← All guides

Crypto volatility explained: why prices swing, and how to measure the swing

Market basics · 11 min read

Financial disclaimer, not financial advice. This article is general information and education only. It does not recommend buying, selling, or holding any crypto asset, fund, or product, and it is not personalised financial, tax, or legal advice. Crypto assets are volatile and speculative and can lose all of their value. Consider your own circumstances and, where appropriate, speak with a qualified professional in your jurisdiction before making any financial decision.

Ask what volatility is and the usual answer is a feeling: prices move fast and sleeping well requires not looking. In finance the term has a precise definition, it is the statistical spread of an asset's returns, typically expressed as the annualized standard deviation of returns. Once that definition is in hand, a quoted figure such as "55% volatility" becomes a quantity: it can be recomputed from raw price data, checked against a quoted index, and translated into the expected size of a daily or monthly move. This guide covers why crypto's swings are structurally larger, then builds the measurement toolkit: a hand calculation anyone can repeat, the annualization convention most people get wrong for crypto, the conversion from an annual figure to a daily and monthly expected range, and the difference between what happened and what the options market expects.

Volatility is a spread of returns, not a direction

Volatility says nothing about whether prices go up or down. It measures how far daily returns scatter around their average. A high number means returns spread over a wide range: gains and losses alike arrive in large sizes. That is the most misread property of the term, a rising figure does not mean a drop is coming; it means a big move is priced in, direction unknown.

To put crypto's scale in context, consider typical annualized volatility by asset class. Bitcoin runs roughly 50 to 70% annualized; Ethereum roughly 60 to 90%; the S&P 500 roughly 15 to 20%; gold roughly 12 to 15%; US Treasury bonds roughly 3 to 5%; and a dollar stablecoin under 1%, brief depeg episodes aside. Small-cap altcoins can exceed 100 to 200%. Bitcoin's volatility has been measured at about 3.6 times that of gold and about 5.1 times that of global equities, ratios that are declining as the market matures, while the ordering never changes.

Why crypto swings harder: the structural causes

Crypto does not swing wildly because the technology is broken. It swings because of structural features that traditional markets either lack or have dampened over centuries. Six carry most of the weight.

The market is small. The entire crypto market sits around 2.5 trillion dollars, against about 45 trillion for the S&P 500 alone. In a smaller market, the same capital flow produces a proportionally larger price impact, a stone makes bigger waves in a pond than in an ocean. This size relationship is the root cause of most of crypto's excess volatility, and it moderates as the market grows.

Trading never stops and nothing halts it. Crypto markets run around the clock, 365 days a year, with no closing bell and no circuit breaker. A crash that starts at 3 AM on a Sunday keeps going until buyers show up.

Leverage is extreme and liquidations are automatic. Crypto exchanges offer 10x, 50x, even 100x leverage on futures contracts. When prices move against leveraged positions, forced liquidations dump additional selling pressure onto the market, a feedback loop in which falling prices trigger liquidations that cause more falling.

There is no earnings anchor. Most cryptocurrencies generate no revenue and pay no dividends, so there is no earnings call to anchor valuation; prices reflect what people believe the asset will be worth, and beliefs change fast. Sentiment is priced at internet speed, so FOMO drives buying frenzies and fear drives panic selling, both magnified by social media.

Regulation arrives as binary shocks. Decisions from dozens of uncoordinated jurisdictions land as sudden events with massive market impact, mining bans on one side, spot ETF approvals on the other.

Liquidity is thin and concentrated. Large holders can move markets; on-chain data has shown roughly two thousand addresses controlling about two-fifths of all bitcoin, so a whale moving coins to an exchange is often read as a sell signal. For altcoins the problem is simpler: thin order books mean a modest market order can move a small-cap token sharply, and slippage alone can cost 2 to 5% on a single trade.

None of this is permanent. Bitcoin's volatility has trended down for years: annualized volatility frequently exceeded 200% in 2011, was typically 60 to 80% by 2020, and hovered around 45 to 65% in 2025 to 2026, as the market grew and institutional participation through ETFs expanded. The trend is toward a calmer market, one that still sits at roughly three times the S&P 500.

Measure it yourself: realized volatility, worked step by step

Realized (historical) volatility is what already happened, computed from past prices by the standard "close-to-close" method: take each day's percentage change, compute the standard deviation of those daily returns, then annualize. Here is the calculation on a labelled hypothetical example, ten daily returns, computed for this article.

Hypothetical daily returns over ten days, in percent: 2.0, −3.0, 1.5, −1.0, 4.0, −2.5, 0.5, 3.0, −4.0, 1.0.

Step 1: find the average. The mean of the ten returns is 0.15%. This is the centre the returns scatter around.

Step 2: measure the scatter. Subtract the mean from each return and square the result, squaring removes the signs and weights big deviations more heavily. Averaging the squared deviations (dividing by 9, the sample convention) gives about 7.17; its square root, about 2.68%, is the daily standard deviation: a typical day in this series sits about 2.7% from the average.

