Portfolio Selection Using Sharpe Model
and Genetic Algorithms

Academy of Economic Science, Bucharest

1. Abstract

This paper presents general information regarding Portfolio Theory and Genetic Algorithms that we used for a Portfolio Selection Application written in Java.  It is intended to be a online application, that can be used for evaluating investment opportunities in stock market. This application works on a real sample data, stored in a relational database; Now it’s a Microsoft Access database and we intend to port it to MySql later.

The application is developed using a Layered Architectural Patterns, that is data access, business object layer, and user interface layer have sharply separated functionality.

For the selection of portfolios we developed a highly customizable Genetic Algorithms library that was used for the search of the solution space.

Contents

1.Abstract

2.Portfolio Theory Overview

     2.1 Introduction

     2.2 Harry Markowitz Model

     2.3 William Sharpe Model

     2.4 Other Ideas

     2.5 The Impact of Technology on Portfolio Theory

3. Evolutionary Algorithms Overview

4. UML Diagrams

        4.1 Use case Diagram

        4.2 Package Diagram

        4.3 Class Diagrams

            4.3.1 Data Access Layer Class Diagram

            4.3.2 Business Object Layer Class Diagram

            4.3.3 User Interface Layer Class Diagram

            4.3.4 Genetic Algorithms Class Diagram

5. Sharpe Model Function

6. Relationa Database Diagram

    6.1 SQL queries for Sharpe’s Model Coefficients

7.Bibliography

SourceForge Logo