Correlation and Why It Matters for Portfolio Construction

Diversification is the most frequently cited principle in investing, but it is also the most frequently misunderstood. Many investors believe that owning many different assets automatically provides diversification. It does not. What matters is not how many assets you hold, but how those assets relate to one another — specifically, their correlations.

Correlation is the statistical measure that determines whether diversification actually works. Two assets with high correlation move in lockstep and provide little diversification benefit regardless of how different they look on paper. Two assets with low or negative correlation offset each other's movements, reducing portfolio volatility and smoothing the ride.

This article explains what correlation is, how it behaves in practice (including its troubling tendency to spike during crises), and how tactical allocation strategies use correlation insights to build more robust portfolios.

What Is Correlation?

The Basics

Correlation measures the degree to which two assets move together. It is expressed as a coefficient ranging from −1 to +1:

• - +1: Perfect positive correlation. The two assets move in exactly the same direction, in the same proportion, every period. No diversification benefit.
• - 0: Zero correlation. The two assets move independently. Strong diversification benefit.
• - −1: Perfect negative correlation. The two assets move in exactly opposite directions. Maximum diversification — in theory, you could construct a zero-risk portfolio.

In practice, no major asset class has a correlation of exactly +1, 0, or −1 with any other. Real-world correlations fall somewhere in between and, critically, change over time.

Examples of Asset Class Correlations

Historical correlations between major asset classes give a sense of the landscape:

| Asset Pair | Typical Correlation | |---|---| | US large-cap stocks / US small-cap stocks | +0.85 to +0.95 | | US stocks / International developed stocks | +0.70 to +0.85 | | US stocks / Emerging market stocks | +0.60 to +0.80 | | US stocks / US Treasury bonds | −0.20 to +0.30 | | US stocks / Gold | −0.10 to +0.15 | | US stocks / Commodities | +0.10 to +0.40 | | US Treasury bonds / Gold | −0.10 to +0.20 |

Several patterns emerge from this table. Equities across different markets are highly correlated — they all respond to global risk appetite, interest rates, and economic growth expectations. Adding international stocks to a US stock portfolio provides some diversification, but far less than most investors expect.

The most valuable diversifiers are assets that have low or negative correlation with equities: Treasury bonds and gold. These are the assets that tactical allocation strategies typically use as defensive holdings.

How Correlation Affects Portfolio Risk

The Diversification Formula

The volatility of a two-asset portfolio is not simply the weighted average of the two assets' volatilities. It is calculated as:

σ_portfolio = √(w₁²σ₁² + w₂²σ₂² + 2w₁w₂σ₁σ₂ρ)

Where w₁ and w₂ are the portfolio weights, σ₁ and σ₂ are the asset volatilities, and ρ is the correlation between them.

The critical term is 2w₁w₂σ₁σ₂ρ. When ρ is positive, this term adds to portfolio volatility. When ρ is negative, it subtracts from it. When ρ is zero, the entire term disappears, and portfolio volatility is simply the root of the weighted sum of squared volatilities — always less than the weighted average.

This formula explains why diversification works: combining assets with low correlation produces a portfolio that is less volatile than any individual component (on a risk-per-unit-of-return basis). The lower the correlation, the greater the risk reduction.

A Practical Example

Consider combining US equities (15% volatility) and Treasury bonds (7% volatility) in equal proportions:

• - At correlation +0.80: Portfolio volatility ≈ 10.4%
• - At correlation 0.00: Portfolio volatility ≈ 8.3%
• - At correlation −0.30: Portfolio volatility ≈ 7.2%

The difference between a correlation of +0.80 and −0.30 is enormous: 10.4% vs 7.2% volatility for the same portfolio weights. This is why the sign and magnitude of correlation matter so much for portfolio construction.

The Problem: Correlations Are Not Stable

Correlation Regimes

If correlations were constant, portfolio construction would be straightforward. Estimate the correlation matrix once, build the optimal portfolio, and leave it alone. Unfortunately, correlations are unstable. They shift based on macroeconomic regimes, monetary policy, and market conditions.

