Recognition and Disclosure of Intangible Assets – A Meta-Analysis Review and Framework Anne JENY * ESSEC Business School, Cergy-Pontoise, France Rucsandra MOLDOVAN John Molson School of Business, Concordia University, Montréal, Québec, Canada Abstract We review over one hundred recent empirical archival papers on internally-developed intangible assets. The knowledge economy based on intangible and intellectual capital demands a re-examination of the accounting treatment for intangibles. We use a two- dimensional matrix framework to organize our review; the first dimension is the recognition of intangible-related amounts, either in the balance sheet or the income statement, versus disclosure of such information in the notes to financial statements or other corporate documents; the actors that are part of the financial reporting environment represent the second dimension. Where the number of papers permits, we summarize the findings using meta-analysis. The framework and meta-analyses allow us to highlight a consensus in the empirical results and point out numerous avenues for future research in this area. Keywords: Intangible assets; Research and development; Recognition; Capitalization; Disclosure; Meta-analysis *Corresponding Author: Ave Bernard Hirsch, B.P. 50105, Cergy-Pontoise Cedex, 95021, FRANCE Tél.: + 331 34432803 Fax: + 331 34432811 [email protected]This version, October 2017, do not quote without the permission of the authors Aknowledgment: We thank Jin Jiang for excellent research assistance, Ann Gallon for her editing work and participants at the European Accounting Association Annual Meeting 2016 Maastricht, The Netherlands and the Association Francophone de Comptabilité 2016 Toulouse, France conferences for helpful comments. Anne Jeny acknowledges the financial support from the Research Center of ESSEC Business School.
57
Embed
Recognition and Disclosure of Intangible Assets A Meta ... Moldovan UQAM 2017.pdfIn this situation, the ... disclosure overload (EFRAG, 2012), may overwhelm users ... then there are
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Recognition and Disclosure of Intangible Assets – A Meta-Analysis Review
and Framework
Anne JENY*
ESSEC Business School, Cergy-Pontoise, France
Rucsandra MOLDOVAN
John Molson School of Business, Concordia University, Montréal, Québec, Canada
Abstract
We review over one hundred recent empirical archival papers on internally-developed
intangible assets. The knowledge economy based on intangible and intellectual capital
demands a re-examination of the accounting treatment for intangibles. We use a two-
dimensional matrix framework to organize our review; the first dimension is the recognition
of intangible-related amounts, either in the balance sheet or the income statement, versus
disclosure of such information in the notes to financial statements or other corporate
documents; the actors that are part of the financial reporting environment represent the
second dimension. Where the number of papers permits, we summarize the findings using
meta-analysis. The framework and meta-analyses allow us to highlight a consensus in the
empirical results and point out numerous avenues for future research in this area.
Keywords: Intangible assets; Research and development; Recognition; Capitalization;
Disclosure; Meta-analysis
*Corresponding Author:
Ave Bernard Hirsch, B.P. 50105, Cergy-Pontoise Cedex, 95021, FRANCE
2016; Lehavy, Li, & Merkley, 2011). Therefore, the recognition versus disclosure debate on
intangibles is overlaid by the issues of how much disclosure of intangible assets users find
useful, and whether disclosure should be mandatory or left entirely up to management, as is
the case under the current financial reporting standards.
Examining recognition and disclosure in the context of intangible assets also allows
us to contribute to the discussions on the role of financial statements, which the IASB has
been revising as part of its Disclosure Initiative. Before any arguments concerning the failure
of the current accounting model (e.g., Lev, 2008), the role of financial statements in today’s
2 Clor-Proell & Maines (2014) conduct an experiment with financial managers over the recognition versus
disclosure of contingent liabilities and find that managers in public companies put in more effort and are less
strategically biased under recognition than disclosure. At the same time, cognitive effort and bias are unaffected
by this choice for managers in private firms. 3 See Healy & Palepu (2001) for an extensive literature review on empirical disclosures studies.
7
knowledge economy should first be clarified. Skinner (2008b) maintains that as long as
intangible-intensive firms are able to attract financing even though the accounts do not
recognize many of their intangibles, then there are no problems with accounting for
intangibles, and therefore no problems with the current accounting model.
