How correlated is the crypto market?
We measured how closely the 10 largest cryptocurrencies actually moved together, using real daily closes rather than intuition. The short answer: far more than a diversified portfolio would like.
average pairwise correlation of the 10 largest cryptocurrencies over the last year, on a scale where 1.0 is lockstep and 0 is unrelated.
Key findings
- The 10 biggest cryptocurrencies averaged a 0.80 pairwise correlation over the last year (365 trading days). The market trades much more like one asset than like 10.
- The most alike pair was ETH/SOL at 0.89; even the least alike, BNB/DOT, sat at 0.67 β far from independent.
- Over last 5 years the average was 0.68, so this is not a one-year fluke β the assets have moved together for years.
- Diversification ceiling: at 0.80 correlation, an equal split of all 10 is only ~9.4% less volatile than holding one. The absolute floor, even with infinite coins, is ~10.5% (βΟ).
- Were they uncorrelated instead, those same 10 coins would cut volatility by ~68.4% β the gap between that and 9.4% is what correlation quietly costs a "diversified" bag.
- Most volatile of the group: DOGE at ~189% annualised; calmest: BTC at ~81%. Deepest crash in our data: DOT, 98.6% from peak.
The matrix
Every pair of the 10 majors over the last year, from daily returns. Deeper colour is a tighter link. There is nowhere in this grid that two of these assets genuinely go their own way.
| Asset | BTC | ETH | XRP | BNB | SOL | DOGE | ADA | LINK | LTC | DOT |
|---|---|---|---|---|---|---|---|---|---|---|
| BTC | β | 0.88 | 0.86 | 0.76 | 0.85 | 0.78 | 0.79 | 0.81 | 0.73 | 0.67 |
| ETH | 0.88 | β | 0.84 | 0.78 | 0.89 | 0.83 | 0.83 | 0.88 | 0.76 | 0.74 |
| XRP | 0.86 | 0.84 | β | 0.73 | 0.84 | 0.84 | 0.83 | 0.84 | 0.75 | 0.74 |
| BNB | 0.76 | 0.78 | 0.73 | β | 0.77 | 0.74 | 0.74 | 0.76 | 0.70 | 0.67 |
| SOL | 0.85 | 0.89 | 0.84 | 0.77 | β | 0.84 | 0.84 | 0.88 | 0.79 | 0.77 |
| DOGE | 0.78 | 0.83 | 0.84 | 0.74 | 0.84 | β | 0.87 | 0.86 | 0.81 | 0.81 |
| ADA | 0.79 | 0.83 | 0.83 | 0.74 | 0.84 | 0.87 | β | 0.86 | 0.78 | 0.83 |
| LINK | 0.81 | 0.88 | 0.84 | 0.76 | 0.88 | 0.86 | 0.86 | β | 0.80 | 0.81 |
| LTC | 0.73 | 0.76 | 0.75 | 0.70 | 0.79 | 0.81 | 0.78 | 0.80 | β | 0.82 |
| DOT | 0.67 | 0.74 | 0.74 | 0.67 | 0.77 | 0.81 | 0.83 | 0.81 | 0.82 | β |
Why more coins barely helps
Diversification only removes the risk that assets do not share. For equal-weight holdings that all correlate at Ο, a portfolio's volatility relative to a single asset is β(1/N + (1β1/N)Β·Ο). As you add coins the 1/N term shrinks toward zero, but the (1β1/N)Β·Ο term does not β it converges on Ο. So the most volatility any number of these coins can ever remove is 1 β βΟ, and at 0.80 that ceiling is about 10.5%.
| If the 10 coins correlated at⦠| Volatility cut vs one coin |
|---|---|
| 0.00 | 68.4% |
| 0.20 | 47.1% |
| 0.40 | 32.2% |
| 0.60 | 20% |
| 0.80β crypto today | 9.4% |
| 0.90 | 4.6% |
Assumes equal weights and equal volatilities β a simplification, but the direction is right and the ceiling is real. The interactive tool below uses each asset's true covariance.
Methodology
- Data: our own daily closing prices for the 10 assets, through 2026-08-10. Correlations use the 365-day window for the headline figure; the last 5 years window is shown for context.
- Method: Pearson correlation of daily returns, not prices β two assets in a shared uptrend have correlated price levels almost by definition, so measuring levels would overstate the link.
- Windows: each window uses only assets that existed for all of it; a coin younger than the window is dropped, never back-filled.
- Volatility & drawdown figures are annualised standard deviation of daily returns and peak-to-trough falls, bounded by the data we hold β not all-time claims.
Cite or share this study
The figures update as new data comes in. If you use them, a link back to this page keeps the citation live and lets your readers see the current numbers.
TheCryptoTools (2026). How Correlated Is the Crypto Market? Retrieved from https://thecryptotools.com/research/crypto-correlation/
Prefer an interactive version your readers can play with? The correlation matrix tool lets anyone pick their own coins and windows, and the embeddable widgets can drop a live calculator straight into an article.
Frequently asked questions
οΌAre cryptocurrencies correlated with each other?
Strongly. Over the last year the 10 largest crypto assets moved with an average pairwise correlation of 0.80 β where 1.0 is lockstep and 0 is unrelated. In practice the market trades far more like one asset than like 10 independent ones.
οΌDoes holding many coins diversify a crypto portfolio?
Much less than most people assume. At a correlation of 0.80, an equal split of 10 of these coins is only about 9.4% less volatile than holding one of them. Even an infinite number of equally correlated coins would cut volatility by at most 10.5% β that is the mathematical ceiling, βΟ.
οΌWhich crypto pair is most and least correlated?
Over the last year the most alike pair was ETH/SOL at 0.89, and the least alike was BNB/DOT at 0.67. Even the least correlated major pair is far above zero.
οΌDo these correlations hold in a crash?
No β and that is the most important caveat. Correlations rise sharply in a sell-off: coins that drift apart in calm markets tend to fall together in a bad week, exactly when the diversification was supposed to help. Treat a low correlation as a fair-weather property, not a guarantee.