{"schema_version":1,"experiment_id":"eoa_inverse_innovation_exp11_mechanism_context_external20_20260804","research_id":"eoa_inverse_innovation_exp11_external_scrutiny_20260804","cell_id":"computability_boundary_mapping__economics_finance","opaque_id":"computability_boundary_mapping__economics_finance__C","search_lanes":{"direct_problem":{"queries":["universal smart contract verification undecidable arbitrary programs termination safety property","algorithmic trading strategy formal verification solvency all market paths undecidable","smart contract liquidity solvency verification formal methods paper","financial regulator algorithmic trading pre-deployment testing risk controls official"],"source_ids":["S3","S4","S5","S7"],"no_result_note":"No retained source documented a supervisor or venue demanding an exact terminating Boolean solvency classifier for arbitrary code and every market path; the component risks and verification limitations were documented separately."},"closest_prior_art":{"queries":["smart contract liquidity solvency verification formal methods paper","decidable verification subclass arbitrary programs automatic verification","expressive logic undecidable smart contract decidable layer verification","Solidity SMTChecker unproved unsupported timeout"],"source_ids":["S2","S3","S4","S8"],"no_result_note":null},"historical_terminology":{"queries":["Turing 1936 computable numbers Entscheidungsproblem PDF original","Rice 1953 Classes of recursively enumerable sets and their decision problems PDF","universal halting totality program verification decidability"],"source_ids":["S1","S2"],"no_result_note":"Historical results establish general decision-procedure boundaries but do not themselves establish a semantics-preserving reduction for financial solvency."},"products_practices_standards":{"queries":["site:soliditylang.org SMTChecker unknown unsupported timeout documentation","site:eur-lex.europa.eu 2017/589 algorithmic trading testing environments conformance","site:fca.org.uk algorithmic trading controls development testing governance review","site:sec.gov Rule 15c3-5 automated trading market access pre-trade risk controls testing"],"source_ids":["S4","S5","S6"],"no_result_note":null},"non_english_regional":{"queries":["智能合约 形式化验证 不可判定 超时 未知","Unentscheidbarkeit Smart Contracts formale Verifikation Solvenz","Règlement délégué 2017/589 environnement de test trading algorithmique"],"source_ids":["S6","S7"],"no_result_note":"The retained Chinese Ethereum documentation independently described nontermination, decidability, specification, and performance limitations; multilingual EUR-Lex versions confirmed the regional regulatory terminology."},"composition_subproblems":{"queries":["bounded horizon finite state trading strategy verification solvency","sound incomplete static analysis unknown timeout unsupported result","formal smart contract liquidity every reachable state withdrawal verification","algorithmic trading separated testing environment approval change records controlled deployment"],"source_ids":["S2","S3","S4","S5","S6","S8"],"no_result_note":null}},"sources":[{"source_id":"S1","title":"On Computable Numbers, with an Application to the Entscheidungsproblem","url":"https://www.astro.puc.cl/~rparra/tools/PAPERS/turing_1936.pdf","publisher":"Proceedings of the London Mathematical Society","date_or_year":"1936-1937","source_type":"PRIMARY_RESEARCH","language":"English","claims_supported":["Classical computability theory supplies proof techniques showing that some universal terminating decision procedures cannot exist.","An impossibility conclusion depends on a declared computational representation and a valid reduction, not merely on poor empirical performance."]},{"source_id":"S2","title":"Decidable Verification of Uninterpreted Programs","url":"https://arxiv.org/abs/1811.00192","publisher":"Proceedings of the ACM on Programming Languages / arXiv","date_or_year":"2019","source_type":"PRIMARY_RESEARCH","language":"English","claims_supported":["Automatic verification of the paper's unrestricted uninterpreted-program class is undecidable.","The authors identify machine-recognizable coherent and k-coherent subclasses with decidable verification procedures.","Restricting the admitted program class can convert a general verification problem into a decidable one, with complexity analysis following afterward."]},{"source_id":"S3","title":"Solvent: liquidity verification of smart contracts","url":"https://arxiv.org/abs/2404.17864","publisher":"arXiv","date_or_year":"2024","source_type":"PRIMARY_RESEARCH","language":"English","claims_supported":["Liquidity is a directly relevant smart-contract property concerning whether users can recover crypto-assets.","The tool asks whether, from every reachable state, a suitable transaction sequence permits withdrawal.","Existing Solidity verification tools were reported as ineffective for expressing and verifying these liquidity properties, motivating a specialized verifier and benchmark evaluation."]