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):

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

What the bands are not

Worth settling early, because it is the costliest misunderstanding: the bands are not support or resistance. Support is a zone where real buy orders sit; a Bollinger band is a calculation redone on every candle that shifts with price itself. Price bouncing off the lower band does not mean there is demand there: it means it has moved two deviations from its mean, and that the mean is about to move towards it.

%B: where you are inside the channel

%B turns price position into a number: 0% at the lower band, 50% at the SMA20, 100% at the upper. It can exceed 100% or fall below 0% when price leaves the channel, readings that occur more often in crypto than in traditional markets.

Interpretation depends on the regime

Here is the key that decides whether the indicator helps or misleads. In a range, a %B near the extremes suggests mean reversion with reasonable reliability. In a strong trend, price can "walk the band" —%B pinned near 100%— for weeks.

Selling there on "overbought" is the classic error, exactly the one made with the RSI above 70. In those phases the band does not mark a ceiling: it marks that the trend is working. So before reading %B you have to decide which regime the market is in —and that is answered by a trend indicator, not by the bands.

Bandwidth and the squeeze: the star signal

Bandwidth measures the channel's relative width. Contraction means low volatility; expansion, high volatility. And here appears volatility's most exploitable property: it is cyclical and tends to revert. Periods of extreme calm precede violent moves, and vice versa.

BBWP: the comparable version

The squeeze occurs when Bandwidth falls to lows. The problem is that its absolute value is not comparable across assets or across different eras of the same asset. The fix is BBWP (Bollinger Band Width Percentile): the percentile of current width against its own history. A BBWP below 20% indicates volatility is more compressed than in 80% of its history. The spring is loaded.

That percentile idea deserves a guide of its own, because it is what lets you compare the volatility of two different assets: the BBWP beside the RSI and the four combinations that come out of crossing them.

What the squeeze does not say

It does not say the direction of the expansion. That has to be sought in momentum, volume and the structure of the larger trend. Nor does it say when: compression can resolve in hours or persist for weeks. The tradable signal is not the squeeze itself but its resolution —Bandwidth expansion accompanied by a breakout on volume—.

There is an alternative way to spot it: when the Bollinger Bands move inside a Keltner channel —built on ATR rather than standard deviation— the compression is especially marked.

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. Where the bands say "a move is coming", the ATR says "this big".

Its most practical use in crypto is as a confirmation margin: a break of support or resistance that fails to clear a fraction of the ATR —half, for instance— has a high probability of being a false break, ordinary market noise rather than a real move. The ATR is also the basis of Supertrend, which uses it to place dynamic support that adapts itself to current volatility.

The bands on each asset

On Bitcoin the channel behaves cleanly and weekly squeezes have preceded significant directional legs.

On Ethereum it pays to watch compression on two charts at once: the dollar chart and the ETH/BTC ratio. When the bands compress on both simultaneously, what is coming is not just a move but most likely a change of leadership between the two assets.

On XRP the squeeze carries more value than anywhere else in the market. Its pattern of months-long ranges interrupted by vertical moves makes the compression warning especially informative: on an asset that spends most of its time ranging, detecting the end of that range is worth far more than on one always in motion.

Common mistakes

• Selling every touch of the upper band in an uptrend. Walking the band is a sign of strength, not of a top.
• Reading the squeeze as directional. It warns of expansion, never of which way.
• Using the bands as rigid support and resistance. They are moving statistical levels, not supply and demand zones.
• Comparing Bandwidth across assets without normalizing. That is what BBWP is for.
• Reading %B without knowing the market regime. The same reading means opposite things in a range and in a trend.

Frequently asked questions