Source: https://forexiapro.com/gold-trading/gold-liquidity-sweep/

# Gold Liquidity Sweep: Six London XAUUSD Days and One That Kept Going

Dylan Forexia14 November 20267 min read

A gold liquidity sweep is a run past an obvious level on XAUUSD, into the orders that sit behind it, followed by a candle that closes back inside. We call it a stop hunt, and most traders call it a sweep. Over six London sessions in 2026 gold pushed 3 to 20 pips past the night's range and then ran 42 to 125 pips the other way. On a seventh day, it closed back and then went on past the level.

[![Gold liquidity sweep: a run past an obvious high that takes the stops behind it, then a close back inside, or a move that keeps going](/seo/gold-liquidity-sweep/gold-liquidity-sweep-level-sweep-close-back-diagram.webp)](/seo/gold-liquidity-sweep/gold-liquidity-sweep-level-sweep-close-back-diagram.webp)

Diagram: a sweep runs past an obvious level, takes the orders behind it, and either closes back or keeps going. The real days are below.

These days were marked up afterwards, for study. None is a signal, and a reversal seen afterwards is not a setup. On gold $1 is 1 pip, and all times are New York. The guide to [how to trade gold](/gold-trading/) has the rest, and [the stop hunt](/glossary/the-stop-hunt/) is our word for the move.

## What is a gold liquidity sweep, and which levels does gold sweep?

It is the same move as on a currency pair, only measured in dollars. Traders mark the obvious levels and put stops just past them. Price runs through, those orders fill someone else's, and a candle finishes back on the right side. [The close back](/glossary/the-close-back/) is that candle.

Two kinds of level matter on gold. The first is the edge of the night's range, the Asian Box, from the 17:00 session break to 03:00. The second is the previous day's high or low. On 25 September 2026 gold took both at once.

How often does it happen? From 6 April to 1 October 2026 we took each edge of the Asian Box and each previous-day high or low that gold traded past by at least 3% of its average day. There were 184. A candle closed back on 151 of them by 17:00. After 96 of those close backs, a later candle went past the furthest price anyway. So a close back is common, and it is not a promise.

Gold sweeps about as often as EUR/USD. Over 2021 to 2026 the median hunt past the Asian range was 7 pips on gold and 16 on EUR/USD, and the same share of the average day, 23%, on both. Gold's moves are larger in pips and about the same in proportion.

## What did six London sweeps on gold look like?

Six, in one table. "Past" is how far the extreme went beyond the level. "Ran" is the distance from that extreme to the far extreme of the move that followed.

| Day | Asian range, pips | Swept | Time | Past, pips | Ran, pips | Reversal |
| --- | --- | --- | --- | --- | --- | --- |
| Thu 14 May | 44 | High | 03:30 | 6 | 75 | Cleared the range |
| Tue 2 June | 76 | High | 03:00 | 3 | 63 | Stopped inside it |
| Wed 15 July | 41 | Low | 04:00 | 4 | 64 | Cleared the range |
| Mon 20 July | 47 | High | 07:30 | 12 | 42 | Stopped inside it |
| Fri 4 September | 27 | High | 03:45 | 4 | 125 | Cleared the range |
| Fri 25 September | 40 | High | 07:00 | 20 | 61 | Cleared the range |

Five swept the high and one the low. Four cleared the whole range and two stopped inside it. A sweep that did not clear the range is not a failure of the structure, but it is a smaller move, and it is the reason we grade them. On 2 June the push was 3 pips past a 76-pip range, and gold was back to 4478.53 within the range, 63 pips from the high of 4541.39.

## A small push, a big fall: Friday 4 September 2026

Asia left a 27-pip range, 4459.93 to 4487.02. At 03:45, in London's first hour, a candle pushed 4 pips past the high, to 4490.63, the breakout the crowd buys. It closed back inside, and gold reached 4365.31 by 09:00, 125 pips under the high and through the whole range.

The 08:30 candle carried Nonfarm Payrolls, which the calendar lists as medium impact, and a 97-pip candle printed on the release, 17.5 times the size of the candles before it. So the fall had a release in the middle of it, and the sweep came before it. We read the sweep as the stops above the box filling a seller, and no data we hold shows stop orders, so that is a reading of the structure.

[![Gold liquidity sweep in London on 4 September 2026: a push 4 pips past the Asian high to 4490.63 at 03:45, then a fall of 125 pips to 4365.31 by 09:00 with a 97-pip candle at 08:30](/seo/gold-liquidity-sweep/gold-liquidity-sweep-london-4-september-2026.webp)](/seo/gold-liquidity-sweep/gold-liquidity-sweep-london-4-september-2026.webp)

Gold, 15-minute candles, Friday 4 September 2026. The Asian range was 27 pips, 4459.93 to 4487.02. London pushed 4 pips past the high to 4490.63 at 03:45 and gold fell 125 pips to 4365.31 by 09:00. The stop-loss bands are a reading of the structure, not order-book data.

## Two levels in one: Friday 25 September 2026

Asia left a 40-pip range, 4255.93 to 4295.44, and Thursday's high sat at 4302.86. At 07:00 London pushed 20 pips past the Asian high, to 4315.55, which was also 13 pips through Thursday's high. It reversed 61 pips to 4254.40 in 3 hours, through the whole Asian range. No medium or high-impact release lined up with it. This is the day that [London open stop hunt](/forex-sessions/london-open-stop-hunt/) draws on gold.