We apply a number of limits to simplify and streamline our survey. First, we restrict
our focus to all intangible assets except goodwill. Similar to Skinner (2008b), we consider
that recognition and measurement of goodwill relates to accounting for business
combinations rather than strictly intangibles.4 Compared to previous reviews of the literature
on intangible assets (e.g., Wyatt, 2008), we take a broader perspective on financial reporting
for intangibles, considering both the recognition intangible asset amounts (by capitalization
or expensing), and disclosure of information about intangible assets. Second, our survey only
includes papers that have been published in accounting journals or focus mainly on
accounting for intangibles. Intangible assets, particularly R&D investments, are studied by
papers in a varied range of areas, from biotechnology advancements to strategic and
operations management. The point of interest to us, however, is the way companies account
for and externally report these investments. Therefore, we limit the survey to accounting
papers that touch upon the accounting treatment of intangible assets. Third, we only consider
empirical papers (archival or experimental), without including analytical research in this
area.5 Focusing on empirical research allows us to discuss the depth of the reporting
environment and the role of other stakeholders in the reporting decision. Lastly, we limit our
4 “Accounting standard-setters have also devoted a great deal of attention to accounting for goodwill, which is a
topic that I leave aside because it is largely separable from the discussion in many of the proposals on
intangibles accounting and because its recognition and measurement is related to accounting for business
combinations, which I see as taking the discussion too far afield. I would note though that a loose definition of
goodwill - as the excess of a business' economic value over its book value - is taken by commentators as
evidence of the failure of the current accounting model to correctly recognize intangibles” (Skinner, 2008b). 5 We believe that modelling-based research necessarily assumes away a lot of the complexities of the
environment in which managers decide how to account for intangible assets (Beyer, Cohen, Lys, & Walther,
2010).
8
survey’s main focus to recently-published and working papers, so as to complement existing
literature reviews in this area (e.g., Cañibano et al., 2000; Wyatt, 2008; Zéghal & Maaloul,
2011).
Our survey contributes to future research in several ways. First, our discussion of
recognition versus disclosure of internally-developed intangible assets emphasizes a number
of areas where more research could shed additional light and advance understanding. Some
of the fundamental questions that still remain unanswered are: the role of auditors, the impact
of IAS 38 on firms' behavior, international comparisons of intangible asset accounting
treatment and their consequences.
Second, by aggregating and summarizing the evidence on recognition and disclosure
of intangible assets, we contribute to the ongoing debate between actors asking for more
recognition and disclosure related to intangible assets (e.g., Lev, 2008) and actors who
believe the status quo is perfectly adequate for the accounting treatment of intangibles to
achieve its stated purpose (e.g., Skinner, 2008a,b). Empirical evidence seems to favor
recognition of certain internally-developed intangibles (development costs and brands) as
assets, even though this could be a channel for earnings management. But the same is true of
any accounting choice, and once again we are faced with a trade-off between reliability and
relevance in accounting information.
Third, this review contributes input for the standard-setters’ work on a disclosure
framework as part of the Disclosure Initiative. At a time when the current accounting model
is regarded by some (Lev, 2008) as insufficient and inconsonant with the knowledge-based
business models, the IASB is revisiting some of the conceptual underpinnings of the financial
statements. Our literature survey is relevant because it highlights the attitudes of most of the
actors involved in the financial reporting environment vis-à-vis the main issue that could
make financial statements less useful in today’s world.
9
We continue by describing our organizing framework for analyzing the empirical
accounting literature on intangible assets and the meta-analysis methodology in section 2.
Section 3 summarizes the papers relating to standard-setters and auditors. Section 4 reviews
the papers on preparers, and section 5 the papers on financial statement users. Section 6
concludes and discusses avenues for future research.
2. Organizing framework and research methodology
2.1 Framework for organizing the literature on intangible assets
We organize the literature review along two dimensions. The first dimension concerns
the accounting treatment of internally-developed intangible assets from a recognition versus
disclosure perspective.6 The tension between recognition and disclosure arises from the
perceived differential in reliability and the question of whether users actually read and
understand disclosures. In the context of intangible assets, where recognition rules are
relatively strict, and perhaps out of step with the knowledge economy, “disclosures can
bridge the gap between a firm’s financial statement numbers and its underlying business
fundamentals” (Merkley, 2014).
The second dimension of the framework is represented by stakeholders who have
some interest in the matter of intangible asset recognition and disclosure: (1) standard-setters
and regulators, (2) preparers of financial information, (3) financial statement users (investors,
financial analysts and creditors) and (4) auditors. Examining the literature from the
perspective of each player reveals that some areas have been well researched while other
areas have not yet been fully explored, and draws attention to different interests related to
intangible assets that are not yet clearly understood (e.g. potential biases, strategic decisions).
6 Recognition refers to the number recognized on the face of financial statements, either as an asset on the
balance sheet or as expense on the income statement. Disclosure refers to the narrative or numerical information
provided in the notes to financial statements, other parts of the annual report, and other public corporate
documents (Schipper, 2007).
10
[FIGURE 1 ABOUT HERE]
As economies and the business world evolve towards a more intangible nature, the
spotlight is cast on the parties involved in the financial reporting process and environment,
and the facilitating or debilitating role they play in the provision of relevant financial
information. Both recognition and disclosure of information related to intangible assets result
from interaction between at least four actors. Preparers refer to the accounting standards and
principles they abide by, and weigh up the costs and benefits of disclosing more information.
