Bollinger Bands: What They Are

What are Bollinger Bands?

Bollinger Bands were developed by John Bollinger in the 1980s and are now found in almost every charting tool. They consist of three lines: a moving average in the middle, and two bands above and below it that sit as far from the average as price volatility dictates.

The core idea is that "high" and "low" are judged relative to recent history rather than as absolute numbers. Price near the upper band is high relative to the last 20 periods; price near the lower band is relatively low. The width of the bands shows how active the market is. The crypto technical analysis guide places the indicator in the wider toolkit.

Formula: middle band SMA(20), outer bands ±2 standard deviations

The standard (20, 2) setting is calculated as follows:

Here σ is the standard deviation of the same 20 closes: how far the closes stray from their average. The indicator's creator specifies the population standard deviation (the sum of squares is divided by N, not N − 1).

When closes are scattered around the average, σ grows and the bands open; when closes cluster, σ shrinks and the bands narrow. The share of prices the bands contain is not fixed, because prices do not follow a normal distribution; the bands are a relative measure, not a probability boundary.

Worked example

Simple case: with a middle band (SMA 20) of 100 and a standard deviation of 5, the upper band is 100 + 2 × 5 = 110 and the lower band 100 − 2 × 5 = 90.

The standard deviation itself: to see the logic, take a small five-close window: 96, 98, 100, 102, 104 (illustrative numbers). The mean is 100. The squared differences from the mean are 16, 4, 0, 4 and 16; they sum to 40, divided by five gives a variance of 8, and the square root gives σ ≈ 2.83. With this window the bands are 100 ± 2 × 2.83, about 105.66 and 94.34. The real indicator runs the same steps over 20 closes.

Band width: squeeze and expansion

One of the companion measures Bollinger defined is BandWidth:

BandWidth = (Upper band − Lower band) / Middle band

In the example, (110 − 90) / 100 = 0.20.

• Squeeze: BandWidth falls to very low levels relative to its own history. Volatility has stayed low for a while. Quiet periods are usually followed by more active ones, but a squeeze does not tell you the direction; it only suggests a move may be building.
• Expansion: the bands open quickly, consistent with a strong move getting under way.

For the direction of the break, traders look at other evidence such as trading volume, MACD or breaks of horizontal support and resistance.

The %B indicator

%B expresses where price sits relative to the bands as a single number:

%B = (Close − Lower band) / (Upper band − Lower band)

• %B = 1: price is at the upper band;
• %B = 0.5: price is at the middle band;
• %B = 0: price is at the lower band.

With a close of 105 and bands at 110/90, %B = (105 − 90) / 20 = 0.75. The value can rise above 1 (price outside the upper band) or fall below 0. %B makes it easy to compare assets at very different price levels and to read price alongside oscillators such as RSI.

How to read them: a band touch is not a buy or sell signal

The most common mistake is to read a touch of the upper band as "expensive, time to sell" and a touch of the lower band as "cheap, time to buy". Bollinger's own description stresses that a band touch is not a signal by itself and needs confirmation from another indicator.

A sounder reading:

• In a strong advance, repeated touches of the upper band are consistent with trend strength, not weakness.
• If an upper-band touch comes with fading momentum, the scenario of a tiring advance gains weight.
• The middle band (SMA 20) is a reference watched during pullbacks within a trend.
• Candlestick patterns help judge the strength of a reaction at the band edges.

Bands behave differently by asset. Crypto is usually more volatile than large-cap stocks, so its bands are wider in percentage terms; the same BandWidth of 0.20 can mean a calm phase for a cryptocurrency and an active one for a stock. That is why squeezes and expansions are judged against each asset's own history.

A related envelope is the Keltner Channel: it is drawn around a 20-period exponential average at twice the average true range (ATR) instead of standard deviations. Some analysts treat Bollinger Bands moving inside the Keltner Channel as another definition of a squeeze.

Limits: walking the band in a strong trend

• Walking the band: in strong trends, price can travel along the upper (or lower) band for days. Treating each touch as a turning point means trading against the trend.
• Lag: the middle band is a moving average and follows price.
• Setting sensitivity: (20, 2) is a starting point; a shorter period makes touches more frequent. Bollinger's rules raise the multiplier to 2.1 at 50 periods and lower it to 1.9 at 10 periods. Change the setting and every reading changes.
• Technical indicators can mislead: false signals are common in volatile markets, so bands are read together with trend-following tools such as Supertrend.

Live examples

To see Bollinger (20, 2) values on live data, the Bitcoin technical analysis and support and resistance levels page lists the upper, middle and lower bands next to price; the same values appear on the Ethereum technical analysis page.

For stocks and commodities, NVIDIA technical analysis and Gold (XAU) technical analysis show the same calculation. To compare many assets on one screen, use the crypto technical summary table and the stocks and commodities technical summary.