Digital assets are frequently described as uncorrelated with traditional markets. The description holds in calm conditions and breaks down precisely when the diversification would be most useful.
Correlation is not a fixed property
Correlation measures how two assets moved together over a chosen window, and the answer changes with the window selected.
Over long, quiet periods, crypto and equities are driven by different things and the measured relationship is weak.
During sharp drawdowns the same calculation over a shorter window produces a much stronger relationship, using the same assets and the same method.
A single published correlation figure is therefore a summary of a period rather than a stable characteristic, and it can be made to say almost anything by choosing the dates.
Stress transmits through positions, not fundamentals
Investors hold both asset classes in the same portfolios, and when losses appear in one, risk limits and margin requirements apply to the whole book.
Meeting those demands means selling whatever can be sold, which is often the most liquid holding rather than the one causing the problem.
An asset that trades around the clock is unusually easy to sell at the moment a margin call arrives, which puts it near the front of the queue for liquidation.
Nothing about the assets has converged. The connection is the balance sheet of the people who own them.
Leverage compresses the timescale
Crypto markets carry substantial leverage, and liquidations execute automatically without regard to conditions.
A shock arriving from elsewhere can therefore trigger forced selling within minutes, rather than over days of discretionary decisions.
The speed makes the co-movement more visible, because both markets fall inside the same short window rather than one following the other over subsequent sessions.
Crypto also trades continuously, so it absorbs news released while equity markets are closed and often registers the reaction first, which can make it look like the source of a move it merely reported early.
Macroeconomic conditions affect both directly
Interest rate expectations change the appeal of any asset that produces no income, which includes most digital assets and long-duration equities alike.
Both therefore react to the same policy signals, and the reaction runs in the same direction more often than not.
This channel operates continuously rather than only in crises, and it strengthened as institutional participation in crypto markets grew.
What this means for diversification claims
An asset that behaves independently most of the time and moves with everything else during dislocations offers less protection than its long-run correlation implies.
Averaged correlation figures conceal this, because the calm periods dominate the sample and the episodes that matter are brief.
The more useful measure is how the assets behaved together during past drawdowns specifically, which is a different and considerably less flattering number.