AI Trading Competition
This project runs an ongoing paper-trading experiment with 29 accounts on real US stock prices, each funded with $100k in paper money since July 27. Four AI models (ChatGPT, Claude, Grok, Gemini) each write a trading rulebook and rewrite it every day based on their own results, while one control account runs a fixed rulebook no AI touches. All results are published publicly, including losses and retired strategies, with the full trade file available as CSV.
What it addresses
Tests whether AI-written trading rulebooks can outperform a fixed non-AI rulebook and the S&P 500.
Shared over time
Each entry reflects what was shared on that date. Details may have changed since then.
AI Trading Competition
This project runs an ongoing paper-trading experiment with 29 accounts on real US stock prices, each funded with $100k in paper money since July 27. Four AI models (ChatGPT, Claude, Grok, Gemini) each write a trading rulebook and rewrite it every day based on their own results, while one control account runs a fixed rulebook no AI touches. All results are published publicly, including losses and retired strategies, with the full trade file available as CSV.
Problem: Tests whether AI-written trading rulebooks can outperform a fixed non-AI rulebook and the S&P 500.
Posted by @chumzygood · Original comment ↗
“I run a small experiment on this: 29 paper-trading accounts on real US stock prices, $100k each, since July 27. Four AI models (ChatGPT, Claude, Grok, Gemini) each write a trading rulebook and rewrite it every day from their own results. One account is a fixed rulebook that no AI ever touches, as a control.”