HEURISTICS FOR DEBT MANAGEMENT: AN EMPIRICAL STUDY

Six heuristics for allocating funds are evaluated using a well-established utility function and are compared to an optimal allocation rule. The first three heuristic rules prioritize on single system parameters: growth rate, initial state, and periodic demand, respectively. The latter three rules prioritize on multiplicative combinations of these parameters. System behavior under these rules is simulated for five discrete distributions of growth rates, four termination periods, and three initial reservoir volumes. The rules are evaluated on accuracy and effort. The results show that rules including growth rate in their calculations perform within 80%, others within 60% of the optimal solution with substantially less effort.