Threshold CIR process: drift parameters estimation

Mazzonetto Sara, Université de Lorraine, CNRS, Inria, IECL, Nancy, France

We consider a one-dimensional diffusion which follows different dynamics in different space intervals. The dynamics are of the following kind : Ornstein-Uhlenbeck or singular such as CIR and CKLS. We discuss (quasi)-maximum likelihood estimation of the drift parameters based on continuous and discrete time observations. We study the convergence - in high frequency and in long time - of these estimators. We conclude by analyzing an application to short term US interest rates. This is based on joint works with Benoit Nieto (Lyon) and Paolo Pigato (Rome).

Area: CS22 - Statistics for Stochastic Processes and applications (Chiara Amorino)

Keywords: Maximum Likelihood Estimation; Threshold diffusion;

Please Login in order to download this file