Bollinger Bands Explained: Time Entries With Volatility

Bollinger Bands are a volatility indicator made of a 20-period moving average with an upper and lower band drawn two standard deviations away. The bands widen when price swings grow and narrow when the stock goes quiet, so they show whether a market is coiling for a move or already stretched. They do not predict direction on their own. This article is educational only, not financial advice.

Key takeaways

• Bollinger Bands use a 20-period simple moving average with bands set 2 standard deviations above and below it, and the bands expand and contract with volatility.
• A squeeze, when band width falls to a multi-month low, signals compression but not direction, so you need a trigger such as a close outside the band on strong volume.
• Touching the upper or lower band is not a trade signal, because strong trends can ride a band for weeks.
• Mean-reversion trades work best in range-bound stocks, with a reversal candle that closes back inside the band and the middle band as the target.
• Pair the bands with volume, a momentum indicator and a defined stop, and expect the setup to fail some of the time.

How are Bollinger Bands calculated?

Bollinger Bands consist of three lines, built by John Bollinger in the 1980s:

1. Middle band: a 20-period simple moving average (20 days on a daily chart).
2. Upper band: the middle band plus 2 standard deviations of price over the same 20 periods.
3. Lower band: the middle band minus 2 standard deviations.

Upper = SMA(20) + 2 x standard deviation
Lower = SMA(20) - 2 x standard deviation
Band width = (Upper - Lower) / Middle

Standard deviation measures how far closing prices have scattered around their average. Quiet trading means a small number and tight bands. Wild swings mean a large number and wide bands. That is why the bands are best read as a volatility gauge first and a support-and-resistance tool second.

If prices were perfectly normally distributed, about 95% of closes would fall inside a 2-standard-deviation band. Real markets have fatter tails, so treat 95% as a rough guide only. The point is that a close outside the band is relatively unusual and worth attention.

The default (20, 2) suits daily swing charts. Shorter windows react faster but give more false signals, and longer windows smooth more but lag. Start with the default before changing anything.

What is a Bollinger Band squeeze?

A Bollinger Band squeeze is a period when the bands contract to their narrowest in months, showing volatility has dried up. Volatility tends to cycle between quiet and active phases, so a tight range often precedes a larger move. The squeeze tells you that energy is stored, not which way it will go.

How to spot one:

• Band width sits at or near its lowest reading of the past 3 to 6 months.
• Price is moving in a tight range, often inside a flag, triangle or flat base.
• Volume is drifting lower (see volume dry-up).

To trade it, wait for a directional trigger: a daily close outside the band, on volume of at least 1.5x the 50-day average, ideally with momentum confirming. A related setup is covered in the Bollinger Band squeeze breakout strategy, and Keltner Channels offer another way to define a squeeze.

Worked example: squeeze breakout

These numbers are hypothetical. Stock XYZ trades between $48 and $52 for three weeks and band width falls to its lowest level this year. On day 22 it closes at $53.40, above the upper band, on volume 2.5x average.

• Entry: $53.40 on the close, or the next session if it holds above the band.
• Stop: $50.90, below the middle of the range and the middle band. Risk is $2.50 per share.
• Target: the height of the range ($4) projected from the breakout, about $57.40. That is $4.00 of reward for $2.50 of risk, or 1.6:1.

A 1.6:1 payoff is thin, so many traders take partial profit at the measured target and trail the rest, or skip trades with less than 2:1. If the stock closes back inside the range, the breakout has failed, and the stop is there to limit the damage.

Can you trade mean reversion with Bollinger Bands?

Yes, but only in the right conditions, because price tends to drift back toward the middle band and yet a band touch alone is not a signal. In a strong uptrend, a stock can hug the upper band for weeks, and shorting each touch is a reliable way to lose money.

A higher-quality mean-reversion long has these features:

• The bands are roughly flat, meaning the stock is range-bound and not trending hard.
• Price tags or slightly pierces the lower band, then prints a reversal candle such as a hammer or bullish engulfing.
• The candle closes back inside the band.
• Volume does not show panic selling that continues.

Worked example: lower-band bounce

Hypothetical stock ABC has moved sideways for two months, with the middle band at $36. It closes at $31.20 at the lower band. The next day it prints a hammer with a low of $31.10 and closes at $32.50, inside the band.

• Entry: $32.80, above the hammer's high.
• Stop: $30.90, below the hammer's low. Risk is $1.90.
• Target: the middle band at $36. Reward is $3.20, or about 1.7:1.

That is acceptable but not great. Skip the trade if the middle band is much closer than the stop distance, since reward-to-risk below about 1.5:1 rarely survives losing streaks.

How do band slope and width show trend strength?

When both bands slope up and price rides the upper half, the trend is strong, and pullbacks to the middle band are potential buy zones. When both slope down with price in the lower half, the trend is weak, and rallies to the middle band are where sellers tend to appear. Flattening bands suggest the trend is fading into a range.

Band width adds a second read:

How do Bollinger Bands differ from Keltner Channels and ATR stops?

Bollinger Bands measure volatility with standard deviation of closing prices, so one large candle widens them immediately. Keltner Channels use the average true range (ATR) around a moving average, which responds more smoothly. Many traders define a squeeze as the moment Bollinger Bands move inside the Keltner Channels, which is a stricter test than band width alone.

This matters for stops too. Because bands expand after a big move, a stop placed at the opposite band can be far away just when volatility is highest. An ATR-based stop scales to the stock's normal daily range instead. See ATR trailing stops for one way to do it.

A practical rule is to use the bands for context and the trigger, and use ATR or structure for the stop. For example, if the band-based stop would risk 8% of the entry price, the trade probably needs a smaller size or should be skipped, because a stop that wide will not hit a good reward-to-risk ratio on a typical setup.

Common Bollinger Band mistakes

• Treating every band touch as a signal. Always require a confirming candle and context.
• Fading a strong trend. Mean-reversion trades fight the tape when the bands are sloping steeply.
• Entering a squeeze early. Without a trigger, you are guessing the direction.
• Ignoring events. Earnings and news can push price far outside the bands in a single candle, and those moves are not clean technical setups.
• Skipping the stop. A band can be walked for weeks. Define the exit before you enter.
• Using the bands alone. Combine them with volume, a momentum read such as RSI (see the 2-period RSI strategy) and the broader trend.

Quick checklist

Squeeze breakout: band width near a multi-month low, tight range, close outside the band on volume 1.5x or more, stop below the base.

Mean reversion: flat bands, outer-band touch, reversal candle that closes back inside, target at the middle band, stop beyond the wick.

Trend riding: both bands sloping the same way, price along one band, and a stop below the middle band for longs.

The bottom line

Bollinger Bands turn volatility into something you can see: tight bands mean compression, wide bands mean the move is underway, and a touch of the outer band is only a hint. Use the squeeze for breakout setups, use mean reversion only in ranges, and always set a stop and size the position for a loss. No indicator works every time, past performance does not guarantee future results, and you should do your own research.

StockSetups scans the US-equities universe after the close, detects chart patterns from trendline geometry, and sorts long setups into four lanes, including Setting up and Breaking out. Paid plans add trade plans with entry, stop, target and a conviction score, so you can check a squeeze candidate against volume and trend quickly.

For stop and target placement, see average true range (ATR).