Bitcoin Low-Volatility Periods: Dates and Duration
Compare observed quiet spells in the stored daily-close archive, with known and unknown boundaries shown explicitly.
Quiet threshold: 30.0% or lower. Fixed descriptive threshold for Bitcoin's 30-day annualised realised volatility; a period needs at least seven consecutive quiet daily readings.
Dated Bitcoin quiet periods
8 shown · latest 40 at most · UTC dates| Start | Last quiet date | Quiet readings | Boundaries | Minimum vol | Episode price change | 7 calendar days after last quiet date |
|---|---|---|---|---|---|---|
| 2026-08-01 | 2026-08-18 | 18 days | Naturally ended | 21.9% | +3.03% | +21.34% 2026-08-25 · observed |
| 2026-05-17 | 2026-06-01 | 16 days | Naturally ended | 25.0% | -7.81% | -11.66% 2026-06-08 · observed |
| 2026-01-01 | 2026-01-19 | 19 days | Naturally ended | 25.0% | +4.27% | -4.62% 2026-01-26 · observed |
| 2025-09-18 | 2025-10-08 | 21 days | Naturally ended | 23.1% | +5.32% | -10.17% 2025-10-15 · observed |
| 2025-07-17 | 2025-08-21 | 36 days | Naturally ended | 21.2% | -5.60% | +0.06% 2025-08-28 · observed |
| 2023-09-28 | 2023-10-19 | 22 days | Naturally ended | 22.3% | +6.26% | +18.94% 2023-10-26 · observed |
| 2023-07-21 | 2023-08-16 | 27 days | Naturally ended | 16.8% | -3.92% | -8.00% 2023-08-23 · observed |
| 2022-12-30 | 2023-01-11 | 13 days | Naturally ended | 25.8% | +8.04% | +15.24% 2023-01-18 · observed |
Archive coverage. Read window: 2016-10-26 through 2026-10-25, the cutoff for completed UTC observations (today's unfinished day is excluded). Actual source dates: 2019-01-01 to 2026-10-25. Complete daily closes: 2835; missing or excluded days from the source beginning through the cutoff: 20. Qualified volatility readings: 2333; complete days without a qualified reading during warmup: 502. Complete-close coverage: 2019-01-01 to 2026-10-25. Qualified reading coverage: 2019-01-31 to 2026-10-25.
Source. Stored Binance spot BTCUSDT hourly candles only, aggregated to complete UTC daily closes; prices are in USDT. The 23:00 close qualifies only with all 24 distinct, aligned hours and finite positive closes. No interpolation or live fallback. This hourly archive may begin in 2019, while the parent volatility history uses daily API candles that may begin in 2017. Their start dates and gaps can differ; the two sources are not merged.
Method. Thirty consecutive daily log returns, ln(close / previous close), require 31 consecutive complete daily closes. Volatility is their sample standard deviation multiplied by sqrt(365), annualised. Quiet means at or below 30.0%; seven consecutive quiet readings form an episode. Missing or incomplete days reset the 30-return window.
Boundaries and duration. “Naturally ended” means the next day's known volatility exceeded the threshold. “Interrupted” means a missing or incomplete day broke observation; it does not establish a volatility exit. The first valid reading after warmup has an unknown start (left-censored); “ongoing” marks the latest archived boundary (right-censored), even if that archive is old. Unknown boundaries make duration “at least” the observed reading count. Median and longest durations use the entire observed population, including censored episodes; they do not estimate complete spell lengths. Episode price change compares the start and last quiet daily closes.
Seven-day observations. Only naturally ended episodes with a known start qualify. Compare the last quiet close with the close exactly seven calendar days later, requiring every intervening complete day. A dated dash says “history gap” for missing history, “not yet observed” when the endpoint is after the completed-day cutoff, or “unknown boundary” for ineligible episodes. These are observed price changes, not strategy returns or forecasts; they do not show that quiet volatility caused a later move.