Standard-setters must make sure their standards can withstand the challenges of “the business
world of tomorrow” without becoming entirely obsolete so that new standards must be issued
whenever a change occurs (Tokar, 2015).7 Investors, creditors, and financial analysts demand
decision-useful information from preparers and engage with standard-setters to ensure their
information needs are met. Auditors need to keep up with the ever-changing business
environment while simultaneously balancing the requirements of the applicable standards
with their duty towards shareholders and their independence of the audited company’s
management. The interactions are complex and the actors’ sometimes conflicting incentives
lead to awkward situations, or situations that put the least powerful party at a disadvantage.
Understanding these actors’ stakes in accounting for intangible assets, a relevant topic to the
new business models of the digitalized economy, places the accounting community in a better
position to tackle the accounting for intangible assets.
For each category of stakeholder, we discuss the results of the meta-analyses,
provided enough studies are available for that category. For standard-setters and auditors, the
number of studies is limited and meta-analysis is not suitable, and so those categories’
attitude towards intangible assets is discussed in a traditional literature review format.
7 In order to begin to address these challenges, the IASB, for example, conducts meetings with practitioners and
academics on “help shape the future of financial reporting” (https://www.icas.com/events/help-shape-the-future-
of-financial-reporting last accessed on November 29, 2015).
Meta-analysis is a statistical technique for summarizing quantitative empirical studies.
It provides a comprehensive way of analyzing a relationship between two variables that has
been examined in at least two prior studies, and a coherent way of making inferences from
the findings of several studies, thereby overcoming some of the shortcomings of narrative
literature reviews.8 Trotman & Wood (1991) emphasize that meta-analysis “leads to more
valid inferences about the knowledge of a set of studies than can be derived from a narrative
literature review.” We analyze the literature on intangible assets using meta-analysis when
the number of published articles is sufficient. Where the same sample is analyzed more than
once, we use the main result in order to ensure sample independence9.
Following prior accounting literature reviews that used meta-analysis (e.g., Hay,
Knechel, & Wong, 2006; Khlif & Chalmers, 2015) and considering the available information
included in our sample of studies, we use the Stouffer Combined test (Wolf, 1986). This
technique uses individual z-stats or converts individual p-values to z-scores and computes an
overall Z-statistic that can be used to test the direction and significance of an effect for the
relationship between two variables. We compute the Z-statistic using the Lipták-Stouffer
method (Lipták, 1958; Stouffer, DeVinney, & Suchmen, 1949) also known as the weighted
Z-test, that weights each z-score based on the sample size n from which it is derived. The
overall Z is then converted into an overall p-value which will be used to assess its
significance.
8 Our aim is to conduct a comprehensive review of the intangible asset literature; hence we use meta-analysis in
an exploratory manner, without developing any ex-ante hypotheses. 9 One study could report several results, usually one for the main research question/hypothesis and two, three
others for secondary research questions/hypotheses (e.g., interaction results). We use only one result (usually the
one for the first hypothesis) from each study included in the literature review.
12
𝑍 =∑ 𝑛𝑖 × 𝑍𝑖𝑘𝑖=1
√∑ 𝑛𝑖2𝑘
𝑖=1
(1)
Rosenthal’s effect size formula uses Z to provide a measure of the overall correlation
between the two variables subject to the meta-analysis (Rosenthal, 1991) and is computed as
follows.
𝐸𝑆(𝑟) =𝑍
√𝑁
(2)
where N is the total sample.
We use the file drawer test to assess the robustness of our meta-analyses. First, we
compute Rosenthal’s Fail-Safe N (FSN) (Rosenthal, 1979) to determine the number of
studies with non-significant results needed to reverse conclusions about a significant
association with a 95% confidence level (Khlif & Chalmers, 2015).
𝐹𝑆𝑁 =𝑘2 × 𝑍2
2.706− 𝑘
(3)
where k is the number of studies included in the analysis. The benchmark for FSN is the
critical number of studies, FSNC. If FSN is larger than FSNC, then the effect is robust.
𝐹𝑆𝑁𝐶 = 5 × 𝑘 + 10 (4)
For each meta-analysis, we also test the homogeneity of the studies by computing a
chi-square statistic with k-1 degrees of freedom. The null hypothesis is that the samples
included in the meta-analysis are homogeneous. If the null hypothesis is corroborated, then
the variation in individual effects is due to statistical errors rather than moderating factors.
