Volatility Estimators

The Volatility Estimators indicator computes a chosen statistical volatility estimate over a lookback window and plots it as a single percentage line. A method selector covers five published estimators — Close-to-Close (the default), Zero-Trend, Parkinson, Garman-Klass, and Rogers-Satchell — with the range-based members drawing on each bar's high, low, open, and close rather than closes alone. Each is implemented in its standard published form over a default 10-bar window, with an annualize toggle scaling the per-bar figure by the square root of the periods per year.

How to Trade the Volatility Estimators?

• Rising line: realized volatility is expanding — wider stops, smaller sizes, likelier breakout follow-through.
• Falling line: volatility compressing — a quieting regime favoring mean-reversion tactics and tighter targets.
• Low readings vs the instrument's own history: the compression squeeze traders watch — quiet and expansive phases alternate.
• Estimator disagreement: a close-only estimate vs a range-based one (via a second instance) shows whether movement comes from close-to-close jumps or intrabar range.

A gauge, not a signal generator: it says how much the market is moving, never in which direction.

Volatility Estimators Settings

• Method (default Close-to-Close): the estimator plotted. Close-to-Close and Zero-Trend need only closes; Parkinson, Garman-Klass, and Rogers-Satchell fold in the bar's range, stabilizing on fewer bars at the cost of assuming no opening gaps.
• Length (default 10): bars in the estimation window; longer smooths, shorter reacts quickly but jitters.
• Annualize (default on): scales the per-bar estimate by the square root of the periods per year set in the adjacent field (default 252, the daily-bar convention); off plots raw per-bar values.

Frequently Asked Questions

Which volatility estimator should I use?

For gap-prone instruments such as daily equities, close-based estimates are more honest — range-based formulas assume away opening gaps and understate. For near-continuous markets, they give smoother readings from fewer bars. The default carries the fewest assumptions.

What does the Zero-Trend method do differently?

It is the close-to-close calculation without subtracting the mean return — squared log returns sum directly, assuming zero drift over the window. In trending stretches it reads higher, because trend movement counts as volatility instead of being absorbed into the mean.

How do these estimators differ from ATR?

ATR averages the bar's true range in price units, an absolute measure. These measure the dispersion of log returns in percent, optionally annualized — comparable across instruments and relatable to options-style volatility figures.