Approximating closed form solutions to a class of feedback policies

Abstract

Dynamic optimization problems cover a large class of problems in theoretical and applied economics. A simple iterative algorithm with fast convergence is proposed. It is demonstrated that the algorithm in a few steps produce excellent analytic (closed form) approximations including error bounds to a class of nonlinear problems.The algorithmic scheme is also well suited to produce numerical solutions. The notions of dynamic and potential rents are operationalized. The algorithm is utilizing a relation balancing these concepts. The result is particularly strong in the case of zero discounting where the exact CU-optimal policy is determined in a single step. Applying a particular seed in the general convergent scheme reproduces in a simple way results (formulas) published in the last decade in bioeconomics.