Markets run on stories. Someone should audit them.
Market Scholar was built on a simple observation: a company’s price is driven less by its filings than by the narrative wrapped around them — and that narrative is rarely examined with the same rigor as the numbers. We build the missing layer.
Principles
Measure, don’t opine
We score how a narrative holds up against the primary record. We don’t tell you what to trade — we show you what’s true to the filings.
Evidence over vibes
Every score traces back to a claim, a source, and a timestamp. If we can’t point to the record, we don’t publish the number.
Honest about limits
We separate what’s validated from what’s exploratory, and we say which is which. The forensic layer is measurement; it is not a promise of returns.
A method, not a black box
Our approach is grounded in three patent-pending methods: multi-dimensional source-credibility assessment, narrative lifecycle and decay tracking, and an inference-time technique that prevents look-ahead bias in analysis.
The forensic layer is validated as predictive of narrative structure, and we’re deliberate about that boundary: measurement of how a story holds up is a different claim from a forecast of returns. We hold ourselves to the former.
Built to sit inside a regulated workflow
Read-only by design
The application reads from our forensic index. It does not place trades, move money, or take actions on your behalf.
Your queries are yours
We don’t use customer queries to train models. Enterprise deployments support data isolation.
Access controls
SSO and audit logging are available for enterprise so you can see who looked at what, and when.
Primary-source provenance
Every score is anchored to the filing or article that produced it, so findings can be checked, not just trusted.
Pursuing formal certifications as we scale; we’ll share status candidly rather than imply more than is in place. Ask us where we are today.