Premium
Across‐sample Incomparability of R 2 s and Additional Evidence on Value Relevance Changes Over Time
Author(s) -
Gu Zhaoyang
Publication year - 2007
Publication title -
journal of business finance and accounting
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.282
H-Index - 77
eISSN - 1468-5957
pISSN - 0306-686X
DOI - 10.1111/j.1468-5957.2007.02044.x
Subject(s) - relevance (law) , explanatory power , measure (data warehouse) , econometrics , value (mathematics) , sample (material) , scale (ratio) , contrast (vision) , mathematics , regression analysis , statistics , economics , computer science , philosophy , chemistry , epistemology , chromatography , artificial intelligence , political science , law , physics , quantum mechanics , database
Given the increasing popularity of across‐sample R 2 comparisons in accounting research, this paper illustrates why the regression R 2 s are incomparable across samples and the general nature of this problem. The regression residual dispersion with proper control for scale is proposed as the alternative measure of explanatory power for across‐sample comparisons. In market‐on‐accounting variable regressions, this measure can be conveniently interpreted as the degree of accounting‐based pricing errors and be used as a measure of value relevance of accounting information. As an application, the issue of over‐time value relevance changes is re‐visited. In contrast to prior mixed findings based on the R 2 measure, a decline of value relevance since the early 1970s is robustly detected using the alternative measure.