BB vs KC
Bollinger Bands vs Keltner Channels
Bollinger Bands place standard-deviation envelopes around a moving average, while Keltner Channels place ATR-based envelopes around an exponential average. Bollinger Bands react more to closing-price dispersion and outliers; Keltner Channels usually expand and contract more smoothly because ATR measures range.
What changes in practice
| Question | Bollinger Bands | Keltner Channels | Decision |
|---|---|---|---|
| Volatility measure | Standard deviation of the selected price source. | Average True Range, including gaps and intrabar range. | The bands respond to dispersion; the channels respond to traveled range. |
| Centerline | Usually a simple moving average. | Usually an exponential moving average. | Even equal envelope widths can sit around different centers. |
| Outlier response | A large close can increase standard deviation sharply and widen the envelope. | A large true range raises ATR but smoothing often creates a steadier adjustment. | Bollinger width tends to be more sensitive to abrupt dispersion. |
| Squeeze use | Often compared inside the Keltner envelope to define unusually compressed dispersion. | Provides the smoother range benchmark around the Bollinger envelope. | The classic squeeze is a relationship between both tools, not one narrow band alone. |
| Main failure mode | Shorting every upper-band touch in a trend. | Treating a channel breakout as exceptional when ATR has not caught up to a volatility shock. | Envelope contact describes location, not automatic mean reversion or continuation. |
A clean workflow
- 01
Choose the hypothesis
Use Bollinger Bands when the question concerns statistical dispersion from a mean. Use Keltner Channels when the question concerns movement relative to average traveled range.
- 02
Measure compression consistently
For a squeeze, require the Bollinger envelope to contract inside the Keltner envelope and define the settings before viewing the outcome. Eyeballing narrowness changes the rule from chart to chart.
- 03
Wait for expansion and direction
Compression predicts that volatility may expand, not which direction wins. Use structure, closing acceptance, volume or flow to decide whether a break deserves action.
- 04
Model costs after the release
The first bars after compression can carry wider spread and slippage. A historical close beyond an envelope is not the same as an achievable live fill at that close.
Common mistakes
- Assuming every Bollinger Band touch is a mean-reversion signal.
- Calling any visually narrow period a squeeze without a fixed comparative rule.
- Comparing default bands and channels while ignoring different centerlines and multipliers.
- Predicting breakout direction from compression alone instead of waiting for acceptance evidence.
Questions
Which is better for identifying a squeeze?+
The common squeeze definition uses both: Bollinger Bands contract inside Keltner Channels. That relationship compares closing-price dispersion with average true range. A narrow version of either indicator alone can still be useful, but it is a different rule and should be tested separately.
Why are Bollinger Bands wider after a large candle?+
A large deviation from the moving average increases the rolling standard deviation, especially when the lookback is short. The new outlier remains in the window for several bars, so the envelope can stay wide even after immediate price movement calms.
Do Keltner Channels use standard deviation?+
The standard construction does not. Keltner Channels use a moving average as the center and a multiple of Average True Range for the envelope. Implementations can differ in centerline and ATR smoothing, so those settings should be matched before comparing charts.
Is a close outside the bands a breakout signal?+
It can describe strong expansion, but it is not automatically a trade. In a trend, price can continue walking an outer band; in a failed break it can return immediately. Location, volume, follow-through and the planned invalidation decide which hypothesis is testable.
Updated 2026-08-28 · Educational reference, not financial advice.