{"paper_id":"2411.05807","title":"Schur Complementary Allocation: A Unification of Hierarchical Risk Parity and Minimum Variance Portfolios","authors":["Peter Cotton"],"paper_date":"2024-10-29","paper_type":"theoretical","paper_url":"https://arxiv.org/pdf/2411.05807v1","page_url":"https://thequant.space/flowcharts/schur-complementary-allocation-a-unification-of-hierarchica/","scores":{"empirical_rigor":3,"math_complexity":7.5,"quadrant":"Lab Rats"},"proposes":{"status":"available","statement":"The paper introduces a new family of top‑down hierarchical allocation methods called Hierarchical Minimum Variance (HMV), which augment sub‑covariance matrices with Schur‑complement information.","quote":"Hierarchical Minimum Variance (HMV). Using the same way of representing hierarchical allocation as in 2.1, we now introduce a new family of top-down methods with the following change: (4.1) w(Σ) ∝ 1 ν(A′)w (A′′) 1 ν(D′)w (D′′) ! where, in contrast with 2.1, the A′ = A′(Σ) and A′′ = A′′(Σ) are modifications of A designed to incorporate some information from the full covariance ma","summary":"Despite many attempts to make optimization-based portfolio construction in the spirit of Markowitz robust and approachable, it is far from universally adopted. Meanwhile, the collection of more heuristic divide-and-conquer approaches was revitalized by Lopez de Prado where Hierarchical Risk Parity…","strategy":null},"assumptions":{"status":"available","items":[{"assumption":"Prior work tacitly assumes that the original sub‑covariance matrices A and D are passed unchanged to recursive steps.","type":"model","stated":"explicit","quote":"All prior work to our knowledge has tacitly assumed that the original A and D sub-covariance matrices should be passed to the recursive step unchanged although this does, in fact, warrant financial and mathematical scrutiny.","support":"full"},{"assumption":"The method assumes that weak shrinkage is applied to the covariance matrix before computing the minimum‑variance terminal portfolio.","type":"model","stated":"explicit","quote":"As a compromise we use “weak” shrinkage defined in Appendix 10. We also use this weakening to define the portfolio fitness ν(), namely as the variance of the minimum variance portfolio when this weak shrinkage is applied.","support":"full"},{"assumption":"Portfolio allocation packages may require the covariance matrix to be positive definite.","type":"data","stated":"implied","quote":"As a practical matter this depends on the assumptions made by portfolio allocation packages employed, or the explicit checks they perform (some, but not all, demand positive definite matrices, for example).","support":"full"}]},"datasets":{"status":"available","items":[{"name":"Synthetic constant‑off‑diagonal covariance matrix (p=500)","provider":null,"kind":"synthetic","access":"unknown","period":null,"universe":null,"quote":"An approximately constant off-diagonal covariance matrix dimension p = 500 is generated, with off-diagonal correlations averaging close to ρ = 0.35. See rand_symm_cov in [Cotton, 2022] for preicse details. 2.","support":"full"}],"known":[],"sample":{"start":null,"end":null,"market":null,"asset_class":null,"frequency":null},"data_resolution":"synthetic"},"transaction_costs":{"status":"not_applicable","modeled":"n/a","headline_net_of_costs":null,"detail":null,"quote":null,"source":"evidence_card","rule":"R2","candidates":{"front_matter":"unclear","evidence_card":"n/a","evidence_card_quoted":false},"meaning":"the paper reports no trading or execution result"},"results":{"status":"available","evaluation_type":"unclear","findings":[{"finding":"Including Schur‑complement information (γ>0) steadily reduces out‑of‑sample portfolio variance compared with the traditional top‑down allocation (γ=0).","kind":"performance","quote":"As can been seen from Figure 1, the inclusion of information using the Schur complement (i.e. taking γ > 0) steadily reduces portfolio variance.","support":"full"},{"finding":"The variance reduction corresponds to an economic benefit of roughly 10 basis points of additional annual return for institutional investors.","kind":"performance","quote":"The difference is material for institutional investors, and corresponds to a roughly equivalent increase in portfolio return on the order of 10 basis points (bps). For perspective, those gains carry economic value on the order of tens of millions per year, however, for funds that are roughly on the same scale as, say, the larger bespoke ETFs.","support":"full"},{"finding":"All tested values of γ outperform a naive optimization and equal‑weighting benchmark.","kind":"performance","quote":"In this comparison it is worth noting that all the choices of γ are superior to a naive optimization, in keeping with [De Prado, 2016] and also superior to equal-weighting of assets.","support":"full"}],"reported_performance":null,"out_of_sample":{"method":"unclear","headline_out_of_sample":null}},"independent_evidence":{"status":"available","supports":[],"contradicts":[],"qualifies":[],"compared_with":10,"citations":{"cited_by_in_corpus":4,"by_independent_authors":4,"how_cited":{"used_as_baseline":0,"uses_method_or_data":2,"extends":0,"discusses_findings":0,"mentions":0,"unclassified":2},"independent_papers":[{"paper":{"paper_id":"2507.04176","title":"skfolio: Portfolio Optimization in Python","paper_date":"2025-07-05","authors":["Carlo Nicolini","Matteo Manzi","Hugo Delatte"],"empirical_rigor":7,"page_url":"https://thequant.space/flowcharts/skfolio-portfolio-optimization-in-python/"},"how":"uses_method_or_data","quote":null},{"paper":{"paper_id":"2504.05743","title":"Causal Portfolio Optimization: Principles and Sensitivity-Based Solutions","paper_date":"2025-04-08","authors":["Alejandro Rodriguez Dominguez"],"empirical_rigor":3,"page_url":"https://thequant.space/flowcharts/causal-portfolio-optimization-principles-and-sensitivity-ba/"},"how":"uses_method_or_data","quote":null}],"scope":"Citations between papers in this corpus only; citations from outside it are not counted."}},"implementation":{"status":"available","public":false,"code_url":null,"note":"No repository presented by the authors as their own was found in the paper. Absence here is not proof that no code exists."},"replication":{"status":"none_found","card":null,"note":"No replication of this paper has been run by thequant.space yet."},"extraction":{"text_source":"fulltext","model":"openai/gpt-oss-120b","version":2,"extracted_at":"2026-11-14T19:54:20Z"},"provenance":{"claims":"Everything here is what the authors state in the paper. Nothing was re-run or reproduced by the platform.","quotes":"Assumptions, datasets, cost detail and findings were extracted by a language model. Every quote is verbatim paper text. Where an item has `support`, \"full\" means every figure in it appears in the quoted sentences and \"partial\" means a figure is stated elsewhere in the paper; items with a figure found nowhere in the paper were discarded.","datasets_known":"datasets.known lists well-known vendors and public datasets by a stable id, found by a dictionary over the full text where the sentence says the data is used; it is independent of the model and covers only names in the dictionary.","flags":"transaction_costs.modeled, results.evaluation_type and results.out_of_sample.method are platform classifications (structured-decision models); see https://thequant.space/score-guide/#penalty-flags","comparisons":"supports / contradicts / qualifies are a language model's reading of two papers' quoted findings, restricted to closely related papers that share no author. They are pointers to read, not verified replications, and coverage is partial.","implementation":"code_url is a repository the paper itself presents as its own (rule-scored, ambiguous cases judged by a model), checked for liveness when it was recorded."},"docs":"https://thequant.space/agents/"}