Вы находитесь на странице: 1из 18

Accounting Research Journal

The effects of comprehensive income on investors’ judgments: An investigation of


one-statement vs. two-statement presentation formats
Ning Du Kevin Stevens John McEnroe
Article information:
To cite this document:
Ning Du Kevin Stevens John McEnroe , (2015),"The effects of comprehensive income on investors’
judgments", Accounting Research Journal, Vol. 28 Iss 3 pp. 284 - 299
Downloaded by The University of British Columbia Library At 04:17 03 February 2016 (PT)

Permanent link to this document:


http://dx.doi.org/10.1108/ARJ-11-2013-0083
Downloaded on: 03 February 2016, At: 04:16 (PT)
References: this document contains references to 30 other documents.
To copy this document: permissions@emeraldinsight.com
The fulltext of this document has been downloaded 370 times since 2015*
Users who downloaded this article also downloaded:
Mahfoudh Abdul Karem Mahfoudh Al-Musali, Ku Nor Izah Ku Ismail, (2015),"Board diversity and
intellectual capital performance: The moderating role of the effectiveness of board meetings",
Accounting Research Journal, Vol. 28 Iss 3 pp. 268-283 http://dx.doi.org/10.1108/ARJ-01-2014-0006
Michael Seamer, Adrian Melia, (2015),"Remunerating non-executive directors with stock options:
who is ignoring the regulator?", Accounting Research Journal, Vol. 28 Iss 3 pp. 251-267 http://
dx.doi.org/10.1108/ARJ-12-2013-0092
Teng Zhou, Jacqueline Birt, Michaela Rankin, (2015),"The value relevance of exploration
and evaluation expenditures", Accounting Research Journal, Vol. 28 Iss 3 pp. 228-250 http://
dx.doi.org/10.1108/ARJ-09-2013-0067

Access to this document was granted through an Emerald subscription provided by emerald-
srm:512739 []
For Authors
If you would like to write for this, or any other Emerald publication, then please use our Emerald
for Authors service information about how to choose which publication to write for and submission
guidelines are available for all. Please visit www.emeraldinsight.com/authors for more information.
About Emerald www.emeraldinsight.com
Emerald is a global publisher linking research and practice to the benefit of society. The company
manages a portfolio of more than 290 journals and over 2,350 books and book series volumes, as
well as providing an extensive range of online products and additional customer resources and
services.
Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the
Committee on Publication Ethics (COPE) and also works with Portico and the LOCKSS initiative for
digital archive preservation.
Downloaded by The University of British Columbia Library At 04:17 03 February 2016 (PT)

download.
*Related content and download information correct at time of
The current issue and full text archive of this journal is available on Emerald Insight at:
www.emeraldinsight.com/1030-9616.htm

ARJ
28,3
The effects of comprehensive
income on investors’ judgments
An investigation of one-statement vs.
284 two-statement presentation formats
Received 15 November 2013
Ning Du, Kevin Stevens and John McEnroe
Revised 14 May 2014 School of Accountancy and MIS, DePaul University, Chicago, Illinois, USA
Downloaded by The University of British Columbia Library At 04:17 03 February 2016 (PT)

11 August 2014
Accepted 21 August 2014

Abstract
Purpose – This paper aims to understand the effects of different presentation formats on
nonprofessional investors’ judgments. Both International Financial Reporting Standards and US
Generally Accepted Accounting Principles require an entity to present items of net income and other
comprehensive income (OCI) either in one continuous or in two separate, but consecutive, statements
but limited understanding exists about their differential effects on evaluation of company performance.
Design/methodology/approach – To investigate this research question, we used a two (Financial
Position) x two (Format) randomized between-subjects experiment. Ninety-four graduate students
assumed the role of investor and participated in this study.
Findings – Results of the experiment suggest that participants are more likely to incorporate OCI
information presented in the one-statement format than in the two-statement format. Further analysis
suggests that participants both assign more weight to OCI and perceive OCI to be relatively more
important in the one-statement format than in the two-statement format, especially when the entity
suffers an economic loss.
Originality/value – Results from this study provide evidence to the Financial Accounting Standards
Board and International Accounting Standards Board that should be useful in evaluating the
effectiveness of alternative comprehensive income reporting formats and should be of interest to
accounting rule-making bodies, investors, publicly traded entities and financial analysts, among others.
Keywords Judgment, Assimilation, Net income, Other comprehensive income,
Nonprofessional investors
Paper type Research paper

1. Introduction
Both the International Accounting Standards Board (IASB) and the Financial
Accounting Standards Board (FASB) currently require firms to disclose other
comprehensive income (OCI) in a performance-based reporting option and allow the
choice of reporting OCI either in one continuous statement (i.e. the statement of
comprehensive income) or in two separate, but consecutive, statements. US Generally
Accepted Accounting Principles (US GAAP) has historically allowed comprehensive
income (CI) to be reported in a nonperformance-based statement such as stockholders’
Accounting Research Journal
equity. Given the ample evidence suggesting that managers strategically choose the
Vol. 28 No. 3, 2015
pp. 284-299
equity format option for self-serving reasons (Bamber et al., 2010; Hunton et al., 2006; Lee
© Emerald Group Publishing Limited et al., 2006), it is not surprising that, in June 2011, the US FASB eliminated that format
1030-9616
DOI 10.1108/ARJ-11-2013-0083 and achieved convergence of guidance on the CI presentation between US GAAP (see
Accounting Standard Codification (ASC) Topic 220) and International Financial Effects of
Reporting Standards (IFRS; see IAS 1, 2010, Presentation of Financial Statements). comprehensive
Prior research on OCI mostly examines the difference between a performance-based
statement (e.g. the two-statement option) and a nonperformance-based statement (i.e.
income
owner’s equity). This study focuses on the performance-based option and examines the
difference between the one-statement and two-statement formats. In spite of reporting
the same total OCI number, these two formats present the OCI information in different 285
locations. Unlike the one-statement format, which combines net income (NI) and
components of OCI in one continuous statement, the two-statement approach, by
leaving the income statement in its current form, adds a second statement which reports
Downloaded by The University of British Columbia Library At 04:17 03 February 2016 (PT)

