Correlation Risk: Why Your "Diversified" Portfolio Isn't

Correlation risk is the danger that positions you believe are diversified actually move together, because diversification depends on how your holdings behave relative to each other, not simply on how many of them you own.

Number of Holdings vs Real Diversification

A portfolio of 20 stocks sounds diversified by the usual rule of thumb (see Portfolio Diversification: How Many Stocks Is Enough?), but if 18 of those 20 are large-cap U.S. technology companies, the portfolio behaves much more like a single concentrated tech bet than like 20 independent positions. Correlation, not count, is what determines whether your risk is actually spread out.

What Correlation Measures

Correlation is expressed as a coefficient between -1 and +1:

Correlation	Meaning
+1.0	Perfectly move together
+0.7 to +0.9	Strongly move together (common among stocks in the same sector)
0	No relationship
-0.3 to -0.7	Tend to move in opposite directions
-1.0	Perfectly move opposite

Most individual stocks within the same sector carry correlations in the 0.6-0.9 range during normal conditions, and correlations across all equities tend to rise further (often above 0.8-0.9) during sharp market-wide selloffs — precisely the moment diversification matters most, and precisely when it tends to fail.

Hidden Correlation Sources

Correlation risk often hides in places that aren't obvious from ticker symbols alone:

• Sector concentration: multiple stocks across different tickers but the same industry (semiconductors, regional banks, biotech) tend to move on the same catalysts.
• Factor exposure: growth stocks across different sectors can still be highly correlated because they share sensitivity to interest rates and risk appetite.
• Supply chain and customer relationships: a chipmaker and its largest customer can move together even if they're classified in different sectors.
• Macro sensitivity: companies with heavy debt loads, or heavy export exposure to a specific country, can all react similarly to a single macro headline (a rate decision, a currency move).
• Crypto and high-beta tech: during risk-off events, crypto assets and speculative tech stocks have shown rising correlation, undermining the idea that crypto is a fully separate asset class from equities.

Worked Example

Consider a portfolio with five positions: a chipmaker, a cloud software company, an EV maker, a fintech company, and a large-cap AI-adjacent hardware company. Individually, these look like five different "sectors." In practice, all five are:

• Growth-oriented and rate-sensitive
• Heavily owned by the same institutional and retail flows chasing similar themes
• Prone to selling off together on the same macro catalyst (a hot CPI print, a hawkish Fed statement)

A portfolio like this can carry an effective correlation closer to a single-sector tech fund than to five independent bets, even though it technically spans five GICS sectors.

How to Actually Reduce Correlation Risk

1. Check sector weightings explicitly, not just position count — no single sector should dominate an unintentionally large share of the portfolio.
2. Add genuinely uncorrelated or negatively correlated assets — this is where bonds, certain commodities, or defensive sectors (utilities, consumer staples) can serve a real structural purpose beyond just "another position."
3. Stress-test against a market-wide selloff scenario, not just normal-condition correlations — remember correlations tend to spike toward 1.0 in a crisis, so diversification benefits measured in calm markets often overstate protection during the drawdowns that matter most.
4. Watch factor exposure, not just sector labels — two stocks in different GICS sectors can still share heavy correlation if they're both high-growth, high-multiple, low-profitability names reacting to the same rate environment.
5. Revisit periodically — correlations are not static; a portfolio that was well-diversified a year ago can drift toward concentration as individual holdings' businesses or valuations evolve.

Summary

A portfolio with many positions is not automatically diversified — real diversification depends on how those positions correlate with each other, especially during the market-wide stress periods when correlations across almost all equities tend to rise. Checking sector and factor concentration explicitly, and stress-testing against a broad selloff rather than only normal-market conditions, is what separates genuine diversification from a portfolio that's diversified in name only.

---

Related reading:

• Portfolio Diversification: How Many Stocks Is Enough?
• Sector Rotation: How to Trade Market Cycles
• Small Cap vs Large Cap Stocks: Risk/Reward Tradeoffs
• Position Sizing: How to Calculate How Much to Risk Per Trade