Skip to content

Summary

An index's volatility is not the average of its constituents' volatilities: it is the constituents' volatilities diluted by correlation. When correlation is low, stocks move independently and index vol is much lower than constituent vol; when correlation is high (crises, macro shocks), the gap compresses. This gap — the "correlation spread" — is the basis of dispersion trading, a modern structural topic the four books touch only indirectly via the volatility smile and vol regime discussions.

Concept (one paragraph)

Index variance is approximately the weighted sum of constituent variances plus all pairwise covariance terms. Define implied correlation as the ratio of index implied variance to a weighted average of constituent implied variances. The spread between a basket of single-stock implied vols and the index implied vol is therefore a view on correlation: the classic dispersion trade — short single-stock vol / long index vol — is short implied correlation (it profits when realized correlation comes in below implied); the opposite orientation is long correlation. The books' smile material explains why index smiles are steep (crash fear priced in index puts) while single-stock smiles differ [^hull-smile]; dispersion formalizes the difference as a tradable quantity rather than an oddity; for the modern trade's mechanics see the practitioner primer.[^bnp-dispersion]

Mechanics (one paragraph)

The classic dispersion trade is short a basket of single-stock straddles (or strangles) against a long index straddle (or variance swap), sized so that vega is hedged — the trade's P&L then isolates realized correlation versus implied correlation. When realized correlation comes in below what was implied (typical in calm markets), individual stocks realize more idiosyncratic movement than the index, and the short-stock-vol / long-index-vol orientation collects theta faster on the single-name side than it loses on the index side — i.e., it profits from realized correlation coming in below implied (short correlation). Position sizing is correlation-sensitive: vega-neutral is not correlation-neutral, and the trade typically shows losses in high-correlation events, so it is usually run with defined-risk structures (e.g., buying index puts as crash insurance rather than naked long index straddles).

Edge Source (one paragraph)

The edge candidates are: (1) a structural correlation risk premium — implied correlation tends to exceed subsequent realized correlation in calm regimes, analogous to the index-level variance risk premium the books describe; [^tomic-vol] (2) overpricing of idiosyncratic event moves — single-stock implied vol clusters around earnings dates, and the aggregate of event-implied vol often exceeds realized; (3) flow effects — index options carry crash-hedging demand that inflates index vol relative to a pure diversification model. All three are regime-dependent: the premium has episodically vanished or inverted (e.g., when idiosyncratic stock vol collapses or correlation spikes), which is why the trade belongs in the regime layer, not as an unconditional sell.

Correlation as a Regime Variable

  • Correlation regimes map naturally onto the regime layer: low-correlation/bull-grind states favor dispersion harvesting; high-correlation stress states are exactly when the trade loses and when short-premium strategies generally bleed. See Regime Definition for where correlation state should be encoded alongside realized-vol and term-structure states.
  • Practical signals used by practitioners (approximate, unverified): rolling realized correlation of index constituents, implied-correlation indices published for major benchmarks, and the index-versus-constituent implied vol spread itself.
Regime Typical correlation Index vs constituent vol Dispersion stance
Calm grind Low Index vol ≪ stock vol Harvest: short correlation (short stock vol vs long index vol)
Transition Rising Gap narrowing Reduce size; tighten risk
Stress High (0.7+, approximate) Gap compressed or inverted Trade loses; stand down or flip

Links

References

  • [^hull-smile] Hull, J., Options, Futures and Other Derivatives, 8th ed., 2012 — chapter on volatility smiles.
  • [^natenberg-vol] Natenberg, S., Option Volatility and Pricing, 1994 — volatility chapters.
  • [^tomic-vol] Chen, M. & Sebastian, J., The Option Trader's Hedge Fund, 2012 — volatility chapters.
  • Driessen, J., Maenhout, P. & Vilkov, G., "The Price of Correlation Risk: Evidence from Equity Options" (2009), Journal of Finance — canonical empirical study of the correlation risk premium.
  • Avellaneda, M. & Cont, R. literature on variance swaps; Cboe indices (e.g., Implied Correlation Index, CBOE S&P 500 Implied Correlation Index, ticker KCJ-class) documentation: https://www.cboe.com/tradable_products/vix/ (see related indices)