Review of Accounting Studies, 8, 213–220, 2003
# 2003 Kluwer Academic Publishers.Manufactured in The Netherlands.
Discussion of ‘‘Differential Market Reactions to
Revenue and Expense Surprises’’
PHILIP G. BERGER email@example.com
Graduate School of Business, University of Chicago, 1101 E. 58th St., Chicago, IL 60637
Abstract. Ertimur et al.(2003, this issue) study the difference in the market’s reaction to revenue versus
expense surprises.The discussion ﬁrst reviews their main ﬁndings and assesses the paper’s potential
contributions.Alternative explanations are then considered for the base-line result that the market reacts
more to revenue surprises than to expense surprises.The hypothesized reasons for revenue surprises to
matter more are critiqued, as are the tests of the hypotheses, and potential extensions that would link these
test to ﬁnancial statement analysis are suggested.Finally, two aspects of the assessment of how the
reaction to revenue and expense surprises differs across value and growth ﬁrms are discussed: The
deﬁnitions of value and growth ﬁrms and the potential beneﬁts of assessing why analyst revenue forecasts
are (not) observed for many value (growth) ﬁrms.
This discussion reviews Ertimur et al.’s (2003) analysis of the differential investor
reaction to revenue versus expense surprises at the preliminary earnings announce-
ment date.The next section reviews the main ﬁndings and Section 3 assesses the
paper’s potential contributions.Alternative explanations for the base-line results are
explored in Section 4, two issues raised by the comparison of value versus growth
ﬁrms are discussed in Section 5, and Section 6 concludes.
2. Main Findings
The base-line result replicates Swaminathan and Weintrop’s (1991) ﬁnding that the
market reacts more strongly to a dollar of surprise in revenues than to a dollar of
expense surprise.Three explanations for the stronger market reaction to sales
surprises are then pursued.The persistence explanation hypothesizes that investors
react more strongly to a sales surprise because these surprises are less transitory than
expense surprises.The homogeneity explanation proposes that the differential
investor reaction is attributable to revenues being more homogeneous than expenses,
which the authors argue may lead investors to perceive the revenue signal as less
noisy (for some unspeciﬁed reason).The earnings management explanation argues
the differential investor reaction could be attributable to accounting manipulations
being easier to implement and more difﬁcult to detect than manipulations of sales.
The authors claim to ﬁnd empirical support for all three of these explanations.
The paper’s other main thrust is to examine how the response to revenue and
expense surprises varies between value and growth ﬁrms.One main ﬁnding here is