{"paper_id":"2505.12806","title":"Hierarchical Representations for Evolving Acyclic Vector Autoregressions (HEAVe)","authors":["Cameron Cornell","Lewis Mitchell","Matthew Roughan"],"paper_date":"2025-05-19","paper_type":"empirical","paper_url":"https://arxiv.org/pdf/2505.12806v1","page_url":"https://thequant.space/flowcharts/hierarchical-representations-for-evolving-acyclic-vector-aut/","scores":{"empirical_rigor":6,"math_complexity":8.5,"quadrant":"Holy Grail"},"proposes":{"status":"available","statement":"The paper proposes an evolutionary approach to fitting acyclic vector autoregressive processes and introduces a novel hierarchical representation for causal networks.","quote":"We propose an evolutionary approach to fitting acyclic vector autoregressive processes and introduces a novel hierarchical representation that directly models structural elements within a time series system.","summary":"Causal networks offer an intuitive framework to understand influence structures within time series systems. However, the presence of cycles can obscure dynamic relationships and hinder hierarchical analysis. These networks are typically identified through multivariate predictive modelling, but…","strategy":"HEAVe hierarchical evolutionary algorithm"},"assumptions":{"status":"available","items":[{"assumption":"The analysis is limited to VAR(1) processes, assuming higher-order lags do not occur independently of a lag‑1 dependence.","type":"model","stated":"explicit","quote":"To streamline our discussion and estimation of causal networks, we limit our analysis to VAR(1) processes and omit the index p from our discussion, with A = A1 unless otherwise specified."},{"assumption":"Any causal link i→j will first manifest in order‑1 effects; lag p > 1 effects will not occur independently of a p = 1 dependence.","type":"model","stated":"explicit","quote":"For the empirical networks in Section 5 we operate under the assumption that any causal link i →j will first manifest in order-1 effects, and that lag p > 1 effects will not occur independently of a p = 1 dependence."},{"assumption":"The error term εt is often assumed to be Gaussian, though this is not a strict requirement.","type":"statistical","stated":"explicit","quote":"The εt is often assumed to be Gaussian, however this is not a strict requirement."}]},"datasets":{"status":"available","items":[{"name":"cryptocurrency return series","provider":null,"kind":"crypto-onchain","access":"unknown","period":null,"universe":"100 cryptocurrency return series","quote":"When applied to a dataset of 100 cryptocurrency return series, our method generates acyclic causal networks capturing key structural properties of the unconstrained model."},{"name":"simulated datasets","provider":null,"kind":"synthetic","access":"unknown","period":null,"universe":null,"quote":"On simulated datasets, our model retains most of the predictive accuracy of unconstrained models and outperforms permutation-based alternatives."}],"known":[],"sample":{"start":null,"end":null,"market":"cryptocurrency return series","asset_class":"crypto","frequency":null},"data_resolution":"daily"},"transaction_costs":{"status":"not_applicable","modeled":"n/a","headline_net_of_costs":null,"detail":null,"quote":null,"source":"front_matter","rule":"R3","candidates":{"front_matter":"n/a","evidence_card":"n/a","evidence_card_quoted":false},"meaning":"the paper reports no trading or execution result"},"results":{"status":"available","evaluation_type":"forecast-accuracy","findings":[{"finding":"The identified acyclic networks retain 92% of the predictive accuracy of unconstrained models while containing 90.2% as many links.","kind":"predictive-accuracy","quote":"In an empirical study of 100 cryptocurrencies, the identified acyclic networks retain 92% of the predictive accuracy of unconstrained models while"},{"finding":"HEAVe retains 92% of the predictive power of unconstrained VAR models in simulations, with 5.6‑8.3% improved prediction and 6.1‑14.1% improved hierarchy identification.","kind":"predictive-accuracy","quote":"A simulation study demonstrating the capacity for HEAVe to produce acyclic VAR models that retain 92% of the predictive power of unconstrained VAR. The results show clear advantages to the hierarchical approach, with 5.6-8.3% improved prediction, 6.1-14.1% improved hierarchy identification..."},{"finding":"The hierarchical evolutionary algorithm outperforms permutation‑based evolutionary algorithms in retaining predictive capacity and achieving higher link precision and more accurate hierarchy detection.","kind":"performance","quote":"HEAVe is validated on synthetic networks of varying sizes, where it retains a larger percentage of the unconstrained model’s predictive capacity compared to permutation‑based EAs."}],"reported_performance":{"strategy":"HEAVe hierarchical evolutionary algorithm","sharpe":null,"sharpe_benchmark":null,"annual_return_pct":null,"total_return_pct":null,"max_drawdown_pct":null},"out_of_sample":{"method":"none","headline_out_of_sample":false}},"independent_evidence":{"status":"none_found","supports":[],"contradicts":[],"qualifies":[],"compared_with":14,"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":"2026-11-14T17:37:52Z"},"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/"}