Bollinger Bands and Keltner Channels look similar on a chart. Both create a channel around a moving average. Both expand during volatile periods and contract during quiet ones. But they measure fundamentally different things, and the relationship between them reveals one of the most reliable patterns in technical analysis: the volatility squeeze.
How They Differ
Bollinger Bands use standard deviation to set the channel width. The upper band is the moving average plus a multiple of the standard deviation of closing prices. The lower band is the average minus the same multiple. Because standard deviation responds to directional moves (not just range), Bollinger Bands expand rapidly during trends.
Keltner Channels use Average True Range (ATR) for channel width. The upper band is the EMA plus a multiple of ATR. The lower band is the EMA minus ATR. Because ATR measures the average range of each candle (high minus low, adjusted for gaps), Keltner Channels respond to volatility more smoothly. They expand during volatile candle ranges, not necessarily during directional moves.
The practical difference: during a trending move with low intra-bar volatility (price moving steadily in one direction without large wicks), Bollinger Bands expand (because standard deviation increases) while Keltner Channels may not (because individual candle ranges are normal). During a ranging market with high intra-bar volatility (price whipsawing within a range), Keltner Channels expand (large candle ranges) while Bollinger Bands may contract (standard deviation of closes is low).
The Squeeze
A squeeze occurs when Bollinger Bands contract inside Keltner Channels. Specifically: the lower Bollinger Band is above the lower Keltner Channel and the upper Bollinger Band is below the upper Keltner Channel. This means price volatility (measured by standard deviation) has compressed below the recent range volatility (measured by ATR).
Squeezes indicate that the market is coiling. Price is moving in an increasingly narrow range, building energy for a directional breakout. The squeeze does not predict the direction of the breakout, but it reliably identifies that a breakout is imminent.
Our volatility squeeze strategy detects when the squeeze fires (Bollinger Bands release from inside Keltner Channels). The direction is determined by a momentum oscillator: positive momentum at the release indicates a bullish breakout, negative momentum indicates bearish.
Our Implementation
The strategy uses Bollinger Bands with length 20 and standard deviation 2.0 alongside Keltner Channels with length 20 and scalar 1.5 (times ATR). The Keltner scalar of 1.5 is deliberately narrower than the Bollinger standard deviation of 2.0, which means the Bollinger Bands are usually wider than the Keltner Channels under normal conditions. A squeeze occurs when this relationship inverts: volatility compresses enough that the Bollinger Bands fit inside the normally-narrower Keltner Channels.
A 12-period momentum oscillator determines breakout direction. When the squeeze fires (was squeezed on the previous bar, released on the current bar) and momentum is positive, the strategy generates a long signal with base confidence 0.65. If the 4-hour ADX is above 20 (indicating a developing trend on the higher timeframe), confidence increases by 0.15 to 0.80.
Stop-loss is set at 2.0 times ATR and take-profit at 3.0 times ATR. The wider stops account for the fact that breakouts from squeezes often produce volatile initial moves with significant pullbacks before continuing in the breakout direction.
Why Squeeze Matters in Crypto
Crypto markets spend roughly 60 to 70 percent of their time in ranging conditions and 30 to 40 percent in trending conditions. The transition from ranging to trending is where the largest moves originate. The squeeze identifies these transition points by detecting when volatility has compressed to levels that historically precede expansion.
In our backtest data, squeeze-fired entries have a higher hit rate than entries generated purely by price-based indicators. The squeeze adds a timing dimension: it does not just identify direction (which RSI or MACD can do) but identifies when a directional move is likely to begin. This timing advantage is most valuable in crypto because the transition from low to high volatility happens faster and more dramatically than in traditional markets.
Keltner Channels as a Standalone Tool
Beyond the squeeze, Keltner Channels work well as breakout indicators. Our keltner_breakout strategy generates signals when price closes above the upper channel or below the lower channel. ADX above 20 adds confidence that the breakout is trend-driven rather than noise. Volume above 1.5 times the 20-period average provides additional confirmation.
The base confidence is 0.45 for a channel break alone, increasing to 0.60 with ADX confirmation, 0.70 with volume, and up to 0.80 with higher timeframe trend alignment. This layered confidence system means the strategy is most aggressive when multiple conditions align and most conservative when only the price break is present.
Keltner Channels have an advantage over Bollinger Bands for breakout strategies because the ATR-based width is smoother. Bollinger Bands can produce false squeeze signals when a single large candle temporarily spikes the standard deviation. Keltner Channels smooth over individual candle outliers because ATR is an average of ranges, not a point-in-time calculation.
Choosing Between Them
For mean reversion strategies, Bollinger Bands are superior because the standard deviation bands directly measure how unusual the current price is relative to recent prices. Our Bollinger Band mean reversion strategy produces Sharpe ratios of 9 to 19 precisely because the bands capture genuine statistical deviations.
For breakout strategies, Keltner Channels are superior because the ATR-based bands provide smoother levels that are less susceptible to whipsaws from individual volatile candles.
For detecting regime transitions, the combination (squeeze) is superior to either alone because it captures the specific condition where trend-based volatility has compressed below range-based volatility, a reliable precursor to directional breakouts.
Our platform supports all three approaches. The mean reversion strategy uses Bollinger Bands. The keltner breakout strategy uses Keltner Channels. The volatility squeeze strategy uses both. Each tool serves its specific purpose within the portfolio.