{"schema_version":1,"experiment_id":"eoa_inverse_innovation_exp11_mechanism_context_external20_20260804","research_id":"eoa_inverse_innovation_exp11_external_scrutiny_20260804","cell_id":"layer_decay_and_expiration_management__economics_finance","opaque_id":"layer_decay_and_expiration_management__economics_finance__A","search_lanes":{"direct_problem":{"queries":["credit loss management overlays stale accumulation double counting risk sunset review","post model adjustments overlays expected credit losses governance removal review","IFRS 9 management overlays expiry revalidation audit trail","COVID overlays banks release retained stale ECL overlays"],"source_ids":["SRC2","SRC5","SRC6"],"no_result_note":null},"closest_prior_art":{"queries":["site:bankofengland.co.uk post model adjustment framework monitoring review removal trigger PMA inventory","site:eba.europa.eu IFRS 9 overlays governance temporary nature monitoring removal","post model adjustment inventory lifecycle system bank overlay governance tool","PMA framework trigger unwind review frequency overlays bank"],"source_ids":["SRC1","SRC4","SRC6"],"no_result_note":null},"historical_terminology":{"queries":["credit loss expert credit judgment qualitative adjustment Q factors sunset governance CECL","model overlay capital planning governance track analyze overlay performance","IFRS 9 management adjustment post-model adjustment top level adjustment"],"source_ids":["SRC3","SRC4","SRC5"],"no_result_note":null},"products_practices_standards":{"queries":["IFRS 9 overlay management software workflow approval audit trail post model adjustment","management overlay register expiry date owner IFRS 9","PRA SS1/23 PDF Principle 5.1 post model adjustments reduced removed root cause analysis","IFRS 9 Software overlays documentation sign-off audit trail"],"source_ids":["SRC1","SRC4","SRC7","SRC8"],"no_result_note":null},"non_english_regional":{"queries":["IFRS 9 \"surcouches\" management overlays provisions gouvernance retrait","IFRS 9 \"ajustes post-modelo\" overlays gestión revisión eliminación","IFRS 9 \"Management Overlays\" Abbau Überprüfung Wertberichtigungen","ECL 覆盖 管理层调整 IFRS 9 叠加 治理"],"source_ids":["SRC3","SRC7"],"no_result_note":null},"composition_subproblems":{"queries":["post model adjustment multiple PMAs duplication hierarchy overlapping overlays","PMA inventory owner review date removal criteria audit trail","overlay archive reconstruction reversible retirement expected credit loss","model adjustment dependencies decommissioning audit history"],"source_ids":["SRC1","SRC4","SRC6","SRC8"],"no_result_note":null}},"sources":[{"source_id":"SRC1","title":"Model risk management principles for banks — Supervisory Statement SS1/23","url":"https://www.bankofengland.co.uk/-/media/boe/files/prudential-regulation/supervisory-statement/2026/liaf0126app5.pdf","publisher":"Bank of England, Prudential Regulation Authority","date_or_year":"2026 update of 2023 statement","source_type":"OFFICIAL_GUIDANCE","language":"English","claims_supported":["In-scope banks must maintain model inventories, governance details, validation dates, limitations, dependencies, and decommissioned-model information.","Principle 5.1 requires a firm-wide PMA process covering completeness, calculation, continued relevance, independent review, reduction or removal criteria, prolonged-use triggers, recurring-PMA trends, remediation, documentation, and approval authority.","The guidance substantially anticipates recurring lifecycle review and retirement of overlays, although it does not prescribe vintage-level archival, reversible deactivation, or restoration testing."]},{"source_id":"SRC2","title":"IFRS 9 2023 Monitoring Report","url":"https://service.betterregulation.com/sites/default/files/2023-11/Final%20Report%20on%20IFRS9%20implementation%20by%20EU%20institutions.pdf","publisher":"European Banking Authority","date_or_year":"2023","source_type":"PRIMARY_RESEARCH","language":"English","claims_supported":["Most sampled institutions implemented overlays, and almost all continued using them at the end of 2021.","Institutions generally retained COVID-19 overlays introduced in 2020, while replacing or complementing them with adjustments for later risks.","Some institutions had discrepancies between overlay justifications and well-defined risk factors; overlay shares exceeded 50% of ECL for some portfolios.","The EBA calls for scrutiny of overlay purpose, calibration, temporary nature, model incorporation, governance, and backtesting."]