Squeeze breakout
XAUUSD · 1H chart · strategy s14_squeeze · 2 variations
Buys gold on the 1-hour chart when a quiet, tightly coiled period ends with price pushing up, in the direction of the daily trend.
What it is about
Markets switch between quiet phases and busy phases. The 'squeeze' (made popular by trader John Carter) measures this with two sets of bands drawn around a 20-candle average. Bollinger Bands widen and narrow with how spread out recent closing prices are; Keltner Channels are set by ATR, the average candle size. When the Bollinger Bands shrink inside the Keltner Channels, price is unusually compressed: that is the squeeze.
Compression rarely lasts. When the squeeze releases, price often starts a move that carries on for a while, because orders that piled up during the quiet spell get triggered together. The strategy joins the release only when price is above its 20-candle average and gold's daily trend is up.
It does not try to pick the top. It simply holds the trade for 20 hourly candles (a little under one trading day) and then gets out, unless the stop is hit first.
How it trades
Gold only, on 1-hour candles, at any hour gold is open (Sunday 18:00 to Friday 17:00 New York time). About 4 to 5 trades a month.
- A bit under half the trades win (46% for the default). Winners are somewhat larger than losers because trades get up to 20 hours to run.
- About 4.6 trades a month; each trade lasts at most 20 hourly candles, so usually less than a day.
- Backtest 2018-today at zero cost: +117.2R (2018-23 +56.1R, 2024 to Oct 2026 +61.1R), positive in all 9 calendar years.
- Largest drop from a peak (drawdown): about 8.3R, roughly eight full losses in a row.
- Some years are thin: 2018 made +5.0R and 2021 +3.1R.
- Buys only, built and tested while gold was in a long bull market (2018-2025). In a falling or choppy gold market it finds fewer good trades and earns less.
- The same rules faded on NAS100 in 2024-26, so this is used on gold only; it is not an edge that works everywhere.
- A trade can be open over the weekend close or big news; a price gap can fill the stop worse than planned. Overnight financing (swap) costs are not included.
- 85% of nearby settings made money in both test periods, which is encouraging, but all results are simulated.
Where it comes from
A classic volatility-breakout setup (John Carter's squeeze: Bollinger Bands inside Keltner Channels), tested in the project's October 2026 batch of documented strategies. Settings were picked on 2018-2023 (train) and checked on 2024 to Oct 2026 (test).
Main variation: s14-A ★ Main
The variations
Each variation is the same strategy with a different approach, trading its own paper account. Click one to see exactly what it changes.
The default: no trailing stop and no target. Every trade is held for 20 hourly candles unless the 2 ATR stop is hit, which gives winners room to run.
s14-BGold, 1R trailing stopSame entries, but once a trade is 1R in profit the stop trails 1R behind the best price, so many trades close early with a small profit instead of drifting back. The idea is more winners and a smoother ride.
Variations side by side
Live result of all 2 together: 0.0R (zero cost).
| Live paper account (R, zero cost) | Backtest 2018 – today (R, zero cost) | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| s14-A★ MainSqueeze breakout | XAUUSD | 1H | – | 0 | – | – | +117.2R | +0.24R | 9/9 | 4.6 | 46% | −8.3R |
| s14-BSqueeze breakout | XAUUSD | 1H | – | 0 | – | – | +81.5R | +0.17R | 9/9 | 4.6 | 51% | −13.3R |
Main = the variation with the most backtest R per R of max drawdown (2018 – today, zero cost). It is picked on that history, so the live accounts are the real test.