On the relation between firm characteristics and volatility dynamics with an application to the 2007-2009 financial crisis

Despite the extensive literature on the analysis of firm equity volatility, relatively little is known about the relation between firm characteristics and volatility dynamics. This is partly due to the lack of an appropriate modelling framework in which these research questions can be addressed adequately. This work proposes a Hierarchical Factor GARCH model for multivariate volatility analysis in large panels of assets. The novelty consists of augmenting the dynamic specification with equations that link the volatility dynamics parameters of each firm to observed and unobserved characteristics. The hierarchical approach has features that are useful for both economic and forecasting applications. It permits one to investigate how variation of firm variables explains variation in volatility dynamics. Moreover, it allows for a parsimonious parameterization of a multivariate system that is independent of its dimension, yet capable of retaining flexibility in the individual series dynamics thanks to the random effect structure. The model is estimated via Maximum Likelihood. The proposed methodology is used to analyse the volatility dynamics of top U.S. financial institutions during the 2007-2009 crisis using a financia