Normalized financial statements
Income statements, balance sheets, and cash flows. Normalized from SEC CompanyFacts into a SQLite database you can work with locally.
SEC fundamentals in SQLite
Download a published snapshot, open it with SQLite, and join financial statements with company metadata. No hosted API required.
Coverage varies by company, period, and field. Missing values stay NULL; they are never filled with zero.
Normalized financial statements
Download the compressed SQLite database and decompress it before opening.
- File
- api_all_periods.sqlite3.gz
- Uploaded (UTC)
- 28 Oct 2026
- Download size
- 96.96 MB
Hosted on Cloudflare R2. This development endpoint is rate-limited. Dataset licensing and commercial contribution terms are still undecided.
| Table | What you’ll find |
|---|---|
| companies | CIK, ticker, company name, SIC, and available company metadata |
| income_statements | Revenue, expenses, earnings, and per-share metrics |
| balance_sheets | Assets, liabilities, equity, and related balances |
| cash_flow_statements | Operating, investing, and financing cash flows |
| dataset_metadata | Snapshot and normalization version information |
Mapping trees
Explore the current industry mappings. Trees are generated from the repository’s JSON at deployment. Nesting does not imply that child values should be added together; * marks concepts under review.
Loading mapping trees…
Query financial statements
- Download and decompress a database release.
- Open the SQLite file in your preferred client.
- Query by CIK, period, and reporting currency.
Financial values are stored as decimal text. Cast explicitly when performing numeric calculations. Q1–Q4 and FY identify periods; frequency distinguishes annual and quarter rows.
SELECT date, fiscalYear, revenue,
reportedCurrency
FROM income_statements
WHERE cik = '0000320193'
AND period = 'FY'
AND reportedCurrency = 'USD'
ORDER BY date DESC
LIMIT 5;
Contribute to the mappings
Help improve the mappings. Browse an industry’s current structure, inspect unmapped concepts, and suggest a classification with filing evidence. Accounting knowledge is enough to get started.
Before you use the data
Mappings are provisional and normalization may include candidate selection, calculations, and sign adjustments. Validation does not establish accounting correctness. Historical coverage is uneven. Check material values against the original filings.
Public access to this repository does not imply a license to reuse or redistribute the dataset. Dataset licensing and commercial contribution terms are still undecided.