the components of OCI. Because the one-statement format tends to bury NI within CI (i.e.
a subtotal in the middle of a continuous statement of CI), many have criticized the
one-statement approach for awarding NI and OCI the same prominence in the income
statement in spite of their conceptual differences. Therefore, we want to understand
whether, compared to the two-statement approach, the one-statement presentation of
OCI may increase the salience of OCI while diminishing the importance of NI. In other
words, will the one-statement format, by presenting the NI and OCI in close proximity,
draw investors’ attention away from NI as the “bottom line” of a business and affect
nonprofessional investors’ evaluation of a company’s financial performance?
OCI is conceptually different from NI; NI summarizes the current financial
performance results of operating a company and is a clear indicator of current financial
performance, but OCI presents information about a company’s potential income and
cash flows from transactions generally to be finalized sometime in the future. Because of
its complexity, determining how to use and/or how much to use the OCI information is
not an easy task, especially for nonprofessional investors. As argued by Maines and
McDaniel (2000), because nonprofessional investors are limited in their understanding
of financial analysis, the format of the financial statement is important in that it provides
signals about the nature and importance of CI information (Maines and McDaniel, 2000,
p. 10). Thus, the location of OCI becomes an important cue for assessing the importance
of OCI information when investors evaluate a company’s financial performance.
Based on the inclusion-exclusion model in the psychology literature (Schwarz and
Bless, 1992), we argue that investors are more likely to incorporate OCI in their
evaluations of financial performance when OCI is represented in the one-statement, than
the two-statement, format. To test these hypotheses, we use a two (financial position) x
two (format) randomized between-subjects experiment. We manipulate financial
position as gain or loss of CI, and we manipulate format as one (continuous) statement or
two (separate) statements. Ninety-four graduate students assumed the role of investors
and participated in this study. Results of the experiment suggest that higher perceived
financial performance is related to the one-statement format, potentially because of the
close proximity of OCI and NI, which may have led to a higher assigned weight to OCI
in the performance evaluation. This effect is particularly strong when the company
suffers a financial loss.
This study contributes to the prior literature in several ways. First, this investigation
extends prior research on the reporting format of CI. Unlike prior studies, which tend to
focus on comparing the performance option and the nonperformance option, we
investigate the subtle differences between performance-based options. Our results
suggest that despite both being performance-based, the two currently allowed OCI
ARJ presentation formats may lead to different judgments among nonprofessional investors.
28,3 In addition, we find that the format effect is quite complex: though investors may prefer
the clarity of one continuous statement, the one-statement reporting format may dilute
the focus on NI as the most important performance measure of a company. Moreover,
this study extends Maines and McDaniel (2000) by providing evidence about how the
location of OCI information affects nonprofessional investors’ differential weighting of
286 OCI information across the different performance-based statement. Lastly, the results
from this study are timely and provide evidence to the FASB and IASB in evaluating the
effectiveness of alternative CI reporting formats and should be of interest to accounting
rule-making bodies, investors, publicly traded entities and financial analysts, among
Downloaded by The University of British Columbia Library At 04:17 03 February 2016 (PT)

others.

2. Literature review and hypothesis development


2.1 History of CI
Prior to 1997, OCI was not required to be reported anywhere in the financial statements,
and many of its components bypassed the income statement and went directly to
owners’ equity. In 1997, the FASB issued SFAS 130 (ASC Topic 220), Reporting
Comprehensive Income (FASB, 1997), requiring companies to report and display CI and
its components in a primary financial statement. FAS 130 (ASC Topic 220) defines CI as
consisting of NI and OCI. Other Comprehensive Income refers to revenues, expenses,
gains and losses that under US GAAP are included in CI but excluded from NI (i.e.
unrealized holding gains and losses on available for-sale securities, additional minimum
pension liability adjustments, currency translations, gains and losses of cash flow
hedges and asset revaluations). In 2007, the IASB published a revised version of the IAS
1 – Presentation of Financial Statements to bring OCI into line with the reporting
requirement of SFAS 130. Figure 1 presents a diagram illustrating the timeline of events
regarding OIC formats.
CI is derived from the concept of an all-inclusive income statement which refers to all
the changes in assets and liabilities other than those that involve transactions with
owners. It satisfies the desire of financial statement users for one figure, which includes
all components of income that lead to changes in the overall financial position of
organizations. Statement of Financial Accounting Concepts No. 6 (SFAC 6, 1985),
Elements of Financial Statements, defines CI as:
[…] the change in equity of a business enterprise during a period from transactions and other
events and circumstances from non-owner sources. It includes all changes in equity during a
period except those resulting from investments by owners and distribution to owners.
The concept of CI is closely related to the “clean” vs “dirty” surplus concept. Under the
clean surplus approach, all income items must pass through the income statement; they
sometimes are referred to as items that are reported above the line (the NI line). Thus,
earned surplus (equivalent to retained earnings) was “clean” of these items (Lee et al.,
2006). Under the dirty surplus approach, certain items bypass the income statement and
are directly reported in the statement of owners’ equity. These items were given the
name “Other Comprehensive Income (OCI)”. Accordingly, the notion of dirty surplus
includes items that are reported below the NI line.
The lack of a conceptual definition of the difference between NI items and OCI items
creates the OCI controversy. Most transactions recorded in OCI reflect changes in the
Effects of
In 1997, the FASB issued comprehensive
SFAS 130 (ASC Topic
220), requiring OCI to be
income
reported in one-statement,
two-statement or the
owners’ equity format.
287
In 2007, the IASB
published revised
IAS 1—
Downloaded by The University of British Columbia Library At 04:17 03 February 2016 (PT)

