MethodologyThe rubric

How we grade the market.

No black box. Every name that reaches the cut runs the same gauntlet — the same rubric, the same signals, the same disclosure of misses.

UNFAIR · SIGNALS RESEARCHLiving document

The cadence

We scan the entire investable US equity universe — roughly 6,000 names — twice every trading day: a pre-open pass at 9:45 AM ET and a mid-afternoon pass at 3:30 PM ET. We are deliberately not a real-time firehose. The two windows are when filings, flow, and price have enough signal to be worth reading, and infrequent enough that you are not whipsawed by noise.

The four verdicts

Every name that surfaces gets exactly one verdict. No "maybe," no watchlist of 400.

The eight grade-cards

Behind each verdict sit eight graded dimensions. Each is scored on the same A–F scale, against the same definition, every time — so the score is auditable, not vibes.

Grade-cardWhat it measures
MoatDurability of the competitive edge
Balance sheetLeverage, liquidity, net cash vs. net debt
ExecutionTrack record of beats, raises, and guidance credibility
ValuationPrice paid relative to growth and quality
InsiderForm 4 buying — especially clustered, open-market purchases
Institutional13F accumulation vs. distribution by credible holders
Short interestCrowding, days-to-cover, and the direction of the trend
CatalystWhether there is a dated, live event that can re-rate the name

The smart-money signals

Fundamentals tell you what a business is worth. The signals tell you what informed money is doing about it before the headline. We cross-check four:

Ender

Ender is the AI analyst that turns the grades and signals into a readable thesis. It reads the same primary sources a buy-side desk reads — filings, transcripts, flow — and writes the memo with a citation on every material claim. It is an analyst, not an oracle: it shows its work so you can disagree with it.

On the misses

We publish the calls that didn't work. The Field Notes include post-mortems, and the Receipts table shows losers next to winners with net alpha measured against SPY over the same dates. A methodology you can only see when it's winning isn't a methodology.

Conviction over coverage. We would rather hand you twelve names we can defend than four hundred we can't.

This page describes an analytical process and is not investment advice, a solicitation, or a recommendation. Grades and verdicts are opinions generated from public data and may be wrong. Past performance does not guarantee future results.

Read Field Notes Request an invite →