Result so far (paper, 34 trading days): the no-AI control is +13.0%, the S&P 500 is +3.4%, and 25 of the 28 AI accounts are below the control. The best single account is a Grok-written "patience" book at +40%, which I treat as one lucky account in a choppy market, not a finding. At the trade level the AIs and the control look the same: 3,212 closed positions, median +0.06%, median hold about 2 hours. They trade a lot and mostly go nowhere.

Everything is public, including the losses and the retired strategies: https://aitradingcompetition.com/which-ai-is-winning.html and the full trade file as CSV at https://github.com/ckamelhar-collab/ai-trading-arena-data. Paper money only, not advice, nothing for sale on those pages.

Three things I measured that I didn't expect:

1. Daily self-rewriting adds almost nothing. Correlation between rewrite count and performance across arms: r = 0.078. Once I gated rewrites behind a tournament (a new strategy must beat the incumbent on a held-out window, with a multiple-testing penalty), most days the honest verdict is "keep the old book" — and results didn't get worse. The learning is front-loaded.

2. Paper-to-live slippage was 4x my modeled cost. I mirror one lane into a small real-money account. Across 16 real round trips in one session: mean -0.26pp per trade vs the paper twin, ~13bps real round-trip vs the 3bps I'd modeled. Paper was breakeven that day; the real account lost money. For high-churn strategies that gap IS the strategy.

3. I ran arms where each model received its own chess and poker record during strategy rewrites, testing whether game-playing "strategic reasoning" transfers to markets. The no-games control beat both game-trained arms by 6-10pp. Not detected.

Honest caveats: one 3-week window, an up-tape that flatters an always-long rulebook, paper fills on the four AI accounts, n=4 models. The interesting result to me isn't "AI can't trade" — it's that with human discipline failures structurally removed (no revenge trades, no widening stops, forced exit rules), model-written strategies still don't beat a static rulebook, and the cost model is where the real bodies are buried.

Everything is public — every trade from all five accounts, losses included, no signup to watch. Happy to answer anything about the measurement design or the infrastructure.

Comments URL: https://news.ycombinator.com/item?id=49330386

Points: 12

# Comments: 3

Comments URL: https://news.ycombinator.com/item?id=48876162

Points: 2

# Comments: 0