},{"source_id":"SRC3","title":"Overlays and in-model adjustments: identifying best practices for capturing novel risks","url":"https://www.bankingsupervision.europa.eu/press/blog/2023/html/ssm.blog230526~29af0452d6.sl.html","publisher":"European Central Bank Banking Supervision","date_or_year":"2023","source_type":"OFFICIAL_GUIDANCE","language":"English content on Slovenian regional interface","claims_supported":["An ECB targeted review covered 51 supervised banks and found widespread use of overlays for novel credit risks.","The ECB found weak umbrella and total-ECL overlays lacking granularity and risk sensitivity and linked provisioning practice to capital, profitability, pricing, and resilience.","The source documents historical synonyms including management adjustment, post-model adjustment, and top-level adjustment."]},{"source_id":"SRC4","title":"Federal Reserve Guidance on Supervisory Assessment of Capital Planning and Positions for LISCC Firms and Large and Complex Firms","url":"https://www.federalreserve.gov/frrs/guidance/federal-reserve-guidance-on-supervisory-assessment-of-capital-planning-and-positions-for-liscc-firms.htm","publisher":"Board of Governors of the Federal Reserve System","date_or_year":"2015","source_type":"OFFICIAL_GUIDANCE","language":"English","claims_supported":["Banks should use a consistent firm-wide overlay process, document rationale and assumptions, report results with and without overlays, and subject overlays to independent challenge and materiality-based approval.","Firms should track and analyze overlay performance, reduce reliance by fixing underlying models, monitor aggregate overlay use, and treat persistent overrides as evidence for redevelopment.","This older capital-planning practice is close prior art for recurring overlay governance but is not specific to IFRS 9 overlay-vintage retirement and archival."]},{"source_id":"SRC5","title":"Feedback analysis—Measuring expected credit losses, Agenda Paper 27A","url":"https://www.ifrs.org/content/dam/ifrs/meetings/2024/march/iasb/ap27a-feedback-analysis-measuring-ecl.pdf","publisher":"IFRS Foundation / International Accounting Standards Board staff","date_or_year":"2024","source_type":"OFFICIAL_GUIDANCE","language":"English","claims_supported":["Stakeholders reported that PMAs had increased, were large and subjective, and reduced comparability and transparency.","Respondents reported that entities repurpose PMAs rather than release them and that release or reversal timing is a major challenge.","The paper records weaker umbrella-overlay practices and concludes that PMA use and reversal must remain consistent with IFRS 9 ECL objectives."]},{"source_id":"SRC6","title":"PMA implementation: Don't let overlays become oversights","url":"https://www.deloitte.com/uk/en/services/consulting-risk/blogs/2025/pma-implementation-dont-let-overlays-become-oversights.html","publisher":"Deloitte UK","date_or_year":"2025","source_type":"TRADE_PROFESSIONAL","language":"English","claims_supported":["A practical PMA lifecycle covers identification, quantification, implementation, evidence, root-cause analysis, granularity, sensitivity analysis, and governance.","The source explicitly warns that multiple PMAs can apply to one account and calls for responsibility and hierarchy to assess duplicated impacts.","It identifies Heads of Impairment and Risk Managers as operational adopters."]},{"source_id":"SRC7","title":"IFRS aktuell, Ausgabe Dezember 2021","url":"https://www.pwc.at/de/newsletter/ifrs/2021/ifrs-aktuell-12-2021.pdf","publisher":"PwC Austria","date_or_year":"2021","source_type":"TRADE_PROFESSIONAL","language":"German","claims_supported":["Austrian reporting guidance required transparent disclosure of material management-overlay amounts, allowance effects, rationales, and methodologies.","It also warned against double counting in impairment estimates and required specific information about material ECL adjustments.","This regional evidence shows that overlay traceability and double-counting concerns predate the recent lifecycle proposal."]},{"source_id":"SRC8","title":"Estimator 9 — IFRS 9 Expected Credit Loss Software","url":"https://contracthive.ai/products/estimator9","publisher":"ContractHive / FutureHives","date_or_year":"2026","source_type":"FIRST_PARTY_PRODUCT","language":"English","claims_supported":["A marketed IFRS 9 product already supports overlays with documentation, sign-off, audit trails, lineage, backtesting, validation, and reproducible reruns.","The product identifies credit-risk managers, finance leads, CROs, auditors, lenders, and insurers as users or adopters.","The product does not publicly claim overlay expiry states, dependency-safe retirement, reversible deactivation, preservation holds, or archive restoration tests."]