The question

Bitcoin is often described as a leveraged bet on technology stocks: when the Nasdaq sells off, bitcoin sells off harder, and when tech rallies, bitcoin rallies more. If that is true, bitcoin is not a separate asset at all but a high-beta version of one you may already own. If it is only sometimes true, the interesting question becomes when, and how to tell. This page keeps a running answer from daily closes, recomputed by a scheduled job after each US session, and the meter shows the same statistics live.

Why it matters

A holding only diversifies a portfolio when its returns move independently of the rest. A tech-heavy investor who adds bitcoin during a coupled regime adds volatility without adding diversification. The same investor during a decoupled regime gets a different asset. Neither state is permanent, so a single correlation figure is less useful than knowing how often each regime occurs, how long it lasts, and how abruptly it ends.

Method

The sample is — trading sessions from — to —. Bitcoin trades every calendar day and the index does not, so the two series are first intersected on the dates each actually traded; otherwise bitcoin's Monday return spans one day while the index's spans three. Returns are daily log returns. Correlation, beta and R² are computed over the last 30, 90 and 252 sessions. Regimes are read off a rolling 90-session correlation: above 0.5 counts as coupled, below 0.2 as decoupled, and the band between as loosely coupled. A break is the fall in that rolling correlation over any 20-session span; the three largest are listed. Volatility is the sample standard deviation of 30 sessions of log returns, annualized by √252. Drawdowns run on each asset's own history, since a drawdown belongs to one price path.

What the numbers show

The rolling 90-session correlation, with the coupled band shaded green and the decoupled band red:

The correlation was above 0.5 on — of sessions and below 0.2 on —, with the remaining — in between. The longest coupled stretch was —; the longest decoupled stretch was —. Over the full year the beta against the Nasdaq is — with an R² of —, meaning the index explains about — of the day-to-day variance in bitcoin's returns. Against the S&P 500 the 90-session correlation is —.

When the coupling broke

The largest falls in the rolling correlation over any 20-session span. A large fall means the two stopped moving together quickly, whichever direction each then went.

The shape of the relationship

Each square is one of the last — sessions: the Nasdaq's return across, bitcoin's up. The line is the least-squares fit; its slope, —, is the 90-session beta. The axes are scaled separately because bitcoin's returns run several times the index's, so read the slope as a shape rather than an angle.

Volatility

Bitcoin's 30-session realized volatility is — annualized against the Nasdaq's —, a ratio of —. Over the year the averages were — and —. This is the "leveraged" half of the leveraged-bet claim: even when the two move together, bitcoin moves several times as far.

Drawdowns

Bitcoin is — from its peak in this window. Its deepest drawdown was —; the Nasdaq's was —. The two closed in the same direction on — of the last 252 sessions and — of the whole sample.

In words

These sentences are generated from the figures above by a fixed template. They add nothing the numbers do not say.

Caveats

The sample is about a year of daily closes, which is short: a single episode can dominate the regime shares. Correlation is not causation; nothing here says which side leads or why they move together. The 0.5 and 0.2 thresholds are conventions, chosen to be legible rather than derived. Bitcoin's daily close is a UTC-day bar while the index closes at 16:00 New York time, so the pairing is approximate by hours. Prices come from Financial Modeling Prep and CoinGecko and may be delayed, revised, or wrong.

Standing

Market data for reference only. Nothing on this page is investment advice, a recommendation, or a solicitation to trade. The regime labels describe the past; they say nothing about the next session.