Maximizing the Sortino Ratio: downside risk management and alpha yielding

This paper presents an empirical analysis of portfolio optimization aimed at maximizing the Sortino ratio. We used a data set comprising weekly returns for 166 stocks traded in the Brazilian stock market from 2014 to 2024, to examine the capacity of this approach to generate excess returns and control downside risk in the long run. The portfolios use two different portfolio rebalancing frequencies (approximately three and six months) and three different target returns (minimum acceptable return thresholds - MAR), as required by the Sortino ratio. We find that especially portfolios using a three-month rebalancing window exhibited superior risk-adjusted performance, demonstrating the expected risk-return trade-off (elevated risk compensated by greater returns) for three target returns. In addition, we explore the sensitivity of portfolio construction to different MARs and check portfolio robustness, offering practitioners guidance on parameter selection that aligns with specific investment objectives in a multidimensional approach.