ETF correlation matrix
This ETF correlation matrix shows how closely the monthly returns of up to 10 funds moved together, over the same months for every fund, and how each one performed. Pick the ETFs and a period: useful to spot funds that overlap and the ones that actually diversify.
Data updated October 26, 2026
Nov 2021 – Oct 2026 · 60 months, the same for every fund
| GLD | TLT | AGG | VNQ | VEA | VWO | SPY | |
|---|---|---|---|---|---|---|---|
| TLT | .370.37 | ||||||
| AGG | .390.39 | .950.95 | |||||
| VNQ | .270.27 | .690.69 | .720.72 | ||||
| VEA | .460.46 | .660.66 | .760.76 | .800.80 | |||
| VWO | .470.47 | .640.64 | .670.67 | .610.61 | .810.81 | ||
| SPY | .170.17 | .550.55 | .630.63 | .780.78 | .810.81 | .640.64 | |
| QQQ | .100.10 | .520.52 | .560.56 | .600.60 | .680.68 | .580.58 | .910.91 |
Most alike
AGG and TLT 0.95
Holding both adds little diversification.
Most different
QQQ and GLD 0.10
They have moved largely independently.
Best diversifier here
GLD 0.32
Lowest average correlation with the other funds.
Average correlation with the other funds
Lower means the fund did more to smooth the rest of the set.
- GLD0.32
- QQQ0.56
- TLT0.63
- VWO0.63
- VNQ0.64
- SPY0.64
- AGG0.67
- VEA0.71
Returns with dividends reinvested, each in its fund’s trading currency. Historical correlations change over time, especially in crises. Not investment advice.
Performance over the same months
Growth of 10,000 in each fund from Nov 2021 to Oct 2026, dividends reinvested.
| Fund | 10,000 became | Per year | Volatility | Worst fall |
|---|---|---|---|---|
| GLD | 22,199 | +17.3% | 17.3% | −23.8% |
| QQQ | 19,169 | +13.9% | 20.7% | −32.6% |
| SPY | 17,846 | +12.3% | 15.4% | −23.9% |
| VEA | 15,616 | +9.3% | 16.2% | −27.9% |
| VWO | 13,404 | +6.0% | 14.7% | −27.6% |
| VNQ | 10,285 | +0.6% | 19.0% | −32.7% |
| AGG | 9,787 | −0.4% | 6.6% | −15.8% |
| TLT | 6,436 | −8.4% | 14.6% | −42.0% |
Key ETF pairs: how their correlation changed over time
One number hides how much a correlation moves. Each chart shows the correlation of the last 36 monthly returns, month by month: the pairs investors ask about most, from the full common history of each fund.
Stocks and bonds
SPY vs AGG- 3 years
- 0.47
- 5 years
- 0.63
- 10 years
- 0.40
- Since 2006
- 0.24
Over the last 5 years SPY and AGG have moved partly together (0.63). Measured over rolling 36-month windows, the correlation ranged from −0.47 (Dec 2012) to 0.73 (Dec 2024). The last 3 years read higher than the full history (0.24 since 2006): recently the two have diversified each other less.
Bonds are the classic cushion for stock falls. That works when falls come from slowing growth; when they come from rising rates and inflation, as in 2022, both can fall together.
Stocks and long Treasuries
SPY vs TLT- 3 years
- 0.47
- 5 years
- 0.55
- 10 years
- 0.18
- Since 2006
- −0.08
Over the last 5 years SPY and TLT have moved partly together (0.55). Measured over rolling 36-month windows, the correlation ranged from −0.79 (Jan 2013) to 0.68 (Dec 2024). The last 3 years read higher than the full history (−0.08 since 2006): recently the two have diversified each other less.
Long-dated Treasuries react to interest rates far more than broad bond funds: the strongest cushion in a growth scare, the largest falls when rates rise.
Stocks and gold
SPY vs GLD- 3 years
- 0.02
- 5 years
- 0.17
- 10 years
- 0.13
- Since 2006
- 0.09
Over the last 5 years SPY and GLD have moved only loosely together (0.17). Measured over rolling 36-month windows, the correlation ranged from −0.27 (Sep 2017) to 0.44 (Aug 2014). The last 3 years are close to the full-history figure (0.09 since 2006).
Gold has no link to company earnings: it moves more with real interest rates, the dollar and demand for safety.
