Worst-of Options Pricing and Implied Correlation using Copula Methods

Abstract

The importance of managing portfolio correlations has created a market for derivatives which allow investors to trade correlation risk. These products present a unique pricing problem, as effective pricing models must consider the dependencies between the option’s underlying assets. In this paper, we use various copula-GARCH models and a regime-switching model to price bivariate worst-of options. We assess the ability of different copula families to model the relationship between two assets, and compare their pricing outputs across option strikes and maturities. Finally, we extract implied correlation values from our prices and evaluate their performance as estimators for future correlation values.