Liquidity, leverage and crypto-native blowups. Fed pivots, rate shocks, LUNA, FTX, forced liquidations and ETF flows caused every 20%+ swing.
Headlines without a money channel. Wars, coups, downgrades, shutdowns — even the 2024 halving — mostly got a 1–5% wobble that was gone within days.
Geopolitics only bites through oil → inflation → the Fed. If a war doesn't change rate expectations or dollar liquidity, BTC ignores it.
of the biggest days had no news at all. Of every day BTC closed ±6% or more since Oct 2021, had no identifiable catalyst — just leverage and positioning. Multi-day swings are different: only lacked a driver.
Every number here is computed live from the three tables and the 5-year price scan below. Hover or tap any bar to see the events behind it.
What drove each of the days BTC closed ±6% or more since Oct 2021
Average same-day BTC move of the big movers, by how expected the event was
How much of each table the market saw coming
Geopolitics only moves BTC through oil, inflation and the Fed. No money channel, no move.
2026 Iran war: Hormuz closed → Brent $73 → $120 → inflation back above 4% → the Fed's first hike since 2023 (Sep 16).
Examples: Hamas attack, Wagner mutiny, Assad's fall, Maduro's capture, the halving.
Measured on (Binance daily closes). BTC trades the surprise and the guidance, not the decision itself.
The Fed publishes its meeting calendar long in advance, and every CPI and jobs report shifts the CME FedWatch odds.
Fed officials stop speaking publicly.
The meeting minutes can move BTC on their own: the Jan 5, 2022 minutes started a −30% slide.
One bar per meeting; height = the market-implied odds of the outcome the Fed delivered, about a day before. The decision itself is almost never the surprise — the guidance is: the Dec 2024 cut was 97% priced and BTC still fell 5.6% on the dot plot.
Buy the rumour, sell the news: BTC rallied into every cut and dipped on most cut days, while hike days were mostly green. Hover a cell for every meeting. Small samples (12 hikes, 6 cuts) — patterns, not rules.
Every US session since Oct 2021 where the S&P 500 fell 2.5%+ or the Nasdaq 3%+, against BTC's move the same day. Dots below the dashed line: BTC fell harder than stocks.
Correlation of daily moves, per year (1 = moves in lockstep, 0 = unrelated)
The biggest BTC losses on stock-crash days were crypto's own blowups landing the same day (LUNA, Celsius, 10/10). On the two worst stock days of 2025, BTC had already dropped the evening before — it trades 24/7 — and then held steady while stocks crashed.
Median BTC move around crypto adoption, crypto crackdowns, bank failures and government / fiat trust shocks since Oct 2021. Hover a cell for every event.
Policy news barely moves BTC: adoption headlines are usually sold, crackdowns cause a one-day dip. The “hedge” shows up only when the authorities answer a crisis with money: the Fed's bank backstop (+7.5% that day), the freeze of Russia's reserves (+14.5%), the Treasury's long-bond buybacks (+7.1%). Bank failures without a rescue pushed BTC down. Medians are used because a few same-day moves overlap with other news (e.g. China's Feb 6, 2026 ban landed on the rebound after the Warsh crash).
Median BTC move around big AI news, crypto launches & protocol upgrades, quantum-computing scares, Strategy's largest BTC buys and government BTC sales since Oct 2021. Hover a cell for every event.
None of these are catalysts on the day. They move BTC before: launches and upgrades were bought into (BTC up the week before 16 of 24, then flat on the day), and Strategy announces its big buys after rallies (up the week before 12 of 16): it raises money when its own stock is high. Quantum headlines dip BTC on the day (down 7 of 10) but the scare is gone within a week: Willow day −3.8%, the week after +9%. Germany's 50K BTC sale was priced before the coins hit the market: −11% in the week into July 8, 2024, then +14–15% in the week after the last coins left. ChatGPT's launch day (+4.4%) was really Powell's “slow the hikes” speech the same afternoon. Medians are used because some days overlap with bigger news (Core v30 shipped on the 10/10 crash day).
Every Nvidia quarterly report since Nov 2021. Nvidia reports after the US close, so both moves are measured over the same 24 hours: 4pm ET on report day to 4pm ET the next day.
BTC does track the Nasdaq on big stock-crash days, but Nvidia's own news doesn't carry over: the dots sit in all four corners. BTC's average move after earnings is the same as on any other day. Where AI does touch crypto is miners renting their power and data centres to AI companies, and even those deals didn't move BTC (see AI news above).
Deribit, the biggest BTC options exchange, settles options at 08:00 UTC on the last Friday of each month; the quarterly expiries (Mar, Jun, Sep, Dec) are the largest. BTC's daily candle that Friday vs every other Friday since Oct 2021.
Expiry days are quieter, which fits the idea that dealers hedging around big strikes “pin” the price into settlement. They also closed red more often than other Fridays, but 59 days is a small sample. The weekend after a quarterly expiry is no wilder than any other weekend (avg 1.6% vs 1.7%). The popular “max pain” price can't be checked here: it needs past open interest per strike, which isn't public in daily form. Deribit expiry rules.
For each big mover: how much of BTC's move was still ahead one hour after the news broke, and how long half the move took (hourly candles, move measured to BTC's extreme within 7 days). Click a row in table 1 for the full signal timeline.
Reading it: the news itself was usually enough. Only the moves that happened in the first hour were “too late”: war news (Russia's invasion, Iran's drone attack) and a surprise court ruling (Grayscale). And in those wars the warning was public days earlier. Warnings beat the news in slow-motion blowups (FTX: CZ's tweet 2 days earlier; LUNA: the first depeg) and in macro shocks with a visible first domino (China's rare-earth controls before 10/10, the BOJ hike before the yen unwind, a hot CPI before the Celsius week). A warning only counts if acting on it beat waiting for the news by at least 1 point. Hindsight caveat: a warning counts only if it was public at a known hour and pointed the way BTC went; warnings older than 14 days are shown as context, because acting on them mostly caught unrelated drift. Not trading advice.
The biggest BTC moves since Oct 2021 and what triggered them. Newest first. The main row is the full move over days; the Instant reaction row under it is the same-hour / same-day move only. Expected? shows what markets priced beforehand (prediction-market or FedWatch odds where they existed). Under it, Options shows the daily move BTC options priced the day before the trigger — and whether traders were bracing for a catalyst (more than usual) or calm. Click any row for the full story.
| Date | Event | % |
|---|
Wars, conflicts, political shocks and even the halving: events where you'd expect a big move — and BTC barely flinched (or went the other way). Click any row for why.
| Date | Event | % |
|---|
Found by scanning all 1,826 Binance daily candles since Oct 2021 for ±6% daily closes, ±8% intraday wicks, and rolling 7-day (±15%) and 30-day (±25%) swings, then matching every hit against known news. These are the ones where nothing credible explains the move. Claimed cause is what got blamed at the time; Best guess is the likely mechanics. Click any row for why the story doesn't hold.
| Date | Event | % |
|---|
Prices are approximate (daily closes / intraday extremes from public market data; 2026 figures from contemporary news reports). Not financial advice.