Market Selection and Learning Under Model Misspecification

This paper studies market selection in an Arrow-Debreu economy with complete markets where agents learn over misspecified models. Under model misspecification, standard Bayesian learning loses its formal justification and biased learning processes may provide a selection advantage. Given the natural connection between selection outcomes and long-run asset prices, understanding which biased learning processes are evolutionary fit is instrumental to build a parsimonious long-run asset valuation model robust to misspecification. Leveraging two cases of model misspecification and four learning processes, our analysis reveals a general difficulty in ranking learning behaviors with respect to their survival prospects. For example, the advantage of predictions averaging disappears when the true data generating process does not belong to the same family of models agents use to learn. Rules that generically guarantee survival, appear to require an unreasonable amount of knowledge about all the agents that compose the market ecology. The goal of a parsimonious long-run asset valuation model robust to model misspecification remains out of reach