Bitcoin trades nonstop, but a decade of data shows its volatility increasingly follows Wall Street’s clock.
A recent study found that the nine hours from 13:00 to 21:59 UTC accounted for 50.6% of Bitcoin’s daily realized variance between 2022 and 2025, up from 38.4% in 2016-2018. Those hours represent just 37.5% of a full day, suggesting price discovery has become increasingly concentrated while Bitcoin itself continues trading around the clock.
The research, based on 87,672 hourly observations from Kraken’s XBT/USD market between 2016 and 2025, found the shift extends beyond higher activity during American business hours. Bitcoin’s volatility peak now moves when New York changes its clocks and fades when the New York Stock Exchange closes, tying the market’s most active period increasingly closely to the US equity calendar.
That pattern has developed alongside the institutionalization of crypto through regulated futures, publicly traded companies and US-listed investment products. The study stops short of assigning the change to any single channel, leaving ETF creation and redemption, futures activity, market-maker hedging and other institutional flows as potential drivers.
Bitcoin’s volatility clock started following New York
The most compelling evidence comes from what happens when Wall Street’s schedule itself changes.
Researchers used US daylight-saving transitions to distinguish Wall Street’s influence from activity that simply happens to occur during similar hours. The US equity market opens at 9:30 a.m. New York time, which shifts by one hour in UTC when American clocks change, while Asian trading hours and strategies operating on fixed UTC schedules remain unaffected.
During 2022-2025, Bitcoin’s most volatile hour moved from 14:00 UTC during US daylight-saving time to 15:00 UTC during standard time, moving with the shift in the American trading session. The variance-weighted center of the US window also moved 0.33 hours later, a statistically significant change.
No comparable pattern appeared in 2016-2018, when Bitcoin’s intraday volatility profile showed no distinct response to US clock changes.
NYSE holidays provided a second test. On weekday US market holidays during the recent period, Bitcoin’s share of variance occurring during US hours fell by 13.9 percentage points compared with matched trading days, dropping from 55.7% to about 41.9%.
The resulting distribution moved close enough to the 37.5% benchmark for volatility spread evenly through the day that the difference was statistically insignificant.
Those changes make the effect harder to explain through Asian or European trading alone. If the concentration were driven primarily by automated strategies operating on fixed UTC schedules, shifting the New York session would not be expected to move Bitcoin’s volatility peak with it.
The market’s center of volatility has also moved deeper into the American session. A window-free measure used by the study placed Bitcoin’s variance-weighted center at 14.1 UTC in 2016-2018, close to the overlap between London trading and the New York open. By 2022-2025, it had moved to 17.1 UTC, while the study’s concentration index increased more than 40%.
The ETF launch didn’t start the shift
The timing complicates the assumption that US spot Bitcoin ETFs suddenly transformed the market after their January 2024 debut.
A statistical change-point analysis identified November 2021 as the single major break in Bitcoin’s US-hours volatility trend. Researchers found no comparable local break around either the December 2017 introduction of CME Bitcoin futures or the approval of US spot ETFs in January 2024.
The distinction becomes stark depending on how the data are measured.
A broad comparison of all observations before and after the ETF launch produces an apparent 9.6-percentage-point increase in Bitcoin’s US-hours variance share. Restricting the analysis to a symmetric 12-month window around the event reduces the change to just 0.1 percentage point, with no statistically significant break.
CME futures show the same problem. An apparent 7.1-point increase using the full before-and-after sample shrinks to 0.3 point when the analysis focuses on the period immediately surrounding the launch.
The study tested that statistical problem by selecting 1,000 random dates and treating each as though a market-changing event had occurred. The broad before-and-after method produced statistical significance at the 0.1% level for all 1,000 fake events, showing how a long-running trend can make almost any chosen milestone appear responsible for a structural change.
The symmetric approach produced the same result in just 4.7% of cases.
That leaves institutionalization as a broader explanation rather than assigning the shift to one product. The findings identify the US cash session as the timing anchor while leaving unresolved whether ETF flows, derivatives activity, market-maker hedging or other institutional channels are doing the most to pull Bitcoin’s price discovery toward New York.
Crypto’s 24/7 market is developing office hours
The shift is also appearing across the weekly calendar.
Bitcoin’s weekend-to-weekday volatility ratio fell from 0.96 in 2016, when Saturdays and Sundays were almost as volatile as weekdays, to 0.60 in 2024 and 0.64 in 2025. The equivalent trading-volume ratio dropped from 0.78 to 0.43 in 2024 before edging to 0.46 last year.
The trend extends beyond Bitcoin. Similar increases in the US-hours share of variance appeared in six of seven other long-listed crypto assets tested on Kraken, including Ethereum, $XRP, Solana, Cardano, Dogecoin and Chainlink.
$XRP’s share rose from 37.2% in its earliest two-year period to 46.2% in its latest, while ETH moved from 41.8% to 48.2%. Litecoin was the only asset without a statistically significant trend.
For trading firms, the change carries a more immediate consequence than the symbolism of a 24-hour market adopting Wall Street hours. Risk models that assume volatility is distributed relatively evenly across the day can understate exposure during the US session and overstate it overnight.
Thinner weekends can further widen the gap between continuously traded spot crypto and futures or options linked to traditional-market calendars, complicating hedging when institutional venues are less active.
The study relies primarily on one exchange and stops at the end of 2025, leaving multi-venue confirmation as the next test. Order-book and trade-level data would also be needed to determine how much of the pattern comes from ETF creations and redemptions, futures positioning, market-maker hedging or other institutional flows.
For trading desks, that distinction could determine whether Bitcoin’s emerging volatility clock becomes useful for margin requirements, liquidity provisioning and hedging schedules. A market that remains open every hour now gives them a narrower question to manage: which hours increasingly carry the risk.
