How to read Bollinger Bands: position, width and distance

What are Bollinger Bands?

They place an upper and a lower boundary around a moving average of closing prices. In the supplied method, the middle is SMA20 and each boundary is two population standard deviations from it. This describes where price sits relative to its recent average and dispersion, not where it must go next.

Upper = SMA20 + 2 × σ. Lower = SMA20 − 2 × σ. Here σ is the population standard deviation of the same 20 closes: the square root of the average squared distance from their mean. Using a different deviation convention changes the bands.

SMA20: what it is, how it is obtained and why it matters

SMA20 is the sum of the latest 20 closing prices divided by 20. Each close receives equal weight. On this chart the inputs are four-hour closes, not daily or weekly closes. When the window advances, its oldest close drops out and a new close enters.

It supplies the middle band and a consistent reference for above- or below-average positioning. A crossing describes a change in that relationship; it does not by itself validate a trading rule. The supplied chart reads 19 candles before the visible window so that its first plotted average has all the required inputs.

Band width: measuring separation, not direction

Band width = (upper − lower) / SMA20 × 100. It expresses the distance between the bands as a percentage of the middle. Wider bands describe greater dispersion of the closes used in this calculation; narrower bands describe less dispersion.

This makes width useful for comparing the same series through time, independently of whether the last close is near the upper or lower band. A percentile places the current width within a chosen historical reference. Its interpretation depends on that period and on the eligible observations; it is not the probability that prices will rise or fall.

%B: locating the close inside or outside the bands

%B = (close − lower) / (upper − lower). Zero locates the close on the lower band; one locates it on the upper band. Values below zero or above one place it outside the interval. The reported 0.20 reading belongs to the lower half.

Use %B to describe relative location without confusing it with width: location asks where the close is, whereas width asks how far apart the boundaries are. If the boundaries coincide, the denominator is zero and this ratio is undefined.

Distance to the bands: a different denominator

Distance to a reference = (reference / close − 1) × 100. The reference can be the upper band, middle or lower band. Positive means above the close; negative means below it. This explains the signed distances supplied alongside the final prices without calculating a new market result.

Distances help compare the close with each current reference. Unlike band width, their denominator is the close. Unlike %B, each distance concerns one reference. None of these distances is an expected return, and moving bands do not remain fixed after a new candle.

A useful reading order

Identify the instrument, timeframe and data cutoff first. Then read the close relative to SMA20, its %B position, the width and its reference period. Finally distinguish wick touches from closing breaches. A price can continue along a band during a trend; touching it does not establish an imminent reversal.

The linked article supplies 180 visible candles and a separate 180-day width reference whose exact observation count is not reported. It does not supply a forward-return study. Historical similarity does not guarantee the same future outcome.

USD(S)-M perpetual: identify the source

The chart uses Binance’s BTCUSDT perpetual future, not a blend of spot markets. The existing guide explains the identified instrument, how our own system builds its prices from aggTrades, and why that provenance matters.

What BTCUSDT USD(S)-M perpetual means · How its prices are obtained · Uses and limits of the data

Application and source

Bitcoin closes in the lower half of its Bollinger Bands, with width at 4.69%.

Definitions from the RavenInvestor report dated October 23, 2026, 17:12 UTC, and its original chart. Declared data source: Binance BTCUSDT aggTrades.

Educational financial information, not personalized investment advice or a recommendation to trade. Historical observations do not guarantee future results.