Information aggregation in competitive markets

Abstract

We study when equilibrium prices can aggregate information in an auction market with a large population of traders. Our main result identifies a property of information—the betweenness property—that is both necessary and sufficient for information aggregation. The characterization provides novel predictions about equilibrium prices in complex, multidimensional environments.

Full Information Equivalence in Large Elections

“…The second standard assumption is the MLRP, which is substantially more demanding than the linear refinement condition. In particular, we next introduce a substantial weakening of the MLRP, the betweenness condition analogous to the betweenness property identified in the auction environment in Siga and Mihm (2021) (henceforth SM), that is strictly more demanding than our linear refinement condition.…”

“…However, such result may well lead to false positives when the modeler assigns one state for each alternative (by coalescing multiple states where the alternative is best into a single aggregate state): with more states than alternatives, many environments do not allow FIE. Second, we connect our linear refinement condition to the MLRP, by establishing that the betweenness property identified in Siga and Mihm (2021) for an auction, is sufficient, but not necessary, for the linear refinement condition. As Siga and Mihm establish that the betweenness property is strictly weaker than the MLRP, our linear refinement generalizes the conditions in the literature.…”

“…The similarities are somehow surprising given obvious differences in the mechanisms, incentives, and outcomes. Closest to ours is Siga and Mihm (2021) (SM), who characterize information structures that permit fully informative equilibria in large common‐value auctions and allows us to contrast the consequences of the differences in the models. In the auction mechanism, individual actions (i.e., bids) are aggregated into a final outcome—the price—which is given by a specific quantile of the bids, and information is aggregated if the price reveals the value in every state.…”