Thirty vol is high for one stock and low for another. A utility that normally trades at 15 would be in a panic at 30; a small-cap biotech at 30 would be asleep. So the question is never whether IV is high, but whether it is high for this stock, against its own history. Two measures answer that, and they are easy to confuse.
IV rank
IV rank places today's 30-day IV inside the range of the past year. Take the lowest and highest readings of the year, and express today's reading as a percentage of the distance between them. If the year's low was 18, the high was 58 and today is 28, the rank is (28 minus 18) divided by (58 minus 18), which is 25. A rank of 0 means IV is at its one-year low; 100 means it is at its high.
The weakness is in the two numbers it depends on. One bad week can set the year's high at 58, and for the next twelve months every reading is measured against that spike. The stock may spend nine of those months between 20 and 30, a perfectly normal range for it, while the rank reads a permanently low 5 to 30 and tells you options are cheap. They are not cheap. They are normal, and the rank has been stretched by a single event.
The chart puts dollars on it. It prices the same at-the-money straddle at different implied volatilities, so you can see what the spike did to the price of options against what a normal day costs.
A 30-day at-the-money straddle on a $100 stock against implied volatility. Hover at 20% and 26%, a normal range, then at 70%, a spike. The engine cannot draw a year of IV history; the link opens the Rank tab, which does.
At the spike the straddle cost almost three times its normal price. Measured against that, every ordinary day looks like a bargain, which is exactly the distortion the rank builds in.
IV percentile
IV percentile ignores the extremes and counts days. Of the roughly 252 trading days in the past year, on what share was IV below today's reading? Using the same stock, if 160 of the 252 days had IV under 28, the percentile is 63. Today's reading is higher than it was on almost two thirds of the year, which is a very different message from a rank of 25.
Percentile is more honest because a spike counts as a few days, not as the end of the ruler. It answers the question you actually care about: how often has this stock been cheaper to insure than it is now? The Terminal's guide reads: below the 30th percentile is cheap against the stock's own history, above the 70th is rich. Both measures are shown, because rank is the older convention and many traders still quote it, but when the two disagree, believe the percentile.
Cheap and rich, with care
A low percentile says options are cheap for this stock, not cheap in absolute terms and not necessarily a buy. If the stock has genuinely changed, say a takeover completed or a volatile business sold off, its old history stops being the right yardstick. A high percentile ahead of earnings is expected, not an anomaly. The measures tell you where today sits against the record; the reason for it is your job. Pair them with the cone from the last lesson, which does the same percentile thinking for realized volatility, and you have both sides of the picture.
The maths: the two formulas side by side
IV rank = (IV today - IV low) / (IV high - IV low) × 100, where the low and high are the lowest and highest 30-day IV readings of the past year.
IV percentile = (days with IV below today) / (days in the window) × 100.
With the numbers from above: rank is (28 - 18) / (58 - 18) × 100 = 10 / 40 × 100 = 25. Percentile is 160 / 252 × 100, about 63.5. Rank uses two data points out of the year; percentile uses all 252, which is why one spike moves the rank so much and the percentile so little. Conventions differ slightly between screens (below, or at and below; 252 days or a calendar year), so expect the same stock to read a point or two apart in different places.
Your turn
Open MSFT. Read the one-year percentile from Vol Snapshot, then the rank from IV Intelligence's Rank tab, and notice whether they agree. The last live question goes back to the Cone tab for the realized side of the same idea.