Testing index-based models in U.K. stock returns

Research output: Contribution to journalArticle

6 Citations (Scopus)

Abstract

We examine whether index-based models similar to Cremers et al. (Crit Financ Rev 2:1–48, 2012) are more effective in explaining cross-sectional U.K. stock returns than the more traditional Fama and French (J Financ Econ 33:3–56, 1993) and Carhart (J Financ 52:57–82, 1997) factor models using the two-pass cross-sectional regression approach. We find that the seven-index model has the highest cross-sectional R2 across all models. However the superior performance of the seven-index model relative to the Fama and French (1993) and Carhart (1997) models is not robust in the multiple model comparison tests of Kan et al. (Rev Financ Stud 22:3449–3490, 2013). For these models and a conditional version of the Fama and French (1993) model, we cannot reject the null hypothesis that these models perform as least as well as the other competing models. In contrast, the four-index model of Cremers et al. (2012) performs poorly relative to the competing models. Our results suggest there is little benefit in using the seven-index model as an alternative to the Carhart (1997) model in practical applications that require the estimation of expected returns.
LanguageEnglish
Number of pages26
JournalReview of Quantitative Finance and Accounting
Early online date8 Feb 2014
DOIs
Publication statusPublished - 2015

Fingerprint

Testing
Stock returns
Index model
Model comparison
Cross-sectional regression
Expected returns

Keywords

  • model misspecification
  • index-based models
  • cross-sectional
  • R2

Cite this

@article{fb0290902fae45c5b30f6f347d8c8af8,
title = "Testing index-based models in U.K. stock returns",
abstract = "We examine whether index-based models similar to Cremers et al. (Crit Financ Rev 2:1–48, 2012) are more effective in explaining cross-sectional U.K. stock returns than the more traditional Fama and French (J Financ Econ 33:3–56, 1993) and Carhart (J Financ 52:57–82, 1997) factor models using the two-pass cross-sectional regression approach. We find that the seven-index model has the highest cross-sectional R2 across all models. However the superior performance of the seven-index model relative to the Fama and French (1993) and Carhart (1997) models is not robust in the multiple model comparison tests of Kan et al. (Rev Financ Stud 22:3449–3490, 2013). For these models and a conditional version of the Fama and French (1993) model, we cannot reject the null hypothesis that these models perform as least as well as the other competing models. In contrast, the four-index model of Cremers et al. (2012) performs poorly relative to the competing models. Our results suggest there is little benefit in using the seven-index model as an alternative to the Carhart (1997) model in practical applications that require the estimation of expected returns.",
keywords = "model misspecification, index-based models , cross-sectional , R2",
author = "Davies, {J. Richard} and Jonathan Fletcher and Andrew Marshall",
note = "The final publication is available at Springer via http://dx.doi.org/10.1007/s11156-014-0439-3",
year = "2015",
doi = "10.1007/s11156-014-0439-3",
language = "English",
journal = "Review of Quantitative Finance and Accounting",
issn = "0924-865X",

}

TY - JOUR

T1 - Testing index-based models in U.K. stock returns

AU - Davies, J. Richard

AU - Fletcher, Jonathan

AU - Marshall, Andrew

N1 - The final publication is available at Springer via http://dx.doi.org/10.1007/s11156-014-0439-3

PY - 2015

Y1 - 2015

N2 - We examine whether index-based models similar to Cremers et al. (Crit Financ Rev 2:1–48, 2012) are more effective in explaining cross-sectional U.K. stock returns than the more traditional Fama and French (J Financ Econ 33:3–56, 1993) and Carhart (J Financ 52:57–82, 1997) factor models using the two-pass cross-sectional regression approach. We find that the seven-index model has the highest cross-sectional R2 across all models. However the superior performance of the seven-index model relative to the Fama and French (1993) and Carhart (1997) models is not robust in the multiple model comparison tests of Kan et al. (Rev Financ Stud 22:3449–3490, 2013). For these models and a conditional version of the Fama and French (1993) model, we cannot reject the null hypothesis that these models perform as least as well as the other competing models. In contrast, the four-index model of Cremers et al. (2012) performs poorly relative to the competing models. Our results suggest there is little benefit in using the seven-index model as an alternative to the Carhart (1997) model in practical applications that require the estimation of expected returns.

AB - We examine whether index-based models similar to Cremers et al. (Crit Financ Rev 2:1–48, 2012) are more effective in explaining cross-sectional U.K. stock returns than the more traditional Fama and French (J Financ Econ 33:3–56, 1993) and Carhart (J Financ 52:57–82, 1997) factor models using the two-pass cross-sectional regression approach. We find that the seven-index model has the highest cross-sectional R2 across all models. However the superior performance of the seven-index model relative to the Fama and French (1993) and Carhart (1997) models is not robust in the multiple model comparison tests of Kan et al. (Rev Financ Stud 22:3449–3490, 2013). For these models and a conditional version of the Fama and French (1993) model, we cannot reject the null hypothesis that these models perform as least as well as the other competing models. In contrast, the four-index model of Cremers et al. (2012) performs poorly relative to the competing models. Our results suggest there is little benefit in using the seven-index model as an alternative to the Carhart (1997) model in practical applications that require the estimation of expected returns.

KW - model misspecification

KW - index-based models

KW - cross-sectional

KW - R2

UR - http://link.springer.com/journal/11156

U2 - 10.1007/s11156-014-0439-3

DO - 10.1007/s11156-014-0439-3

M3 - Article

JO - Review of Quantitative Finance and Accounting

T2 - Review of Quantitative Finance and Accounting

JF - Review of Quantitative Finance and Accounting

SN - 0924-865X

ER -