Ambiguity in Dynamic Environments – AmbiDyn
The work is of theoretical nature, using techniques from decision theory, game theory, contract theory and general equilibrium theory.
In the project «Endogenous feedbacks in self-confirming equilibria with uncertainty«, Marieke Pahlke (Bielefeld) and Frederic Koessler (PSE) study ambiguity aversion and the problem of designing ambiguous feedbacks in games. They consider general n-person games in which each player only receives coarse information about others' behavior. For each player, his information induces a set of possible beliefs about the feasible action distributions of the other players. An equilibrium is a refinement of the notion of self-confirming equilibrium in which every player maximizes his maxmin expected utility given the set of beliefs compatible with his feedbacks. The set of equilibria is characterized in different classes of games as a function of players’ feedbacks, and the optimal solution for a designer who is able to choose players’ feedbacks is analyzed. In games with binary actions, players’ feedbacks can be summarized by a directed network in which player j is informed about player i's behavior if and only if there exists a link from player i to player j. In public good and coordination games, Marieke Pahlke and Frederic Koessler characterize networks of information feedbacks that maximize total welfare. They also consider and compare optimal feedback design for other forms of ambiguity aversion.
In the project Efficiency and Welfare under Knightian Uncertainty, Chiaki Hara (Kyoto) Sujoy
Mukerji (Queen Mary) and Frank Riedel (Bielefeld) and Jean-Marc Tallon (PSE) study the optimal risk sharing arrangements in an economy with smooth ambiguity averters. They show how to construct an associated economy with a given prior and expected utility maximizers. The setup also invites to redefine the notion of contingent good and possibly condition the allocation on the “true” data generating process.
Work to be done on strategic communication under ambiguity
On stochastic independence under ambiguity, F. Ceron & V. Vergopoulos Economic Theory (2020)
Market Allocations under Ambiguity: A Survey, A. Billot, S. Mukerji & J.-M. Tallon, Revue Economique (2020)
and several working papers under submisson
Uncertainty plays a fundamental role in financial markets and strategic conflicts. In many real-world situations, it is not possible to model uncertainty by means of well-defined probability distributions. Instead, participants face Knightian uncertainty: they do not know the exact probability distribution over possible outcomes. In view of this, it is important to incorporate such Knightian uncertainty or ambiguity into the analysis of markets and strategic conflicts. This project aims to clarify the foundations for ambiguity in dynamic financial markets and strategic interactions. As a by-product, we aim to unify different (extant) approaches to modeling preferences and decision making in games under ambiguity. The issue of dynamic consistency will now be of particular concern. We apply our newly developed theoretical tools to a concrete economic problems by studying in detail the consequences of Knightian uncertainty in strategic interactions like dynamic contracts and elections on the one hand, as well as in markets under uncertainty, including portfolio choice and asset pricing with heterogeneous agents and incomplete information.
Project coordination
Jean-Marc TALLON (ECOLE D'ECONOMIE DE PARIS)
The author of this summary is the project coordinator, who is responsible for the content of this summary. The ANR declines any responsibility as for its contents.
Partner
ECOLE D'ECONOMIE DE PARIS
Help of the ANR 128,165 euros
Beginning and duration of the scientific project:
February 2019
- 36 Months
