Cambridge, MA, Dec. 18, 2024 (GLOBE NEWSWIRE) -- Many governments are addressing climate change by managing carbon emissions through a system of permits, sold in auctions at a uniform price — or the lowest winning bid, which represents the price that all winning bidders typically pay for each unit. New research by MIT Sloan School of Management associate professor Negin Golrezaei suggests that switching to a pay-as-bid auction, where each buyer pays the price they bid, could boost revenue and provide stronger incentives for companies to invest in greener technologies.
“Learning in Repeated Multi-Unit Pay-As-Bid Auctions,” published in the academic journal Manufacturing & Service Operations Management, and co-written with MIT Operations Research Center PhD student Rigel Galgana, is the first to explore how participants can learn optimal bidding strategies in multi-unit auctions over time, without prior knowledge of competitors’ strategies or values. The paper builds on Golrezaei’s earlier paper “Learning and Collusion in Multi-unit Auctions.”
“The social cost of carbon is estimated at $190 per ton by the U.S. Environmental Protection Agency, yet auction prices often fall far short of this figure,” said Golrezaei. “By adopting a market-driven approach like pay-as-bid auctions, we could better align permit prices with the true societal cost of carbon, making carbon markets more effective in driving emissions reductions.”
A key contribution of Golrezaei’s research is the development of an adaptable algorithm that addresses market dynamics. Until now, effectively evaluating the market under pay-as-bid auctions has been difficult due to the repeated interactions among participants — including power generation companies, industrial facilities, and the aviation sector — which influences market dynamics, revenue, and overall welfare. Repeated interactions can also create opportunities for collusion further complicating the design of these markets. This novel data-driven approach overcomes these challenges, enabling companies to learn how to bid effectively in this competitive environment.
“Our algorithm not only helps companies refine their bidding strategies but also offers valuable insights into the dynamics of the auction itself— how each bid is influenced by and influences others. Understanding these dynamics is crucial for assessing the potential impact of pay-as-bid auctions in carbon markets,” said Golrezaei.
“The principles behind the pay-as-bid research apply far beyond carbon markets. Multi-unit auctions drive billions of dollars across sectors ranging from electricity to telecommunications. By improving pricing accuracy and market efficiency, the principles from the study could lead to better outcomes in these sectors as well,” she continued.
Golrezaei noted that these improvements are also evident in the online advertising market, which has shifted from second price (single-unit version of uniform price auctions) auctions to first price (single-unit version of pay-as-bid auctions) auctions, resulting in increased revenues.
“Our work introduces data-driven algorithms that empower fair, efficient bidding across diverse industries, viewing multi-unit auctions as dynamic and adaptive systems,” said Golrezaei.
Attachment
Casey Bayer MIT Sloan School of Management 914.584.9095 bayerc@mit.edu Patricia Favreau MIT Sloan School of Management 617.595.8533 pfavreau@mit.edu Matthew Aliberti MIT Sloan School of Management 781.558.3436 malib@mit.edu
