Otimizando carteiras de investimento em ações com Programação Linear

Abstract

The stock market, characterized by high volatility and economic uncertainty, requires the use of quantitative strategies — that is, those based on numerical data and mathematical models — to guide investment allocation. In this context, this study proposed the application of Operations Research and Linear Programming, solved using the Simplex Method, with the objective of optimizing investment portfolios for different risk profiles: conservative, moderate, and ag- gressive. The methodology involved the collection and analysis of data from stocks that make up the Ibovespa Index. Subsequently, a mathematical optimization model was developed and implemented in Python using the MIP library. The model aimed to maximize the expected return of portfolios while respecting the risk constraints defined for each investor profile. As a main result, based on the analysis of the portfolios generated for the year 2023 from a universe of 87 assets, three low-risk portfolios with the highest returns were identified and recommended for any type of investor. Excluding these options, the model produced distinct outcomes for each profile, according to their respective risk levels.