A research filter, not a predictor — not financial advice. Grades rank growth stocks on evidence-based quality signals from end-of-day data; they are not buy/sell calls or price forecasts. Always do your own research.

Editorial policy

FrontierLab publishes three kinds of pages, and we're explicit about how each is made.

1. Data pages (automatic)

Per-stock, per-sector, screen, and track-record pages are rendered from our own daily data — grades, flags, calibration. No opinions are generated; they show what the numbers say, refreshed each morning.

2. Editorial notes (AI-drafted, human-approved)

Short market notes are drafted by AI strictly from FrontierLab's own data (the day's grade changes, new/cleared flags, calibration moves) and reviewed and approved by a human before publishing. Nothing is auto-published. Every note links to the data it cites and is labeled "AI-assisted, human-reviewed."

3. Research reports (AI-assisted, human-shaped)

Longer studies (e.g. "did the grades work this quarter?") are built from our computed tables, AI-assisted in drafting, then shaped and approved by a human. Every figure carries its sample size and window.

What we will not do

No mass-published, unreviewed AI articles. Velocity is capped on purpose.
No price targets, "fair values," or buy/sell calls.
No fake precision — small samples are labeled "too early to tell."

Corrections

If a published page is wrong, we fix it and note the correction. The grade-at-entry record in the paper game is immutable, so the track record can't be quietly rewritten.

Methodology → · About → · Disclaimer →