Correlation 101: How Asset Relationships Shape Portfolio Risk

beginnerPublished: 2025-12-29

Stocks and bonds had -0.3 to -0.5 correlation during 2000-2020 (moved in opposite directions, Source: Russell Investments correlation analysis), which meant bonds rose in 5 of 6 years when stocks were negative (2008: stocks -37%, bonds +20%). But during 2022 rate hikes, correlation flipped to +0.5 (both fell together: stocks -18%, bonds -13%). The practical antidote: correlation tells you whether your diversification will work when you need it (negative = assets cushion each other, positive = assets amplify losses), and monitoring correlation regime changes prevents false diversification.

What Correlation Measures (Range from -1.0 to +1.0)

Correlation is a statistical measure of how two assets move together, ranging from -1.0 (perfect inverse relationship) to +1.0 (perfect positive relationship), with 0 meaning no relationship.

+1.0 (perfect positive): Assets move in lockstep (when A rises 10%, B rises 10%; when A falls 20%, B falls 20%). Provides zero diversification benefit (both crash together, both rally together). Example: S&P 500 and Nasdaq have +0.90 to +0.95 correlation (almost perfect positive—tech-heavy Nasdaq moves with broader market).

0 (no correlation): Assets move independently (A's movement tells you nothing about B's movement). Provides moderate diversification benefit (when A crashes, B might be flat or up or down—unpredictable). Example: stocks and gold have ~0 to -0.2 correlation historically (gold sometimes acts as safe haven, sometimes doesn't).

-1.0 (perfect negative): Assets move in exact opposite directions (when A rises 10%, B falls 10%). Provides maximum diversification benefit (when one crashes, other rallies by same amount—perfect hedge). Extremely rare in practice (almost no real-world asset pairs have -1.0 correlation over long periods).

The point is: negative correlation is your friend (assets cushion each other's declines). Positive correlation reduces diversification benefit (assets amplify each other's moves). Zero correlation is neutral (better than positive, worse than negative). You want portfolios built from assets with low or negative correlation to reduce total volatility.

Stock-Bond Correlation Changes by Regime (Why 60/40 Worked Until 2022)

Stock-bond correlation is not constant (it varies by economic regime, especially inflation environment). This changes the diversification benefit of holding both assets.

Low inflation / low rates regime (2000-2020): stock-bond correlation was -0.3 to -0.5 (negative). When stocks fell, investors fled to safety (bought Treasuries), bond prices rose. This created powerful diversification: 60/40 portfolios fell much less than 100% stocks during crashes (2008: 60/40 fell -22% vs stocks -37%).

High inflation / rising rates regime (1970s, 2021-2023): stock-bond correlation turned +0.3 to +0.5 (positive). Both stocks and bonds fell as the Fed raised rates (stocks hurt by slowing economy and higher discount rates, bonds hurt by rising yields). 2022 was a rare year where both fell significantly (stocks -18%, bonds -13%). 60/40 portfolios fell -16% (barely helped).

Since 1940 (long-term view): stock-bond correlation has been negative only 38% of the time (positive correlation is more common historically). But during crisis periods (2008, 2020, 2011 debt ceiling), correlation often turns sharply negative exactly when you need protection (bonds rally as stocks crash).

Example (2008 financial crisis, negative correlation): S&P 500 fell -37%. 10-year Treasuries returned +20% (massive flight to safety, yields collapsed, bond prices soared). Correlation was approximately -0.5 during the crisis (strong negative). A 60/40 portfolio fell only -22% because the 40% bond allocation gained 20%, offsetting 40% × 20% = 8% of stock losses.

Example (2022 rate hikes, positive correlation): S&P 500 fell -18%. Bloomberg Aggregate Bond Index fell -13% (Fed raised rates from 0.25% to 4.5%, bond prices tanked). Correlation was approximately +0.5 (positive). A 60/40 portfolio fell -16% (60% × -18% + 40% × -13% = -10.8% - 5.2% = -16%). Bonds didn't cushion (they amplified losses).

The durable lesson: stock-bond correlation is regime-dependent. In low-inflation, stable-rate environments (1990s, 2000s, 2010s), correlation tends negative (bonds diversify stocks). In high-inflation, rising-rate environments (1970s, 2020s), correlation turns positive (bonds stop diversifying). Over very long horizons (30+ years), the negative correlation periods (crises) matter most because they prevent catastrophic drawdowns when you're near retirement or need to withdraw.

Correlation Examples Across Domestic Asset Classes

Different asset pairs have characteristic correlation ranges based on their drivers. Understanding these helps you build portfolios that actually diversify (not just hold many assets that all move together).