𝜒2𝑠𝑡𝑎𝑡𝑖𝑠𝑡𝑖𝑐 =∑(𝑧𝑠𝑐𝑜𝑟𝑒𝑖 −∑𝑧𝑠𝑐𝑜𝑟𝑒𝑖
𝑛𝑖)2
(5)
We conduct two sensitivity tests. First, since our main results are based on studies
with geographically-diverse samples which could contribute to the heterogeneity of
individual effects, we run a test that excludes all non-U.S. samples. Second, we test the
13
sensitivity to outliers of the effects uncovered, by eliminating the observation with the
highest z-score (i.e., the highest individual effect) for each meta-analysis.
2.3 Papers reviewed and sample of papers used in the meta-analysis
We begin our endeavor of reviewing the literature on intangible assets by collecting
all the papers related to this topic. We conduct a keyword search on EBSCO and ProQuest,
the two largest published-article databases, the search is limited to journals in accounting and
finance, and on SSRN for the most recent working papers on this topic. The keywords used
are “intangible”, “intangible asset”, “research and development”, “R&D”, “intellectual
capital”, “IAS 38”, “software development” and associated forms of these words. The search
query included the title, abstract, and keywords of published articles in these databases. We
also check the list of references in previous literature reviews on intangibles, to include any
relevant papers that may not have appeared in our database search. We further limit the
results by keeping only papers using an empirical archival research method and non-goodwill
related papers. The search yielded 102 different relevant papers that we include in our
review, of which 3 are working papers. Table 1 provides an overview of all these papers
organized in our review framework discussed above. Out of the 102 papers we review, 69
papers study the recognition of intangible assets (68%) and 33 examine the disclosure of
intangibles (32%). The vast majority of the papers take the point of view of financial
statement users (58 papers, 57%) and preparers (38 papers, 37%). Very few examine the role
of auditors (4 papers) or standard setters (2 papers).
[TABLE 1 ABOUT HERE]
The sample used for meta-analyses is further restricted by the research methodology
(i.e., empirical archival studies only) and by the published status of the paper. Following
Habib (2013), we include only published papers in the sample for meta-analysis, since
14
unpublished manuscripts have not yet received the vetting of the review process and not all
are publicly available, which induces sample selection bias. Another 19 papers look at
intangible assets in specific settings, which impedes our classification of the variables used
(i.e., we cannot classify those papers in a meaningful way given our framework), and 12
papers examine variable pairs that are not examined in other papers (i.e., only one paper per
variable pair). After these exclusions (Table 2, Panel A), the restricted sample of papers used
for meta-analyses contains 63 papers corresponding to 164 observations at the paper-variable
pair level.
Table 2, Panel B presents the distribution of this restricted sample by publication.
There are 25 papers (40%) published in high-quality journals (i.e., The Accounting Review,
Journal of Accounting Research, Journal of Accounting and Economics, Contemporary
Accounting Research and Review of Accounting Studies).10
Table 2, Panel C presents the
sample distribution by the country or geographic region from which the study sample is
drawn. Most studies examine U.S. companies (44 papers, 67%), followed by the United
Kingdom (8 papers, 12%), Australia (4 papers, 6%) and France (2 papers, 3%).
[TABLE 2 ABOUT HERE]
3. Insights from standard-setters and auditors on intangible assets11
3.1 Standard-setters’ stance on intangible asset recognition and disclosure
Recognition and disclosure of internally-developed intangible assets differs under
IFRS and U.S. GAAP, but both sets of standards struggle with the question of how to
incorporate the economic properties of intangible assets into the financial reporting system
(Powell, 2003).
10
According to usual academic journal rankings. 11
The number of papers published on the views of standard setters and auditors on intangibles is low (2 papers
and 4 papers, respectively), hence we cannot apply the meta-analysis methodology; we discuss their insights
separately based on our organizing framework.
15
IAS 38 Intangible assets requires most internally-developed intangible assets, such as
customer lists, trademarks, brands, mastheads, etc., to be expensed. Since their cost cannot be
distinguished from the normal cost of doing business (IAS 38 par. 16), reliable measurement
for these assets is difficult. For R&D projects, however, the standard distinguishes between
costs incurred in the research phase and costs incurred in the development phase. While the
distinction involves considerable judgment, the general discriminating principle is the
probability of future economic benefits. Since the research phase has highly uncertain
outcomes, the standard mandates expensing of research costs. The development phase,
however, is an application phase to advance the project to a ready-for-use or sale state, at
which point future economic benefits are probable. Development costs must also meet six
recognition criteria before being capitalized. The identification phase along with the six
recognition criteria for development costs constitute a high recognition threshold which
means that, although companies applying IFRS capitalize some of the development costs,
most R&D costs are expensed.
