{"id":"5622ffa79dbb7fdc2632983f38c8f724c7bf31e3ba3aa0ecce9d8ea053739c8f","title":"The authors propose a council for assigning credit among trading agents","argument":"In their paper, Yunhua Pei, Zerui Ge, Jin Zheng and John Cartlidge propose Market Regime Council (MRC), a method for deciding how much influence specialist agents should have in a multi-agent system managing portfolios. The authors say these systems lack a principled way to assign credit among agents, may face cold-start dominance when market regimes shift, and offer limited transparency into how allocations are formed. MRC calculates Shapley credits across individual agents and coalitions, then uses those credits to adjust agent weights over time. It also uses a Bayesian adaptive mixture to stabilize early periods, regime-dependent multipliers to change agent authority, and a five-layer causal trace to record each rebalance. In tests spanning 1,037 trading days, 13 crypto assets and five seeds, the authors report a Sharpe ratio of 1.51 and cumulative return of 440.1%. They say MRC ranked first on CR, SR and IR among active baselines and had the lowest MDD among active methods. Ablations, which test components separately, led them to attribute the gains to combining coalition outputs with Shapley-weighted integration, rather than to any single stage. The reported results concern this portfolio-management setup; performance outside the reported assets and seeds is not established by the abstract.","topicIds":["intelligence-and-discovery"],"domain":"science","claimType":"attributed reading","limits":"The reported performance results are for portfolio management across 13 crypto assets and five seeds; the abstract does not report results outside that setup. The abstract does not specify the identities of the active baselines or give the full ablation results.","speaker":{"name":"Pei, Yunhua, Ge, Zerui, Zheng, Jin, Cartlidge, John","roleAtSource":"Article author"},"source":{"title":"Market Regime Council for Dynamic Credit Assignment in Multi-Agent LLM Decision Systems","publisher":"arXiv cs.AI","format":"Article","url":"https://arxiv.org/abs/2605.24490","publishedAt":"2026-05-23","dateBasis":"Date basis not provided."},"evidence":[{"text":"The authors propose MRC as an online credit-assignment mechanism for coalitions of specialist agents, with portfolio management as the task.","excerpt":"We propose Market Regime Council (MRC), an online credit-assignment mechanism over coalitions of specialist agents, with portfolio management as the downstream task.","url":"https://arxiv.org/abs/2605.24490","relationship":"Source-stated"},{"text":"MRC calculates Shapley credits across individual and coalition outputs to set agent weights, and uses further mechanisms to adjust authority and record rebalances.","excerpt":"MRC recomputes coalition-based Shapley weights from exponentially weighted performance histories, uses a Bayesian adaptive mixture to stabilize early periods, applies regime-dependent multipliers to adjust agent authority, and records each rebalance through a five-layer causal trace.","url":"https://arxiv.org/abs/2605.24490","relationship":"Source-stated"},{"text":"In tests across 1,037 trading days, 13 crypto assets and five seeds, the authors report a Sharpe ratio of 1.51 and cumulative return of 440.1%.","excerpt":"Over 1,037 trading days across 13 crypto assets and five seeds, MRC achieves a Sharpe ratio of 1.51 and a cumulative return of 440.1%","url":"https://arxiv.org/abs/2605.24490","relationship":"Source-stated"},{"text":"The authors report that MRC ranked first on CR, SR and IR among active baselines and had the lowest MDD among active methods.","excerpt":"ranking first on CR, SR, and IR among active baselines and attaining the lowest MDD among active methods.","url":"https://arxiv.org/abs/2605.24490","relationship":"Source-stated"},{"text":"The authors attribute the gains to Shapley-weighted integration across coalition outputs, rather than any one stage in isolation.","excerpt":"the gains come from Shapley-weighted integration across coalition outputs rather than from any single stage in isolation.","url":"https://arxiv.org/abs/2605.24490","relationship":"Source-stated"}],"sourceMedia":{"status":"not_found","sourceUrl":"https://arxiv.org/abs/2605.24490","checkedAt":"2026-11-12T01:56:07.942Z","items":[],"note":"No suitable media was found by the supported page extraction."