Financial Stability Implications of Generative AI: Taming the Animal Spirits
A new study shows that large language models, a type of generative AI, could make financial markets more stable by avoiding herd behavior that often causes bubbles and crashes. In lab experiments, AI agents made more rational trading decisions than human professionals, relying on private information and making fewer mistakes. However, when signals were confusing, AI showed human-like biases. The authors warn that real-world factors and regulations could affect these outcomes, calling for more research.
What it examines
This paper studies how generative AI, especially large language models, affects financial stability by comparing AI agents' trading decisions to those of human professionals. Using lab-style experiments, it explores whether AI reduces herd behavior and asset bubbles, focusing on rationality and decision-making in financial markets.
What it concludes
AI agents show more rational trading and less herd behavior than humans, which could mean fewer market bubbles and more stable financial markets. These findings suggest AI-powered trading tools may improve market discipline, but further research is needed to understand long-term effects and guide safe AI adoption in finance.
Evidence objects
Groundbreaking research reveals that large language models (LLMs) in finance make more rational, information-driven trading decisions than humans, avoiding herd behavior that often triggers asset bubbles and financial crises.
key_findings bullet 1 · key_findings · validation V0
Innovative lab-style experiments show AI agents commit fewer trading errors and resist unnecessary imitation, introducing the concept of 'AI aversion to herding'a surprising contrast to human professionals tendency to follow market trends.
key_findings bullet 2 · key_findings · validation V0
Despite their strengths, AI agents can inherit human-like biases when signals are counterintuitive, and the study warns that real-world market complexities and regulations may alter these promising trends, requiring further investigation.
key_findings bullet 3 · key_findings · validation V0
This paper uniquely applies classic behavioral finance experiments to LLM agents, offering a fresh micro-foundational analysis of AI decision-making in investment management. Its originality lies in exploring LLMs behavioral biases and rationality, not just predictive power, providing novel insights into financial stability, herd behavior, and future AI adoption in markets.
key_findings bullet 4 · key_findings · validation V0
Raw abstract and provenance
Abstract: This paper investigates the impact of the adoption of generative AI on financial stability. We conduct laboratory-style experiments using large language models to replicate classic studies on herd behavior in trading decisions. Our results show that AI agents make more rational decisions than humans, relying predominantly on private information over market trends. Increased reliance on AI-powered… ▽ More This paper investigates the impact of the adoption of generative AI on financial stability. We conduct laboratory-style experiments using large language models to replicate classic studies on herd behavior in trading decisions. Our results show that AI agents make more rational decisions than humans, relying predominantly on private information over market trends. Increased reliance on AI-powered trading advice could therefore potentially lead to fewer asset price bubbles arising from animal spirits that trade by following the herd. However, exploring variations in the experimental settings reveals that AI agents can be induced to herd optimally when explicitly guided to make profit-maximizing decisions. While optimal herding improves market discipline, this behavior still carries potential implications for financial stability. In other experimental variations, we show that AI agents are not purely algorithmic, but have inherited some elements of human conditioning and bias. △ Less
Source row: 821 · abstract type: unknown