How do correlation trading strategies work, and why is implied correlation often higher than realized correlation?

Correlation trading involves taking positions that profit from changes in the co-movement between assets. While related to dispersion trading, correlation can also be traded more directly through correlation swaps, basket options, and structured products.

Implied vs Realized Correlation:

Implied correlation is extracted from index and component option prices:

rho_implied = (sigma_index^2 - Sum(w_i^2 x sigma_i^2)) / (Sum_i_!=_j(w_i x w_j x sigma_i x sigma_j))

Historically, implied correlation exceeds realized correlation by 5-15 points because of the correlation risk premium --- investors pay to hedge against the scenario where all assets fall together.

Direct Correlation Instruments:

Loading diagram...

Worked Example:

Crestline Capital enters a 6-month correlation swap on the top 5 stocks in the Northfield 100 Index:

• Strike (fixed) correlation: 0.55
• Notional: $10M\n- Position: short correlation (pays if realized < strike)\n\nOver 6 months, the average pairwise realized correlation is 0.38.\n\nPayoff = (0.55 - 0.38) x$10M = +$1,700,000\n\nWhy Implied Correlation Is Typically Elevated:\n1. Hedging demand: Portfolio managers buy index puts, pushing up index IV and implied correlation\n2. Crash premium: During crises, stocks become highly correlated ("correlations go to 1 in a crisis"), and investors pay for this crash protection\n3. Structured product issuance: Banks sell worst-of products that embed short correlation, hedging by buying correlation in the interdealer market\n\nPortfolio Management Applications:\n- Hedge against correlation spikes during systemic events\n- Express views on macro regime changes (risk-on vs risk-off)\n- Enhance yield through selling the correlation risk premium in calm markets\n\nRisk Warning:\nCorrelation trades have extreme negative skew when you are short correlation. The 2008 financial crisis saw realized correlations surge to 0.85+, causing enormous losses for short correlation books.\n\nPractice correlation analysis in our CFA Derivatives question bank.