Multistage Capital Budgeting Under Uncertainty

Abstract

The problem of allocating a fixed set of time-phased investment allowances among competing investment projects and proposals has received great attention in the literature since Weingartner [11] established a linear/integer programming formulation. Reviews of the literature can be found in Weingartner [12], Hunter [6], Hodges [5], and Bernhard [2]. Most investment situations are stochastic in nature. Also, few project selection decisions are once and for all, and frequently, a number of continuation decisions may be foreseen during the evolution of a project. In this paper a method of incorporating these features into a capital budgeting framework is described. The method is an extension of Weingartner's mathematical programming formulation and allows chains of events to be represented with stochastic decision trees. The method was first developed by the authors when they were considering problems of portfolio selection in a research and devel