Working Paper: NBER ID: w11123
Authors: Ali Hortacsu; Steven L. Puller
Abstract: We examine the bidding behavior of firms competing on ERCOT, the hourly electricity balancing market in Texas. We characterize an equilibrium model of bidding into this uniform-price divisible-good auction market. Using detailed firm-level data on bids and marginal costs of generation, we find that firms with large stakes in the market performed close to theoretical benchmarks of static, profit-maximizing bidding derived from our model. However, several smaller firms utilized excessively steep bid schedules that deviated significantly from our theoretical benchmarks, in a manner that could not be empirically accounted for by the presence of technological adjustment costs, transmission constraints, or collusive behavior. Our results suggest that payoff scale matters in firms' willingness and ability to participate in complex, strategic market environments. Finally, although smaller firms moved closer to theoretical bidding benchmarks over time, their bidding patterns contributed to productive inefficiency in this newly restructured market, along with efficiency losses due to the close-to optimal exercise of market power by larger firms.
Keywords: electricity markets; bidding behavior; ERCOT; Bayesian Nash Equilibrium; market power
JEL Codes: L1; L2; L5; L9
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| adjustment costs and technological constraints (D24) | bidding behavior (D44) |
| firm size (L25) | bidding efficiency (D44) |
| firm size (L25) | strategic bidding effectiveness (D44) |
| smaller firms (L25) | productive inefficiency (D61) |
| firm size and scale of stakes (L25) | performance in strategic environment (L10) |