Presentation of
Financial
Statements,
requiring OCI to
be reported in
In 2011, the FASB issued one- statement or
ASU No. 2011-05 to two-statement
amend SFAS 130 (ASC
220), eliminating the
equity option for CI
reporting and requiring
OCI to be reported in one-
statement or two-
statement

Figure 1.
Timeline of events
regarding OCI
reporting

fair value of various assets and consist mostly of unrealized gains or losses driven by
external market factors. Thus, CI measures cash flow plus changes in net asset value.
Even though future income and cash flows are valuable information, nonprofessional
investors may not understand the nature of OCI and may have difficulty in viewing it as
a reliable performance measure. Recent standards, such as those on financial
instruments and post-employment benefits, increase the prominence of fair value as the
primary measurement attribute and thus expand the use of OCI. For example, requiring
companies to record changes in fair value for many financial instruments (including
most loans and issued debt) through OCI magnifies the difference between NI and OCI
in the financial statements. However, the more important conceptual issues such as
whether OCI is a performance measure, and concomitantly, what the items included in
OCI are intended to represent (i.e. whether they fulfill the criteria of income or expense
or, rather, represent other non-owner changes in equity) remain problematic and
controversial. Not surprisingly, the lack of a clear conceptual definition of what should
be reported in NI and what should be reported in OCI complicates the decision processes
of financial statement users. For example, despite recent convergence efforts, the FASB
and the IASB differ sharply in reclassifying (recycling) OCI items into NI. Under US
GAAP, amounts initially recognized in OCI are reclassified eventually into profit or loss,
ARJ whereas under IFRS different items are reclassified in different ways (e.g. actuarial
28,3 gains and losses related to employee benefit plans recognized initially in OCI are not
reclassified into profit or loss).

2.2 OCI reporting formats


288 The presentation of CI has always been controversial. In 1997, the FASB issued SFAS
130 (ASC Topic 220) which specified three options for reporting CI: one-statement,
two-statement or the equity format. The one-statement (combined statement of income
and CI) format displays the components of OCI below the NI total in a traditional income
Downloaded by The University of British Columbia Library At 04:17 03 February 2016 (PT)

statement. The two-statement approach presents each component of OCI in a separate


statement of CI that begins with NI. Both formats end with the same total CI. The equity
format, the most popular approach among corporate managers, reports CI as part of the
statement of changes in stockholders’ equity. While all three options were acceptable,
the FASB encouraged reporting OCI in the performance-based statement employing
either the one- or two-statement format and discouraged the use of the equity format
because, as the FASB concludes, the equity format lacks transparency by “hiding” CI in
a nonperformance-based statement. However, the IASB did not give companies as much
discretion as the FASB. In the revised IAS 1 (2007), IASB only allows the use of
performance-based reporting format to report OCI, i.e. the one-statement or
two-statement presentation but does not permit the equity option. On June 16, 2011, the
FASB (2011) issued Accounting Standards Update No. 2011-05, Presentation of
Comprehensive Income to amend ASC Topic 220, Comprehensive Income. This standard
eliminates the option to present items of OCI in the statement of changes in stockholders’
equity, but allows the choice of presenting items of NI and OCI either in one continuous
statement (i.e. the statement of CI) or in two separate, but consecutive, statements.

2.3 Prior academic research on OCI


Accounting researchers have been interested in understanding the effects of CI and its
presentation on users of financial statements. Prior archival researchers have examined
company’s actual choices in selecting CI reporting formats. Jordan and Clark (2002) and
Pandit and Phillips (2004) find the majority (more than 60 per cent) of the companies
used the equity option, followed by the one-statement format and then finally, the
two-statement format. In addition, their evidence shows that companies with negative
OCI were more likely to adopt the equity option, implying that managers choose a
nonperformance-based financial statement such as stockholders’ equity to “bury” the
loss in equity rather than having the loss prominently displayed in the income
statements.
Another stream of archival research focuses on the effects of presentation formats of
OCI information on the entity’s market return; however, the results are inconclusive. For
example, some studies find that CI is more relevant than NI as a performance indicator
because of its stronger correlation to stock return than NI (Biddle and Choi, 2006; Cahan
et al., 2000; Chambers et al., 2007). In contrast, Dhaliwal et al. (1999) documented no
evidence of a strong association between CI and US stock market returns. In a more
recent study, Bamber et al. (2010) investigated a broad cross-section of firms and find
that over 80 per cent of their sample S&P 500 firms report CI in the statement of equity.
They conclude, managers with stronger equity-based incentives and less job security
are concerned about volatility in CI when it is reported in the income statement, and, Effects of
thus, prefer the equity option. comprehensive
Behavioral researchers focus on the differential effects of reporting formats on the
judgments of financial statement users such as analysts and individual investors.
income
Hunton et al. (2006) examine whether the display of CI affects managers’ selective sale of
available-for-sale (AFS) securities to manage earnings. In their study, financial
executive participants prefer the less transparent equity format for earnings 289
management (EM), but when the authors require participants to report total CI in a
single continuous statement, participants are reluctant to engage in EM attempts. With
a focus on financial analysts’ stock valuations, Hirst and Hopkins (1998) compare the
Downloaded by The University of British Columbia Library At 04:17 03 February 2016 (PT)

equity to the two-statement approach. They find that analysts fail to identify unrealized
gain and loss (UGL) information for AFS when this information is presented in a
statement of stockholders’ equity, but that the reporting of UGL in a separate
two-statement of CI enhances financial analysts’ ability to detect EM through selective
sale of AFS.
Maines and McDaniel (2000) argue that the reporting format of CI is unlikely to affect
the judgments of more sophisticated market participants, such as financial analysts,
because analysts understand the role of CI in firm evaluation and will conduct an
extensive search for this information (Hunton and McEwen, 1997). In contrast, Maines
and McDaniel (2000) investigate nonprofessional investors’ responses to different CI
reporting formats. Ninety-five MBA students participated in their study and analyzed
an insurance company’s financial statements which included low and high volatility
produced by unrealized gains of AFS market securities. They find that participants are
sensitive to the negative effects of volatility on corporate and management performance
and placed significant weight on volatility when OCI is presented in a separate
two-statement format but not when it is in an equity format. They conclude that
the reporting format has no effect on investors’ acquisition and evaluation of OCI
information but does have an effect on the weighting of OCI information. In summary,
prior studies focus on the equity option, but none of the above studies investigates the
differences between the one-statement and two-statement formats.

