Bitcoin volatility.
Compare Bitcoin’s historical daily price dispersion across complete 30-, 90- and 365-day windows. These annualised estimates do not predict price direction.
The last 30 days have been calmer than the 90-day baseline by 1.7%
Latest completed candle:
Reference data: Binance BTC/USDT perpetual daily closes. This is historical market data, not a product recommendation or access statement.
A dash means a complete daily window or a valid comparison baseline is unavailable; it does not mean zero volatility.
Closed daily log returns with sample standard deviation and √365 annualisation. Independent of the separate Crypto Fear & Greed Index
What realised volatility measures
Realised volatility describes the dispersion of past returns, not whether Bitcoin will rise or fall. This page uses completed UTC daily BTC/USDT perpetual reference candles. It is a contract-price estimate, not a consolidated spot-market benchmark, and the current incomplete day is excluded.
For each day, calculate ln(close / previous close). Take the sample standard deviation of exactly 30, 90 or 365 consecutive daily returns, then multiply by √365 × 100 to express an annualised percentage. A window needs one additional prior close. The daily percentage shown is the annualised estimate divided by √365; neither percentage is a guaranteed price range or expected return.
An independent comparison, separate from the sentiment index
Alternative.me describes volatility and maximum drawdown as a 25% component compared with recent 30- and 90-day averages. Its published description does not supply a complete reproducible scoring formula. This page’s 30-day/90-day realised-volatility ratio does not replicate that component or calculate a Fear & Greed score.
A ratio of 1 compares equal estimates; above 1 means the 30-day estimate exceeds the 90-day estimate. Our descriptive bands are below 0.75, 0.75 to below 1.25, 1.25 to below 1.75, and 1.75 or above. These are site-defined labels, not standard risk grades, probabilities or trading signals. A zero or unavailable baseline leaves the ratio unavailable.
Limitations worth knowing
Close-to-close estimates omit intraday movements that reverse before the daily close. Results depend on the instrument, price source, daily boundary, window and estimator. A sharp historical return can raise the result and its later exit from the window can lower it without a new market shock. Historical dispersion does not establish future profit or a suitable position size.
Only complete consecutive daily windows are shown. Missing or insufficient history appears as a dash, not zero or a shorter period relabelled as a year. The chart shows the available continuous sequence of complete 30-day windows. The observation date is the end of the latest closed candle; retained older data is labelled, and data older than seven days is not shown.
Frequently asked questions
Is high volatility bad?
High realised volatility means past returns were more dispersed. It says neither which direction comes next nor whether a trade is attractive. Position risk also depends on size, liquidity, leverage, costs and execution; the same notional remains the same notional when volatility changes.
Why is the number annualised when the window is only 30 days?
Annualisation puts historical estimates on a common scale using √365 for these daily crypto returns. It is a scaling convention, not an interest rate, confidence interval or forecast that the next year will repeat the sample.
How does this differ from implied volatility?
Realised volatility uses historical prices. Implied volatility is inferred from options prices under a pricing model; gamma measures how an option’s delta changes with the underlying price. This page calculates neither implied volatility nor gamma. Their relationship to realised volatility varies and does not guarantee an option seller’s profit.
Does the 24/7 nature of crypto change the calculation?
This implementation uses completed UTC daily candles and 365 daily periods per year because crypto trades every day. Other instruments, sampling choices and annualisation conventions can differ; the convention must be stated rather than assumed universal.
Why does your figure differ slightly from another site's?
First compare the exact instrument and price source, observation date, window, daily boundary, sample versus population standard deviation and annualisation convention. An hourly or high-low estimator is a different measure, and spot and perpetual prices can differ.
How often does this page update?
The calculation is rendered on page requests and has a five-minute process cache, but that is not the age of the market data. The shared yearly candle feed can be cached for an hour and can retain older observations during an upstream failure. Check the displayed candle date and stale notice; reloading does not create a new daily close.