Bollinger bands
A moving average with a standard-deviation envelope around it — used in this product for its WIDTH, not as a set of levels.
- Assumes:
- No other page
- Markets:
- Crypto · Equities · Indices
- Timeframes:
- 1h · 4h · 1d
What it means
Bollinger bands are three lines: a simple moving average of the closes, and a band each side of it placed a number of standard deviations away. Conventionally that is twenty bars and two standard deviations.
Because the offset is a standard deviation, the bands widen when recent closes have been scattered and narrow when they have been tightly grouped. The distance between them is a volatility measurement.
How we detect it
The engine computes the bands with the conventional twenty-bar, two-deviation settings, and uses them for exactly one thing: the bandwidth series that the compression and expansion read is built from.
It does NOT treat a touch of a band as an event. There is no detection here for price reaching the upper band, closing outside it, or walking along it.
That omission is a decision. A band touch is a statement about the last twenty closes, not about a level anyone is defending, and this product's level vocabulary comes from swing clusters instead.
When the read carries information
- There are enough closes for the moving average and the deviation to be defined at all.
- The interest is in the width. That is the only question the engine asks of these bands.
- The instrument's price series is clean; a standard deviation is easily distorted by a bad print.
When it misleads
- A standard deviation assumes a distribution that market returns do not have, so the proportion of closes falling outside the bands is not the textbook number.
- In a strong trend price can sit on a band for many bars, which readers often mistake for a level being tested.
- The bands are a lagging construction: they widen after the volatility arrives, never before it.
How it reaches you
A reader will not see Bollinger levels quoted on our surfaces. What they will see is the compression or expansion state that the bandwidth produces.
The bands are a measurement, and this publication never turns one into an instruction.
A worked example
Constructed example — not market data
A wide swing whose range narrows bar by bar until the series ends at its tightest. Run over this series the engine classifies volatility as compressed, and it reaches that word by ranking the current Bollinger bandwidth against this series' own history rather than against a fixed level — which is why the series has to be long enough to have a history to be ranked against.
Worked example
A worked example for this concept has cleared review and is shown with it.