Convergence of least squares learning in self-referential discontinuous stochastic models. (Q5956277): Difference between revisions

Dynamic stochastic model with discontinuous decision rules are studied. The author is looking for sufficient conditions under which adaptive agents learn an unknown parameter to pin down a rational expectation equilibrium. The following two main questions are on the discussion. The first one is whether the estimator of the value function is consistent and whether the individual learns to take the optimal section asymptotically. The second one is whether the market outcome induced the learning behaviour of the individual agents converges to the rational expectations equilibrium outcome. An approach is based on the results by \textit{A. Macet} and \textit{T. Sargent} [J. Econ. Theory 48, No. 2, 337--368 (1989; Zbl 0672.90023)] and by \textit{M. Woodford} [Econometrica 58, No. 2, 277--307 (1990; Zbl 0731.90021)].