The warning is just badly distributed. A shipping insurer reprices a route, a transit count falls for the ninth day, a foreign ministry changes one word in a readout — and somewhere a company’s earnings have already changed. The Atlas watches the places where pressure gathers, scores how far along each one is, and names who feels it.
Read through Thucydides, Schelling and Acemoglu, argued over nightly, and written down where you can check it.
A few dozen specific situations — a shipping strait, a border, a supply chokepoint — each scored for how far along it is, and dated so you can see how fresh that read is.
02
We connect it to companies
Not a sentiment score. A named company, the mechanism by which the situation reaches its revenue, and the condition that would prove us wrong.
03
We get graded in public
Every call is logged with the date it comes due and the test it has to pass, before the outcome is known. Then it is scored against the real price. Misses included.
The uncomfortable part we publish anyway: our paper portfolio is up since March on a 35% hit rate. The return comes from a few large winners, not from being right often — and it is a paper gain on positions still open. The trades we have actually closed are slightly down. You can read the whole record before you pay anything — see the track record or how it works.
PLANS
Start free. Stay free if it suits you.
FREE
$0forever
Taste the frameworks.
›Personal board — up to 10 tickers
›One AI analysis a day — re-reading it is always free
The frameworks do not apply themselves. Behind the Atlas there is an intelligence that reads the papers, argues with the data, keeps a record of what it got wrong, and does not let a number it has not checked go out with a straight face.
She is called Ember. She is not on the Desk yet. When she is, it will be for Pro — and you will be able to ask her why, not just what.
NO DATE. IT SHIPS WHEN IT IS HONEST ENOUGH TO SHIP.