Fin-Analyst at FinMMEval 2026 Task 3: A Live Hybrid Trading Agent with LLM Specialists and Rule-Based Signals
Quick Answer
This paper shows that Fin-Analyst, a hybrid trading agent for FinMMEval 2026, achieved a +13.51% return on Tesla (TSLA) and outperformed the Buy-and-Hold strategy with a Sharpe ratio of 4.10.
Quick Take
Despite a flat performance for Bitcoin (BTC), the agent's event-driven signals proved influential, highlighting the need for memory-aware models in volatile markets.
Key Points
- Fin-Analyst ranked first on TSLA with a +13.51% return.
- The agent achieved a Sharpe ratio of 4.10 and 88% win rate.
- BTC performance remained flat, indicating noise trading issues.
- Event-driven 8-K disclosures were identified as key TSLA signals.
- Memoryless agents repeated incorrect calls, necessitating -based improvements.
Paper Resources
📖 Reader Mode
~2 min readAbstract:Large language model (LLM) trading agents show promising performance in equity markets, yet remain narrowly focused on US equities with little evidence from live deployment. We present Fin-Analyst, a hybrid agent for FinMMEval 2026 Task 3: an eight-specialist LLM pipeline over news, SEC filings, fundamentals, analyst forecasts, technical indicators, and social sentiment, aggregated by a Meta-Agent for Tesla (TSLA), and a lightweight rule based three-signal vote for Bitcoin (BTC). On the final official leaderboard (accessed 2026-07-05), Fin-Analyst ranks first of all agents on TSLA with a +13.51% return, +28.33 points over Buy-and-Hold (Sharpe 4.10, 88% win rate), while the BTC vote ends flat yet well above a sharply falling baseline. Relative to the interim performance, the asset ranking reversed, indicating that short live windows yield volatility-sensitive rankings. Ablation identifies event-driven 8-K disclosures as the most influential TSLA signal. Error analysis shows that the memoryless agents repeat wrong calls for days at a time, and that the fixed-threshold BTC rules lost money by trading on noise in a sideways market while the LLM pipeline gained under similar conditions, motivating a memory-aware, LLM-based successor for both assets.
| Comments: | 14 pages, 7 tables, 1 figure. CLEF 2026 FinMMEval Task 3 Working Notes |
| Subjects: | Computation and Language (cs.CL); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2607.12233 [cs.CL] |
| (or arXiv:2607.12233v1 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2607.12233 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Lingzi Hong [view email]
[v1]
Tue, 14 Jul 2026 00:27:07 UTC (155 KB)
— Originally published at arxiv.org
Want this in your inbox every morning?
Daily brief at your local 8am — bilingual EN/中文, free.
More from arXiv cs.CL
See more →TriAgent: Divergence-Aware Committees for Cost-Efficient Financial Sentiment Analysis
TriAgent introduces a cost-efficient multi-agent system for financial sentiment analysis, combining VADER, FinBERT, and Qwen2.5. It achieves an F1 score of ~0.87 with significant savings of $9.3M/year at a 10M-user scale compared to GPT-4o-mini, while also detecting hallucinations with an AUC of 0.90.