{"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__C","search_lanes":{"direct_problem":{"queries":["credit loss management overlays lifecycle expiry sunset review governance IFRS 9","post model adjustments management overlays credit impairment periodic review removal documentation","loan loss provision overlays stale double counting management adjustments audit"],"source_ids":["SRC1","SRC2","SRC3","SRC6","SRC8"],"no_result_note":null},"closest_prior_art":{"queries":["IFRS 9 post-model adjustments governance framework rationale monitoring exit criteria audit trail owner approval","management overlay governance temporary exit criteria model limitation expected credit loss PwC EY Deloitte","\"post-model adjustments\" \"exit criteria\"","\"PMAs will be removed when the models are updated\""],"source_ids":["SRC3","SRC4","SRC5","SRC7"],"no_result_note":null},"historical_terminology":{"queries":["CECL qualitative adjustments overlay governance review expiration model risk","loan loss provision overlays stale double counting management adjustments audit","\"management adjustment overlay\" loan loss documentation"],"source_ids":["SRC5","SRC6"],"no_result_note":null},"products_practices_standards":{"queries":["CECL qualitative adjustment governance inventory tracking approval removal criteria documentation regulator","site:sc.com annual report \"PMA\" \"removed\" ECL model updated","bank annual report management overlay \"will be removed\" model"],"source_ids":["SRC6","SRC7"],"no_result_note":null},"non_english_regional":{"queries":["ajustes post modelo superposiciones IFRS 9 revisión vencimiento gestión","Management Overlays IFRS 9 Befristung Überprüfung Auflösung"],"source_ids":["SRC8"],"no_result_note":"The Spanish IASB document provided directly relevant regional terminology and concerns; the German query produced no comparably probative retained source."},"composition_subproblems":{"queries":["IFRS 9 post-model adjustments governance framework rationale monitoring exit criteria audit trail owner approval","management overlay governance temporary exit criteria model limitation expected credit loss PwC EY Deloitte","\"overlay\" \"sunset criteria\" credit loss","\"overlay register\" IFRS 9 management"],"source_ids":["SRC3","SRC4","SRC5","SRC6","SRC7"],"no_result_note":null}},"sources":[{"source_id":"SRC1","title":"IFRS 9 Implementation by EU Institutions: 2023 Monitoring Report","url":"https://eba.europa.eu/sites/default/files/2023-11/25b12d35-9c28-4335-a589-166c77198920/Final%20Report%20on%20IFRS9%20implementation%20by%20EU%20institutions.pdf","publisher":"European Banking Authority","date_or_year":"November 2023","source_type":"PRIMARY_RESEARCH","language":"English","claims_supported":["In a 37-institution EU sample, post-model overlays remained widely used and were often described as temporary.","Institutions retained COVID-19 overlays and replaced or complemented them with adjustments for later geopolitical, inflation, interest-rate, ESG, and data risks.","Some institutions had weakly defined rationales, overlays exceeded 50% of portfolio ECL in some cases, and overlays were often not backtested.","The EBA calls for structured methodology, strict governance, internal controls, periodic review, and backtesting of overlays."]},{"source_id":"SRC2","title":"Thematic Feedback on Accounting for IFRS 9 ECL and Climate Risk","url":"https://www.bankofengland.co.uk/-/media/boe/files/prudential-regulation/letter/2024/thematic-feedback-on-accounting-for-ifrs-9-ecl-and-climate-risk.pdf","publisher":"Bank of England, Prudential Regulation Authority","date_or_year":"27 September 2024","source_type":"OFFICIAL_GUIDANCE","language":"English","claims_supported":["The PRA found continuing variation in PMA practice and emphasizes challenging PMA completeness, model-redevelopment progress, monitoring, and governance.","Major UK firms continue to rely on PMAs for risks and model limitations, while model redevelopment has reduced some PMAs.","Senior cross-functional committees, formal operating boundaries, monitoring, and remediation plans are identifiable governance mechanisms."]