On the structural estimation of an optimal portfolio rule

We adopt a structural approach to estimate the parameters of the optimal asset allocation rule dictated by a standard dynamic portfolio choice problem. In doing so, we propose two novel approaches to estimates the preference/incentive parameters of the model.We model the asset selection problem of a portfolio manager who balances two investment motives; on the one hand, the desire to follow an asset allocation in line with the risk-return trade-off of the client, and on the other hand, the relative performance concerns that bear on the shoulders of the portfolio manager, due to the compensation and/or employment risk that he/she faces.We solve the model analytically, and estimate its parameters using NAV data from Chilean pension funds. Our results suggest that portfolio managers decisions from our sample are heavily motivated by the relative performance concerns they face. In particular, our point estimates suggest that manages are very reluctant to take almost any bet against the asset allocation of their peers.