},"assumptions":[],"openQuestion":{"attribution":"Open question","text":"How would MRC perform outside the 13 crypto assets and five seeds reported in the paper?"},"whatWouldChangeView":{"attribution":"Change condition not recorded","text":""},"editorialComparisons":[],"live":{"title":"The authors propose a council for assigning credit among trading agents","summary":"In their paper, Yunhua Pei, Zerui Ge, Jin Zheng and John Cartlidge propose Market Regime Council (MRC), a method for deciding how much influence specialist agents should have in a multi-agent system managing portfolios. The authors say these systems lack a principled way to assign credit among agents, may face cold-start dominance when market regimes shift, and offer limited transparency into how allocations are formed. MRC calculates Shapley credits across individual agents and coalitions, then uses those credits to adjust agent weights over time. It also uses a Bayesian adaptive mixture to stabilize early periods, regime-dependent multipliers to change agent authority, and a five-layer causal trace to record each rebalance. In tests spanning 1,037 trading days, 13 crypto assets and five seeds, the authors report a Sharpe ratio of 1.51 and cumulative return of 440.1%. They say MRC ranked first on CR, SR and IR among active baselines and had the lowest MDD among active methods. Ablations, which test components separately, led them to attribute the gains to combining coalition outputs with Shapley-weighted integration, rather than to any single stage. The reported results concern this portfolio-management setup; performance outside the reported assets and seeds is not established by the abstract.","question":"How would MRC perform outside the 13 crypto assets and five seeds reported in the paper?","claims":[{"text":"The authors propose MRC as an online credit-assignment mechanism for coalitions of specialist agents, with portfolio management as the task.","excerpt":"We propose Market Regime Council (MRC), an online credit-assignment mechanism over coalitions of specialist agents, with portfolio management as the downstream task.","support":"source-stated"},{"text":"MRC calculates Shapley credits across individual and coalition outputs to set agent weights, and uses further mechanisms to adjust authority and record rebalances.","excerpt":"MRC recomputes coalition-based Shapley weights from exponentially weighted performance histories, uses a Bayesian adaptive mixture to stabilize early periods, applies regime-dependent multipliers to adjust agent authority, and records each rebalance through a five-layer causal trace.","support":"source-stated"},{"text":"In tests across 1,037 trading days, 13 crypto assets and five seeds, the authors report a Sharpe ratio of 1.51 and cumulative return of 440.1%.","excerpt":"Over 1,037 trading days across 13 crypto assets and five seeds, MRC achieves a Sharpe ratio of 1.51 and a cumulative return of 440.1%","support":"source-stated"},{"text":"The authors report that MRC ranked first on CR, SR and IR among active baselines and had the lowest MDD among active methods.","excerpt":"ranking first on CR, SR, and IR among active baselines and attaining the lowest MDD among active methods.","support":"source-stated"},{"text":"The authors attribute the gains to Shapley-weighted integration across coalition outputs, rather than any one stage in isolation.","excerpt":"the gains come from Shapley-weighted integration across coalition outputs rather than from any single stage in isolation.","support":"source-stated"}],"limitations":["The reported performance results are for portfolio management across 13 crypto assets and five seeds; the abstract does not report results outside that setup.","The abstract does not specify the identities of the active baselines or give the full ablation results."],"mediaId":"f914e192cdf7791f73ff74476705fa83766062bafb33aabef30613ab6d54c27b","correctionReason":null,"id":"5622ffa79dbb7fdc2632983f38c8f724c7bf31e3ba3aa0ecce9d8ea053739c8f","publicationId":"c40fe5bb-3ab1-49ed-8e9b-e2b7732933b1","version":1,"sourceUrl":"https://arxiv.org/abs/2605.24490","sourceTitle":"Market