Investment portfolio formation
via multicriteria decision aid: a
Brazilian stock market study
Marcio Pereira Basilio, Jéssica Galdino de Freitas,
Milton George Fonseca Kämpffe and Ricardo Bordeaux Rego
Department of Production Engineering,
Federal Fluminense University, Niter oi, Brazil
Abstract
Purpose – The purpose of this paper is to identify how multicriteria decision aid(MCDA) can assist the
investment portfolios formation, increasing the reliability of decision-making.
Design/methodology/approach – To develop this paper, a simulation-based approach is used.
Information about the assets traded on the spot market of the São Paulo Stock Exchange -
BM&FBOVESPA was selected. They had 100 per cent participation in the 246 trading sessions carried out
in 2015 and had an average number of business/day greater or equal to 1,000. The stratification resulted in
the selection of 111 assets. Aiming assets evaluation, data are collected from 21 financial indicators.
Subsequently, the principal component analysis (PCA) is used to reduce the mass of collected data without
the loss of essential information. PROMETHEE II method is used for assets ranking; it belongs to the
group of methods for MCDA. At the end of these stages, four groups of investment portfolios are created
for simulation.
Findings – After the construction of portfolios, a simulation was performed with real data of the assets from
January 03, 2011 to November 14, 2016. It resulted in a comparison in which it was observed that 100 per cent
of portfolios showed positive returns on the investment. The result of portfolios’ group composed of assets
based on the 21 financial indicators was higher than the other one formed from PCA criteria. Both of them
were higher than Ibovespa result in the same period.
Research limitations/implications – As a contribution to new research, the model presents an
opportunity for improvement through linear programming methodologies with the objective of optimizing the
results, as the results obtained with the model were not optimized.
Practical implications – This research presents an alternative logic to the traditional one, as it seeks
the reduction of investment risk based on the results of the management of the companies, reflected
through their indicators. The model implies a change in how companies, financial institutions and small
and medium investors choose their assets to form investment portfolios. The authors believe that the
model has the potential to attract investors looking for long-term gains, such as public servants,
retirees, professionals and others who seek to build heritage to overcome the adversities of the uncertain
future. The model offers these investors the opportunity to choose which companies to invest in, based
on established indicators in the literature, whose information is available in the market. The model
systematizes the information and builds a ranking of the best companies so that the investor can make a
conscious decision, thus avoiding what experts call a “herd effect”, which makes the majority of
investors decide according to the oscillation of the market, thus ignoring the financial fundamentals of
companies.
Originality/value – This study presents a proprietary methodology by merging the PCA tool with MCDA
to build efficient investment portfolios.
Keywords Brazil, PROMETHEE, Stock market, Multicriteria decision aid,
Principal components analysis, Stock portfolio
Paper type Technical paper
JM2
13,2
394
Received 10 February 2017
Revised 30 May 2017
Accepted 17 July 2017
Journal of Modelling in
Management
Vol. 13 No. 2, 2018
pp. 394-417
© Emerald Publishing Limited
1746-5664
DOI 10.1108/JM2-02-2017-0021
The current issue and full text archive of this journal is available on Emerald Insight at:
www.emeraldinsight.com/1746-5664.htm