IV Rank and IV Percentile: What They Measure
IV rank shows where an option's implied volatility sits between its 52-week low and high, on a scale of 0 to 100. IV percentile is the share of the past year's trading days on which implied volatility was lower than it is now, and one spike in the year can make the two disagree sharply.
IV rank tells you where implied volatility sits inside its own 52-week range, from 0 at the year’s low to 100 at the year’s high. IV percentile answers a different question: on what share of the past year’s trading days was implied volatility lower than it is today.
Both exist because a raw number like “32% implied volatility” means nothing on its own. It is high for a utility stock and low for a small biotech, so options traders compare each market with its own history instead.
How it works
IV rank uses only two past numbers, the year’s low and high.
IV rank = (current IV - 52-week low) / (52-week high - 52-week low) x 100
If implied volatility ranged from 20% to 60% over the year and sits at 30% today, the rank is (30 - 20) / (60 - 20) x 100 = 25.
IV percentile uses every day in the year.
IV percentile = days in the past 252 with IV below today’s / 252 x 100
If 189 of the last 252 readings were below today’s, the percentile is 75.
The options education site tastylive defines them this way on its explainer page, read on 26 September 2026: rank against the high and low of the last 52 weeks, percentile as the share of the last 52 weeks’ trading days, counted as 252, on which implied volatility was below today’s. Brokers differ in small details, such as whether today’s reading is inside the window, so the same stock can show slightly different values on two platforms.
Why one spike bends the rank
The rank is anchored to a single day at each end. A one-day panic sets the 52-week high, and for the next year every reading is measured against it. Implied volatility can sit well above its usual level and still show a low rank, because the scale was stretched by that one day.
The percentile does not care how high the spike went. It only counts how many days were lower. That makes it steadier after a shock, but it has its own blind spot: it says nothing about how far above or below the typical day the current reading is.
So the two disagree in a predictable way. After a spike, rank reads low and percentile reads high. After a long calm stretch with one low print, rank can read high while percentile reads low.
What traders use it for
The common use is deciding whether to buy or sell premium. A high reading is read as options being expensive against their own history, which is the case usually made for selling them through a credit spread or a short strangle. A low reading is read as options being cheap for buying.
Neither number says why volatility is where it is. Implied volatility often climbs into an earnings date and falls straight after it, the IV crush. A high rank ahead of a known event is the price of that event, not a mispricing.
Check the options chain as well. Rank and percentile are one summary number for the whole stock, while each strike and expiry has its own implied volatility, as the guide to reading an options chain shows.
A worked example
26 February 2018, the VIX. The VIX is Cboe’s index of 30-day implied volatility on S&P 500 options, so it is a clean public series for this test.
- Close that day: 15.80.
- Lowest close in the 252 sessions ending that day: 9.14, on 3 November 2017.
- Highest close in the same window: 37.32, on 5 February 2018.
IV rank = (15.80 - 9.14) / (37.32 - 9.14) x 100 = 6.66 / 28.18 x 100 = 23.6.
IV percentile: 235 of the 252 closes before that day were lower than 15.80, so the percentile is 235 / 252 x 100 = 93.3.
One measure called volatility low and the other called it nearly as high as it had been all year. The cause was the spike three weeks earlier: 2017 had been unusually calm, so 15.80 was high against almost every day of the year, but low against one day in February.
The original data
The rule. Daily VIX closes from Cboe, 2 January 1990 to 25 September 2026. IV rank on each day uses the 252 closes ending that day. IV percentile counts the 252 closes before it that were strictly lower. Both need a full year of history, so the readings start on 31 December 1990: 9,029 days. Every reading is in the VIX rank and percentile file.
Rank sits low most of the time. The median IV rank was 22.2 and the median IV percentile 44.4. IV rank was below 50 on 81.9% of days (7,395), while IV percentile was below 50 on 54.4% (4,914). Rank was under 20 on 45.9% of days; it reached 80 or more on only 4.2% (380 days). Percentile reached 80 or more on 20.7% (1,873 days).
| Reading | IV rank, share of days | IV percentile, share of days |
|---|---|---|
| 0 to under 20 | 45.9% | 29.2% |
| 20 to under 40 | 27.9% | 17.3% |
| 40 to under 60 | 14.1% | 16.3% |
| 60 to under 80 | 7.9% | 16.4% |
| 80 to 100 | 4.2% | 20.7% |
They disagree often. On 1,085 days, 12.0% of the record, IV rank was under 30 while IV percentile was over 50.
High readings did fall back more often. Across all days, the VIX was lower 21 sessions later 53.7% of the time (4,837 of 9,008). When IV rank was 80 or more it was lower 21 sessions later on 285 of 380 days, 75.0%. With IV percentile at 90 or more the share was 74.6% (850 of 1,140). Neighboring days in a spike overlap heavily, so these are fewer independent cases than the day counts suggest, and they describe the VIX, not the profit on any options trade.
Today’s readings, 25 September 2026. The VIX closed at 14.87 against a 52-week range of 13.47 (24 December 2025) to 31.05 (27 March 2026): IV rank 8.0 and IV percentile 7.5. The five Cboe single-stock volatility indexes spread much wider:
| Index (stock) | Close | IV rank | IV percentile |
|---|---|---|---|
| VXIBM (IBM) | 46.38 | 55.6 | 82.5 |
| VXGS (Goldman Sachs) | 34.84 | 40.8 | 62.3 |
| VXAZN (Amazon) | 34.27 | 30.7 | 39.3 |
| VXGOG (Alphabet) | 33.62 | 35.2 | 39.3 |
| VXAPL (Apple) | 23.67 | 29.7 | 9.5 |
On IBM the percentile ran 26.9 points above the rank; on Apple the rank ran 20.2 points above the percentile.
The topic is barely covered on video. In the 24,971-video study behind this site, 3 titles mention IV rank or IV percentile, from 3 channels, at a median of 1,551 views.
When it fails
The high rolls out of the window
The VIX’s current 52-week high, 31.05 on 27 March 2026, drops out of the window next spring. When it does, the rank will be measured against a lower high and can jump without implied volatility moving at all. Check the date of the high before reading a change in rank as a change in the market.
A high reading has a reason
Volatility is often high because something is scheduled: earnings, a drug trial result, a court ruling. The reading is then a price for a real risk, and selling it means taking that risk on.
High can go higher
The fade was common but far from universal: on a quarter of the days with an IV rank of 80 or more (95 of 380), the VIX was not lower 21 sessions later. A short-premium position sized as if the reading must fall carries that quarter in full.
Different platforms, different numbers
Platforms build their implied volatility figure in different ways, from different expiries and strikes. Two screens can show different readings for the same stock, so compare readings only within one platform.
Related
Implied volatility is the number both measures rank, and that page covers where it comes from.
IV crush explains the most common reason a high reading collapses, right after a scheduled event.
A credit spread is one of the trades a high reading is usually offered as a reason for, and that page covers its defined risk.
The volatility risk premium is the longer-run gap between implied and realized volatility that premium sellers are trying to collect.
Read IV rank and IV percentile together, and check the date of the 52-week high before trusting either one. If a single spike set the high, the rank understates how rich the options are, so I look at the percentile before calling volatility cheap.
— Michael Whitman
This page is educational, not financial advice. Test every idea on your own charts before risking money.