Abstract

Journal of Actuarial Practice

Volume 12, 2005


An Application of Control Theory to the Individual Aggregate Cost Method

Alexandros A. Zimbidis and Steven Haberman

Abstract

The paper investigates the individual aggregate cost method (also known as the individual spread-gain method), which is normally applicable in small pension funds or fully contributory schemes, using a control theoretical framework. We construct the difference equations describing the mechanisms of the respective funding method and then calculate the optimal control path of the contribution rate assuming (first) a stochastic and (second) a deterministic pattern for the future investment rates of return. For the first case, the optimal solution is achieved through a linear approximation and using stochastic optimization techniques. It is proved that the contribution rate is (optimally) controlled through the control of the valuation rate (which is determined incorporating a certain feedback mechanism of the past contribution rate). The optimal solution for the deterministic case is obtained using standard calculus and the method of Lagrange multipliers.

Key words and phrases: individual aggregate funding, linear approximation, optimal stochastic control, Lagrange multipliers

Corresponding Author:

Alexandros A. Zimbidis

Athens University of Economics and Business

Department of Statistics

76 Patission St.,

Athens 10434

GREECE

E-mail: aaz@aueb.gr


© Copyright Absalom Press, Inc.