Regime Council for Dynamic Credit Assignment in Multi-Agent LLM Decision Systems","publisher":"arXiv cs.AI","publishedAt":null,"originalPublished":{"value":"2026-05-23","precision":"day","basis":"publisher_citation_date"},"firstObservedAt":"2026-11-11T05:05:09.255Z","createdAt":"2026-11-11T23:48:32.358Z","historical":true,"entityIds":["arxiv"],"topicIds":["intelligence-and-discovery"],"media":[{"id":"f914e192cdf7791f73ff74476705fa83766062bafb33aabef30613ab6d54c27b","kind":"article","url":"https://arxiv.org/abs/2605.24490","sourceUrl":"https://arxiv.org/abs/2605.24490","label":"Read original","provenance":"source_link","verification":{"status":"verified","checkedAt":"2026-11-11T23:47:28.574Z","method":"acquired_source","contentType":"text/html"},"recommendationReason":"Recommended by Jev for understanding this source; checked by Luna."}],"sourceMedia":{"status":"not_found","sourceUrl":"https://arxiv.org/abs/2605.24490","checkedAt":"2026-11-12T01:56:07.942Z","items":[],"note":"No suitable media was found by the supported page extraction."},"editorialDecision":{"outcome":"accept","reasonCodes":["material_development"],"rubricVersion":"stream-admission-v5","mediaId":"f914e192cdf7791f73ff74476705fa83766062bafb33aabef30613ab6d54c27b","confidence":0.73,"provider":"typesafe","model":"jev-1.13.0","inputHash":"6f4cda2daf079ed9bb817dfd31a00f2528da9112df6d6d70b27605f4ec340e4e","decidedAt":"2026-11-11T23:47:29.646Z"},"provenance":{"textHash":"df6e2958ca538b5cf7ce859be50ade8559cd4fd1c8f1720b3d0ac4dca2fcc6b3","observationId":"356397bc6d63195efc6b071235263166f4c164fc7b4449c77fba7d037e4e9443","retrievedAt":"2026-11-11T23:47:28.483Z","kind":"publisher-page","preparer":{"model":"gpt-6-luna","runId":"03a2853c-ec8a-4979-8d98-870e2950d74b"},"reviewer":{"model":"gpt-6-luna","runId":"5954d789-2197-4c4d-8421-2354e9799c6f"}},"audit":{"status":"uncertain","reviewedAt":"2026-11-11T23:48:32.358Z","model":"anthropic/claude-opus-5.5"},"sourceAttribution":{"authors":[{"name":"Pei, Yunhua","url":null,"kind":"unknown"},{"name":"Ge, Zerui","url":null,"kind":"unknown"},{"name":"Zheng, Jin","url":null,"kind":"unknown"},{"name":"Cartlidge, John","url":null,"kind":"unknown"}],"authorRole":"article-author","hosts":[],"status":"stated","authorRoster":{"truncated":false,"omitted":0,"assertionGroupsOmitted":0},"participation":[{"actor":{"name":"Pei, Yunhua","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Pei, Yunhua","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2605.24490","locator":"head/meta[name=author|citation_author]"},"assertionId":"ad8ae4f3f53d070702739ac1dc423932348fc2b4e76a09167f24411e1b8a78e2:participation:0","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-12T01:56:07.942Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"ad8ae4f3f53d070702739ac1dc423932348fc2b4e76a09167f24411e1b8a78e2","sourceUrl":"https://arxiv.org/abs/2605.24490","contentSha256":"d11bd5e22f56e4a4decd2ba4b26ce6553fb6c42cfe5cd393a8892685f4e581d0","textHash":"df6e2958ca538b5cf7ce859be50ade8559cd4fd1c8f1720b3d0ac4dca2fcc6b3","observedAt":"2026-11-12T01:56:07.942Z"}},{"actor":{"name":"Ge, Zerui","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Ge, Zerui","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2605.24490","locator":"head/meta[name=author|citation_author]"},"assertionId":"ad8ae4f3f53d070702739ac1dc423932348fc2b4e76a09167f24411e1b8a78e2:participation:1","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-12T01:56:07.942Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"ad8ae4f3f53d070702739ac1dc423932348fc2b4e76a09167f24411e1b8a78e2","sourceUrl":"https://arxiv.org/abs/2605.24490","contentSha256":"d11bd5e22f56e4a4decd2ba4b26ce6553fb6c42cfe5cd393a8892685f4e581d0","textHash":"df6e2958ca538b5cf7ce859be50ade8559cd4fd1c8f1720b3d0ac4dca2fcc6b3","observedAt":"2026-11-12T01:56:07.942Z"}},{"actor":{"name":"Zheng, Jin","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Zheng, Jin","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2605.24490","locator":"head/meta[name=author|citation_author]"},"assertionId":"ad8ae4f3f53d070702739ac1dc423932348fc2b4e76a09167f24411e1b8a78e2:participation:2","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