Realised volatility: what actually moved
Realised (historical) volatility measures past price variation, not direction. Our volatility screener uses daily log returns r = ln(close today / close yesterday). RV = sample stdev(r) × √365: subtract the mean return, sum squared deviations, divide by n − 1, take the square root, then annualise. Multiply by 100 to display a percentage. Crypto trades every day, so we use 365 rather than 252 trading days.
Implied volatility: what options price
Implied volatility (IV) is the volatility input inferred from an option's price under a pricing model. It describes the cost of future uncertainty, not a promised move. Our 30-day ATM IV comparison uses at-the-money options. Between listed expiries T1 and T2, interpolate total variance w = σ²T linearly to T = 30/365 years, then IV30 = √(w30 / T). An eligible exact 30-day maturity needs no interpolation. See the IV, smile and skew guide.
Why they differ: volatility risk premium
Option buyers pay for insurance against event risk and jumps. Sellers take uncertain hedging costs and tail losses, so IV often sits above subsequent realised volatility: the volatility risk premium. Supply, demand and liquidity also affect prices. The gap against trailing RV shown here is only a proxy; it does not measure the eventual premium earned over the option's life.
Match the horizons and annualisation
Compare 30-day RV with 30-day IV on the same annualised basis. The 7-day RV reading tells you whether the last week was unusually active; it is not a substitute for 30-day RV in this spread. RV looks backward over completed daily candles; IV looks forward from its observation time. Matching the length does not make their dates or information sets identical.
Reading the spread
- IV well above RV: protection may be expensive relative to recent movement, or the market may anticipate a catalyst absent from the trailing window. It does not guarantee that selling options is profitable.
- IV below RV: a recent shock may still dominate RV while options price calmer conditions ahead. Unexpected jumps can also make RV overtake IV before quotes adjust.
Worked example with the cached live readings
One or both 30-day readings are unavailable, so no live spread is calculated. Open the RV screener and IV comparison for current coverage.
For a hypothetical arithmetic example, 36% IV minus 32% RV equals +4 vol points. This is a difference in annualised volatility percentage points, not a 4% relative change or an expected return.
Limits
- RV uses one exchange's daily candles: Binance USDT perpetual futures, completed UTC days. It is not a consolidated spot index and close-to-close returns miss intraday swings. The screener permits partial coverage (at least 80% of days); missing candles can affect the sample.
- IV uses recorded eligible venues (Binance and Bybit where publicly available) and the latest matched BTC/ETH daily 08:00 UTC observation. It is not a streaming option quote. Venue coverage and eligible expiries can change.
- The box reads only the parent pages' existing in-process caches. A cold process or stale cache shows Unavailable independently for each side; missing does not mean zero. RV's end date and IV's observation time can differ.
- The spread is descriptive. Transaction costs, funding, hedging and jump risk matter. This is not investment advice.
Checked against the exchanges' own documentation on 20 October 2026: Binance API: daily klines · Bybit: implied volatility. Not financial advice.