Title: Estimating long memory in volatility by means of factor models
Authors: Roxana Halbleib - University of Konstanz (Germany) [presenting]
Giorgio Calzolari - University of Firenze (Italy)
Aygul Zagidullina - University of Konstanz (Germany)
Abstract: A method is provided for modeling the long-memory of realized volatilities by means of factor models. Applying the model on large panels of realized volatilities increases the precision of the estimates and shows that the long memory of volatilities can be captured by the aggregation of short memory dynamic factors with different degrees of persistence. The model outperforms standard approaches, such as ARFIMA and HAR model both on simulated and real data.