S&P 500 and Nasdaq (both US equity indexes):

  • Correlation: +0.85 to +0.95 (very high positive)
  • Interpretation: Nasdaq is tech-heavy, S&P 500 is ~30% tech (Apple, Microsoft, Nvidia). They move together closely (tech drives both).
  • Diversification benefit: minimal (holding both is redundant—just own total market index instead)

Large-cap stocks and small-cap stocks:

  • Correlation: +0.75 to +0.85 (high positive)
  • Interpretation: Both are equities, both driven by economic growth, but small-caps are more volatile and sensitive to credit conditions
  • Diversification benefit: modest (small-caps outperform in expansions, underperform in recessions, but move with large-caps directionally)

Stocks and REITs (Real Estate Investment Trusts):

  • Correlation: +0.6 to +0.8 (moderate-high positive)
  • Interpretation: REITs are equity-like (publicly traded) but driven by real estate fundamentals (rents, cap rates, rates). They move with stocks but less correlated than other equity sectors.
  • Diversification benefit: moderate (REITs provide some diversification, especially during low-rate environments, but crash with stocks in severe downturns like 2008)

Stocks and gold:

  • Correlation: ~0 to -0.2 (low or slightly negative)
  • Interpretation: Gold is a safe-haven asset and inflation hedge. Sometimes rises when stocks fall (2008, 2020), sometimes doesn't (2022: both fell). Correlation varies by regime.
  • Diversification benefit: moderate (gold provides diversification during crises but is unreliable—correlation flips between negative and positive)

Stocks and investment-grade bonds (low-inflation regime):

  • Correlation: -0.3 to -0.5 (negative, diversification benefit)
  • Interpretation: Bonds rally when stocks fall (flight to safety, Fed cuts rates)
  • Diversification benefit: high (this is the foundation of 60/40 portfolios)

Stocks and bonds (high-inflation regime):

  • Correlation: +0.3 to +0.5 (positive, reduced diversification)
  • Interpretation: Both fall when Fed raises rates aggressively (stocks hurt by slowing economy, bonds hurt by rising yields)
  • Diversification benefit: low (2022 example: both fell together)

Stocks and Treasury bills (cash):

  • Correlation: ~0 (no relationship)
  • Interpretation: Cash is stable (no volatility), provides no correlation-based diversification but guarantees value won't fall
  • Diversification benefit: low (cash doesn't rally when stocks crash, but it doesn't fall either—preserves capital)

Why this matters: if you build a portfolio of S&P 500 + Nasdaq + tech sector fund, you have +0.90+ correlation across holdings (all move together, zero diversification). If you build S&P 500 + bonds + small allocation to gold, you have mix of negative, zero, and moderate positive correlations (actual diversification that reduces total volatility).

How Correlation Affects Portfolio Volatility (The Math Behind Diversification)

Portfolio volatility depends on individual asset volatilities AND correlations between assets. Lower correlation = lower portfolio volatility (even if individual assets are volatile).

Example (two assets, high positive correlation):

  • Asset A (stocks): 10% return, 18% std dev
  • Asset B (REITs): 8% return, 20% std dev
  • Correlation: +0.8 (high positive)
  • 50/50 portfolio: expected return = 9% (simple average), but volatility ≈ 17% (only slightly lower than stocks alone because high correlation means both move together)

Example (two assets, zero correlation):

  • Asset A (stocks): 10% return, 18% std dev
  • Asset B (gold): 4% return, 15% std dev
  • Correlation: 0 (no relationship)
  • 50/50 portfolio: expected return = 7%, volatility ≈ 12% (much lower than weighted average because zero correlation provides diversification benefit)

Example (two assets, negative correlation):

  • Asset A (stocks): 10% return, 18% std dev
  • Asset B (bonds): 5% return, 6% std dev
  • Correlation: -0.4 (negative)
  • 60/40 portfolio: expected return = 8%, volatility ≈ 10% (significantly lower than weighted average ~13% because negative correlation amplifies diversification)

The point is: correlation matters more than number of holdings. You can own 100 stocks and have zero diversification benefit (if all 100 are tech stocks with +0.95 correlation). Or you can own 3 assets (US stocks, bonds, international stocks) with low correlations and achieve excellent diversification (reduce volatility by 30-40% vs single asset).

Correlation Breakdown During Crises (When Diversification Fails)

During severe market crashes, correlations often spike toward +1.0 (all risky assets fall together, only ultra-safe assets like Treasuries rise). This is called correlation breakdown (diversification fails exactly when you need it most).

Example (March 2020 COVID crash): stocks fell -34% in one month. REITs fell -40%, small-caps fell -35%, international fell -32%, high-yield bonds fell -15%. Correlation across all risk assets spiked to +0.90+ (everything crashed together). Only Treasuries rallied (10-year Treasuries +5-10% during selloff). Cash was flat.

