Platform

A forensic engine for the story behind the price

Market Scholar treats a market narrative as something measurable. It reads the primary record, scores each claim, and resolves dozens of signals into a single verdict — with every number traceable to its source.

The pipeline

Ingest, audit, surface

01

Ingest

SEC filings, earnings, analyst coverage, options and short-interest data, and a 450K-article news corpus — normalized and time-stamped so analysis can’t peek at the future.

02

Audit

A scoring engine decomposes each narrative into claims and runs them against the disclosures, scoring verification, drift, coordination, authority, divergence and decay six times a day.

03

Surface

Results land in a search engine and a monitoring layer: look up any company for a verdict, or watch your universe for narratives turning against the filings.

The signals

What we measure

Verification Match (VMS)

0–100

Decomposes a narrative into discrete claims and checks each against the structured data in the 10-K and 10-Q. A high score means the story is literally true to the filings — which is not the same as the story being safe.

Epistemic Drift

0–100

Measures the gap between what a narrative implies and what the disclosures actually support. Drift fires on selective framing and omission, not only on falsehood.

Coordination

class + score

Detects orchestrated narratives from cross-source timing and phrasing — distinguishing organic spread from a campaign that arrived everywhere at once.

Source Authority (SRS)

0–100

A forensic credibility score for each outlet, earned from its historical record of claims that later held up against the filings — not from brand or reach.

Fair-Value Divergence (FVD)

%

How far price has run from a pure-fundamentals anchor built off the filings. A prudent reference point — not a price target, and never a sizing input.

Narrative Decay

half-life

Physics-based modeling of a narrative’s energy: how long it persists and when it exhausts. Built on a decay framework, validated against forward structure.

The output

A verdict, with its reasons

Every company resolves to one of a small set of forensic verdicts. The verdict is a summary — the components are the evidence.

Structurally Supported

Claims align with the filings and valuation divergence is contained.

Narrative Risk

Elevated structural pressure that has not yet resolved into a clean trap.

Narrative Trap

Price detached from fundamentals while claims still verify, with coordination present.

Coordinated Watch

A coordination signature is present across sources.

Exhausted Narrative

The story has spent most of its energy.

Monitoring

No dominant signal. Tracked for changes.

A note on framing: the “Narrative Trap” verdict is valuation-anchored, not a fraud label — a high verification score means the claims check out. For falsity screening, read the components (drift, coordination, omission), not the headline verdict.

Research

Patent-pending methods

63/971,470

Analyst Credibility Assessment

Multi-dimensional scoring of analyst and source claims against the primary record.

63/971,478

Narrative Lifecycle Tracking

Decay monitoring that models how a narrative loses energy over time.

Application pending

Divergence Decoding

Inference-time temporal unlearning to prevent look-ahead bias in analysis.

Put the engine to work on your universe