Bollinger Bands Explained: How the Indicator Works and How to Trade It

Bollinger Bands plot a moving average with an upper and lower band based on the recent standard deviation of price, so the envelope widens when measured volatility rises and tightens when it falls. Developed by John Bollinger in the 1980s, the bands show whether price is high or low relative to its own recent range; they do not assign a probability to the next move. This page is for beginning and intermediate traders who want a working understanding of the tool, not just a definition. Read the calculation and limitations below, then test any proposed use on historical and simulated data before risking capital. John Bollinger's official overview is the primary reference for the indicator's purpose and standard construction.

The three components of Bollinger Bands

The indicator draws three lines over price:

• Middle band. A simple moving average, typically 20 periods. It represents the short-term consensus price and anchors the whole structure.
• Upper band. The middle band plus two standard deviations of price over the same lookback. It is a relative upper reference, not a price ceiling.
• Lower band. The middle band minus two standard deviations. It is a relative lower reference, not a price floor.

The familiar “about 95% within two standard deviations” rule applies to observations from a fixed normal distribution. It must not be read as a promise that 95% of prices will remain inside Bollinger Bands: market data are not normally distributed, and both the mean and the bands roll forward with every bar. A band touch therefore shows only that price has reached a relative extreme under the chosen lookback and settings. It is not, by itself, evidence of reversal or continuation.

How Bollinger Bands are calculated

The math takes three steps for each new bar:

1. Compute the 20-period simple moving average of closing prices. This is the middle band.
2. Compute the standard deviation of the same 20 closes, which measures how widely prices have scattered around that average.
3. Add and subtract two standard deviations from the middle band to plot the upper and lower bands.

A compact example: if the 20-day average is $50 and the standard deviation is $1.50, the upper band sits at $53 and the lower band at $47. If volatility doubles, the same average produces bands at $56 and $44. Nothing about the average changed; only the dispersion did. That is the property that makes the bands adaptive where fixed-percentage envelopes are static.

Settings are adjustable, but changing the lookback or multiplier changes the question the indicator answers. The conventional reference configuration is a 20-period average with bands two standard deviations away. Any alternative should be selected before testing and evaluated out of sample rather than tuned until past charts look persuasive.

How to interpret the bands

Band width and the volatility cycle

The distance between the bands is a reading of recent price dispersion. Wide bands follow larger recent moves; narrow bands follow quieter stretches. A contraction describes what has already happened. It does not establish when volatility will expand, how large the next move will be, or which direction price will take.

The squeeze

A Bollinger Band “squeeze” is the informal name for unusually narrow bands relative to a chosen history. Some traders monitor what happens after a squeeze, but the pattern is not a directional forecast and does not guarantee an imminent expansion. A valid test must define “narrow,” the entry, the comparison benchmark, transaction costs, and how false breaks are handled before results are known.

Walking the band

In strong trends, price can ride the upper band for many bars in a row. Each touch looks overbought, yet the market keeps climbing. This "walking the band" behavior is the single most important correction to the beginner's instinct that an upper-band touch means sell. In a confirmed uptrend, repeated upper-band contact is strength, not weakness.

Mean reversion signals

In a range, traders sometimes treat the outer bands as candidate extremes and the middle band as a reference point. That is a hypothesis about the market regime, not a property of the formula. A price can remain near or outside a band, and a move toward the middle band is never assured.

Trading strategies using Bollinger Bands

• Range hypothesis. Test whether reactions near an outer band add information after the range and exit rules are defined objectively.
• Compression hypothesis. Test whether a chosen band-width percentile followed by a defined close has useful results after false breaks and costs.
• Trend-context hypothesis. Evaluate whether the middle band adds information to a separately defined trend rather than assuming every pullback is an entry.
• Second-low hypothesis. Compare a second low inside the band with a first low outside it, but require independent confirmation and measure failures as carefully as successes.

Every strategy above needs an exit plan and position sizing before entry. The bands identify conditions; they do not manage risk for you. How your orders actually fill also affects results, especially on breakout entries in fast markets; the SEC's investor publication on trade execution explains why the price you click is not always the price you get.

A worked session example

Hypothetical example: a stock trades between $28 and $30 for six weeks while its band width falls to the lowest reading of the year. It then closes at $30.40 above the upper band on unusually high volume. A predefined test might enter at that close, use $29.20 as an invalidation level, and trail a rule based on the middle band. If the stock later reaches $34 and the rule exits near $33, that is one successful observation—not proof of an edge. The same test must include failed breakouts, slippage, spreads, and every signal in the sample.