|
Activity Number:
|
447
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 6, 2008 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Nonparametric Statistics
|
| Abstract - #301180 |
|
Title:
|
A Simple Semiparametric Method for Estimating ARCH Models
|
|
Author(s):
|
Li Wang*+ and Lijian Yang and Jianhua Huang
|
|
Companies:
|
The University of Georgia and Michigan State University and Texas A&M University
|
|
Address:
|
223 Statistics Building, Athens, GA, 30602,
|
|
Keywords:
|
B-spline ; foreign exchange returns ; knots ; news impact curve ; semiparametric regression ; volatility
|
|
Abstract:
|
For the past two decades, stochastic volatility models have been extensively investigated in various ways. Among those models, the non/semi parametric (G)ARCH model is particularly interesting due to the great flexibility of the models to be used without restricting a certain shapes of the volatility functions. Linton and Mammen (2005) considered a class of semiparametric ARCH models. They proposed an estimation method based on kernel smoothing and profiled likelihood, which is theoretical in nature with an analysis that relied on a complicated solution of a linear Fredholm integral equations. To make the semiparametric ARCH models more appealing, we developed a simple semiparametric method based on polynomial splines and an fast algorithm to implement the method in practice. The estimation procedure has been illustrated by the daily foreign exchange return data.
|
