On Pricing Weather Derivatives and Extracting Risk Loading Factors

The purpose of this work is to introduce a novel method for pricing weather derivatives. The method takes a functional data analytic approach to modeling the temperature indices via appropriate basis functions. Uncertainty in the estimates, useful for risk loading, is obtained by simple functional descriptive statistics. Except for the choice of the basis, the proposed method is non-parametric. Closed-form formulas are obtained for CME based HDD and CDD futures and options. The functional data approach allows interesting interpretation and analysis of the temperature indices; examples are the study of the rates of change in the indices, and exploring the temperature index relationship to non-temperature data.