Step 4: annualize. Random daily wiggles partly cancel over time, and the scaling factor that converts a daily spread to an annual one is the square root of the number of days. Crypto trades every day, so the factor is the square root of 365 (about 19.1). Multiply: 2.68% × 19.1 ≈ 51.2% annualized. That is the finished figure: this hypothetical series has about 51% realized volatility, squarely in crypto's normal band.

The annualization trap: 365 days, not 252

This is the one detail that most people get wrong. Stock-market convention annualizes daily volatility with the square root of about 252, the number of trading days in a year, excluding weekends and holidays. Guides written for equities carry that convention in muscle memory, and it leaks into crypto calculators and explainer articles. But crypto trades 365 days a year, and using 252 understates the answer: for the same hypothetical series, the 252-day convention gives about 42.5% instead of 51.2%, roughly a sixth lower, for no reason other than the wrong calendar.

The practical rule is simple: any realized-volatility figure computed from daily closes should be annualized with √365 for crypto. When reading a quoted figure, check which convention the source used, and if the source is silent, treat a round-looking number with suspicion, because the silent convention is usually 252.

From an annual number to a move you can picture

An annualized figure answers "how volatile is this?" but not "how far could it move tomorrow?" The conversion is the same square-root scaling run in reverse: divide the annualized figure by √365 to get the expected daily move, then multiply by the square root of the horizon in days for any other period. The results are one-standard-deviation moves: under the textbook model, the price stays inside the range about two times out of three.

Annualized volatilityExpected daily moveExpected 7-day moveExpected 30-day move
30%about ±1.6%about ±4.1%about ±8.6%
40%about ±2.1%about ±5.5%about ±11.5%
60%about ±3.1%about ±8.3%about ±17.2%
90%about ±4.7%about ±12.5%about ±25.8%

These figures are documented conversion arithmetic (each row verified for this article), with a standing warning: crypto has fatter tails than the textbook model, so moves beyond two standard deviations happen more often than the model says. The range shows what a normal bad week looks like; it does not cap an abnormal one. If the implied daily move is ±2%, a 1.5% dip is noise and a 6% drop is news.

Realized vs implied: what DVOL adds

Realized volatility is backward-looking: what the price already did. Implied volatility is the forecast baked into options prices, what the market expects, derived from the price of insurance traders actually pay. Deribit's DVOL index is bitcoin's 30-day implied volatility index, expressed as an annualized percentage; a reading of 40 implies a typical move of roughly ±2% a day and ±11.5% over a month, and says nothing about direction.

Three properties of the realized-implied relationship are worth knowing. First, implied usually sits above realized: option sellers take the risk of a sharp move and charge for it, the volatility risk premium. In one analysis of Deribit data, DVOL sat above the following 30 days' realized volatility 65% of the time, by 3.5 points on average. The gap is the normal price of insurance, not a signal. Second, in the remaining 35% of cases the market moved more than options had priced, the episodes people remember. Third, a rising implied index does not mean a drop is coming: it climbs in bull runs too, as traders chase calls. A documented October 2026 DVOL reading of 38.29 implied about ±2.0% per day and ±11.0% over 30 days, a calm month by bitcoin's standards.

Put to use, the pair works like a weather report: if the implied daily move doubles from 2% to 4%, a 2% stop gets hit by noise, so position size gets cut; low implied readings make buying protection cheap; and rising leverage against low implied volatility is the classic setup for a sharp move, direction unknown.

Reading any volatility figure: a short checklist

  1. Realized or implied? Past fact or forward-looking market price? The number means different things in each case.
  2. Which window? A 30-day realized figure answers a different question than an annual one. Match the window to the question you are asking.
  3. Annualized how? For crypto the factor is √365. If the source is silent about its convention, assume 252 and mentally add back the difference.
  4. Converted to a horizon? Run the table: an annual number is abstract, a daily or monthly expected move is actionable.
  5. Direction not implied. A high reading is an expected move size, not a forecast of up or down.
  6. Tails are fat. The one-standard-deviation range covers about two moves in three in theory; crypto exceeds the range more often than theory says. Size for the outlier, not the average.

What success looks like: given any quoted volatility figure, you can name its type, window, and annualization, and translate it into an expected daily and 30-day move. A figure that survives all six questions is information; one that fails the first two is atmosphere.

FAQ

Is high volatility always bad? No, volatility is directionless. The relevant question is whether the size of the swings matches the time horizon and position size involved.

Why does bitcoin's volatility keep falling? The downtrend tracks a larger market, more institutional capital arriving through ETFs, and hedging tools that let traders buy protection instead of panic-selling.

Do stablecoins have zero volatility? Not quite: a dollar stablecoin runs under 1% annualized, but depeg episodes have occurred (a brief slide to about $0.97 in May 2022), so "stable" means pegged with rare exceptions, not risk-free.

Can the stock formula be used in a spreadsheet? Yes, the close-to-close standard deviation is the same computation for any asset. The only crypto adjustment is the annualization factor: √365, not √252.

Reminder: Crypto Investing is educational content only, not financial advice. Crypto assets are volatile and can lose all value. Product terms, fees, and rules change, read the current prospectus, terms, or manufacturer documentation for any product you are considering.