Financial-condition grades for SEC filers
Every page is rendered from these files, and all of them are free to download. Each company page comes from its own artifact at tickers/TICKER.json, which carries every input with its receipt: tag, value, period end, filed date, form, accession and filing URL. The Downloads section of a company page offers its PDF, its artifact, and its inputs and rescans as CSV. The rescan history covers every rescan on file; the time machine browses it and the point-in-time reader opens any company on any rescan date.
| File | Size | Contents |
|---|---|---|
| index.csv | 1.6 MB | every ticker, one row: status, letter, percentile, composite, size band, warnings, price measures |
| inputs.csv.gz | 748 KB | every computable input of every graded company: value, percentile, peers, period end |
| history/rescans.csv.gz | 11.9 MB | long format: as_of, ticker, letter, percentile, composite, band, momentum, one row per company per rescan |
| history/cube.json | 7.5 MB | per ticker, arrays of letters and percentiles aligned to the list of rescan dates; what the time machine reads |
| assay.sqlite3.gz | 26.4 MB | one SQLite database: universe, rescans, inputs, warnings, with indexes on ticker |
| index.json | 6.3 MB | every ticker: status, letter, percentile, warnings, price measures, size band, filing gap |
| universe.json | 2.4 MB | every ticker with its eligibility status and reason |
| grade_history.json | 1.6 MB | one change point per letter or band change, per ticker |
| filing_calendar.json | 1.3 MB | expected periodic filings per eligible ticker: class, period, due date, status |
| methodology.json | 7 KB | the run's method parameters and factor formulas |
| forecasts.json | 0 KB | the forecast log and scorecard |
Field conventions: percentiles run 0 to 1 with lower stronger; money in US dollars as reported; dates ISO 8601; a missing value is null with a reason beside it. Data as is, without warranty.