S&P 500 and Nasdaq-100
SPY vs QQQ- 3 years
- 0.90
- 5 years
- 0.91
- 10 years
- 0.91
- Since 2006
- 0.91
Over the last 5 years SPY and QQQ have moved almost as one (0.91). Measured over rolling 36-month windows, the correlation ranged from 0.84 (Nov 2018) to 0.95 (Dec 2020). The last 3 years are close to the full-history figure (0.91 since 2006).
Most Nasdaq-100 companies are also in the S&P 500, where the largest of them carry a big weight, so the two funds overlap heavily.
US and international stocks
SPY vs VEA- 3 years
- 0.67
- 5 years
- 0.81
- 10 years
- —
- Since 2020
- 0.82
Over the last 5 years SPY and VEA have moved closely together (0.81). Measured over rolling 36-month windows, the correlation ranged from 0.62 (Feb 2026) to 0.90 (May 2024). The last 3 years are close to the full-history figure (0.82 since 2020).
Developed markets outside the US share the same global cycle but differ in sectors and currency, which keeps the link below 1.
Stocks and real estate
SPY vs VNQ- 3 years
- 0.63
- 5 years
- 0.78
- 10 years
- —
- Since 2020
- 0.80
Over the last 5 years SPY and VNQ have moved closely together (0.78). Measured over rolling 36-month windows, the correlation ranged from 0.63 (Oct 2026) to 0.91 (Nov 2023). The last 3 years are close to the full-history figure (0.80 since 2020).
REITs are listed companies too: they move with the stock market, with an extra sensitivity to interest rates.
How to read the ETF correlation matrix
Values above 0.9 mean two funds have moved almost as one: holding both changes little. Values between 0.3 and 0.7 give partial diversification; values near zero or below are the ones that smooth a portfolio.
“Grouped by similarity” reorders the funds so that the ones moving together sit next to each other: blocks of dark cells are groups that behave as one, and a fund with pale cells across its row is the one diversifying the rest.
Compare the 3-year and the full-history view: when correlations jump in recent years, the diversification you are counting on may be weaker than the long-run figure suggests. The performance chart below the matrix uses the same months, so returns and correlations describe the same stretch of history.
ETF correlation: questions
Are stocks and bonds correlated?
Between SPY (S&P 500) and AGG (US aggregate bonds) the correlation of monthly returns was 0.47 over the last 3 years, 0.40 over 10 years and 0.24 over the full common history since 2006. Over rolling 36-month windows it ranged from −0.47 to 0.73: bonds have not always cushioned stock falls by the same amount.
Is gold correlated with the S&P 500?
Between SPY and GLD (physical gold) the correlation of monthly returns was 0.02 over the last 3 years, 0.13 over 10 years and 0.09 over the full common history since 2006. Over the last 5 years the two have moved only loosely together.
How correlated are SPY and QQQ?
The correlation of monthly returns between SPY and QQQ was 0.90 over the last 3 years, 0.91 over 10 years and 0.91 over the full common history since 2006. With figures this high, holding both mostly adds to the weight of large US technology companies rather than diversifying.
What does correlation between two ETFs mean?
It measures how closely their returns move together, from +1 (always in the same direction and proportion) through 0 (no relation) to −1 (always opposite). Two funds with a correlation near 1 behave almost as one; adding the second brings little diversification.
Why monthly returns?
US and European funds trade on different days and hours, so daily returns of funds on different exchanges are not aligned and understate their correlation. Month-end returns line up across markets and give a cleaner picture for portfolio decisions.
Is a low correlation always good?
A low or negative correlation smooths a portfolio’s ups and downs, which is the point of diversifying. But correlations change over time and often rise in a crisis, just when diversification is needed most: check the shorter periods as well as the longest one.
Which ETFs are available?
The ETFs covered by our ETF pages, US-listed and European UCITS, across stocks, bonds, gold, commodities, real estate and crypto. Returns include reinvested dividends and are in each fund’s trading currency.
More free tools
- ETF returns and comparisonsReturns, drawdowns, head-to-head
- Efficient frontierBest mix for each level of risk
- Portfolio backtestTest a mix of ETFs on history
- Rolling returnsEvery holding period, not one start date
- DCA calculatorMonthly investing with real returns
- Rebalancing calculatorTrades to get back to target weights
Source: MarketStack end-of-day prices, total return with dividends reinvested. Figures are historical and descriptive; they are not forecasts and not investment advice. MEDGE Capital does not recommend buying or selling any instrument. Risk disclaimer.