2.4 Hypotheses development


In both one-statement and two-statement formats, the NI and OCI are shown together as
components of CI, included in the same CI category, and summed with the total
presented; however, the two formats differ significantly in where the information is
presented. Specifically, the one-statement format presents the breakdown of OCI
directly after the NI number, and both NI and OCI information is often presented on the
same page and appears in a continuous statement. In contrast, in the two-statement
format, the NI statement and the statement of CI (the second statement) are often
presented on different, separate pages. In a recent study, Hodge et al. (2010) examine
financial statement format effects with a focus on information proximity. Guided by the
proximity compatibility principle, they predict users will more effective and efficiently
integrate financial information that is presented in close proximity such as on a single
statement (or page) rather than on separate statements (Wickens and Carswell, 1995;
Carswell and Wickens, 1996). In their study, they presented financial information such
as cash flow and non-cash assets to graduate accounting students, and found that
presenting information on a single page results in reduced forecast errors and faster
ARJ learning in forecasting future cash flow. Results from Hodge et al. (2010) suggest that the
28,3 choice of reporting OCI in the one-statement versus the two-statement format will affect
investors’ judgments because these two formats differ in their information proximity.
The effects of information proximity can be explained by the Inclusion-Exclusion
Model proposed by Schwarz and Bless (1992 and 2007). That is, the way a piece of
information influences judgments depends on how it is categorized. The
290 Inclusion-Exclusion Model holds that assimilation effects result when the stimulus
information is included in the temporary representation that individuals form of the
target category; thus, judgment reflects the information that is included in the
representation used. For example, the addition of some positive information results in a
Downloaded by The University of British Columbia Library At 04:17 03 February 2016 (PT)

more positive judgment, whereas the addition of some negative information results in a
more negative judgment. Because individuals do not retrieve or use all knowledge, but,
rather, rely on the subset of potentially relevant information that is most accessible at
the time of judgment, the categorization of information is susceptible to contextual
influences. One such influence is the boundary of information. The inclusion-exclusion
model argues that fuzzier boundaries of stimulus and target, or a shorter distance
between these two, would promote assimilation because closeness implies that both
stimulus and target belong to the same category, and, therefore, increases the likelihood
of inclusion, whereas stronger boundaries and greater distance would lead to the
opposite effect. Evidence from prior accounting research also supports the linkage
between flexible categorization and fluid information assimilation. For example, Lipe
and Salterio (2000) document that direct links between information items help
decision-makers mentally “chunk” these items, thereby increasing the weight placed on
the items in forming judgments. Hopkins (1996) documents that financial analysts use
accounting classification to interpret information. In particular, his evidence suggests
that whether hybrid financial instruments are categorized as a liability or owners’
equity affects financial analysts’ common stock valuations.
Similarly, we argue that the placement and location of OCI affects how investors
categorize OCI information, in other words, whether or not they classify OCI in the
category of a performance measure. We propose that OCI is more likely to be assimilated
into the category of performance measure in the one-statement format because of its
close proximity to NI. Investors are familiar with the traditional income statement
format. When OCI is added directly below the NI in a traditional income statement,
investors are more likely to view OCI to be a similar measure as NI and, thus, categorize
OCI as a performance indicator. On the contrary, when OCI is included in a separate
statement, investors may have difficulty in classifying the nature of the indicator and
will not be sure whether it indeed belongs to the traditional income statement. Thus, OCI
in the two-statement format is less likely to be viewed as a similar measure to NI and,
therefore, will be less likely to be judged as performance related. Assuming a company
has positive OCI because of unrealized gains, if OCI is viewed as a performance measure,
the OCI will cast a more positive light on the company’s operations regardless of their
profits or losses in operating income. Therefore, our Hypothesis 1 focuses on the impact
of positive OCI information on investors’ evaluation of financial performance:
H1. Nonprofessional investors are more likely to judge that a company has higher
financial performance when positive OCI information is presented in the
one-statement format than in the two-statement format.
Nonprofessional investors have relatively ill-defined valuation models (Maines and Effects of
McDaniel, 2000); thus, most of them lack a clear understanding of the difference between comprehensive
NI items and OCI items and are not sure to what extent to rely on OCI in their
performance evaluation. In their study, they find that alternative OCI presentation
income
formats affect nonprofessional investors’ weighting of volatility information. Similarly,
we argue that when individuals form a judgment about operating results based on NI,
they construct a mental representation of corporate performance; the close association 291
between OCI and NI implies that both items belong to the same category which, in turn,
increases the likelihood of including OCI in a mental representation of the performance
judgment. Thus, the second hypothesis focuses on the weight assigned to OCI by
Downloaded by The University of British Columbia Library At 04:17 03 February 2016 (PT)

investors across different formats:


H2. Nonprofessional investors assign greater weight to OCI in the one-statement
format than in the two-statement format.

3. Method
3.1 Participants
The participants are 94 graduate students from a large private university in the USA.
These students are currently enrolled in the master of accounting program. We recruited
the students in their intermediate accounting course. Students volunteered to participate
in response to in-class announcements for extra course credit. Demographic information
collected at the end of the experiment indicates that most participants are likely to have
the necessary knowledge and experience to complete the task. For example, the majority
of participants have both work and investment experience. On average, participants
have about 1.83 years of working experience and 1.07 years of investment experience.
We also ask them to indicate how much knowledge they have about CI. The mean rating
is 5.52 on an 11-point scale (1 being very little, 11 being a great deal). The rating of 5.52
suggests that participants are not confident about their understanding of OCI, a result
we surmise reflects the general competence of individual nonprofessional investors.
Among the participants, 24 are males and 70 are females, representative of the graduate
school cohort. Despite the difference in gender numbers, we do not perceive this to affect
the results of our study (a MACOVA analysis indicates that the gender factor is not
significant).

