The optimal consumption function in a Brownian model of accumulation. Part C: a dynamical system formulation

Foldes, L. (2017). The optimal consumption function in a Brownian model of accumulation. Part C: a dynamical system formulation. (Systemic Risk Centre Discussion Papers 68). Systemic Risk Centre, The London School of Economics and Political Science.
Copy

This Paper continues the study of the Optimal Consumption Function in a Brownian Model of Accumulation, see Part A [2001] and Part B [2014]; a further Part D, dealing with the effects of perturbations of the Brownian model, is in preparation. We begin here with a review of the o.d.e. system S which was used in Part B for the proof of the existence of an optimal consumption function. This system is non-linear, two dimensional but bilaterally asymptotically autonomous with limiting systems as log-capital tends to plus/minus infinity, each of which has a unique saddle point. An important part is played in the existence proof by the sets of forward/backward ‘special’ solutions, i.e. solutions of S converging to the asymptotic saddle points, and by their representing functions f and g. A ‘star’ solution, which is both a forward and a backward special solution, corresponds to an optimal consumption function. It is shown here that the sets of special solutions of S are C(1) sub-manifolds of R(3), hence that the functions f and g are continuously differentiable. The argument involves the construction of an imbedding of S in a 3-D autonomous dynamical system such that the asymptotic saddle points are mapped to saddle points of the 3-D system and the sets of forward/backward special solutions are mapped into stable/unstable manifolds. The usual Stable/Unstable Manifold Theorem for hyperbolic stationary points then yields the required C(1) properties locally (i.e. near saddle points), and these properties can be extended globally. A ‘star’ solution of S then corresponds to a saddle connection in the 3-D system. A stability result for the saddle connection is given for a special case.

picture_as_pdf

subject
Published Version

Download

Export as

EndNote BibTeX Reference Manager Refer Atom Dublin Core JSON Multiline CSV
Export