What the graph suggests — hypotheses that tie companies together through what the layer's graph knows they share

Built only from the relational graph: the members of Congress who traded several of them, the funds that hold several, the posts that name several, the lenses that landed on several in the same week. The app never stated a relationship; every link here was inferred by the data layer from ordinary reads and writes. Each hypothesis carries a check and a date, and is judged when it comes due. Not advice.

show: all · baskets · clusters · lead–lag · divergences

No hypotheses yet. The engine runs every 30 minutes over the week's hottest companies; a hypothesis needs two or more companies that share actors across two or more lenses.