BTC volatility dashboard: read realized and implied stress

Reading each number on the dashboard.

The dashboard shows three realized volatility (RV) numbers: 7-day, 30-day and 90-day, all annualized. Realized volatility is computed from actual log returns, not implied from options prices. The 7-day reflects current regime, the 30-day reflects the prevailing trend, the 90-day reflects the structural baseline. The relationship between the three tells most of the story.

Four common bands at a glance.

BTC RV typically falls into four bands. Below 30% annualized signals a sustained calm regime — rare and often a coiling state before larger moves. 30%-55% is the modal band, covering most of 2023-2025. 55%-90% indicates an active phase, common during late-cycle rallies. Above 90% marks crisis or euphoria peaks: the March 2020 COVID crash, the May 2021 leverage flush, the November 2022 FTX collapse, the August 2024 yen-carry unwind all printed above 90% on 7-day RV.

What realized volatility actually is.

Realized volatility is the standard deviation of log returns over a rolling window, annualized by multiplying by the square root of the number of periods per year. For daily data, that means RV equals stdev(log returns) times sqrt(365), where the standard deviation subtracts the window's mean log return and divides by N (the population form): RV = √(Σ(rᵢ − r̄)² / N) × √365. That is the same calculation the live dashboard on the Chinese page runs. The result is a statistical measure of how much BTC has actually moved, distinct from implied volatility, which captures what options markets expect. Kaiko, Glassnode and CoinMetrics all publish RV time series for cross-reference.

BTC RV history in broad strokes.

2013-2016 — early wild era.

On Bitstamp BTC/USD daily closes, 30-day RV was above 90% on about a third of days in 2013-2014 and above 150% on roughly one day in eight, with a peak above 300% in May 2013. Liquidity was thin, regulatory clarity was nonexistent, and Mt. Gox's collapse in early 2014 set the tone. Modern reads of this era should treat the numbers as descriptive of a different market, not directly comparable.

2017-2020 — first mainstream wave through 312.

During the late-2017 ICO rally 30-day RV sat mostly between about 77% and 120% (Binance BTCUSDT, September-December 2017), reaching about 152% at year-end; it swung widely through the 2018 bear, then spiked above 200% after March 2020's "Black Thursday" COVID crash (about 208% on 6 April 2020).

2021-2022 — pre-institutional adoption.

The April 2021 ATH ($64,854 on 14 April) was reached with 30-day RV only around 53%-62%. The May 2021 leverage washout pushed it to about 118% (24 May). After the FTX collapse, 30-day RV peaked near 82% on 23 November 2022 and had fallen back to about 28% by 31 December — the shock faded faster than earlier ones.

2023-2024 — around ETF approval.

RV compressed to multi-year lows in mid-2023 (30-day RV near 16% on 13 August 2023). After the January 2024 spot ETF launch, 7-day RV rose from about 49% to 76%-78% on 12-14 January as BTC sold off, but 30-day RV stayed in the 45%-54% range through 23 January. A common reading is that broader institutional access has damped volatility, but treat that as an interpretation, not a law.

2025-2026 — current phase.

From January 2025 to September 2026, 30-day RV spent half its time between about 31% and 46% (median about 39%), with a spike to about 82% in late February and early March 2026. Common explanations for the lower baseline are ETF flow absorbing more of the marginal demand and deeper regulated derivatives markets — plausible, but an interpretation rather than a measured cause.

Reading the three numbers.

First — the 7-day to 30-day relationship.

When 7d RV is materially above 30d RV (say 1.3x or higher), the market is in an accelerating regime. When 7d is below 30d (say 0.7x or lower), the market is calming down. The crossover points often coincide with regime shifts.

Second — 30-day RV's percentile within one year.

A 30d RV of 50% in absolute terms means little until you place it in context: rank it against the past 365 daily readings and look at the percentile, not the raw number.

Third — use RV as a stop-loss width yardstick.

A 30d RV of 50% annualized implies daily standard deviation of roughly 2.6%. Stop losses tighter than 1.5x that daily stdev are statistically likely to be hit by noise. RV-aware sizing is the cleanest way to set risk without curve-fitting.

Volatility view of key BTC events.

March 12, 2020 — Black Thursday.

BTC fell about 44% intraday on 12 March (Binance BTCUSDT open 7,934, low 4,410) and about 52% from that open to the 13 March low of 3,782. 7-day RV peaked near 374% annualized on 18 March.

April 2021 — first cycle top.

30-day RV was only about 53%-62% around the 14 April top — volatility was not elevated at the peak. The May 2021 washout then pushed 7-day RV to about 174% (24-25 May).

May 2022 — Terra/Luna collapse.

7-day RV rose from about 60% to about 108% (10 May) as the Anchor depeg and UST collapse propagated, and to about 148% during the June follow-through. BTC fell from $40,024 (4 May) to $26,700 (12 May) in eight days.

November 2022 — FTX collapse.

7-day RV touched about 152% (10 November). The bottom came when 30-day RV peaked and started compressing — a pattern repeated in subsequent shocks.

January 11, 2024 — ETF approval day.

