Dispersion trading: hedge ratios, rebalancing, and risk checks
Learn how to size index and component volatility legs, align vega and variance exposure, rebalance the basket, and stress correlation and liquidity
Direct answer
A dispersion trade compares volatility exposure in an index with volatility exposure in a basket of its components. The practical challenge is not naming a long and short leg; it is choosing weights, aligning maturities, managing Greeks, rebalancing the basket, and surviving correlation and liquidity shocks. The framework below is an educational risk map, not a trade recommendation.
What this article adds
Implied correlation and dispersion trading explains why index variance contains component variance and covariance. This companion focuses on implementation questions:
Start with the variance identity
For an index with weights wᵢ, component variances σᵢ², and pairwise correlations ρᵢⱼ, a simplified variance expansion is:
σ_index² = Σ(wᵢ²σᵢ²) + ΣΣ(wᵢwⱼρᵢⱼσᵢσⱼ)
The cross terms represent co-movement. A component-volatility leg and an index-volatility leg therefore do not offset by matching the number of option contracts or the notional value alone. The intended hedge must specify whether it is matched by variance notional, vega, expected move, premium, or another risk measure.
Choose the hedge objective before the ratio
Write one sentence that describes what should be neutral at entry:
1. Vega-neutral: a small parallel volatility move has limited first-order P&L impact 2. Variance-notional neutral: modeled variance exposures are scaled to a chosen amount 3. Premium-neutral: initial option premium is approximately balanced 4. Delta-neutral: spot movement is hedged at the selected snapshot
These objectives are not interchangeable. A book can be vega-neutral while retaining gamma, theta, skew, volga, dividend, or event risk. If the objective is unclear, the hedge ratio cannot be audited after the trade.
A simple vega-ratio example
Suppose the index leg has an absolute vega of −$120,000 for a one-volatility-point parallel move. The selected component basket has aggregate vega of +$30,000 for the same move per basket unit.
The first-order vega ratio is:
index hedge ratio = 120,000 ÷ 30,000 = 4 basket units
This says four basket units offset the measured vega at that snapshot. It does not say the position is variance-neutral, delta-neutral, or safe through an earnings event. Recalculate when prices, implied volatility, time to expiration, or constituents change.
Build the basket from exposure, not convenience
Document:
Using only the most liquid names can create a basket that is easy to trade but no longer represents the index variance you intended to compare. Using every name can create operational and execution risk. The residual is a risk item, not a footnote.
- index methodology and current component weights
- included and omitted names, with a residual exposure estimate
- option expirations, strikes, exercise styles, and settlement methods
- price, volatility, dividend, and borrow data timestamps
- liquidity limits and the maximum size per component
Why vega matching does not finish the hedge
After the initial ratio, inspect at least these exposures:
A component basket can gain idiosyncratic volatility around earnings while the index leg behaves differently. A parallel-volatility hedge does not neutralize a jump, a skew change, or a correlation shock.
- delta and beta to index and component spot moves
- gamma from different strikes and maturities
- theta and weekend or holiday decay
- skew and wing sensitivity
- volga or sensitivity to a change in volatility itself
- dividend and stock-borrow assumptions
- earnings, macro releases, and sector concentration
Rebalancing rules should be mechanical
Before entry, define:
1. the Greek or variance measure that triggers a rebalance 2. an absolute and percentage threshold 3. the minimum liquidity and maximum transaction cost allowed 4. whether rebalancing occurs on a schedule, after events, or both 5. how dividends, rolls, corporate actions, and missing quotes are handled 6. who can override the rule and how the override is recorded
Rebalancing too often can turn a modeled edge into spread and commission loss. Rebalancing too late can leave the book exposed to a changed index weight, event calendar, or volatility surface.
Stress the correlation regime directly
At minimum, scenario-test:
Long dispersion is often described as long component volatility and short index volatility, but synchronized selloffs can raise index volatility and correlation together. A favorable long-run thesis does not remove short-horizon path risk.
- index and component volatility up together
- correlation rising while component volatility also rises
- correlation falling with a few large single-name gaps
- index skew steepening and component skew flattening
- one or more names becoming untradeable
- bid-ask spreads widening while the short leg is difficult to close
Measure realized results with the same definitions
Do not compare an implied correlation built from one basket and maturity with realized correlation from a different universe and sampling interval. Keep the same:
- constituents and weights
- return frequency and observation window
- price and corporate-action treatment
- volatility and covariance estimator
- option maturity and quote convention
Attribute P&L to component variance, index variance, correlation, delta hedging, skew, events, financing, borrow, spreads, and residual names. A single headline “dispersion return” can hide offsetting gains and losses.
TryMark dispersion worksheet
Keep one dated record with:
1. Index and component definitions, weights, and omitted residual 2. Contract terms, prices, Greeks, and hedge objective 3. Ratio calculation, quote timestamps, and liquidity assumptions 4. Rebalance thresholds, event calendar, and roll plan 5. Stress results for volatility, correlation, skew, gaps, and spreads 6. Actual fills, financing, borrow, hedging trades, and attribution
This guide explains dispersion-trade mechanics for education. It does not identify a profitable strategy, forecast correlation, or replace the terms of the listed or OTC instruments. Option vega, option gamma, and option bid-ask spread cover the supporting risk dimensions.
Common questions
What is a dispersion hedge ratio?
It is the scaling between an index-volatility leg and a component-volatility basket for a stated objective, such as vega or variance notional. There is no universal ratio independent of contracts and market conditions.
Is a long dispersion trade always short correlation?
It often has short-correlation exposure in a stylized construction, but Greeks, events, weights, skew, jumps, rebalancing, and execution prevent it from being a pure single-factor position.
Why not match the number of option contracts?
Contracts have different multipliers, prices, vegas, maturities, strikes, liquidity, and settlement terms. Equal counts can leave the intended index and component exposures badly mismatched.
How often should a dispersion basket be rebalanced?
There is no universal schedule. Define a threshold and cost rule that matches the objective, liquidity, event calendar, index reconstitution, and acceptable tracking error, then test it historically and under stress.
Can dispersion lose when component volatility rises?
Yes. A synchronized selloff can raise index volatility and correlation, while the short index leg, skew, hedging costs, and liquidity changes dominate the component-volatility gains.
Sources and further reading
Quick check
Read the guide? Check yourself with 3 questions
Question 01
Which statement best matches this guide — What this article adds?
Choose an answer to see the explanation
Options glossary
A risk-neutral price of average co-movement inferred from index and component option variances; it is not a guaranteed forecast of future realized correlation.
Read the deeper guideVegaAn estimate of how much an option value may change for a one-point change in implied volatility, all else equal; it does not measure realized volatility directly.
Read the deeper guideGammaAn estimate of how much delta may change for a small move in the underlying, all else equal; it often becomes more concentrated near expiration and near the money.
Read the deeper guide