Newton's Method for Optimal Control Involving Nonlinear Econometric Models

O. Patrick Kreidl

Newton's method for optimal control is a well-established nonlinear programming algorithm for addressing optimization problems involving a discrete-time dynamic system. We consider problems where the system is described by a nonlinear, simultaneous, econometric model and the objective is to minimize a quadratic cost function over a finite-time horizon. We outline the sequence of computations required within each iteration of the method, including the historic solution to the linear-quadratic tracking problem, and provide general comments regarding algorithm convergence.