?
Education only. Not financial advice. Signals are clues from public data, not predictions.
Alternative data, explained

The hidden signals behind the stocks you know.

Offbeat public data, like cardboard box prices, airport lines, and FDA reports, that hints where companies are heading. Explained in plain English.

Search about 14,000 US stocks and ETFs. Curated names get deep signals; everything else gets a light pack.

Deep-signal picks

Loading live data…

  1. Offbeat data

    We pull free public numbers most investors skip: box prices, TSA lines, FDA reports, developer downloads.

  2. Which way it leans

    Green or red shows if the number rose or fell vs its recent past. We label that a tailwind or headwind for the company.

  3. Why it matters

    Every signal says, in plain English, how it connects to the business and the usual time frame: weeks to a quarter.

New to stocks? Read the 2-minute beginner's guide →
Education only. Not financial advice. These are clues from public data, not predictions or buy/sell calls.

Pick a stock

Curated stocks get deep custom signals. You can also look up any US ticker for a lighter pack (price, news, Wikipedia, Hacker News, SEC inventory).

Green = number is up Red = number is down Gray = flat / noise

Green and red track whether the data moved up or down vs its recent baseline. Separate Tailwind / Headwind badges say whether that move looks favorable or unfavorable for the company (some signals flip — e.g. more inventory can be a headwind even when the number is up).

Track record

Did the leans play out?

We log every stock's lean and check what the price did 2 weeks, 1 month, and 1 quarter later, against the S&P 500 too.

See the track record →

Free weekly email

The strangest signals, once a week

Every Monday: the 3 to 5 oddest moves in our data (owner complaints, box prices, developer chatter…), one story explained in plain English, and our honest track record. No stock tips, ever.

We email a confirmation link first, and use your address only for this digest. Privacy · See this week's issue · Education only, not financial advice.

How it works

1

Collect

A scheduled job pulls fresh numbers from public sources like the St. Louis Fed, the SEC, the FDA, the TSA, Wikipedia, and the GDELT news index. No logins, no paywalls, no scraping where it isn't allowed.

2

Compare

Each signal is compared with its own recent past (last 28 days vs the 28 before, or this year vs last year). Small moves inside a noise band are called neutral.

3

Explain

Every card says what the data is, why it might matter for that company, and what it's showing right now, written from the actual numbers. If a source fails, we say so instead of guessing.

Read the full method and source list →

Scout finds

A scheduled correlation scout scans free public series against curated stock prices (Tiingo) and ranks economically motivated pairings. Top hits below — correlation is not causation, and multiple testing is a real risk. Full method on the how it works page.

Green = number is up Red = number is down Gray = flat

Same rule as the rest of the site: green/red track whether the public number moved up or down vs its recent baseline — not a buy/sell tip.

Loading scout results…

Live signals from this scout also appear on the linked stock pages. Full tables live in the repo under reports/correlation-scout.md.

Signals we'd love to add

Some of these started as ideas. Where we found a free, legal, frequent public source, we wired a live proxy into a curated stock (linked below). The rest stay labeled ideas — we never invent numbers.