Bitcoin Fell on 92% of Stocks' Worst Days Since 2022, Sentora Finds

Bitcoin fell on 92% of the U.S. stock market's worst days since 2022, according to Nov. 13 research from Sentora, a finding that directly challenges the long-running argument that crypto offers protection during equity selloffs. Ether performed worse still, declining on 95% of those sessions. The numbers come from a study comparing Bitcoin and Ether returns against U.S. stock indices from May 2018 through November 2026.

Sentora examined the diversification claim after the Nov. 5 selloff. Over that window, the S&P 500 lost 4.4% and the Nasdaq-100 fell 6.4%, while Bitcoin dropped 5% and Ether declined 7.2%. In plain terms, both major cryptocurrencies lost roughly as much as the indices they were supposed to hedge against.

How often did crypto fall when stocks did?

Across the full sample, Sentora identified 107 sessions within the S&P 500's worst-performing 5%, each involving a loss of at least 1.74%. Bitcoin fell in 81% of those observations; Ether declined in 79%. On ordinary trading days, each cryptocurrency fell roughly 49% of the time.

That gap is the study's central point. Calculated from the reported figures, the frequency of losses during severe equity declines ran 32 percentage points higher for Bitcoin and 30 points higher for Ether than on normal days. Restricting the analysis to 2022 onward, using that period's own worst-5% threshold across 61 observations, produced the headline 92% and 95% loss frequencies.

The correlation data, notably, tells a calmer story than the tail-risk data. Over the entire period, Bitcoin's correlation with the S&P 500 measured 0.29 and Ether's 0.31. Both readings were slightly negative in 2018–19 but reached roughly 0.50 in 2026. Average correlations can look moderate while the worst days still cluster; the takeaway is that diversification claims built on averages ignore what happens when investors need the hedge most.

Which asset held up better?

Ether lost more ground than Bitcoin during equity stress. Since 2022, Ether underperformed Bitcoin on 79% of the worst equity sessions. Median dollar losses stood at 3.6% for Bitcoin and 5.6% for Ether — a two-percentage-point difference on the reported medians.

The ETH/BTC exchange rate reinforces the pattern. Its median return was negative 1.52% on the worst equity days, against positive 0.67% on the best, a 2.19-percentage-point gap. Sentora said the pattern survived the removal of March 2020 from the analysis. Bitcoin's relative strength did not mean it avoided losses: both assets posted negative median dollar returns during the recent sample's worst equity sessions.

Why do U.S. trading hours matter?

Sentora's hourly analysis covered the period since U.S. spot Bitcoin ETFs launched in January 2024. The firm defined its cash-session window as 10 a.m. through the actual New York stock-market close, excluding the opening half-hour to use complete hourly observations.

That window represented just 17.1% of observed hours, yet it accounted for 33% of Bitcoin's price variation and 30% of Ether's. In a typical U.S. trading hour, Bitcoin's average absolute move ran 2.09 times its weekend-hour move; Ether's ratio was 1.78. Removing the largest 1% of hourly moves barely changed either result. The study measured the timing of volatility only — it did not establish that stock trading caused the crypto moves, a caveat worth holding onto.

What did the weekly flows show?

The November selloff carried substantial losses for leveraged traders. CoinGlass recorded 24-hour crypto liquidations of approximately $1.08 billion on Nov. 5, including $954.38 million in long positions and $127.30 million in shorts.

Sentora's accompanying weekly briefing reported $3.09 billion in net Bitcoin exchange inflows and $1.10 billion for Ether — a combined $4.19 billion. The firm noted that inflows generally leave more coins available to trade, though the transfers do not prove that holders sold them. Stablecoin supply, meanwhile, fell by $4.75 billion to $287.77 billion during the week, after declining every day for more than two weeks.

For allocators weighing crypto as portfolio insurance, the operative question is no longer whether correlations are moderate on average, but whether the asset holds up on the days that define a portfolio's worst year — and on the evidence since 2022, it has not.