IV rank vs IV percentile
By Faizan Ahmad. Updated .
IV rank tells you where today's implied volatility sits between its lowest and highest level of the past year, on a 0 to 100 scale. IV percentile tells you the percentage of days in that year when implied volatility was lower than today. Both answer the same question, "are this stock's options cheap or expensive compared with its own history?", but IV rank can be thrown off by a single spike, while IV percentile counts every day.
What are the formulas?
IV rank = (current IV - 1-year low IV) ÷ (1-year high IV - 1-year low IV) × 100
IV percentile = days in the past year with IV below today's ÷ total days × 100
Both use a year of daily implied volatility readings. Platforms differ on which IV they track (a 30-day reading, or the nearest at-the-money options), which is why two sites can show different numbers for the same stock on the same day.
A worked example: when rank and percentile disagree
Illustrative numbers for a hypothetical stock with 252 trading days of history. Most of the year its IV sat between 25% and 45%, but one surprise sent it to 85% for a week. Today it is 40%, and 189 of the 252 days had IV below 40%.
- IV rank = (40 - 25) ÷ (85 - 25) × 100 = 25
- IV percentile = 189 ÷ 252 × 100 = 75
Rank says options are cheap. Percentile says they are more expensive than on three days out of four. Percentile is closer to the truth here: that one week at 85% stretched the range so far that 40% looks low on a scale that the rest of the year never came near.
Now remove the spike. Same stock, same 40% today, but the year's high was 45%. IV rank becomes (40 - 25) ÷ (45 - 25) × 100 = 75, and the percentile stays 75. Without an outlier the two measures agree; with one, rank gets distorted.
IV rank vs IV percentile: which should you use?
| IV rank | IV percentile | |
|---|---|---|
| Uses | Only the year's high and low | Every day in the year |
| Effect of one spike | Large: it stretches the range | Small: it's one week out of 52 |
| A reading of 50 means | Halfway between the low and the high | Half the days had lower IV |
| Easy to calculate by hand | Yes, needs two numbers | Needs the full daily history |
| Best for | Stocks with a smooth volatility history | Stocks with past spikes: earnings gaps, crashes, buyouts |
If you look at only one, look at percentile. If the two disagree by a lot, there is almost always a spike in the past year, and it's worth finding out what caused it.
Why isn't raw IV enough?
Because every stock has its own normal. On October 30, 2026, Tradytics showed these implied volatility readings:
| Stock | Implied volatility | Change that day |
|---|---|---|
| Microsoft (MSFT) | 37.3% | Up 5.1 points |
| Apple (AAPL) | 42.4% | Up 12.3 points |
| Nvidia (NVDA) | 45.7% | Up 5.9 points |
| AMD | 74.3% | Up 6.6 points |
| Amazon (AMZN) | 39.5% | Down 27.3 points |
The day's highest readings were TDOC at 389% and SLS at 258%. None of these numbers tells you whether the options are expensive for that stock. AMD's 74.3% is not "high" just because it's double Microsoft's; it can only be judged against AMD's own year. Rank and percentile put every stock on its own 0 to 100 scale so you can compare like with like.
The daily changes matter too. Apple's 12.3-point jump came ahead of its earnings report, and Amazon's 27.3-point drop came after its report, according to Tradytics' AI story for the day. Both measures swing hard around earnings, which is the next point.
How do traders use them, and where do they mislead?
- High readings (say above 50): options are rich against their own history, so selling premium or using spreads gets more attention, and outright buying costs more.
- Low readings (say below 30): options are cheap against their history, so buying calls, puts or straddles costs less.
- Earnings distort both. IV climbs into every report, so a high reading the week before earnings is normal. Compare it with the same point before past reports, and expect IV crush after.
- They look backward. A 95th-percentile reading can go higher in a real selloff. Neither measure is a timing signal.
Once you know where IV stands, vega tells you how many dollars a move in it means for your position, and the expected move tells you what that IV is pricing in price terms.
How to see it on Tradytics
Each stock's research options page, such as AMD options, shows implied volatility by expiry and by strike along with the expected move. The options market page has a Big movers card that compares each stock's option prices with their normal, the same relative question IV rank and percentile answer, and flags stocks with earnings ahead.
Common questions
What IV rank is good for selling options?
There's no magic cutoff. A common rule of thumb treats readings above 50 as rich and below 30 as cheap, but earnings dates and news matter more than any single threshold.
Can IV rank go above 100 or below 0?
Not if the window includes today. A new one-year high gives a rank of 100, and a new low gives 0.
Why does my broker show a different IV rank from another site?
They're likely using a different IV input, a different lookback, or closing versus intraday values.
Can I calculate IV rank for the whole market?
Yes. Cboe publishes daily VIX history back to 1990, so you can take the past year's high and low and plug them into the formula.
Sources
- Cboe, VIX historical data: daily VIX values from 1990 to the present, enough to compute rank and percentile for the S&P 500's implied volatility.
- Cboe, VIX Volatility Products: the VIX as a measure of expected near-term volatility from S&P 500 option prices.
- SPY options traded $335.96M in premium on 2026-11-02: $193.64M in calls and $142.32M in puts, a net bullish lean of $32.08M.
- Put/call ratio 0.73.
- Options price a move of about ±$6.07 by 2026-11-02.
- Max pain for 2026-11-02 is $765.0.
Updated 2026-11-02 16:05 UTC