As electricity market participants increasingly adopt learning‑based agents for their bidding strategies, the markets are becoming algorithmic. Evidence from algorithmic markets in other domains shows that tacit collusion can arise purely through independent learning. Moreover, electricity markets are typically oligopolistic and feature repeated interaction among a small number of participants, making them structurally susceptible to non‑competitive behavior. In light of these observations, this paper investigates whether tacit collusion may emerge in electricity markets where participants' actions are controlled by autonomous learning‑based algorithms. We model strategic bidding as a repeated game with imperfect public monitoring and simulate participants' emergent behavior using multi‑agent reinforcement learning. To assess whether the resulting behavior constitutes tacit collusion, we propose a multi‑dimensional set of criteria that go beyond simple profit comparisons against Nash equilibria. Experimental results demonstrate that the danger is realistic for electricity markets: in certain cases agents learn to sustain supra‑competitive outcomes that align with multiple tacit collusion indicators, even though the agents were never instructed to collude.
Blogger's Review: The study serves as a timely warning that algorithmic trading can unintentionally foster anti‑competitive outcomes in power markets, urging both regulators and industry players to develop technical safeguards and policy frameworks.