Bitcoin volatility.
How violently Bitcoin has actually moved, measured against its own baseline. The same measure the Fear & Greed Index weights at 25%.
The last 30 days have been more volatile than the 90-day baseline by 6.9%
Close-to-close log returns, annualised by the square root of periods per year. This is the measure weighted at 25% in the Crypto Fear & Greed Index
What realised volatility measures
Realised volatility is a measure of how much an asset actually moved, worked out after the fact from its own price history. It is not a forecast and it is not an opinion about direction — a market that fell hard and a market that rallied hard can carry exactly the same volatility reading.
The number above is calculated the standard way. Take the closing price of each candle, work out the natural logarithm of each day's change against the one before it, and take the standard deviation of that set of returns. That gives a per-day figure, which is then annualised by multiplying by the square root of the number of periods in a year. Annualising is a convention, not a prediction: a 60% annualised reading means that IF the past month's daily movement continued unchanged for a year, the spread of outcomes would be about 60%.
Bitcoin has historically run somewhere between 30% and 100% annualised, spiking far higher around liquidation cascades, exchange failures and macro shocks. Equity indices for comparison typically sit in the teens. That gap is the reason position sizing on Bitcoin has to be treated as a different exercise from position sizing on stocks, not merely a more aggressive version of it.
How the Fear & Greed Index uses it
Volatility is the single heaviest input to the Crypto Fear & Greed Index, tied with market momentum at 25% of the composite score. The index does not ask whether volatility is high in absolute terms. It asks whether current volatility is unusual compared with the recent past — specifically, it measures the current reading against its own 30-day and 90-day averages.
That comparison is the ratio shown above. A ratio near 1.00 means the last month looks much like the last quarter, and the volatility factor contributes little in either direction. A ratio well above 1.00 means the market has become jumpier than its own recent baseline, and the index reads that as fear. A ratio well below 1.00 means things have gone quiet relative to the quarter, which the index treats as complacency and scores toward greed.
This is worth understanding because it explains a result that surprises people: Bitcoin can be extremely volatile in absolute terms and still contribute a neutral volatility score, simply because it has been that volatile for months. The index is measuring change in character, not level.
Limitations worth knowing
Realised volatility is backward-looking by construction. It tells you what has already happened and carries no information about what happens next. Options markets price implied volatility, which is a forward-looking estimate, and the two frequently disagree — that disagreement is itself a traded market.
Annualising a 30-day window amplifies whatever that window happened to contain. One violent liquidation day inside a 30-day sample can lift the annualised figure substantially and then drop it just as sharply thirty days later, with nothing having changed in the market on the day it fell. Short windows are more responsive and noisier; the 90-day and one-year figures are shown alongside precisely so the 30-day number can be read in context rather than on its own.
Finally, this measure is close-to-close. It does not see what happened inside each candle, so a day that crashed and fully recovered before the close registers as a quiet day. Estimators that use the high and low — Parkinson, Garman-Klass — capture more of that intraday range, at the cost of being more sensitive to thin or illiquid books.
Frequently asked questions
Is high volatility bad?
It is neither good nor bad on its own — it is a measure of dispersion, not direction. What it does change is risk: the same position size carries materially more exposure when volatility doubles. Traders who size positions as a fixed percentage of account value are quietly taking far more risk in a high-volatility regime than a low one, which is why volatility-adjusted sizing exists.
Why is the number annualised when the window is only 30 days?
Annualising puts every window on the same scale so a 30-day and a 90-day reading can be compared directly. It is a unit convention, like quoting an interest rate per year on a one-month deposit. It does not imply a forecast that the next year will look like the last month.
How does this differ from implied volatility?
Realised volatility is computed from prices that have already printed. Implied volatility is extracted from options prices and represents what the market is currently charging for future movement. Implied typically trades above realised — that spread is the premium sellers earn for carrying the risk — and the gap widens ahead of events the market expects to be turbulent.
Does the 24/7 nature of crypto change the calculation?
It simplifies it. Equity volatility calculations have to handle overnight gaps and non-trading days, and analysts argue about whether to annualise using 252 trading days or 365 calendar days. Crypto trades continuously, so calendar days and trading days are the same thing and the annualisation factor is unambiguous.
Why does your figure differ slightly from another site's?
Almost always because of window length, candle source or estimator. A 30-day close-to-close figure from daily Binance candles will not match a 20-day figure, a figure built from hourly candles, or a Parkinson estimate using highs and lows. The method used here is stated above so the number can be reproduced rather than taken on trust.
How often does this page update?
The underlying candles are cached for five minutes, so a reload picks up new data at most that far behind. Realised volatility over a 30-day window moves very slowly in any case — meaningful change in the reading takes days, not minutes.