A Similarity-Based Approach to Prediction

I. GILBOA, O. Lieberman, D. Schmeidler

Journal of Econometrics

mai 2011, vol. 162, n°1, pp.124-131

Départements : Economie et Sciences de la décision, GREGHEC (CNRS)

Assume we are asked to predict a real-valued variable yt based on certain characteristics xt = (x1t , . . . , xdt), and on a database consisting of (x1i, . . . , xdi , yi) for i = 1, . . . , n. Analogical reasoning suggests to combine past observations of x and y with the current values of x to generate an assessment of y by similarity-weighted averaging. Specifically, the predicted value of y, yst , is the weighted average of all previously observed values yi, where the weight of yi, for every i = 1, . . . , n, is the similarity between the vector x1t, . . . , xdt, associated with yt , and the previously observed vector, x1i , . . . , xdi. The ''empirical similarity'' approach suggests estimation of the similarity function from past data. We discuss this approach as a statistical method of prediction, study its relationship to the statistical literature, and extend it to the estimation of probabilities and of density functions.Keywords:Density estimationEmpirical similarityKernelSpatial models

Aggregation of multiple prior opinions


Journal of Economic Theory

novembre 2011, vol. 146, n°6, pp.2563-2582

Départements : Economie et Sciences de la décision, GREGHEC (CNRS)

Mots clés : Aggregation of opinions, Ambiguity, Multiple priors

Experts are asked to provide their advice in a situation of uncertainty. They adopt the decision maker's utility function, but each has a potentially different set of prior probabilities, and so does the decision maker. The decision maker and the experts maximize the minimal expected utility with respect to their sets of priors. We show that a natural Pareto condition is equivalent to the existence of a set ¿ of probability vectors over the experts, interpreted as possible allocations of weights to the experts, such that (i) the decision maker's set of priors is precisely all the weighted-averages of priors, where an expert's prior is taken from her set and the weight vector is taken from ¿; (ii) the decision maker's valuation of an act is the minimal weighted valuation, over all weight vectors in ¿, of the experts' valuations

Aligning Ambition and Incentives


Journal of Law, Economics and Organization

octobre 2011, vol. 27, n°3, pp.655-688

Départements : Economie et Sciences de la décision, GREGHEC (CNRS)

Mots clés : Reputation, Asymmetric learning, Relative performance contracts, Transparency

Labor turnover creates longer term career concerns incentives that motivate employees in addition to the short-term monetary incentives provided by the current employer. We analyze how these incentives interact and derive implications for the design of incentive contracts and organizational choice. The main insights stem from a trade-off between “good monetary incentives” and “good reputational incentives.” We show that the principal optimally designs contracts to create ambiguity about agents’ abilities. This may make it optimal to contract on relative performance measures, even though the extant rationales for such schemes are absent. Linking the structure of contracts to organizational design, we show that it can be optimal for the principal to adopt an opaque organization where performance is not verifiable, despite the constraints that this imposes on contracts

Belief-free equilibria in games with incomplete information: characterization and existence

J. Hörner, S. LOVO, T. TOMALA

Journal of Economic Theory

septembre 2011, vol. 146, n°5, pp.1770-1795

Départements : Finance, GREGHEC (CNRS), Economie et Sciences de la décision

Mots clés : Repeated game with incomplete information, Harsanyi doctrine, Belief-free equilibria

We generalize the results of Hörner and Lovo (2009) [15] to N-player games with arbitrary information structure. First, we characterize the set of belief-free equilibrium payoffs under low discounting as the set of feasible payoffs that are individually rational, jointly rational, and incentive compatible. Second, we provide necessary and sufficient conditions on the information structure for this set to be non-empty

Evidence for an endogenous rebound effect impacting long-run car use elasticity to fuel price


Economics Bulletin

octobre 2011, vol. 31, n°4, pp.2777-2786

Départements : Economie et Sciences de la décision, GREGHEC (CNRS)

This paper presents a structural equation model of household fleet fuel efficiency and car use. It allows to weigh the contribution of car equipment changes and car use adjustments to the price elasticity of household demand for fuel. This model is implemented using a panel dataset of 322 households that were present in each annual wave of the French Car Fleet survey from 1999 to 2007. The longitudinal dimension of this dataset enables to assess the short and long-run adjustments at the household level over a period of fuel price increase. The estimated price elasticities of the demand for fuel are fully consistent with the literature: -0.30 in the short run and -0.76 in the long run. Regarding car use elasticities, accounting for an endogenous rebound effect allowed a striking finding: the sensitivity of household car use to fuel price changes is lower on the long run than on the short run. This paper thus not only provides the latest estimations of elasticities for France, in the early 2000's, it also shows that, on the long run, French households have managed to mitigate the impact of increasing fuel prices on their car mobility by using more fuel efficient cars

Experienced vs. Described Uncertainty: Do We Need Two Prospect Theory Specifications?


