Stevens Institute of Technology; The Fin AI; Columbia University; Harvard University, 2024
InvestorBench
Backtested single-asset trading decisions by an LLM agent framework (with memory) across stocks, crypto and an ETF, with many LLMs swapped in as the backbone.
About this benchmark
- Size
- 3 tasks: 7 stocks, 2 cryptos (BTC, ETH), 1 ETF (NIFTY-based); 13 LLM backbones
- Public split
- Paper says datasets and environments are open source; no repository URL found on any fetched primary page
- Grading
- Cumulative return and Sharpe ratio vs buy-and-hold
- Licence
- MIT (data, per paper's ethics statement)
- Submission
- none
- Products allowed
- Not stated
- Latest version
- ACL 2025 Long Papers (July 2025); arXiv v1 (2024-12-24)
- What it does not cover
- Live or contamination-free periods (test windows are 2020-2023 history) and multi-asset portfolios
Results
10 modelsEach score shows who reported it: the benchmark's authors (quoted from their paper), the benchmark's official board, or Moneda (run with the benchmark's own grader and rules). We do not copy other aggregators.
InvestorBench ACL 2025 Long Papers (July 2025); arXiv v1 (2024-12-24)
cumulative return %, higher is better. 10 models.
Updated 24 Dec 2024
InvestorBench ACL 2025 Long Papers (July 2025); arXiv v1 (2024-12-24)
cumulative return %, higher is better. 10 models.
Updated 24 Dec 2024
- 1Qwen2.5-72B-Instruct46.2
- 2GPT-443.7
- 3GPT-4o39.0
- 4Llama-3.1-70B-Instruct38.9
- 5Yi-1.5-34B-Chat38.0
- 6Qwen-2.5-Instruct-7B29.5
- 7DeepSeek-V2-Lite (15.7B)28.7
- 8DeepSeek-67B-Chat26.9
- 9Llama-3.1-8B-Instruct25.5
- 10GPT-o1-preview25.1
| 1 | Qwen2.5-72B-Instruct | Qwen | 46.2 | stock task, average over 7 stocks, 2020-10-01 to 2021-05-06; buy-and-hold 34.099 | Paper Table 2 Reported by the authors, 24 Dec 2024 |
|---|---|---|---|---|---|
| 2 | GPT-4 | OpenAI | 43.7 | stock task, 7-stock average | Paper Table 2 Reported by the authors, 24 Dec 2024 |
| 3 | GPT-4o | OpenAI | 39.0 | stock task, 7-stock average | Paper Table 2 Reported by the authors, 24 Dec 2024 |
| 4 | Llama-3.1-70B-Instruct | Meta | 38.9 | stock task, 7-stock average | Paper Table 2 Reported by the authors, 24 Dec 2024 |
| 5 | Yi-1.5-34B-Chat | Not stated | 38.0 | stock task, 7-stock average | Paper Table 2 Reported by the authors, 24 Dec 2024 |
| 6 | Qwen-2.5-Instruct-7B | Qwen | 29.5 | stock task, 7-stock average | Paper Table 2 Reported by the authors, 24 Dec 2024 |
| 7 | DeepSeek-V2-Lite (15.7B) | DeepSeek | 28.7 | stock task, 7-stock average | Paper Table 2 Reported by the authors, 24 Dec 2024 |
| 8 | DeepSeek-67B-Chat | DeepSeek | 26.9 | stock task, 7-stock average | Paper Table 2 Reported by the authors, 24 Dec 2024 |
| 9 | Llama-3.1-8B-Instruct | Meta | 25.5 | stock task, 7-stock average | Paper Table 2 Reported by the authors, 24 Dec 2024 |
| 10 | GPT-o1-preview | OpenAI | 25.1 | stock task, 7-stock average | Paper Table 2 Reported by the authors, 24 Dec 2024 |
Papers and links
Last checked 1 Nov 2026. Something out of date or wrong? Send a correction.