},{"source_id":"S4","title":"SMTChecker and Formal Verification — Solidity documentation","url":"https://docs.soliditylang.org/en/latest/smtchecker.html","publisher":"Solidity Project","date_or_year":"2026 (accessed 2026-08-04)","source_type":"FIRST_PARTY_PRODUCT","language":"English","claims_supported":["A deployed smart-contract analysis product explicitly reports proved and unproved targets and separately reports unsupported language features.","The tool uses resource limits or configurable timeouts, so exhaustion is operationally distinct from proof of safety or failure.","Unsupported constructs are over-approximated to preserve soundness but may produce false positives, and external calls may be modeled as unknown.","Bounded and constrained-Horn-clause engines have different scopes, costs, and completeness characteristics."]},{"source_id":"S5","title":"Multi-firm review of algorithmic trading controls: high-level observations","url":"https://www.fca.org.uk/publications/multi-firm-reviews/algorithmic-trading-controls-high-level-observations","publisher":"UK Financial Conduct Authority","date_or_year":"2025-08-21","source_type":"OFFICIAL_GUIDANCE","language":"English","claims_supported":["Regulated trading firms conduct pre-deployment conformance and simulation testing, use pre-trade controls, and maintain deployment approval processes.","The FCA found variation and deficiencies in documentation, governance, technical understanding, testing sophistication, ownership, and accountability.","Most reviewed firms used documented deployment processes, while some approvals involved senior management, compliance, risk, or multiple business areas.","Testing against stressed scenarios is practiced but does not purport to establish correctness over every possible market path."]},{"source_id":"S6","title":"Commission Delegated Regulation (EU) 2017/589 (RTS 6)","url":"https://eur-lex.europa.eu/eli/reg_del/2017/589/oj","publisher":"European Union / EUR-Lex","date_or_year":"2017","source_type":"OFFICIAL_STANDARD","language":"Multilingual EU text","claims_supported":["EU law requires conformance testing before deployment or material updates of specified algorithmic trading systems and strategies.","Required testing must occur in an environment separated from production, and firms retain responsibility for the testing.","Controlled deployment uses predefined limits on instruments, prices, order quantities, positions, and venues.","Material software changes, their authors, approvers, and nature must be recorded; risk management produces validation reporting."]},{"source_id":"S7","title":"智能合约的形式化验证 (Formal verification of smart contracts)","url":"https://ethereum.org/zh/developers/docs/smart-contracts/formal-verification/","publisher":"ethereum.org","date_or_year":"Updated 2026-06-26","source_type":"OFFICIAL_GUIDANCE","language":"Chinese","claims_supported":["Formal verification proves conformance only to a specified model and specification; a poor specification can leave violations undetected.","State explosion, path explosion, and solver cost constrain verification performance.","Program verifiers may fail to decide a property because computation may not terminate, even when the contract is well specified.","Interactive theorem proving can require human assistance, which changes the operational computation model."]},{"source_id":"S8","title":"Formal requirement enforcement on smart contracts based on linear dynamic logic","url":"https://research.ibm.com/publications/formal-requirement-enforcement-on-smart-contracts-based-on-linear-dynamic-logic","publisher":"IBM Research","date_or_year":"2018-07-01","source_type":"PRIMARY_RESEARCH","language":"English","claims_supported":["Expressive formal logic for smart contracts may be undecidable, preventing automatic generation of executable programs.","The proposed response separates a decidable specification layer from an implementation-rule layer while systematically relating them.","The approach was implemented and evaluated on Hyperledger Fabric, showing that a restricted, proof-oriented contract workflow is technically feasible."]}],"problem_evidence":{"status":"PARTLY_SUPPORTED","finding":"The underlying mismatch is real: regulated firms must test executable trading algorithms before deployment; smart-contract liquidity is an active verification target; and production verification tooling exposes timeouts, unproved targets, unsupported constructs, abstractions, and unknown external behavior. However, the bounded search found no direct evidence that a supervisor or venue currently promises an exact, terminating safe/unsafe solvency decision for every arbitrary strategy and every admissible market path, nor evidence that timeout or abstention is routinely collapsed into a Boolean verdict. The proposal therefore combines a well-supported technical risk with an unverified institutional-demand hypothesis.","source_ids":["S2","S3","S4","S5","S6","S7","S8"],"uncertainty":"A semantically faithful reduction from the proposed financial solvency predicate to an undecidable program property has not been supplied. Finite precision, gas limits, bounded horizons, admission-language restrictions, or finite scenario sets could make an actual deployed requirement decidable, leaving complexity rather than computability as the binding issue."