Mapping Real-Time Regulatory Impacts to Asset Vulnerability Under Strict API Budgets

Systems that connect regulatory change to the companies it touches are usually built on silently smoothed data: gaps are imputed, single-source values are presented as settled facts, and the boundary between a disclosed number, a calculated number, and an inferred number is invisible to the reader. This report describes the methodology of Meridian, a working provenance-first research system that maps real-time regulatory instruments and changes (supervisory rules, sanctions and trade-control lists, official filings) onto a universe of roughly 1,300 issuers and infers potential — not measured — impact on the affected assets. Three commitments organise the work. First, a provenance-first data model in which every value carries its own source key, retrieval and publication timestamps, an immutable hashed raw artifact, a parser version, and an explicit verification class (source fact, company disclosure, official regulatory fact, calculated metric, evidence-based inference, or unverified); missing values are shown as missing, with a reason, and are never estimated. Second, an evidence-first linkage engine that connects an instrument to a company only through a quotable, sourced business fact (a listing venue, a supervisory classification, a recorded sector label, a disclosed geographic revenue split), records supporting and counter-evidence, and keeps source fact, calculation, and labelled inference distinct at every step. Third, an operating discipline for strict API budgets: a per-provider daily-cap governor that refuses to exceed advertised free-tier limits, a need-driven rotation that spends breadth-first before cross-provider confirmation, and an "outage is not absence" rule that prevents a provider's rate-limit ban from being mistaken for a data gap. We describe the data model, the linkage method, the budget governor, the epistemic-honesty design, and a weighted coverage-tiering and completeness-gating scheme, and we discuss the limitations that follow from choosing honesty over coverage. We report no empirical impact findings; all worked examples are illustrative of mechanism, not results. Preprint; not yet peer reviewed.