{"schema_version":1,"experiment_id":"eoa_inverse_innovation_exp04_retrieval_first_paired20_20260802","cell_id":"invariant_mode_decomposition_design__economics_finance","arm":"RETRIEVAL_FIRST","round_index":0,"hypotheses":[{"hypothesis_id":"H1","title":"Modal intraday-liquidity buffers","problem":"Bank liquidity dashboards can look adequate while a coupled withdrawal-and-collateral mode approaches instability.","affected_stakeholder":"Bank treasury teams and uninsured depositors","workflow_boundary":"Intraday monitoring through liquidity-buffer allocation","failure_mode":"Aggregate ratios conceal an amplifying combination of depositor outflows, collateral haircuts, and payment obligations.","unit_of_analysis":"Bank-hour state vector","causal_lever":"Allocate contingent liquidity to exposures loading most heavily on locally unstable modes, subject to residual alarms.","archetype_mapping":"Estimate the balance-sheet transition operator, classify growing modes, and intervene on their constituent exposures rather than on isolated ratios.","expected_value":"Potential earlier run detection and less idle liquidity for the same risk tolerance; prevalence and magnitude are unknown.","falsifiable_claim":"In retrospective stress episodes, a modal policy triggers at least one business day earlier than standard ratio thresholds without increasing false alarms by more than 10%.","diversity_rationale":"Targets dynamic stability in an intraday control loop, unlike the later hypotheses' contract assembly, underwriting, policy response, and liquidation workflows.","mechanism_slugs":["eigendecomposition_workflow","modal_stability_analysis","modal_sensitivity_sweep","spectral_gap_monitor"],"search_questions":["Have banks tested eigenmode-based intraday liquidity allocation rather than only systemic-risk description?","Do modal instability signals add lead time beyond liquidity coverage and concentration ratios?"]},{"hypothesis_id":"H2","title":"Spectral consent sequencing for restructurings","problem":"Viable debt restructurings can stall because fragmented veto rights create an anti-commons.","affected_stakeholder":"Distressed borrowers and restructuring coordinators","workflow_boundary":"Creditor mapping through consent solicitation","failure_mode":"Teams negotiate by claim size while small but structurally central veto holders sustain deadlock.","unit_of_analysis":"Creditor-right edge within one restructuring","causal_lever":"Sequence outreach, side payments, or rights purchases by spectral centrality and sensitivity of the consent network.","archetype_mapping":"Treat consent dependencies as a connectivity operator; use its dominant mode to locate exclusion rights whose resolution most reduces deadlock.","expected_value":"Potentially fewer negotiation rounds and lower coordination cost; adopter demand and effect size are unknown.","falsifiable_claim":"Across completed restructurings, spectral sequencing predicts the next binding holdout better than claim-size ranking and reduces simulated consent rounds by at least 15%.","diversity_rationale":"Uses network position and transaction sequencing at a case-edge level rather than temporal instability, data compression, behavioral offset, or market control.","mechanism_slugs":["network_spectral_centrality_analysis","power_iteration_probe","modal_sensitivity_sweep"],"search_questions":["Has spectral centrality been used operationally to sequence creditor consents?","Does centrality outperform claim size or legal seniority in identifying binding holdouts?"]},{"hypothesis_id":"H3","title":"Residual-safe thin-file underwriting","problem":"Document-heavy credit screening imposes access friction on thin-file firms whose evidence is distributed across correlated records.","affected_stakeholder":"Creditworthy small firms lacking conventional credit histories","workflow_boundary":"Loan-document intake through referral to manual underwriting","failure_mode":"Coordinate-level checklists duplicate correlated evidence while low-variance but decision-critical exceptions are discarded.","unit_of_analysis":"Loan application","causal_lever":"Replace redundant document checks with validated evidence modes while routing structured residuals to human review.","archetype_mapping":"Use PCA or SVD to define compact evidence modes, then retain enough modes to meet predictive and subgroup residual tolerances.","expected_value":"Potentially lower application burden without sacrificing loss control; prevalence and demand are unknown.","falsifiable_claim":"A modal intake protocol reduces required applicant fields by at least 25% while keeping out-of-sample default discrimination and subgroup false-negative rates within two percentage points of the full checklist.","diversity_rationale":"Focuses static information compression and residual governance at the application level, distinct from dynamic control and network-ranking hypotheses.","mechanism_slugs":["principal_component_analysis","singular_value_decomposition","residual_reconstruction_test","spectral_decomposition_report"],"search_questions":["Do existing thin-file underwriting systems explicitly use residual structure to preserve rare evidence?","Can modal document bundles reduce applicant effort without widening approval disparities?"]},{"hypothesis_id":"H4","title":"Offset-resistant policy bundles","problem":"Forward-looking firms and households can pre-adjust several margins that jointly neutralize an announced subsidy, tax, or credit intervention.","affected_stakeholder":"Fiscal and monetary policy designers","workflow_boundary":"Policy announcement through first-round behavioral response","failure_mode":"Elasticities are evaluated one margin at a time, missing a persistent combination of anticipatory offsets.","unit_of_analysis":"Policy-event by agent-segment response vector","causal_lever":"Choose bundles that load weakly on empirically observed offset modes and strongly on the target outcome mode.","archetype_mapping":"Recover response modes from announcement-window observations, perturb candidate policy controls in modal coordinates, and bound conclusions to the measured regime.","expected_value":"Potentially greater net pass-through from a fixed policy budget; world novelty and effect size are unknown.","falsifiable_claim":"Out-of-sample, bundle designs selected by modal sensitivity retain at least 20% more of their engineered effect after anticipatory responses than designs selected from marginal elasticities.","diversity_rationale":"Examines strategic pre-adjustment across policy events, unlike balance-sheet instability, consent topology, underwriting compression, or automated liquidation.","mechanism_slugs":["mode_shape_testing","modal_sensitivity_sweep","reduced_order_model","spectral_decomposition_report"],"search_questions":["Have policy designers estimated joint anticipatory-response modes rather than separate elasticities?","Do mode-targeted bundles improve realized pass-through in natural experiments?"]},{"hypothesis_id":"H5","title":"Mode-aware margin liquidation","problem":"Asset-by-asset margin rules can trigger mutually reinforcing liquidations across correlated portfolios.","affected_stakeholder":"Leveraged investors, brokers, and market-risk supervisors","workflow_boundary":"Margin breach detection through automated collateral sale","failure_mode":"Local liquidation minimizes one account's deficit while amplifying a shared price-impact mode across accounts and assets.","unit_of_analysis":"Account-asset liquidation instruction per stress interval","causal_lever":"Stagger or substitute sales to damp the most sensitive market-impact mode while satisfying account-level solvency constraints.","archetype_mapping":"Factor the account-to-asset liquidation map with SVD, classify locally amplifying modes, and run a reduced-order controller with reconstruction checks.","expected_value":"Potentially lower fire-sale impact and fewer cascading breaches at unchanged credit protection; feasibility and magnitude are unknown.","falsifiable_claim":"In replayed stress intervals, mode-aware liquidation lowers peak aggregate price impact by at least 15% without increasing broker credit exposure or unresolved deficits.","diversity_rationale":"Applies constrained real-time control to account-asset instructions, materially differing in stakeholder, workflow, unit, failure mechanism, and lever from H1-H4.","mechanism_slugs":["singular_value_decomposition","modal_stability_analysis","reduced_order_model","residual_reconstruction_test"],"search_questions":["Do broker liquidation engines optimize against systemwide modal price impact?","Can a reduced-order liquidation controller meet solvency constraints under realistic latency and residual error?"]}]}