A derivative based estimator for semiparametric index models

Bas Donkers and Marcia Schafgans

No EI 2003-08, Econometric Institute Research Papers from Erasmus University Rotterdam, Erasmus School of Economics (ESE), Econometric Institute

Abstract: This paper proposes a semiparametric estimator for single- and multiple index models. It provides an extension of the average derivative estimator to the multiple index model setting. The estimator uses the average of the outer product of derivatives and is shown to be root-N consistent and asymptotically normal. Unlike the average derivative estimator, our estimator still works in the single-index setting when the expected derivative is zero (symmetry). Compared to other estimators for multiple index models, the proposed estimator has the advantage of ease of computation. While many econometric models can be regarded as multiple index models with known number of indices, our estimator in addition provides for a natural framework within which to test for the number of indices required.

Keywords: average derivatives; index models; outer product of derivatives; rank testing; semiparametric estimation (search for similar items in EconPapers)
Date: 2003-03-26
References: View references in EconPapers View complete reference list from CitEc
Citations: View citations in EconPapers (3)

Downloads: (external link)
https://repub.eur.nl/pub/1698/feweco20030326115042.pdf (application/pdf)

Related works:
Working Paper: A Derivative Based Estimator for Semiparametric Index Models (2003)
Working Paper: A Derivative Based Estimator for Semiparametric Index Models (2003)
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:ems:eureir:1698

Access Statistics for this paper

More papers in Econometric Institute Research Papers from Erasmus University Rotterdam, Erasmus School of Economics (ESE), Econometric Institute Contact information at EDIRC.
Bibliographic data for series maintained by RePub ( this e-mail address is bad, please contact ).