HEURISTIC ALGORITHMS FOR CAPITAL-BUDGETING DECISIONS IN MULTIDIVISION FIRMS

This paper presents two versions of a heuristic algorithm to solve a model of the capital-budgeting problem in a decentralized multidivision firm involving no more than two exchanges of information between headquarters and divisions. Headquarters makes an allocation of funds to each division based upon its cash demand and its potential growth rate. Each division determines which projects to accept. Then, an additional iteration is performed to define the solution. More than one thousand examples were simulated resulting in an average relative error of less than one percent.