Dated, falsifiable investment predictions in targeted protein degradation, published before the events they describe and scored afterwards whether they were right or wrong.
No notes published yet. This repository currently carries the methodology, the tooling and the reference data; the first note is in pre-publication review. PROTOCOL.md is the methodology — start there.
PROTOCOL.md— the method: coverage tiers, note anatomy, quality gates, scoringreference/— gathered data: base rates, competitive benchmarks, clinical datamodels/— rNPV models as JSON, so every valuation is reproduciblescripts/— the tooling (SEC EDGAR, ClinicalTrials.gov, rNPV, Brier scoring)log/predictions.csv— the scorecard, filling in as notes publishARCHIVE.md— the track record, generated from the prediction loglog/catalyst-calendar.csv— dated catalysts under watch, each with a confidence level and the primary source the date came from
Companion Substack: SUBSTACK_URL_HERE — notes are published there and committed here the same day.
The commit timestamp is the evidence. Notes are never edited after publication; corrections are new notes that link to the original.
A note only appears here once it has cleared pre-publication review. Review records are published alongside the notes they cover, including the ones that stopped a note going out.
Research produced with AI assistance. Every view published here is one I have reviewed and stand behind, and every number carries a source.
For educational and analytical purposes only. Not investment advice, not a recommendation to buy or sell any security. The author is not a registered investment adviser. The author holds no position in any security discussed unless expressly stated.