Journalism & education — not investment advice. Full disclaimer ↓

How It Works

One builder, a newsroom of AI analysts.

Most stock write-ups are one person with one hunch.
My Friend The Trend works like a newsroom.

The method

Every candidate trend has to survive three questions, in order:

— Is it real? A genuine change in behaviour, backed by observable evidence — sell-throughs, search spikes, sellouts, damage estimates — not vibes.

— Is it still under the radar? Has financial media already connected the trend to the stock? Has the price already moved on it? If the story has been written, the edge is gone.

— Is it big enough to matter? Is the trend large relative to the company's revenue — and is anything else drowning the signal out?

The scouts

Five AI scouts work independent beats and can't see each other's work:

— The Feed — viral consumer products

— The Register — search & commerce momentum

— The Marquee — culture & entertainment

— The Barometer — the physical world (weather, outages, repair demand)

— The Docket — policy & regulation

Each returns structured trend dossiers — dated evidence, real sources — and is required to report the leads it discarded, and why. The reject pile is part of the product.

The desks

Two desks then try to kill every idea before it ships.

The Skeptic hunts for proof the market already knows — analyst notes, price action, management guidance. When in doubt it rules "priced in." A wrong "under the radar" costs you money; a wrong "priced in" only costs us one idea.

The Accountant does the arithmetic — is the trend material to this specific company, and is the stock actually investable (US-listed, liquid, real)? It also writes the kill list: the concrete things that would prove the thesis wrong. No kill list, no thesis.

The editor

An editor reads across every dossier and desk verdict to find the thread — the meta-trend that shows up on several beats at once (the GLP-1 economy did, in issue #001) — picks the two-to-four theses worth shipping, and files the rest honestly: real-but-priced trends under Too Late, real-but-no-vehicle trends on the Watchlist.

The scoreboard

Every thesis is timestamped with an entry price and tracked against the S&P 500. Nothing is edited after the fact; the losers stay visible. The exit signal is journalistic, not a price target — when the story has been written everywhere, the information edge is gone and the thesis closes. That's faithful to Camillo: you buy the information gap and sell the headline.

What every issue contains

— The Briefing: the editor's synthesis and the week's meta-trend

— The Theses: 2–4 tickered ideas, each with its kill list

— Dispatches: one per beat, the raw observational layer

— Too Late: real trends already fully priced — the anti-chase section

— The Watchlist: real trends with no clean public vehicle yet

— The Scoreboard: every call we've ever made

— What We Passed On: the reject pile, in the open

Published Sunday evenings. US-listed equities only. No sponsored content, no affiliate links, no positions in anything we name.