Many electricity markets, both in the U.S. and around the world, use auctions to procure power and these auctions are complex strategic environments. Firms that participate in these auctions to sell power can vary in their level of strategic sophistication which impacts not only the firms' profitability but the overall efficiency of the market. For this reason, understanding how heterogeneity in strategic sophistication affects efficiency is a primary concern for market design and public policy.
In this project, the PIs develop a model of bidding behavior that builds on game theoretic models from auction theory and bounded rationality models, to provide a rigorous framework to analyze strategic sophistication in auction settings. The PIs apply the model to real-world data, using detailed firm-level data from the Texas electricity market to categorize bidders' strategic behavior. The analysis estimates the increase in the cost of electricity production that results from some firms bidding with lower levels of strategic sophistication, and it quantifies the profit gains of using alternative bidding strategies. To do so, the PIs develop an algorithm to compute equilibrium using a set of firms, their characteristics, and marginal costs. Using this algorithm, the project will document the extent to which mergers that increase strategic sophistication can improve the efficiency of the market even if the mergers increase market concentration.
