Aggregation of not independent experts’ opinions under ambiguity

• The paper considers an aggregation rule of experts’ opinions expressed through different probability distributions and a decision-maker that adopts a multiple priors decision model.
• It is introduced an approach to form a consensus distribution that adopts the quantile function in a setting with multiple priors.
• It shows the condition for preserving stochastic dominance.

We consider an aggregation scheme of opinions expressed through different probability distributions or multiple priors decision model. The decision-maker adopts entropy maximization as a measure of risk diversification and a rational form of prudence for valuing uncertain outcomes. We show a new aggregation rule based on the composite value function that is able to represent asymmetric attitude on extreme events (optimism with respect to windfall gains and pessimism with respect to catastrophic events) and a rational prudence on ordinary events. We define when the new rule preserves stochastic dominance.