{"paper_id":"2311.13564","title":"High order universal portfolios","authors":["Gabriel Turinici"],"paper_date":"2023-11-22","paper_type":"theoretical","paper_url":"https://arxiv.org/pdf/2311.13564v3","page_url":"https://thequant.space/flowcharts/high-order-universal-portfolios/","scores":{"empirical_rigor":4,"math_complexity":8,"quadrant":"Lab Rats"},"proposes":{"status":"available","statement":"The paper proposes high order universal portfolios (HOUP) built by recursively adding the universal portfolio as a synthetic asset to the market.","quote":"We introduce in this paper the high order universal portfolios (HOUP) constructed from Cover’s initial suggestion by recursive additions to the underlying market.","summary":"The Cover universal portfolio (UP from now on) has many interesting theoretical and numerical properties and was investigated for a long time. Building on it, we explore what happens when we add this UP to the market as a new synthetic asset and construct by recurrence higher order UPs. We…","strategy":null},"assumptions":{"status":"available","items":[{"assumption":"All asset prices are strictly positive at any time.","type":"market","stated":"explicit","quote":"It is assumed that all asset prices are strictly positive at any time."},{"assumption":"The distribution over the simplex SK is uniform.","type":"model","stated":"explicit","quote":"we will use the uniform law over SK and denote Ew[·] or Ew∈SK[·] the average over it."},{"assumption":"Perturbation analysis is performed for small epsilon (ϵ) and first‑order terms are considered.","type":"model","stated":"explicit","quote":"¯µ = … Then, up to the first order in ϵ the Sharpe ratio of the log‑return of UP 2(M) is better (i.e., larger) than the Sharpe ratio of the log‑return of UP 1(M)."},{"assumption":"Log‑normal dynamics are used for simulated assets.","type":"model","stated":"explicit","quote":"We choose a log-normal model for k = 1, 2 : dSk/Sk = µkdt + σkdWk(t), Sk0 = 1,"},{"assumption":"Proportional transaction costs of 0.1%, 1% and 2% are applied with daily re‑balancing.","type":"execution","stated":"explicit","quote":"We consider proportional transaction costs and investigate fee levels of 0.1%, 1% and 2% assuming daily re‑balancing."},{"assumption":"The empirical dataset consists of daily quotations from 1965 to 1987 for 36 NYSE stocks.","type":"data","stated":"explicit","quote":"We implemented the high order universal portfolios on data from Cover (1991) which contains 22 years of daily quotations from 1965 to 1987 (T = 5650) for K = 36 stock symbols quoted on the New York Stock Exchange."}]},"datasets":{"status":"available","items":[{"name":"Old NYSE","provider":"Cover (1991) / Helmbold et al. (1998) / Kalai and Vempala (2","kind":"market-prices","access":"public","period":"1965-1987","universe":"36 NYSE stocks","quote":"We implemented the high order universal portfolios on data from Cover (1991) which contains 22 years of daily quotations from 1965 to 1987 (T = 5650) for K = 36 stock symbols quoted on the New York Stock Exchange."}],"known":[],"sample":{"start":null,"end":null,"market":null,"asset_class":null,"frequency":null},"data_resolution":"daily"},"transaction_costs":{"status":"available","modeled":"yes","headline_net_of_costs":null,"detail":"Proportional transaction costs of 0.1%, 1% and 2% are applied assuming daily re‑balancing.","quote":"We consider proportional transaction costs and investigate fee levels of 0.1%, 1% and 2% assuming daily re‑balancing.","source":"front_matter","rule":"R3","candidates":{"front_matter":"yes","evidence_card":"yes","evidence_card_quoted":true},"card_detail":"Proportional transaction costs of 0.1%, 1% and 2% are applied assuming daily re‑balancing.","card_quote":"We consider proportional transaction costs and investigate fee levels of 0.1%, 1% and 2% assuming daily re‑balancing.","meaning":"costs are part of the reported headline results"},"results":{"status":"available","evaluation_type":"trading-pnl","findings":[{"finding":"The second‑order HOUP has a higher Sharpe ratio than the standard UP under the perturbation regime.","kind":"theoretical","quote":"up to the first order in ϵ the Sharpe ratio of the log‑return of UP 2(M) is better (i.e., larger) than the Sharpe ratio of the log‑return of UP 1(M)."},{"finding":"HOUPs improve over the Cover UP in the toy example and benchmark experiments.","kind":"performance","quote":"The results presented in Figure 1 confirm the basic properties of the Cover UP and the remark 1: the UP improves over the individual performances and HOUP improves when compared to the UP."},{"finding":"The 10‑th order universal portfolio consistently outperforms the first‑order UP.","kind":"performance","quote":"The 10‑th order universal portfolio UP 10 appears to consistently outperform UP 1"},{"finding":"Higher‑order UPs achieve larger empirical Sharpe ratios than UP1 across parameter settings.","kind":"performance","quote":"First we note that the Sharpe ratio of UP ℓ is (in median) larger than the Sharpe ratio of UP 1 for all choices of the parameters, consistent with Proposition 3."},{"finding":"Transaction costs of 0.1% still allow HOUPs to beat the best individual stock, but performance advantage disappears above 1% fees.","kind":"performance","quote":"while the fee level of 0.1% still provides performances that are better than that of the best individual stock, this advantage breaks down for fee levels above 1%."}],"reported_performance":null,"out_of_sample":{"method":"none","headline_out_of_sample":null}},"independent_evidence":{"status":"none_found","supports":[],"contradicts":[],"qualifies":[],"compared_with":25,"citations":{"cited_by_in_corpus":0,"by_independent_authors":0,"how_cited":{"used_as_baseline":0,"uses_method_or_data":0,"extends":0,"discusses_findings":0,"mentions":0,"unclassified":0},"independent_papers":[],"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":1,"extracted_at":"2026-11-14T18:19:18Z"},"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/"}