Reading an implied-volatility number
An implied-volatility reading is an annualized percentage, but most moves you care about happen over days or weeks — so the first skill is converting it to a horizon you can use. A quick rule: divide the annual figure by about 16 to get a rough one-day expected move (16 is the square root of the ~256 trading days in a year). So 20% annual implied volatility implies a daily move of roughly 1.25%. For a month, divide by about 3.5 (the square root of 12): that same 20% implies a monthly move near 5.8%.
The second skill is context. A 20% reading means nothing in isolation — it is high for a sleepy bond ETF and low for a single tech stock into earnings. What matters is where today’s number sits relative to its own history. Is it in the top decile of the last year, or the bottom quartile? That percentile framing is how professionals decide whether volatility is cheap or expensive.
A worked example
NVDA’s VolDex® reads 55 into earnings; a broad ETF reads 14. Convert first: 55 ÷ 16 ≈ a 3.4% expected daily move for NVDA; 14 ÷ 16 ≈ 0.9% for the ETF.
But is 55 “high”? Only against NVDA’s own record. If NVDA routinely runs 45–70, a 55 is unremarkable — mid-range. Meanwhile the ETF at 14 might sit in the 90th percentile of its own year, which is genuinely stretched. The lower raw number is the more extreme reading. Percentile beats level, every time.
The third skill is to watch the change, not just the level. A jump in implied volatility often signals that new risk has been recognized; a steady bleed lower usually accompanies grinding, complacent markets. The direction and speed of the move carry as much information as the number itself. And finally, remember what the number is not: it is not a forecast of direction, and it is not a promise. It is the market’s probability-weighted expectation of magnitude — and, like any expectation, often wrong, which is precisely why the gap between implied and realized is tradeable.
Common pitfalls
Comparing raw IV across names. A 30 on one stock and a 30 on another are not the same signal; each must be ranked against its own history.
Reacting to the level, not the change. A reading holding steady at a high level is very different from one spiking there today.
Treating IV as a direction call. It measures expected magnitude, not which way the price will go.
One distinction to keep straight as the habit forms: IV percentile (the share of days over the past year when implied volatility was lower than today) and IV rank (where today sits between the year’s low and high) can disagree, especially after a single large spike. Percentile is usually the steadier read, because one outlier day stretches the range but barely moves the distribution.
What to do with this
Read every implied-volatility number in three moves: convert it to your trading horizon, rank it against its own history as a percentile, then check whether it is rising or falling. Only after those three steps does a raw figure become a decision. Next, we look at why puts and calls almost never carry the same implied volatility — the skew.
Next lesson · continue the courseSkew: why puts and calls cost different amounts →
Percentiles turn a raw reading into a verdict. Every index and underlying carries live lifetime and 52-week percentiles inside ETF Analytics.