Doctoral thesis

Measuring and modelling realized volatility : from tick-by-tick to long memory

    06.07.2005

121

Thèse de doctorat: Università della Svizzera italiana, 2005 (jury note: summa cum laude)

English This study develops new realized volatility and correlation estimators which, while fully exploiting all the available information contained in tick-by-tick data, effectively correct for the bias induced by microstructure effects. Building on such high frequency measures, it also proposes new conditional volatility models able to provide accurate and easy-to-implement volatility forecasts.
Language
  • English
Classification
Economics
License
License undefined
Identifiers
Persistent URL
https://n2t.net/ark:/12658/srd1317904
Statistics

Document views: 68 File downloads:
  • 2005ECO004.pdf: 91