},{"source_id":"SRC3","title":"Post-model Adjustments for Expected Credit Losses During COVID-19","url":"https://www.pwc.com.au/assurance/ifrs/assets/post-model-adjustments-for-expected-credit-losses-during-covid19.pdf","publisher":"PwC Australia","date_or_year":"May 2020","source_type":"TRADE_PROFESSIONAL","language":"English","claims_supported":["Overlay governance should document the limitation, rationale, assumptions, data, affected exposures, reviewers, and how the overlay will be consumed over time.","Overlay performance can be assessed through backtesting and KPI monitoring.","End-to-end review should test for double counting across overlays and modeled adjustments.","Absent documented rationale, determining whether an overlay remains necessary in later periods becomes more difficult."]},{"source_id":"SRC4","title":"Entering a “New Normal” for IFRS 9 ECL Measurement","url":"https://www.deloitte.com/content/dam/assets-zone2/cy/en/docs/services/audit-assurance/2024/CY_Audit_IFRS-9_ECL_measurement_in_the%20Covid19_era_onwards_NoExp.pdf","publisher":"Deloitte IFRS 9 Working Group","date_or_year":"March 2021","source_type":"TRADE_PROFESSIONAL","language":"English","claims_supported":["Leading banks link each overlay to a limitation or reason, independently challenge it, and establish criteria for reduction or change.","Leading banks actively manage PMA portfolios by regularly reviewing scope, duration, and unwinding conditions.","Monitoring, analysis, documentation, and eventual model remediation are presented as central elements of the ECL control framework."]},{"source_id":"SRC5","title":"Policies and Internal Controls—Allowance for Credit Losses","url":"https://publishedguides.ncua.gov/examiner/content/examinersguide/AllowanceCreditLoss/RiskManagement/PoliciesInternalControls.htm","publisher":"National Credit Union Administration","date_or_year":"5 June 2023","source_type":"OFFICIAL_GUIDANCE","language":"English","claims_supported":["CECL policies should identify roles, periodic evaluation, qualitative adjustments, controls, and documentation.","Controls should verify each reporting period that inappropriate layering—more than one amount for the same probable loss—has not occurred.","Board reporting, independent validation, and documentation of decisions, assumptions, inputs, and results are prudent practices."]},{"source_id":"SRC6","title":"Staff Accounting Bulletin No. 102: Selected Loan Loss Allowance Methodology and Documentation Issues","url":"https://www.sec.gov/interps/account/sab102.htm","publisher":"U.S. Securities and Exchange Commission","date_or_year":"2001","source_type":"OFFICIAL_GUIDANCE","language":"English","claims_supported":["Older U.S. allowance terminology already required periodic methodology review, written support, defined responsibilities, approval, validation, and documentation of management adjustments.","A documented bridge from portfolio review to the reported allowance preserves auditability and provides a benchmark for later judgment.","Frequent management or credit-committee adjustments should prompt analysis of their causes and reassessment of the methodology."]},{"source_id":"SRC7","title":"Standard Chartered Annual Report 2024—Risk Review and Capital Review","url":"https://www.sc.com/en/uploads/sites/66/content/docs/annual-report-2024-risk-review-and-capital-review.pdf","publisher":"Standard Chartered PLC","date_or_year":"2025 reporting on 2024","source_type":"FIRST_PARTY_PRODUCT","language":"English","claims_supported":["An identifiable bank has formal standards for PMA identification, calculation, review, approval, removal, and remediation tracking.","Model-performance PMAs are removed when estimates return within monitoring or validation thresholds.","Judgmental adjustments require committee approval, quarterly review and reapproval, and release when the covered risk is no longer relevant.","Independent validation and named senior risk and finance committees provide identifiable adopters and authorizers."]},{"source_id":"SRC8","title":"Solicitud de Información: Revisión Posterior a la Implementación de la NIIF 9—Deterioro de Valor","url":"https://www.ifrs.org/content/dam/ifrs/project/pir-9-impairment/es-rfi-iasb-2023-1-ifrs9-impairment.pdf","publisher":"IFRS Foundation, International Accounting Standards Board","date_or_year":"May 2023","source_type":"OFFICIAL_GUIDANCE","language":"Spanish","claims_supported":["Spanish-language IFRS materials use términos such as ajustes posteriores al modelo, superposiciones or solapamientos de gestión, and anulaciones de modelos.","Stakeholders reported increasing use of these adjustments, subjective judgment, inconsistent governance, significant cross-entity variation, and inadequate entity-specific disclosure.","The regional terminology search confirms the issue is not confined to the English phrase management overlay."]