Contagion on the Trading Floor: How Adversarial Signals Spread in Multi-Agent Trading Systems

arXiv:2609.19789 · cs.AI · Submitted 2026-09-17 · Read on arXiv

cs.AI

Submitted: 2026-09-17

Updated: 2026-09-17

Comments: Published in ECML PKDD 2026. 25 pages, including 7 pages of supplementary material

Journal ref: Machine Learning and Knowledge Discovery in Databases. Research Track, ECML PKDD 2026, LNCS 16946, Springer (2027) 206-223

DOI: 10.1007/978-3-032-37673-2_12

License: http://creativecommons.org/licenses/by/4.0/

The gist: Multi-agent trading systems built on large language models (LLMs) are beginning to appear in quantitative finance, yet their robustness to adversarial inputs is largely unknown.

Terminology

Abstract

Multi-agent trading systems built on large language models (LLMs) are beginning to appear in quantitative finance, yet their robustness to adversarial inputs is largely unknown. We study the vulnerability of LLM trading stacks to black-box, input-only attacks that enter solely via admissible social-media feeds. We introduce the Generic Multi-Agent Trading System (GMATS), a framework that captures modern multiagent trading architectures and instantiate a class of black-box poisoning attackers that treat an LLM as a post generator and inject budget-constrained, plausibly benign social-media content into the analyst's evidence stream. We define contagion metrics that trace how adversarial content propagates through the stack, including belief-shift scores at analyst and coordinator layers and attack-clean deltas on standard backtest metrics. Experiments on a safe offline benchmark with historical market and social data show that even simple input-only attackers can materially degrade risk-return profiles, sharply reducing Sharpe ratios. At the same time, we find that suitably designed multi-agent topologies and coordinator prompts can dampen adversarial shocks and improve average robustness under identical poisoning budgets.

Sources

Related papers