3.2 Procedure and task


In the experiment, we asked the participants to assume the role of investor. Their task
was to evaluate the provided financial statements and assess the financial performance
of XYZ Company. The financial statements included NI and CI information, and were
constructed from OCI examples in PWC Dataline (www.pwc.com). In addition, the
instruction provides a short explanation about the difference between CI and NI by
including the following paragraph:
Comprehensive income includes both Net Income and Other Comprehensive Income. Net
income summarizes the current financial results of operating a company and reflects complete
transactions (or nearly complete) that produce currently available net earnings and cash flows
for use by a company’s management. However, most transactions recorded in Other
Comprehensive Income reflect changes in fair value and consist of unrealized gains or losses
driven by external market factors. Other Comprehensive Income consists of long-term and less
recurring items and may or may not affect future cash flows of an entity. In other words, Other
ARJ Comprehensive Income eventually may or may not be realized as Net Income in the future.
Unlike the Statement of Net Income, the Statement of Comprehensive Income includes both net
28,3 income and other comprehensive income.
After reviewing the provided information, participants made a series of judgments
about Company XYZ. Finally, participants finished by answering a short series of
manipulation check questions and were debriefed. The entire experiment took 20
292 minutes on average to complete.

3.3 Independent variables


Downloaded by The University of British Columbia Library At 04:17 03 February 2016 (PT)

We use a two (financial position) x two (format) between-subjects design. The financial
position variable is manipulated at two levels (gain or loss) and the format variable at
two levels (one-statement or two-statement). Participants are randomly assigned to
experimental conditions. We manipulate financial position as gain or loss of CI. The loss
condition is included for two reasons:
(1) to ensure that the predicted effects are consistent across profits and losses; and
(2) to ensure participants attend to the experimental stimulus and the differential
between gains and losses.

We achieve this by holding OCI constant at $840,000 but varying profit or loss in the
income statement. Both conditions have the same revenue of $3,550,000. We create a
mirror image of profit or loss by varying the expense section, with a NI of $1,400,000 for
the gain condition and a net loss of $1,400,000 in the loss condition. Thus, we have a CI
of $2,240,000 in the gain condition and $560,000 in the loss condition.
We manipulate the format variable as one continuous statement or two separate
statements. In the one-statement condition, OCI follows directly after NI and earnings
per share. The heading reads as “Statement of Comprehensive Income (XYZ Inc.) for
Year ended December 31, 2011”. All information is presented on one single page. The
two-statement condition contains two separate pages. The first page includes
“Statement of Net Income for Year ended December 31, 2011”. The second page includes
“Statement of Comprehensive Income for Year ended December 31, 2011”, which starts
with NI, followed directly by OCI. In both formats, the total of CI is presented at the end
of the statements.

3.4 Dependent variables


Dependent variables include the strength of financial performance and percentage
weight assigned to OCI. We also collected judgments on the degree of relatedness
between NI and OCI and the importance of OCI.

4. Results
4.1 Manipulation checks
To check the financial position and format manipulation, we ask participants to
examine the received information and list the amounts of NI or loss, OCI or loss and total
CI or loss for XYZ Inc. All participants identified the right amounts in their assigned
conditions.
4.2 Multivariate analysis of covariance Effects of
As the four dependent variables are potentially correlated, we rely on multivariate comprehensive
analysis of covariance (MACOVA) for the initial analysis. First, we examine the
assumption of equality of covariance matrices across treatment groups. The Box
income
statistic is significant with a Box’s M of 73.72 (F ⫽ 2.26, p ⬍ 0.05). However, we are not
concerned with violations of the equality assumption, because of our approximately
equal group sizes (Hair et al., 1998, p. 348). Thus, we proceed to the MACOVA analysis. 293
To control for any potentially extraneous influence on dependent variables, we include
gender, investment experience, work experience and the ratings on self-knowledge of
OCI as covariates in our analysis. The results of MACOVA are presented in Table I and
Downloaded by The University of British Columbia Library At 04:17 03 February 2016 (PT)

show that both independent factors are significant. Wilks’ lambdas are 0.82 (F ⫽ 4.90,
p ⬍ 0.01) for format and 0.41for gains or loss (F ⫽ p ⬍ 0.01). Their interaction are
marginally significant (Wilks’ lambda ⫽ 0.91, F ⫽ 2.06, p ⫽ 0.09). In addition,
MACOVA indicates none of the four control variables is significant. Specifically, for
self-knowledge, Wilks’ lambda is 0.95 (F ⫽ 2.85, p ⬎ 0.05); for work experience, Wilks’
lambda is 0.94 (F ⫽ 2.64, p ⬎ 0.05); for investment experience, Wilks’ lambda is 0.96
(F ⫽ 1.63, p ⬎ 0.05); for gender, Wilks’ lambda is 0.95 (F ⫽ 1.92, p ⬎ 0.05). Because of the
significant results of our independent variables, we follow up with analysis of variance
(ANOVA) for each of the dependent variables.

