An application and case history of a dynamic R&D portfolio selection model

A variety of formal mathematical models have been proposed which are based on a portfolio approach to a total R&D program. However, there have been few descriptions of applications appearing in the subsequent literature. This paper presents an application of stochastic integer-programming formulation to a portfolio of projects (each of which were planned with the aid of a decision-tree structure). Follow-up studies undertaken one year later are described in an attempt to assess the accuracy of the data and adequacy of the model in practice.