Skip to main content
Use this page when a response is more than a source document. It explains how to read a normalized, classified, or calculated result without extending it beyond the evidence returned. The route reference remains the contract for fields and parameters.
For a consequential result, retain the route, parameters, period, source fields, methodology version when supplied, and every returned capability, warning, degradation, or statusNote field. Save the retrieval time and a stable result extract when policy permits. This makes it possible to distinguish a changed source, changed method, and changed request when a result later differs.

​
Shared terms

​
Market schedules

For GET /v1/market/calendar, read coverage, confidence, statusNote, and sessionStatus together. configured_holidays means the requested market has configured holiday coverage. weekend_only does not establish local-holiday coverage. tentative means the route cannot confirm the day from configured coverage. XNYS and XNAS are configured-holidays markets. See market calendar.

​
Filing-backed and derived results

Anchor filing-backed results to an accession number and filing URL. A page-level trace supports the page, not an exact resolved source span. Segment labels and hierarchies retain the issuer’s disclosure context.

​
Review before reuse

  1. Identify whether the result is source-backed, derived, or a combination, then open the linked source where the decision needs primary evidence.
  2. Match the issuer, period, units, methodology version, and any route-specific input choices before comparing two results or reusing one in a model.
  3. Preserve the returned status fields with the value. Null, unknown, partial, degraded, and unavailable are states to explain, not values to coerce to zero.
  4. For a calculation, retain the inputs and the result together. Recalculate only when the route documents the necessary inputs and method; otherwise label the result as supplied by the route.
  5. When a route returns limited coverage or a changed methodology, re-evaluate the decision rather than silently joining it to earlier outputs.
Similar labels can represent different periods, source selections, or calculations. This registry describes interpretation boundaries; it does not create comparability where the returned evidence does not support it. Show unknown, partial, degraded, and unavailable states in your application. Read this alongside Freshness and source evidence and Coverage and historical depth before relying on a calculation.