Rogers-Satchell Estimator

Rogers-Satchell Estimator plots drift-independent realized volatility: each bar contributes ln(H/C)·ln(H/O) + ln(L/C)·ln(L/O), a term that prices two-sided movement inside the bar; a candle running open-to-close in a straight line scores zero. The clean build of the Rogers-Satchell estimator averages the term over a 20-bar window, annualizes at 252 bars per year, and adds workflow: a close-to-close drift diagnostic, percentile-ranked alerts, and a dashboard.

How to Trade the Rogers-Satchell Estimator?

• Read the spread: a persistent gap between the close-to-close comparison and the Rogers-Satchell line shows how much of the reading is drift or gaps rather than two-sided movement.
• Compression: the percentile rank crossing under 20 marks quiet two-sided conditions that often precede range expansion — and because drift cancels, a market can read compressed while still grinding one way.
• Expansion: the rank crossing above 80 flags an expanded regime — a sizing input, not a directional call.

Rogers-Satchell Estimator Settings

• Window Length (default 20): bars the per-bar variance is averaged over; four log relations per bar stabilize short windows.
• Annualize (default enabled) and Bars Per Year (default 252): scaling to annualized volatility; match the constant when comparing tools.
• Close-to-Close Comparison (default enabled): the drift-sensitive baseline on identical settings.
• Percentile Lookback (default 100) with High Percentile (default 80) and Low Percentile (default 20): the rank window and alert thresholds.
• Show Percentile Levels (default disabled): dotted levels at the two thresholds.
• Show Dashboard (default enabled) with Position (default Top Right) and Size (default Small); Style inputs set colors and the Gradient Fill (default enabled).

Frequently Asked Questions

When should I use Yang-Zhang instead?

When gaps matter. This estimator sees only inside the bar, so overnight moves never enter it and gap-heavy instruments read low; the Yang-Zhang Estimator adds the overnight and the open-to-close legs around the same intraday core.

Why does a strong trend day read as quiet?

By design: pairing each extreme with both endpoints cancels drift, so the estimator prices movement around the bar's route, not the distance covered. That keeps it unbiased in trends — and the comparison line exposes it.

What do the percentile thresholds add?

Absolute volatility levels do not line up across instruments or annualization settings. Ranking the reading against its own last 100 bars lets the same 80/20 thresholds screen any instrument — what the two alerts fire on.