A1247
Title: Structural estimation of nonlinear rational expectations models with recursive preferences
Authors: Bart Claassen - University of Groningen (Netherlands) [presenting]
Diego Ronchetti - Audencia Business School (France)
Abstract: The risk and the timing premia for the U.S. market from 1952 to 2019 are measured separately by consistently estimating the risk aversion and the endurance to cope with uncertainty over time of a representative agent with Epstein-Zin preferences in a Markovian setting. To minimize the risk of misspecification, we employ a novel Local Generalized Method of Moments built upon conditional moment restrictions for a nonparametric time series model, with moment functions identified by a functional equation in contraction mapping. The point estimate for the relative risk-aversion parameter is substantially lower and more plausible than previously empirically obtained through other empirical estimation techniques. We assess the finite sample performance of our estimator in a Monte Carlo study of the Bansal-Kiku-Yaron long-run risks model. We find that the method performs well in finite samples. Furthermore, we argue that this method can be applied to any functional form of recursive preferences that belongs to the Chew-Dekel class.