-12T01:56:07.942Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"ad8ae4f3f53d070702739ac1dc423932348fc2b4e76a09167f24411e1b8a78e2","sourceUrl":"https://arxiv.org/abs/2605.24490","contentSha256":"d11bd5e22f56e4a4decd2ba4b26ce6553fb6c42cfe5cd393a8892685f4e581d0","textHash":"df6e2958ca538b5cf7ce859be50ade8559cd4fd1c8f1720b3d0ac4dca2fcc6b3","observedAt":"2026-11-12T01:56:07.942Z"}},{"actor":{"name":"Cartlidge, John","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Cartlidge, John","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2605.24490","locator":"head/meta[name=author|citation_author]"},"assertionId":"ad8ae4f3f53d070702739ac1dc423932348fc2b4e76a09167f24411e1b8a78e2:participation:3","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-12T01:56:07.942Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"ad8ae4f3f53d070702739ac1dc423932348fc2b4e76a09167f24411e1b8a78e2","sourceUrl":"https://arxiv.org/abs/2605.24490","contentSha256":"d11bd5e22f56e4a4decd2ba4b26ce6553fb6c42cfe5cd393a8892685f4e581d0","textHash":"df6e2958ca538b5cf7ce859be50ade8559cd4fd1c8f1720b3d0ac4dca2fcc6b3","observedAt":"2026-11-12T01:56:07.942Z"}}],"participationCoverage":{"total":4,"returned":4,"complete":true,"fullAssertions":"source-metadata-endpoint"},"distribution":null,"publisher":null,"sourceUrl":"https://arxiv.org/abs/2605.24490","snapshotId":"ad8ae4f3f53d070702739ac1dc423932348fc2b4e76a09167f24411e1b8a78e2","contentSha256":"d11bd5e22f56e4a4decd2ba4b26ce6553fb6c42cfe5cd393a8892685f4e581d0","textHash":"df6e2958ca538b5cf7ce859be50ade8559cd4fd1c8f1720b3d0ac4dca2fcc6b3","observedAt":"2026-11-12T01:56:07.942Z","basis":"publisher-page-metadata","sourceLinks":47,"linksTruncated":false,"metadataUrl":"/api/pipeline/stream/5622ffa79dbb7fdc2632983f38c8f724c7bf31e3ba3aa0ecce9d8ea053739c8f/source"}},"canonicalUrl":"https://escapemeridian.com/readings/the-authors-propose-a-council-for-assigning-credit-among-5622ffa7","formats":{"html":"/readings/the-authors-propose-a-council-for-assigning-credit-among-5622ffa7","markdown":"/readings/the-authors-propose-a-council-for-assigning-credit-among-5622ffa7.md","json":"/readings/the-authors-propose-a-council-for-assigning-credit-among-5622ffa7.json"},"attribution":"Headline and summary by Escape Meridian, prepared by Luna from the attributed source. A source check is not independent confirmation of the source’s claims.","maintainedGuideLinks":[],"readerContext":{"version":"reader-context-v1","attribution":"Reference context selected by Escape Meridian; separate from source evidence. Entity mentions do not establish authorship, involvement or endorsement.","terms":[],"entities":[{"id":"arxiv","name":"arXiv","kind":"organization","context":"arXiv is a research manuscript repository where authors share papers that may not yet have been peer reviewed.","referenceUrls":["https://arxiv.org/","https://info.arxiv.org/help/api/index.html"],"path":"/entities/arxiv","json":"/entities/arxiv.json","markdown":"/entities/arxiv.md"},{"id":"time","name":"TIME","kind":"organization","context":"TIME is a news magazine publishing reporting on world affairs, business, technology and the workers behind digital services.","referenceUrls":["https://time.com/"],"path":"/entities/time","json":"/entities/time.json","markdown":"/entities/time.md"}]},"researchQuestions":[],"withdrawn":false,"references":{"schemaVersion":1,"target":{"type":"reading","id":"5622ffa79dbb7fdc2632983f38c8f724c7bf31e3ba3aa0ecce9d8ea053739c8f","revision":"c40fe5bb-3ab1-49ed-8e9b-e2b7732933b1","revisionKind":"publication-id","title":"The authors propose a council for assigning credit among trading agents","url":"https://escapemeridian.com/readings/the-authors-propose-a-council-for-assigning-credit-among-5622ffa7"},"source":{"type":"source","id":"https://arxiv.org/abs/2605.24490","revision":"df6e2958ca538b5cf7ce859be50ade8559cd4fd1c8f1720b3d0ac4dca2fcc6b3","revisionKind":"sha256-source-text","title":"Market Regime Council for Dynamic Credit Assignment in Multi-Agent LLM Decision Systems","url":"https://arxiv.org/abs/2605.24490"},"status":"available","items":[],"next":null,"scope":"Explicit citations in public readings and public file attachments to this reading. Shared topics are not citations. Other artifact types need an explicit reference adapter."