Management Science

octobre 2011, vol. 57, n°10, pp.1879-1895

Départements : Economie et Sciences de la décision, GREGHEC (CNRS)

Mots clés : experience-based decisions; description-based decisions; rare events; risk; uncertainty; prospecttheory; utility; loss aversion; decision weights; probability weighting; source of uncertainty; ambiguityHistory:

This paper reports on the results of an experimental elicitation at the individual level of all prospect theory components, i.e. utility, loss aversion, and weighting functions, in two decision contexts: situations where alternatives are described as probability distributions and situations where the decision maker must experience unknown probability distributions through sampling before choice. For description-based decisions, our results are fully consistent with prospect theory's empirical findings under risk. Furthermore, no significant differences are detected across contexts as regards utility and loss aversion. Whereas decision weights exhibit similar qualitative properties across contexts typically found under prospect theory, our data suggest that, for gains at least, the subjective treatment of uncertainty in experience-based and description-based decisions is significantly different. More specifically, we observe a less pronounced overweighting of small probabilities and a more pronounced underweighting of moderate and high probabilities for experience-based decisions. On the contrary, for losses, no significant differences were observed in the evaluation of prospects across contexts.

Fault Reporting in Partially Known Networks and Folk Theorems


Operations Research

mai-juin 2011, vol. 59, n°3, pp.754-763

Départements : Economie et Sciences de la décision, GREGHEC (CNRS)

Mots clés : Networks/graphs, Communication networks, Communications, Communication protocols, Games/group decisions, Noncooperative, Repeated games

We consider a group of players who perform tasks repeatedly. The players are nodes of a communication network and observe their neighbors' actions. Players have partial knowledge of the network and only know their set of neighbors. We study the existence of protocols for fault reporting: whenever a player chooses a faulty action, the communication protocol starts and the output publicly reveals the identity of the faulty player. We consider two setups. In the first one, players do not share authentication keys. We show that existence of a protocol for fault reporting is equivalent to the 2-vertex-connectedness of the network: no single vertex deletion disconnects the graph. In the second setup, we allow players to share authentication keys. We show that existence of a distribution of the keys and of a protocol for fault reporting is equivalent to the 2-edge-connectedness of the network: no single edge deletion disconnects the graph. We give applications to the implementation of socially optimal outcomes in repeated games.

Finitely repeated games with semi-standard monitoring

P. Contou-Carrère, T. TOMALA

Journal of Mathematical Economics

janvier 2011, vol. 47, n°1, pp.14-21

Départements : Economie et Sciences de la décision, GREGHEC (CNRS)

Mots clés : Finitely repeated games, Semi-standard monitoring, Folk Theorem

This paper studies finitely repeated games with semi-standard monitoring played in pure strategies. In these games, each player’s action set is endowed with a partition, and the equivalence classes of the actions played are publicly observed. We characterize the limit set of equilibrium payoffs as the duration of the game increases

General properties of long-run supergames

T. TOMALA, J. Renault

Dynamic Games and Applications

2011, vol. 1, n°2, pp.319-350

Départements : Economie et Sciences de la décision, GREGHEC (CNRS)

Mots clés : Repeated games, Signals, Folk theorem

Supergames are repeated games in which a fixed known finite one-shot game is repeated over and over. Information about the actions chosen at each stage is provided by a signalling technology. This paper studies the main properties that are valid over this whole class of games and both surveys known results and provides new ones.

Intermediaries, Credibility and Incentives to Collude

E.-A. PEYRACHE, L. Quesada

Journal of Economics and Management Strategy

décembre 2011, vol. 20, n°4, pp.1099-1133

Départements : Economie et Sciences de la décision, GREGHEC (CNRS)

Mots clés : Intermediaries, Collusion and soft information, Credibility and market structure

A seller contracts and potentially colludes with a certification intermediary. We investigate the intermediary’s incentives to collude, her pricing strategy, and the extent to which buyers rely on the intermediary’s announcements. The probability of collusion is an endogenous variable, determined by the intermediary’s pricing strategy. The extent to which the market relies on the intermediary’s reports, the certification price and the intermediary’s profit decrease as the intermediary becomes less patient. By making certification mandatory, the intermediary loses her ability to screen out low-quality sellers, which increases the probability of collusion