The earlier hunt is similar. On Thursday 14 May the Asian range was 44 pips, 4668.16 to 4712.44. At 03:30, half an hour into London, a candle ran 6 pips over, to 4718.57, and closed back inside. US retail sales, a high-impact release, came at 08:30, and gold fell slowly, to 4643.97 by 16:30, 75 pips off the high.

[![Gold stop hunt in London on 14 May 2026: a 6-pip run over the Asian high to 4718.57 at 03:30, a close back inside, then 75 pips down to 4643.97 by 16:30](/seo/gold-trading/gold-stop-hunt-london-14-may-2026.webp)](/seo/gold-trading/gold-stop-hunt-london-14-may-2026.webp)

Gold, Thursday 14 May 2026. London ran 6 pips over the Asian high and closed back inside. Price fell to 4643.97 by 16:30, and the day ended at 4648.90. The stop bands are a reading of the structure, not order-book data.

## When the sweep kept going: Thursday 24 September 2026

On 24 September gold looked like a clean sweep of the low, and it was not. Asia left a 30-pip range. London broke the bottom at 04:15, to 4261.99, and pulled back to 4273.60. At 05:00 a candle ran to 4258.57 and closed at 4262.88, under a dollar back over the low.

The next candle closed at 4255.20, under it again. By 06:30 gold was at 4243.93, 18 pips past the first low. It did rally, 44 pips to 4288.32 by 12:15, and ended near 4274. So anyone who took the first close back as the end was early. Looking back, the 04:45 candle had already closed under the first low, at 4261.15, and the highs after 04:15 kept getting lower. [The liquidity sweep backtest](/stop-hunt/liquidity-sweep-backtest/) has the same day and two others.

That is the other half of the count above: of 151 close backs, 96 were followed by a later candle past the furthest price. A close back is a read, and gold gave a false one here for about 90 minutes.

## What can a gold sweep not tell you?

Quite a lot, and the days above are picked, so it is worth saying plainly.

1. **They are marked days, not every sweep.** We chose them from our marked-up set, and a sweep seen afterwards is not a setup. The 184 and 151 count is the honest base rate, and the 96 shows how often a close back was not the end.
2. **The "ran" figure was not a trade.** It is the distance from the extreme to the far extreme. No one entered at the high, and the day's low was not known at the time.
3. **A release can be the reason.** On 4 September the 08:30 payrolls candle was 97 pips. We read the sweep before it as the stops filling, and the release as the fuel, and a chart cannot separate the two.
4. **It does not make a method.** We ran our setup on gold as a fixed rule with the real spread and costs, over 3,884 trades from 2012 to 2 October 2026, and it lost between 0.089R and 0.146R a trade in every period. R is the amount risked on a trade. Our [stop hunt guide](/stop-hunt/) has the test and its limits.

From [the three seats](/glossary/the-three-seats/): retail sees a break of the night's high and buys it, with a stop just under. The market maker, as we read it, sees a pool of orders past the level. The observer sees whether a candle closes back, and that only after it has.

For the New York version of the same read, see [gold stop hunts at the New York open](/gold-trading/gold-stop-hunt-new-york-open/), and for the hours when it happens, [XAUUSD trading hours](/gold-trading/xauusd-trading-hours/). Course 3 of ForexiaPRO, [Reverse Price Action](/pro), teaches the hunt and the close back step by step.

## Tools for this

- [Pip calculator](/tools/pip-calculator/)
- [Lot size calculator](/tools/forex-lot-size-calculator/)
- [Risk-reward calculator](/tools/risk-reward-calculator/)

## Frequently asked questions

### What is a gold liquidity sweep?

A run past an obvious level on gold, an edge of the night's range or yesterday's high or low, into the orders waiting behind it, followed by a candle that closes back inside. We call it a stop hunt, and most traders call it a liquidity sweep. The close back is the read, and it is not a promise that price keeps going the other way.

### How often does gold sweep a level and come back?

From 6 April to 1 October 2026 we took each edge of the Asian Box and each previous-day high or low that gold traded past by at least 3% of its average day. There were 184. A candle closed back on 151 by 17:00, and after 96 of those a later candle went past the furthest price anyway.

### How big is a gold sweep?

Over 2021 to 2026 the median hunt past the Asian range was 7 pips on gold against 16 on EUR/USD, about 23% of the average day on both. On the six London days below the push was 3 to 20 pips past a range of 27 to 76 pips.

### Is a gold liquidity sweep a trade?

Not by itself. Run as a fixed rule over 3,884 trades from 2012 to 2 October 2026, with the real spread and costs, our setup on gold lost between 0.089R and 0.146R a trade in every period. The days below were marked afterwards and are not signals.

## More from the blog

### [Gold M and W Pattern: Sixteen Real XAUUSD Days, Two in Full](/gold-trading/gold-m-and-w-reversals/)

The gold M and W pattern on real days: sixteen marked XAUUSD days from May to September 2026 with the second leg and the reversal, and two worked in full.

### [Forex Weekly Outlook: How We Map the Week on Sunday](/weekly-cycle/forex-weekly-outlook/)

Our forex weekly outlook starts with last week's high and low, the COT stance and the calendar, not a forecast. One real Sunday and what it could not say.

### [Forex Weekly Cycle Examples: Real Weeks Read Day by Day](/weekly-cycle/forex-weekly-cycle-examples/)

Forex weekly cycle examples from real chart markups: weeks on time, a day early, a day late, partly followed and one that never turned, read day by day.