Modelo e método matemático para análise de investimentos de menor variância no mercado acionário

Abstract

This study seeks to accomplish a simulation that aims to apply a method of optimizing a asset distribution in an investment portfolio. Thus, literature on Linear Programming and the Markowitz Model was used to improve a portfolio with 12 stocks chosen from among the most traded on IBOVESPA, seeking, through covariance calculations, an ideal distribution that guaranteed lower risk and higher return. Furthermore, time series were used to predict the variation of these assets during the month of September and, therefore, in the comparison between the optimized and non-optimized portfolio, it was observed that the second obtained a better expected return. However, there was also a comparison between the actual variation values for the month of September (2023) and those predicted by the model, in which the optimized portfolio obtained a higher return than the non-optimized portfolio, contrary to expectations. Afterwards, the asset variation forecast was also made for the month of October (2023), and the non-optimized portfolio, in the same way, obtained a better return than the optimized portfolio, as well as the September forecast. However, the method applied proved to be efficient, since, after finding the period variation and applying these values to real data, a return considered significant and with risk minimization was obtained.