Estimating sensitivities of temperature-based weather derivatives

Pricing of temperature-based weather derivatives has been studied in the literature; however, there is no analysis of the estimation of the sensitivities of weather derivatives in a stochastic model of temperatures. We use pathwise derivative and kernel methods to derive Monte Carlo estimators for the sensitivity (Greeks) of temperature-based weather derivatives. These sensitivities can be used by investors for choosing the most suitable weather contracts for partial hedging or speculation. Temperature data from New York, Atlanta and Chicago are used in the discussion of numerical results.