Discriminatory versus uniform electricity auctions in a duopolistic competition scenario with learning agents

Electricity industries are worldwide transitioning from centrally regulated systems to decentralized and organized markets, i.e., power exchanges. The specific characteristics of electricity markets, i.e., inelastic demand, reduced number of sellers and repeated interaction among sellers, require proper attention in the policy design so to guarantee market efficiency. In particular, this paper investigates the properties of the clearinghouse double-auction comparing two standard price mechanisms, i.e., discriminatory and uniform. A learning-in-games approach is used to derive results in the context of "infinitely" repeated games. We consider an inelastic demand and we model the seller\´s decision-making process with an adaptive evolutionary learning algorithms proposed by Ma-rimon and McGrattan. This algorithm is characterized by requiring minimal information and by being game-structure independent. Two duopolistic scenarios are examined. A low demand situation, where one seller can satisfy all the demand, and a high demand situation, where both sellers are necessary to match demand. Simulations lead to the conclusion that market power is higher in the uniform-auction case in both economic scenarios.