The disclosure requirements in IAS 38 mainly concern the accounting policies for
recognized classes of intangible assets. Required disclosures also include the amount of
expensed R&D expenditure during the period. Without actually making it mandatory, the
standard encourages disclosures about the fully amortized intangible assets still in use and a
description of the significant intangible assets controlled by the entity but not recognized as
assets because they did not meet the recognition criteria (IAS 38 par. 128).
Under U.S. GAAP, the accounting treatment for internally-developed intangibles is
conservative and requires immediate expensing. However, recognition of purchased
intangibles is allowed (Ciftci & Darrough, 2015). In the early 2000s, the FASB worked on a
project related to “Disclosure of information about intangible assets not recognized in the
financial statements” intended to expand note disclosure on internally-developed intangible
16
assets (FASB, 2001). In the AAA comment letter on this project, Skinner et al. (2003) note
that “voluntary disclosure of intangibles information is not widespread” suggesting that the
costs of measuring intangibles and proprietary costs outweigh the benefits of disclosure.12
Chen, Gavious and Lev (2015) show, however, that under the IAS 38 capitalization of
development costs requirement, Israeli companies that switched from U.S. GAAP to IFRS
disclosed more R&D-related information than previously, and more than companies that
continue to apply U.S. GAAP. This finding indicates that when the information is available,
managers are more likely to disclose it, which further suggests that the cost of producing this
information is probably the highest hurdle against disclosure.
The IASB’s Disclosure Initiative Project includes a review of disclosure requirements
in the existing financial reporting standards. Although a redraft of the disclosure requirements
in IAS 38 is not yet available, if IAS 16 Property, plant and equipment is any indication, the
IASB is likely to require more details about the business model as related to the particular
item being disclosed and the risks associated with that item for the entity, on top of the
disclosures of measurement basis and changes during the year already included in most
standards (IASB, 2015).
Any upcoming changes related to disclosure requirements in IAS 38 could be
carefully used to answer questions related to the usefulness of such disclosure and the costs
incurred in providing it. Changes in U.S. accounting standards related to intangible assets are
relatively old.13
However, the switch to IFRS in 2005 in the EU and worldwide is a fruitful
event to exploit for examining preparers and investors’ reaction to the change. Stolowy,
Haller and Klockhaus (2001) detail the differences between French and German GAAP and
IAS 38 and note that, for example, capitalization of internally-generated brands was possible
12
The FASB removed this project from its technical agenda in 2004 stating that any action on this topic will be
taken jointly with the IASB. 13
SFAS 2 “Accounting for research and development costs” was issued in 1974. SFAS 86 “Accounting for the
costs of computer software to be sold, leased, or otherwise marketed” was issued in 1985.
17
under French GAAP, and that allocation to brands of the difference arising on first
consolidation was a widely-used practice in France before the IFRS adoption in 2005. Future
research could, for example, use such differences to examine the relevance of IAS 38
compared to local GAAP.
Nixon (1997) surveyed senior UK accountants for their views on the treatment of
R&D expenditure. While most respondents prefer to expense all R&D costs immediately
given that the ex ante benefits are too uncertain, there is a strong consensus that the ex post
benefits of R&D expenditure are positive. Two other perspectives emerge: (1) disclosure is
much more important than the accounting treatment of R&D expenditure and (2) financial
statements are not viewed as the primary channel of communication ons R&D. These
perceptions suggest that regulators need to move beyond a narrow focus on the technical
issues related to intangibles, and consider the role of financial statements in the wider
communication process that occurs between companies and users.
3.2 External auditors
Very few papers have studied the role of auditors in intangible assets’ recognition and
disclosure. In the U.S. context of R&D expensing, Godfrey and Hamilton (2005) find that
more R&D-intensive firms are more likely to choose auditors who specialize in auditing
R&D contracts. Additionally, R&D-intensive firms tend to appoint top-tier auditors. The
results are particularly strong for small firms where auditor choice is not constrained by the
need to appoint a top-tier auditor to ensure the auditor’s financial independence of the client.
Tutticci, Krishnan and Percy (2007) and Krishnan and Wang (2014) examine
questions related to auditing and R&D capitalization. Using an Australian sample, Tutticci et
al. (2007) find that that external monitoring by a Big 5 auditor and the Australian Security
Commission decreases managers’ tendency to capitalize R&D costs. Furthermore, a well-
18
known or “brand-name” auditor leads to a stronger relationship between capitalized R&D
and stock returns, consistent with the high audit quality associated with brand-name auditors.
Complementing these findings, using a U.S. sample over the period 2004-2009, Krishnan and
Wang (2014) find that audit fees are smaller for companies that capitalize software
development costs, after controlling for traditional measures of client risk. This result is
consistent with the idea that capitalized software development costs are informative about
audit risk.