4.3 ANOVA
H1 predicts that nonprofessional investors are more likely to judge higher financial
performance when positive OCI information is presented in the one-statement format
than in the two-statement format. To test this hypothesis, we asked participants to
answer the question on an 11-point scale (1 being very weak, 11 being very strong):
“How strong is the company’s financial performance?” The mean ratings for financial
performance are provided in Table II and plotted in Figure 2(a). Table III presents the
results for ANOVA. If participants incorporate the positive OCI number in their
performance evaluation, the OCI information will positively affect their judgments. To
the extent that participants paid attention to the experimental stimulus, they would
perceive higher performance in the gain condition than in the loss condition. As
expected, the participants recognized the financial performance in the gain condition
(mean rating ⫽ 7.78) to be significantly higher than in the loss condition (mean rating ⫽
3.67) (F (1,90) ⫽ 122.06, p ⬍ 0.01). We also predicted that in the one-statement format, the
OCI information is more likely to be assimilated into NI; thus, participants are more
likely to categorize the OCI information as a performance measure. Because the total of

Source Wilk’s lambda F (3.90) Significance

Format 0.82 4.900 0.00


Gain or loss 0.41 31.98 0.00
Format by gain or loss 0.91 2.06 0.09

Note: This table shows the results of a MANOVA with four dependent variables: financial
performance, degree of relatedness, weight assigned to OCI and importance of OCI; we include gender, Table I.
investment experience, work experience and the ratings on self-knowledge of OCI as covariates in our Multivariate analysis
analysis of variance
ARJ Gain Loss
28,3 Tests of between-subjects One-statement Two-statement One-statement Two-statement
effects (H1 and H2) N ⫽ 25 N ⫽ 23 N ⫽ 22 N ⫽ 24

Weight assigned to OCI 34.8% (10.75) 38.96% (14.67) 37.73% (20.85) 25.31% (14.03)
Importance of OCI 7.16 (2.36) 5.70 (3.44) 6.91 (2.50) 5.88 (2.25)
294 Financial performance 8.04 (1.34) 7.09 (1.12) 4.45 (2.48) 2.95 (1.58)
Degree of relatedness 6.00 (2.20) 4.43 (2.27) 6.27 (2.22) 5.33 (2.57)

Notes: This table presents the means (std) for the dependent variables: rating for relatedness between
NI and OCI, percentage weight assigned to OCI, NI importance rating and rating for financial
Downloaded by The University of British Columbia Library At 04:17 03 February 2016 (PT)

performance; the dependent variables are based on the following four questions: Please indicate to what
degree net income and other comprehensive income are related?; In your evaluation of Company XYZ,
how did you weight the two different numbers – net income and other comprehensive income? Please
give a percentage about the attention you paid to each number. Note that both percentages should add
Table II. up to 100%; please rate the importance of Other Comprehensive Income when evaluating XYZ (on an
Means (variances) for 11-point scale); How strong is the financial performance of Company XYZ? (on an 11-point scale); the
dependent variables independent variables are Format, which is manipulated at two levels: one-statement and
(tests of H1 and H2) two-statements, and gain/loss

OCI is positive, the OCI will be perceived as a positive cue and affects participants’
evaluation positively. Thus, the assimilation process will result in higher perceived
performance in the one-statement than in the two-statement formats, regardless of profit
or loss condition. The results support this prediction as the average mean performance
rating in the one-statement condition (6.36) is significantly higher than in the
two-statement condition (4.98) (F(1.90) ⫽ 12.30, p ⬍ 0.01). The interaction between the
independent variables is not significant. Thus, the results support H1.
To provide further support to H1 we ask participants to evaluate the relatedness of
the NI and OCI. Specifically, we asked the following question on an 11-point scale:
“Please indicate to what degree Net Income and Other Comprehensive Income are
related? (1 being not at all related, 11 being very closely related)”. The mean ratings are
provided in Table II and plotted in Figure 2(b). Table III presents the results for the
ANOVA. The results of the ANOVA show a significant main effect for the format
variable but not for the gain/loss variable. Participants perceive NI and OCI to be
significantly more related to each other in one continuous statement (average mean ⫽
6.13) than two separate statements (average mean ⫽ 4.89) (F (1,90) ⫽ 6.81, two-sided
p ⫽ 0.01) regardless of gain or loss position. No significant interaction is observed for the
interaction of the two independent variables. Thus, the results further support H1.
H2 predicts that participants will assign greater weight to OCI in the one-statement
format than in the two-statement format. To test this hypothesis, we asked the following
question:
In your evaluation of Company XYZ, how did you weight the two different numbers – Net
Income and Other Comprehensive Income? Please give a percentage to each number. Note that
both percentages should add up to 100 per cent.
Thus, the dependent variable is the weight assigned to OCI. Table II provides the mean
percentages of OCI weights which are plotted in Figure 2(c). The overall weight is less
than 40 per cent. This suggests that, in general, participants recognize the importance of
Effects of
comprehensive
income

295
Downloaded by The University of British Columbia Library At 04:17 03 February 2016 (PT)

Figure 2.
Various judgments
regarding CI format

NI and pay much less attention to OCI information than NI. The results of ANOVA
presented in Table III show a non-significant format effect, but a marginal significant
effect of gain or loss (F(1,90) ⫽ 2.86, p ⫽ 0.09). Apparently, participants value OCI
slightly higher in the gain than loss condition, as they assigned a mean rating of 36.80
per cent in the gain condition and a mean rating of 31.25 per cent in the loss condition. A
close inspection of the graph suggests that in the loss condition, the two-statement
format generates the lowest rating of 25 per cent. This suggests that participants are
reluctant to rely on the positive number of OCI when a company suffers an economic
loss. In addition, the results of ANOVA show a significant interaction between the two
independent variables (F (1,90) ⫽ 6.84, p ⫽ 0.01). Given the significant interaction, we
conduct pairwise comparisons across gain and loss conditions. Results of pairwise
comparisons (see Table IV) suggest that in the gain condition, participants assign
similar weights to OCI as the difference between one- and two-statement formats is not
significantly different (mean difference ⫽ ⫺4.16 per cent, p ⬎ 0.05); however, in the loss
condition participants assign significantly higher weight in the one-statement than in
ARJ Source df MS F p (2-sided)
28,3
Future performance
Format 1 35.16 12.30 0.00
Gain/Loss 1 348.81 122.05 0.00
Format X gain/loss 1 1.72 0.60 0.43
296 Within cells (error) 90 2.85
Degree of relatedness
Format 1 36.771 6.81 0.01
Gain/loss 1 8.042 1.49 0.22
Downloaded by The University of British Columbia Library At 04:17 03 February 2016 (PT)

