A Probabilistic Objective Function for R & D Portfolio Selection

Abstract

This paper contributes to the literature on the problem of allocation of resources to a set of risky investments. Our objective is to develop the ideas in the context of a research and development laboratory. A mathematical programming approach to the resource allocation problem is taken, and various forms for an objective function under risk are discussed. A probabilistic objective function appropriate to R & D is isolated and tested on a small hypothetical example. Parametric linear programming is used to yield a near-optimum allocation.