VEGA FIELD GUIDE · 07 · VOLATILITY
IV rank and IV percentile
A 35% IV can feel lively for one stock and sleepy for another. IV rank and IV percentile give each symbol its own historical yardstick, which makes cross-sectional research much more useful.
IV rank measures position between the trailing low and high.
IV percentile measures the share of historical observations at or below the current reading.
Outliers can pull rank sharply while leaving percentile relatively stable.
Raw IV needs a home field
Implied volatility reflects the characteristic movement, event calendar, liquidity, and investor base of its underlying. A utility and a biotechnology company can carry very different normal ranges. Historical normalization lets each name compete against its own past.
Vega uses an explicit tenor such as 30-day ATM IV and an explicit history such as 252 market sessions. Changing either one changes the question. A 30-session rank is a short regime signal. A 252-session rank supplies a one-year frame.
Percentile asks how often history sat below the current observation.
IV rank: position inside the range
IV rank places the current reading between the lowest and highest values in the lookback. A reading of 0% sits at the trailing low. A reading of 100% sits at the trailing high. Values can exceed that range when a current observation sets a fresh extreme before the stored comparison window rolls forward.
Rank is intuitive and responsive. One unusually high print can widen the denominator for months, which makes future readings appear subdued. Data-quality gates are especially important around the high and low anchors.
100 x (current IV - trailing minimum IV) / (trailing maximum IV - trailing minimum IV)IV percentile: position inside the distribution
IV percentile counts how many valid historical observations were less than or equal to today’s IV. It cares about frequency. A current reading above nearly every quiet day can have a high percentile even when one crisis spike keeps its range rank moderate.
Percentile works well for statements such as “today is higher than 85% of the last year.” Ties, missing sessions, and minimum-history rules should be handled consistently. Vega publishes availability so a young listing does not masquerade as a full-history comparison.
100 x count(historical IV <= current IV) / count(valid historical IV observations)Worked example: rank 75, percentile 85
A stock’s 252-session IV30 range runs from 18% to 34%, and today’s IV30 is 30%. Its IV rank is (30 - 18) / (34 - 18) = 75%. Among 252 valid observations, 214 were at or below 30%, so its empirical percentile is 214 / 252 = 84.9%.
The percentile sits above the rank because the stock spent much of the year near the lower half of its range. A few brief high-IV sessions stretched the top of the range. Both readings are correct, and together they describe the shape of the history.
| Measure | Calculation | Result |
|---|---|---|
| IV rank | (30 - 18) / (34 - 18) | 75.0% |
| IV percentile | 214 / 252 | 84.9% |
| Interpretation | Range plus frequency | High-volatility regime |
Use rank as a first cut
- Screen IV rank and percentile together. Large disagreements often reveal an outlier-shaped history worth opening.
- Add realized volatility, variance-risk premium, skew, and term slope before forming a trade thesis.
- Filter known earnings windows when the goal is a regime comparison across ordinary sessions.
- Open the symbol history and hover over the series to identify the dates behind the trailing extremes.
- Treat a high reading as a research lead. Structure, direction, and risk limits decide the position.
Rank today’s IV across the universe
Sort and filter by IV rank, percentile, realized volatility, company group, and event window.