{"paper_id":"2510.05809","title":"Coherent estimation of risk measures","authors":["Martin Aichele","Igor Cialenco","Damian Jelito","Marcin Pitera"],"paper_date":"2025-10-07","paper_type":"empirical","paper_url":"https://arxiv.org/pdf/2510.05809v3","page_url":"https://thequant.space/flowcharts/coherent-estimation-of-risk-measures/","scores":{"empirical_rigor":4,"math_complexity":8.5,"quadrant":"Lab Rats"},"proposes":{"status":"available","statement":"The paper proposes a statistical framework for coherent risk estimators (CREs) with a robust representation and structural classification of all CREs.","quote":"The key novel contribution of this paper is the development of the robust representation and structural classification of all CREs, see Theorem 4.1.","summary":"We develop a statistical framework for risk estimation, inspired by the axiomatic theory of risk measures. Coherent risk estimators – functionals of P&L samples inheriting the economic properties of risk measures – are defined and characterized through robust representations linked to L-estimators.…","strategy":null},"assumptions":{"status":"available","items":[{"assumption":"The underlying sample X is an i.i.d. sequence.","type":"statistical","stated":"explicit","quote":"For simplicity, throughout this work we assume that X is an i.i.d. sequence."},{"assumption":"Some results assume the CRE is comonotonic.","type":"model","stated":"explicit","quote":"assuming additionally that a CRE is comonotonic, we show that it can be represented as a single L-estimator; see Theorem 4.10."},{"assumption":"Consistency results are conjectured to extend under ergodic stationarity (e.g., AR, MA, GARCH).","type":"statistical","stated":"explicit","quote":"We conjecture that the definition of consistency, together with Proposition 5.3, can be established under suitable ergodic stationarity assumptions, see (Hayashi, 2011, Section 2)."}]},"datasets":{"status":"available","items":[{"name":"Fama–French dataset","provider":null,"kind":"market-prices","access":"unknown","period":null,"universe":null,"quote":"Finally, we checked the subadditivity violation for the empirical VaR estimator using the Fama–French dataset."},{"name":"Simulated data","provider":null,"kind":"synthetic","access":"unknown","period":null,"universe":null,"quote":"Numerical illustrations based on simulated and market data demonstrate that coherence of a risk measure does not necessarily carry over to its estimators."}],"known":[{"id":"french-library","name":"Kenneth French Data Library","provider":"Kenneth R. French","kind":"market-prices","access":"public","quote":"In addition to simulated data, we use empirical market data from the Fama–French data library; see French (2026) for details.","mentions":1,"used_in":1}],"sample":{"start":null,"end":null,"market":null,"asset_class":null,"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":"unclear","findings":[{"finding":"Subadditivity was violated in 22.4% of the 4,950 portfolio pairs for the empirical VaR estimator on the Fama–French dataset.","kind":"negative","quote":"Subadditivity was violated in 22.4% of the total 4950 pairs."},{"finding":"For ρ = 0.9, the GPD plug‑in ES estimator violated subadditivity in more than 10% of runs, with 8% of violations exceeding a 10% excess over the sum of individual ES estimates.","kind":"negative","quote":"For ρ = 0.9, the subadditivity is violated in more than 10% of the runs. Moreover, for 8% of the violations, estimated ES for x + x′ exceeded the sum of individual ES estimates by more than 10%."},{"finding":"Estimator #3 provides the most conservative risk estimates and the best exceedance‑frequency control, but at the cost of a positive bias and higher scaled RMSE.","kind":"performance","quote":"Estimator #3, which produces more conservative risk estimates and achieves improved control of the exceedance frequency. The improved coverage of estimator #3, however, comes at the cost of a positive (conservative) bias and a deterioration in scaled RMSE."},{"finding":"Estimator #2 has the smallest absolute statistical bias across scenarios, while estimator #1 shows superior variability‑adjusted error.","kind":"performance","quote":"Estimator #2, in turn, exhibits the smallest absolute statistical bias throughout the considered scenarios."},{"finding":"Parametric plug‑in estimators for Gaussian distributions may fail monotonicity, and semi‑parametric ES estimators for the GPD are not subadditive with positive probability.","kind":"theoretical","quote":"parametric plug‑in estimators for Gaussian distributions may fail monotonicity; semi‑parametric ES estimators for the GPD are not subadditive with positive probability."}],"reported_performance":null,"out_of_sample":{"method":"none","headline_out_of_sample":null}},"independent_evidence":{"status":"none_found","supports":[],"contradicts":[],"qualifies":[],"compared_with":47,"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":"2026-11-14T17:25:23Z"},"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/"}