4. Preparers
4.1 Meta-analysis results
Table 3 presents the results of the meta-analyses conducted on studies that examine
variables related to preparers and intangible assets. Table 3, Panel A lists the 17 papers
included that relate to preparers, either examining the consequences of recognition or
disclosure of intangible assets or examining the determinants of decisions related to
recognition or disclosure of intangible assets.
Table 3, Panel B presents the results of the meta-analyses. For the question of the
consequences of recognition or disclosure of intangible assets, we generally find significant
positive associations between future firm profitability and variables that proxy for internally-
Eberhart, Maxwell and Siddique, Journal of Accounting Research, 2008
2
Total 102 69 33
This table provides an overview of all the papers that we review organized and classified in our framework. We review 102 distinct papers. Papers that are classified into two
boxes are preceded by #.
44
Table 2. Composition of the sample of studies included in the meta-analyses
Panel A: Sample selection for meta-analyses
Number of
studies Percent
Papers reviewed 102
(-) Papers using other than empirical archival quantitative
research methods -5
(-) Unpublished papers -3
Initial sample of papers considered for meta-analyses 94 100%
(-) Cannot classify dependent or independent variable -19
(-) Only one study to examine the relation between two
variables -12
Final sample of papers for meta-analyses 63 67.02%
This table describes the sample selection for the studies included in the meta-analyses. The final sample of
studies represents 158 observations at the paper-variable pair level.
Panel B: Sample by journal
Journal name and abbreviation Frequency Percent
Accounting and Finance (AF) 3 4.76
Canadian Accounting Perspectives (CAP) 1 1.59
Contemporary Accounting Research (CAR) 3 4.76
European Accounting Review (EAR) 4 6.35
Financial Management (FM) 1 1.59
Journal of Accounting, Auditing and Finance (JAAF) 7 11.11
Journal of Accounting and Economics (JAE) 4 6.35
Journal of Accounting and Public Policy (JAPP) 3 4.76
Journal of Accounting Research (JAR) 9 14.29
Journal of Business Finance & Accounting (JBFA) 9 14.29
Journal of Empirical Finance (JEF) 1 1.59
Journal of Financial Economics (JFE) 1 1.59
Journal of International Accounting Research (JIAR) 1 1.59
Journal of International Financial Management and Accounting
(JIFMA) 1 1.59
Review of Accounting Studies (RAS) 4 6.35
Review of Quantitative Finance and Accounting (RQFA) 2 3.17
The Accounting Review (TAR) 5 7.94
The International Journal of Accounting (TIJA) 4 6.35
Total 63 100%
Papers in high-quality journals 25 40%
This table presents the distribution of the final sample of papers included in the meta-analyses by publication
journal. Journals in boldface font are considered high-quality.
45
Panel C: Sample by country
Country/Region Frequency Percentage
Australia 4 6.06%
Canada 1 1.52%
Continental Europe 1 1.52%
France 2 3.03%
International (emerging) 1 1.52%
International (including U.S.) 1 1.52%
Japan 1 1.52%
South Korea 1 1.52%
Spain 1 1.52%
Taiwan 1 1.52%
United Kingdom 8 12.12%
United States 44 66.67%
Total 66 100%
This table presents the distribution of the final sample of papers included in the meta-analyses by country or
region from which the sample of companies is drawn. Bah & Dumontier (2001) is counted four times since their
sample contains observations from four countries and they conduct the analyses per country.
46
Table 3. Meta-analyses of studies related to intangible assets and preparers
Panel A: List of papers
Authors
Publication
Year Journal Country Sample period
Sample
size
Preparers – Consequences of the recognition of or disclosure related to intangible assets
Ahmed and Falk 2006 JAPP Australia 1992 - 1999 1172
Amir, Guan, and Livne 2007 JBFA U.S. 1972 - 2002 37263
Anagnostopoulou and Levis 2008 TIJA UK 1990 - 2003 15488
Brown and Kimbrough 2011 RAS U.S. 1980 - 2006 119436
Cazavan-Jeny, Jeanjean, and Joos 2011 JAPP France 1992 - 2001 1060
Ciftci and Cready 2011 JAE U.S. 1975 - 2003 122636
Pandit, Wasley, and Zach 2011 JAAF U.S. 1972 - 2000 20391
Ritter and Wells 2006 AF Australia 1979 - 1997 1078
Shust 2015 JAAF U.S. 1988 - 2010 77003
Sougiannis 1994 TAR U.S. 1975 - 1985 66
Weiss, Falk, and Zion 2013 AF U.S. 1990 - 2005 528
Preparers – Determinants of recognition of or disclosure related to intangible assets
Bah and Dumontier 2001 JBFA Continental Europe 1996 - 1996 204
Japan 1996 - 1996 353
UK 1996 - 1996 233
U.S. 1996 - 1996 1069
Cazavan-Jeny, Jeanjean, and Joos 2011 JAPP France 1992 - 2001 1060
This table presents the results of the meta-analyses of variables related to intangible assets and preparers. Sample (N) is the total number of observations added up across the
studies that examine a pair of variables. Number of studies (k) refers to the number of independent samples (i.e., papers). FSN is Rosenthal’s Fail Safe N and FSNC is the
critical number of studies in the file drawer. For each meta-analysis (i.e., variable pair), the 2 column indicates the chi-square statistic with k-1 degrees of freedom for testing
whether the empirical correlations are homogeneous. Statistical significance is indicated as follows: *** p-value < 0.01; ** p-value < 0.05; * p-value < 0.1; n.s. denotes p-
value not significant at conventional levels.