},"revisions":[{"id":"5622ffa79dbb7fdc2632983f38c8f724c7bf31e3ba3aa0ecce9d8ea053739c8f","title":"The authors propose a council for assigning credit among trading agents","argument":"In their paper, Yunhua Pei, Zerui Ge, Jin Zheng and John Cartlidge propose Market Regime Council (MRC), a method for deciding how much influence specialist agents should have in a multi-agent system managing portfolios. The authors say these systems lack a principled way to assign credit among agents, may face cold-start dominance when market regimes shift, and offer limited transparency into how allocations are formed. MRC calculates Shapley credits across individual agents and coalitions, then uses those credits to adjust agent weights over time. It also uses a Bayesian adaptive mixture to stabilize early periods, regime-dependent multipliers to change agent authority, and a five-layer causal trace to record each rebalance. In tests spanning 1,037 trading days, 13 crypto assets and five seeds, the authors report a Sharpe ratio of 1.51 and cumulative return of 440.1%. They say MRC ranked first on CR, SR and IR among active baselines and had the lowest MDD among active methods. Ablations, which test components separately, led them to attribute the gains to combining coalition outputs with Shapley-weighted integration, rather than to any single stage. The reported results concern this portfolio-management setup; performance outside the reported assets and seeds is not established by the abstract.","topicIds":["intelligence-and-discovery"],"domain":"science","claimType":"attributed reading","limits":"The reported performance results are for portfolio management across 13 crypto assets and five seeds; the abstract does not report results outside that setup. The abstract does not specify the identities of the active baselines or give the full ablation results.","speaker":{"name":"Pei, Yunhua, Ge, Zerui, Zheng, Jin, Cartlidge, John","roleAtSource":"Article author"},"source":{"title":"Market Regime Council for Dynamic Credit Assignment in Multi-Agent LLM Decision Systems","publisher":"arXiv cs.AI","format":"Article","url":"https://arxiv.org/abs/2605.24490","publishedAt":"2026-05-23","dateBasis":"Date basis not provided."},"evidence":[{"text":"The authors propose MRC as an online credit-assignment mechanism for coalitions of specialist agents, with portfolio management as the task.","excerpt":"We propose Market Regime Council (MRC), an online credit-assignment mechanism over coalitions of specialist agents, with portfolio management as the downstream task.","url":"https://arxiv.org/abs/2605.24490","relationship":"Source-stated"},{"text":"MRC calculates Shapley credits across individual and coalition outputs to set agent weights, and uses further mechanisms to adjust authority and record rebalances.","excerpt":"MRC recomputes coalition-based Shapley weights from exponentially weighted performance histories, uses a Bayesian adaptive mixture to stabilize early periods, applies regime-dependent multipliers to adjust agent authority, and records each rebalance through a five-layer causal trace.","url":"https://arxiv.org/abs/2605.24490","relationship":"Source-stated"},{"text":"In tests across 1,037 trading days, 13 crypto assets and five seeds, the authors report a Sharpe ratio of 1.51 and cumulative return of 440.1%.","excerpt":"Over 1,037 trading days across 13 crypto assets and five seeds, MRC achieves a Sharpe ratio of 1.51 and a cumulative return of 440.1%","url":"https://arxiv.org/abs/2605.24490","relationship":"Source-stated"},{"text":"The authors report that MRC ranked first on CR, SR and IR among active baselines and had the lowest MDD among active methods.","excerpt":"ranking first on CR, SR, and IR among active baselines and attaining the lowest MDD among active methods.","url":"https://arxiv.org/abs/2605.24490","relationship":"Source-stated"},{"text":"The authors attribute the gains to Shapley-weighted integration across coalition outputs, rather than any one stage in isolation.","excerpt":"the gains come from Shapley-weighted integration across coalition outputs rather than from any single stage in isolation.","url":"https://arxiv.org/abs/2605.24490","relationship":"Source-stated"}],"sourceMedia":{"status":"not_found","sourceUrl":"https://arxiv.org/abs/2605.24490","checkedAt":"2026-11-12T01:56:07.942Z","items":[],"note":"No suitable media was found by the supported page extraction."