HNHacker News
TopNewBestAskShowJobs

areimann

5 karma · joined November 12, 2020

submissionscomments
areimann··on Show HN: Daily ETF holdings for 2,200+ ETFs as one API
That is an interesting one to dive into! Thank you for sharing!
areimann··on StockFit API – structured SEC EDGAR data with a free tier
I've been building StockFit API for about a year — a REST API that turns SEC EDGAR filings into clean JSON you can actually use.

The backstory: I kept starting side projects that needed financial data and kept burning 80% of my time just wrestling with EDGAR instead of building. XBRL is technically public and free, but "free" is doing a lot of work there — tagging is inconsistent across filers, fiscal years drift, Q4 isn't even a standalone period (you derive it from the 10-K minus the 9-month Q3 figure or the next best available data point), and the same financial concept maps to half a dozen different tags. Every fintech dev I know has written and thrown away their own EDGAR parser at least once.

So I built the thing I wanted. Under the hood:

- Arelle for XBRL processing - DuckDB for the fact store (~250M facts, ~5M filings) - Daily ingestion from EDGAR, with Q4 reconstruction at read time - Curated fact mapping on top of raw XBRL concepts so you don't have to know that us-gaap:Revenues, us-gaap:SalesRevenueNet, and us-gaap:RevenueFromContractWithCustomerExcludingAssessedTax all mean roughly the same thing depending on the filer - REST API with OpenAPI docs and an MCP server if you want to use it from Claude/Cursor/etc

Coverage right now: financials, ownership, ETF/mutual fund exposure, insider transactions (Forms 3/4/5), institutional 13F holdings, executive comp from proxy filings, the raw filings themselves, and screeners for Stocks/ETFs.

The free tier is 100 requests/day, no credit card. Paid tiers start around $19/month if you need more.

Things I'd genuinely value feedback on:

1. The API shape — I made a bunch of opinionated decision. Happy to change things if they could be better. 2. The Q4 reconstruction logic — it works, but I've seen edge cases with non-December fiscal year ends (AAPL=Sep, COST=Aug) - I'm reasonably sure I've covered all of them. 3. The docs - are they good the way they are

Not things I'd love feedback on (I know, they're on the roadmap):

- Historical point-in-time data earlier than ~2010 is spotty - Real-time quotes — this is fundamentals data, not a market-data feed - International filings — US-only for now

Link: https://developer.stockfit.io Docs: https://api.stockfit.io/docs

Happy to answer questions or chat about the project if anyone's interested in building on top of it. Thanks for reading!

areimann··on StockFit – A long term investor app
By spending your own time and doing the research, you can absolutely beat the market. We can't guarantee that you will, but with StockFit you can greatly increase the odds. The app can help you along the way by giving you the information you need to find excellent companies best positioned for the long term.