Beyond Forecasting: The Belief-to-Trade Layer in Prediction-Market Agents
Quick Answer
Raven-Agent is introduced as the first autonomous trading agent for prediction markets, demonstrating the only positive return and positive risk-adjusted return among tested policies in a controlled replay.
Quick Take
Raven-Agent is introduced as the first autonomous trading agent for prediction markets, demonstrating the only positive return and positive risk-adjusted return among tested policies in a controlled replay. This highlights the gap between forecasting and trading performance, emphasizing the need for a belief-to-trade layer in AI models.
Key Points
- Raven-Agent is the first autonomous trading agent for prediction markets.
- Achieved the only positive return among all tested trading policies.
- Demonstrated the need for a belief-to-trade layer in AI models.
- Highlights the gap between calibrated probability scores and trading results.
- Code for Raven-Agent has been released for public access.
Paper Resources
📖 Reader Mode
~1 min readAbstract:Forecasting future events has attracted growing attention as a testbed for general-purpose AI. A natural way to ground this evaluation is let the models trade in the prediction markets. Trading, however, requires more than forecasting. Moreover, recent benchmarks report a substantial gap between calibrated probability scores and the trading results. We propose Raven-Agent, to the best of our knowledge, the first autonomous trading agent for prediction markets. On a controlled replay over an archived decision set, our architecture achieves the only positive return and the only positive risk-adjusted return among all tested policies. We have released our code in this https URL .
| Comments: | 10 pages, 4 figures |
| Subjects: | Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2607.03015 [cs.AI] |
| (or arXiv:2607.03015v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2607.03015 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Yuxuan Wang [view email]
[v1]
Fri, 3 Jul 2026 06:49:48 UTC (813 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.AI
See more →Procedural Memory Distillation: Online Reflection for Self-Improving Language Models
Procedural Memory Distillation (PMD) enhances reinforcement learning by converting cross-episode signals into reusable memory, improving Qwen3-8B and OLMo3-Instruct-7B models by 3.8-5.5% on SCIKNOWEVAL and 7.9-13.6% on . The co-evolution of policy and memory allows for more effective self-supervision, demonstrating significant performance gains when both components are active.