Why this happens: in panic selloffs, investors sell everything liquid to raise cash (forced liquidations, margin calls, redemptions). Fundamentals stop mattering (correlations converge to +1.0 for all risky assets). The only diversifiers are ultra-safe assets: Treasuries, cash, gold (sometimes).

2008 financial crisis: similar pattern. Stocks, REITs, commodities, high-yield bonds all fell -30% to -50% (correlations spiked to +0.85+). Only Treasuries and cash preserved value. Gold fell initially (liquidation), then rallied later.

The durable lesson: don't expect diversification across risky assets (stocks, REITs, commodities, high-yield bonds) to protect you in crashes (they all fall together). The only reliable diversifiers during crises are investment-grade bonds (especially Treasuries) and cash. This is why 60/40 portfolios include bonds (not just for different return profile, but for crisis protection).

Portfolio Construction Using Correlation (Build Uncorrelated Assets)

To build a diversified portfolio, combine assets with low or negative correlation (not just different asset classes). The goal is reduce total volatility without sacrificing too much return.

Core diversified portfolio (3-fund model):

  • 60% US total stock market (VTI or equivalent): 10% expected return, 18% std dev
  • 30% US total bond market (BND or equivalent): 5% expected return, 6% std dev
  • 10% international stocks (VXUS or equivalent): 9% expected return, 20% std dev

Correlations:

  • US stocks and bonds: -0.3 (low-inflation regime) or +0.3 (high-inflation regime)
  • US stocks and international: +0.70 to +0.80 (high positive, moderate diversification)
  • Bonds and international: -0.1 to +0.2 (low correlation)

Portfolio result: expected return ~8.5%, portfolio volatility ~11% (compared to 18% for 100% US stocks). The negative stock-bond correlation drives most diversification benefit (international adds modest benefit).

Adding uncorrelated assets (advanced):

  • 5-10% gold: adds uncorrelated/slightly negative asset (correlation ~0 to -0.2 with stocks). Provides crisis protection but lowers expected return.
  • 5-10% commodities: low correlation with stocks (~+0.3), inflation hedge, but adds volatility and complexity.
  • 5-10% TIPS (inflation-protected bonds): low correlation with nominal bonds (~+0.5), protects against inflation spikes.

What NOT to add (false diversification):

  • Multiple US stock funds (S&P 500 + Nasdaq + growth fund): all have +0.90+ correlation (redundant, not diversified)
  • Sector bets (tech fund + healthcare fund + energy fund): sectors have +0.60 to +0.80 correlation with market (stock-picking without diversification benefit)
  • Multiple bond funds (intermediate + long-term + short-term Treasuries): all have +0.80+ correlation (over-diversified for no benefit, just added complexity)

The test: calculate correlations between your holdings (use Morningstar Portfolio Manager, Portfolio Visualizer, or broker tools). If average correlation > +0.70, you're not diversified (just holding many things that move together). Target average correlation < +0.50 (mix of positive, zero, and negative correlations provides actual diversification).

Correlation Changes Over Time (Monitor Regime Shifts)

Correlation is not static (it drifts over time based on economic regime). What diversified in 2000-2020 (negative stock-bond correlation) stopped diversifying in 2021-2023 (positive correlation). You need to monitor correlation regime and adjust if needed.

Signs stock-bond correlation is turning positive (diversification benefit fading):

  • Inflation rising above 3-4% (Fed forced to raise rates aggressively → both stocks and bonds fall)
  • Fed hiking cycle (rates rising → bond prices fall, stocks fall from higher discount rates)
  • Yield curve steepening after inversion (signals Fed tightening → both assets at risk)

Signs stock-bond correlation is negative (diversification benefit present):

  • Inflation low and stable (1-2% range, Fed on hold or cutting)
  • Fed cutting cycle or paused (rates falling or stable → bonds rally when stocks fall)
  • Economic uncertainty or recession (flight to safety → bonds rally, stocks fall)

Portfolio adjustments based on correlation regime:

  • If stock-bond correlation turns positive: reduce overall risk (shift to 50/50 or 40/60 from 60/40), add cash allocation (10-20%), consider TIPS (protect against inflation without correlation to stocks). Don't abandon diversification (just adjust expectations that bonds won't cushion as much).
  • If stock-bond correlation is negative: maintain 60/40 or 70/30 (diversification is working), rebalance mechanically (sell stocks after rallies, buy bonds; reverse after crashes).