}],"problem_evidence":{"status":"SUPPORTED","finding":"The accumulation-governance problem exists, although age alone is not evidence of staleness. The EBA found retained COVID overlays being replaced or complemented by adjustments for successive shocks, weakly defined risk linkages in some institutions, material overlay shares, and scarce overlay backtesting. PwC identifies double-counting and later-period necessity as explicit control questions, while NCUA defines duplicate allowance layering as an inappropriate practice. Official IFRS feedback also records governance and comparability concerns.","source_ids":["SRC1","SRC3","SRC5","SRC8"],"uncertainty":"Public sources do not quantify how often active overlays specifically lack an owner, expiry condition, supersession marker, or reconstruction trail. The evidence establishes a material governance risk and documented weak practices, not the prevalence of every proposed failure field."},"adopter_evidence":{"status":"SUPPORTED","finding":"Credit-risk and finance functions, independent model validation, impairment committees, model assessment committees, boards, auditors, and prudential supervisors are identifiable adopters or authorizers. Standard Chartered publicly identifies CMAC, GMV, and its IFRS 9 Impairment Committee as responsible for PMA review, approval, tracking, reapproval, and release.","source_ids":["SRC2","SRC5","SRC6","SRC7"],"uncertainty":"Committee names and authority allocations vary by institution and jurisdiction; implementation would require mapping the lifecycle tool to each institution's accounting-control and model-risk framework."},"implementation_evidence":{"status":"SUPPORTED","finding":"Substantial elements are demonstrably implementable and already practiced: documented rationale and exposure mapping, independent challenge, periodic reapproval, threshold- or event-based release, duration and unwind review, duplicate-layer checks, performance monitoring, remediation tracking, and retained supporting documentation. A shadow inventory and reviewer comparison can be run without changing booked reserves.","source_ids":["SRC3","SRC4","SRC5","SRC7"],"uncertainty":"The retained sources do not demonstrate routine use of the proposal's full state-machine implementation, especially explicit supersession links, orphan detection, archival reconstruction packages, preservation exceptions, or controlled rollback windows."},"prior_art":{"disposition":"ESTABLISHED_PRACTICE","closest_analogues":[{"name":"Leading-bank PMA portfolio lifecycle governance","source_ids":["SRC4"],"same_problem":true,"same_causal_lever":true,"overlap":"Links each overlay to a limitation, requires independent challenge and change criteria, and regularly reviews PMA scope, duration, and unwind conditions while models are remediated.","remaining_difference":"The proposal makes lifecycle states and artifact-level fields more explicit and adds supersession, dependency, reconstruction, preservation-exception, archival, and restore controls not described in the source."},{"name":"Standard Chartered PMA and judgmental-adjustment governance","source_ids":["SRC7"],"same_problem":true,"same_causal_lever":true,"overlap":"Uses prescribed identification, calculation, approval, monitoring, remediation, quarterly reapproval, threshold-based removal, and event-resolution release processes under named committees.","remaining_difference":"The public disclosure does not establish an overlay-level identity map, owner and age index, explicit supersession graph, orphan detector, reconstruction archive, or rollback procedure."},{"name":"PwC overlay establishment and monitoring control questions","source_ids":["SRC3"],"same_problem":true,"same_causal_lever":true,"overlap":"Requires documented rationale, affected exposures, reviewers, consumption path, performance monitoring, double-counting review, and continuing-necessity assessment.","remaining_difference":"It is control guidance rather than a formal lifecycle-state register with archival and restoration semantics."},{"name":"U.S. allowance adjustment and anti-layering controls","source_ids":["SRC5","SRC6"],"same_problem":true,"same_causal_lever":false,"overlap":"Requires periodic review, documented adjustments and rationale, defined responsibility, validation, approval, and checks against duplicate loss layering.","remaining_difference":"These controls govern the allowance process generally and do not prescribe explicit lifecycle states for individual post-model overlays."