I study the interaction of resale, aggregate shocks, and decentralized trade when buyers and sellers are forward looking. I develop a fully dynamic structural model of a sequential auctions market with resale and aggregate shocks. I then estimate the model on data from classic car auctions. The model matches the strong positive correlation between prices and the state of the economy that is also present in the data. Only half of the variation in prices over the business cycle stems from the direct effect of the aggregate wealth shock. The remaining half is due to the dynamic decisions of buyers and sellers. A series of counterfactual experiments shows that allowing for instantaneous resale substantially increases average car prices and their volatility over the business cycle compared to allowing no resale at all, a result that is driven primarily by the possibility to speculate. A second set of counterfactual simulations reveals that centralizing trade lowers prices and increases the volume of trade. Price volatility remains unchanged in this scenario, even with frequent resale opportunities because there is less scope for speculation in a centralized market.
Download the paper