The short answer
On 18 years of gold, the ICT "Power of Three" and the "Judas swing" are not an automatic daily signal: taken on every day, they went each way about equally. The edge has to come from where and when you use them, and that is something you can measure. One clear finding: when London broke the Asian range, the move kept going on 54% of days.
The idea
ICT (Inner Circle Trader) teaches that every day has three phases. In the Asian session the market builds a range. In the London session it makes a fake move out of one side of that range, to trap traders. Then the New York session runs the other way. The "Judas swing" is a similar idea: the first move after midnight is the fake one, and the rest of the day goes the opposite way. On some days the chart looks exactly like the textbook.
How we tested it
We took 18 years of gold prices (2008 to 2026, 4,844 trading days) and also Bitcoin (2018 to 2026). In New York time: Asia is 20:00 to 00:00, London 02:00 to 05:00, and New York runs until 16:00. Then we simply counted how often the day did what the model says.
What happened
On gold:
- London broke above the Asian range, and New York then fell: 48.9% of 2,026 days
- London broke below the Asian range, and New York then rose: 50.0% of 1,761 days
- Judas swing, the rest of the day went against the first move after midnight: 50.0% of 4,841 days
- For comparison: on 50.3% of all days, the rest of the day simply went up
We also checked a second version of the Judas swing: London breaks the Asian high or low, and the day ends back inside the Asian range. It happened on 46% of 4,268 days, so on 54% of those days the London breakout kept going. Useful to know if you trade London breakouts.
Bitcoin showed a similar picture: 48.2%, 52.3% and 50.1%.
What it means for you
- Don't take every Judas swing blindly. Every perfect example you have seen is real; the skill is in picking the right days. Add a filter (a killzone, the higher-timeframe direction, a key level) and test that filter
- When London broke the Asian range, the move continued more often than it turned back (54% of days). A useful hint for breakout traders
- Count the days your setup worked and the days it failed on your own chart, then keep the version that works
Our free indicator ICT Killzones Honest Stats draws the sessions and counts on your own chart how often the Judas swing and the killzones really worked: https://www.mql5.com/en/market/product/200542
Details for those who want them
- Data: XAUUSD hourly from Dukascopy, 2008 to 2026; BTC 2018 to 2026; all times New York time
- Judas swing: the move from 00:00 to 05:00 against the move from 05:00 to 16:00
- The average move in the direction the model predicts was +0.02%, -0.03% and -0.03% for the three tests, very small on their own
Have a strategy, an indicator or an EA you want tested like this? Write it in the comments.
If one of our free indicators or tools helped you, a short review on its Market page helps other traders find it.
Next week: a buy-only gold grid without a stop loss, run through 2008 to 2026.