},"assumptions":[],"openQuestion":{"attribution":"Open question","text":"How would MRC perform outside the 13 crypto assets and five seeds reported in the paper?"},"whatWouldChangeView":{"attribution":"Change condition not recorded","text":""},"editorialComparisons":[],"live":{"title":"The authors propose a council for assigning credit among trading agents","summary":"In their paper, Yunhua Pei, Zerui Ge, Jin Zheng and John Cartlidge propose Market Regime Council (MRC), a method for deciding how much influence specialist agents should have in a multi-agent system managing portfolios. The authors say these systems lack a principled way to assign credit among agents, may face cold-start dominance when market regimes shift, and offer limited transparency into how allocations are formed. MRC calculates Shapley credits across individual agents and coalitions, then uses those credits to adjust agent weights over time. It also uses a Bayesian adaptive mixture to stabilize early periods, regime-dependent multipliers to change agent authority, and a five-layer causal trace to record each rebalance. In tests spanning 1,037 trading days, 13 crypto assets and five seeds, the authors report a Sharpe ratio of 1.51 and cumulative return of 440.1%. They say MRC ranked first on CR, SR and IR among active baselines and had the lowest MDD among active methods. Ablations, which test components separately, led them to attribute the gains to combining coalition outputs with Shapley-weighted integration, rather than to any single stage. The reported results concern this portfolio-management setup; performance outside the reported assets and seeds is not established by the abstract.","question":"How would MRC perform outside the 13 crypto assets and five seeds reported in the paper?","claims":[{"text":"The authors propose MRC as an online credit-assignment mechanism for coalitions of specialist agents, with portfolio management as the task.","excerpt":"We propose Market Regime Council (MRC), an online credit-assignment mechanism over coalitions of specialist agents, with portfolio management as the downstream task.","support":"source-stated"},{"text":"MRC calculates Shapley credits across individual and coalition outputs to set agent weights, and uses further mechanisms to adjust authority and record rebalances.","excerpt":"MRC recomputes coalition-based Shapley weights from exponentially weighted performance histories, uses a Bayesian adaptive mixture to stabilize early periods, applies regime-dependent multipliers to adjust agent authority, and records each rebalance through a five-layer causal trace.","support":"source-stated"},{"text":"In tests across 1,037 trading days, 13 crypto assets and five seeds, the authors report a Sharpe ratio of 1.51 and cumulative return of 440.1%.","excerpt":"Over 1,037 trading days across 13 crypto assets and five seeds, MRC achieves a Sharpe ratio of 1.51 and a cumulative return of 440.1%","support":"source-stated"},{"text":"The authors report that MRC ranked first on CR, SR and IR among active baselines and had the lowest MDD among active methods.","excerpt":"ranking first on CR, SR, and IR among active baselines and attaining the lowest MDD among active methods.","support":"source-stated"},{"text":"The authors attribute the gains to Shapley-weighted integration across coalition outputs, rather than any one stage in isolation.","excerpt":"the gains come from Shapley-weighted integration across coalition outputs rather than from any single stage in isolation.","support":"source-stated"}],"limitations":["The reported performance results are for portfolio management across 13 crypto assets and five seeds; the abstract does not report results outside that setup.","The abstract does not specify the identities of the active baselines or give the full ablation results."],"mediaId":"f914e192cdf7791f73ff74476705fa83766062bafb33aabef30613ab6d54c27b","correctionReason":null,"id":"5622ffa79dbb7fdc2632983f38c8f724c7bf31e3ba3aa0ecce9d8ea053739c8f","publicationId":"c40fe5bb-3ab1-49ed-8e9b-e2b7732933b1","version":1,"sourceUrl":"https://arxiv.org/abs/2605.24490","sourceTitle":"Market