A new look at the swing contract: From linear programming to particle swarm optimization

As the energy market has grown in importance in recent decades, researchers have paid increasing attention to swing option contracts. Early studies evaluated the swing contract as if it were a financial derivative contract, by ignoring its storage constraints. Aided by recent advances in artificial intelligence (AI) and machine learning (ML) technologies, recent studies were able to incorporate storage limitations. We make two discoveries in this paper. First, we contribute to the literature by proposing an AI methodology-particle swarm optimization (PSO)-for the evaluation of the swing contract. Compared to the other ML methodologies in the literature, PSO has an advantage by expanding to include more features. Secondly, we study the relative impact of the price process (exogenously given) that underlies the swing contract and the storage constraints that affect a quantity decision process (endogenously decided), and discover that the latter has a much greater impact than the former, indicating the limitation of the earlier literature that focused only on price dynamics.