When Bollinger Bands contract to a narrow width, volatility is low, and some traders expect a breakout. The squeeze is a popular setup, but the direction of the breakout is not predicted by the bands.
A method to test: define a squeeze as band width at its lowest point over the last 100 bars, then record the size and direction of the move over the following bars. Compare with random periods to see whether squeezes are followed by larger moves than usual.
If squeezes do lead to larger moves, a straddle approach, entering in the direction of the first break, might be worth testing. False breaks are common, so a confirmation rule, such as a close outside the band, may help.
News releases often end squeezes, so a squeeze before a scheduled event may simply reflect waiting for data.
The width threshold and lookback need to be set without too much optimisation.
Has anyone measured squeeze outcomes on intraday charts?
Bollinger Band squeezes on hourly charts
-
FXS London Scalper
- Posts: 975
- Joined: Sun Oct 18, 2026 12:25 pm
Bollinger Band squeezes on hourly charts
FXS Official Team | London Session Desk
-
Amanhasnoname
- Posts: 2060
- Joined: Fri Oct 30, 2026 9:45 am
Re: Bollinger Band squeezes on hourly charts
The key point is the one you already made: the squeeze tells you volatility is compressed, not which way it will expand. Low volatility does tend to be followed by higher volatility, so the test of whether squeezes lead to larger moves than random periods is a sensible baseline. Just make sure the comparison accounts for time of day, because on hourly charts a squeeze often forms simply because the Asian session is quiet, and the expansion that follows is just London arriving rather than anything the bands predicted.
For the direction problem, a close outside the band is a reasonable confirmation, but it costs you part of the move and still gets caught by fakeouts. Another option worth testing is waiting for the first break, then entering on a retest of the broken band or the range edge, with the stop on the other side of the squeeze range. That keeps risk defined by the structure itself rather than an arbitrary pip value.
On the news point, filtering out squeezes that end within an hour of a scheduled release would show whether the setup holds up on its own. Keep the lookback and width threshold simple, check a few nearby values to make sure results don't collapse when you shift them slightly, and run it on more than one pair before trusting the numbers.
For the direction problem, a close outside the band is a reasonable confirmation, but it costs you part of the move and still gets caught by fakeouts. Another option worth testing is waiting for the first break, then entering on a retest of the broken band or the range edge, with the stop on the other side of the squeeze range. That keeps risk defined by the structure itself rather than an arbitrary pip value.
On the news point, filtering out squeezes that end within an hour of a scheduled release would show whether the setup holds up on its own. Keep the lookback and width threshold simple, check a few nearby values to make sure results don't collapse when you shift them slightly, and run it on more than one pair before trusting the numbers.