The Bollinger Squeeze: What It Predicts, and What It Does Not

A Bollinger squeeze is a period when the bands pull unusually close together because price has gone quiet. It is one of the few chart patterns with a solid idea behind it: volatility moves in cycles, and long stretches of calm tend to end in a sharp move. The catch is in that sentence. A squeeze tells you a move is likely to come. It says almost nothing about which way.

If you have not read it yet, the basics of the three lines, %b and bandwidth are in [Bollinger Bands explained](/blog/bollinger-bands-explained).

How to measure a squeeze instead of eyeballing it

Bands that look narrow on one chart look wide on another, so a squeeze is defined by comparison, not by appearance. The tool is bandwidth:

bandwidth = (upper band − lower band) / middle band

John Bollinger's original definition is simple: a squeeze is when bandwidth falls to its lowest level of roughly the last six months (about 125 daily bars). Some traders use the lowest 10% or 20% of readings over that window, which gives more signals and weaker ones.

A worked example. Gold on the daily chart, middle band at 2,400:

| Date | Upper | Lower | Bandwidth |

|---|---|---|---|

| Six weeks ago | 2,520 | 2,280 | 10.0% |

| Three weeks ago | 2,460 | 2,340 | 5.0% |

| Today | 2,436 | 2,364 | 3.0% |

If 3.0% is the lowest bandwidth in six months, the market is in a squeeze. Nothing has happened yet. That is the point.

The TTM version: bands inside Keltner Channels

A popular variant compares Bollinger Bands with Keltner Channels, which are built from average true range instead of standard deviation. When the Bollinger Bands sit inside the Keltner Channels, the market is in a squeeze; when they move back outside, the squeeze has "fired". It is the same idea with a different measuring stick: standard deviation shrinks faster than average range when closes stop moving, so the bands slipping inside the channels flags unusual calm.

Neither version is more correct. The six-month bandwidth low is stricter and rarer; the Keltner comparison fires more often.

What a squeeze does not tell you

Direction. A squeeze is a statement about volatility. Price can leave a squeeze up or down, and the bands themselves do not lean either way. Anything that tells you direction has to come from somewhere else: the trend on a higher timeframe, where the squeeze sits relative to a major level, or what volume does on the break.

Timing. A squeeze can last for days or for weeks. Entering early because "it has to break soon" means sitting in a position that pays you nothing while your stop sits inside a range that keeps getting narrower.

Size of the move. Low volatility tends to be followed by higher volatility, but "higher" can mean a modest swing, not the explosive breakout in the screenshots people share.

The head fake

Bollinger gave a name to the most frustrating part of trading squeezes: the head fake. Price leaves the squeeze in one direction, triggers everyone waiting for the breakout, then reverses and makes the real move the other way. It happens because a quiet market has liquidity bunched on both sides of the range, and the first push often exists to take it.

You cannot avoid head fakes. You can make them cheaper:

1. Wait for a close outside the range, not a touch. An intraday spike through the band means little; a daily close beyond the range high or low means more.
2. Size for the expansion. When volatility expands, a stop that fit the squeeze is suddenly too tight. Put the stop on the other side of the squeeze range and shrink the position so the money at risk stays the same. The arithmetic is in [position sizing](/learn/risk/position-sizing).
3. Decide in advance whether you take the second break. If the first break fails and price closes back inside, a break the other way is a legitimate setup, not revenge. Write that rule down before the squeeze resolves, not after.

What our own test found

In a large test of textbook conditions on the 20 largest crypto assets, on daily, weekly and 4-hour charts with doubled trading costs, we included two squeeze conditions: a Bollinger squeeze release (bandwidth rising right after a squeeze, price above the middle band) and a TTM-style squeeze turning off with price rising. The question was whether entries taken while the condition was true did better than entries taken while it was false.

Neither passed. After correcting for the number of conditions tested (306 in total), neither squeeze condition showed a reliable difference on any of the three timeframes. Of all the Bollinger-based conditions, the only one that did show a reliable effect was the plain one: price sitting in the upper half of its bands on the 4-hour chart. And even that one failed when we used it to block trades, because it removed more good entries than bad ones.

That does not make squeezes useless. It means the squeeze is not an edge on its own, at least not for trend entries in large-cap crypto. What it reliably does is warn you that the quiet part is ending, which is exactly when your stop distance and your position size need a second look.

A practical checklist

• Measure bandwidth against its own six-month history before calling anything a squeeze.
• Get direction from something other than the bands: higher timeframe trend, key levels, volume on the break.
• Wait for a close outside the range, and expect at least one head fake.
• Place the stop beyond the opposite side of the squeeze range, then reduce size to keep risk fixed.
• Judge the setup by the decision, not by whether this particular breakout ran. Over a few dozen trades, the [process](/learn/psychology/process-vs-outcome) matters more than any single result.

Frequently asked questions

How long does a Bollinger squeeze last? There is no typical duration. On daily charts squeezes often last one to several weeks, and the longest ones tend to be the most frustrating because traders enter early. Waiting for the break costs some of the move; entering early costs time and usually several small stops.

Which timeframe is best for the squeeze? It works the same way on any timeframe, but the meaning differs. A squeeze on the daily chart reflects weeks of calm and tends to resolve into a larger move. A squeeze on the 5-minute chart can simply be lunch hour.

Can you tell the direction of a squeeze breakout in advance? Not from the bands. The direction of the trend on a higher timeframe is the most common tiebreaker, and a squeeze that forms just under a major resistance level is a different situation from one in open space. Neither is a guarantee.

What is the difference between the Bollinger squeeze and the TTM squeeze? The Bollinger squeeze uses bandwidth at a six-month low. The TTM squeeze checks whether Bollinger Bands sit inside Keltner Channels and adds a momentum histogram for direction. Both measure the same thing: unusually low volatility.

Indikora is an AI-powered trading coach for crypto, forex, gold and indices. It reads your own trades, shows the volatility context behind each entry, and holds you to the stop you planned before the break. Try it free: https://indikora.com

This article is for educational purposes only and is not financial advice.