Asset Correlations Updated yearly

How assets move together - in normal times and in crashes. Averages for ~2010–2025, crash windows, and a longest-term matrix. Or build your own with any tickers and any window.

Build your own correlation matrixLive

Add 2-8 tickers (any stock, ETF or crypto - plus US housing from the Case-Shiller index), pick the return interval and lookback, and the matrix computes from live price history.

Pearson correlation of overlapping periodic returns (dividend-adjusted closes; "US housing" uses the monthly Case-Shiller national index from FRED, so adding it forces monthly returns). Short windows and few observations make correlations noisy - the observation count is shown so you can judge. Daily data reaches back ~10 years; weekly and monthly go back decades.

Normal periods

Crash periods

−1.0 (inverse)

+1.0 (correlated)

Longest-term correlation matrix

Each cell uses the maximum overlapping data window for that pair. Years shown below each value.

50+ yrs 30–49 yrs 10–29 yrs <10 yrs

Data availability by asset

Database sources + expand

About this page

Correlation matrices for stocks, bonds, gold, real estate and crypto in two states of the world: normal markets and crises (2008, COVID, 2022), plus a builder to compute the matrix for any tickers you choose.

The crisis matrices carry the lesson: correlations between risk assets lurch toward 1 exactly when diversification is needed most. Assets that look independent in calm years sold off together in 2008 and March 2020; historically only Treasuries, cash and occasionally gold kept their distance.

Frequently asked questions

Why do correlations rise in a crash?

Because panics are liquidity events: investors sell whatever can be sold to raise cash or meet margin calls, dragging unrelated assets down together. The fundamentals differ; the seller is the same.

What actually diversifies a stock portfolio?

Historically long-term Treasuries, cash and to a lesser degree gold and managed futures. In 2022 even Treasuries failed the job while commodities passed, a reminder that the diversifier depends on the crisis.

What does a correlation of 0.5 mean?

The two assets share about half their directional movement. Zero means no linear relationship; diversification benefits grow as correlation falls, and turn negative correlations into genuine hedges.