Towards Understanding Life Cycle Saving of Boundedly Rational Agents : A Model with Feasibility Goals

This paper develops a new life cycle model that aims to describe the savings and asset allocation choices of boundedly rational agents. In this model, agents make forward-looking decisions without the requirement of anticipating their actual future decisions. Instead, agents pursue two simple so-called feasibility goals. The first goal entails assuring feasibility of a certain minimum consumption level in a future worst-case scenario. The second concerns feasibility of a certain standard of living in a normal scenario. The feasibility goals framework represents a particularly natural alternative to the standard rational benchmark model. The framework is highly tractable and parsimonious. In particular, it is able to explain important empirical patterns of asset allocation that are puzzling from the point of view of existing models