49
Panel C: Sensitivity test – U.S. samples only
Dependent variable
Independent
variable
Sample
(N)
Number
studies
(k)
Effect
size
Z-
statistic
p-
value Sig FSN FSNC 2
Sig
2
Preparers – Consequences of the recognition of or disclosure related to intangible assets
This table presents the results of a sensitivity test of the meta-analyses of studies related to intangible assets and preparers that excludes all non-U.S. samples. Sample (N) is
the total number of observations added up across the studies that examine a pair of variables. Number of studies (k) refers to the number of independent samples (i.e., papers).
FSN is Rosenthal’s Fail Safe N and FSNC is the critical number of studies in the file drawer. For each meta-analysis (i.e., variable pair), the 2 column indicates the chi-square
statistic with k-1 degrees of freedom for testing whether the empirical correlations are homogeneous. Statistical significance is indicated as follows: *** p-value < 0.01; ** p-
value < 0.05; * p-value < 0.1; n.s. denotes p-value not significant at conventional levels.
50
Panel D: Sensitivity test – eliminate the largest empirical correlation
Dependent variable
Independent
variable
Sample
(N)
Number
studies
(k)
Effect
size
Z-
statistic p-value Sig FSN FSNC 2
Sig
2
Preparers – Consequences of the recognition of or disclosure related to intangible assets
This table presents the results of a sensitivity test of the meta-analyses on variables related to preparers that eliminates the highest empirical correlation. Sample (N) is the
total number of observations added up across the studies that examine a pair of variables. Number of studies (k) refers to the number of independent samples (i.e., papers).
FSN is Rosenthal’s Fail Safe N and FSNC is the critical number of studies in the file drawer. For each meta-analysis (i.e., variable pair), the 2 column indicates the chi-square
statistic with k-1 degrees of freedom for testing whether the empirical correlations are homogeneous. Statistical significance is indicated as follows: *** p-value < 0.01; ** p-
value < 0.05; * p-value < 0.1; n.s. denotes p-value not significant at conventional levels.
51
Table 4. Meta-analyses of studies related to intangible assets and financial statement users
Panel A: List of papers
Authors
Publication
Year Journal Country
Sample
period
Sample
size
Financial Statement Users - Analysts
Barron, Byard, Kile, and Riedl 2002 JAR U.S. 1986 - 1998 1103
Barth, Kasznik, and McNichols 2001 JAR U.S. 1983 - 1994 10631
Chambers, Jennings, and Thompson II 2002 RAS U.S. 1979 - 1998 89419
Ciftci, Lev and Radhakrishnan 2011 JAAF U.S. 1979 - 1997 7591
Gu and Wang 2005 JBFA U.S. 1981 - 1998 6167
Jones 2007 CAR U.S. 1997 - 1997 119
Matolcsy and Waytt 2006 AF U.S. 1990 - 1997 421
Merkley 2014 TAR U.S. 1996 - 2007 22482
Rajgopal, Venkatachalam, and Kotha 2003 JAR U.S. 1999 - 2000 434
Financial Statement Users - Bondholders
Eberhart, Maxwell, and Siddique 2008 JAR U.S. 1990 - 1998 72
Shi 2003 JAE U.S. 1991 - 1994 81
Financial Statement Users - Investors
Aboody and Lev 1998 JAR U.S. 1987 - 1995 778
Ahmed and Falk 2006 JAPP Australia 1992 - 1999 1172
Ali, Ciftici, and Cready 2012 JBFA U.S. 1975 - 2006 38853
Barth and Clinch 1998 JAR Australia 1991 - 1995 1750
Barth, Clement, Foster, and Kasznik 1998 RAS U.S. 1991 - 1996 595
Boone and Raman 2004 JAAF U.S. 1994 - 1997 52
Brown and Kimbrough 2011 RAS U.S. 1980 - 2006 119436
Bulitz and Ettredge 1989 TAR U.S. 1974 - 1983 2832
Callimaci and Landry 2004 CAP Canada 1997 - 1999 109
Cazavan-Jeny and Jeanjean 2006 EAR France 1993 - 2002 770
Chambers, Jennings, and Thompson II 2002 RAS U.S. 1979 - 1998 89419
Chan, Martin, and Kensinger 1990 JFE U.S. 1979 - 1985 79
Chauvin and Hirschey 1993 FM U.S. 1988 - 1990 4653
