Bollinger Bands on crypto: buy the breakout, not the lower band
Buying a close above the upper band made money with all 9 settings on 4-hour bars, the highest median of any indicator: +52.9% a year after costs. Buying below the lower band failed to make money with 23 of 24 settings.
How it works
Bollinger Bands draw a band two standard deviations above and below a 20-bar simple moving average. The band widens when prices swing more and narrows when they calm down.
Traders use it in two opposite ways: buying when the price closes below the lower band, expecting a return to the middle, or buying when it closes above the upper band, treating that as the start of a move.
What we tested
- Breakout: 20, 30 or 50 bars; 1.5, 2 or 2.5 standard deviations; buy on a close above the upper band, sell on a close below the middle line: 9 settings.
- Lower-band buying: 20, 30 or 50 bars; 1.5 to 3 standard deviations; buy on a close below the lower band, sell at the middle line or at the upper band: 24 settings.
Every setting was run on eight bar lengths, from 1 minute to 4 hours, on Upbit's Korean-won market: 12 large coins from January 2024 to October 2026 (from January 2025 on 5- and 10-minute bars; 8 coins from April 2026 on 1- and 3-minute bars). Signals are taken at a bar's close and filled at the next bar's open, spot buying only, with equal money in each coin. Every trade pays Upbit's 0.05% fee plus one price tick of slippage. That makes 33 single-indicator settings, and 3,715 more combinations where the best of them were given a filter or an exit rule. Full method.
Results by setting and bar length
Each cell is the yearly return after costs over the whole period. An outlined cell passed the two-period test: at least +10% a year in both the first and the second period, with the second keeping at least half of the first.
Return per year after costs, %
Sea level (0) is break-even after costs: blue is under water, green to brown is land.
Bollinger breakout
Settings above water after costs: 9 of 9 on 4-hour bars, 4 of 9 on 1-hour bars, and none on 30-minute bars or shorter. Passed the two-period test: 1 on 1-hour bars, 6 on 4-hour bars.
| Setting | 1m | 3m | 5m | 10m | 15m | 30m | 1h | 4h |
|---|---|---|---|---|---|---|---|---|
| BB(20, 1.5σ) buy above upper band, sell below middle | −100.0 | −100.0 | −100.0 | −99.9 | −97.9 | −83.9 | −43.7 | +36.9 |
| BB(20, 2σ) buy above upper band, sell below middle | −100.0 | −100.0 | −100.0 | −99.1 | −93.6 | −71.5 | −24.4 | +43.7 |
| BB(20, 2.5σ) buy above upper band, sell below middle | −100.0 | −99.6 | −98.4 | −89.5 | −76.5 | −47.3 | +0.5 | +50.9 |
| BB(30, 1.5σ) buy above upper band, sell below middle | −100.0 | −100.0 | −100.0 | −99.4 | −94.2 | −68.7 | −27.2 | +57.2 |
| BB(30, 2σ) buy above upper band, sell below middle | −100.0 | −100.0 | −99.9 | −97.3 | −87.0 | −51.3 | −7.2 | +56.4 |
| BB(30, 2.5σ) buy above upper band, sell below middle | −100.0 | −99.7 | −97.9 | −87.6 | −71.4 | −32.4 | +6.2 | +57.5 |
| BB(50, 1.5σ) buy above upper band, sell below middle | −100.0 | −100.0 | −99.9 | −96.9 | −82.2 | −46.3 | −0.3 | +52.9 |
| BB(50, 2σ) buy above upper band, sell below middle | −100.0 | −100.0 | −99.2 | −91.0 | −68.5 | −29.1 | +15.6 | +49.6 |
| BB(50, 2.5σ) buy above upper band, sell below middle | −100.0 | −99.4 | −95.3 | −78.0 | −52.0 | −10.5 | +20.6 | +53.1 |
Bollinger mean reversion
Settings above water after costs: 1 of 24 on 4-hour bars, 1 of 24 on 1-hour bars, and none on 30-minute bars or shorter. None passed the two-period test.