Format X gain/loss 1 2.296 0.42 0.51


Within cells (error) 90 2.858
Weight assigned to OCI
Format 1 399.75 1.69 0.19
Gain/loss 1 673.20 2.86 0.09
Format X gain/loss 1 1,609.66 6.84 0.01
Within cells (error) 90 235.24
Importance of OCI
Format 1 36.59 5.12 0.02
Gain/loss 1 0.030 0.00 0.94
Format X gain/loss 1 1.08 0.15 0.69
Within cells (error) 90 7.141

Notes: H1 predicts that nonprofessional investors are more likely to judge higher financial
performance when positive OCI information is presented in the one-statement format than in the
two-statement format. We test this hypothesis using the following question: How strong is the financial
performance of Company XYZ? (on an 11-point scale). The independent variables are Format, which is
manipulated at two levels: one statement and two statements, and gain/loss. We also use the following
question to test H1: Please indicate to what degree net income and other comprehensive income are
related? (on an 11-point scale); H2 predicts that nonprofessional investors assign greater weight to other
comprehensive income in the one-statement format than the two-statement format. We test this by the
following question: In your evaluation of Company XYZ, how did you weight the two different
numbers – net income and other comprehensive income? Please give a percentage about the attention
Table III. you paid to each number. Note that both percentages should add up to 100%. In addition, we ask
Tests of between- nonprofessional investors judge OCI information more important in the one-statement format than the
subjects effects (H1 two-statement format. We test H3 using this question: Please rate the importance of other
and H2) comprehensive income when evaluating XYZ (on an 11-point scale)

the two-statement format (mean difference ⫽ 12.42 per cent, p ⬍ 0.01). Apparently,
when the company suffers economic losses, participants are more likely to rely on the
positive OCI cue if it is presented in the one-statement format. Therefore, the
one-statement format indeed increases the salience of OCI information, but only in
the loss situation. Thus, the results partially support our Hypothesis 2.
To provide further support to H2, we asked the participants to rate the importance of
OCI on an 11-point scale (1 being not important, 11 very important). The mean ratings
are provided in Table II and plotted in Figure 2(c). Table III presents the results for the
ANOVA. The overall rating for OCI’s importance is 6.41 (below 7), suggesting that
participants do not perceive OCI as an important financial performance measure.
Results of ANOVA show a significant main effect for the format variable (F (1,90) ⫽ Effects of
5.12, p ⬍ 0.05) but no significant effect for gain or loss or the interaction of the two comprehensive
independent variables. Apparently, participants believe OCI to be more important in the
one-statement condition (mean ⫽ 7.04) than in the two-statement condition (mean ⫽
income
5.78). The results support H2.

297
5. Conclusions
The results of the experiment suggest that participants are more likely to
incorporate OCI information in the one-statement format than in the two-statement
Downloaded by The University of British Columbia Library At 04:17 03 February 2016 (PT)

format and are in support of the common concern that the one-statement format may
draw the attention of financial statement users away from NI to OCI as the “bottom
line” of a business, especially when the company suffers economic losses. However,
this experimental investigation, like others, is limited in several respects. We
manipulated the format as one continuous statement versus two separate
statements. In the two-statement format, the NI total is carried over from the
previous page. Even though the FASB does not require the statement of CI to begin
with NI, some firms may choose to report items of NI on the second statement of CI.
Future research could test whether differences in formatting within the
two-statement format make a difference to users. In addition, we manipulate
financial position by varying the sign of NI but holding OCI constant as a positive
figure. As people often react differently to gains and losses (Du, 2009), future studies
should investigate the format effects when OCI is negative. Moreover, this study
uses graduate accounting students rather than real investors. Because our
experiment requires the participants to have some, but not extensive, knowledge of
CI, we believe graduate accounting students are an appropriate proxy for
nonprofessional investors. However, we mention at this point that many accounting
and auditing studies have employed both undergraduate and graduate students
(Budescu and Du, 2007; Du, 2009; Du et al., 2011; Du and Budescu, 2005; Elliott et al.,
2007). Despite the extensive use of students in various behavioral experiments,
future studies should explore how real market participants react to different formats
of presenting OCI information.

Condition Comparison Mean Diff. Std. Sig. (2-sided)

Gain One vs. two-statement ⫺4.16 4.43 0.35


Loss One vs. two-statement 12.42 4.527 0.00
One statement Gain vs. loss ⫺2.93 4.48 0.52
Two statement Gain vs. loss 13.64 4.47 0.00
Table IV.
Notes: H2 predicts that the weight assigned to other comprehensive income will be more sensitive to Percentage weight
gain or loss for one-statement than two-statement condition. To test this hypothesis, we asked the assigned to other
following question: In your evaluation of Company XYZ, how did you weight the two different comprehensive
numbers – net income and other comprehensive income? Please give a percentage about the attention income (test of H2)
you paid to each number. Note that both percentages should add up to 100%. The dependent variable (Figure 2(c): Results
is weight assigned to other comprehensive income. The independent variables are Format, which is of pairwise
manipulated at two levels: one statement and two statements, and gain/loss comparisons)
ARJ References
28,3 Bamber, L., Jiang, J., Petroni, K. and Wang, I. (2010), “Comprehensive income: who’s afraid of
performance reporting?”, The Accounting Review, Vol. 85 No. 1, pp. 97-126.
Biddle, G.C. and Choi, J.H. (2006), “Is comprehensive income useful?”, Journal of Contemporary
Accounting and Economics, Vol. 2 No. 1, pp. 1-32.
Budescu, D.V. and Du, N. (2007), “The coherence and consistency of investors’ probability
298 judgments”, Management Science, Vol. 53 No. 11, pp. 1731-1744.
Cahan, S.F., Courtneay, S.M., Gronewoller, P.L. and Upton, D.R. (2000), “Value-relevance of
mandated comprehensive income disclosures”, Journal of Business Finance and
Accounting, Vol. 27 Nos 9/10, pp. 1272-1301.
Downloaded by The University of British Columbia Library At 04:17 03 February 2016 (PT)