Ciftci and Cready 2011 JAE U.S. 1975 - 2003 122636
Ciftci, Darrough, and Mashruwala 2014 EAR U.S. 1975 - 2007 171894
Donelson and Resutek 2012 RAS U.S. 1973 - 2008 56145
Ely and Waymire 1999 JAR U.S. 1927 - 1927 146
Ely, Simko, and Thomas 2003 JAAF U.S. 1988 - 1998 193
Franzen and Radhakrishnan 2009 JAPP U.S. 1982 - 2002 47167
Givoly and Shi 2008 JAAF U.S. 1986 - 1998 390
Green, Stark, and Thomas 1996 JBFA UK 1990 - 1992 230
Gu and Li 2010 JAAF U.S. 1995 - 2004 4966
Guo, Lev, and Zhou 2004 JAR U.S. 1995 - 1997 265
Han and Manry 2004 TIJA South Korea 1988 - 1998 3191
Hirschey and Richardson 2004 JEF U.S. 1989 - 1995 1720
Hirschey and Weygandt 1985 JAR U.S. 1977 - 1977 390
Hirschey, Richardson, and Scholz 2001 RQFA U.S. 1989 - 1995 1290
Kallapur and Kwan 2004 TAR UK 1984 - 1998 232
Lev and Sougiannis 1996 JAE U.S. 1975 - 1991 3000
Lev and Sougiannis 1999 JBFA U.S. 1975 - 1989 1200
Merkley 2014 TAR U.S. 1996 - 2007 22482
Oswald and Zarowin 2007 EAR UK 1991 - 1999 1002
Palmon and Yezegel 2012 CAR U.S. 1993 - 2004 8620
52
Ritter and Wells 2006 AF Australia 1979 - 1997 1078
Shah, Stark, and Akbar 2009 TIJA UK 1990 - 1998 9752
Shevlin 1991 TAR U.S. 1980 - 1985 145
Shortridge 2004 JBFA U.S. 1985 - 1996 172
Shust 2015 JAAF U.S. 1988 - 2010 77003
Sougiannis 1994 TAR U.S. 1975 - 1985 66
Tutticci, Krishnan, and Percy 2007 JIAR Australia 1992 - 2002 386
Xu, Magnan, and André 2007 CAR U.S. 1998 - 2004 1232
This table provides the results of the meta-analyses of studies related to intangible assets and financial statement users. Sample (N) is the total number of observations added
up across the studies that examine a pair of variables. Number of studies (k) refers to the number of independent samples (i.e., papers). FSN is Rosenthal’s Fail Safe N and
FSNC is the critical number of studies in the file drawer. For each meta-analysis (i.e., variable pair), the 2 column indicates the chi-square statistic with k-1 degrees of
freedom for testing whether the empirical correlations are homogeneous. Statistical significance is indicated as follows: *** p-value < 0.01; ** p-value < 0.05; * p-value <
0.1; n.s. denotes p-value not significant at conventional levels.
This table presents the results of a sensitivity test of the meta-analyses of studies related to intangible assets and financial statement users that excludes all non-U.S. samples.
Sample (N) is the total number of observations added up across the studies that examine a pair of variables. Number of studies (k) refers to the number of independent
56
samples (i.e., papers). FSN is Rosenthal’s Fail Safe N and FSNC is the critical number of studies in the file drawer. For each meta-analysis (i.e., variable pair), the 2 column
indicates the chi-square statistic with k-1 degrees of freedom for testing whether the empirical correlations are homogeneous. Statistical significance is indicated as follows:
*** p-value < 0.01; ** p-value < 0.05; * p-value < 0.1; n.s. denotes p-value not significant at conventional levels.
57
Panel D: Sensitivity test – eliminate the largest empirical correlation
This table provides the results of a sensitivity test of the meta-analyses of studies related to intangible assets and financial statement users that eliminates the highest empirical
correlation. Sample (N) is the total number of observations added up across the studies that examine a pair of variables. Number of studies (k) refers to the number of
independent samples (i.e., papers). FSN is Rosenthal’s Fail Safe N and FSNC is the critical number of studies in the file drawer. For each meta-analysis (i.e., variable pair), the
2 column indicates the chi-square statistic with k-1 degrees of freedom for testing whether the empirical correlations are homogeneous. Statistical significance is indicated as