}],"contrastive_claim_remaining":"Relative to an institution's existing policy- and committee-based PMA governance, an artifact-level register with explicit active, review-pending, and archived states plus owner, dates, supersession, dependencies, reconstruction package, and controlled restoration will reduce stale or duplicative active overlays, reviewer disagreement, and reconstruction time without causing reviewers to omit necessary uncertainty coverage.","contrastive_claim_falsifier":"The claim is falsified if representative institutions already maintain materially equivalent artifact-level lifecycle and reconstruction controls, or if a blinded shadow comparison shows no meaningful reduction in stale selections, duplication, disagreement, or reconstruction time—or shows more missed risk coverage—relative to existing governance.","confidence":"HIGH","search_limitations":"The bounded search covered six lanes and eight opened sources across regulators, standard-setting materials, professional practice, and a bank disclosure. It was English-heavy apart from a Spanish IFRS source; internal bank policies, vendor implementations, patents, paywalled material, and unpublished audit findings were unavailable. The search cannot establish world novelty, patentability, freedom to operate, market size, or realized impact."},"researchability_gates":{"externally_supported_problem":{"status":"PASS","rationale":"Regulatory benchmarking, supervisory feedback, professional guidance, and CECL controls support persistent overlay governance, backtesting, documentation, and duplicate-layer risks.","source_ids":["SRC1","SRC2","SRC3","SRC5","SRC8"]},"identifiable_adopter_or_authorizer":{"status":"PASS","rationale":"Banks' risk and finance functions and named validation and impairment committees are identifiable; Standard Chartered discloses specific authorizing and reviewing bodies.","source_ids":["SRC5","SRC6","SRC7"]},"distinct_testable_incremental_claim":{"status":"PASS","rationale":"Although lifecycle governance is established practice, the incremental claim about an artifact-level state register with supersession and reconstructable archival controls is distinct, measurable, and falsifiable against current governance.","source_ids":["SRC3","SRC4","SRC7"]},"bounded_next_evidence_step":{"status":"PASS","rationale":"One portfolio and one reporting cycle can be tested in shadow mode using historical or current overlays, blinded reviewers, stale-selection outcomes, disagreement rates, duplication findings, and reconstruction time.","source_ids":["SRC3","SRC5","SRC7"]},"no_unresolved_safety_or_authority_stop":{"status":"PASS","rationale":"A read-only shadow pilot leaves booked reserves and committee authority unchanged. Premature release risk is controlled by prohibiting automatic removal, preserving evidence, and halting on missing dependencies or unexplained material effects.","source_ids":["SRC2","SRC6","SRC7"]},"adequate_search_evidence":{"status":"PASS","rationale":"All six required lanes were searched adversarially, including older ALLL terminology, CECL controls, bank practice, Spanish IFRS terminology, lifecycle combinations, and duplicate-layer and reconstruction subproblems. Eight direct sources from seven publisher organizations were opened, with multiple official and first-party sources.","source_ids":["SRC1","SRC2","SRC3","SRC4","SRC5","SRC6","SRC7","SRC8"]}},"strict_success":false,"screen_survival":false,"remaining_research_value":"MODERATE","recommended_next_step":"Run the authorized shadow pilot on one portfolio for one reporting cycle. Pre-register lifecycle fields and stale-overlay criteria; have reviewers independently assess overlays first under the current process and then with the lifecycle register; measure stale or duplicate selections, disagreement, missed uncertainty coverage, and time to reconstruct prior-period estimates. Do not change booked reserves or destroy evidence.","world_novelty_boundary":"The search shows that the core problem-intervention package—actively governed PMA portfolios with documented rationale, periodic challenge, release criteria, tracking, and model remediation—is established practice. The potentially researchable boundary is the narrower artifact-level state, supersession, dependency, reconstruction, archival, and controlled-restore implementation and its incremental effect over strong existing governance. No claim of world novelty, patentability, freedom to operate, market size, or realized impact is made."}