What Is Correlation in Investing?

Correlation is a statistical measure of how two assets move in relation to each other. It is expressed as a number between −1 and +1. A correlation of +1 means two assets move in perfect lockstep — when one rises 5%, the other rises 5%, always. A correlation of −1 means two assets move in exact opposition — when one rises 5%, the other falls 5%, always. A correlation of 0 means the two assets move independently — knowing what one did tells you nothing about what the other did.

For investors building a portfolio, correlation is not an abstraction. It is the mathematical input that determines whether combining two assets actually reduces portfolio risk — and by how much. Understanding correlation is understanding why diversification works at the mathematical level, not just as a general principle.

Why Correlation Is the Key to Diversification

The intuitive case for diversification is simple: don't put all your eggs in one basket. The mathematical case is more precise and more powerful. When you combine two assets whose returns are not perfectly correlated, the resulting portfolio historically has had less volatility than the weighted average of the two assets' individual volatilities. This reduction in volatility is not free — it comes at the cost of not being fully concentrated in whichever asset turns out to perform better. But the potential reduction in risk is quantifiable, even if not guaranteed in any specific market environment.

The extreme case illustrates the point clearly. Two assets with a correlation of −1 can, in the right proportions, produce a portfolio with zero volatility — neither asset alone has zero volatility, but their movements cancel each other out exactly. No combination of positively correlated assets can achieve this. Two assets with a correlation of +1 produce a portfolio whose volatility is exactly the weighted average of the two — no diversification benefit at all, regardless of how you split between them.

This is the mathematical insight that Harry Markowitz formalized in 1952 in the paper that became Modern Portfolio Theory: the expected return of a portfolio is the weighted average of the expected returns of its components, but the portfolio's volatility is lower than the weighted average of its components' volatilities — as long as they are not perfectly correlated. The lower the correlation between assets, the more dramatic the potential volatility reduction. See Modern Portfolio Theory in Plain English for the full framework.

The Correlation Coefficient: Reading the Number

In practical portfolio analysis, correlation coefficients between major asset classes tend to cluster in ranges that have meaningful implications.

Strong negative correlation (−1.0 to −0.5): Rare in practice among broad asset classes. High-quality government bonds and equities have historically shown mildly negative to low-positive correlation, particularly during equity market stress — which is the basis for the traditional stock-bond portfolio.

Low or near-zero correlation (−0.3 to +0.3): Assets in this range provide meaningful diversification benefit. Adding a near-zero-correlation asset to a portfolio reduces portfolio volatility proportionally to the weight allocated to it, with minimal drag from return correlation.

Moderate positive correlation (+0.3 to +0.7): Some diversification benefit, but less than the above ranges. Most equity asset classes — US large cap, US small cap, international developed markets — fall in this range relative to each other, which is why an all-equity portfolio is still reasonably diversified but does not benefit from the deeper diversification of adding uncorrelated asset classes.

High positive correlation (+0.7 to +1.0): Very limited diversification benefit. Combining two highly correlated assets mostly averages their returns without meaningfully reducing risk. Owning both a broad US equity ETF and a US large-cap ETF provides almost no diversification — the assets move together.

Historical Correlations Between Major Asset Classes

Historical correlation estimates between broad asset classes give a sense of where diversification has come from in practice. These figures are based on long-period historical data and should be understood as general reference points — correlations change over time and across market regimes.

These figures illustrate a consistent pattern: most equity asset classes are moderately to highly correlated with each other, which limits the diversification benefit of holding multiple equity categories. The most meaningful diversification benefit historically has come from adding bonds and alternative asset classes — not from subdividing equity exposure across categories.

It is worth noting that the US equity / bond correlation has varied substantially across different historical periods. During the deflationary recessions of the 1990s and 2000s, bonds and equities were negatively correlated — bonds rose when stocks fell, providing genuine portfolio protection. During the inflationary environment of the 1970s and early 2020s, both asset classes fell simultaneously, as rising interest rates hurt bonds and slowing growth hurt equities. As of early 2026, the equity/bond correlation has largely re-stabilized following the inflationary shocks of the mid-2020s — but investors should remain aware that regime shifts can occur rapidly, and this relationship is more volatile than it appeared in the pre-2022 era. The bond allocation's diversification benefit depends heavily on the economic regime. See Bonds for how this plays out in practice for early retirement portfolios.

