Beyond Linearity: Dynamic Evidence on Beta Instability and State-Dependent Risk Pricing

This paper reassesses the empirical validity of the Capital Asset Pricing Model (CAPM) by examining the structural instability of the Security Market Line (SML) under asymmetric volatility, endogenous regime transitions, and behavioral frictions. Using high-frequency data (2015–2024) from four major U.S. large-cap firms, we implement an integrated empirical framework combining quantile regression, rolling-window beta estimation, and a Markov-Switching AR (1) model. Our findings reveal pronounced distortions in the beta–return relationship, which becomes non-linear, time-varying, and contingent on latent market regimes. These results challenge the CAPM’s core assumptions of linearity, stationarity, and rational expectations. While the empirical scope centers on U.S. equities, the identified dynamics—rooted in behavioral asymmetries and volatility shocks—are particularly relevant for emerging markets, where such distortions are typically amplified. The study offers a diagnostic framework that may inform risk monitoring, portfolio management, and asset pricing in complex and unstable market environments, including those in emerging economies.