapping candles is weaker than one followed by a strong move away.
- Log failed sweeps as breakouts. Over time, your journal shows which conditions turned “sweeps” into trend days.
- Use the session clock as a filter. A sweep of the Asian low at 03:00 New York time has a different context from the same sweep at 13:30, when London has closed and liquidity is thinning. Record the time of every sweep you study.
Quick checklist
- Did I mark this liquidity pool before the session?
- What is the higher-timeframe bias and draw on liquidity?
- Has price closed back inside the level on my confirmation timeframe?
- Has short-term structure shifted with displacement?
- Is the sweep small enough relative to ATR to still look like a probe?
- Is my entry area (order block, FVG or retest) defined?
- Is my stop beyond the sweep extreme, with size calculated from it?
- Is high-impact news due during the trade?
- Am I inside a session window I’ve tested?
- Will I record this setup whether it works or not?
Frequently asked questions
What is a liquidity sweep in gold trading?
It’s a move through an obvious high or low — where many stop and breakout orders are likely clustered — that quickly reverses back inside the prior range. Traders also call it a stop hunt, stop run or false breakout.
How do I know whether it's a sweep or a real breakout?
You can’t know in advance. The difference is confirmed by what happens next: a sweep closes back inside the range and fails to hold beyond the level; a breakout holds, builds acceptance and often retests the level from the other side.
Where should the stop go on a sweep trade?
Most traders place invalidation beyond the extreme of the sweep, plus a buffer. If price returns beyond that extreme, the sweep interpretation is wrong.
Are sweeps more common at certain times?
On XAUUSD, sweeps of the Asian range and previous-day extremes are frequently discussed around the London open and the New York open, and around major data releases. They can happen at any time.
Is a liquidity sweep the same as a Judas swing?
They’re closely related. In ICT terms, a Judas swing is an early-session false move, often in the London open, that runs liquidity in one direction before the main move of the day goes the other way. Most Judas swings are sweeps, but not every sweep is a Judas swing.
What timeframe is best for trading liquidity sweeps on gold?
Many intraday traders mark liquidity on H1 or H4 and look for the rejection on M15 or M5. Higher-timeframe sweeps of daily or weekly levels are rarer but usually more meaningful. There’s no verified best timeframe, so test the one you’ll actually trade.
Sources and further reading
- CME Group — Gold futures overview — Contract specifications and trading hours for COMEX gold
- CFTC — Commitments of Traders — Weekly futures positioning, including gold
- Investor.gov — Stop order — How stop orders work, and their limits in fast markets
External sites are independent of XAUUSD Global. Check dates and data yourself.