Regime Council for Dynamic Credit Assignment in Multi-Agent LLM Decision Systems","publisher":"arXiv cs.AI","publishedAt":null,"originalPublished":{"value":"2026-05-23","precision":"day","basis":"publisher_citation_date"},"firstObservedAt":"2026-11-11T05:05:09.255Z","createdAt":"2026-11-11T23:48:32.358Z","historical":true,"entityIds":["arxiv"],"topicIds":["intelligence-and-discovery"],"media":[{"id":"f914e192cdf7791f73ff74476705fa83766062bafb33aabef30613ab6d54c27b","kind":"article","url":"https://arxiv.org/abs/2605.24490","sourceUrl":"https://arxiv.org/abs/2605.24490","label":"Read original","provenance":"source_link","verification":{"status":"verified","checkedAt":"2026-11-11T23:47:28.574Z","method":"acquired_source","contentType":"text/html"},"recommendationReason":"Recommended by Jev for understanding this source; checked by Luna."}],"sourceMedia":{"status":"not_found","sourceUrl":"https://arxiv.org/abs/2605.24490","checkedAt":"2026-11-12T01:56:07.942Z","items":[],"note":"No suitable media was found by the supported page extraction."},"editorialDecision":{"outcome":"accept","reasonCodes":["material_development"],"rubricVersion":"stream-admission-v5","mediaId":"f914e192cdf7791f73ff74476705fa83766062bafb33aabef30613ab6d54c27b","confidence":0.73,"provider":"typesafe","model":"jev-1.13.0","inputHash":"6f4cda2daf079ed9bb817dfd31a00f2528da9112df6d6d70b27605f4ec340e4e","decidedAt":"2026-11-11T23:47:29.646Z"},"provenance":{"textHash":"df6e2958ca538b5cf7ce859be50ade8559cd4fd1c8f1720b3d0ac4dca2fcc6b3","observationId":"356397bc6d63195efc6b071235263166f4c164fc7b4449c77fba7d037e4e9443","retrievedAt":"2026-11-11T23:47:28.483Z","kind":"publisher-page","preparer":{"model":"gpt-6-luna","runId":"03a2853c-ec8a-4979-8d98-870e2950d74b"},"reviewer":{"model":"gpt-6-luna","runId":"5954d789-2197-4c4d-8421-2354e9799c6f"}},"audit":{"status":"uncertain","reviewedAt":"2026-11-11T23:48:32.358Z","model":"anthropic/claude-opus-5.5"},"sourceAttribution":{"authors":[{"name":"Pei, Yunhua","url":null,"kind":"unknown"},{"name":"Ge, Zerui","url":null,"kind":"unknown"},{"name":"Zheng, Jin","url":null,"kind":"unknown"},{"name":"Cartlidge, John","url":null,"kind":"unknown"}],"authorRole":"article-author","hosts":[],"status":"stated","authorRoster":{"truncated":false,"omitted":0,"assertionGroupsOmitted":0},"participation":[{"actor":{"name":"Pei, Yunhua","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Pei, Yunhua","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2605.24490","locator":"head/meta[name=author|citation_author]"},"assertionId":"ad8ae4f3f53d070702739ac1dc423932348fc2b4e76a09167f24411e1b8a78e2:participation:0","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-12T01:56:07.942Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"ad8ae4f3f53d070702739ac1dc423932348fc2b4e76a09167f24411e1b8a78e2","sourceUrl":"https://arxiv.org/abs/2605.24490","contentSha256":"d11bd5e22f56e4a4decd2ba4b26ce6553fb6c42cfe5cd393a8892685f4e581d0","textHash":"df6e2958ca538b5cf7ce859be50ade8559cd4fd1c8f1720b3d0ac4dca2fcc6b3","observedAt":"2026-11-12T01:56:07.942Z"}},{"actor":{"name":"Ge, Zerui","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Ge, Zerui","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2605.24490","locator":"head/meta[name=author|citation_author]"},"assertionId":"ad8ae4f3f53d070702739ac1dc423932348fc2b4e76a09167f24411e1b8a78e2:participation:1","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-12T01:56:07.942Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"ad8ae4f3f53d070702739ac1dc423932348fc2b4e76a09167f24411e1b8a78e2","sourceUrl":"https://arxiv.org/abs/2605.24490","contentSha256":"d11bd5e22f56e4a4decd2ba4b26ce6553fb6c42cfe5cd393a8892685f4e581d0","textHash":"df6e2958ca538b5cf7ce859be50ade8559cd4fd1c8f1720b3d0ac4dca2fcc6b3","observedAt":"2026-11-12T01:56:07.942Z"}},{"actor":{"name":"Zheng, Jin","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Zheng, Jin","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2605.24490","locator":"head/meta[name=author|citation_author]"},"assertionId":"ad8ae4f3f53d070702739ac1dc423932348fc2b4e76a09167f24411e1b8a