Title: Modelling realized covariance matrices by means of factor models
Authors: Roxana Halbleib - University of Konstanz (Germany) [presenting]
Giorgio Calzolari - University of Firenze (Italy)
Abstract: A latent factor model is proposed to capture the dynamics of daily realized covariance matrix series with forecasting purposes. The long memory in the series is captured by means of aggregating latent factors with short memory, where the factors are extracted from the common dynamics of realized variance and covariance series. Our approach accommodates the positive-definiteness of the variance-covariance matrix forecasts within a very parsimonious framework. For estimation purposes, we implement the quasi maximum likelihood approach applied on the Kitagawa state-space filtering procedure. We provide Monte Carlo evidence on the accuracy of the estimates and real data evidence on the very good performance of the model to forecast variance-covariance matrices one-step and multi-step ahead.