}],"problem_evidence":{"status":"SUPPORTED","finding":"The stale-overlay problem exists. IASB feedback records that entities repurpose PMAs rather than release them and struggle with reversal timing. EBA monitoring found COVID-era overlays retained and later replaced or complemented, sometimes with weak links to defined risks and material portfolio impacts. Deloitte separately identifies duplicated impacts when multiple PMAs affect the same account. These sources support persistence, overlap, opacity, and materiality, although they do not quantify the net provision bias caused specifically by stale accumulation.","source_ids":["SRC2","SRC5","SRC6","SRC7"],"uncertainty":"Public evidence does not establish how often retained overlays are truly stale rather than still justified, nor the frequency or magnitude of double-counted provisions, capital misallocation, review burden, or reserve cliffs across representative lenders."},"adopter_evidence":{"status":"SUPPORTED","finding":"Identifiable adopters and authorizers exist: bank credit-risk and finance teams operate PMAs; model owners, validators, Heads of Impairment, risk managers, CROs, audit committees, risk committees, senior management, and other appropriate authorities review or approve them. PRA and Federal Reserve guidance place responsibility within regulated firms and require authority proportionate to materiality.","source_ids":["SRC1","SRC4","SRC6","SRC8"],"uncertainty":"Exact committee mandates and records-retention authority vary by jurisdiction and institution; the named product customers do not prove adoption of the full proposed lifecycle."},"implementation_evidence":{"status":"PARTLY_SUPPORTED","finding":"Most core governance elements are feasible and already prescribed or sold: inventory, ownership, rationale, materiality, independent review, recurring-use triggers, reduction/removal criteria, before-and-after reporting, sign-off, lineage, audit trail, backtesting, and duplication checks. No retained source demonstrates the complete proposed package of overlay-vintage identities, age-adjusted evidence scoring, dependency and hold gates, reversible deactivation, successor-linked archival, and periodic restoration tests in routine ECL production.","source_ids":["SRC1","SRC4","SRC6","SRC8"],"uncertainty":"Vendor claims were not independently verified, and no public implementation study measured operational cost, reviewer agreement, false-positive stale flags, restoration success, or provision volatility for the full package."},"prior_art":{"disposition":"ADJACENT_PRIOR_ART","closest_analogues":[{"name":"PRA SS1/23 Principle 5.1 PMA governance and removal process","source_ids":["SRC1"],"same_problem":true,"same_causal_lever":true,"overlap":"Requires a firm-wide PMA process, clear rationale, materiality-based authority, transparency, independent review of continued relevance, calculation and completeness controls, explicit reduction/removal criteria, prolonged-use triggers, and remediation for recurring PMAs.","remaining_difference":"It does not prescribe a separate vintage-level lifecycle register, age-adjusted review ranking, overlay-to-report dependency gates, legal-hold exceptions, reversible shadow retirement, successor-linked archives, or restoration tests."},{"name":"Federal Reserve model-overlay governance","source_ids":["SRC4"],"same_problem":false,"same_causal_lever":true,"overlap":"Requires firm-wide overlay controls, documentation, pre-use challenge, performance tracking, aggregate reporting, sensitivity analysis, and reduction of recurring overlay dependence through model remediation.","remaining_difference":"It addresses capital-planning overlays broadly rather than stale IFRS 9 ECL vintages and does not define retirement, preservation, archival, rollback, or restoration mechanics."},{"name":"Deloitte PMA lifecycle and multiple-PMA duplication control","source_ids":["SRC6"],"same_problem":true,"same_causal_lever":true,"overlap":"Describes a PMA lifecycle and explicitly requires responsibility and hierarchy for detecting duplicated impacts from multiple PMAs, alongside evidence, root-cause, granularity, and implementation controls.","remaining_difference":"It does not publicly specify recurring expiry states, a recoverable history tier, dependency-safe retirement, preservation exceptions, reversible deactivation, or restoration testing."