Comparação entre Modelos de optimização de carteiras de investimento baseados em utilidade e risco: uma análise aplicada aos activos da bolsa de valores Sul-Africana

Abstract

Portfolio optimization is a fundamental practice in the financial market, seeking to balance return and risk through the selection and efficient allocation of assets. Through mathematical and sta- tistical models such as Modern Portfolio Theory, investors can construct portfolios that maximize expected returns for a given level of risk or minimize risk for a desired return. All investors face the dilemma of return versus risk, that is, normally, an investment in a portfolio that promises to produce a high return is associated with high risk, on the other hand, an investment that generates a lower return will be less risky. Therefore, the investor has to decide which assets to invest in, how much to allocate to each and, finally, when to invest. Data on share prices in different South African companies were used, diversifying according to the sector, in the period from 26/03/2018 to 23/03/23. To achieve the objectives, two different models were used to construct and optimize portfolios, namely, utility-based model (Mean-Variance) and risk-based model (CVaR). Although the performance comparison chart of the Media-Variance and Cvar models shows that the Mean- Variance model presents higher returns than the CVaR model, it was found that the results obtained with the Mean-Variance model suggested a more selective asset allocation , excluding some com- panies from the investment portfolio. On the other hand, CVaR indicated a more inclusive strategy, recommending investment in all JSE companies, with greater emphasis on some of them.