Volatility Swap

A volatility swap pays on the difference between annualized realized volatility and a fixed strike, using a defined currency amount per volatility point.

A volatility swap is a derivative contract whose cash settlement depends on the difference between an underlying asset’s annualized realized volatility and a fixed volatility strike. A long-volatility buyer receives money when realized volatility finishes above the strike and pays when it finishes below the strike, subject to the contract’s notional, cap, and calculation terms.

The strike is agreed when the trade is executed. Option-implied volatility helps inform that strike, but the strike is not simply “the implied volatility” of one option.

Key Takeaways

  • A volatility swap references the magnitude of price changes, not whether the underlying asset rises or falls.
  • Its payoff is linear in realized volatility, usually stated with a currency amount per volatility point.
  • Volatility points and decimal volatility are different units: 20 points means 20%, or 0.20 in decimal form.
  • The contract must define observation dates, price sources, return formula, annualization, strike, notional, cap, and disruption treatment.
  • A volatility swap differs from a variance swap, whose payoff depends on volatility squared and responds more strongly to extreme moves.
  • Long volatility can lose money if realized volatility finishes below the strike, even if markets felt volatile for part of the observation period.
  • Counterparty, mark-to-market, liquidity, collateral, measurement, model, and basis risks remain material.

Contract Anatomy

TermWhat it controls
UnderlyingThe stock, index, currency, rate, commodity, or basket whose returns are measured
Observation periodThe dates over which realized volatility is calculated
Observation scheduleDaily or other specified price observations
Volatility strikeFixed annualized volatility level agreed at inception
Volatility or vega notionalCurrency amount gained or lost for each volatility-point difference
Cap or floorContractual limit on recognized volatility or settlement, if any
Settlement termsValuation date, payment date, currency, and cash-flow direction
Disruption provisionsTreatment of missing prices, exchange closures, and extraordinary events

A trade described only as “one-year index volatility” is incomplete. The exact observation and adjustment rules determine the floating value.

Payoff Formula

For a long-volatility position:

$$ \text{Payoff}_{\text{long volatility}} =N_{\text{vol}}\left(\sigma_{\text{realized}}-K_{\text{vol}}\right) $$

where:

  • (N_{\text{vol}}) is the currency amount per volatility point;
  • (\sigma_{\text{realized}}) is annualized realized volatility in volatility points; and
  • (K_{\text{vol}}) is the fixed volatility strike in the same units.

The short-volatility payoff is the negative of this amount. Filed specifications may call (N_{\text{vol}}) vega notional, although this cash-flow multiplier should not be confused with every use of option vega. If the contract instead states volatility in decimals, all inputs and the notional must use that convention consistently. Mixing 0.20 with 20 volatility points creates a 100-fold unit error.

Volatility Points, Notional, and Break-Even

Suppose a swap is quoted at USD 10,000 per volatility point with a 20-point strike. Each one-point change in final realized volatility changes the settlement by USD 10,000:

Realized volatilityDifference from strikeLong-volatility payoff
12 points-8 points-USD 80,000
20 points0 pointsUSD 0
27 points+7 points+USD 70,000
35 points+15 points+USD 150,000

The contractual break-even before fees, funding, collateral remuneration, taxes, or unwind costs is the 20-point strike. The economic break-even can be higher for the buyer after those costs.

Because realized volatility cannot be negative, the long side’s lowest simplified settlement without another adjustment is:

$$ N_{\text{vol}}(0-K_{\text{vol}}) $$

For the example, that is -USD 200,000. The upper settlement is not similarly bounded unless the contract has a volatility cap or payment cap. This payoff bound does not limit losses or liquidity demands arising from other positions, closeout terms, or a counterparty default.

How Realized Volatility Is Measured

Realized volatility is the square root of realized variance. If observed prices are (S_0,S_1,\ldots,S_n), a log return is:

$$ r_i=\ln\left(\frac{S_i}{S_{i-1}}\right) $$

Under one simplified zero-mean convention:

$$ \sigma_{\text{realized}} =100\sqrt{\frac{A}{n}\sum_{i=1}^{n}r_i^2} $$

where (A) is the annualization factor and multiplying by 100 expresses the result in volatility points. A result of 20 means 20% annualized volatility.

