aiqbench analysis
The paper introduces AStockArena, a benchmark that audits LLM trading agents on whether their reported trades are traceable, backed by evidence, and consistent with account states, rather than on returns alone. Its headline finding, that 72.99% of 1,070 trading points show at least one audit gap, is notable, though the consensus-based expert annotation and the 13-model sample limit how far it generalises. For AI readers, the useful point is that process-level evaluation of agent decisions may be a necessary complement to outcome metrics wherever LLMs act with real capital.
AStockArena: Beyond Returns, Can We Entrust Real Capital to the Decisions of Financial Agents?
arXiv:2611.00515v1 Announce Type: cross Abstract: Large language models (LLMs) are increasingly used in dynamic financial trading, yet real-time trading benchmarks mainly evaluate agents by cumulative returns and risk-adjusted performance. Such outcome metrics cannot show whether trades are faithfully reported, supported by available evidence, and consistent with account states. We introduce AStockArena, a real-time decision-auditing benchmark for continuous trading in China's A-share market. Under shared market conditions and execution constraints, we evaluate 13 open- and closed-source models over 3,198 sequential decisions. AStockArena reconstructsthe actions actually posted to accounts by aligning reports, trade requests, execution responses, and account states, and audits them with eight checks spanning execution traceability (O), evidence sufficiency (RK), and financial meta-awareness (B), combining record verification with consensus expert annotation. Of 1,070 trading points, 72.99% exhibit at least one audit gap; model-level rates range from52.2% to 91.2%, with evidence gaps most prevalent. Historical trade-filtering replay further shows thatportfolio outcomes are sensitive to these flagged trades. Decision-process auditing thus offers a verifiable complement to return-based evaluation of financial agents.
Title:AStockArena: Beyond Returns, Can We Entrust Real Capital to the Decisions of Financial Agents?
Abstract:Large language models (LLMs) are increasingly used in dynamic financial trading, yet real-time trading benchmarks mainly evaluate agents by cumulative returns and risk-adjusted performance. Such outcome metrics cannot show whether trades are faithfully reported, supported by available evidence, and consistent with account states. We introduce AStockArena, a real-time decision-auditing benchmark for continuous trading in China's A-share market. Under shared market conditions and execution constraints, we evaluate 13 open- and closed-source models over 3,198 sequential decisions. AStockArena reconstructsthe actions actually posted to accounts by aligning reports, trade requests, execution responses, and account states, and audits them with eight checks spanning execution traceability (O), evidence sufficiency (RK), and financial meta-awareness (B), combining record verification with consensus expert annotation. Of 1,070 trading points, 72.99% exhibit at least one audit gap; model-level rates range from52.2% to 91.2%, with evidence gaps most prevalent. Historical trade-filtering replay further shows thatportfolio outcomes are sensitive to these flagged trades. Decision-process auditing thus offers a verifiable complement to return-based evaluation of financial agents.
Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
arXiv:2611.00515 [cs.CL]
(or arXiv:2611.00515v1 [cs.CL] for this version)
https://doi.org/10.48550/arXiv.2611.00515 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 276 words