ICT Judas Swing on Gold: A Mechanical Version of the London False Move, Tested

The short answer

We turned the ICT Judas swing into a rule: during the London window (02:00 to 05:00 New York time), if gold trades beyond the Asian range (20:00 to midnight New York time) and an hourly bar closes back inside it, trade the reversal toward the other side of the range, with the stop beyond the false move.

• Raw version, two years of hourly data: 224 trades, 27.2% winners, −$540 per ounce after costs.
• With a midnight-open filter: 102 trades, 39.2% winners, +$40 per ounce, profit factor 1.05. That is close to break-even, not a strong edge.
• On 5-minute data (60 days): the raw version won only 4 of 28 trades; the filtered version made 13 trades and lost $7 per ounce.

What the Judas swing is

In ICT’s Power of 3 model, a trading day has three phases: accumulation (a quiet range, often the Asian session), manipulation (a false move that takes stops) and distribution (the real move). The Judas swing is the manipulation: the betrayal move that lures traders the wrong way before the day’s direction appears.

On gold, the classic picture is:

1. Asia builds a range overnight.
2. Around the London open, price runs above the Asian high, triggering breakout buys and short stops.
3. The move fails, price falls back into the range, and the London and New York sessions push lower.

The full framework, including the reference opens and the defensive “failed manipulation” rule, is taught in our Power of 3 lesson, and the wider ICT toolkit in ICT trading concepts for XAUUSD. If you want the opposite trade, our Asian range breakout lesson covers following the break.

Exact rules

All times are New York time, converted with the US daylight saving rule for each date, so the window shifts correctly in GMT when the clocks change.

How we tested it

The results

Hourly, two years

5-minute check (60 days)

Costs (filtered hourly version)

What the results say, in plain English

• Most London sweeps were not Judas swings. When gold ran beyond the Asian range and briefly closed back inside, it more often resumed in the direction of the sweep. The raw rule’s stop was hit on 156 of 224 trades.
• The midnight open did real work. Requiring the reversal to close on the “right” side of the midnight open removed 122 trades and turned −$540 into +$40. That supports the idea that the midnight open matters as context.
• But +$40 over two years is not an edge you can lean on. A profit factor of 1.05 disappears at a $0.60 cost (+$9), and the filtered version lost in year 1.
• Bigger targets did not help. A fixed 2R target lost more than targeting the far side of the Asian range.
• 5-minute data agreed. The raw version won only 4 of 28 trades over ten weeks.

Why the mechanical version struggles

The Judas swing is easy to label in hindsight: once the day is over, the false move is obvious. In real time, the same first push can be the start of a trend day. Discretionary ICT traders try to tell the difference using:

• Daily bias: only fade a London high sweep if the higher timeframe points down.
• Displacement: a sharp, energetic move back through the range, often leaving a fair value gap.
• Draw on liquidity: a clear target on the other side, such as the previous day’s low.
• Calendar: avoiding days with major US data, when London’s move may be undone at 08:30.

These may be where any edge lives. Our test shows the bare pattern is not enough. Compare the related tests of Candle Range Theory and Turtle Soup, which fade false breaks of other levels and also lost in their raw form.

When the Judas swing fails

• Trend days. London breaks the Asian high, pauses, then trends all day.
• Tiny Asian ranges. A $5 range is swept by noise; the stop and target are both small relative to costs.
• Data days. A London reversal can be wiped out by the 08:30 New York release.
• Late signals. A reversal close at 04:00 New York time leaves little room before the session’s real move has already begun.

Risk sizing

The average stop was about $12.94 per ounce (raw) and $16.23 (filtered). Risking $100 on the filtered version:

• $100 / ($16.23 + $0.30) is about 6 ounces (0.06 lots).
• The worst losing run in the filtered test was 10 trades, about $1,000 at that size.

Use the position size calculator for every setup.

Checklist before you trade a Judas swing

• Mark the Asian high and low and the midnight open before London opens, in New York time, so the levels are fixed before you see the move.
• Write your daily bias and the side you expect to be swept. If London sweeps the other side, that is information, not a trade.
• Wait for the close back inside. A wick through the Asian high that has not closed back inside the range is still a breakout.
• Use the midnight open as a filter. It was the one condition that clearly improved our results.
• Check the 8:30 New York calendar. On data days, London’s move can be reversed at the release.
• Measure the stop and the target. If the Asian range is tiny, the reward may not cover costs; if the sweep is huge, the risk may be too large.

Test it yourself

Mark the Asian range and the midnight open every day for a month. Before London opens, write down your bias and the level you expect to be swept. Then record what happened. Compare your journaled results with the mechanical numbers above: if your bias and level selection beat a 1.05 profit factor over 50 or more setups, your judgement is adding value. The strategies tested hub has the rest of the series.

Practise it with other traders

ICT concepts are best tested in the open, where hindsight cannot creep in. In the XAUUSD Global community, members post their London bias and levels before the session and review them afterwards. The first month is free, then $19.99 a month.

Educational content only, not financial advice. Trading gold with leverage carries a high risk of loss.

Frequently asked questions