Title: Estimating Beta pricing models with or without an Intercept : results from simulations
Authors : Bänziger-Aiba, Armin
Gramespacher, Thomas
Published in : International Research Journal of Finance and Economics
Volume(Issue) : 2015
Issue : 140
Pages : 76
Pages to: 82
Publisher / Ed. Institution : EuroJournals
Issue Date: Oct-2015
License (according to publishing contract) : Licence according to publishing contract
Type of review: Peer review (Publication)
Language : English
Subjects : Simulation study; Two-pass approach; Asset pricing econometrics; Cross-section regression
Subject (DDC) : 330: Economics
Abstract: The two-pass (cross-sectional) regression approach is widely used for estimating risk premia and testing factor pricing models. We investigate on two problems of the methodology, namely the error-in-variable bias and the large standard errors of the estimates. To mitigate these effects, we propose to run the first- and/or second-pass regression without an intercept(constant), hence imposing the theoretical restriction of the factor pricing model. We use simulations to assess the (finite sample) properties of estimators in the different regression specifications. Findings show that the error-in- variables bias and the standard errors of the estimates can be reduced considerably. Nevertheless, the standard errors remain substantial, even in large samples, so that statistical inference remains a challenging endeavour.
Departement: School of Management and Law
Publication type: Article in scientific Journal
ISSN: 1450-2887
URI: https://digitalcollection.zhaw.ch/handle/11475/12734
Appears in Collections:Publikationen School of Management and Law

Files in This Item:
There are no files associated with this item.


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.