Asset Markets as an Equilibrium Selection Mechanism: Coordination Failure, Game Form Auctions, and Tacit Communication

Coordination in the Presence of Asset Markets

“…13 These findings suggest that, consistent with our predictions, the presence of an asset market in conjunction with a minimum-input coordination game significantly lowers output levels for all group sizes and periods. These results stand in contrast to the findings of Blume and Ortmann (2007), Cooper et al (1992) and Van Huyck et al (1993), which show that other forms of pre-play communication, including other kinds of pre-play markets, produce coordination on higher output levels. To better understand what drives these results we turn to study two asset market dimensions that can influence behavior: portfolio incentives and security prices.…”

“…We use ordered probit regressions, which take into account the ordinal nature of output, and estimate a model that includes group random effects. We find that output is lower for Large groups, across both treatments (columns 2 and 3), consistent with prior literature (e.g., Van Huyck et al (1990)). …”

“…11 As we 11 This is higher than is typically the case for large groups, which usually converge to the least efficient equilibrium. However, our experiment differs from most previous experiments in important ways: it uses fewer input choices (e.g., 4 vs. 7 in Van Huyck et al (1990)), provides subjects with the full history of choices by other players, and also shows players in each Large group the performance of a Small group (Bornstein, Gneezy and Nagel 2002). While these factors may influence behavior in our experiment, relative to other experiments using similar coordination games, they are constant across all our treatments.…”

Preference Anomalies, Preference Elicitation and the Discovered Preference Hypothesis

“…As in the second-price auction, the optimal bid in the English auction -that is, the price at which it is optimal for the bidder to withdraw from the auction -is the agent's true value. 12 Yet, in induced-value English auctions (Coppinger et al 1980;Kagel et al 1987;McCabe et al1990;van Huyck et al 1993) convergence of bids to true values is much faster than in second-price auctions. 13 It is unlikely that the English and second-price auctions would differ in their ability to foster value learning, even if there were anything to learn about the monetary value of a money voucher.…”

Self-Serving Cheap Talk: A Test Of Aumann's Conjecture

“…Van Huyck et al (1993) employ a two-stage design to test forward induction: the first stage consists of an auction for the right to play in a median-effort coordination game, where there are ranked equilibria and deviations from the median action are costly. There is a high degree of convergence toward the payoff-dominant equilibrium using the 1st-stage auction, while this convergence is never observed without an auction.…”