A Note on Recursive Two-Stage Evaluation Model

In this technical note, the two-stage-evaluation (TSE) model for decision making under ambiguity (He 2019) is extend to intertemporal setting in an axiomatic approach. First set of axiom employed are some axioms commonly adopted for axiomtizing intertemporal non-expected utility models in the literature. Besides these regular axioms, dynamic consistency as well as assumptions for a static TSE for one stage consumption plan at any given path of the history are assumed. It is shown that these axioms holds if and only if these exists a recursively defined evaluation utility model representing decision maker(DM)’s preferences over consumption plans conditional on arriving at any node in an event tree. Such a recursive form implies that one can apply dynamic programming technique (rolling back the decision tree) to solve for a dynamic decision making problem where the DM’s preference over consumption plans conditional events in the decision tree is represented by a TSE model allowing for ambiguity aversion depending on the sources of ambiguity. By assuming a preference regarding ambiguity aversion across different hedges at different nodes, one can also obtain a recursive TSE model with hedge independent ambiguity attitude.