How long it has been reporting.
The Lindy heuristic says a thing that has already survived a long time can be expected to survive longer. Testing that honestly needs an archive that can prove what was knowable on each date — so this measures each issuer's observed reporting span, how many of the years inside that span it actually reported, and its longest gap. Nothing is estimated, and no future value is projected.
Grain: issuer · annual periods only · point-in-time by observed_at · deterministic · on the licence-clean AF-FUND spine · 4 provenance guarantees
Panel dictionary.
| Field | Type | Description |
|---|---|---|
dataco_id | string | Certvas ID of the issuer. |
first_period_end / latest_period_end | date | Earliest and latest annual period the issuer has reported in the archive. |
years_reported | number | Calendar years between those two periods — the observed span. |
n_annual_periods | integer | Distinct annual periods held for the issuer. |
n_facts | integer | Underlying reported facts behind those periods. |
continuity_ratio | number | Distinct reported years divided by calendar years spanned. Below 1.0 means a year inside the span is missing. |
max_gap_years | integer | Longest run of consecutive absent years inside the span. An unbroken record is 0. |
lindy_expected_additional_years | number | The Lindy reading of the observed age, stated explicitly. Equal to years_reported by construction. |
first_observed_at / latest_observed_at | date | When the issuer was first and most recently seen in the archive. |
Read this as a lower bound, not a measurement. The signal measures persistence of reporting in the Certvas archive. A short history can mean a young issuer or late-started coverage, and from inside the data those are indistinguishable. Right now the honest reading is that this panel says as much about Certvas's coverage depth as about issuer longevity — an issuer showing three periods across a ten-year span has not stopped reporting, we simply do not yet hold the middle. That gap closes as coverage deepens, with no change to this signal.
lindy_expected_additional_years is not a forecast. It is a mechanical restatement of observed age under a stated assumption, published as its own column precisely so the assumption is visible and auditable rather than buried inside a composite score. Certvas does not assert that issuer lifetimes are Lindy-distributed.
What this signal deliberately does not contain. No market capitalisation or price term — Certvas holds no market-data licence, and that boundary is deliberate. No adoption or sentiment term, because no licence-clean source for it exists in our register. No subjective risk weights, because multiplying real data by asserted constants presents a guess as a computation. And no projected values for years that have not happened, because Certvas's core promise is that where a value is unknown it is absent rather than guessed. A projection column would contradict that promise inside the product itself.
Scope. Annual periods only — mixing annual with interim makes continuity_ratio mean different things for different issuers while still looking comparable. An issuer with fewer than two annual periods is excluded rather than scored low, because its span is unmeasured, not short.