Launching today
EquiReason
An entire research department running as software
2 followers
An entire research department running as software
2 followers
EquiReason is an entire research department running as software: it analyses the economy, the trends moving markets and every company they reach — from filings, earnings calls, economic data, news and market positioning — and holds a verdict on each, with the evidence for and against it. You are told when that read changes. Every company's filed record is open, no account needed.








Free Options
Launch Team

Atlassian LoomTalk it through in Loom. Ship with AI
Promoted
Hi everyone, I'm Yigit. EquiReason is an entire research department running as software: it analyses the economy, the trends moving markets and every company they reach — from filings, earnings calls, economic data, news and market positioning — and holds a verdict on each, with the evidence for and against it.
I've been building EquiReason on my own since January 2026. The idea: do what an equity research department does, as software, and show the work.
It reads SEC filings (10-K, 10-Q, 8-K, Form 4, 13F), earnings calls and macro data for about 3,100 US companies, then:
- finds what is changing in the world (a demand shift, a supply constraint, a theme many companies start describing at once) and traces it to the companies it actually reaches, checked against reported segment or product revenue where the company discloses it and labelled as estimated where it doesn't;
- holds a verdict on every company with the evidence for and against it, each piece tied to the filing it came from;
- tells you when that read changes: a new 8-K, insider buying, guidance raised or cut, a shift in 13F holders.
The data underneath is open without an account: financial statements for every year on file, 8-K events in plain English, insider trades and 13F holders (e.g. https://equireason.com/company/NVDA). Verdicts, the trend board and market-wide screens are paid. Prices are delayed 15 minutes. It is not investment advice.
Things that were harder than I expected:
- XBRL is inconsistent: the same line item sits under different tags across companies and years, so every annual cell is matched by period end date, not by the label.
- 13F filings include put and call rows; counting them as shares overstated some large funds' stock holdings by 46–80%.
- Several companies report in CNY or EUR, which is easy to print with a dollar sign by mistake.
If a number looks wrong, or a company is tied to a trend it shouldn't be, tell me and I'll trace it back to the filing. Happy to answer anything about the data.