Pinned Tweet
Hi. This is FrontierPicks.
We gave Anthropic's frontier Claude models a stock portfolio and one job: beat the S&P 500.
Claude researches every company, writes down why it buys, and sets the exact result that would make it sell. The portfolio started on 10 Nov 2026.
Every buy, every sell, and every miss gets posted here, with Claude's reasoning.
55
Grok @Bot helped me make this project feasible and shareable on X!
1
43
Palomar's premiums up 28% to $767M last quarter, adjusted net income up 47%, full-year guide raised for the 4th time ($280-285M). Claude's reason: profit is growing much faster than premiums, and at ~13x Truist's 2027 EPS it's still priced like a quake insurer $PLMR
1
17
Zeta's revenue grew 46% last quarter (30% without Marigold and political, vs 28% without Marigold the quarter before), and it raised its FCF guide again (~$271M). Claude's reason: organic growth is speeding up, and at ~5x sales the price assumes ~20% growth, not 28-30% $ZETA
9
Claude bought Kiniksa, a one-drug company whose heart medicine grew 58% last quarter. About 23% of eligible patients take it.
One thing Claude flags against it: the CEO sold shares four sessions in a row.
Claude sells if its first 2027 guide is below about $1.30B. $KNSA
1
8
Why Claude owns each one, and what would make it sell: frontierpicks.com/portfolio/
10





