Portfolio selection problem using generalized differential evolution 3

A Adebiyi Ayodele,K Ayo Charles

doi:10.12988/ams.2015.5188

Abstract

This Portfolio selection Problem (PSP) remains an intractable research problem in finance and economics and often regarded as NP-hard problem in optimization and computational intelligence. This paper solved the extended Markowitz mean- variance portfolio selection model with an efficient Metaheuristics method of Generalized Differential Evolution 3 (GDE3). The extended Markowitz mean- variance portfolio selection model consists of four constraints: bounds on holdings, cardinality, minimum transaction lots, and expert opinion. There is no research in literature that had ever engaged the set of four constraints with GDE3 to solve PSP. This paper is the first to conduct the study in this direction. The first three sets of constraints have been presented in other researches in literatures. This paper introduced expert opinion constraint to existing portfolio selection models and solved with GDE3. The computational results obtained in this research study show improved performance when compared with other Metaheuristics methods of Genetic algorithm (GA), Simulated Annealing (SA), Tabu Search (TS) and Particle Swarm Optimization (PSO).

Highlights

The ability of financial practitioners, individuals and corporate investors to select appropriate assets to build a portfolio in order to minimize risk and maximize expected returns remained a difficult task to perform over the years
This paper contrast significantly from other researches in literatures being that it is the first ever to use the set of constraints of bounds on holdings, cardinality, minimum transaction lots, and expert opinion with efficient Metaheuristics method Generalized Differential Evolution 3 (GDE3) to find a solution to the extended Markowitz portfolio selection problem
The model developed in this paper is an improvement of the extended Markowitz portfolio model

Summary

Introduction

The ability of financial practitioners, individuals and corporate investors to select appropriate assets to build a portfolio in order to minimize risk and maximize expected returns remained a difficult task to perform over the years. Among the few works in literature that introduced one constraint or the other to Markowitz mean-variance portfolio selection model are as follows: In the research study of [6, 7] they used cardinality and bounding constraints with efficient metaheuristics method. The work of [3] used four sets of constraints, bounds on holding, cardinality, minimum transaction lots and sector capitalization in the extending Markowitz mean-variance model. This paper contrast significantly from other researches in literatures being that it is the first ever to use the set of constraints of bounds on holdings, cardinality, minimum transaction lots, and expert opinion with efficient Metaheuristics method GDE3 to find a solution to the extended Markowitz portfolio selection problem.

Portfolio Selection Problem

Proposed Model

Methodology

Computational Result and Discussion

Findings

Conclusion