Individual dynamic risk modeling: integrating systematic and idiosyncratic structures with financial big data

Abstract

This paper introduces a dynamic risk model that incorporates both systematic and idiosyncratic structures. We call it the factor and idiosyncratic GARCH-X (FIGARCH-X) model. To get advantages of financial big data, the FIGARCH-X model incorporates a high-dimensional dynamic factor model. To overcome the curse of dimensionality, we introduce a parametric model for each individual stock. Specifically, we apply a GARCH-X structural model that incorporates unexplained systematic risk and integrated volatility of the index as its covariates. For the parameter estimation, we propose a two-step estimation procedure. In the first step, we estimate the latent factor components, and in the second step, we estimate the parameters of the individual stock model. We perform Monte Carlo simulations to validate the FIGARCH-X model and two-step estimation procedure. In the empirical study, we find that the integration of both factor and idiosyncratic structures helps accurately measure individual stock risk.