Berry-Esseen Bound and Cramér-Type Moderate Deviation of the MLE for Ornstein-Uhlenbeck Process with Discrete Observations
摘要
In this paper, we study asymptotic properties of the approximated maximum likelihood estimator (MLE) for the drift coefficient in an Ornstein-Uhlenbeck process with discrete observations. By the change of measure method and asymptotic analysis technique, we establish an exponential nonuniform Berry-Esseen bound of the approximated MLE. Then, the Cramér-type moderate deviation can be obtained. As applications, the global and local powers for the hypothesis test are shown to approach one at exponential rates. Simulation experiments are conducted to confirm the theoretical results.