The governing formula may use different mean, denominator, observation-count, holiday, or disruption rules. Some contracts also prescribe treatment for stale prices, market closures, corporate actions, or early termination. Realized volatility should therefore be reproduced from the confirmation, not inferred from a charting platform’s default setting.

Combining Volatility Across Subperiods

Volatility should not ordinarily be averaged arithmetically across periods. Variance aggregates first; volatility is then the square root of the combined variance.

Assume a one-year observation period has equal numbers of observations in two halves. Realized volatility is 30% annualized in the first half and 10% annualized in the second. Under a simplified equal-weight convention, full-period variance is:

$$ \sigma^2_{\text{full}} =0.5(0.30^2)+0.5(0.10^2) =0.05 $$

Therefore, full-period realized volatility is:

$$ \sigma_{\text{full}}=\sqrt{0.05}=22.36\% $$

The answer is not the 20% arithmetic average of 30% and 10%. With a 20-point strike and USD 10,000 per point, the simplified long-volatility payoff is approximately:

$$ \$10{,}000\times(22.36-20)=\$23{,}600 $$

The example assumes equal observation weights and identical annualization conventions. Actual settlement must be rebuilt from the contract’s individual price observations.

Observation and Adjustment Checklist

Contract choicePotential effect
Log returns versus simple returnsChanges each measured observation
Zero mean versus sample-mean adjustmentChanges measured variance before taking its square root
Actual versus expected observation countChanges annualization when observations are missing
Closing price versus special settlement valueCan change the final return materially
Dividend or corporate-action adjustmentCan prevent or create measured moves around ex-dates
Immediate versus forward startDetermines which price changes enter the calculation
Disrupted-day treatmentMay omit, postpone, or replace an observation

For review or dispute work, preserve source prices, timestamps, index versions, observation counts, adjustment notices, and the calculation output. A chart’s displayed historical volatility is not sufficient evidence unless its methodology matches the confirmation.

Worked Example

Assume a one-year long volatility swap has:

  • 20-point volatility strike
  • USD 10,000 volatility notional per point
  • 25-point realized volatility at settlement

The payoff is:

$$ \text{Payoff} =\$10{,}000\times(25-20) =\$50{,}000 $$

The long-volatility buyer receives USD 50,000. If realized volatility finishes at 16 points instead, the payoff is:

$$ \$10{,}000\times(16-20)=-\$40{,}000 $$

The buyer then pays USD 40,000. These simplified calculations exclude collateral, discounting, fees, cap effects, and early termination.

Volatility Strike Versus Implied Volatility

Implied volatility is backed out from an option price for a particular strike and maturity under a pricing model. An underlying can have many implied volatilities at once because options across strikes form a volatility smile or skew.

A volatility-swap strike is one fixed contractual number for the swap’s observation period. Dealers may use the wider option surface, expected future variance, replication costs, convexity, liquidity, and supply and demand to determine it.

The strike should not be described as the current at-the-money implied volatility without qualification. Even if an at-the-money quote and a swap strike happen to be close, they represent different contracts and exposures.

Why Volatility and Variance Strikes Differ

Volatility is the square root of variance. Because the square-root function is nonlinear, the fair volatility-swap strike is generally not obtained by simply taking the square root of a fair variance-swap strike. The difference is commonly associated with a convexity adjustment.

This distinction also affects hedging. Under idealized assumptions, a variance payoff has a close relationship to a broad strip of options. A volatility payoff applies a square root to realized variance, making direct replication more difficult and model-dependent.

Simplified Convexity Example

Suppose there are only two equally likely variance outcomes: 100 variance points or 900 variance points. These correspond to realized volatilities of 10 and 30 points.

  • Expected variance is ((100+900)/2=500) points.
  • The square root of expected variance is approximately 22.36 volatility points.
  • Expected realized volatility is ((10+30)/2=20) points.

The 2.36-point difference appears because taking an expectation and taking a square root do not commute. Since the square-root function is concave, the expected square root is no greater than the square root of the expected value. This is the core intuition behind the convexity adjustment between volatility and variance strikes.

