|
Activity Number:
|
515
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 5, 2009 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Business and Economic Statistics Section
|
| Abstract - #304458 |
|
Title:
|
Recursive Estimation Using Combined Optimal Estimating Functions
|
|
Author(s):
|
Melody Ghahramani*+ and Aerambamoorthy Thavaneswaran
|
|
Companies:
|
University of Winnipeg and University of Manitoba
|
|
Address:
|
, Winnipeg, MB, R3B 2E9, Canada
|
|
Keywords:
|
recursive ; combined estimating functions ; nonlinear time series ; GARCH ; information ; doubly stochastic
|
|
Abstract:
|
In this talk, we discuss combining estimating functions for the parameter of discrete time stochastic processes where the conditional mean and the conditional variance depend on the same parameter with applications to volatility models. The optimal combined estimating function for the parameter is shown to contain more Godambe information than each of the component estimating functions. As an application, nonlinear recursive estimation of the parameter in nonlinear time series models such as autoregressive processes with GARCH errors, based on the optimal combined estimating function is derived. Pre-filtered optimal estimation of the parameter of doubly stochastic time series based on the optimal combined estimating function will also be discussed.
|
