Bitcoin–Nasdaq correlation: a reproducible measurement guide

Bitcoin–Nasdaq correlation describes the linear association between two specified return series over a specified window. It is not a permanent property of Bitcoin. This guide explains how to define the measurement; the July 2026 claim audit addresses an earlier numerical headline separately.

Define the series

Choose one equity proxy: the Nasdaq Composite, Nasdaq-100, QQQ or an index future. They are different instruments. Document the source, currency, price field and whether distributions or other adjustments are included. Choose one Bitcoin venue or a documented index rather than changing sources across the sample.

Bitcoin trades through nights and weekends while US equities have trading sessions and holidays. Align both observations to the same session-close instant, using an exchange calendar and America/New_York daylight-saving rules before converting to UTC. A daily Bitcoin midnight close is not automatically comparable to a US equity close. Use matched sessions; do not forward-fill equity weekends into artificial zero returns.

Calculate and disclose

Calculate returns between consecutive matched observations using either simple returns P(t)/P(t−1)−1 or log returns, consistently for both series. Pearson r is the sum of paired demeaned products divided by the product of their root sums of squared deviations. Zero variation makes it undefined, not zero correlation.

Report the actual paired count, dates, sampling frequency, missing observations and window. For example, 30 matched trading sessions differ from 30 calendar days. Comparing 30-, 60- and 90-session windows is a sensitivity analysis, not three independent confirmations. Returns that overlap across rolling windows are dependent.

Interpret cautiously

Correlation near zero means little measured linear association; it does not prove independence. A high positive reading does not establish that equities caused Bitcoin's move. Shared exposures, unrelated shocks, sample selection and outliers can all influence the statistic. Compare scatter plots and the stability of results before assigning an economic explanation.

A reported macro regime is an interpretation requiring additional evidence. Correlation alone does not establish a transmission channel, a hedge ratio or a reliable forecast. This guide therefore makes no current correlation estimate and does not prescribe a position size.

For the statistical definition and limitations, see NIST's correlation reference and scatter-plot guidance. For a separate example of why time cutoffs matter, see the reproducible Bitcoin baseline study. Revised 24 September 2026.