AI & ML Paradigm Challenge

A few dominant AI models are destroying the diversity of thought required for the global stock market to function correctly.

April 25, 2026

Original Paper

The Model-Efficient Market Hypothesis Price Discovery, Consensus Hallucination, and the Limits of Informational Efficiency in LLM-Mediated Markets

Jeevan Renjith

SSRN · 6605178

The Takeaway

Financial markets rely on millions of people having different opinions to find the true price of a stock. When everyone uses the same LLMs for analysis, that cognitive independence vanishes. This leads to a consensus hallucination where the market reacts to the AI's bias rather than reality. The Efficient Market Hypothesis breaks down when the wisdom of the crowd is replaced by a single algorithm. We may be entering an era of extreme market fragility driven by AI uniformity.

From the abstract

The Efficient Market Hypothesis (Fama, 1970) rests on an axiom so foundational that it is almost never named: cognitive independence. Prices are informative because investors process information through separate, uncorrelated reasoning processes; disagreement is the substrate of discovery. We argue that the diffusion of large language models across the investment industry dissolves this axiom. When a dominant fraction of portfolio managers, analysts, and research platforms delegates the step bet