Authors

R. P. Byron

Date of this Version

December 1991

Document Type

Discussion Paper

Publication Details

Raymond P. Byron (1991) Differencing as a Test of Specification

School of Business Discussion Paper ; No. 17, Dec. 1991

© Copyright Raymond P. Byron and the School of Business, Bond University

Abstract

Extract:

Plosser and Schwert [1977,1978] and Plosser, Schwert and White [1982] introduced the idea of differencing a regression equation as a means of testing the specification of that equation. The idea is a nice one because of its simplicity - "if a model is correctly specified, the estimators from the differenced and undifferenced models should have the same probability limit." If the model is misspecified, regression should lead to different results.