How Different Risks Combine in a Portfolio: Correlation and Diversification

How Different Risks Combine in a Portfolio: Understanding Correlation

When you hold multiple investments, the overall portfolio risk depends not just on the individual asset risks but on how those assets move together—their correlation. Two assets might individually have moderate risk, but if they're highly correlated (move together), combining them provides little diversification benefit. Understanding how different risk types combine is foundational to building portfolios that survive diverse market conditions.

A portfolio's total risk emerges from correlations between its components. This is why professional investors obsess over correlation structures: two positions with identical risk characteristics but negative correlation provide superior diversification than two identical positions with positive correlation. Multiple risk types portfolio construction is the science of reducing overall risk through correlation management.

Quick definition: Portfolio risk from combining multiple assets depends on each asset's individual risk and the correlation between assets. Low or negative correlation reduces portfolio risk below the simple average of individual risks.

Key takeaways

• Correlation measures how two investments move together, ranging from −1 (perfect negative) to +1 (perfect positive)
• Low or negative correlation reduces portfolio risk below the arithmetic average of individual risks
• Correlation structures change during market stress—traditional diversification fails when most needed
• Systematic (beta) risk cannot be diversified away; idiosyncratic risk can
• Building a resilient portfolio requires understanding multiple risk types and their interactions

Understanding Correlation

Correlation is a statistical measure, ranging from −1 to +1, that describes how two variables move relative to each other.

• Correlation = +1: Perfect positive correlation; assets move together identically. A 10% gain in Asset A coincides with a 10% gain in Asset B. Combining them provides zero diversification benefit.

• Correlation = 0: No correlation; assets move independently. Changes in Asset A tell you nothing about Asset B's movement. Some diversification benefit exists.

• Correlation = −1: Perfect negative correlation; assets move oppositely. A 10% gain in Asset A coincides with a 10% loss in Asset B. Maximum diversification benefit.

Empirically, most asset classes show positive correlations during normal times (0.3 to 0.7), reducing but not eliminating diversification benefits. However, correlations are unstable—they change across time periods, market regimes, and especially during crises when everything falls together (correlation approaches +1).

The Mathematics of Risk Combination

When combining two assets with risks σ₁ and σ₂ and correlation ρ, the combined portfolio risk is:

Portfolio Risk = √[w₁² × σ₁² + w₂² × σ₂² + 2 × w₁ × w₂ × ρ × σ₁ × σ₂]

Where w₁ and w₂ are the weights in each asset.

Let's work through an example:

• Asset 1 (Stock): Risk = 16%, Weight = 50%
• Asset 2 (Bond): Risk = 5%, Weight = 50%
• Correlation = 0.3

Portfolio Risk = √[(0.5)² × (0.16)² + (0.5)² × (0.05)² + 2 × 0.5 × 0.5 × 0.3 × 0.16 × 0.05]
Portfolio Risk = √[0.0064 + 0.000625 + 0.0012]
Portfolio Risk = √0.008425 = 0.0918 = 9.18%

Notice: the simple average of 16% and 5% is 10.5%, but the actual portfolio risk is only 9.18% because of the positive but less-than-perfect correlation. The correlation benefit reduces portfolio risk by 1.3% annually.

If correlation were 0 (no relationship):

Portfolio Risk = √[0.0064 + 0.000625 + 0] = 0.0808 = 8.08%

If correlation were +1 (perfect positive):

Portfolio Risk = √[0.0064 + 0.000625 + 2 × 0.5 × 0.5 × 1 × 0.16 × 0.05] = √0.01 = 10%

The range from 8.08% (correlation = 0) to 10% (correlation = +1) shows correlation's profound impact. This is why professional investors prioritize low-correlation diversifiers—they reduce portfolio risk below the theoretical average.

Types of Risk and Their Combinations

Professional portfolio theory divides risks into categories that combine differently:

Systematic Risk (Market Risk) Systematic risk is the risk inherent to holding equities, bonds, or alternative assets—moves that affect entire markets together. During economic recessions, most stocks decline together. During rising interest rates, most bonds decline together. This risk cannot be diversified away within its asset class; it's the source of returns you earn for bearing it.

• Stock market systematic risk: 12–14% volatility
• Bond market systematic risk: 3–5% volatility
• Commodity systematic risk: 15–20% volatility

Idiosyncratic Risk (Specific Risk) Idiosyncratic risk is company-specific or sector-specific—a firm's scandal, management change, or competitive threat. This risk can be diversified away by holding many companies. Vanguard research shows that holding 20–30 diverse stocks eliminates 95% of idiosyncratic risk, leaving only systematic risk.

