Abstract
Acommon probleminmanymaturemarketsishowtodealwithcongestion—asituationinwhichtransaction requestsfrommarketparticipantscannotbeaccommodatedinanexpeditedmanner.Thispaperexaminesthe congestion problem in sequential Dutch auction markets. Transactions in these markets typically involve perishable goods, making marketclearing speedcrucial. Traditionally, sequentialDutchauctions havebeen implemented with fast-paced auction clocks that process equally attractive bids in the order they arrive and only award the first bidder as the winner of each round, which can lead to serious congestion in case of a demandsurge.Weproposeahybridmechanismthatcapitalizesonthediscretenatureoftheauctionclockand batchestogetherthehighestbids, allowingmultipletransactionsatthesamepriceineachround.Toevaluate the performance of the hybrid mechanism, we first develop a game-theoretic model comparing the hybrid mechanism to the traditional sequential Dutch auction mechanism. Our model predicts that the hybrid mechanism will achieve higher operational efficiency without compromising allocative efficiency. We then complement the theoretical analysis by evaluating the hybrid mechanism through a quasi-natural field experiment. Theempiricalanalysisofthe fielddata showsthatthehybridmechanismcan significantlyspeed upthemarketclearingprocessandincreasepricestabilitywithoutaffectingtheexpectedrevenue.Ourfindings shednewlightonthedesignandoperationofmulti-unitauctions.
Original language | English (US) |
---|---|
Pages (from-to) | 457-490 |
Number of pages | 34 |
Journal | MIS Quarterly: Management Information Systems |
Volume | 46 |
Issue number | 1 |
DOIs | |
State | Published - Mar 2022 |
Externally published | Yes |
Bibliographical note
Publisher Copyright:© 2022 University of Minnesota. All rights reserved.
Keywords
- analytical model
- digital transformation
- discrete bids
- game theory
- hybrid mechanism
- market congestion
- Mechanism design
- quasi-natural field experiment
- sequential Dutch auctions