List of Abstracts

Tuesday, the 14th of December 2021

9:30AM-10:50AM: Allocations & Markets under Uncertainty I

Frank Riedel (Toulouse School of Economics): «Knight and Walras meet Chateauneuf»

In recent work, we developed a general equilibrium theory of markets under sublinear expectations as they arise naturally under Knightian uncertainty, but also under other market settings with frictions, e.g. transaction costs. In the talk, we will explore the corresponding theory of general equilibrium in the case when price frictions are modeled by concave Choquet capacities.

John Quah (John Hopkins University): «A Theory of Revealed Indirect Preference» (co-authored with G. Hu, J. Li & R. Tang).

We call a preference over menus an indirect preference if there exists a preference over the objects that make up the menus, and a menu is ranked over another whenever it contains an element that is preferred to every element in the other menu. Suppose an observer has information on an agent's ranking over some menus; we characterize those rankings that guarantee the existence of a preference over objects that induces the observed menu rankings. Our result has the following applications. (1) It gives a characterization of rankings over prices that could be extended to a bona fide indirect utility function. (2) It leads to a generalization of Afriat's Theorem that allows for imperfectly observed choices. (3) It leads to a test of the multiple preferences model. (4) It helps us characterize a model of choice generated by minimax regret.

11:10AM-12:30PM: Allocations and Markets under Uncertainty II

Sujoy Mukerji (Queen Mary University): «Efficient Allocations under Ambiguous Model Uncertainty» (co-authored with C. Hara, F. Riedel & J.M. Tallon).

«Efficient Allocations under Ambiguous Model Uncertainty» (co-authored with C. Hara, F. Riedel & J.M. Tallon).

We investigate consequences of ambiguity on ex ante e¢ cient alloca- tions in an exchange economy. The ambiguity we consider is embodied in the model uncertainty perceived by the decision maker: they are unsure what would be the appropriate probability measure to apply to evaluate consumption contingent on a state space and keep in consideration a set P of alternative probabilistic laws p. We study the case where the typ- ical consumer in the economy is ambiguity averse with smooth ambiguity preferences (Klibano§ et al. 2005) and P is point identiÖed, i.e., the true law p 2 P can be recovered empirically from events in , a framework axiomatized in Denti and Pomatto (2021). Our particular focus is on the systematic di§erence that ambiguity aversion brings to the nature of e¢ - cient allocation, compared to e¢ cient allocations in an economy where all agents are ambiguity neutral, expected utility maximizers. The literature studying properties of e¢ cient allocations under ambiguity sensitive pref- erences has focussed, almost exclusively, on the case where aggregate risk is absent, or if present, the aggregate risk is unambiguous. These studies provide conditions under which the e¢ cient allocation retains the hall- mark properties of e¢ cient allocations under expected utility with com- mon belief. We allow for the case where the aggregate risk is ambiguous, and identify systematic di§erences, in particular, in the well studied set ups where under expected utility the Pareto e¢ cient consumption shar- ing rule is an a¢ ne function of the aggregate endowment. Understanding what is di§erent will tell us what may be missing in normative prescrip- tions for sharing uncertainty e¢ ciently when we think purely in terms of expected utility agents. The understanding may also shed light on market phenomena which may have arisen because of agentsídesire to hedge and share ambiguity exposure.

Jan Werner (University of Minnesota): «Optimal Allocations with Alpha-MaxMin Utilities, Choquet Expected Utilities, and Prospect Theory» (co-authored with P. Beissner).

The analysis of optimal risk sharing has been thus far largely re- stricted to non-expected utility models with concave utility functions, where con- cavity is an expression of ambiguity aversion and/or risk aversion. This paper extends the analysis to α-maxmin expected utility, Choquet expected utility, and Cumulative Prospect Theory, which accommodate ambiguity seeking and risk seeking. We introduce a novel methodology of the quasidifferential calculus of Demyanov and Rubinov (1986, 1992) and argue that it is particularly well- suited for the analysis of these three classes of utility functions which are neither concave nor differentiable. We provide characterizations of quasidifferentials of these utility functions, derive first-order conditions for Pareto optimal allocations under uncertainty, and analyze implications of these conditions for risk sharing with and without aggregate risk.