78e2:participation:2","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-12T01:56:07.942Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"ad8ae4f3f53d070702739ac1dc423932348fc2b4e76a09167f24411e1b8a78e2","sourceUrl":"https://arxiv.org/abs/2605.24490","contentSha256":"d11bd5e22f56e4a4decd2ba4b26ce6553fb6c42cfe5cd393a8892685f4e581d0","textHash":"df6e2958ca538b5cf7ce859be50ade8559cd4fd1c8f1720b3d0ac4dca2fcc6b3","observedAt":"2026-11-12T01:56:07.942Z"}},{"actor":{"name":"Cartlidge, John","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Cartlidge, John","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2605.24490","locator":"head/meta[name=author|citation_author]"},"assertionId":"ad8ae4f3f53d070702739ac1dc423932348fc2b4e76a09167f24411e1b8a78e2:participation:3","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-12T01:56:07.942Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"ad8ae4f3f53d070702739ac1dc423932348fc2b4e76a09167f24411e1b8a78e2","sourceUrl":"https://arxiv.org/abs/2605.24490","contentSha256":"d11bd5e22f56e4a4decd2ba4b26ce6553fb6c42cfe5cd393a8892685f4e581d0","textHash":"df6e2958ca538b5cf7ce859be50ade8559cd4fd1c8f1720b3d0ac4dca2fcc6b3","observedAt":"2026-11-12T01:56:07.942Z"}}],"participationCoverage":{"total":4,"returned":4,"complete":true,"fullAssertions":"source-metadata-endpoint"},"distribution":null,"publisher":null,"sourceUrl":"https://arxiv.org/abs/2605.24490","snapshotId":"ad8ae4f3f53d070702739ac1dc423932348fc2b4e76a09167f24411e1b8a78e2","contentSha256":"d11bd5e22f56e4a4decd2ba4b26ce6553fb6c42cfe5cd393a8892685f4e581d0","textHash":"df6e2958ca538b5cf7ce859be50ade8559cd4fd1c8f1720b3d0ac4dca2fcc6b3","observedAt":"2026-11-12T01:56:07.942Z","basis":"publisher-page-metadata","sourceLinks":47,"linksTruncated":false,"metadataUrl":"/api/pipeline/stream/5622ffa79dbb7fdc2632983f38c8f724c7bf31e3ba3aa0ecce9d8ea053739c8f/source"}}}],"audits":[{"publicationId":"c40fe5bb-3ab1-49ed-8e9b-e2b7732933b1","verdict":"uncertain","findings":[{"dimension":"attribution","status":"pass","explanation":"The reading credits the paper and its findings to the four listed authors, and frames results as author reports ('the authors report', 'they say'). It does not present them as established fact.","excerpt":"Authors:Yunhua Pei, Zerui Ge, Jin Zheng, John Cartlidge"},{"dimension":"summary","status":"uncertain","explanation":"The reading is faithful but hard for a general reader to follow. It drops that the agents are LLM agents; the title and summary say only 'trading agents' and 'multi-agent system'. Repair: say 'multi-agent LLM'. CR, SR, IR and MDD are left as bare abbreviations. Sharpe ratio, Shapley credits, cold-start dominance, regime shifts and seeds are not explained in plain language. Smallest repair: add 'LLM', spell out cumulative return, Sharpe ratio, information ratio and maximum drawdown, and add tooltips for the remaining terms. The question is a fair open issue.","excerpt":"Multi-agent LLM decision systems for portfolio management still lack a principled way to assign credit across specialist agents"},{"dimension":"date","status":"pass","explanation":"originalPublished is 2026-05-23, which matches the v1 submission date. It is not the 10 Nov 2026 v2 revision or the November feed date.","excerpt":"[Submitted on 23 May 2026 (v1), last revised 10 Nov 2026 (this version, v2)]"},{"dimension":"support","status":"pass","explanation":"Every claim matches the abstract: the mechanism, the Shapley weighting, the Bayesian mixture, the regime multipliers, the five-layer trace, the test scope, the metrics, the rankings and the ablation attribution. The limitations stay within the abstract and add no unsupported benefits or costs. The draft does not turn the backtest result into real-world reliability.","excerpt":"Ablation results show that the gains come from Shapley-weighted integration across coalition outputs rather than from any single stage in isolation."},{"dimension":"media","status":"pass","explanation":"The selected media is a verified, direct link to the arXiv abstract page of this paper. It is the original source.","excerpt":null}],"createdAt":"2026-11-11T23:48:32.358Z","model":"anthropic/claude-opus-5.5"}],"correctionUrl":"https://escapemeridian.com/corrections/"}