Correlation risk is when positions you think are independent actually move together. A trader holding BHP, RIO, and FMG believes they have three positions—they don't. They have one bet on iron ore prices wrapped in three different wrappers. When correlation hits 0.8 or higher, your separate trades become a single leveraged position, and your risk compounds instead of spreading. The portfolio variance formula tells the truth: two positions with 0.9 correlation carry nearly the same risk as one position at double the size.
The math exposes the illusion. Say you risk $2,000 per position on a $100,000 account—that's 2% per trade. You take three positions you believe are diversified. If those positions have zero correlation, your combined risk is $3,464 (square root of 2000² + 2000² + 2000²). But if correlation is 0.9, combined risk jumps to $5,692—nearly triple the size of one position. You thought you had 6% total risk spread across three independent bets. You actually have 5.7% risk on one highly leveraged bet. The market moves against your thesis once, and all three positions bleed together.
Take a real example with tech stocks. You allocate 2% risk to AAPL, 2% to MSFT, 2% to NVDA—thinking you've spread $6,000 across three positions on your $100,000 account. Pull a 60-day correlation matrix and you see AAPL-MSFT at 0.75, MSFT-NVDA at 0.82, AAPL-NVDA at 0.68. Average correlation near 0.75. Your actual portfolio risk isn't $6,000 spread independently—it's closer to $5,200 on a single bet that tech sentiment stays positive. The Fed hints at holding rates, tech sells off, and you watch three positions drop 3% simultaneously. Your 6% allocation just cost you $15,600 instead of the $6,000 you modelled for one position failing. This is correlation risk crystallising into real loss.
Traders fail here by confusing different tickers with different exposures. They hold index futures, tech ETFs, and mega-cap stocks, telling themselves they're diversified across instruments. But correlation coefficient doesn't care about wrapper type—it measures price movement synchronisation. Banks collapsed in 2008 because risk models assumed diversification across mortgage securities with different ratings. Correlation spiked to 0.95 when housing turned, and the entire book moved as one. Retail traders repeat this by holding AUD/USD, AUD/JPY, and NZD/USD, believing three currency pairs spread risk. All three have AUD exposure. The Reserve Bank cuts rates once and all three pairs drop together. You took 6% notional risk thinking you had three 2% independent positions. The market showed you that you had 5.4% on one bet: long the Australian dollar. Check your covariance matrix before you size, or the market will teach you about correlation the expensive way.
This content is educational only and does not constitute financial advice. Past performance is not indicative of future results. Always seek licensed financial advice before trading.