A0987
Title: Unrestricted maximum likelihood estimation of multivariate realized volatility models
Authors: Jan Vogler - Ruhr Universitaet Bochum (Germany) [presenting]
Vasyl Golosnoy - Ruhr Universitaet Bochum (Germany)
Abstract: The popular conditional autoregressive Wishart (CAW) model for dynamics of realized covariance matrices provides a flexible parametrization. However, the number of parameters grows quadratically with the number of assets, which causes enormous computational difficulties in higher dimensions. Therefore, its unrestricted maximum likelihood (ML) estimation up to now has been conducted only for small portfolios with around five assets. It is elaborated on unrestricted ML estimation of the CAW model in higher dimensions abound 30 assets, which is a sufficient number for portfolio diversification. This can be done by providing various explicit analytical results for computing the gradient for log-likelihood optimization.