Portfolio Optimization Strategy for Concentrated Portfolios: Models and Time Horizons

Abstract

Recent academic literature has noted that high conviction, or concentrated, portfolios (i.e., those with approximately 15 or fewer assets) often outperform larger, theoretically better diversified, portfolios. Recent research has also suggested that some of the gains from traditional portfolio optimization are in fact from the selection of relatively lower volatility stocks through the optimization process. To help wealth managers navigate the new uncertainty in the academic literature, the authors conduct a large-scale test of four approaches to portfolio construction that could be applied to high conviction portfolios. This study found that the traditional mean–variance-optimization approach worked well with very small portfolios but that a minimum Black Swan risk portfolio worked better when more holdings were included in the portfolios. The authors discuss the results and possible implications for winning approaches for wealth management.