Abstract

This dissertation evaluates the impact of endogenously generated public and private information on price formation in a bargaining-in-markets framework. Bargaining is assumed to be repeated and take place under persistent two sided uncertainty. Three models are developed assuming adaptive boundedly rational agents who take into account relative differences between their own outcomes and those of the market when forming expectations. These relative differences, labeled market distance, are used to determine the weight allocated to private and public information when forming expectations. These models of expectations are then embedded into a repeated bargaining game using a modified version of the asymmetric bargaining solution to characterize equilibrium. It is shown theoretically that bargaining power emerges from an agent’s position in the information structure, relative to his bargaining opponent. Using agent based simulations it is shown that each model generates average prices that drift (lower or higher) from the initial to ending trading period. The particular pattern and magnitude of drift observed depends on the model being evaluated and the pattern of bargaining matches. The predictions from the theoretical model and computational results are then tested using experimental data collected and first reported in Phillips and Menkhaus (2007). Results indicate that the model containing price expectations formed in a self-serving biased manner most closely fits the data in terms of individual outcomes and aggregate change in average prices across experimental time periods.