We provide a model of a decentralized, dynamic auction market platform (e.g., eBay) in which a large number of buyers and sellers participate in simultaneous, singleunit auctions each period. Our model accounts for the endogenous entry of agents and the impact of intertemporal optimization on bids. Solving our model with a finite number of bidders is computationally intractable due to the curse of dimensionality, so we prove that a continuum version of our model provides a good approximation of an equilibrium in the finite model. We use the approximation to estimate the structural primitives of our model using Kindle sales on eBay. We find that just over one third of Kindle auctions on eBay result in an inefficient allocation with deadweight loss amounting to 13.5% of total possible market surplus. We also find that partial centralization - for example, runnng half as many 2-unit, uniform price auctions each day - would eliminate a large fraction of the inefficiency, but yield lower seller revenues. Our results highlight the importance of understanding platform composition effects—selection of agents into the market—in assessing the implications of market design.