blog · ~6 min read

Correlation Risk: When Diversification Fails

Correlation risk is the danger that assets which normally move independently suddenly move together, destroying your diversification right when you need it most.

T By tradernewbie · AI-drafted, human-reviewed
#risk-management#portfolio

Correlation Risk: When Diversification Fails

"In a crisis, all correlations go to 1." Diversification is a fair-weather friend. Correlation risk is the storm that exposes it.

Correlation risk is the danger that assets which usually move independently suddenly start moving together — usually downward. It's the hidden flaw in many "diversified" portfolios.

What correlation actually measures

Correlation ranges from −1 to +1:

Value Meaning
+1 Always move together
0 No relationship
−1 Always move opposite

A portfolio that holds five assets with +0.9 correlation isn't five independent bets — it's effectively one big bet. In calm markets the difference may not show. In a selloff, every position falls together.

When correlations break

Most of the time, gold and equities have low correlation. But in a liquidity crisis — March 2020, 2008, August 2015 — investors sell everything for cash. Correlations spike toward +1 across the board:

  • Stocks fall
  • Gold falls
  • Bonds fall
  • Crypto falls
  • Even "safe havens" get sold to meet margin calls

The portfolio you thought was diversified turns out to have one hidden risk: liquidity risk. There's no asset that protects you when everyone needs cash at once.

Hidden correlation sources

1. Common currency exposure

Long EUR/USD and long GBP/USD aren't two trades — they're both bets against the dollar. If USD rallies, both lose.

2. Common sector exposure

Long Apple, Microsoft, and Nvidia looks like three stocks. In a tech selloff, it's one trade.

3. Common macro driver

Long oil and long Canadian dollar (CAD) are correlated because Canada is a major oil exporter. One factor moves both.

4. Common counterparty

If every trade is held at one broker or in one custody, you have one point of institutional failure. Counterparty risk correlates across all positions.

Measuring your real exposure

  1. Pull daily returns of every instrument you trade
  2. Compute a correlation matrix
  3. Flag any pair above +0.7 — treat them as the same position for risk purposes
  4. Sum risk across correlated positions, not just by trade count

If you have $100 risk on EUR/USD and $100 risk on GBP/USD with +0.7 correlation, your effective risk is closer to $170, not $200 — but still far more than $100.

Defenses against correlation risk

Defense How it helps
Hard cap on total open risk (e.g., 5%) Limits damage when correlations spike
Pair-wise correlation limit (no two trades > +0.7) Forces genuine diversification
Holding cash The ultimate uncorrelated asset
Hedging with index shorts Offsets correlated long exposure
Reducing size during high-vol regimes When correlations rise, shrink the book

Practical steps

  • Verify position sizing with the position size calculator — and check your combined risk, not just per-trade
  • Log the asset class of each trade in a journal and review monthly for hidden clusters
  • Stress-test your book: "If the market drops 5% tomorrow, what's my total loss?" If it's more than you can stomach, you have correlation risk

Diversification is real risk management, but only when correlations stay low. Plan for the day they don't.

AI-assisted content · Not financial advice · Trade at your own risk