Skip to main content
Top

2020 | Book

Multicriteria Portfolio Construction with Python

insite
SEARCH

About this book

This book covers topics in portfolio management and multicriteria decision analysis (MCDA), presenting a transparent and unified methodology for the portfolio construction process. The most important feature of the book includes the proposed methodological framework that integrates two individual subsystems, the portfolio selection subsystem and the portfolio optimization subsystem. An additional highlight of the book includes the detailed, step-by-step implementation of the proposed multicriteria algorithms in Python. The implementation is presented in detail; each step is elaborately described, from the input of the data to the extraction of the results. Algorithms are organized into small cells of code, accompanied by targeted remarks and comments, in order to help the reader to fully understand their mechanics. Readers are provided with a link to access the source code through GitHub.

This Work may also be considered as a reference which presents the state-of-art research on portfolio construction with multiple and complex investment objectives and constraints. The book consists of eight chapters. A brief introduction is provided in Chapter 1. The fundamental issues of modern portfolio theory are discussed in Chapter 2. In Chapter 3, the various multicriteria decision aid methods, either discrete or continuous, are concisely described. In Chapter 4, a comprehensive review of the published literature in the field of multicriteria portfolio management is considered. In Chapter 5, an integrated and original multicriteria portfolio construction methodology is developed. Chapter 6 presents the web-based information system, in which the suggested methodological framework has been implemented. In Chapter 7, the experimental application of the proposed methodology is discussed and in Chapter 8, the authors provide overall conclusions.

The readership of the book aims to be a diverse group, including fund managers, risk managers, investment advisors, bankers, private investors, analytics scientists, operations researchers scientists, and computer engineers, to name just several. Portions of the book may be used as instructional for either advanced undergraduate or post-graduate courses in investment analysis, portfolio engineering, decision science, computer science, or financial engineering.

Table of Contents

Frontmatter
Chapter 1. Introduction
Abstract
Nowadays, one of the major problems of the financial sector is the creation and management of an efficient investment portfolio under the complex environment of globalized society, rapidly increasing competition and sweeping economic changes at national and international level. In general, an investment portfolio is a basket of assets, the selection of which was based on specific economic objectives and constraints, aiming at the generation of profits for the investor.
Elissaios Sarmas, Panos Xidonas, Haris Doukas
Chapter 2. The Portfolio Management Problem
Abstract
This chapter focuses on the portfolio management problem, and more specifically, the mean–variance model introduced by Harry Markowitz. The problem of portfolio optimization was introduced as a quadratic mathematical programming problem. Since then, the development in this field was rapid, as many scientists have attempted to improve this methodology and cure its weaknesses. In discussing the various portfolio theory techniques, we will adhere to standard notation such as in Xidonas et al. The presentation of the mean–variance methodological framework is developed in three sections.
Elissaios Sarmas, Panos Xidonas, Haris Doukas
Chapter 3. Multicriteria Decision Analysis Methods
Abstract
The discussion of the portfolio management problem in the previous chapter signified the need for new methodological decision support frameworks, in order to overcome the existing problems and cure the inadequacies of the conventional mean–variance model.
Elissaios Sarmas, Panos Xidonas, Haris Doukas
Chapter 4. Literature Review
Abstract
In the basis of the Markowitz mean–variance methodology [112, 113], a series of new models have been developed, such as the single index models, the multi-index models, the average correlation models, the mixed models, and the utility models. Additionally, different criteria have been proposed such as geometric mean return, skewness, etc.
Elissaios Sarmas, Panos Xidonas, Haris Doukas
Chapter 5. The Proposed Methodology
Abstract
As already mentioned, the purpose of this book is the development of an integrated methodological framework for portfolio management. The process of portfolio management is a very complex problem, as it consists of two phases which demand a series of decisions. The first phase focuses on portfolio selection, i.e. the selection of the strongest investment opportunities. The second phase includes portfolio optimization, the determination of the most efficient allocation of the available capital to the selected securities in order to maximize return.
Elissaios Sarmas, Panos Xidonas, Haris Doukas
Chapter 6. Information System in Python
Abstract
The presentation of the proposed methodology emphasized the need for modern information systems in order to implement the necessary methods. The purpose of the information system is to implement efficiently the algorithms described in the previous section in order to support the decision-making process.
Elissaios Sarmas, Panos Xidonas, Haris Doukas
Chapter 7. Empirical Testing
Abstract
In this chapter we present the empirical testing of the proposed methodology on real data. This step is very important, because it is necessary that the methodology is explicitly tested to verify its validity. Consequently, a large experimental application of the proposed methodological framework was conducted, including securities from four stock exchanges (NYSE, NASDAQ, Paris, and Tokyo). Thus, every step of the methodological framework is examined, while the results are tested with an out-of-sample validation process.
Elissaios Sarmas, Panos Xidonas, Haris Doukas
Chapter 8. Concluding Remarks and Future Prospects
Abstract
In conclusion, we highlighted the inability of conventional methods to deal effectively with the portfolio management problem due to many reasons, such as the failure of combining the multiple conflicting criteria which must be considered during the portfolio construction process, the inadequacy to integrate the preferential profile of the decision-maker during the decision-making process, and the lack of flexibility to allow the user to intervene in the process and set his own policy restrictions.
Elissaios Sarmas, Panos Xidonas, Haris Doukas
Backmatter
Metadata
Title
Multicriteria Portfolio Construction with Python
Authors
Elissaios Sarmas
Panos Xidonas
Haris Doukas
Copyright Year
2020
Electronic ISBN
978-3-030-53743-2
Print ISBN
978-3-030-53742-5
DOI
https://doi.org/10.1007/978-3-030-53743-2