An ambiguity averse decision maker evaluates plans under a restricted family of what we call structured models and unstructured alternatives that are statistically close to them. The structured models can include parametric models in which parameter values vary over time in ways that the decision maker cannot describe probabilistically. Because he suspects that all parametric models are misspeciﬁed, the decision maker also evaluates plans under alternative probability distributions with much less structure.
October 2018 | Working Paper
Tags: Risk, Robustness and Ambiguity, Uncertainty|