Cross-Section Regression Models of the Long-Run Relationship Between Market and Accounting Values - The Case for a Log-Linear Model

Willett, Roger J.; Falta, Michael

doi:10.2139/ssrn.1164644

Download This Paper

Open PDF in Browser

Add Paper to My Library

Cross-Section Regression Models of the Long-Run Relationship Between Market and Accounting Values - The Case for a Log-Linear Model

33 Pages Posted: 21 Jul 2008

See all articles by Roger J. Willett

Roger J. Willett

Victoria University of Wellington - Te Herenga Waka - Victoria Business School

Michael Falta

University of Otago

Date Written: July 21, 2008

Abstract

Basic annual cross sections of market on accounting values are estimated in levels and returns using a 30 US firm by 50 year panel of data. Log transformations of the levels data are shown to produce an improved statistical specification. Deflated returns models are shown to suffer from other problems of inference. The nature of the implicit assumption in cross section analysis that firm's in the sample have a common data generating processes is discussed.

Keywords: Cross section models, Capital market research, Misspecification, Dynamic models

JEL Classification: C21, C22, C23, D40, D49, G12, M41

Suggested Citation: Suggested Citation

Willett, Roger J. and Falta, Michael, Cross-Section Regression Models of the Long-Run Relationship Between Market and Accounting Values - The Case for a Log-Linear Model (July 21, 2008). Available at SSRN: https://ssrn.com/abstract=1164644 or http://dx.doi.org/10.2139/ssrn.1164644