Bollinger Bands, created by John Bollinger in the 80s, are one of the most elegant technical indicators. They frame price with two curves following volatility, drawing a channel that expands or contracts. Read well, they tell you both where price is cheap or expensive, and when an explosive move is likely near. Read poorly, they make you take counter-trend positions in pumps.
How they're constructed
Bollinger Bands have three components:
- Middle line: an SMA 20 (simple moving average over 20 periods).
- Upper band: SMA 20 + 2 × standard deviation of last 20 closes.
- Lower band: SMA 20 − 2 × standard deviation of last 20 closes.
Standard deviation measures statistical dispersion. The more prices vary, the larger the deviation, the wider the bands. The calmer prices are, the narrower the bands.
Statistically, roughly 95% of prices should stay inside the bands (standard deviation property in normal distribution). In practice in crypto, lower — say 85-90% — because crypto markets don't follow perfect normal distribution.
The three main signals
1. The squeeze (tightening)
When bands tighten sharply (volatility very low), you're witnessing a squeeze. The market "compresses" its energy. In most cases, a squeeze is followed by an explosive move — in either direction.
The squeeze indicates the direction of the future move, just that it'll be strong. To determine direction, watch:
- The context of trend (what does EMA 200 say?);
- The candles: the first clear candle breaking out of either side indicates probable direction;
- The volume: a breakout with strong volume confirms.
Squeezes are particularly relevant on 4h and 1D. On 5m, squeezes are frequent but rarely lead to exploitable moves.
2. The walk on the band
This is counter-intuitive: in a strong trend, price can stay glued to a band for long. A powerful uptrend can see price "walking" the upper band for days. A powerful downtrend, the lower band.
Classic beginner trap: see "price touches upper band, it's overvalued, I short". Wrong. In strong trend, touching the upper band means "the trend is strong, not that it's topped".
Golden rule: never take a position against-trend just because price touches a band. It's only a reversal signal in a range-bound market.
3. Return to mean
In a range-bound market (no clear trend), price oscillates regularly between the two bands, returning toward the SMA 20 middle before heading out again. This is where mean reversion strategies on Bollinger work:
- Price touches lower band in a range → long toward the middle line or upper band.
- Price touches upper band in a range → short toward the middle line or lower band.
Absolute condition: verify you're actually in a range (ADX < 20, sideways structure) before taking these signals. In a trend, they're disastrous.
Settings
Standard is (20, 2): SMA 20, ±2 standard deviations. It works well on most TFs and markets. Some variants:
- (20, 2.5): wider bands, fewer signals but more reliable.
- (10, 2): more reactive bands, useful on short TFs.
- (50, 2): highly smoothed, for identifying structural zones on 1D.
As always, stick with standard until mastery, then experiment if you have a specific need.
Pitfalls
Pitfall #1: "price exceeds the band, so it's overvalued". False in most cases. Band overshoots are normal in trends. Don't transform that reading into a counter-trend entry signal.
Pitfall #2: ignoring trend context. Bollinger Bands don't work the same in trend and range. Always verify the regime before interpreting.
Pitfall #3: believing bands predict. Bands describe recent volatility. They don't know what happens next. A squeeze isn't a prediction, it's an observation ("volatility is low") with historical predictive value.
Pitfall #4: trade every touch. Price touches the bands very regularly. If you trade every touch, you'll make 100 trades a week — and lose on every false signal. Be selective.
Bollinger combined with other indicators
Bollinger Bands are more powerful combined than isolated. Some useful combos:
- Bands + RSI: RSI divergent when price makes a new high outside the bands = far more reliable reversal signal.
- Bands + Volume: a squeeze breakout with massive volume = high confidence. Squeeze breakout no volume = probably false.
- Bands + ADX: ADX climbing above 25 during a squeeze = the starting trend is probably exploitable.
- Bands + static levels: touching lower band at the same place as a horizontal support = confluence, far more reliable than a band touch alone.
Confluences are always more robust than individual signals — that's the general TA principle.
Concrete use case: the squeeze break
Here's a classic setup I like:
- I identify a squeeze: bands tighten sharply on the 4h of a coin (visible to the eye). I note ADX is low (< 18).
- I wait for the breakout: I don't enter during the squeeze. I wait for a clear candle closing outside one of the two bands.
- I verify context: is the breakout direction coherent with the 1D trend? Does volume accompany? If yes to both, I have a signal.
- I execute: entry on pullback toward the broken band (now support/resistance), stop opposite the squeeze, target at least the current band width after opening.
This setup doesn't work every time, but when it does, risk/reward is excellent because the entry is precise and the stop is tight.
In DYOR
DYOR displays Bollinger Bands on coin detailed view. In the Trendscanner, you can filter on:
- Price outside the bands: coins currently making a strong extension.
- Active squeeze: coins where bands are tight right now (potential breakouts coming).
Always combine with other filters (trend, volume) to find the best setups.
To go further
- ATR — another volatility measure, more raw;
- Moving Averages — the middle line of Bollinger Bands is an SMA;
- Support and Resistance — bands often act as dynamic support/resistance.