Research in International Business and Finance 38 (2016) 509–516

Contents lists available at ScienceDirect

Research in International Business

and Finance
journal homepage: www.elsevier.com/locate/ribaf

Full length Article

Guiding through the Fog: Does annual report readability

reveal earnings management?
Aymen Ajina a , Mhamed Laouiti b , Badreddine Msolli c,∗
a
Arab East Colleges Riyadh and FSEG-University of Sousse, Tunisia
b
Faculty of Economic Sciences and Management of Mahdia, Tunisia
c
ESSCA School of Management—LUNAM University, France

a r t i c l e i n f o a b s t r a c t

Article history: This paper attempts to find a relationship between earnings management and annual report
Received 9 March 2016 readability. The study was conducted on the French stock market between 2010 and 2013
Received in revised form 5 June 2016 and includes 163 firms listed on the CACALL shares index. We assess the readability of
Accepted 8 July 2016
annual reports using the Gunning-Fog Index and earnings management is measured by
Available online 12 July 2016
discretionary accruals estimated using the models of Dechow et al. (1995) and Raman and
Shahrur (2008). Our results show the existence of a positive and significant relation between
JEL classification:
the level of discretionary accounting adjustments and Fog Index. In fact, companies man-
G14
aging their earnings tend to make annual report readability more complex. In this way,
G18
M41 we also find a significant link between the Fog Index and the other financial indicators of
M45 our model (Ownership dispersion, Profitability, Leverage and Firm size), so supporting our
main result.
Keywords: © 2016 Elsevier B.V. All rights reserved.
Earnings management
Annual report readability
French stock market

1. Introduction

Legality plays a significant role in and can be said to characterize financial communication. Its aim is to protect the
shareholder by guaranteeing “exact, precise and sincere” information (1 August 2003, the [French] Financial Security Law).
Financial institutions continue to bolster regulations by strengthening the financial standards transparency directive.
Corporate disclosure usually involves three elements: content (what), timing (when) and presentation (how) (Courtis,
2004), the usefulness of which depends on readability and understandability. Thus, according to the Act NRE-New Economic
Regulations- (2001), we expected a bolstering of the regulations and the recommendations in favor of the readability of the
information emitted within financial communication.
In 2010 and following the Securities and Exchange Commission (SEC) in the United States (U.S.), the annual report of the
AMF (Autorité des Marchés Financiers) [“Financial Markets Regulator”] in France focused on readability. The AMF published
an action plan to find solutions to the public crisis of confidence and political expectations. It also aimed to improve the
readability of financial information for the “layperson”. Taking stock of the legal situation, some companies dedicated funds
to making annual reports more readable.

∗ Corresponding author.
E-mail address: badreddine.MSOLLI@essca.fr (B. Msolli).

http://dx.doi.org/10.1016/j.ribaf.2016.07.021
0275-5319/© 2016 Elsevier B.V. All rights reserved.
510 A. Ajina et al. / Research in International Business and Finance 38 (2016) 509–516

