On buffered threshold GARCH models


wiltsai - Posted on 15 January 2015

Project Description: 

This paper proposes a conditional heteroscedastic model with a new piecewise linear structure such that the regime-switching mechanism has a buffered zone where regime-switching is delayed. The Gaussian quasi-maximum likelihood estimation (QMLE) is considered, and its asymptotic behaviors, including the strong consistency and the asymptotic distribution, are derived. Its finite sample performance is evaluated by Monte Carlo simulation experiments, and an empirical example is reported to give further support to the new model.

Researcher name: 
Dr. Philip L.H. Yu
Researcher position: 
Associate Professor
Researcher department: 
Department of Statistics and Actuarial Science
Researcher email: 
Research Project Details
Project Duration: 
09/2014 to 06/2015
Remarks: 
Carry out intensive simulation to evaluate performance of the novel model