The stock-bond correlation is the most important example. From the 1960s through the late 1990s, stocks and bonds were positively correlated — they tended to rise and fall together, driven by inflation and interest rate cycles. From approximately 2000 to 2021, the correlation turned negative — stocks and bonds moved in opposite directions, as central banks cut rates during recessions (boosting bond prices) and raised rates during expansions (hurting bond prices but supporting stocks).

In 2022, the correlation flipped back to positive as both stocks and bonds sold off simultaneously in response to aggressive rate hikes. Investors who had built portfolios assuming negative stock-bond correlation — including the classic 60/40 portfolio — experienced unexpected losses as their supposed diversifier (bonds) fell alongside their equities.

Correlation Spikes During Crises

The most dangerous property of correlations is their tendency to increase during market crises. During normal times, assets may show moderate correlations. During panics, correlations spike toward 1.0 as investors sell everything to raise cash. This phenomenon, sometimes called "correlation convergence," means that diversification fails precisely when it is needed most.

Academic research by Longin and Solnik (2001) confirmed that correlations increase in volatile bear markets but not in volatile bull markets. The implication is that portfolios designed for normal conditions will be less diversified than expected during the conditions that pose the greatest risk.

This is one of the fundamental reasons that tactical asset allocation exists. Static diversification based on historical correlations is unreliable during crises. Tactical strategies address this by actively reducing risk exposure when conditions deteriorate, rather than relying on passive diversification to absorb losses.

Correlation at the Strategy Level

Why Strategy Correlation Matters

Tactical investors who build multi-strategy portfolios face a second layer of correlation: the correlation between strategies, not just between asset classes.

Two momentum strategies that trade the same universe of assets will be highly correlated — they are likely to hold similar positions and experience similar drawdowns. Combining them provides little diversification benefit.

A momentum strategy and a defensive allocation strategy, however, may have low correlation because they respond to different market signals. The momentum strategy captures upside trends; the defensive strategy limits downside participation. Their return streams are driven by different factors, and blending them smooths the overall equity curve.

Measuring Strategy Correlation on Portfoliowiser

When you build a multi-strategy portfolio on Portfoliowiser, the platform helps you assess how different strategies interact. The equity curves of individual strategies can be overlaid, making it visually apparent when strategies move together and when they diverge.

Key principles for multi-strategy correlation:

1. 1. Avoid redundancy. If two strategies hold the same assets with the same timing, adding both to the portfolio does not improve diversification.

1. 2. Combine different signal types. Momentum-based strategies, mean-reversion strategies, and defensive strategies tend to have lower correlations with each other because they are driven by different market phenomena.

1. 3. Diversify across asset universes. A US equity momentum strategy and an international equity momentum strategy have lower correlation than two strategies trading the same US equity universe.

1. 4. Include defensive components. Strategies with explicit defensive mechanisms (canary signals, trend filters, bond overlays) tend to have lower correlation with aggressive equity strategies during drawdowns — exactly when it matters.

Tactical Responses to Correlation Instability

Dynamic Allocation

The most direct response to unstable correlations is dynamic allocation — adjusting portfolio weights as correlations change. If stock-bond correlation turns positive (as in 2022), a tactical strategy can reduce bond exposure and increase allocation to assets with more reliable negative correlation, such as gold or managed futures.

This is one of the key advantages of TAA over static allocation. A 60/40 portfolio is built on the assumption that bonds will diversify stocks. When that assumption fails, the static portfolio has no mechanism to adapt. A tactical portfolio can shift its defensive allocation to assets that are actually providing diversification in the current regime.

Trend-Based Defensive Switching

Many tactical strategies do not explicitly model correlations. Instead, they use trend and momentum signals that implicitly capture correlation shifts. When equities are in a downtrend, the strategy moves to defensive assets. When bonds are also in a downtrend (positive stock-bond correlation), the strategy may move further along the defensive spectrum — to gold, cash, or short-term Treasury bills.

