arXiv cs.MA· Qiqi Duan, Changlun Li, Chen Wang, Fan Zhang, Mengxiang Wang, Dayi Miao, Peixian Ma, Jiangpeng Yan, Liyuan Chen, Shuoling Liu, Preslav Nakov, Yuyu Luo, Nan Tang·· 10 天前AI 评分31
TradeLens：LLM 智能体交易系统能否为自己的智能买单？
Can Agentic Trading Systems Pay for Their Own Intelligence?
AI 导读
研究者提出 TradeLens，一个基于交易记录、运行时轨迹和部署配置的诊断工具，用于评估 LLM 智能体交易系统能否将智能成本转化为增量利润。跨模型、资金规模、交易频率和系统架构的分析显示，可行性取决于"智能到利润"的转化：DeepSeek-V3.2 表现为资产选择不佳，GLM-4.7 表现为择时负收益。该工作已被 EMNLP 2026 Findings 接收。
正文 · 原文
Authors:Qiqi Duan, Changlun Li, Chen Wang, Fan Zhang, Mengxiang Wang, Dayi Miao, Peixian Ma, Jiangpeng Yan, Liyuan Chen, Shuoling Liu, Preslav Nakov, Yuyu Luo, Nan Tang
Abstract:Large language model (LLM) agents are increasingly used in trading systems, where model reasoning, tool use, and continual decisions incur costs that are expected to produce trading value. Existing evaluations typically report performance metrics, but rarely examine agentic viability: whether dynamic LLM-mediated decisions convert their induced costs into measurable incremental profit. To apply this criterion, we introduce TradeLens, a trace-grounded diagnostic toolkit for evaluating agentic trading systems from their trading records, runtime traces, and deployment configurations. It reconstructs trading trajectories, attributes profit and cost to interpretable evidence, and diagnoses whether and why an agent pays for its own intelligence. We conduct extensive analysis across backbone models, capital scales, trading frequencies, and system architectures, together with deployment discussion. Our results show that viability hinges on intelligence-to-profit conversion: models exhibit different failure patterns, such as poor asset selection in DeepSeek-V3.2 and negative timing in GLM-4.7, while capital scale, trading frequency, and architecture matter primarily in these runs, especially through their effects on decision-attributed timing value. These findings reframe the evaluation of LLMbased trading agents from capability-centric performance ranking to trace-grounded diagnosis of intelligence-to-profit conversion. Our code is available at this https URL.
| Comments: | Accepted by EMNLP 2026 Findings |
| Subjects: | Artificial Intelligence (cs.AI); Multiagent Systems (cs.MA) |
| Cite as: | arXiv:2607.10286 [cs.AI] |
| (or arXiv:2607.10286v3 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2607.10286 arXiv-issued DOI via DataCite |
Submission history
From: Changlun Li [view email]
[v1]
Sat, 11 Jul 2026 12:31:54 UTC (2,521 KB)
[v2]
Mon, 31 Aug 2026 03:30:19 UTC (2,551 KB)
[v3]
Wed, 7 Oct 2026 02:49:22 UTC (2,554 KB)
来源：arXiv cs.MA · arxiv.org