Um framework para a pré-seleção de ações e montagem de carteira de investimentos utilizando o método multicritério fitradeoff e modelo de markowitz

Abstract

The constant dissemination of information about profitable investments, driven by marketing, has brought the spotlight to the stock market in recent years, showing significant increases in the numbers of individuals registered on the Brazilian stock exchange and historical maxims in the main index of the Brazilian stock exchange (Ibovespa) , the latter being greatly affected by the current crisis caused by COVID-19. In order to support the decision-making of investors in the financial market, this research aims to propose a framework for the pre-selection of shares in a rational and efficient way to set up an investment portfolio. To represent the merit of the research proposals, the framework was applied in a case study, where the actions of the Bovespa index were used as a numerical example. The shares incorporated as alternatives for the decision-making process were divided according to their respective specific or similar operating sectors. Pre-selection was based on the use of quantitative (P / L, P / VP, DY, ROE, Net Debt / Equity, Liquidity, Volatility) and qualitative (Corporate Governance) criteria and supported by the multicriteria decision analysis method Fitradeoff. The portfolio assembly was based on the main results of the pre-selection and on the Markowitz optimization model. The research is classified as applied, quantitative and normative. The results show that the best ratings in the pre-selection come from the best trade-offs among the criteria in line with the preferences of each investor. When building the portfolio, capital allocations vary according to the investor's objectives, just as diversification depends only on the connection between the assets and not on the quantity. Thus, the research presents a structured sequence of steps in order to promote the reproducibility of the proposed framework and support the decision of each investor, considering their preferences and objectives efficiently.