The Correlation Breakdown Problem

The most important practical limitation of correlation-based diversification is that correlations are not stable. They change over time, shift across market regimes, and — most critically — tend to increase sharply during market crises.

This is sometimes called the diversification breakdown problem: the assets that appeared to provide diversification based on historical correlation estimates tend to become more correlated during the periods of market stress when diversification is most needed.

The mechanism is partly mechanical and partly behavioral. During a market crisis, investors sell assets across the board to raise cash or meet margin calls, pushing down asset classes that would normally be uncorrelated with equities. Risk-parity strategies that use leverage amplify this effect. International markets, which may be uncorrelated with US equities during normal periods, are hit by the same global deleveraging.

The practical implication is that correlation estimates based on long-period historical data systematically understate the correlation that will be experienced during the worst market environments — precisely the environments where portfolio construction matters most. This does not mean diversification is ineffective; it means that portfolios should not rely solely on historical correlation estimates holding during stress periods, and stress-testing portfolios against scenarios where correlations increase significantly is a prudent complement to standard allocation analysis.

This is one of the reasons that Modern Portfolio Theory's assumptions — including the assumption that correlations are stable — break down in practice. It is also part of what motivates approaches like the Black-Litterman Model, which builds portfolios from market equilibrium rather than relying entirely on historical correlation inputs.

Correlation and the Efficient Frontier

The efficient frontier — the set of portfolios with the highest expected return for any given level of risk — is determined by the correlation structure of the assets included. Lower correlations between assets push the efficient frontier further to the upper left: more expected return for less risk. Higher correlations collapse the frontier toward the line connecting the two individual assets, reducing or eliminating the diversification benefit of combining them.

This is why adding a truly uncorrelated or negatively correlated asset class to a portfolio can improve its risk-return profile even if that asset class has a lower expected return than existing holdings. The contribution to risk reduction can more than offset the drag from a lower expected return — but only if the lower correlation holds. If the new asset class is actually highly correlated with existing holdings under stress, the expected diversification benefit may not materialize when it matters most.

See What Is the Efficient Frontier? for how correlation determines the shape of the frontier and which portfolios sit on it.

Correlation in a Goal-Based Portfolio

For investors managing multiple financial goals with different time horizons — a retirement goal 20 years away, a college funding goal in 5 years, a real estate purchase in 2 years — correlation analysis applies at multiple levels.

Within each goal's portfolio, the correlation between assets determines how much diversification benefit is achieved for that goal's specific allocation. A retirement portfolio with a long time horizon can tolerate higher-correlation, higher-volatility assets because time provides a buffer for volatility to average out. A short-horizon goal in a preservation-focused portfolio should have asset classes whose correlations under stress remain stable — because there is no time to recover from a correlation breakdown.

Across goals, the separate-portfolio approach of goal-based investing provides a form of implicit diversification: goals with different time horizons hold different allocations, so a market event that hurts the retirement portfolio (primarily equities) may not equally affect a short-horizon goal portfolio (primarily cash and short-duration bonds).

See What Is Asset Allocation? for how time horizon drives the starting allocation for each goal, and What Is an Investment Time Horizon? for the relationship between time horizon and appropriate risk.

California Note

Correlation in investing is a portfolio theory concept, and there is no California-specific tax treatment of correlation itself. However, the correlation structure of a portfolio has California tax implications through its effect on rebalancing frequency and tax-loss harvesting opportunities. A portfolio containing low-correlation assets will experience more frequent divergence between asset classes — more instances where some positions are up and others are down at the same time. For a California investor, this creates both more rebalancing events (which can trigger capital gains in a taxable account at rates up to 13.3%) and more tax-loss harvesting opportunities (selling the positions that are down to realize losses that offset gains elsewhere). A direct indexing approach — owning individual stocks rather than ETFs — exploits within-portfolio correlation variation systematically to generate harvestable losses even in rising markets. It is worth noting that direct indexing and tax-loss harvesting strategies do not guarantee a specific outcome and may result in tracking error relative to the target index — meaning the portfolio's returns may diverge from the index it is designed to approximate. See Direct Indexing for how this works.

This article is for educational purposes only and does not constitute investment, tax, or financial advice. Portfolio construction involves risk, and all investments may lose value. Past performance does not guarantee future results. Historical correlations are not a guarantee of future correlations. Always consult a qualified financial advisor before making investment decisions.