Variance Swap Replication and Realized Variance Annualization

Summary

The author describes estimating a variance swap’s fixed variance strike from observed option prices using a method attributed to Derman, then asks how to compute realized variance for payoff calculation. The central uncertainty is whether realized variance should be annualized for a contract with a stated maturity, especially when the strike estimate is not annualized. The author reports that the short variance swap profit and loss appears implausible and notes a Sharpe ratio, but provides no supporting data or calculation details.

The post highlights a practical consistency issue: the realized variance convention and the variance strike must use compatible scaling when calculating a payoff. However, it is a question rather than an answered explanation. It does not specify the realized-variance estimator, sampling frequency, treatment of days, or contract multiplier, so it cannot establish which annualization convention is correct or validate the reported performance. Readers would need contract terms and a clearly defined variance calculation to resolve the issue.

Key ideas

• The author estimates a variance swap strike from option prices using a cited method.
• The post asks how realized variance should be calculated for the swap payoff.
• The author questions whether realized variance should be annualized over the contract maturity.
• Strike and realized variance need consistent scaling conventions in a payoff calculation.
• The document supplies no definitive estimator or resolution to the annualization question.

Tags

# How to Compute the payoff of Var Swaps, which I have replicated


# How to Compute the payoff of Var Swaps, which I have replicated












I used Derman(1999) method, to calculate the fixed Kvar for Variance Swaps using actual option price data. The first Pic Shows the outcome. (ignore the 0s).

Now the profit and loss of short var swaps is totally wrong, the Sharpe ratio is 2.

I want to ask how the realized Var is actually computed, to calculate payoff. Do I need to annualise it, currently I multiply RealisedVar with 365/45. as Maturity is 45 days. P&L is depends on the multiplication factor. Kvar is not anualised so why should realized var be?

see https://en.wikipedia.org/wiki/Variance_swap#cite_note-FINCAD-1

Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.