Kai Liao

In decentralized online marketplaces, platforms often employ non-price mechanisms to reduce market congestion, enhance revenue, and improve efficiency. This study explores how the choice of trading mechanisms affects the dynamics of supply and demand. We develop a dynamic auction model incorporating endogenous entry, product heterogeneity, and bargaining to analyze the U.S. salvage car market. We characterize a static equilibrium where the demand and supply are shaped both by arriving agents and the existing agents. Using a novel dataset obtained through web scraping, we estimate the demand system of incoming buyers and investigate how their entry and bidding decisions shape the product space at equilibrium. In the counterfactual study, our findings suggest that transitioning from the current mixed selling method—consisting of Pure Sale (a no-reserve auction) and Reserve Sale (a reserve auction followed by contingent negotiation)—to exclusively Pure Sale reduces the demand-side market thickness by 53% and supply-side market thickness by 23%, resulting in a drop in the aggregate buyer-to-seller ratio from 3.39 to 2.24. Despite these reductions, the change would increase overall welfare by 5% due to less congestion and information friction.