Correlation
Correlation matrix calculator who moves together
Empirical correlation from your closes. The heatmap is the number in each cell, not a color standing in for one.
BTC / ETH / SOL
Average pairwise 0.87
Tightest pair 1-2 at 0.90. Loosest pair 2-3 at 0.84.
- Average pairwise
- 0.87
- Tightest
- 0.90
- Loosest
- 0.84
- Assets
- 3
Off-diagonal mean
1-2
2-3
89 returns
Correlation. Cyan is positive, pink is negative. The number is the coefficient.
| BTC | ETH | SOL | |
|---|---|---|---|
| BTC | 1.00 | 0.90 | 0.87 |
| ETH | 0.90 | 1.00 | 0.84 |
| SOL | 0.87 | 0.84 | 1.00 |
| BTC | ETH | SOL | |
|---|---|---|---|
| BTC | 1.00 | 0.90 | 0.87 |
| ETH | 0.90 | 1.00 | 0.84 |
| SOL | 0.87 | 0.84 | 1.00 |
- Each row is daily, 365 periods a year. Annualized figures use 365 periods.
- 90 closes in the series. Computed by Thrive's portfolio engine. Rounded to the digit shown.
Correlation is the hedge you think you have
A 0.9 between two coins means a second position is almost the same bet. A 0.2 means the second name can actually change the book. The sign matters: negative correlation is the rare diversifier.
Correlation moves. A 90-close number can hide a 30-close break. The pair pages on this site print both windows for the names traders actually search.
How this page runs it
Thrive's portfolio engine builds the empirical correlation of the arithmetic returns. The matrix is symmetric, with ones on the diagonal. You can paste from 2 to 6 columns.
Rows must be the same observations. If you mix a weekend crypto series with an ETF, drop the days the ETF did not trade or the correlation is nonsense.
How to run it
- Paste aligned closes, one column per asset.
- Read the off-diagonal cells. Color is only a guide. The number is the result.
- Open a pair page for the live 30-close and 90-close print.
- If two names sit above 0.8, stop counting them as two bets.
FAQ
People also ask
Above 0.8, the second asset barely changes portfolio vol. Between 0.5 and 0.8 you still get some help. Under 0.3 the mix does real work, if the vols are comparable.
This page is the empirical Pearson matrix, which is what mean-variance and the optimizers on the next pages consume. Rank correlation is a different question.
Window length, close source, and whether weekends are included all move the number. This page states the row count. Match that before you call a disagreement an error.
Not by itself. You also need the two volatilities. Beta is correlation times vol of the asset over vol of the benchmark. The beta calculator does that step.
Related
Efficient frontier calculator
Two or more close series. The engine estimates means and covariance, then lays out the mean-variance frontier: the mixes that take the least vol for each return.
Portfolio beta calculator
Beta and information ratio of one series against another. Ethereum against Bitcoin is the sample. Swap in your book and the benchmark you claim to beat.
Volatility calculator
Close-to-close realized volatility from the closes you paste, de-meaned, then annualized with the calendar you pick. One asset or a whole list.
Trading journal
The ratio on a pasted series is a lab result. The journal is the book you actually traded.
Bitcoin and Ethereum
The live pair, 30 closes and 90 closes.
Bitcoin and SPY
Crypto against the S&P 500, matched on shared sessions.
Bitcoin volatility
The live 90-close vol print for the BTC sample.
Ethereum volatility
The live 90-close vol print for the ETH sample.
Solana volatility
The live 90-close vol print for the SOL sample.
BTC and ETH correlation
The pair inside this sample window.
Correlation without a book is a trivia card.
Thrive shows whether the names in your journal actually moved together on the days you were in them.