Anmerkungen:
Nach Informationen von SSRN wurde die ursprüngliche Fassung des Dokuments September 12, 2015 erstellt
Beschreibung:
We develop a volatility estimator that can be directly applied to tick-by-tick data. More specifically, we consider a model that allows for (i) irregular observation times that can be endogenous, (ii) dependent noise that can have diurnal features and be dependent on the latent price process, and (iii) jumps in the latent price process. We show that our estimator yields consistent estimates and enjoys the optimal rate of convergence. Simulation as well as empirical studies demonstrate favorable properties of our proposed estimator