| Setting | 1m | 3m | 5m | 10m | 15m | 30m | 1h | 4h |
|---|---|---|---|---|---|---|---|---|
| BB(20, 1.5σ) buy below lower band, sell at middle band | −100.0 | −100.0 | −100.0 | −99.4 | −92.4 | −72.6 | −53.1 | −37.5 |
| BB(20, 1.5σ) buy below lower band, sell at upper band | −100.0 | −100.0 | −99.6 | −94.1 | −76.8 | −55.2 | −43.2 | −36.2 |
| BB(20, 2σ) buy below lower band, sell at middle band | −100.0 | −100.0 | −99.9 | −95.6 | −79.4 | −53.5 | −36.2 | −27.6 |
| BB(20, 2σ) buy below lower band, sell at upper band | −100.0 | −99.6 | −96.5 | −83.2 | −56.7 | −38.0 | −28.9 | −29.2 |
| BB(20, 2.5σ) buy below lower band, sell at middle band | −100.0 | −98.8 | −93.4 | −73.2 | −45.4 | −30.6 | −19.4 | −9.3 |
| BB(20, 2.5σ) buy below lower band, sell at upper band | −99.6 | −80.0 | −77.5 | −59.2 | −26.8 | −17.6 | −21.2 | −19.3 |
| BB(20, 3σ) buy below lower band, sell at middle band | −98.0 | −63.3 | −48.7 | −34.2 | −13.5 | −10.3 | +1.1 | −1.5 |
| BB(20, 3σ) buy below lower band, sell at upper band | −61.5 | −6.8 | −37.3 | −34.9 | −16.7 | −1.5 | −12.4 | −27.0 |
| BB(30, 1.5σ) buy below lower band, sell at middle band | −100.0 | −100.0 | −100.0 | −97.5 | −84.1 | −59.3 | −43.5 | −33.6 |
| BB(30, 1.5σ) buy below lower band, sell at upper band | −100.0 | −99.8 | −98.1 | −88.9 | −65.9 | −47.3 | −35.8 | −31.3 |
| BB(30, 2σ) buy below lower band, sell at middle band | −100.0 | −100.0 | −99.2 | −90.4 | −68.3 | −44.8 | −33.0 | −26.4 |
| BB(30, 2σ) buy below lower band, sell at upper band | −100.0 | −97.7 | −91.4 | −76.9 | −49.0 | −32.7 | −33.7 | −24.7 |
| BB(30, 2.5σ) buy below lower band, sell at middle band | −100.0 | −98.6 | −91.4 | −70.0 | −42.4 | −29.3 | −15.3 | −14.4 |
| BB(30, 2.5σ) buy below lower band, sell at upper band | −99.4 | −76.5 | −72.3 | −60.7 | −30.6 | −21.1 | −23.9 | −12.0 |
| BB(30, 3σ) buy below lower band, sell at middle band | −98.1 | −64.9 | −57.5 | −37.8 | −12.8 | −9.5 | −2.7 | −1.4 |
| BB(30, 3σ) buy below lower band, sell at upper band | −64.4 | −14.1 | −43.4 | −45.3 | −18.2 | −12.4 | −11.9 | −21.3 |
| BB(50, 1.5σ) buy below lower band, sell at middle band | −100.0 | −100.0 | −99.3 | −91.5 | −70.0 | −51.4 | −39.8 | −27.3 |
| BB(50, 1.5σ) buy below lower band, sell at upper band | −100.0 | −97.8 | −92.0 | −79.0 | −51.6 | −37.6 | −35.9 | −25.4 |
| BB(50, 2σ) buy below lower band, sell at middle band | −100.0 | −99.6 | −95.9 | −80.8 | −57.0 | −37.1 | −32.1 | −19.6 |
| BB(50, 2σ) buy below lower band, sell at upper band | −99.9 | −86.5 | −82.3 | −70.1 | −42.5 | −29.6 | −31.7 | −21.4 |
| BB(50, 2.5σ) buy below lower band, sell at middle band | −100.0 | −95.9 | −84.9 | −61.8 | −38.0 | −24.1 | −21.1 | −10.3 |
| BB(50, 2.5σ) buy below lower band, sell at upper band | −97.9 | −70.2 | −65.0 | −54.7 | −27.0 | −26.1 | −23.0 | −19.7 |
| BB(50, 3σ) buy below lower band, sell at middle band | −97.1 | −65.0 | −59.2 | −36.3 | −24.0 | −7.2 | −3.7 | +0.5 |
| BB(50, 3σ) buy below lower band, sell at upper band | −51.3 | −25.2 | −47.9 | −41.2 | −21.8 | −12.6 | −1.5 | −10.4 |
Adding filters and exit rules
On each bar length, the settings with the best first-period results were combined with up to two filters (trend on a longer chart, ADX, volume, RSI range, Bitcoin's trend) or one filter and an exit rule (stop-loss, trailing stop, take-profit, time limit). Combinations were only built where this indicator ranked among the leaders in the first period.