2:00PM-4:00PM: Time, Uncertainties and Aggregation

Marcus Pivato (University of Cergy): «Bayesian Social Aggregation with Almost-Objective Uncertainty» (co-authored with E.F. Tchouante).

We consider collective decisions under uncertainty, in which different agents may have not only different beliefs, but also different ambiguity attitudes —in particular, they may or may not be subjective expected utility maximizers. We assume that the space of possible states of nature is a Polish space. We consider sequences of acts which are “almost-objectively uncertain” in the sense that asymptotically, all agents almost-agree about the probabilities of the underlying events. We impose a weak ex ante Pareto axiom which applies only to asymptotic preferences along such almost-objective sequences. We show that this axiom implies that the social welfare function is utilitarian (i.e. a weighted sum of individual utility functions). But it does not impose any relationship between individual and collective beliefs, or between individual and collective ambiguity attitudes.

Lukasz Balbus (University of Zielona Gora): «On Collective Intertemporal choice, Altruism and Time-consistent Decision Rules » (co-authored with J.-P. Drugeon & L. Wozny).

We consider a dynamic decision problem of a collective-household with heterogeneous discount factors. We prove existence a time-consistent decision rule in Markovian policies. In fact, decision rules we consider are characterized by monotone and Lipschitz continuous investments policies. We provide sufficient conditions for the validity of the (generalized) first-order approach. Finally, we propose a novel method to approximate the constructed decision rule by a sequence of equilibria of bequest games with paternalistic altruism.

Antoine Billot (LEMMA - University of Paris II): «Deliberative Democracy and Utilitarianism» (co-authored with Xiangyu Qu).

This paper explores the possibility, in case of belief and taste heterogeneity, to aggregate individual preferences through a deliberation process enabling society to reach a consensus. However, we show that the same deliberation process, even characterized by a convergent matrix, may lead to different consensus depending on the updating rule which is chosen by individuals, i.e., deliberation is sufficient to determine social preferences but not univocally. Then, we prove that the Pareto condition allows to choose from possible consensus the one whereby social deliberated beliefs and tastes are of a utilitarian shape.

4:30PM-6:30PM: Finance, Markets & Time

Jean-Marc Bonnisseau (PSE - University of Paris 1 - CES): «On Future Allocations of Scarce Ressources without Explicit Rates of Discount» (co-authored with A. Chateauneuf & J.-P. Drugeon).

n this paper, we show that by merely fixing upper bounds and lower bounds for the stream of consumptions, we can compute the exact planning of these consumptions without explicit time discounting. The same principle applied to future reimbursement of a debt allows to determine the optimal planning, without an explicit time discounting as soon as we assume it to be strictly increasing. Therefore, one gets an allocation satisfying some kind of intergenerational fairness since the highest effort is supported by the first generations and then it decreases for the remaining ones.

Lasse Mononen (University of Zurich): «Recursive Dual-Self Expected Utility and Time Dependent Preferences».

Experimental evidence in intertemporal choice has documented a preference for consumption smoothing that cannot be captured by exponential discounting. We study a general class of recursive dual-self preferences that have a preference for consumption smoothing. We show that these preferences have a simple structure and are characterized by uncertainty about the discounting factor for the future utility with either preference or aversion for the uncertainty at each time period. Especially, under a stationarity axiom, these preferences are uncertainty loving or uncertainty averse. Second, we study time-dependent preferences for consumption smoothing and their identification with countably many time periods. We show that time-dependent utilities are identified if the uncertainty about the discount factors is full-dimensional as under finitely many time periods. However, the behavioral characterization for full-dimensional uncertainty does not extend from finitely many time periods due to a standard technical assumption ruling out infinite utility streams.

Alain Chateauneuf (PSE - University of Paris 1 - CES - IPAG): «Put-Call Parities, Absence of Arbitrage Opportunities & Choquet Pricing Rules» (co-authored with L. Bastianello & B. Cornet).

