Skip to content
  • Overview
  • Curriculum
    • FLUMental maths and numerical fluency
    • TVMTime value, rates and linear products
    • OPTOptions: fundamentals and arbitrage
    • PRCOption pricing models
    • GRKThe Greeks and hedging
    • VOLVolatility
      • 1Implied against realised

        • Implied against realised, and the surface
      • 2Trading volatility

        • Trading volatility: gamma scalping, events and the weekend
      • 3Variance and volatility derivatives

        • Variance swaps, the log contract and what the VIX is
      • 4Relative-value volatility

        • Relative value: dispersion, skew trades and correlation
    • EXOExotics and structured products
    • SCStochastic calculus

Practise

  • Question bank
  • Mental arithmetic
  • Market simulator
  • Arbitrage trees
  • Horse racing
  • Bid book
  • Screening tests
  • Mock papers

Reference

  • Formula reference
  • Search

Your record

  • Review queue
  • Progress
  • Leaderboard
  • Profile
  • Invite friends
AccountSend feedback
  1. Curriculum
  2. /Derivatives and options
  3. /Volatility
  4. /Relative-value volatility

Relative value: dispersion, skew trades and correlation

VOL · Chapter 4·13 min read·Asked at Optiver, SIG, Citadel Securities, Akuna

Assumes Variance swaps, the log contract and what the VIX is.

After this lesson you should be able to

  • Derive implied correlation from index and single-name volatilities.
  • Explain what a dispersion trade is long and short.
  • Describe a skew trade and the risk it carries.

Once you can price volatility, the interesting trades are between volatilities rather than on one. Index against single names is a correlation trade; one strike against another is a skew trade; one expiry against another is a term-structure trade. Each isolates a parameter that is not directly quoted anywhere.

Equation 4.1

Index variance from its constituents

The index is a portfolio, so its variance carries a cross term — and that cross term is where correlation enters.

σI2=∑iwi2σi2+∑i≠jwiwjρijσiσj\sigma_I^2 = \sum_i w_i^2\sigma_i^2 + \sum_{i \ne j} w_iw_j\rho_{ij}\sigma_i\sigma_jσI2​=i∑​wi2​σi2​+i=j∑​wi​wj​ρij​σi​σj​
ρij\rho_{ij}ρij​
Pairwise correlations. Assuming a single ρ\rhoρ gives the implied correlation.
wiw_iwi​
Index weights.

Derivation 4.2

Implied correlation

Assume one common correlation and invert.

  1. σI2≈ρ(∑iwiσi)2+(1−ρ)∑iwi2σi2\sigma_I^2 \approx \rho\left(\sum_i w_i\sigma_i\right)^2 + (1-\rho)\sum_i w_i^2\sigma_i^2σI2​≈ρ(i∑​wi​σi​)2+(1−ρ)i∑​wi2​σi2​

    The two limits: ρ=1\rho = 1ρ=1 gives the weighted average volatility, ρ=0\rho = 0ρ=0 gives the diversified one.

  2. ρ≈σI2−∑iwi2σi2(∑iwiσi)2−∑iwi2σi2\rho \approx \frac{\sigma_I^2 - \sum_i w_i^2\sigma_i^2}{\left(\sum_i w_i\sigma_i\right)^2 - \sum_i w_i^2\sigma_i^2}ρ≈(∑i​wi​σi​)2−∑i​wi2​σi2​σI2​−∑i​wi2​σi2​​
a number that is quoted nowhere but is implied by two things that are\text{a number that is quoted nowhere but is implied by two things that are}a number that is quoted nowhere but is implied by two things that are

Proposition 4.3

The dispersion trade

Sell index volatility and buy the constituents’ volatility, in variance-matched size. Since index variance sits below the weighted-average single-name variance by exactly the diversification the correlation provides, you are short correlation: you profit when the names move but move independently, and lose when they all move together.

Holds when

  • The trade is historically profitable because index implied correlation trades above realised — index put demand bids up index volatility specifically.
  • The losing scenario is a market-wide selloff, when correlation goes to one. It is therefore short the tail.
  • Weights must be variance-matched, not vega-matched, or the exposure drifts as volatilities move.

Example 4.4

An index of two equally weighted names implies 18%18\%18%. Each name implies 30%30\%30%. What is the implied correlation?

Show the worked solutionHide the worked solution

Worked solution

  1. Formula
    σI2=w2σ12+w2σ22+2w2ρσ1σ2\sigma_I^2 = w^2\sigma_1^2 + w^2\sigma_2^2 + 2w^2\rho\sigma_1\sigma_2σI2​=w2σ12​+w2σ22​+2w2ρσ1​σ2​
  2. Substitute
    w=0.5, σ1=σ2=0.30, σI=0.18w = 0.5, \ \sigma_1 = \sigma_2 = 0.30, \ \sigma_I = 0.18w=0.5, σ1​=σ2​=0.30, σI​=0.18
  3. Solve
    0.0324=0.25(0.09)+0.25(0.09)+0.5ρ(0.09)0.0324 = 0.25(0.09) + 0.25(0.09) + 0.5\rho(0.09)0.0324=0.25(0.09)+0.25(0.09)+0.5ρ(0.09)
  4. 0.0324=0.045+0.045ρ0.0324 = 0.045 + 0.045\rho0.0324=0.045+0.045ρ
  5. ρ=0.0324−0.0450.045\rho = \frac{0.0324 - 0.045}{0.045}ρ=0.0450.0324−0.045​
  6. Answer
    ρ=−0.28\rho = -0.28ρ=−0.28

Sanity check. Negative, which is a real answer here: two names at 30%30\%30% that were uncorrelated would give an index at 21.2%21.2\%21.2%, so an index implying 18%18\%18% requires them to offset. If this came from live quotes, the first thing to check is whether the index volatility is stale.

Why index volatility is persistently rich. The flow is one-directional. Institutions hedge portfolios by buying index puts, and there is no equivalent structural buyer of single-name puts across the whole market — so index implied volatility is bid relative to what its constituents imply, and the gap shows up as an implied correlation above what subsequently realises. The dispersion trade harvests that, and it has the shape of every flow-driven premium: a small, steady return with a large and correlated loss when the hedging demand is finally justified.

The rest of this lesson is in Premium

You have read the opening. 10 more sections follow, including 4 worked examples and 3 quick checks.

Start the free 7-day trialSign in

Nothing is charged for 7 days, and you can cancel before then. Or read Forwards, futures and the cost of carry in full, free.

← Variance swaps, the log contract and what the VIX isBack to Volatility →
On this page
  • Index variance from its constituents
  • Implied correlation
  • The dispersion trade
  • Worked example

QuantMax · 141 lessons · 1342 questions · c5c0caa

  • Premium
  • Arbitrage trees
  • Horse racing
  • Invite friends
  • Account
  • About QuantMax
  • Terms
  • Privacy

Firm names identify publicly reported question patterns and nothing more. QuantMax is not affiliated with, endorsed by, or recruiting for any firm named in the curriculum. Everything you do in lessons and the question bank is kept to your account.