Bitcoin just hit a decade-low S&P 500 correlation, but the daily data tells a different story

Buyers holding Bitcoin alongside US shares can not inform whether or not their portfolio has develop into safer merely from a unfavourable correlation between the property’ worth paths. A historic comparability by way of September 4 reveals why: Bitcoin and the S&P 500 can transfer additional aside over a yr whereas nonetheless posting losses on most of the similar buying and selling dates.

Crypto asset supervisor Bitwise’s September 7 Market Compass described its 260-day correlation because the lowest since 2015. Its chart measures the logarithms of worth ranges. In CryptoSlate’s calculation utilizing FRED information, that measure was about −0.62. Correlation between proportion returns over 260 widespread buying and selling observations was +0.40.

The 2 figures measure various things. The primary tracks how the 2 worth paths moved over the interval. The second tracks whether or not their proportion positive aspects and losses tended to maneuver collectively. Neither reveals by itself that Bitcoin will offset losses when shares fall.

The FRED costs are additionally recorded at totally different occasions of day. The comparability makes use of obtainable every day observations, not simultaneous closing costs, so it can not exactly measure how Bitcoin responds to an equity-market shock.

Two costs can diverge whereas losses overlap

Bitwise explicitly labels its chart as a log-level calculation and credit Bloomberg and Bitwise Europe. The calculation describes medium-term motion within the two worth paths. It isn’t a daily-return statistic, nor ought to it’s handled as a mistaken model of 1.

For somebody holding each property, portfolio positive aspects and losses rely upon returns. A unfavourable relationship between worth ranges doesn’t assure unfavourable covariance between returns. In less complicated phrases, two property can drift in several instructions over months whereas nonetheless rising or falling collectively on particular person days. Their volatility and portfolio weights additionally decide how a lot that co-movement impacts the portfolio.

The FRED Bitcoin observations and S&P 500 series present the distinction. Utilizing the final 260 widespread worth dates by way of September 4 produces a log-level correlation of −0.6161. Utilizing 260 proportion returns, from August 25, 2025, by way of September 4, 2026, produces +0.3957. Calculating 260 returns requires one extra beginning worth.

The unfavourable log-level consequence resembles the pattern in Bitwise’s report, however it’s a proxy somewhat than an actual copy of its Bloomberg calculation. FRED’s public S&P historical past covers solely 10 years, so it can not independently set up the complete “since 2015” rating. That historic declare stays Bitwise’s.

The consequence additionally modifications with the time interval. In the identical asynchronous information, return correlation was +0.14 over 30 observations, +0.32 over 60, +0.34 over 126 and +0.13 over 500, all ending September 4. As a result of these home windows overlap, they aren’t impartial proof of a set long-term correlation.

A hypothetical portfolio reveals what these variations can imply for an investor. Changing 5% of the fairness allocation with Bitcoin elevated measured annualized volatility over the 260-return window, from 12.68% to 13.01%. The biggest peak-to-trough decline additionally grew, from 9.10% to 9.98%.

Over 500 returns, the consequence reversed. The identical Bitcoin allocation modestly lowered each measured volatility and most drawdown.

Each home windows finish September 4, 2026. CryptoSlate calculations used the dated Bitcoin CSV and S&P 500 CSV, with weights reset at each widespread commentary. Charges, taxes, and dividends had been excluded. The S&P sequence is a worth index, so these should not whole investment-return comparisons. Most drawdown is the biggest decline from a previous portfolio peak throughout the pattern.

The desk makes use of costs recorded at totally different occasions somewhat than an executable technique rebalanced at one shared closing time.

The 2 home windows produced totally different outcomes. Bitcoin elevated measured volatility and drawdown within the 260-return pattern, then modestly lowered each within the 500-return pattern. A low correlation can diversify one supply of danger with out making a portfolio much less unstable than equities alone.

What occurred when equities fell

However full-period volatility is just a part of the danger: traders additionally have to know what occurred to Bitcoin when equities fell. The S&P 500 fell on 115 widespread dates throughout the 260-return pattern. Bitcoin fell on 76 of them and averaged a 1.03% decline throughout all 115 equity-down observations. Correlation inside that subset was +0.29.

Bitcoin fell much more persistently alongside shares throughout bigger declines. On the 25 observations when the S&P 500 fell greater than 1%, Bitcoin declined on 22 and averaged a 2.59% loss. There have been solely three observations when equities fell greater than 2%. Bitcoin fell on all three, averaging a 4.23% decline.

The outcomes describe what occurred on this pattern, not what precipitated the strikes. Three declines higher than 2% are additionally far too few to estimate how reliably Bitcoin would behave throughout a stock-market crash. Altering the window, loss threshold, or measurement time can considerably change outcomes from such a small pattern.

What the observations do present is that unfavourable pattern correlation coexisted with repeated losses in each property. Buyers can not translate a chart of diverging worth paths instantly into an assumption that one holding will offset the opposite on a foul day.

FRED data its Coinbase Bitcoin commentary at 5 p.m. PST. Its S&P 500 commentary represents the US market shut, sometimes 4 p.m. ET. The 2 costs due to this fact share a calendar date however not a timestamp.

The calculations retain dates with legitimate observations for each property, then measure modifications between consecutive widespread dates. Bitcoin strikes throughout intervening weekends or holidays are included within the interval. Inventory costs should not stuffed ahead to create extra buying and selling observations.

The timing distinction can change the consequence considerably. Shifting Bitcoin’s observations again one calendar day modifications the 260-return correlation to +0.14; transferring them ahead modifications it to +0.03. Over 30 observations, these shifts produce −0.30 and −0.37, respectively.

These shifts take a look at attainable lead-and-lag results; they don’t create synchronized costs. The sensitivity additionally means +0.40 shouldn’t be learn as the precise correlation between Bitcoin and equities on the similar second. Measuring that may require costs recorded at matching occasions.

Bitwise additionally stops wanting describing the decoupling as a hedge. Its report says a repeat of Bitcoin’s roughly 9,800% rise from mid-2015 to late 2017 is unlikely and warns that decouplings normally don’t final.

For traders holding each property, the numbers reply two separate questions. The unfavourable log-level correlation reveals that Bitcoin and the S&P 500 adopted more and more totally different worth paths over the interval. Their returns present that the 2 property nonetheless misplaced cash collectively on most of the similar days.

Portfolio danger is dependent upon the second query in addition to allocation dimension, volatility, measurement time, and the interval being studied.