January 24, 2025 (NO COMMENTS)

Alexandre Antonov, Alexander Lipton and Marcos Lopez de Prado compare and contrast two portfolio allocation methods: the classical Markowitz approach and the hierarchical risk parity (HRP) approach. For both methods they derive analytical formulas for the noise in the optimal weights corresponding to the estimation uncertainty of the covariance matrix. They demonstrate that HRP is less noisy and therefore more robust than the traditional Markowitz model. They also