S&P 500 Correlation Leaderboard
These are the S&P 500 correlation pairs: the stocks whose daily returns have moved most closely together, and most opposite to each other, over the last 30, 60 or 90 trading days. Highly correlated stocks tend to react to the same news, so holding several of them can add less diversification than it looks like.
For the cross-asset view (sectors, indices, commodities, crypto) and the Nasdaq-100 version, see Correlations. To see how many S&P 500 stocks are above their 200-day average, see S&P 500 breadth.
Last updated: Oct 30, 2026, 2:17 AM UTC
S&P 500 Correlation Leaderboard
498 stocks, 60-day returnsThe most and least correlated pairs across every S&P 500 stock with price history.
Most Correlated
- 1
- 2
- 3
- 4
- 5
- 6
- 7
- 8
- 9
- 10
- 11
- 12
- 13
- 14
- 15
- 16
- 17
- 18
- 19
- 20
- 21
- 22
- 23
- 24
- 25
Most Inverse
- 1
- 2
- 3
- 4
- 5
- 6
- 7
- 8
- 9
- 10
- 11
- 12
- 13
- 14
- 15
- 16
- 17
- 18
- 19
- 20
- 21
- 22
- 23
- 24
- 25
About This Data
- Correlation is computed on daily returns over the trailing 30, 60, or 90 trading days (toggle above), aligned by calendar date.
- Only S&P 500 stocks with a full price history are included - a small handful of newly-added constituents may be missing until their history backfills.
- Each ticker appears in at most 3 rows per column, and same-company share classes (e.g. dual-class stock) are excluded, so the list reflects distinct relationships rather than one dominant pair repeated.
- Updated once a day.
S&P 500 correlation pairs: common questions
- What is a correlation pair in the S&P 500?
- A correlation pair is two S&P 500 stocks whose daily returns tend to move together (a correlation near +1) or in opposite directions (near -1). A value near 0 means their moves are essentially unrelated.
- Which S&P 500 stocks are the most correlated?
- The "Most Correlated" list above shows the pairs moving together most closely right now, recomputed daily, and "Most Inverse" shows the pairs moving most opposite. Pairs in the same industry, such as banks, chipmakers or oil producers, usually rank highest because the same news moves them together.
- Why do the most and least correlated pairs change?
- Correlation is measured on a trailing window of daily returns (30, 60 or 90 trading days, selectable above). When the market is driven by a different theme, such as rates, earnings or a sector rotation, the pairs that move together change with it.
- How can I use correlation pairs?
- Highly correlated holdings behave like one bet, so owning several of them adds less diversification than it appears to. Inversely correlated pairs can offset each other, and some traders compare a correlated pair when one stock lags the other. This is educational data, not investment advice.
New to correlation? Read Asset Correlation and Relative Strength Explained.