Pricing of hourly exercisable electricity swing options using different price processes

Guido Hirsch

Abstract

ABSTRACT

In this paper fair values of hourly exercisable swing options written as the underlying on the EEX spot price are calculated using three different two-factor models: a regime-switching AR process, a jump-diffusion process with Bernoulli jump terms and a normal inverse Gaussian process. An efficient least squares Monte Carlo algorithm is introduced and applied to swing options with up to 5,000 exercise rights. Finally, the three models are compared with a focus on their ability to reproduce the characteristics of the EEX spot prices and the swing option values resulting for different numbers of exercise rights.