John Bollinger published the bands in the 1980s and the default settings have barely changed: a 20 period simple moving average, with an upper and lower band placed two standard deviations of closing price away from it. The middle line is trend. The distance between the outer lines is volatility. When the market goes quiet the bands pull in tight, and when it moves they flare out. That is the entire mechanism, and it is worth understanding properly because almost every misuse comes from forgetting what a standard deviation of a price series actually measures.
Why the bands are not statistical boundaries
The textbook line is that two standard deviations contain roughly 95 percent of observations in a normal distribution. Price returns are not normally distributed, the sample is rolling and short, and the standard deviation is computed from the same 20 bars that set the average. So the bands do not contain 95 percent of anything in a way you can trade against. In a quiet range price will sit inside them almost all the time. In a strong trend price will spend long stretches pinned to one band while the band itself keeps moving. Both are normal outputs of the formula.
This is why "price hit the upper band, so sell" is the most expensive misreading in technical analysis. The upper band is where price is unusually high relative to its own recent behaviour, which in an uptrend describes every strong day. It says the move is extended, not that it is over. If you want that idea properly separated from the tool, read the volatility measures piece alongside this one.
The squeeze
The setup Bollinger Bands are genuinely good at is the squeeze: bandwidth contracts to a level that is low relative to its own recent history, which means realised volatility has dried up. Quiet periods do not last, and the squeeze flags that a directional move is more likely to start soon than it was a week ago. It does not tell you the direction. Anyone claiming otherwise is reading the bands for information that is not in them.
The measurement matters more than the eyeball. Bandwidth, defined as the distance between the bands divided by the middle line, turns the visual into a number you can rank against the last few months. A squeeze in the bottom decile of the last six months of bandwidth is a real observation. A squeeze that just looks narrow on your screen is a function of your chart's vertical scale.
The first breakout out of a squeeze frequently fails, particularly on lower timeframes and around session opens. A common practical rule is to wait for the second push, or to require the breakout to hold through a close rather than a wick. Neither rule is magic, and both cost you the fastest moves in exchange for skipping a chunk of the false ones.
Band walks and what they say about trend
When price closes repeatedly outside or against the upper band while the middle line rises, that is a band walk and it is a trend signal, not an exhaustion signal. Bollinger's own guidance is that tags of the upper band in an uptrend confirm strength. The practical filter is the middle line: if the 20 period average is rising and price is holding above it, upper band tags are continuation. If the middle line is flat, the same tags in a range are much closer to a mean reversion condition. This is the point where a plain moving average and the bands do the same job and the bands add the volatility context.
The %b calculation formalises the position: it expresses where price sits between the bands, so 1.0 is the upper band, 0.0 is the lower and 0.5 is the middle line. Values above 1 mean price closed outside the band. Tracking %b is more useful than tracking touches because it lets you compare today's extension with the extension on the previous swing, which is where divergence style readings come from.
Settings, and when to change them
Twenty periods and two deviations is the default for a reason: it is what the tool was designed and tested around. Shortening the average makes the bands react faster and produce more tags, which usually means more noise rather than more signal. Lengthening it smooths the middle line and widens the useful holding period. If you shorten the period, Bollinger's own suggestion was to reduce the deviation multiplier slightly, and if you lengthen it, to raise it, because the sample size changes what two deviations captures.
Timeframe changes the character more than the settings do. On a daily chart the bands describe a regime that lasts weeks. On a five minute chart they mostly describe the last hour, and session transitions distort them badly, which is worth reading against chart timeframes if you trade intraday. A squeeze on M5 at the London open is not the same object as a squeeze on D1.
What to pair them with
Bands measure extension. They do not measure participation or momentum, so they work best next to something that does. Volume or tick volume tells you whether a band break has anything behind it. A momentum reading such as the RSI tells you whether the extension is accelerating or fading. Horizontal support and resistance tells you whether the band tag is happening at a level anyone cares about, which is usually the difference between a tradeable location and a random one.
The comparison people ask about most is Keltner Channels, which use average true range around an exponential average instead of standard deviation. Keltner bands are smoother and less prone to the self-referential widening that standard deviation produces after a single large bar. Some traders run both and treat the bands sitting inside the Keltner channel as a stricter squeeze definition. That is a reasonable use of two tools that measure volatility in different ways rather than a secret.
Trading leveraged instruments is high risk and no indicator changes that. Bollinger Bands describe conditions. Position size, stop placement and the rest of your risk rules decide what those conditions cost you when the reading is wrong, and it will be wrong regularly.
"The band is a measuring stick, not a barrier. Price does not bounce off it any more than a thermometer stops the weather from getting hotter."
— Alex Onta, Executive Director, SINGUARD
Key Takeaways
- The bands are a 20 period average with two standard deviations either side, so they measure how extended price is against its own recent behaviour, not where it must stop.
- A squeeze signals that volatility has contracted and a directional move is more likely soon. It carries no information at all about which direction.
- Repeated tags of one band while the middle line trends are continuation, not exhaustion. The slope of the middle line is what separates the two readings.
- Bands measure extension only, so they need a momentum or participation reading beside them before a tag becomes a trade location.
Frequently Asked Questions
What are the best Bollinger Bands settings?
Twenty periods with a two standard deviation multiplier is the original default and remains the sensible starting point. Shorter averages generate more band tags without adding information, and any change to the period should be paired with a small adjustment to the multiplier.
Does price always return to the middle band?
No. Price returns to the middle line often in ranging conditions and can stay away from it for long stretches in a trend. Treating the middle line as a target is a range assumption that fails badly when the market is trending.
What is the difference between Bollinger Bands and Keltner Channels?
Bollinger Bands use standard deviation of closing prices around a simple moving average. Keltner Channels use average true range around an exponential moving average, which produces smoother bands that react less sharply to a single large candle.
About the Author
Alex Onta is an Executive Director at SINGUARD. He built eTrader, the terminal, the mobile apps, eTrader Broker, Copytrading, Business and Community, along with the worldwide clustered-server infrastructure it all runs on, with his brother Roman Onta helping on the design, and he leads that division today. Together with Roman he builds the Prop Firm CRM, the Broker CRM, Scalegram and CopySignals, and the two of them carry worldwide compliance, payment processing and international business structuring side by side. He lives and works in Dubai for most of the year. Meet the executive duo leading Singuard's five divisions.