The point is: correlation regimes last 5-10 years typically. We had negative stock-bond correlation for 20 years (1998-2020), then positive for 2 years (2021-2023), now turning neutral/slightly negative again (2024-2025 as inflation cools). You don't need to trade frequently, but audit correlation annually (especially after major regime changes like inflation spikes or Fed pivots).

Detection Signals (Your Diversification Is Illusory)

You're likely falsely diversified (high correlation disguised as diversification) if:

  • You own S&P 500 + Nasdaq 100 + tech sector fund thinking you're diversified (correlation +0.90+ across all holdings—they're the same assets repackaged). Simplify to one total market fund (VTI or equivalent).
  • Your portfolio fell -18% in 2022 when S&P fell -18% even though you own "10 different funds" (sign all your funds are highly correlated equities). Check correlation matrix—if average > +0.80, you're not diversified.
  • You hold 3 bond funds (short, intermediate, long Treasuries) thinking you're diversified, but they all fell together in 2022 (correlation +0.85+ across bond durations when rates rise). Simplify to one total bond fund (BND or AGG).

You're likely well diversified if:

  • Your portfolio fell -16% in 2022 when stocks fell -18% (sign bonds or other assets cushioned slightly, even in positive correlation regime).
  • You own mix of stocks, bonds, and small allocation to gold/commodities with average correlation < +0.50 (actual diversification across drivers).
  • Your annual returns cluster tightly around 7-9% with rare -20% years (diversification is smoothing volatility, delivering consistent risk-adjusted returns).

The test: use Portfolio Visualizer (free tool) to input your holdings and check correlation matrix. If any pair has > +0.85 correlation, you're holding redundant assets (eliminate one). If average portfolio correlation is +0.60 to +0.80, you have moderate diversification (good). If < +0.50, you have strong diversification (excellent).

Next Step: Check Correlation Matrix for Your Holdings

Action (15 minutes): build a correlation matrix for your portfolio to identify redundant or highly correlated holdings.

Step 1: List your top 5-7 holdings (funds or ETFs). Example: VTI (total stock), BND (total bond), VNQ (REITs), GLD (gold), VXUS (international stock).

Step 2: Use free tool to generate correlation matrix:

  • Portfolio Visualizer (portfoliovisualizer.com): click "Asset Correlations," enter tickers, select time period (10 years), view matrix.
  • Morningstar Portfolio Manager: input holdings, view "correlation" tab.
  • Broker tools: Vanguard, Fidelity, Schwab often have correlation tools in research section.

Step 3: Interpret matrix. Look for:

  • High correlations (> +0.85): redundant holdings (consider eliminating one)
  • Moderate correlations (+0.50 to +0.80): some diversification but limited
  • Low/negative correlations (< +0.50 or negative): strong diversification

Example matrix interpretation:

  • VTI and VXUS: +0.75 (moderate-high, okay to hold both)
  • VTI and BND: -0.30 (negative, excellent diversification)
  • VTI and GLD: 0 (uncorrelated, good diversification)
  • VTI and VNQ: +0.70 (moderate, some benefit but both equities)

Step 4: Simplify or adjust. If you have multiple holdings with +0.90+ correlation, eliminate redundancy (keep one, sell others). If average correlation > +0.70, add uncorrelated asset (bonds if you're all stocks, or gold/commodities if you're stocks+bonds).

Where to start: if you don't want to analyze correlation, use simple 3-fund portfolio (US total stock, total bond, international stock) with allocation based on risk tolerance (60/30/10 for moderate, 80/15/5 for aggressive, 40/50/10 for conservative). This captures most diversification benefit without complexity.

Correlation measures how assets move together and determines whether your portfolio is actually diversified. Negative correlation (stocks and bonds in low-inflation regime: -0.3 to -0.5) provides powerful diversification (bonds rally when stocks crash). Positive correlation (stocks and bonds in high-inflation regime: +0.3 to +0.5) reduces diversification benefit (both fall together like 2022). Holding multiple assets with +0.90 correlation (S&P 500 + Nasdaq + tech fund) is false diversification (all move together). Build portfolios from assets with low or negative correlation (stocks + bonds + international, average correlation < +0.50) to reduce total volatility without sacrificing return. Monitor correlation regime changes (inflation spikes often flip stock-bond correlation from negative to positive) and adjust allocation when needed (add cash or TIPS if correlation turns positive). Check your correlation matrix annually using free tools (Portfolio Visualizer, Morningstar) to ensure you're not holding redundant assets disguised as diversification.

Sources:

Related Articles

Compound Interest: Taxable vs Tax-Advantaged Accounts

beginner

How Economic Cycles Affect Investment Outcomes

intermediate

Inflation-Protected vs Traditional Savings: TIPS and I Bonds Explained

beginner