IV rank and IV percentile

簡単な回答

An IV of 40% is high for a utility and unremarkable for a small biotech, so raw IV numbers need context. IV rank and IV percentile supply it by comparing today's reading with a reference set.

IV rank locates today's IV between the lowest and highest readings of the past year: 0 means at the low, 100 at the high. IV percentile asks a different question — on what fraction of past days was IV below today's level?

Options Band shows a per-stock percentile two ways: a universe percentile comparing the stock against all covered names today, and an own-history percentile against the stock's own record once enough history has accumulated.

詳しく見る

The two measures use the same inputs and can disagree. A worked example with round numbers:

• 52-week IV low: 20%
• 52-week IV high: 60%
• Today's IV30: 30%

IV rank = (30 − 20) ÷ (60 − 20) = 10 ÷ 40 = 25. On the low-to-high ruler, today sits a quarter of the way up.

Now suppose the 60% high came from a single two-week earnings spike and the stock otherwise spent the year between 22% and 28%. Then perhaps 80% of daily readings fell below 30%, making the IV percentile 80. Same day, same stock: rank 25, percentile 80. Rank is anchored to two extreme points; percentile reflects the whole distribution, which is why many analysts treat it as the sturdier of the two.

Options Band's implementation is worth knowing. The universe percentile is cross-sectional — today's IV30 ranked against every covered stock on the stocks list — and is available from day one. The own-history percentile compares a stock against its own daily IV30 records and appears once 60 daily records exist. The site began recording its own history in August 2026 and does not backfill vendor data, so own-history figures reflect only genuinely observed days.

Neither number says what IV will do next; both describe where today stands against a chosen yardstick. The yardstick — universe versus own history, one look-back window versus another — changes the answer, so the label matters as much as the number.

任意: 詳細な正式説明 確認問題には不要

IV rank = 100 × (IV_today − IV_min) ÷ (IV_max − IV_min)

IV percentile = 100 × (days with IV below IV_today) ÷ (days in window)

Rank is a min-max normalization and inherits the fragility of extremes: one anomalous print resets the denominator for a full year and compresses every subsequent reading. Percentile is a rank statistic, indifferent to how far the extremes sit from the body of the distribution, and therefore more robust to outliers — though it says nothing about magnitude: moving from the 50th to the 90th percentile may represent a single vol point in a quiet name.

Both statistics assume the reference window is a fair sample. A stock that changed character mid-window — a merger, an index inclusion, a new business line — drags stale readings into the comparison. Cross-sectional percentiles avoid staleness but mix stocks with structurally different volatility levels, which is exactly what makes them useful for screening and misleading if read as a time-series statement.

A common conflation is treating "rank 25" and "percentile 25" as interchangeable. They coincide only for particular distributions; the divergence itself describes how IV spent the year.