Beyond the Event Horizon: Peak Risk-Adjusted Performance in Post-Event Markets

We develop a dynamic asset pricing model to analyze investor behavior around highuncertainty events such as earnings announcements and FDA approvals. Our key innovations include: (1) a two-risk framework distinguishing between directional news risk (uncertainty about event outcomes) and impact uncertainty (uncertainty about market response magnitude); (2) a threephase volatility process (pre-event rise, event-day peak, post-event dynamics) modeled through GJR-GARCH specifications; and (3) integration of heterogeneous investor beliefs and asymmetric transaction costs. Investors with mean-variance preferences trade an event-related asset and a generic risky asset in a multi-period framework. We solve for equilibrium prices with three investor types: informed, uninformed, and liquidity traders. Our model generates a testable hypothesis predicting that risk-adjusted returns, specifically return-to-variance and Sharpe ratios, peak during the post-event rising phase due to high volatility and biased expectations. Empirical validation using 2000-2024 data from earnings announcements and FDA approvals provides exceptionally strong support for our predictions, with return-to-variance ratios showing 4.4x amplification for FDA approvals and 9.5x enhancement for earnings announcements during the post-event rising phase. The framework provides insights for risk management and investment timing around high-uncertainty events. GitHub: https://github.com/brandonyee-cs/Event-Driven-Model.