},"adopter_evidence":{"status":"SUPPORTED","finding":"Identifiable adopters and authorizers exist. FCA-regulated principal trading firms, investment firms subject to RTS 6, trading venues, direct-market-access providers, compliance and risk functions, and senior deployment approvers already own testing, validation, controlled-deployment, and recordkeeping decisions. A venue or supervisory risk owner can authorize an offline evaluation; only the legally empowered governance body can alter live admission policy.","source_ids":["S5","S6"],"uncertainty":"The sources establish authority for algorithmic-trading controls, not that these bodies currently commission universal solvency classifiers. Smart-contract deployment governance varies across chains and venues."},"implementation_evidence":{"status":"PARTLY_SUPPORTED","finding":"The intervention's building blocks are feasible and precedented: formal research maps undecidable general classes to decidable subclasses; IBM demonstrated a decidable smart-contract specification layer; Solidity distinguishes proved, unproved, unsupported, timeout-sensitive, and externally unknown cases; and EU regulation requires separated testing environments, controlled bounds, approval records, and revalidation. What remains unimplemented is the complete finance-specific package: a faithful solvency semantics, an enforceable fragment recognizer, reviewed reductions or deciders, a stable public status taxonomy, and evidence that users interpret the statuses correctly.","source_ids":["S2","S4","S6","S7","S8"],"uncertainty":"No retained source validates the proposed solvency predicate across real market microstructure, stochastic real-valued paths, institutional rules, oracle failures, and cross-system interactions."},"prior_art":{"disposition":"ADJACENT_PRIOR_ART","closest_analogues":[{"name":"Decidable verification through coherent and k-coherent program subclasses","source_ids":["S2"],"same_problem":false,"same_causal_lever":true,"overlap":"It explicitly proves unrestricted verification undecidable, identifies recognizable restricted subclasses, provides decision procedures, and separates decidability from complexity.","remaining_difference":"It is generic program-verification research, not a financial solvency classifier, and does not provide supervisory governance, status communication, distributional review, or an offline admission-policy pilot."},{"name":"IBM two-layer smart-contract development using decidable Linear Dynamic Logic","source_ids":["S8"],"same_problem":true,"same_causal_lever":true,"overlap":"It confronts undecidable expressive contract logic by restricting the formal specification layer to a decidable logic and linking that layer to executable implementation.","remaining_difference":"It does not classify arbitrary existing strategies over market paths, prove a solvency-specific boundary, distinguish timeout and out-of-scope operational states, or test supervisory interpretation."},{"name":"Solvent liquidity verifier","source_ids":["S3"],"same_problem":true,"same_causal_lever":false,"overlap":"It directly targets a solvency-adjacent financial property over every reachable contract state and supplies a specialized verification tool and benchmark evaluation.","remaining_difference":"It is constructive verification for a particular liquidity formulation, not a map of computability boundaries for arbitrary executable strategies, and it lacks the proposed governance and abstention workflow."},{"name":"Solidity SMTChecker result and scope handling","source_ids":["S4"],"same_problem":false,"same_causal_lever":true,"overlap":"It operationalizes scoped engines, resource limits, unproved results, unsupported-feature warnings, sound over-approximation, and unknown external calls.","remaining_difference":"It does not publish a solvency-specific impossibility certificate or computation-model contract, enforce venue admission subclasses, or measure whether decision-makers confuse non-verdict states with safe or unsafe."},{"name":"RTS 6 and FCA algorithmic-trading testing and deployment controls","source_ids":["S5","S6"],"same_problem":false,"same_causal_lever":false,"overlap":"Existing regulation and practice already provide separated pre-deployment testing, stress scenarios, controlled limits, accountable approvals, inventories, records, and revalidation triggers.","remaining_difference":"These controls are empirical and operational rather than computability-boundary analysis; they neither promise exhaustive path coverage nor define proof-relative unknown and out-of-scope states."