Real options models in electricity capacity generation

This dissertation exposes new real options models inspired by power plant investment problems. The starting point of the work is a real options capacity expansion model for a power industry under perfect competition (Chapter 3). We use the equivalence between perfect competition’s equilibrium and the supremum of the social welfare to formulate a stochastic control problem. The industry we set up has a convex cost function and a welfare which is not additively separable in its several production technologies. We investigate whether the optimality of the myopic behavior (i.e., the stochastic control/optimal stopping equivalence) which usually prevails for such type of real options models holds in our case. The subsequent chapters (4, 5 and 6) consider myopic real options models, i.e., optimal stopping problems. Chapters 4 and 5 are devoted to multi-asset optimal stopping problems with linear reward, i.e., optimal exercise of perpetual multi-asset exchange options. We successively consider the cases where (1) all assets are geometric Brownian motions (Chapter 4) and where (2) several types of Ito-diffusions are involved in the exchange (Chapter 5). Chapter 6 presents possible applications of these results to power plant investment. We close this dissertation by a discussion on real options’ ability to dampen boom and bust cycles in power investment.
(FSA 3) -- UCL, 2011