The Bollinger Band squeeze is one of the most talked-about setups in technical analysis: when the bands pinch tight, volatility is about to expand, so trade the breakout. We tested the squeeze on Nasdaq-100 futures (NQ) from June 2010 to April 2026, and put it side by side with the opposite idea, buying a close below the lower band. The squeeze barely traded. The lower-band fade did most of the work.
John Bollinger, bandwidth and the squeeze
John Bollinger developed his bands in the early 1980s and later wrote Bollinger on Bollinger Bands (2001). The standard setup is a 20-period simple moving average with bands two standard deviations above and below it. Because the bands are based on standard deviation, they widen when the market is volatile and narrow when it's quiet.
Bollinger's BandWidth indicator measures that width: (upper band − lower band) ÷ middle band. His squeeze idea is that low volatility tends to be followed by high volatility, so an unusually narrow BandWidth can come before a big move. A common way to define the squeeze, and the one Bollinger has described, is BandWidth at its lowest level in about six months.
The exact Bollinger Band strategy rules we tested
- Market: NQ continuous front-month futures, daily bars, June 7, 2010 to April 16, 2026.
- Bands: 20-day simple moving average ± 2 standard deviations of the close.
- Squeeze: BandWidth at (or within 0.01% of) its lowest reading of the past 125 trading days, about six months. This happened on 109 days in the whole test.
- Squeeze breakout long: if a squeeze occurred within the last 5 days and the close is above the upper band, buy at the close. Exit at the close when price closes back below the 20-day average.
- Squeeze breakout short: the mirror image: squeeze in the last 5 days, close below the lower band, short, cover on a close above the average.
- Lower-band fade long: no squeeze needed. Buy at the close whenever it's below the lower band. Exit on a close above the 20-day average.
- Stops: none in any variant.
- Costs: 1.0 NQ point per round trip. Results are NQ points per 1 contract; MNQ = $2/point, NQ = $20/point.
Bollinger Band squeeze results on NQ
| Variant | Trades | Win rate | Avg pts | Total pts | Profit factor | Closed-trade DD (pts) | Pts 2010–19 | Pts 2020–26 | PF 2020–26 |
|---|---|---|---|---|---|---|---|---|---|
| Squeeze breakout, long | 19 | 52.6% | 112.59 | 2,139.2 | 3.15 | -365.2 | 122.0 | 2,017.2 | 4.40 |
| Squeeze breakout, short | 10 | 10.0% | -169.97 | -1,699.8 | 0.00 | -1,699.8 | -609.8 | -1,090.0 | 0.00 |
| Lower-band fade, long | 83 | 77.1% | 139.13 | 11,547.8 | 3.20 | -1,694.0 | 3,836.2 | 7,711.5 | 2.76 |
Buy and hold over the same span: 24,677.2 points, worst decline 5,878.2 points.
The squeeze breakout: good numbers, tiny sample
The long squeeze breakout looks great on paper: 52.6% winners, a profit factor of 3.15 and 2,139.2 points. But it traded only 19 times in almost 16 years. That is not a statistically meaningful sample. One or two big winners can make the whole result, and the split proves it: just 122.0 points came in 2010–2019 and 2,017.2 in 2020–2026, with 2022 alone contributing 936.0.
The short squeeze breakout was worse and even smaller: 10 trades, one winner, -1,699.8 points. Its lone winner made only 5.0 points after costs, which is why the profit factor rounds to zero. Ten trades tells us almost nothing except that downside squeeze breaks on NQ didn't follow through in this period.
The lower-band fade: the real result
Buying closes below the lower band traded 83 times with a 77.1% win rate, a profit factor of 3.2 and 11,547.8 points ($23,096 on one MNQ contract). That's still a modest sample, but it's over 30 trades and it held up across time.
Look at the payoff, though. The average winner was 262.29 points and the average loser -275.72 points. Losers were slightly larger than winners. The strategy works because it wins about three times out of four, not because winners are big. That's classic mean reversion, and it's why a few bad trades can hurt: 2019 (-100.2) and 2020 (-439.8) were both losing years.
The robustness check: 2010–2019 vs 2020–2026
NQ rose from about 1,800 to about 26,000 during the test, so later years carry more points. The lower-band fade made 3,836.2 points in 2010–2019 and 7,711.5 in 2020–2026, with a 2020–2026 profit factor of 2.76. It was positive in both decades, and given how much cheaper NQ was before 2020, the early-decade result is arguably the more impressive one.
What this backtest does not model
- Open-trade drawdown. With no stop, trades can sit far underwater before the exit; the closed-trade drawdown understates real risk.
- Continuous-contract roll gaps, which can affect the bands and a few signals.
- Fills at the exact daily close.
- Other squeeze definitions (for example, Bollinger Bands inside Keltner Channels, or intraday charts).
- Overnight gaps, margin and taxes.
This is a historical backtest. Past results don't guarantee future results, and nothing here is financial advice.
Key takeaways
- The Bollinger Band squeeze (125-day BandWidth low) occurred on only 109 days of NQ daily data from 2010 to 2026.
- Squeeze breakout longs: 19 trades, PF 3.15. The sample is too small to be statistically meaningful.
- Squeeze breakout shorts: 10 trades, 1 winner, -1,699.8 points.
- Buying closes below the lower band: 83 trades, 77.1% winners, PF 3.2, positive in both decades.
- Average losses were slightly bigger than average wins, and there's no stop, so size carefully.
Put the research to work on your Desk
Explore the tools behind the research, compare the available access, or open your Desk if you already have an account.
FAQ
Does the Bollinger Band squeeze work on Nasdaq futures?
We can't say with confidence. On NQ daily bars from 2010 to 2026, the long squeeze breakout had a profit factor of 3.15 but only 19 trades, which is too few to be statistically meaningful. The short version lost on 10 trades.
What is Bollinger BandWidth?
BandWidth = (upper band − lower band) ÷ middle band. A squeeze is when BandWidth falls to an unusually low level. We used its lowest reading of the past 125 trading days.
Is buying the lower Bollinger Band a good strategy?
On NQ it was the best variant we tested: 83 trades, 77.1% winners, profit factor 3.2, exiting on a close above the 20-day average. It has no stop, and average losses were slightly larger than average wins.
What Bollinger Band settings did you use?
The standard 20-period simple moving average with bands at 2 standard deviations, on daily closes.
Loomi.Ai provides educational market research, not financial advice. Trading futures and options involves substantial risk of loss, and backtested results are historical and do not guarantee future returns.
Sources
- Loomi backtest: blog_strategy_tests.py + results JSON, Loomi research
- John Bollinger, Wikipedia
- Bollinger Band Squeeze, StockCharts ChartSchool