},{"name":"Estimator 9 governed overlay workflow","source_ids":["SRC8"],"same_problem":false,"same_causal_lever":false,"overlap":"Provides documentation, sign-off, audit trails, lineage, validation, backtesting, and reproducible reruns for overlays inside an ECL platform.","remaining_difference":"Its public materials do not claim staleness detection or managed retirement and archival of overlay vintages."}],"contrastive_claim_remaining":"Relative to ordinary PMA governance or a one-time rationalization, adding an explicit overlay-vintage lifecycle—age-and-evidence review prioritization plus dependency/hold-gated reversible deactivation, successor-linked archival, and restoration testing—will improve stale-overlay detection, traceability, reconstruction success, and reviewer agreement without causing unacceptable simulated provision volatility.","contrastive_claim_falsifier":"The claim is falsified if the comparator already performs these vintage-level controls, or if a controlled shadow pilot shows no improvement in stale-overlay detection, traceability, reconstruction, or reviewer agreement, or breaches a predeclared provision-volatility or control-exception bound.","confidence":"HIGH","search_limitations":"This bounded eight-source search was strongest for UK, EU, and US supervisory material. Non-English coverage retained one German/Austrian source and a Slovenian ECB interface; Spanish, French, and Chinese searches yielded no stronger direct source. Public sources reveal expectations and selected practices, not confidential bank workflows. Product claims were not audited. The search cannot establish world novelty, patentability, freedom to operate, market size, or realized impact."},"researchability_gates":{"externally_supported_problem":{"status":"PASS","rationale":"IASB feedback directly reports PMAs being repurposed rather than released and difficulty timing reversals; EBA monitoring and practitioner evidence support persistence, weak rationales, materiality, and possible overlap.","source_ids":["SRC2","SRC5","SRC6"]},"identifiable_adopter_or_authorizer":{"status":"PASS","rationale":"Regulated banks, credit-risk and finance teams, model owners, validators, senior management, risk and audit committees, Heads of Impairment, and CROs are identifiable operational users or authorizers.","source_ids":["SRC1","SRC4","SRC6","SRC8"]},"distinct_testable_incremental_claim":{"status":"PASS","rationale":"Existing guidance substantially covers PMA governance and removal, but the combination of explicit vintages, prioritized staleness review, dependency and hold gates, reversible retirement, recoverable archives, and restoration tests remains distinguishable and measurable.","source_ids":["SRC1","SRC4","SRC6","SRC8"]},"bounded_next_evidence_step":{"status":"PASS","rationale":"A one-portfolio, one-reporting-cycle shadow pilot can inventory overlays, apply lifecycle flags, simulate retirement, test restoration, compare reviewer decisions, and measure traceability and provision movement without changing booked ECL.","source_ids":["SRC1","SRC6","SRC8"]},"no_unresolved_safety_or_authority_stop":{"status":"PASS","rationale":"The evidence supports materiality-based approval, independent challenge, transparent before-and-after reporting, and retention of audit evidence. A shadow-only pilot with no booked reserve or customer-credit changes avoids an immediate accounting or customer-safety stop, provided committee, model-risk, legal, and records-control approvals are obtained.","source_ids":["SRC1","SRC4","SRC7"]},"adequate_search_evidence":{"status":"PASS","rationale":"All six lanes were searched adversarially using current and historical terms, official standards and supervisory practices, product implementations, regional language, and component combinations. Exactly eight opened direct sources span seven publishers, including multiple official and first-party sources.","source_ids":["SRC1","SRC2","SRC3","SRC4","SRC5","SRC6","SRC7","SRC8"]}},"strict_success":true,"screen_survival":true,"remaining_research_value":"MODERATE","recommended_next_step":"Run the authorized shadow pilot on one portfolio and one reporting cycle. Pre-register the comparator, stale-overlay criteria, materiality bound, dependency and hold checks, reviewer-agreement metric, traceability completion rate, restoration-success test, false-positive review rate, and simulated provision-volatility halt threshold. Do not alter booked provisions or destroy evidence.","world_novelty_boundary":"The search establishes close supervisory, professional, and product prior art and leaves only a narrower, falsifiable operational increment. It does not establish world novelty, patentability, freedom to operate, market size, or realized impact."}