Correlation Structure Risk During normal times, correlations between stocks and bonds are 0.0 to 0.2 (low), providing excellent diversification. During crises (2008, 2020), correlations spike to 0.4–0.8 as panic forces selling across all risky assets. A portfolio constructed for normal-times correlation may face unexpected volatility during crisis-times correlation. Regime risk is the danger that correlation structures change to your disadvantage.

The Correlation Breakdown During Crises

One of the most expensive lessons investors learn is that correlations are unstable and increase during market stress. A portfolio diversified for normal times may concentrate risk during crises.

Historical Example: 2008 Financial Crisis Before 2008, typical correlations:

• Stocks & Bonds: 0.1
• Stocks & Real Estate: 0.6
• Stocks & Commodities: 0.2

During the 2008 crisis, correlations spiked:

• Stocks & Bonds: 0.4 (bonds lost value alongside stocks)
• Stocks & Real Estate: 0.8 (real estate fell harder)
• Stocks & Commodities: 0.3 (minimal benefit)

A 60/40 portfolio provided expected risk of ~8.5% based on normal correlation. When correlation shifted, actual volatility reached 18% as both stocks and bonds declined. The diversification benefit evaporated precisely when needed.

Building Portfolios Across Risk Regimes

Professional investors address correlation instability through multiple approaches:

Barbell Strategy Rather than a continuous 60/40 mix, use a barbell: 50% safe assets (Treasury bonds, cash) + 50% growth assets (stocks, alternatives). This maximizes diversification benefit—if stocks fall, bonds often stabilize the portfolio. If bonds are under pressure from rate increases, stocks often benefit. The barbell reduces correlation stress across regimes.

Uncorrelated Diversifiers Beyond stocks and bonds, adding assets with low or negative correlation in various conditions:

• Managed futures: Momentum strategy, profitable in rising inflation, protective in deflation
• Long-volatility strategies: Make money when volatility spikes; hedge during crisis
• Alternative assets: Private equity, real estate, hedge strategies with different return drivers
• Trend-following: Benefit in strong directional markets; underperform in ranging markets

Empirical data from Morningstar and academic research shows that adding 10–20% alternatives (managed futures, hedge funds, private markets) to a 60/40 portfolio reduces tail risk (worst 5% of outcomes) while maintaining expected return.

Correlation vs. Causation in Risk Combination

Investors often conflate correlation with causation. Just because stocks and bonds have correlated recently doesn't mean adding bonds to a stock portfolio offers no benefit. The relationship between assets is complex:

1. Supply-and-demand correlation: When both stocks and bonds are selling (due to liquidity crisis), they fall together—high correlation.

2. Economic regime correlation: During recessions, both stocks and bonds sometimes decline (stocks from weaker earnings, bonds from credit risk). This produces correlation spikes around macroeconomic shocks.

3. Policy-driven correlation: When central banks raise rates aggressively, both stocks and bonds can decline simultaneously (opportunity cost for stocks, duration loss for bonds).

Understanding the source of correlation helps predict when it will break. A stock-bond correlation driven by policy (central bank decisions) might be temporary, reversing when policy shifts. A correlation driven by systemic financial stress might persist, requiring different diversification strategies.

Real-world examples

Example 1: The Barbell Portfolio's Stress Test An investor compares two 60/40 allocations: a traditional continuous 60/40 mix, and a barbell of 30/70 (30% stocks + 70% bonds) with 50% of the bond allocation in Treasury bonds and 50% in long-duration diversifiers.

During 2022's bear market with rising rates:

• Traditional 60/40: Falls 16% (both stocks and bonds decline)
• Barbell: Falls 8% (stocks decline 22%, but long-duration Treasury bonds rise 10%, offsetting loss)

The barbell's higher bond concentration, initially appearing conservative, performed better due to correlation structure. The diversification benefit proved genuine during the stress period.

Example 2: The Commodity Diversifier An investor holds a 100% stock portfolio worth $200,000, experiencing 20% volatility. They add 10% commodities ($20,000), reducing stocks to $180,000. Historical correlation between stocks and commodities is 0.2.

Expected portfolio volatility:

= √[(0.9)² × (0.20)² + (0.1)² × (0.25)² + 2 × 0.9 × 0.1 × 0.2 × 0.20 × 0.25]
= √[0.0324 + 0.000625 + 0.0018]
= √[0.03485] = 18.7%

By adding 10% commodity exposure, volatility decreased from 20% to 18.7%—a 1.3% reduction despite the higher-volatility commodity asset. The low correlation (0.2) creates diversification benefit that outweighs the higher commodity volatility.

However, during the 2022 energy crisis, commodity-stock correlation spiked to 0.6, reducing the diversification benefit. The investor who understood this regime risk would have accepted the temporary loss of diversification value during crisis but maintained the longer-term position due to its structural benefit.

Example 3: The International Diversification Surprise Historically, international developed markets (Europe, Japan) have low correlation with U.S. markets (0.4–0.5), suggesting diversification benefit. An investor allocates 30% of stocks to international developed markets.

