The SEC gives away every company's financials. Parsing them was the hard part

Short summary

The founder of Intrinsiqq explains why parsing SEC filing data is far harder than the finance itself. Filings are free and machine-readable but silently wrong in ways that never throw errors—misaligned fiscal years, cumulative vs. period figures, restatements, and reconstruction arithmetic all produce plausible-but-incorrect numbers. The entire project discipline became detecting wrongness when nothing signals it.

• •SEC filing data is free and structured but silently wrong in non-obvious ways
• •Fiscal year mismatches, cumulative vs. period figures, and restatements create hidden data errors
• •The core challenge is detecting incorrect values when nothing throws or logs an error

Generated with AI, which can make mistakes.