Risk-Neutral Compression and the Geometry of Latent Market Stress

Abstract

Public option surfaces can quote local surface coefficients without fixing the latent Perron loading that moves non-public claims. This paper formulates that boundary as risk-neutral compression. For a square-integrable claim or option-surface summary, the priced structural loading is its projection onto the latent Perron information field under the risk-neutral measure; a public information set observes only a conditional projection of that loading. The main theorem shows that public pricing summaries, option-surface transfers, signed physical/risk-neutral coefficient comparisons, nonlinear variance-compression gaps, and visible-hedging residuals all pass through the same projected object. Calibration of a surface and recovery of a latent loading are therefore separate operations, and public visibility becomes hedgeability only after projection onto the tradeable span. The SPX/NDX OptionMetrics study matched to Oxford-Man realized-volatility data estimates the finite public projection and finds option-surface diagnostics aligned with the rebuilt Perron stress direction.

Citation

Vidal Llauradó, Joan. “Risk-Neutral Compression and the Geometry of Latent Market Stress.” 2026. doi:10.2139/ssrn.6680938