Exceptional-point criticality in real-world information flows

Exceptional points—spectral singularities at which both eigenvalues and eigenvectors of a non-Hermitian operator coalesce—are widely studied in quantum optics, photonics, and condensed matter physics. Here we report their observation in a completely different domain: the collective dynamics of information in financial order-flow systems. We show that the Fisher information metric of bid-ask order flow degenerates at market stress events, driving two-mode coalescence with the algebraic structure of an exceptional point. The theoretical prediction—a susceptibility divergence scaling as |Δ_eff|-1/2 and a complementary curvature spectral gap closing as |Δ_eff|+1/2—is observed over three decades in 5.28 million millisecond-resolution trades from the extreme liquidity collapse on 5 August 2024. These findings establish exceptional-point physics as a new universality class governing critical transitions in information-driven complex systems.