Key takeaways
- Both sides are small and positive. Under a point of edge on daily charts, either direction.
- The long side clears breakeven, the short side does not. And most of that gap is market drift, not the bands.
- The 4-hour sample is far larger and the edges no bigger. The familiar pattern.
- The band is a volatility measure, not a valuation. A close outside it means the move was unusual for recent conditions, nothing more.
What it does
Bollinger Bands place a 20-period simple moving average in the middle and draw two lines two standard deviations above and below it, calculated from the same 20 bars. Because the width is driven by recent volatility, the bands contract in quiet conditions and expand in violent ones.
The signals tested are closes beyond the bands: a close below the lower band read as a long, a close above the upper band read as a short. That is the textbook mean-reversion interpretation, and it assumes the band marks a stretch that tends to snap back.
How this was measured
Every figure below comes from our own tick-derived archive, not from a vendor summary. The method is identical to the one behind our indicator signal study, so the numbers here are directly comparable to every other signal we publish.
- Data. Dukascopy bid-side OHLC for 93 instruments — major and cross FX pairs, gold, silver, stock indices and crypto — from 2010 to August 2026.
- The trade. Enter at the open of the next bar. Stop at 1.5 × ATR(14), targets at 1:1 and 1:2, walking forward up to 40 bars.
- Ties go against you. A bar holding both stop and target is scored a loss every time.
- Real costs. Per-instrument median spreads from our own bid/ask measurement.
- The control. For every signal, a trade with the same direction and the same stop distance at a random bar — five draws each. The gap between the two columns is all the signal is worth.
The result
Breakeven is a 50% win rate at a 1:1 target and 33.3% at 1:2. "Matched random" is a trade in the same direction with the same stop distance, opened at a random bar.
| Signal | Target | Trades | Win rate | Matched random | Edge | Significance | Beats breakeven? |
|---|---|---|---|---|---|---|---|
| Close below lower Bollinger | 1:1 | 7,102 | 52.0% | 51.6% | +0.5 pts | not significant | Yes |
| Close above upper Bollinger | 1:1 | 39,652 | 48.4% | 48.0% | +0.4 pts | not significant | No |
| Close above upper Bollinger | 1:1 | 8,016 | 47.0% | 46.7% | +0.3 pts | not significant | No |
| Close below lower Bollinger | 1:1 | 36,813 | 49.3% | 49.1% | +0.2 pts | not significant | No |
| Close below lower Bollinger | 1:2 | 35,452 | 32.8% | 31.8% | +1.0 pts | *** | No |
| Close below lower Bollinger | 1:2 | 6,806 | 35.4% | 34.4% | +0.9 pts | not significant | Yes |
| Close above upper Bollinger | 1:2 | 7,728 | 30.0% | 29.1% | +0.9 pts | not significant | No |
| Close above upper Bollinger | 1:2 | 38,373 | 31.4% | 30.6% | +0.9 pts | *** | No |
The honest summary is that Bollinger Bands are neither the reversion machine they are sold as nor a trap. Every row sits between zero and one point of edge. Nothing here is significantly negative, and nothing is worth structuring a strategy around.
The one thing worth internalising is what the band actually is. Two standard deviations of the last twenty bars is a statement about recent volatility, not about whether price is expensive. In a strong trend price can ride the upper band for weeks, and the measured absence of a meaningful short edge is exactly what that behaviour predicts.
How to use it
- Read the bands as a volatility regime. Their width is the useful information. Narrow bands mean recent ranges were small, which is a genuine input to position sizing.
- Do not trade the touch on its own. Under a point of edge is not a trigger.
- Prefer the long side if you use it at all. It is the side that clears breakeven, though the control column shows most of that comes from market drift.
- Combine, do not isolate. Everything here is measured with no trend filter, which is how the isolated value is established and not how the tool is usually applied.
What does not work
- "Buy the lower band, sell the upper band" as a system. Both sides come in under a point.
- Reading a band touch as overextension. The band measures recent volatility. Price riding it is normal trend behaviour, not an anomaly.
- Assuming the squeeze is measured here. We tested closes beyond the bands, not band-width compression. Squeeze claims do not inherit these numbers.
Limitations
- Default parameters only. Tuning them on this same data would be curve-fitting.
- One stop rule (1.5 × ATR) and two fixed targets. Trailing stops, time exits and scaling are untested.
- Signals are taken in isolation, with no trend filter or confluence — which is how the isolated value of each is measured, but not how most people trade them.
- Bid data with modelled ask. Spread is a measured median, not a replayed tick-level ask feed.
- Spread only — no commission or swap. Both would push results further down.
- A statistically significant edge of one point is still only one point. Large samples make small effects certain; they do not make them large.
Cite or republish this data
Free to quote, screenshot, or republish — in an article, a video, a newsletter, or a course — as long as you credit FxBacktest with a link back to this page.
Frequently asked
Do Bollinger Bands work?
Weakly and symmetrically. Closing beyond a 2-standard-deviation band produced under one point of edge over a matched random entry on daily charts across 93 instruments, on either side. Nothing measured was significantly negative, and nothing was large enough to trade on its own.
Should I buy when price touches the lower Bollinger band?
Not on the strength of the touch alone. The long side of the band produced under a point of edge over a random long with the same stop distance, and much of the win rate it does show comes from the fact that markets rose over the sample period rather than from the band itself.
What do Bollinger Bands actually measure?
Volatility, not value. The bands sit two standard deviations of the last twenty bars either side of a 20-period simple moving average, so their width reflects how large recent ranges have been. A close outside a band means the move was unusual relative to recent conditions - it does not mean price is expensive or cheap.
Did you test the Bollinger squeeze?
No. We measured closes beyond the upper and lower bands only. Band-width compression, the squeeze, and any breakout rules built on it are untested here and should not inherit these numbers.
How were these Bollinger Band numbers produced?
Enter at the open of the bar after the close beyond the band, stop 1.5 x ATR(14) away, targets at 1:1 and 1:2, walking forward up to 40 bars. Bars containing both stop and target are scored as losses. Real per-instrument median spreads are applied. Every signal is compared against five random-entry draws in the same direction with the same stop distance. Data is Dukascopy bid OHLC for 93 instruments, 2010 to August 2026.