Bollinger Squeeze

How to read squeezes and breakouts

Bands and bandwidth

Bollinger Bands put a 20-period simple moving average between two lines placed two standard deviations above and below that average. These default parameters can be changed by the calculation configuration. Bandwidth, or BBW, divides the gap between the upper and lower lines by the middle line. It expresses width relative to the asset price, making readings within one asset comparable over time. Percent B locates the latest close within the two bands. A value near the upper band shows where price sits now, not where it will go. A close can also lie outside the bands.

Intensity and history

Raw BBW should not rank Bitcoin against a smaller, naturally more volatile token. Instead, SignalTrack compares the current reading with the previous 250 valid BBW values for the same asset and timeframe. Squeeze intensity equals 100 minus that historical percentile. At an intensity of 90 or above, the current bandwidth is among roughly the narrowest tenth of its own reference window. The reference window excludes the current reading so a new observation cannot change its own benchmark. A complete 20-bar band calculation and the preceding 250 readings are required before a squeeze is shown.

From squeeze to breakout

The tightening stage marks a sequence of falling band widths that is approaching the squeeze threshold. In squeeze starts when the threshold is crossed and counts consecutive closed bars. Duration is visible because a long period of compression differs from a one-bar dip. A breakout requires a prior squeeze lasting at least two bars, a bandwidth expansion of at least 25% from that squeeze minimum within three bars, and a closed price outside a Bollinger Band. This definition is deliberately stricter than a simple intrabar touch. The breakout page shows the observed up or down direction only after those conditions have appeared on a closed candle.

Keltner and records

A Keltner confirmation appears when both Bollinger Bands fit completely inside a channel based on a 20-period exponential moving average and 1.5 times average true range. Traders often call this a TTM-style squeeze. The badge is additional evidence of compression, not proof that a breakout will occur. A record badge needs complete history for its stated six or twelve month window; an all-history badge refers only to available source candles. Missing history never becomes a record claim. Liquidity and market-cap filters remove many thin markets, but they cannot eliminate spread, exchange or data-quality risk.

Directional tilt

The direction panel describes five independent clues: where the close lies inside the bands, the MACD histogram and its slope, price relative to a higher-timeframe 200-bar average, the balance of volume on rising versus falling candles, and the asset’s own earlier breakout directions. Missing clues are marked as unavailable and the label stays neutral when evidence is too sparse. These factors can disagree or reverse. A squeeze itself is directionless. Treat “up”, “down” and “neutral” as a compact reading of current conditions, never as a probability or a prediction.

Outcomes and false exits

The history cards measure returns after three, ten and twenty later closed bars when enough candles exist. They also record the largest favourable and adverse excursion in each finished window. A false breakout means a close returned inside the bands within the next three bars; it does not erase the original observed exit. An unfinished window remains pending instead of being counted as a confirmed outcome. Historical samples can be small, vary between assets, and include very different market regimes. Averages and medians describe these samples, not expected returns. These observations are educational market data, not investment advice.