Alocação de ativos financeiros: uma análise comparativa entre os métodos de alocação tradicionais e modelos decorrentes de inteligência artificial

Abstract

This study aims to compare the performance and risk of investment portfolios prepared using traditional methods and Artificial Intelligence (AI) techniques. The traditional methodology was based on fundamental analysis combined with the Markowitz portfolio optimization model, while the AI approach used criteria such as liquidity, sector diversification, and reduced correlation, in addition to the application of algorithms for mean-variance optimization. The portfolios were assembled using historical data from 2022 and evaluated for performance in 2023 using metrics such as annualized return, volatility, Sharpe ratio, Sortino ratio, and Value at Risk (VaR). The results showed that the traditional portfolio had higher profitability, but with greater exposure to risk. On the other hand, the AI portfolio presented lower returns, but with greater resilience in adverse scenarios, reinforcing the importance of diversification and multidimensional analysis provided by AI. The study concludes that the convergence between traditional and artificial intelligence methods may represent a promising strategy for more efficient and adaptive investment management.