pflab_hierarchical | Hierarchical portfolios that need no expected returns and no matrix inversion: HRP, HERC (any of 20 risk measures incl. VaR, drawdown at risk, entropic risk, fourth moments) or Schur complementary allocation, on any prior, codependence distance and linkage. | |
pflab_nested_clusters | Nested clusters optimisation: assets are clustered, a mean-risk optimizer (objective, risk measure, prior) weights the assets inside each cluster, and a second one weights the clusters using their out-of-sample returns from k-fold cross-validation. | |
pflab_clustering | Codependence and distance matrices of the assets (Pearson, Kendall, Spearman, covariance, distance correlation or mutual information), their hierarchical clustering (linkage, number of clusters by the gap statistic or max_clusters, flat clusters, dendrogram coordinates) and an asset ordering (optimal leaf order or spectral seriation). | |
