The Bollinger Band squeeze is one of the most confidently taught setups anywhere: when the bands narrow, volatility is "coiling", and a big move is coming.
We measured band width for every observation, ranked it against that same stock's own trailing 120 sessions, and asked our question.
| Band width | sample | touched +20% | vs the market |
|---|---|---|---|
| Bottom 20%. The squeeze | 8,297 / 7,014 | 15.5% / 26.0% | 0.88× / 0.90× |
| Bottom 10%, extreme squeeze | 4,769 / 4,033 | 15.5% / 26.3% | 0.88× / 0.91× |
| Top 20%, already expanded | 7,544 / 6,965 | 20.8% / 33.6% | 1.18× / 1.17× |
The squeeze measured below the market in both periods, and tightening it from the bottom 20% to the bottom 10% did not help at all, 0.88× either way. Meanwhile the stocks whose bands were already wide beat the market in both periods.
That is the exact opposite of the taught rule, on 24,113 measured observations.
The mechanism, which is straightforward
A 20% move in 30 sessions requires roughly 0.6% of movement per session, every session. A stock in a genuine volatility squeeze is, by construction, a stock that is not moving. Waiting for it to start is a bet on a change of character. Buying one already moving is a bet on continuation. Continuation is the more common event.
This is the same U-shape you met in What Makes a Stock Move 20%, quiet stocks touched +20% in 9.2% of cases against 29.0% for wild ones, arriving from a different instrument.
The Bollinger reading that does work
| %B position | sample | touched +20% | vs the market |
|---|---|---|---|
| Riding above the upper band | 2,253 / 2,658 | 23.8% / 34.8% | 1.36× / 1.21× |
| Upper half of the bands | 14,755 / 13,492 | 18.1% / 29.2% | 1.03× / 1.02× |
| Lower half | 16,949 / 13,419 | 16.2% / 27.4% | 0.92× / 0.95× |
| Below the lower band | 2,093 / 1,479 | 18.9% / 27.2% | 1.07× / 0.95× |
A stock trading outside its upper band. The classic "too extended, wait for a pullback", measured 1.36× and 1.21×. Buying the lower band measured 0.92× and 0.95×.
Bollinger Bands are useful on PSX. Just not in the direction they are taught. The band that pays is the upper one, and the squeeze pays less than nothing.
Do this now, practical activity
- Open SuperCharts and add Bollinger Bands from the indicator list.
- Find a stock whose bands are visibly narrow. Note the date and price.
- Find one whose price is riding outside the upper band. Note that too.
- Come back in two weeks and compare. One of these is a 0.88× setup and the other is a 1.36× one.
Honest limit: two stocks over two weeks is a coin toss, not evidence. Do it anyway, the value is in learning to place a chart into a measured category instead of reacting to how it looks.
Measured: stocks in the bottom 20% of their own band width touched +20% in 15.5% of 8,297 cases (0.88×) and 26.0% of 7,014 held-out cases (0.90×). Stocks riding above the upper band measured 1.36× and 1.21×.