PRICING TEMPERATURE DERIVATIVES UNDER WEATHER FORECASTS

Abstract

We investigate the pricing of temperature derivatives under weather forecasts modeled by initially enlarged filtrations. For this purpose, we introduce a mean-reverting temperature model with seasonality and derive expressions for the so-called forward temperature. Although our analysis focuses on cumulative average temperature (CAT) futures, the presented derivation techniques can likewise be applied to other weather derivatives such as heating degree day (HDD) or cooling degree day (CDD) futures, for instance. We also treat option pricing and utility maximizing portfolio selection in temperature markets under additional information on future weather behavior. We finally prove an anticipative sufficient stochastic minimum principle in an enlarged filtration setup and apply the result to minimal variance hedging of temperature derivatives under weather forecasts. In this context, we derive explicit minimal variance hedging portfolios for different weather-related claims.