I had Claude design a stock strategy and backtested it on a 2-year holdout
QuantAscent · reddit · 2026-10-16
A user shares a full writeup of having Claude design a low-beta stock strategy targeting S&P 500-like risk-adjusted returns, then backtesting on a 2-year holdout. Claude impressed by voluntarily holding out data for an "honesty check," starting at 2016 to avoid survivorship bias, and dropping momentum when data contradicted it. It faltered by clinging to pre-chosen metrics whose IC was noise. Results: near-index risk-adjusted returns, notably smaller worst drawdown, slightly under the index on Sharpe. Conclusion: AI is useful for building the strategy-creation process, but requires strong human pushback. Full numbers on Substack, 75-minute uncut session on YouTube.
More from AGI Musings
- Stanford and Databricks use AI agents to reproduce medical findings and test analytical robustness — ChengleiSi · 2026-10-23
- ISMS Fellow Avi Goldfarb: AI disruption raises returns to exploring research directions — avicgoldfarb · 2026-10-23
- signull on AI-found PHP exploits: economic liability is the forcing function legacy software needs — signulll · 2026-10-23
- TheStalwart: Talking computers are sparking a new philosophical boom, echoing ancient Greece — toptickcrypto · 2026-10-23
- AI cyberattacks could paradoxically accelerate AI development in an offense-defense arms race — kimmonismus · 2026-10-23
- Economist on AI detector Pangram: not a filter, but a useful verification signal — paulnovosad · 2026-10-23