# Bollinger band squeeze screener for NSE

> Why volatility contraction predicts expansion, how to measure a squeeze with band width, the NR7 alternative, and ready-made NSE squeeze scans with the breakout confirmations.

Canonical: https://patternsradar.com/blog/bollinger-band-squeeze-screener

Published 2026-08-15. Related scans: https://patternsradar.com/screener/squeeze.md, https://patternsradar.com/screener/bollinger-breakout.md, https://patternsradar.com/screener/nr7.md, https://patternsradar.com/screener/narrow-range-day.md, https://patternsradar.com/screener/range-expansion-up.md

Direction is nearly impossible to predict. Volatility is not — it is the one market property with genuinely reliable structure: **it clusters**. Quiet days bunch together, loud days follow, and the hinge between the two regimes is tradable. The Bollinger band squeeze is the best-known way to screen for that hinge.

## The mechanics

Bollinger bands sit two standard deviations either side of a 20-day average, so their **width is a direct reading of recent volatility** — wide bands after loud weeks, narrow bands after quiet ones. A squeeze is bands unusually narrow *for that stock*:

```
where bb().width < 0.08
```

Width here is normalised (band spread over the middle band), so one threshold means the same thing for a ₹80 stock and an ₹8,000 one. Under 0.08, the stock's entire recent argument spans less than 8% — an argument running out of disagreement.

The premise: a contracted range is a coiled market. Sellers who wanted out are out, buyers who wanted in are in, and the next real order flow — either direction — moves price disproportionately because there is nobody left mid-range to absorb it.

## The squeeze says nothing about direction

This is the most-ignored sentence in every squeeze tutorial. Compression predicts *expansion*, not *up*. Two ways to handle that:

**Filter for favourable ground.** The prebuilt **[Volatility squeeze](https://patternsradar.com/screener/squeeze.md)** scan requires `close > sma(200)` — squeezes resolving inside long-term uptrends break upward more often than not, because the pond is tilted.

**Or wait for the resolution.** The squeeze's *exit* is a signal of its own — price escaping the upper band with volume behind it:

```
where close > bb().upper and rel_volume > 1.5
```

That is the **[Bollinger band breakout](https://patternsradar.com/screener/bollinger-breakout.md)** scan. The trade-off is the standard one: the squeeze entry is early with ambiguity, the breakout entry is confirmed at a worse price. One caution on the breakout side — in a strong trend, price can *ride* the upper band for weeks. A close above it is the start of an expansion, not an overbought reading.

## The single-bar cousins: NR7 and the narrow-range day

The squeeze measures contraction over weeks. The same logic compresses to a single session:

- **[NR7](https://patternsradar.com/screener/nr7.md)** — today's high-low range is the narrowest of the last seven sessions. Toby Crabel's original range-contraction setup: bracket the little bar, take whichever side breaks.
- **[Narrow range day](https://patternsradar.com/screener/narrow-range-day.md)** — today's range under half the stock's own ATR, the "coil" in its per-stock-calibrated form.

And the other bookend — the loud day the quiet ones lead to — is the **[Range expansion up](https://patternsradar.com/screener/range-expansion-up.md)** scan: a session twice the stock's normal size, resolved in the buyers' favour. Expansion days set the tone for what follows more often than chance allows.

## Measure the premise

"Squeezes precede moves" is a testable claim, and the hit-rate replay on each scan page tests it: every session of the past year the scan matched, and the forward return distribution after. The interesting comparison is [the squeeze](https://patternsradar.com/screener/squeeze.md) versus [the breakout](https://patternsradar.com/screener/bollinger-breakout.md) over the same year — how much hit rate does confirmation buy, and how much entry price does it cost? That number, not anyone's opinion, is the answer to "which entry should I use." The rest of the compression-and-release family lives at the [volatility hub](https://patternsradar.com/scans/volatility.md).