2-Year Test Shows SOTA LLMs Suck at Stock Trading
garrytan · x · 2026-08-01
A deep analysis evaluating SOTA LLMs over a 2-year stock trading period yields negative conclusions. The study reveals that:
- Poor Performance: No model comes close to beating a simple static baseline.
- Reasoning Doesn't Help: Higher reasoning capabilities do not translate into better trading decisions.
- Behavioral Bias: When losing money, models tend to trade less frequently rather than optimizing their strategies.
Related event: LLMs Fail to Beat Equal-Weight Baseline in Stock Trading(2 posts)→
More from Research
- Stanford and GXL Host re:AGENT Hackathon for End-to-End Agentic Science — BrianHie · 2026-08-01
- OpenAI Highlights Ten AI-Driven Breakthroughs in Math and Theoretical CS — alphacolony21 · 2026-08-01
- Karpathy: LLMs Lack Cultural Accumulation and Self-Play — i_dg23 · 2026-08-01
- Reverse Engineering Claude's Tokenizer: Anthropic Isn't 'Bitter Lesson-pilled' — stanfordnlp · 2026-08-01
- Why Self-Play Works for Go but Struggles with LLMs — i_dg23 · 2026-08-01
- AgenticASR: Refining Real-World Speech Recognition via Agentic Approaches and Benchmark — solyarisoftware · 2026-08-01