Despite this legal rigidity, corporate scandals and accounting fraud are still current. Recurrent financial scandals cause a
reliable loss in the global market. Accounting problems are widely cited as the main reason for the loss of confidence of the
investors who follow these scandals (Machuga and Teitel, 2007).
In this case, financial information is not completely transparent, whether the communication channel used is “formal” or
“informal”. “Informal” channels refers to press releases or interviews with business leaders, which rarely convey “innocent”
or unfiltered information that has not previously been approved. In fact, this kind of medium contains certain metaphors
that indirectly reflect or create realities (Tomoni, 2012). ‘Formal’ channels refers to the fact that annual reports play a role
other than their original aim and help disseminate statements handled mainly when they are less readable and therefore
complicated to read (Li, 2008; Bloomfield, 2008). In both cases, fraudulent information is difficult to detect. So, the aim of
readability is mainly to make information broadcasted on the financial markets accessible and understandable.
On the theoretical level, readability is always measured by reading scores, known as the “reading index”, originally based
on primary school assessments. In this context, the Gunning-Fog Index (Fog Index), developed by Gunning (1952) and derived
from the computational linguistics literature, is a well-known and simple formula for measuring readability. A reading index
is a calculated score that varies between two extremes. At the first end of the scale is very basic reading comprehension.
At the other end, in contrast, reading comprehension is very advanced, corresponding essentially to scientific documents
(Courtis, 1997).
The application of these scores to annual reports dates back relatively far (Jones and Shoemaker, 1994) and they are still
in use and still show their efficiency as a search tool, particularly in the financial and accounting literature (Tetlock and Paul,
2007; Li, 2008; Biddle et al., 2009; Miller, 2010; Lehavy et al., 2012; Lawrence, 2013; Ramanna and Watts, 2011).
We identified two categories of studies focusing on the relationship between the complexity of annual reports and certain
financial indicators.
The first presents annual report readability as a determining factor of future financial performance, especially to explain
the behavior of stock prices. Antweiler and Frank (2004), Tetlock and Paul (2007), and Li (2008) show that future stock
returns, strategic events and earnings, can be explained by the wording selected by managers to describe their operations
and the language newspaper columnists use to report stock market events.
The second type of studies focuses more the other aspect of readability. In fact, some authors study the impact of discloser
complexity on asymmetry information and the liquidity of the stock market. Miller (2010), Lawrence (2013) and De Franco
et al. (2013) find evidence that complex language discourages individual investor trading due to increased information
processing costs. Supporting this evidence, trading volumes are lower when equity analyst reports have a higher Fog Index
(De Franco et al., 2013).
Li (2008) finds a negative relationship between the Fog Index and the level of earnings. The author concludes that
disclosures of his sample are more difficult to understand. According to Bloomfield (2008), it is unclear whether this result is
due to managers providing complex disclosures to obfuscate bad performance or due to be the fact that bad news is simply
harder to communicate.
Our study combines the previous two categories of works and is also based on Bloomfield’s proposition (2008). In fact,
we try to analyze the possible link between annual report readability and earning management practices. In this way, we
suppose that managers who practice earnings management will tend to obfuscate their act by publishing complex annual
reports.
In other words, we try to answer the following question: Is the complexity of annual reports linked to companies that
practice earnings management? Is there a link between the complexity of annual reports and company practice of earnings
management?
According to the previous question, our paper makes a number of important contributions. We contribute to the limited
number of studies that has examined the linguistic complexity of corporate disclosures and its relation to other variables,
particularly to earnings management indicators.
Furthermore, our study is the first of its kind conducted on the French market. In fact, the French case appeared to be an
interesting subject for study because of its corporate governance environment and the annual reports readability importance
granted by the French authorities (AMF, 2015; the NRE law, 2001; Financial Security Law, 1 August, 2003).
The findings show the existence of a positive and significant relation between the level of the discretionary accounting
adjustments estimated using the models of Dechow et al. (1995), Raman and Shahrur (2008) and Fog Index. This means that
managers provide complex disclosures to obfuscate this kind of practice. Some reports indicate that several managers try to
conceal the profits by means of complexity in annual reports, confirming the theoretical hypothesis of “big bath accounting”.
The remainder of this paper is organized as follows: the second section presents the theoretical setting of the relationship
between earnings management and annual report readability. The third section presents the sample and the methodology
used, followed by the results and discussions in section four. The last section concludes the paper.

2. Theoretical foundations

Corporate governance is a subject of increasing importance for shareholders and generally for all stakeholders. Due
to numerous financial scandals and the economic crisis, the importance of this topic has increased significantly over the
past two decades. The worldwide failure in financial reporting has largely denounced weak internal controls of companies.
 A. Ajina et al. / Research in International Business and Finance 38 (2016) 509–516 511

