Correlation Matrix
Rolling cross-asset correlations — what's moving together and what's diversifying, the relative-value backdrop.
Cross-Asset Correlation Matrix
60-session rolling Pearson on daily returns · ▲▼ marks pairs that shifted ±0.30+ vs the prior 60 sessions (hover any cell for the change)
| S&P 500 | NASDAQ | Small caps | Long bonds | Gold | Dollar | Bitcoin | Ether | Crude | Copper | HY credit | VIX | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| S&P 500 | 0.91 | 0.72 | 0.17 | 0.55 | -0.21 | 0.40 | 0.34 | -0.47 | 0.43 | 0.58 | -0.84 | |
| NASDAQ | 0.91 | 0.62 | -0.05▼ | 0.49 | -0.06 | 0.42 | 0.32 | -0.46 | 0.51 | 0.48 | -0.79 | |
| Small caps | 0.72 | 0.62 | 0.36 | 0.50 | -0.34 | 0.25 | 0.26 | -0.29 | 0.27▼ | 0.62 | -0.70 | |
| Long bonds | 0.17 | -0.05▼ | 0.36 | 0.20 | -0.41 | 0.08 | 0.16 | -0.24▲ | 0.04 | 0.60 | -0.14 | |
| Gold | 0.55 | 0.49 | 0.50 | 0.20 | -0.45 | 0.52 | 0.50 | -0.07 | 0.42 | 0.49 | -0.56 | |
| Dollar | -0.21 | -0.06 | -0.34 | -0.41 | -0.45 | -0.31 | -0.42 | -0.08 | -0.09▲ | -0.38 | 0.23 | |
| Bitcoin | 0.40 | 0.42 | 0.25 | 0.08 | 0.52 | -0.31 | 0.89 | -0.24 | 0.27 | 0.24 | -0.29 | |
| Ether | 0.34 | 0.32 | 0.26 | 0.16 | 0.50 | -0.42 | 0.89 | -0.18 | 0.19 | 0.25 | -0.32 | |
| Crude | -0.47 | -0.46 | -0.29 | -0.24▲ | -0.07 | -0.08 | -0.24 | -0.18 | -0.32 | -0.37 | 0.42 | |
| Copper | 0.43 | 0.51 | 0.27▼ | 0.04 | 0.42 | -0.09▲ | 0.27 | 0.19 | -0.32 | 0.40 | -0.44 | |
| HY credit | 0.58 | 0.48 | 0.62 | 0.60 | 0.49 | -0.38 | 0.24 | 0.25 | -0.37 | 0.40 | -0.53 | |
| VIX | -0.84 | -0.79 | -0.70 | -0.14 | -0.56 | 0.23 | -0.29 | -0.32 | 0.42 | -0.44 | -0.53 |
Regime shifts — last 60 sessions vs the 60 before
- NASDAQ × Long bonds Correlation breaking DOWN (0.31 → -0.05) — the old relationship is decoupling; hedges built on it need rechecking.
- Long bonds × Crude Correlation locking UP (-0.55 → -0.24) — these two are increasingly the same trade; diversification between them is evaporating.
- Dollar × Copper Correlation locking UP (-0.4 → -0.09) — these two are increasingly the same trade; diversification between them is evaporating.
- Small caps × Copper Correlation breaking DOWN (0.57 → 0.27) — the old relationship is decoupling; hedges built on it need rechecking.
Diversifier ranking — avg |correlation| vs the rest
- Long bonds0.22
- Dollar0.27
- Crude0.28
- Copper0.31
- Ether0.35
- Bitcoin0.36
- Gold0.43
- Small caps0.45
- HY credit0.45
- NASDAQ0.46
- VIX0.48
- S&P 5000.51
Lowest = best portfolio diversifier over the last 60 sessions. Measured, not assumed.
Check your own pairs
Any 2-8 Yahoo tickers — stocks, ETFs, futures (GC=F), FX (EURUSD=X), crypto (SOL-USD), indices (^GSPC)
The 60-second version
Correlation is the relative-value backdrop — what’s moving with what. It tells you where real diversification still exists, where a “different” position is actually the same bet in disguise, and when the whole market is fusing into one risk-on / risk-off trade. The matrix covers equities, bonds, gold, the dollar, crypto, crude, copper, credit and the VIX; the Check-your-own-pairs box computes the identical analysis for any tickers you type — your actual positions, not proxies.
How to read it
- Watch the arrows first — ▲▼ marks pairs whose 60-session correlation moved ±0.30+ versus the prior 60 sessions. A static matrix tells you the weather; the shifts tell you the weather is CHANGING — and correlation regime changes are themselves tradeable information.
- Stock–bond is the keystone — S&P × Long bonds negative means bonds still cushion equity drawdowns (the classic 60/40 hedge); a flip to positive — as in 2022 — removes the shock absorber and changes every portfolio's risk math.
- Crypto–Nasdaq measures the narrative — high correlation = crypto trading as a high-beta tech proxy; decoupling = crypto trading on its own driver (halving cycle, ETF flows, liquidity). Check it before treating BTC as diversification.
- Everything → 1 is the fire alarm — in a panic correlations converge to one and diversification fails exactly when it's needed. A matrix turning uniformly green with the VIX row deep red is the crash signature.
- Use the diversifier ranking to size hedges — the asset with the lowest average |correlation| against the rest is the one actually adding balance to a book right now — measured over the last 60 sessions, not assumed from a textbook.
Check your own pairs
Type any 2-8 Yahoo tickers, comma-separated: single stocks (NVDA, AMD), ETFs (SMH, XLE), futures (GC=F, CL=F), FX (EURUSD=X, JPY=X), crypto (SOL-USD) or indices (^GSPC, ^VIX). You get the same matrix, the same regime-shift detection and the same diversifier ranking on exactly those instruments — useful for checking whether two positions you hold are secretly one trade, or whether the hedge you’re paying for still hedges anything.
Methodology & honesty
Correlations are Pearson coefficients on daily percentage returns over the last 60 common trading sessions (dates where every asset in the set traded — crypto’s weekends are dropped when mixed with traditional assets). The shift comparison uses the 60 sessions before that, so a flagged pair reflects roughly the last three months versus the three months prior. Sixty sessions is the desk standard: long enough to be statistically meaningful, short enough to catch regime changes. Remember correlation is not causation and not beta — two assets can be perfectly correlated while one moves 10× as much. It measures direction agreement only.
In the MTS framework
Correlation is a Motion tool — price relationships across assets. It refines position sizing and confirms (or questions) a thesis: a setup independently confirmed across uncorrelated assets is far stronger than one riding a single crowded correlation, and a portfolio of five “different” trades with pairwise correlations above 0.8 is one trade with five commissions.