If prices of assets traded in a financial market are determined by non-linear pricing rules, different versions of the Call-Put Parity exist. We show that, under monotonicity, parities between call and put options and discount certificates characterize ambiguity-sensitive (Choquet and/or Sipos) pricing rules, i.e. pricing rules that can be represented via discounted expectations with respect to non-additive probability measures. We analyze how non-additivity relates to arbitrage opportunities and we give necessary and sufficient conditions for Choquet and Sipos pricing rules to be arbitrage-free. Finally, we identify violations of the Call-Put Parity with the presence of bid-ask spreads.

WEDNESDAY, the 15th of December 2021

9:30AM-10:50AM: Discounting & Myopia I

Lorenzo Bastianello (LEMMA - University of Paris II): «Future Blindness» (co-authored with M. Menez de Andrade and J. Orrillo).

An agent is future-blind if she neglects all possible gains starting from a fixed date N. We study which topologies over the set l^\infty of bounded real-valued sequences are consistent with such behavior. When N tends to infinity, we construct a sequence of non-Hausdorff topologies that converges to a new topology called the eventually-blind topology. We show that this topology is Hausdorff and that its dual consists of all sequences that are different from 0 for finitely many periods.

Lukasz Wozny (Warsaw School of Economics): «Intergenerational Altruism and Time Consistency» (co-authored with L. Balbus and K. Reffett).

«Intergenerational Altruism and Time Consistency» (co-authored with L. Balbus and K. Reffett).

We consider a class of infinite horizon, stochastic, non-stationary dynastic consumption-savings models with a general forms of recursive, time- varying altruistic preferences including direct and indirect pure altruism as well as paternalistic altruism. It is well-known such models lead to time-inconsistent dynastic preferences. Within this class of economies, we propose a novel set-iterative procedure for characterizing all Markov perfect time-consistent solutions in the space of increasing investments. Our approach involves both: value functions and policy iterations. We prove existence of Markov Perfect equilibria in stationary, periodic and also non-stationary strategies. We provide numerous applications to altruistic growth models, behavioral discounting models and collective household models and also discuss the role of various certainty equivalence operators.

11:10AM-12:30PM: Discounting & Myopia II

Philippe Bich (PSE - University of Paris 1 - CES): «On Multiple Discount Rates and Present Bias» (co-authored with X. B. Dong & B. Wigniolle).

In this paper, we axiomatize the following maxmin version of the quasi-hyperbolic model I(x)= min {p0x0 +(1−p0)(1−δ)x1 +(1−p0)(1−δ)δx2 +(1−p0)(1−δ)δ2x3...} over (p0 ,δ)∈Q, where (x0, . . . , xn, . . . ) is a deterministic consumption stream, and where Q is a set of possible pairs of discount rates δ and present bias parameters p0. When there is no present bias, we recover Chambers and Echenique’s axiomatization of maxmin exponential preferences, and when Q reduces to a singleton, we get Montiel Olea and Strzalecki’s axiomatization of quasi-hyperbolic preferences. To prove our main result, we provide some intertemporal variational representation results of interest for its own sake.

Federico Echenique (California Institute of Technology): «Decreasing Impatience» (co-authored with C. Chambers & A. Miller).

We characterize decreasing impatience, a common behavioral phenomenon in intertemporal choice, and a property with certain normative support in the literature on project evaluation. Discount factors that display decreasing impatience are characterized through a convexity axiom for invest- ments at fixed interest rates. Then we show that they are equivalent to a geometric average of generalized quasi-hyperbolic discount rates. Finally, they emerge through parimutuel preference aggregation of exponential discount factors.

2:00PM-4:00PM: Miscellanates

Yves le Yaouanq (CREST-IP Paris/Ludwig-Maximilians-Universität): «Aggregation and Dynamics of Ambiguous Opinions» (co-authored with M. Frick & R. Iijima).

