Bitcoin Fell in 92% of S&P 500's Worst Slumps Since 2022

Bitcoin fell in 92% of the S&P 500's worst 5% return observations since 2022, and Ether fell in 95% of them, according to research by Sentora. The recent sample covers 61 observations of the index's sharpest losses, not every session when the S&P 500 declined.

The full-sample numbers tell a more measured story. Across more than 2,000 trading intervals from May 2018 through November 2026, Bitcoin's correlation with the S&P 500 was 0.29 and Ether's was 0.31. In 2026, both correlations ran near 0.50, marking a step up from the long-run average.

What do the two measures actually show?

The statistics capture different dimensions of crypto's link to equities. Correlation measures how closely returns move together across an entire sample, while the 92% and 95% figures isolate the index's worst-return episodes — the tail risk that matters most to portfolio construction.

In the full sample, the S&P 500 recorded 107 worst-return observations, each representing a loss of at least 1.74%. Bitcoin fell in 81% of those episodes and Ether in 79%. In the middle 90% of index returns, both assets declined roughly 49% of the time — close to a coin flip, and materially calmer than the tail behavior.

The asymmetry carries direct implications for risk models. An asset with a 0.29 correlation can still be a poor hedge if it sells off almost every time equities crash, because diversification measured by average correlation understates losses in the scenarios investors fear most.

How bad were the drawdowns?

During the recent sample's worst equity observations, Bitcoin's median U.S. dollar return was −3.6%, versus −5.6% for Ether. Ether underperformed Bitcoin in 79% of those episodes, confirming its higher-beta profile relative to the larger asset.

Where does crypto volatility actually happen?

The realized-variance breakdown showed a pronounced concentration during U.S. stock-market hours. From the January 2024 spot Bitcoin ETF launch through the study cutoff, the window from 10 a.m. ET to the market's actual close — normally 4 p.m. ET — accounted for 17.1% of observed time but 33% of Bitcoin's realized variance and 30% of Ether's.

Sentora Research cautioned against overreading that figure. "This is a finding about crypto's own volatility timing. It does not mean one-third of its price variation was caused by stocks," the firm said.

The findings extend earlier coverage of Bitcoin's behavior during deep global stock declines by adding a separate S&P 500 sample and an intraday realized-variance measure. As spot ETF flows tie crypto more tightly to U.S. tradable hours, the 2026 correlation readings near 0.50 suggest the relationship is tightening, and allocation models built on pre-2024 data may understate joint downside risk.