}],"contrastive_claim_remaining":"For a fixed corpus of executable financial strategies, adding a reviewed, model-relative computability classification, machine-enforced admission to supported fragments, and visibly distinct unknown, timeout, out-of-scope, failure, safe, and unsafe states to an existing regulated pre-deployment workflow will reduce false Boolean resolutions and user misinterpretation without producing unsound safe labels, unacceptable review delay, or disproportionate exclusion relative to the existing testing-and-manual-review baseline.","contrastive_claim_falsifier":"The incremental claim is falsified if independent proof reviewers find any unsound class or safety label; the fragment gate admits unsupported specimens; users cannot reliably distinguish the status categories; false Boolean resolution does not fall versus baseline; review delay exceeds a preregistered limit; or exclusion burdens increase beyond a preregistered affected-party threshold.","confidence":"MODERATE","search_limitations":"The search retained exactly eight direct sources and covered theory, research, tools, regulation, practice, historical terminology, and Chinese-language material. It did not exhaust patents, proprietary venue requirements, internal supervisory procurements, closed-source verification products, every jurisdiction, or every synonym. The sources cannot establish world novelty, patentability, freedom to operate, market size, or realized impact."},"researchability_gates":{"externally_supported_problem":{"status":"PASS","rationale":"External evidence supports the researchable core: pre-deployment algorithm controls are mandatory and practiced, liquidity verification is an active problem, and real verification tools expose scope, timeout, abstraction, and non-proof states. The narrower claim that an institution already demands the impossible universal Boolean classifier remains unverified but can itself be tested in requirements audits.","source_ids":["S3","S4","S5","S6","S7"]},"identifiable_adopter_or_authorizer":{"status":"PASS","rationale":"Regulated firms, trading venues, compliance and risk functions, senior deployment approvers, and competent supervisory bodies have identifiable roles in testing and admission governance.","source_ids":["S5","S6"]},"distinct_testable_incremental_claim":{"status":"PASS","rationale":"Prior art establishes the ingredients separately but does not eliminate a credible incremental claim about integrating computability-relative scope enforcement and explicit non-Boolean states into a supervised financial workflow and measuring interpretation, soundness, delay, and exclusion outcomes.","source_ids":["S2","S3","S4","S5","S6","S8"]},"bounded_next_evidence_step":{"status":"PASS","rationale":"A fixed-corpus, time-boxed, offline comparison is bounded and consistent with established separated testing environments. It can use independent proof labels and preregistered measures without live orders, capital, or admission decisions.","source_ids":["S4","S5","S6"]},"no_unresolved_safety_or_authority_stop":{"status":"PASS","rationale":"No unresolved stop prevents an offline study if a venue or supervisory risk owner authorizes it, outputs cannot affect live admission, all non-verdict states remain distinct, records are retained, and the exercise automatically halts on an unsound label or gate failure.","source_ids":["S5","S6"]},"adequate_search_evidence":{"status":"PASS","rationale":"All six required lanes were searched adversarially. The eight retained sources come from seven publisher contexts and include primary research, official regulation and guidance, and first-party product documentation, with every retained source opened.","source_ids":["S1","S2","S3","S4","S5","S6","S7","S8"]}},"strict_success":true,"screen_survival":true,"remaining_research_value":"MODERATE","recommended_next_step":"Pre-register and run an offline, automatically expiring pilot on a fixed, versioned corpus containing finite-state and bounded-horizon specimens, unrestricted or oracle-dependent specimens, known-safe cases, known-failing cases, time-consuming cases, and deliberately out-of-scope cases. Obtain independent formal-methods labels before evaluation. Randomize qualified reviewers between the existing workflow and a prototype exposing the model, assumptions, fragment gate, proof record, and seven distinct outcomes. Measure unsound safe labels, false Boolean resolutions, status-comprehension errors, abstention, review time, escalation load, and exclusion rates by developer type. Permit no live trading or admission effect; halt and revert on any unsound label, collapsed state, fragment-gate failure, or unauthorized use.","world_novelty_boundary":"This bounded search supports only an adjacent-prior-art judgment and a remaining empirical integration claim. It does not establish that the package is novel worldwide, patentable, non-infringing, commercially valuable, scalable, or capable of producing realized supervisory or market impact."}