Submitted:
17 July 2018
Posted:
18 July 2018
You are already at the latest version
Abstract
This paper compares the finite sample performance of three non-parametric threshold estimators via Monte Carlo method. Our results show that the finite sample performance of the three estimators is not robust to the relative position of the threshold level along the distribution of threshold variable, especially when a structural change occurs at the tail part of the distribution.
Keywords:
difference kernel estimator
; integrated difference kernel estimator
; M-estimation
; Monte Carlo
; nonparametric threshold regression
Copyright: This open access article is published under a Creative Commons CC BY 4.0 license, which permit the free download, distribution, and reuse, provided that the author and preprint are cited in any reuse.
