Bollinger Bands: volatility, %B and the squeeze

Bollinger Bands do not predict direction: they measure how far price has stretched from its mean and, above all, when volatility has compressed so much that a strong move is statistically imminent.

Building the channel

Created by John Bollinger in the 1980s, they form a statistical channel around price with the standard parameters (20, 2):

With two standard deviations, roughly 95% of price action stays inside the channel. Price touching a band is not an anomaly: it is the edge of the statistically normal. The anomaly —and the information— lies in how the channel itself behaves.

%B: where you are inside the channel

%B converts price position into a number: 0% at the lower band, 50% at the 20 SMA, 100% at the upper band. It can exceed 100% or fall below 0% when price leaves the channel —readings that occur more often in crypto than in traditional markets.

Its interpretation depends on the market regime: in a sideways range, %B near the extremes suggests mean reversion; in a strong trend, price can "walk the band" (%B pinned at 100%) for weeks —selling there for "overbought" is the classic mistake, the same one made with RSI > 70.

Bandwidth and the squeeze: the star signal

Bandwidth measures the channel's relative width. Contraction = low volatility; expansion = high. And here lies volatility's most exploitable property: it is cyclical and mean-reverting. Periods of extreme calm precede violent moves, and vice versa.

The squeeze occurs when Bandwidth drops to historical lows. A rigorous way to measure it is the BBWP (Bollinger Band Width Percentile): the percentile of current width against its own history. A BBWP below 20% means volatility is more compressed than in 80% of its history —the spring is loaded. What the squeeze does not tell you is the direction of the expansion: for that you look at momentum, volume and the structure of the larger trend.

The complement: ATR

The ATR (Average True Range, also Wilder's) measures volatility in price units: the average true range of the last 14 candles, gaps included. Its most practical use in crypto is as a confirmation margin: a support or resistance break that fails to exceed a fraction of the ATR (say, half) has a high probability of being a false break —normal market noise rather than a real move.

Common mistakes

Selling every touch of the upper band in an uptrend. Reading the squeeze as a directional signal. Using the bands as rigid support/resistance —they are moving statistical levels, not supply and demand zones. And comparing Bandwidth across assets without normalizing (which is what the BBWP is for).

Frequently asked questions