Where the number comes from
An options pricing model takes a set of inputs — the underlying price, the strike, time to expiry, interest rates, expected dividends and volatility — and returns a theoretical premium. Every one of those inputs is observable except volatility.
Implied volatility inverts the problem. Rather than supplying a volatility figure to get a price, you take the price the market is actually paying and solve for the volatility that would justify it. The result is the market's current expectation, expressed as an annualised percentage.
That is why implied volatility behaves like a price rather than a fact. It moves with supply and demand for the options themselves. Heavy buying of protection before an earnings announcement lifts premiums, and IV rises with them — not because the stock has become more volatile yet, but because people are paying more for the possibility that it will.
Three measures that answer different questions
| Measure | Looks at | Derived from | Tells you |
|---|---|---|---|
| Historical volatility | Past price movement | Actual closing prices over a chosen window | How much the underlying has moved |
| Implied volatility | Expected future movement | Current option premiums, solved backwards from a pricing model | How much movement the market is currently pricing in |
| IV rank | Where today's IV sits in its own range | Current IV against its high and low over a lookback period | Whether IV is high or low for this particular underlying |
The comparability problem
Why 30% IV means nothing on its own
Underlyings have their own characteristic volatility. A large, heavily traded industrial and a speculative small-cap explorer do not sit anywhere near each other on a volatility scale, and they never will. An implied volatility of 30% might be historically extreme for the first and unremarkable for the second.
So a raw IV percentage cannot be read across underlyings. Doing so compares their baseline character, not their current condition — which is almost never the question you were actually asking.
IV rank
Places current IV between the lowest and highest readings over a lookback period, usually twelve months, on a 0–100 scale. An IV rank of 80 says today sits near the top of that underlying's own range. Because it uses only the two extremes, a single spike months ago can compress every reading since.
IV percentile
Measures the proportion of days in the period on which IV closed lower than today. It uses the whole distribution rather than the endpoints, so it is more robust to a single outlier — the two can disagree sharply after one violent move.
Lookback window
Both measures depend entirely on the period chosen. The same underlying can show a high rank on a three-month window and a low one on two years. Comparing rank figures calculated over different windows is not a comparison.
What these measures do not tell you
A high IV rank says premiums are expensive relative to that underlying's own history. It does not say they are mispriced. Implied volatility is frequently elevated for a reason that has not resolved yet — a pending announcement, a bid, a regulatory decision — and the subsequent move can justify the price entirely.
Treating a rank figure as a signal in isolation mistakes a description for a forecast. It describes where the price sits; it does not tell you why, and the why is usually the part that matters.
How volatility feeds into margin →Rising implied volatility increases the margin required on written positions. A volatility spike can therefore raise your obligation on a position you have not touched.