Servicios Personalizados
Revista
Articulo
Indicadores
Citado por SciELO
Accesos
Links relacionados
Similares en SciELO
Compartir
Investigación & Desarrollo
versión On-line ISSN 2518-4431
Resumen
VARGAS SANCHEZ, Alejandro y MONJE PRUDENCIO, André Nicolas. OPTIMIZING EQUITY PORTFOLIOS WITH MACHINE LEARNING. Inv. y Des. [online]. 2023, vol.23, n.2, pp.23-45. Epub 31-Dic-2023. ISSN 2518-4431. https://doi.org/10.23881/idupbo.023.2-2e.
Given the growing role and acceptance of Artificial Intelligence in the world of Finance, this research proposes applying Machine Learning techniques to the management of equity investment portfolios, opening up the possibility to enhance the portfolio structuring process to yield optimal empirical results compared to traditional techniques, such as the Maximum Sharpe Ratio portfolio and the Equally Weighted portfolio. In contrast to these traditional techniques, the Affinity Propagation Clustering Technique is applied as the main approach to identify patterns of similar behavior among companies, complemented by the Graphical Lasso algorithm to estimate the data dependency structure, and Multi-Dimensional Scaling to improve the visual representation of the Clusters. Through the results, it is identified that the portfolio that maximizes the measures of risk and return is the one formed using these Machine Learning techniques. It is concluded that by combining these three Machine Learning techniques, a viable and effective alternative is obtained in the management of investment portfolios in the equity market.
Palabras clave : Portfolio Optimization; Machine Learning; Graphical Lasso; Clustering Affinity Propagation; Mutidimensional Scaling.