Accounting and Finance 47 (2007) 1–22
Blackwell Publishing, Ltd.Oxford, UKACFIAccounting and Finance0810-5391© The Authors Journal compilation © AFAANZ465
K. Ahmed, J. GoodwinK. Ahmed, J. Goodwin
An empirical investigation of earnings restatements by
, John Goodwin
School of Business, La Trobe University, Bundoora, 3086, Australia
School of Accounting and Law, RMIT University, Melbourne, 3000, Australia
From 1970 to 2003, we document earnings restatements for the top 500 Australian
ﬁrms, examine the characteristics of restating ﬁrms, and test whether restatements
are value relevant. Of the 195 earnings restatements, 49 per cent decrease prior-
period earnings (negative restatements). Negative restatements are relatively
larger than positive restatements. We identify three reasons for earnings restate-
ments; namely, accounting policy changes, revision of estimates, and errors and
unknown, and they comprise 49, 40 and 11 per cent of the sample, respectively.
Restatement ﬁrms have higher growth opportunities and are smaller than non-
restating ﬁrms from the same industry. Restatements are generally negatively
associated with market and non-market value.
: Earnings restatements; Value relevance; Firm characteristics
: M40, M41
In recent years, the frequency of US earnings restatements has increased,
attracting considerable attention from academics. Much of that attention has
been focused on the relation between earnings restatements and shareholders’
wealth. For example, studies have consistently found negative market reactions
from earnings restatements because of accounting errors (see Kinney and
McDaniel, 1989; Defond and Jiambalvo, 1991; Feroz
., 1991; Dechow
We appreciate the valuable comments from Max Aiken, Bruce Bennett, Craig Deegan,
participants at the RMIT University 2005 seminar series, the 2005 AFAANZ conference
and the two anonymous reviewers. Any errors remain our own.
Received 3 August 2005;
accepted 15 March 2006 by Gary Monroe (Deputy Editor).