This indirect approach is often more robust than explicitly modelling correlations, because correlation estimates are noisy and require long data windows to be statistically reliable. Trend signals, by contrast, react to price movements directly and can adapt more quickly to changing conditions.

Multi-Defensive Baskets

The most resilient tactical strategies maintain a diversified basket of defensive assets rather than relying on a single one. By including Treasury bonds, gold, and cash equivalents in the defensive universe, the strategy ensures that at least one defensive asset is likely to perform well regardless of the correlation regime.

This is particularly important in environments like 2022, where traditional bonds failed as diversifiers. Strategies that could rotate to gold or cash during this period fared much better than those that could only choose between equities and bonds.

Common Mistakes with Correlation

Confusing Low Volatility with Low Correlation

A low-volatility asset is not necessarily a good diversifier. If a low-volatility asset is highly correlated with equities (as some low-volatility equity strategies are), it will still decline during equity drawdowns — just by a smaller amount. True diversification requires low correlation, not just low volatility.

Over-Diversifying into Correlated Assets

Holding ten different equity ETFs (US large-cap, US mid-cap, US small-cap, US value, US growth, etc.) may feel diversified, but all of these are highly correlated with the broad US equity market. The portfolio's risk is nearly identical to holding a single total market fund. Meaningful diversification requires crossing asset class boundaries — adding bonds, gold, commodities, or international assets with genuinely different return drivers.

Ignoring Correlation Instability

Perhaps the most dangerous mistake is treating historical correlation estimates as if they will persist. Correlation matrices estimated from the past 10 years may not reflect the next 10 years. Building a portfolio that is "optimal" under historical correlations but fragile under different correlation regimes is a recipe for disappointment.

Assuming Negative Correlation During Crises

Some assets, like gold, have historically shown negative correlation with equities during crises. But this is a tendency, not a guarantee. During liquidity panics, even gold and bonds can temporarily sell off alongside equities. Tactical strategies should be designed to handle periods where all correlations converge toward 1.0, not just periods where diversification works as expected.

Building Correlation-Aware Portfolios

Step 1: Identify Genuinely Different Return Drivers

Start by selecting assets and strategies driven by different fundamental factors. Equities are driven by economic growth and earnings. Bonds are driven by interest rates and inflation expectations. Gold is driven by real rates and uncertainty. Commodities are driven by supply-demand dynamics. Strategies based on momentum respond to trends; defensive strategies respond to risk signals.

Step 2: Verify Historical Correlation

Examine the historical correlation between your selected components. Prioritise combinations with low or negative correlation, particularly during stress periods. Portfoliowiser's equity curves and performance data make it possible to visually assess how strategies behaved relative to each other during historical crises.

Step 3: Plan for Correlation Failure

Assume that correlations will increase during the next crisis. Ensure your portfolio has mechanisms to respond — whether through tactical signals that reduce exposure, a diversified defensive basket, or a multi-strategy blend that captures different market dynamics.

Step 4: Monitor and Adjust

Correlations change over time. Periodically review whether your portfolio's components are still providing the diversification you expect. If correlations have shifted, consider rebalancing toward components that offer better diversification in the current regime.

Conclusion

Correlation is the engine that makes diversification work — or fail. Understanding how assets and strategies correlate with each other, and how those correlations change under stress, is essential for building portfolios that survive real-world market conditions.

Tactical asset allocation strategies address the fundamental limitation of static diversification: the instability of correlations. By actively adjusting allocations in response to changing market conditions, TAA portfolios can maintain effective diversification even when passive approaches break down.

Build diversified, multi-strategy portfolios and see how strategies interact at app.portfoliowiser.com.

*Disclaimer: This article is for educational purposes only and does not constitute financial advice. Past correlations do not guarantee future correlations. All investing involves risk, including the possible loss of principal. Consult a qualified financial adviser before making investment decisions.*