Heuristic Analysis of a near Optimal Approximation Algorithm for the Determination of Investment Options

The aim of this work is to continue with the ideas in Flores et al. [3] including the proposed greedy algorithm in order to develop them. When cash becomes available in a company, there are several strategies that allow us to benefit from it. The problem is how much to invest, for how long, and using which of the investment options in order to get the maximum profit out of it. A common problem in business administration is that we do not want to keep the money idle in the checking account, neither to over-invest. When the cash function becomes negative an analogous scheme is used as we want to pay as little interests as possible. In this paper we are reporting the implementation of the ideas in [3], and exploring several heuristics that can be used with the greedy algorithm and how well they behave. Finally we develop a hybrid algorithm that takes the best of the greedy algorithm and performs a very limited search. We find in this work that with the greedy algorithm we use, in general is not possible to optimize the profit for a given function; nevertheless the algorithm we use can find profits that are very close to the optimum and in some cases it gets the optimum. The proposed algorithm use a heuristic search based on the greedy scheme and greedy selection criteria to find profits close to the optimum. A software application was developed in order to show that the proposed strategy really works. Although this algorithm is suboptimal, it is very efficient in terms of time. ∗Facultad de Ingenierı́a Eléctrica. División de Estudios de Posgrado. †Facultad de Contabilidad y Administración. División de Estudios de Posgrado.