We extent the DeGroot model to analyze social learning environments where individuals' opinions are ambiguous and take the form of a set of priors (or real numbers) as opposed to a precise probabilistic belief. We axiomatize a linear extension of the DeGroot model where each individual uses a fixed contraction procedure to interpret each neighbor's opinion, and a fixed vector of weights to aggregate her neighbor's opinions. We then relax linearity and consider aggregation procedures under which individuals react cautiously to uncertainty about the interpretation of others' opinions and about the aggregation weights. We study conditions under which opinions reach an unambiguous consensus in the long run, and show that ambiguity in the initial opinion might affect the influence of an agent on the long-run opinions.

Xiangyu Qu (CNRS - CES - LEMMA): «Falsifiability and Non-Falsifiability with α-maxmin» (co-authored with A. Chateauneuf, C. Ventura & V. Vergopoulos).

The class of α-maxmin representations of an agent’s preferences are meant to achieve a separation between the ambiguity he perceives and his attitude toward this ambiguity. Yet the same preferences may admit a multiplicity of α-maxmin repre- sentations. Any representation is thus falsified by the simple possibility of the other representations. The objective of this paper is to characterize the falsifiability and non falsifiability of a given α-maxmin representation within the class of all α-maxmin rep- resentations. This allows to identify the α-maxmin representations that admit a robust interpretation in terms of ambiguity and ambiguity attitudes. Furthermore, it is shown that in many economic applications where either objective beliefs or subjective beliefs is a main concern of a modeller, a MEU model, when falsifiable, is the best model to choose.

Itzhak Gilboa (Tel Aviv University - HEC): «Consumption of Values» (co-authored with S. Minardi & F. Wang).

Consumption decisions are partly influenced by values and ideolo- gies. Consumers care about global warming, child labor, fair trade, etc. Incorporating values into the consumer’s utility function will often vio- late monotonicity, if consumption hurts values in a way that isn’t offset by hedonic benefits. We distinguish between intrinsic and instrumental values, and argue that the former tend to introduce discontinuities near zero. For example, a vegetarian’s preferences would be discontinuous near zero amount of animal meat. We axiomatize a utility representa- tion that captures such preferences and discuss the measurability of the degree to which consumers care about such values.

4:30PM-6:30PM: Choice & Rationality

Marciano Siniscalchi (Northwestern University): «Robust Level-k Reasoning»

I propose a novel version of level-k thinking (Stahl and Wilson, 1994) that reflects a concern for robustness in the face of strategic uncertainty. Level-0 reasoners are allowed to play any strategy. Higher-level reasoners are maxmin decision-makers who optimize against players one step below them in the reasoning hierarchy. I demonstrate the tractability of the proposed approach by characterizing robust level-2 bidding in generalized second-price auctions (Edelman, Ostrowsky, Schwartz, 2007), a mechanism broadly used in online ad auctions.

Jose Heleno Faro (INSPER): «Salience-Biased Nested Logit» (co-authored with A.D. Caluz, J.H. Faro & F.M. Sanches).

We propose a nested logit model with salience bias. In this model, a stochastic choice function features a partition of the set of alternatives (the collection of nests or categories) and Luce values assigned to all alternatives. Once a menu of alternatives is offered, the decision maker's attention is drawn to salient alternatives within each category. First, nests are chosen according to a Luce formula combining only values determined by the best alternatives within the categories. Second, alternatives in the same nest are chosen according to the standard Luce rule. We provide an axiomatic foundation for this class of random choice functions and compare it with the standard nested logit model. Even though this class of random choice functions does not obey regularity, we can derive our model in a framework with random utility under similar conditions used in the usual nested logit model. Additionally, using a parametric version of the utility, we propose estimation procedures to estimate the parameters of the model using aggregate and individual-level data, allowing researchers to use the model in the same contexts as the usual nested logit applications.

Fabio Maccheroni (Bocconi University): «Sequential Neural Decision Processes in Multialternative Choice Problems» (co-authored with C. Baldassi, M. Marinacci & M. Pirazzini).

We introduce an algorithmic decision process for multialternative choice that combines binary comparisons and Markovian exploration. We show that a preferential property, transitivity, makes it testable.