However, in global recessions, correlation spikes to 0.7+. The 2008, 2020, and 2022 downturns saw high international correlation, reducing diversification benefit precisely when needed. This surprise causes many investors to abandon international allocation. The professional response is to acknowledge regime risk but maintain allocation, knowing that diversification benefits return during normal times.

Common mistakes

Assuming historical correlations predict future correlations: A 10-year rolling average of stock-bond correlation doesn't predict the next year's correlation, especially if market conditions shift (e.g., inflation spike, policy change). Use forward-looking estimates based on current conditions.

Treating all diversifiers equally: Not all low-correlation assets reduce tail risk equally. Managed futures might show low correlation but experience their own tail events. Evaluate diversifiers' stress-period behavior, not just normal-period correlation.

Overfitting correlations to recent data: If correlations were low during the previous recession, assuming they'll remain low in the next recession is dangerous. Each recession has unique characteristics; regime assumptions should be stress-tested.

Neglecting second-order correlations: When three assets show low pairwise correlations, their joint correlation (all three moving together) might be high. A diversified portfolio of 30 stocks with low average pairwise correlation can still experience high joint correlation during systemic stress.

Concentrating in "diversifiers" that all share a common risk factor: Many "diversified" portfolios hold stocks, bonds, hedge funds, and commodities—but all benefit from liquidity. During liquidity crises, everything falls. True diversification requires holdings with different underlying risk sources.

FAQ

What's a safe correlation level for diversification?

Correlations below 0.3 provide meaningful diversification benefit in normal times. Correlations below 0 provide benefit even during stress. However, be skeptical of any claimed correlation below 0.2 during crises; most assets show higher crisis correlation. Build portfolios assuming 0.3–0.5 crisis correlation, even if normal correlation is lower.

Should I diversify across many assets or concentrate in the best diversifiers?

Some concentration in best diversifiers is efficient—30% long-volatility strategies might provide better risk reduction than 5% each across six different alternatives. However, concentration in any single diversifier introduces new idiosyncratic risk. Target 5–10% minimums in each diversifier to maintain benefit while managing concentration.

How do I measure correlation in my portfolio?

Calculate rolling 60-month (5-year) correlations between major asset classes in your portfolio. Compare normal-period correlations (bull markets) to crisis-period correlations (bear markets). The difference reveals regime risk. Academic sources like SSRN publish correlation matrices across assets; use these as anchors for your portfolio design.

Does owning a stock index fund reduce idiosyncratic risk?

Yes. An S&P 500 index fund holds 500 companies, eliminating nearly 95% of idiosyncratic risk. You're left holding systematic (market) risk, which cannot be diversified through more stocks. This is why passive index funds are efficient for equity exposure—they eliminate avoidable risk at minimal cost.

What if correlation becomes very high during crisis—does my portfolio fail?

Not necessarily. A well-built portfolio with some uncorrelated or negatively correlated assets will experience higher than expected drawdowns during crisis correlation spikes, but not catastrophic failure. The barbell strategy (high bond allocation + high stock allocation) benefits from crisis correlation spillovers.

Can I use options to hedge correlation risk?

Yes, but expensively. Buying portfolio insurance (put options on your portfolio) protects against tail risk but costs 1–2% annually in premiums. This is sometimes worthwhile (e.g., before known election outcomes, geopolitical risks), but regularly hedging with options destroys returns. Better to build correlation-resilient portfolios structurally.

Should I avoid assets with high historical correlation to stocks?

Not automatically. An asset with 0.7 correlation to stocks but a 12% expected return relative to stocks' 8% return might still be attractive, depending on your risk budget and expected tail behavior. Evaluate risk-adjusted return, not correlation in isolation.

Related concepts

• Expected Value Basics for Investors
• Reframing Risk as the Source of Return
• Fixed Dollar Sizing
• Investment Policy Statement

Summary

Portfolio risk emerges from the correlation structures of its components. Understanding how multiple risk types combine—recognizing that low or negative correlation creates diversification benefit while high correlation concentrates risk—is foundational to building resilient portfolios.

The critical insight is that correlations are unstable and regime-dependent. Normal-period diversification may fail during crises when correlations spike upward. Professional portfolio builders address this by constructing portfolios resilient across multiple correlation regimes: barbell structures, inclusion of uncorrelated diversifiers, and explicit modeling of stress-period correlation behavior.

By measuring and monitoring correlation structures in your portfolio, you move from intuitive diversification ("I own stocks and bonds") to systematic correlation management ("I own stocks and bonds with 0.3 normal correlation and 0.6 crisis correlation, requiring X% uncorrelated alternatives for resilience").

Next

→ Reframing Risk as the Source of Return