This two-outcome example is not a pricing model. Market-implied probabilities, discounting, caps, jumps, volatility-surface dynamics, hedging costs, and risk premiums also affect an executable strike.

How a Cap Changes the Payoff

Assume the worked trade has a 35-point realized-volatility cap. The recognized volatility is:

$$ \sigma_{\text{recognized}}=\min(\sigma_{\text{realized}},35) $$

If observed realized volatility reaches 42 points, settlement still uses 35 under this hypothetical term. The long side receives USD 150,000 rather than USD 220,000:

$$ \$10{,}000\times(35-20)=\$150{,}000 $$

A cap on recognized volatility is not the same as a separate dollar-payment cap. The confirmation must state which variable is limited and how the cap interacts with disrupted observations and early termination.

Volatility Swap Versus Variance Swap

FeatureVolatility swapVariance swap
Floating measureRealized volatilityRealized volatility squared
Simplified payoffLinear in volatilityConvex in volatility
Common size unitCurrency per volatility pointCurrency per variance point or converted vega notional
Response to extreme movesDirect through final realized volatilityMagnified through squared returns
ReplicationMore model-dependentMore directly linked to an option strip under idealized assumptions

With a strike of 20, compare the unscaled payoff differences:

Realized volatilityVolatility-swap differenceVariance-swap difference
10(-10)(10^2-20^2=-300)
20(0)(0)
30(+10)(30^2-20^2=+500)

These are payoff units, not comparable dollar amounts. Each contract requires its own notional convention.

Volatility Swap Versus Options and VIX Products

InstrumentWhat primarily drives payoffImportant distinction
Volatility swapRealized volatility over a defined periodOTC terms and linear final-volatility payoff
Variance swapRealized variance over a defined periodConvex exposure to volatility
OptionUnderlying price relative to strike, with time and volatility effectsPayoff is directional and nonlinear in price
VIX futureFutures price settling under the VIX methodologyStandardized exposure to a forward-looking option-implied volatility index, not the swap’s realized-volatility series

A delta-hedged option position can have volatility exposure, but its outcome also depends on gamma, vega, theta, transaction costs, discrete hedging, and the path of the underlying. A volatility swap packages a different contract-defined exposure.

The VIX Index measures the option-implied market expectation of 30-day S&P 500 volatility under Cboe’s methodology. A VIX future settles through a special VIX quotation. Neither is a contract to pay the realized volatility of an arbitrary stock or portfolio over the same dates as a customized volatility swap.

Why Market Participants Use Volatility Swaps

Expressing a realized-volatility view. A trader can take a position on whether realized volatility will finish above or below a fixed strike.

Hedging volatility-sensitive portfolios. Long volatility may offset some losses in portfolios exposed to market turbulence, but the match depends on the underlying, horizon, path, and contract terms.

Relative-value analysis. Participants can compare volatility across assets, maturities, or related option markets. Different measurement and liquidity conventions prevent these comparisons from being risk-free arbitrage.

Separating direction from movement magnitude. The payoff does not directly depend on whether returns are positive or negative. However, volatility often changes asymmetrically with market direction, so practical exposures are not completely independent of directional conditions.

Path and Timing Matter

Final price alone does not determine realized volatility. Two paths can start and finish at the same values but produce very different volatility:

  • a smooth path with small daily changes produces lower realized volatility
  • a path with repeated large gains and losses produces higher realized volatility

Timing also matters. A severe move before the observation period does not enter the calculation. A late large move can materially change the final annualized result. A position can also show a large interim mark-to-market gain and later settle below strike if subsequent observations are calm.

Interim Valuation and Collateral

Partway through the contract, observed squared returns are known, but final realized volatility is not. The valuation must combine accrued realized variance with a distribution for remaining variance and then apply the square root, cap, discounting, and closeout terms.

This is more than adding “realized volatility so far” to “expected volatility later.” As the two-subperiod example shows, variance combines across time before the square root is taken. The nonlinear step makes interim value sensitive to the distribution of future variance, not only a single average forecast.

