Semiparametric diffusion estimation and application to a stock market index

Wolfgang Härdle, Torsten Kleinow, Alexander Korostelev, Camille Logeay, Eckhard Platen

Research output: Contribution to journalArticle

Abstract

The analysis of diffusion processes in financial models is crucially dependent on the form of the drift and diffusion coefficient functions. A new model for a stock market index process is proposed in which the index is decomposed into an average growth process and an ergodic diffusion. The ergodic diffusion part of the model is not directly observable. A methodology is developed for estimating and testing the coefficient functions of this unobserved diffusion process. The estimation is based on the observations of the index process and uses semiparametric and non-parametric techniques. The testing is performed via the wild bootstrap resampling technique. The method is illustrated on SP 500 index data.

Original languageEnglish
Pages (from-to)81-92
Number of pages12
JournalQuantitative Finance
Volume8
Issue number1
DOIs
Publication statusPublished - Feb 2008

Fingerprint

Stock market index
Diffusion process
Testing
Coefficients
Resampling
Financial models
Methodology
Wild bootstrap

Keywords

  • Bootstrap
  • Continuous-time financial models
  • Diffusion
  • Identification
  • Kernel smoothing
  • Semiparametric methods

Cite this

Härdle, Wolfgang ; Kleinow, Torsten ; Korostelev, Alexander ; Logeay, Camille ; Platen, Eckhard. / Semiparametric diffusion estimation and application to a stock market index. In: Quantitative Finance. 2008 ; Vol. 8, No. 1. pp. 81-92.
@article{9a06aaac9042483094a8b39b36de1e4f,
title = "Semiparametric diffusion estimation and application to a stock market index",
abstract = "The analysis of diffusion processes in financial models is crucially dependent on the form of the drift and diffusion coefficient functions. A new model for a stock market index process is proposed in which the index is decomposed into an average growth process and an ergodic diffusion. The ergodic diffusion part of the model is not directly observable. A methodology is developed for estimating and testing the coefficient functions of this unobserved diffusion process. The estimation is based on the observations of the index process and uses semiparametric and non-parametric techniques. The testing is performed via the wild bootstrap resampling technique. The method is illustrated on SP 500 index data.",
keywords = "Bootstrap, Continuous-time financial models, Diffusion, Identification, Kernel smoothing, Semiparametric methods",
author = "Wolfgang H{\"a}rdle and Torsten Kleinow and Alexander Korostelev and Camille Logeay and Eckhard Platen",
year = "2008",
month = "2",
doi = "10.1080/14697680601026998",
language = "English",
volume = "8",
pages = "81--92",
journal = "Quantitative Finance",
issn = "1469-7688",
publisher = "Routledge",
number = "1",

}

Semiparametric diffusion estimation and application to a stock market index. / Härdle, Wolfgang; Kleinow, Torsten; Korostelev, Alexander; Logeay, Camille; Platen, Eckhard.

In: Quantitative Finance, Vol. 8, No. 1, 02.2008, p. 81-92.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Semiparametric diffusion estimation and application to a stock market index

AU - Härdle, Wolfgang

AU - Kleinow, Torsten

AU - Korostelev, Alexander

AU - Logeay, Camille

AU - Platen, Eckhard

PY - 2008/2

Y1 - 2008/2

N2 - The analysis of diffusion processes in financial models is crucially dependent on the form of the drift and diffusion coefficient functions. A new model for a stock market index process is proposed in which the index is decomposed into an average growth process and an ergodic diffusion. The ergodic diffusion part of the model is not directly observable. A methodology is developed for estimating and testing the coefficient functions of this unobserved diffusion process. The estimation is based on the observations of the index process and uses semiparametric and non-parametric techniques. The testing is performed via the wild bootstrap resampling technique. The method is illustrated on SP 500 index data.

AB - The analysis of diffusion processes in financial models is crucially dependent on the form of the drift and diffusion coefficient functions. A new model for a stock market index process is proposed in which the index is decomposed into an average growth process and an ergodic diffusion. The ergodic diffusion part of the model is not directly observable. A methodology is developed for estimating and testing the coefficient functions of this unobserved diffusion process. The estimation is based on the observations of the index process and uses semiparametric and non-parametric techniques. The testing is performed via the wild bootstrap resampling technique. The method is illustrated on SP 500 index data.

KW - Bootstrap

KW - Continuous-time financial models

KW - Diffusion

KW - Identification

KW - Kernel smoothing

KW - Semiparametric methods

UR - http://www.scopus.com/inward/record.url?scp=37248998954&partnerID=8YFLogxK

U2 - 10.1080/14697680601026998

DO - 10.1080/14697680601026998

M3 - Article

VL - 8

SP - 81

EP - 92

JO - Quantitative Finance

JF - Quantitative Finance

SN - 1469-7688

IS - 1

ER -