65th ISI World Statistics Congress 2025

65th ISI World Statistics Congress 2025

Asymptotically efficient estimation for fractional Ornstein-Uhlenbeck processes under high-frequency observations

Author

MF
Masaaki Fukasawa

Co-author

  • A
    Alexandre Brouste
  • T
    Tetsuya Takabatake

Conference

65th ISI World Statistics Congress 2025

Format: IPS Abstract - WSC 2025

Abstract

Local Asymptotic Normality property for the fractional Ornstein-Uhlenbeck process under high-frequency observations is proved. A novel sharp error bounds for the cumulants of Gaussian
quadratic forms plays a key role in the proof. As a consequence, efficient rates for the joint estimation of all parameters are given and the maximum likelihood sequence of estimators is shown to be asymptotically efficient.