Carswell, C.M. and Wickens, C.D. (1996), “Mixing and matching lower-level codes for object
displays: evidence for two sources of proximity compatibility”, Human Factors, Vol. 38
No. 1, pp. 1-22.
Chambers, D., Linsmeier, T., Shakespeare, C. and Sougiannis, T. (2007), “An evaluation of SFAS
No. 130 comprehensive income disclosures”, Review of Accounting Studies, Vol. 12 No. 4,
pp. 557-593.
Dhaliwal, D., Subramanyaman, K.E. and Tresevant, R. (1999), “Is comprehensive income superior
to net income as a measure of firm performance?”, Journal of Accounting and Economics,
Vol. 26 No. 1, pp. 43-67.
Du, N. (2009), “Do investors react differently to range and point management earnings forecasts?”,
The Journal of Behavioral Finance, Vol. 10 No. 4, pp. 195-203.
Du, N. and Budescu, D. (2005), “The effects of imprecise probabilities and outcomes in evaluating
investment options”, Management Science, Vol. 51 No. 12, pp. 1791-1803.
Du, N., Budescu, D.V., Shelley, M. and Omer, T. (2011), “The appeal of vague financial
forecasts”, Organizational Behavior and Human Decision Processes, Vol. 114 No. 2,
pp. 179-189.
Elliott, W., Hodge, F., Kennedy, J. and Pronk, M. (2007), “Are M.B.A. students a good proxy for
nonprofessional investors?”, The Accounting Review, Vol. 82 No. 1, pp. 139-168.
Financial Accounting Standards Board (FASB) (1997), “Reporting comprehensive income”,
Statement of Financial Accounting Standards No. 130 (ASC Topic 220), FASB, Norwalk,
CT.
Financial Accounting Standards Board (FASB) (2011), “Presentation of comprehensive income”,
Accounting Standards Update (ASU) No. 2011-05, FASB, Norwalk, CT.
Hair, J.F., Anderson, R.E., Tatham, R.L. and Black, W.C. (1998), Multivariate Data Analysis, Fifth
ed., Prentice Hall, Upper Saddle River, NJ.
Hirst, D. and Hopkins, P.E. (1998), “Comprehensive income reporting and analysts’ valuation
judgments”, Journal of Accounting Research, Vol. 36 No. 1, pp. 47-75.
Hodge, F., Hopkins, P.E. and Wood, D. (2010), “The effects of financial statement information
proximity and feedback on cash flow forecasts”, Contemporary Accounting Research,
Vol. 27 No. 1, pp. 101-143.
Hopkins, P.E. (1996), “The effect of financial statement classification of hybrid financial
instruments on financial analysts’ stock price judgments”, Journal of Accounting Research,
Vol. 34 No. 1, pp. 33-50.
Hunton, J., Libby, R. and Mazza, C. (2006), “Financial reporting transparency and earnings
management”, The Accounting Review, Vol. 81 No. 1, pp. 135-157.
Hunton, J.E. and McEwen, R.A. (1997), “An assessment of the relation between analysts’ earnings Effects of
forecast accuracy, motivational incentives and cognitive information search strategy”, The
Accounting Review, Vol. 72 No. 4, pp. 497-515. comprehensive
Jordan, C. and Clark, S. (2002), “Comprehensive income: how is it being reported and what are its income
effects?”, Journal of Applied Business Research, Vol. 18 No. 2, pp. 1-8.
Lee, Y., Petroni, K. and Shen, M. (2006), “Cherry picking, disclosure quality, and comprehensive
income reporting choices: the case of property-liability insurers”, Contemporary 299
Accounting Research, Vol. 23 No. 3, pp. 655-692.
Lipe, M. and Salterio, S. (2000), “The balanced scorecard: judgmental effects of information
organization and diversity”, Accounting Review, Vol. 75 No. 3, pp. 283-298.
Downloaded by The University of British Columbia Library At 04:17 03 February 2016 (PT)

Maines, L. and McDaniel, L. (2000), “Effects of comprehensive-income characteristics on


nonprofessional investors’ judgments: the role of financial statement presentation format”,
The Accounting Review, Vol. 75 No. 2, pp. 179-207.
Pandit, G. and Phillips, J. (2004), “Comprehensive income: reporting preferences of public
companies”, The CPA Journal, Vol. 74 No. 2, p. 40.
Schwarz, N. and Bless, H. (1992), “Constructing reality and its alternatives: assimilation and
contrast effects in social judgment”, in Martin, L.L. and Tesser, A. (Eds), The Construction
of Social Judgment, Erlbaum, Hillsdaie, NJ, pp. 217-245.
Schwarz, N. and Bless, H. “Mental construal processes: the inclusion/exclusion model”, in
Stapel, D.A. and Suls, J. (Eds), Assimilation and Contrast in Social Psychology, Psychology
Press, Philadelphia, pp. 119-142.
Wickens, C.D. and Carswell, C.M. (1995), “The proximity compatibility principle: its psychological
foundation and relevance to display design”, Human Factors, Vol. 37 No. 3, pp. 473-494.

Further reading
American Accounting Association’s Financial Accounting Standards Committee (FASC) (1997),
“An issues paper on comprehensive income”, Accounting Horizons, Vol. 11 No. 2,
pp. 120-126.
International Accounting Standards Board (IASB) (2010), Presentation of Items of Financial
Statements (IAS 1), IASB, London.
PWC Dataline (2012), “A look at current financial reporting issues”, available at: www.pwc.com/
en_US/us/cfodirect/assets/pdf/dataline-2012-01.pdf

Corresponding author
Ning Du can be contacted at: ndu1@depaul.edu

For instructions on how to order reprints of this article, please visit our website:
www.emeraldgrouppublishing.com/licensing/reprints.htm
Or contact us for further details: permissions@emeraldinsight.com

Оценить