The edge of biotech hedge funds: predictable trials, LLMs and subtle information extraction
owl_posting · x · 2026-11-21
Abhishaike Mahajan's first Works in Progress column examines where biotech hedge funds' edge comes from (Baker Brothers, Deerfield, OrbiMed each manage $15B+):
- A surprising share of clinical trial outcomes is predictable to anyone willing to look: many expensive trials have failed catastrophically in the past with little reason to succeed.
- This used to require an army of grad students compiling data; RA Capital even has a dedicated division, TechAtlas.
- LLMs can now replace much of that work; the remaining alpha may lie in a "game of increasingly subtle information extraction" plus context gleaned from talking to the right people.
Related event: New Column Examines How Biotech Hedge Funds Beat the Market(2 posts)→
More from Venture
- AI boogeyman stories make every stock unsafe: stalwarts down 40% in months — robleclerc · 2026-11-21
- Arcfront turns real hospital and factory work records into RL environments and frontier model evals — rohanpaul_ai · 2026-11-21
- Legacy companies are sitting on data gold now that hyperscaling is no longer the easy path — henloitsjoyce · 2026-11-21
- Nvidia plunges 13.17% in a single day to $126.75 — jfischoff · 2026-11-21
- Starcloud built a $2.5M space data center with the first H100 in orbit — chetanp · 2026-11-21
- "$NVDA looks like a memecoin chart," investor quips on wild swings — econoar · 2026-11-21