Economists frequently confine themselves to the qualitative analysis of continuous optimization problems or they restrict their quantitative analysis to inaccurate methods like linearization around the steady state. The fact that the solution is characterized by an inherently unstable adjustment...

We present a non-linear solution method of saddlepoint dynamics in discrete time optimization problems. It is based on the backward attractivity of the stable manifold and is very easy to implement. After an introduction to the general method we present two applications. First we consider the...