IV rank vs IV percentile: two numbers that answer different questions

August 14, 2026

IV rank measures where volatility sits in its range. IV percentile measures how often it has been lower. One outlier day can push them 70 points apart, and most platforms show only one without saying which.

Raw implied volatility cannot be compared across symbols. A biotech at 60% is having a normal week. A utility at 60% is in the middle of a crisis.

So both need normalising against their own history, and there are two standard ways to do it. They are routinely used interchangeably. They are not interchangeable, and the gap between them is largest exactly when it matters most.

IV rank: position in the range

IV rank asks where current IV sits between its lowest and highest points over a lookback window, usually one year.

`` IV rank = (IV_now - IV_low) / (IV_high - IV_low) x 100 ``

An IV rank of 0 means volatility is at its cheapest of the year. 100 means the highest. 50 means exactly halfway between the two extremes.

Note what determines this number: two data points. The highest reading of the year and the lowest. Everything in between is ignored.

IV percentile: position in the distribution

IV percentile asks a different question. Of all the trading days in the lookback window, what fraction had implied volatility below where it is now?

`` IV percentile = (days with IV < IV_now) / (total days) x 100 ``

An IV percentile of 80 means volatility has been lower than this on 80% of days in the past year. This uses every observation, not just the two extremes.

Why they diverge, with a worked example

Take a stock whose implied volatility normally lives between 15 and 25, and which had one crisis day in the past year when IV spiked to 80.

Today IV is 25, at the top of its normal range.

IV rank = (25 − 15) / (80 − 15) × 100 = 15. That reads as very cheap.

IV percentile could easily be 90, because on the overwhelming majority of the past year's days, IV was below 25.

Same stock, same day, same underlying data. One number says volatility is near its cheapest, the other says it is near its dearest. Both are computed correctly.

The reason is structural. That single 80 print set the top of the range, and IV rank divides by that range forever after. One outlier permanently compresses every subsequent reading. IV percentile is barely affected, because one day out of 252 moves a percentile by about 0.4 points.

Which to use

IV percentile is the more robust measure, and for most purposes it is the better default. It uses the full distribution, it degrades gracefully when history contains a shock, and it answers the question people usually mean: is this unusual?

IV rank is more common in retail platforms, and it has one genuine advantage: it tells you about proximity to the extremes. If you specifically care whether volatility is near the top of what this name has recently done, the range is what you want.

The failure mode is not choosing wrong. It is not knowing which one you are looking at, because most interfaces label both "IV Rank."

Four traps

The lookback is a parameter, not a fact. A 1-year IV rank and a 6-month IV rank on the same symbol can differ enormously, particularly if the shock sits just inside one window and just outside the other. A value with no stated lookback is not fully specified.

The window rolls, so the number changes with no new information. When a volatility spike ages out of a one-year window, IV rank jumps upward the next day even though nothing happened. The high fell out of the denominator. This is a real and frequently misread effect: a stock whose IV rank leapt from 15 to 60 overnight may simply have forgotten last year's crisis.

Scale conventions differ. Some systems store these as fractions (0.4728), others as percentages (47.28). Both are reasonable, and mixing them silently produces values wrong by a factor of one hundred that still look like plausible numbers. If a threshold that should fire regularly never fires, or fires constantly, suspect the scale before suspecting the logic.

Neither is a signal. A high IV percentile says options are historically expensive for this name. It does not say they are overpriced. Volatility is often high for good reasons, and "sell premium because IV percentile is above 80" is a rule that will happily sell insurance into a genuine emergency.

What these numbers are actually for

They convert an incomparable number into a comparable one. That is the whole job, and it is a real one.

Once volatility is expressed relative to its own history, you can ask sensible questions: is this name's optionality unusually dear right now, is the whole sector elevated, has this moved because of an event or because the window rolled.

What they cannot do is tell you which direction anything goes, or whether today's elevated reading is justified. They locate the price in its history. Deciding whether that price is right is a separate question, and a harder one.