Dynamically consistent updating of multiple prior beliefs – An algorithmic approach

This paper develops algorithms for dynamically consistent updating of ambiguous beliefs in the maxmin expected utility model of decision making under ambiguity. Dynamic consistency is the requirement that ex-ante contingent choices are respected by updated preferences. Such updating, in this context, implies dependence on the feasible set of payoff vectors available in the problem and/or on an ex-ante optimal act for the problem. Despite this complication, the algorithms are formulated concisely and are easy to implement, thus making dynamically consistent updating operational in the presence of ambiguity.

► We consider the maxmin expected utility model of decision making under ambiguity.
► We focus on dynamically consistent updating for all instances of this model and all non-null events.
► We develop algorithms to implement such dynamically consistent update rules.