A large early move can increase expected settlement and trigger collateral from the short side. Later calm observations can dilute its annualized effect, so an interim gain is not the same as final cash settlement. Collateral reduces unsecured exposure but does not guarantee liquidity, eliminate gap risk, or make a bespoke position easy to replace.

Risks and Limitations

  • Realized-volatility risk: Long positions lose when realized volatility is below strike; short positions lose when it is above strike.
  • Jump and gap risk: Large moves can rapidly change expected settlement and collateral requirements.
  • Measurement risk: Observation time, annualization, holidays, missing prices, and return conventions affect the result.
  • Convexity and model risk: Valuation and hedging require assumptions about the relationship between volatility, variance, and the option surface.
  • Volatility-surface risk: Skew and smile changes affect mark-to-market value even though final settlement uses realized volatility.
  • Counterparty risk: An OTC gain is a contractual claim, subject to collateral and closeout terms.
  • Liquidity risk: Bespoke maturities and underlyings may be costly or impossible to unwind promptly.
  • Collateral risk: Mark-to-market losses can require cash before final settlement.
  • Basis risk: The swap’s underlying and observation period may not match the portfolio exposure being hedged.
  • Cap risk: A cap can materially limit the long side’s payoff during extreme volatility.
  • Unit and aggregation risk: Mixing decimals with volatility points or averaging volatility instead of variance can create large valuation errors.
  • Leverage: Dollar sensitivity per point can be large relative to initial cash posted.

U.S. Regulatory Context

The U.S. regulatory classification depends partly on the reference. A volatility swap based on a single security or narrow-based security index can fall under the SEC security-based-swap framework. A volatility swap based on a broad-based index or another non-security reference can fall under the CFTC swap framework. Product terms and current rules determine the result.

How to Evaluate a Volatility Swap

  1. Confirm the underlying, observation period, price source, currency, and settlement date.
  2. Reproduce the realized-volatility formula, including return type, annualization, and disruption treatment.
  3. Confirm whether volatility is quoted as decimals or points and identify the currency notional per unit.
  4. Review the strike source, cap, floor, early termination, and calculation-agent provisions.
  5. Compare the swap with available option, variance-swap, and listed-volatility exposures without assuming they are interchangeable.
  6. Stress calm, volatile, and jump paths as well as interim collateral calls.
  7. Assess counterparty, closeout, liquidity, model, and basis risk.
  8. Reconcile interim values to accrued observations, remaining implied variance, convexity, caps, collateral, and discounting.

Official and Primary Sources

  • Variance Swap: A contract whose payoff depends on realized volatility squared.
  • Volatility: The return-dispersion measure used in the swap’s floating leg.
  • Implied Volatility: Option-implied information that can help inform, but is not identical to, a volatility-swap strike.
  • Vega: The sensitivity of an option’s value to implied volatility.
  • Volatility Surface: Implied volatilities across strikes and maturities used in pricing and hedging analysis.
  • VIX Futures: Standardized futures on the forward value of the VIX Index.
  • Option: A derivative with strike-based asymmetric payoff and volatility-sensitive value.
  • Counterparty Risk: The risk that the other party cannot meet the swap settlement or collateral obligation.

FAQs

Is the volatility strike the same as implied volatility?

No. The strike is a fixed contractual level for the swap’s observation period. Option-implied volatilities across many strikes can help inform it, but one option’s implied volatility is not the swap strike by definition.

Can a volatility swap gain when the underlying price falls?

Yes. The payoff depends on the magnitude of returns, not their direction. Large negative or positive moves can both increase realized volatility.

Can the underlying finish unchanged while realized volatility is high?

Yes. Repeated gains and losses can offset in price while still producing large return dispersion over the observation period.

Why is a volatility swap not the same as a variance swap?

A volatility swap pays linearly on realized volatility. A variance swap pays on volatility squared, so extreme moves have a larger effect and the notional units differ.

Check Your Understanding

Loading quiz…

This article is general financial education, not personalized investment, trading, or legal advice. Volatility swaps are complex leveraged derivatives whose calculation and governing documentation determine the actual exposure.

Browse Financial Instruments