Signature-Based Models - Theory, Calibration, and Expansions

Explore signature methods as a powerful non-parametric tool for extracting features from path-dependent data in this 68-minute webinar presented by Sara Svaluto-Ferro for the Society for Industrial and Applied Mathematics. Discover how signature methods provide a flexible framework for modeling complex behaviors in financial time series through two key applications: stochastic volatility modeling and moment expansions. Learn to use signatures to express both log-price and squared VIX in linear form, enabling closed-form solutions and accurate calibration for SPX and VIX options. Understand how to leverage the time-extended Itô signature to derive high-order expansions of conditional moments and characteristic functions, with practical applications in short-time asymptotics. Gain insights into advanced mathematical finance techniques and data-driven modeling approaches that are essential for researchers and practitioners working with quantitative finance and stochastic processes.