Investment research has a memory problem. Calls that worked are retold with sharpened timestamps; calls that failed quietly disappear from the archive. Nobody needs to lie for the record to bend — a redated post here, a deleted note there, and a track record becomes a marketing document.
ANTIKYTHERA publishes under a different constraint. From this note forward, every piece of research and every investment call on this site is filed, not merely posted:
The consequence is simple and permanent. We cannot backdate a call, soften a thesis after the fact, or delete a mistake without breaking every hash that follows it and contradicting anchors that live on other people's servers. The record can be embarrassing, but it cannot be quietly edited — which is exactly what makes it worth reading.
When we say a call was made at a price on a date, you can check it — in your browser on the
Verify page, or offline by cloning the repository and recomputing
the chain with python3 scripts/verify.py. When our thinking is wrong, the wrongness
stays on the ledger next to everything else. Over time the chain becomes the only honest kind of
track record: one that was evidenced before the outcomes were known.
Long-form research on the machinery of markets — capital cycles, incentive structures, the gearing between narrative and cash flow — filed to the Research register; specific calls filed to the Ledger with the price at the moment of the call; and a public Scoreboard of the managers we measure ourselves against.
The first analogue computer modelled the heavens with bronze gears and patience. We intend to study capital the same way: mechanically, measurably, and in the open.