LATIDIA · Investigación

Contagio en el piso de trading: cómo se propagan las señales adversarias en los sistemas de trading multiagente

arXiv: 2609.19789v1Tipo de anuncio: nuevo Resumen: Los sistemas de trading multiagente basados en modelos de lenguaje grandes (LLM) están comenzando a aparecer en las finanzas cuantitativas, sin embargo, su robustez ante las entradas adversarias es en gran medida desconocida

La noticia

arXiv:2609.19789v1 Announce Type: new 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