Abstract: 

This paper provides axiomatic foundations for a model in which the perception ofambiguity is optimally chosen at a cognitive cost. 
The model generalizes the Choquet expected utility model and can accommodate the preference patterns presented in
Machina [13], which are known to pose a difficult challenge for ambiguity models. Our model is characterized by a novel axiom that disciplines the preference for hedging. Specifically, it implies that whenever there is no benefit to gain from it, it should be avoided.

For further information please contact michela.raganato@unibocconi.it