Bollinger breakout
| Bars | Combinations that traded | Passed | Best that passed | Return a year | Worst fall | Second period |
|---|---|---|---|---|---|---|
| 1-hour | 579 | 155 | BB(50, 2.5σ) buy above upper band, sell below middle + 4-hour chart above EMA 200 + ADX(14) above 20 | +29.5% | −25.4% | +31.6% |
| 4-hour | 522 | 290 | BB(30, 2.5σ) buy above upper band, sell below middle + close above EMA 200 + stop 2% | +64.3% | −11.3% | +57.4% |
Bollinger mean reversion
| Bars | Combinations that traded | Passed | Best that passed | Return a year | Worst fall | Second period |
|---|---|---|---|---|---|---|
| 1-minute | 81 | 0 | None passed. Best overall: 0.0% a year | |||
| 3-minute | 468 | 0 | None passed. Best overall: +0.3% a year | |||
| 5-minute | 504 | 0 | None passed. Best overall: +3.6% a year | |||
| 10-minute | 519 | 0 | None passed. Best overall: +2.9% a year | |||
| 15-minute | 539 | 0 | None passed. Best overall: +18.2% a year | |||
| 30-minute | 503 | 0 | None passed. Best overall: +20.3% a year | |||
What we learned
- The breakout was the strongest single-indicator rule we found on 4-hour bars. BB(30, 2.5σ) buy above upper band, sell below middle made +57.5% a year after costs, with a worst fall of −20.2%, 628 trades and a 39% win rate. 6 of the 9 settings passed the two-period test, so it is not one lucky number.
- Adding a trend filter and a 2% stop gave the best result of the whole study: BB(30, 2.5σ) buy above upper band, sell below middle + close above EMA 200 + stop 2%, +64.3% a year, worst fall −11.3%, +57.4% in the second period. Why we still do not trust it.
- Breakouts need slow bars. On 1-hour bars 4 of 9 settings made money; on 30-minute bars none did.
- Buying the lower band lost on every bar length. It won about 56% of 4-hour trades, but the losing trades were bigger than the winning ones.
- A nine-year replay from 2017 of BB(30, 2.5σ) with a 2% stop on 4-hour bars made +98% a year with the fee alone and +86% with slippage, with a worst fall of 30 to 33% and one losing year, 2022.
Questions
Do Bollinger Bands work for crypto?
As a breakout signal on 4-hour bars, yes in our backtests: all 9 settings made money after costs. As a buy-the-lower-band signal, no: on 4-hour bars the median was −21.3% a year.
What are the best Bollinger Band settings for 4-hour crypto charts?
30 bars with 2 to 2.5 standard deviations did best (BB(30, 2σ) +56.4%, BB(30, 2.5σ) +57.5%), but every breakout setting from 20 to 50 bars made between +36.9% and +57.5% a year.
Why does buying the lower band fail?
In a falling market the price can keep closing below the lower band, and the rule keeps buying the decline. The wins are small bounces back to the middle line; the losses are the moves that keep going.
Every setup with this indicator in the explorer: 4h 1h 30m 15m