Is This IV High? Rank, Percentile and Skew
Drag a marker along a year of IV and watch rank and percentile disagree, sometimes badly.
Worth reading first: Implied Volatility Is a Price, Not a Prediction
Course catalogue99 lessons
The last lesson finished on a question it could not answer: is 34% implied volatility high? Not without context. This lesson gives you two kinds of context, one across time and one across strikes.
IV rank: where you sit in the year's range
IV rank = (current IV − 52-week low) ÷ (52-week high − 52-week low) × 100A position between two endpoints.
If a stock's implied volatility has ranged from 20% to 60% over the past year and sits at 30% today, its IV rank is 25. You are in the bottom quarter of the range, so premium is cheap by this stock's own standards.
The weakness is right there in the formula. It depends on two numbers, the high and the low. One panic day a year ago can set a high that distorts every reading since.
IV percentile: how many days were lower
IV percentile = days in the past year with IV below today ÷ total days × 100A count, not a range.
Same stock, same day, different question. If 180 of the last 252 trading days had lower implied volatility than today, the IV percentile is 71. Today is higher than 71% of the year, wherever the extremes happened to be.
Because it counts days instead of measuring a span, one outlier barely moves it. That makes it steadier, and it is why a lot of traders prefer it.
Where the two measures disagree
A year of daily implied volatility with one spike in it. Drag today's IV and watch rank and percentile give different answers about the same day.
52-week range: 20.6% to 62.3%
One spike in the year stretches the high, which drags IV rank down for every subsequent day. While IV percentile, which only counts days, barely notices. That is why the two figures can point in opposite directions on the same chain.
| Today's IV | IV rank | IV percentile | What it means |
|---|---|---|---|
| 22% | 11 | 18 | Both agree: cheap |
| 32% | 31 | 82 | Sharp disagreement |
| 45% | 53 | 95 | Rank understates it |
A stock that spent most of the year near 25% IV, with one spike to 70%.
The middle row is worth a moment. IV rank says 31, which sounds unremarkable and maybe a little cheap. IV percentile says 82: today is dearer than four days in five. Both are correct from the same data. The spike stretched the range, which drags down every rank reading afterwards, while the percentile simply counts.
Skew: context across strikes
Now the second kind of context. A stock does not have one implied volatility. It has one per strike, and they are not equal.
The volatility curve across strikes
Implied volatility plotted against strike price. Flatten the skew to zero to see the symmetric smile that textbook models assume, then raise it to see what equity chains actually look like.
0 = flat smile, 1 = steep put skew
Downside strikes carry more implied volatility than upside ones. Crashes are faster and more correlated than rallies, and demand for downside protection is persistent, so the market charges more for it. Flatten the skew to zero and the curve becomes the symmetric smile the textbook model assumes.
Downside puts consistently carry higher implied volatility than calls the same distance away. That is skew, and on equities it is close to universal. There are three reasons for it, and all of them are real.
- Crashes are faster than rallies. Markets fall harder than they rise, so big downside moves really are more likely than a symmetric model assumes.
- Protection has steady buyers. Funds hedge downside all the time, and constant demand lifts the price.
- Correlation rises in selloffs. Everything falls together, which makes index downside particularly expensive.
The shape is a volatility smile when it curves up on both sides, and a skew or smirk when the put side is clearly higher, which is what most equities look like.
What skew means if you sell puts
Something quietly good: you are selling the expensive side of the curve. A put seller supplies exactly what all that hedging demand wants, and gets paid above the at-the-money level for it.
Something else worth knowing: skew is why a 0.20 delta put pays noticeably more than a 0.20 delta call on the same stock. If the put side of the chain has ever looked more generous to you, this is why. It is payment for a genuinely fatter tail, not an inefficiency.
What can go wrong
Using IV rank alone after a volatility spike. It will read low for a year and tell you premium is cheap when it is not.
Selling on high IV rank alone. High readings gather around real trouble. Treat it as one input, not a trigger.
Expecting a flat volatility curve. Any model or tool that assumes one is mispricing the wings.
Reading steep skew as an opportunity. It steepens most when the market is most worried, which is when the extra premium is most deserved.
Key takeaways
- IV rank measures where today sits between the year's low and high. IV percentile counts how many days were lower.
- One spike distorts IV rank for a year afterwards, while IV percentile barely notices. Prefer percentile, and watch when they disagree.
- Implied volatility differs by strike. Downside puts carry more than calls the same distance away, and that is skew.
- Skew exists because crashes are faster than rallies and protection has steady buyers. It is payment, not inefficiency.
- Put sellers are selling the expensive side of the curve, which is a real structural advantage.
Check your understanding
1. A stock ranged 20%-60% IV this year and sits at 30% today. What is its IV rank?
2. IV rank reads 31 and IV percentile reads 82 on the same day. What does that tell you?
3. Why does a 0.20 delta put usually pay more than a 0.20 delta call on the same stock?
Related lessons
Track what you have worked through. No account needed. View transcript
This browser only