סמינר באסטרטגיה

Speaker: Roee Teper, Economics Department, Pittsburgh University

Abstract:

In this paper, we put forth a behavioral foundation for an infinite-horizon subjective state space moadel in a dynamic programing framework.

Like in standard ``exploration and exploitation" models (bandit problems, search, rational inattention, etc.) we are allowing for endogenous resolution of uncertainty that depends on the actions the agents are taking.

In addition, there need not be full resolution of uncertainty between periods. 

When a decision maker consumes outcomes, she learns her relative preference between them; after each consumption history, the decision maker's state space is a refinement of the previous state space.

One interpretation for the model is a firm trying to learn about which projects she should invest in and learns through its experience.  

We identify the (set of possible) conditional preferences induced by each series of consumption