Standard deviation
A statistical measure of how spread out prices are from their average — the core building block behind Bollinger Bands, regression channels, and several other indicators here.
What the numbers mean
The lookback is how many bars of price feed the dispersion calculation. Twenty is the default because it matches the Bollinger Band period — the bands are literally this value multiplied and added to a moving average.
- 20 — the default, matching the Bollinger Band period
- 1 deviation — about 68% of bars in a normal distribution
- 2 deviations — about 95%
- 3 deviations — about 99.7%, in theory; markets produce these far more often than the theory allows
How it is used
Used to quantify how far price is straying from its own average, to compare current volatility with past volatility, and as the raw input to bands, channels and risk models.
- Quantifying how far price is straying from its own average
- Comparing today's volatility with the past
- The raw input to bands, channels and position-sizing models
How it relates to everything else
Standard deviation is the engine inside Bollinger Bands, regression channels, historical volatility and the Relative Volatility Index. Every one of those tools is this number dressed differently, which is why they tend to agree — and why an outlier bar inflates all of them at once for the length of the window.
- Bollinger, regression channels and HV are all built on it
- Treats upside and downside movement as equally undesirable
- Real markets have fat tails: three-deviation days are not rare
What to watch for
It treats upside and downside movement as equally undesirable, and a single outlier bar inflates it for the length of the window.