7-day RV rose to about 76%-78% on 12-14 January as BTC sold off after launch ("sell the news"), then fell back to about 31% by 21 January. Farside's daily table shows heavy GBTC redemptions offsetting new-fund inflows in those first sessions.

August 5, 2024 — global asset flash crash.

7-day RV rose from about 30% in late July to about 54% on 5 August, about 69% on 6 August and about 114% on 8 August as the yen-carry unwind hit crypto alongside other risk assets.

2025-2026 current phase.

From January 2025 to September 2026, 30-day RV mostly stayed between about 31% and 46%, with a spike to about 82% in late February and early March 2026.

BTC vs gold vs equities vs altcoins.

Over the past year (26 September 2025 to 25 September 2026) BTC's 30-day RV had a median of about 41%, roughly 1.6x gold (GLD, about 25% — itself unusually high; gold's ten-year median is about 13%) and more than 3x the S&P 500 (SPY, about 12%). Versus other crypto assets: ETH's and SOL's medians were both about 62%, roughly 1.5x BTC; smaller coins usually run higher still. Method: Binance daily closes for crypto (√365), Yahoo Finance daily closes over 21 trading days for GLD and SPY (√252). Risk-parity sizing across crypto requires accounting for this dispersion.

Three things RV cannot do — common misuses.

1. RV does not predict direction.

RV measures magnitude only. High RV is consistent with both rapid rallies and crashes. Do not conflate high RV with bearish sentiment.

2. Short-window RV is noisy.

7-day RV computed over only 7 daily returns has wide statistical error bars. Treat 7-day swings as suggestive, not conclusive. The 30-day window is the most signal-rich length.

3. RV is historical, not predictive.

It is by definition backward-looking. Implied volatility (from options) is the forward-looking analog, and the gap between RV and IV (the volatility risk premium) is itself an interesting indicator — published by Deribit and Glassnode for BTC and ETH.

Three trading scenarios — how to use RV.

Scenario 1 — spot DCA / long-term holding.

RV is irrelevant. Long-term holders should ignore short-term RV swings entirely. The 90-day RV provides a sense of regime but should not drive accumulation cadence.

Scenario 2 — leveraged long/short, 3x-10x.

RV is critical for sizing and stop placement. A position sized to lose 2% of capital at a stop one standard deviation below entry will hit that stop roughly 16% of the time by chance. Sizing should account for the current 30-day RV regime.

Scenario 3 — grid trading / arbitrage.

Grid strategies thrive in mid-RV ranges (40%-70%) where price oscillates but does not break. Below 30%, grid spacing becomes inefficient; above 90%, the grid is run over by trending moves. RV-aware grid configuration is standard practice for serious algo traders.

What to read alongside RV.

• Funding rate guide — derivatives cost layer.
• Open interest guide — derivatives position size.
• Perp-spot ratio — leverage condition.
• Liquidation map — where the leverage breaks.

FAQ — frequently asked questions.

Why use log returns, not percentage returns?

Log returns are time-additive and statistically cleaner. The difference is negligible for small returns but becomes meaningful at the daily scale.

Why annualize with sqrt(365) instead of sqrt(252)?

Crypto trades 365 days per year, not 252. Annualizing with sqrt(252) is the convention for equities but understates crypto RV by roughly 17% (√252 / √365 ≈ 0.83).

Why does 7-day RV look so jumpy?

Statistical noise. With only 7 observations, sampling variance is high. Use 30-day for stable readings, 7-day for regime-shift detection.

Why not compute IV (implied volatility)?

IV requires options prices. Deribit publishes BTC and ETH IV for free; this dashboard focuses on the public-price-only RV that can be computed without an options data feed.

How often does the data refresh?

This English page is a reading guide without a live widget. The live dashboard on the Chinese version recalculates from Binance BTCUSDT daily candles each time the page is loaded; a new daily candle opens at 00:00 UTC.

What if Binance API is blocked?

The live dashboard uses Binance's public API only. If it is unreachable, the dashboard shows an "API unreachable" notice and an offline snapshot dated 25 September 2026 (7-day 48.4%, 30-day 42.3%, 90-day 38.8%, computed from Binance BTCUSDT daily closes with the same method as the live readings) — those are not current readings, so do not act on them.

Related reading.

For deeper background, compute the same RV on other assets' public daily closes and compare them on one chart — the ranking matters more than any single reading.

Annualized RV reference table — BTC across cycles.

The table below summarizes RV ranges by era, useful for context when reading the current dashboard against history. Method: 30-day RV from daily closes (√365); Bitstamp BTC/USD for 2013-2014, Binance BTCUSDT from September 2017 onward (2025-2026 runs to 25 September 2026).

Two structural takeaways from this table. First, the 2023-2026 compression shows up clearly in the data, though whether it lasts is an open question. Second, every era still produces shock-driven spikes; the compression does not eliminate them. The August 2024 7-day spike was back near 38% by 15 August, about ten days later.

For practitioners building risk frameworks, the practical takeaway is that historical RV percentiles within the current era are more useful than absolute thresholds drawn from earlier cycles. A 60% RV reading in 2025 is high-percentile within the current regime, while the same number in 2018 was routine.