Information disclosure through prices and common knowledge of rationality.

Authors
Publication date
1999
Publication type
Thesis
Summary This thesis studies the influence on the information revealed by prices of the ability of agents to coordinate their decisions. Coordination (or its failure) results from a divinatory learning process consisting in the iterated elimination of dominated strategies, i.e. in the exploitation by each agent of the common knowledge (everyone knows x, everyone knows that everyone knows jc, etc.) of the rationality of all and of the economy. In an economy of exchange at a given period, it is therefore a question of applying an assumption of individual rationality to its limit. Divinatory learning is based on strategic reasoning in a game of excess demand, and not on progressive reasoning or on the accumulation of observations. It succeeds when a single solution is compatible with the double hypothesis of common knowledge. This solution is then by construction a rational expectations equilibrium (ear) which is the traditional equilibrium in the competitive framework with asymmetric information. The success of the coordination thus defines a criterion of robustness of an ear that is added to the existing criticisms of the ear. In particular, learning solutions are similar to variants of the ear and its success presupposes the existence of a strategic basis for the ear, which we know does not always exist in the game of excess demand. After a review of the main results of the ear model, we show, in two simple models of exchange of a risky asset with a single ear, that learning succeeds if and only if all agents have private information of "better quality" than that revealed by the price of the ear. These conditions provide a theoretical justification for the influence of agents' confidence in their private information and the sensitivity of their demand to their expectations of aggregate demand on the information revealed by prices.
Topics of the publication
  • ...
  • No themes identified
Themes detected by scanR from retrieved publications. For more information, see https://scanr.enseignementsup-recherche.gouv.fr