A0325
Title: Prediction, filtering, and smoothing for the integrated variance with intraday returns
Authors: Daisuke Nagakura - Keio University (Japan) [presenting]
Abstract: A state space method is proposed to calculate the linear projection of the integrated variance (IV) on squares of high frequency returns in the presence of market microstructure noise (MMN). Unlike most of the previous literature, we allow MMNs to be serially correlated. We also incorporate the leverage effects into the model. The method is based on a state space representation of squares of high frequency returns, in which the IV is one of the state variables. Our simple numerical experiment shows that our method performs better than other existing state space methods, in particular, in terms of forecasting.