Therefore accounting problems are widely cited as the main reason for the loss of confidence of the investors who followed
these scandals (Machuga and Teitel, 2007).
The main explanation for this interest is the consequence of the increase of agency costs. In fact, this is the direct result
of the separation of ownership and control (Berle and Means, 1932; Coase, 1937; Jensen and Meckling, 1976; Holmstrom,
1979; Fama and Jensen, 1983). The appearance of this problem is obvious in the managers’ incentives to take actions that
maximize their utility to the detriment of shareholder wealth. The existence of such incentives in the financial reporting
process reduces the credibility of the reported earnings numbers.
Initially, organizational problems were considered the logical result of information asymmetry. The theoretical hypoth-
esis of adverse selection confirms the existence of informational asymmetries in the market. According to Copeland and
Galai (1983), the costs of adverse selection result from the existence of informed agents whose gains are at the expense of
uninformed agents.
Taking into account this assumption by researchers following the example of Ross (1977), it showed that some agents
on the market possess private information, which is the root cause of signaling theory on the market.
Indeed, managers send signals to investors to reduce information asymmetry (Morris, 1987). According to Keasey and
Short (1997), there are various possible actions that firms will take in order to signal their value to potential shareholders.
This theory is usually applied to corporate finance decisions such as dividend policy. But this theory can be applied to the
readability issue.
According to the opportunistic perspective, earnings management is considered an opportunistic process used by the
manager to modulate results. The manager’s principal aim is to maximize their utility function. The opportunistic approach
has been discussed in the literature since the emergence of political and contractual accounting theory initiated in the late
1970s by two researchers, Ross Watts and Zimmerman, thus falling into the positive approach to accounting research.
However, management honesty and sincerity in the explanation of all aspects of firm performance are questionable,
since there are conflicts of interest between owners and managers (Jensen and Meckling, 1976). As a manager, management
may be motivated to disclose information that conveys only positive performance and conceal negative information that
might harm the firm’s performance. Consequently, information asymmetry can be seen to exist between management and
the public. Indeed, Adelberg (1979) reports that the management knowingly or unknowingly introduces an interpretive
bias into annual reports and that it is plausible that some managers would obfuscate their failures and underscore their
successes.
There are two ways to distort reality and to cover the possible low performance of managers. The first way to emphasize
news can be seen in the manipulation of verbal or numerical information or the use of metaphors. This can be achieved
by thematic manipulation by emphasizing positive words and themes, or emphasizing positive financial performance; or
by performance comparisons by choosing benchmarks that represent current financial performance in the best possible
situation. Tomoni (2012) analysed 74 communication media (interviews, speeches and press releases) and concluded that
some Roman bankers use metaphors in their speech to narrate reality. The author identifies three types of metaphors (war,
water and machine metaphors). Each metaphor reflects a working atmosphere, a personal state of mind or an economic
situation that the leaders did not report directly.
In the second way, and according to Merkl-Davies and Brennan (2007), managers’ motives can be classified into conceal-
ment and attribution. Concealment can be achieved in various ways, either by reading ease manipulation such as making
the text more difficult to read, or rhetorical manipulation such as using persuasive language.
Studies have attempted to evaluate annual report reading ease, its components and its elements, which have led to some
companies’ annual reports being very difficult to read, so that they can be classified as a text which risks being inaccessible
to the detriment of a large proportion of non-experimented shareholders (Jones and Shoemaker, 1994). Therefore, annual
report complexity strengthens the information asymmetry between managers and investors and causes agency conflicts.
Some studies also investigate annual reports to verify whether they become more difficult to read over time. The results are
not similar (Jones and Shoemaker, 1994).
Bloomfield (2008) suggests that loss firms need to provide more explanations as a result of “management by exception”.
He suggests that the nature of accounting conservatism, recognizing bad news in a more timely fashion than good news,
requires managers to provide more explanations about the future when there are losses. Thus, obfuscation requires conscious
actions to affect readability, whereas the ontological explanations suggest that readability is inherently a function of the
circumstances.
In this context, Rutherford (2003) advances that in the event of information asymmetry, firms with good performance
have tried to find ways of signaling the superiority of their performance by disclosing readable information with greater
clarity, whereas firms with poor performance will obfuscate negative information by using complex and difficult wording.
Thus, any costs incurred in monitoring managers’ activities and bonding them are necessary (Morris, 1987). Furthermore,
Morris (1987) stated that producing readable accounting reports and making financial statements easier to understand
are among the devices used in monitoring and bonding the managers (agents). These studies show that readability is an
important measure for judging the quality of financial documents. A recent study used the Fog Index or word count as a
readability measure explaining the degree of complexity of relationships (Loughran and McDonald, 2011; De Franco et al.,
2013; Li, 2008). Reports indicate that some managers try to conceal earnings by using a high level of complexity in annual
reports.
512 A. Ajina et al. / Research in International Business and Finance 38 (2016) 509–516

Thus, firms may manipulate the content and presentation of information in various ways, essentially using what is known
as “impression management” (Godfrey et al., 2003). Using this practice, firms can manipulate verbal information through
reading ease manipulation (e.g., to make the text difficult to read) or through the rhetorical manipulation method/practice
(e.g., using persuasive language). The reading ease manipulation method is used with the intention of hiding bad news.
In his examination of the link between annual report readability and earnings persistence, Li (2008) uses the Fog Index
and word count to estimate readability. He finds a negative relationship between Fog and the level of earnings. Also, the
longest document presents a very low future income. The author confirms that when managers notice the drop in the results,
they must attempt to hide future profits by increasing the complexity of their financial statements. In this case, the quality
of this financial information is indirectly affected by this difficulty.
Based on the discussion above, we present the following testable hypothesis (stated in the alternative form):

Hypothesis. Firms that have managed earnings have annual report disclosures that are less readable, ceteris paribus.

3. Sample and methodology

3.1. Data

Our sample includes French firms listed on the CACALL shares index from 2010 to 2013. Recently, the annual report of the
AMF (2009) began to focus on readability. The AMF published an action plan to find solutions to the public confidence crisis;
its main aim is also to improve financial information readability for “laypersons”. Public accounts must also be readable.
Indeed, the organic law relative to the finance laws imposes on the State the transmission of reliable, readable and usable
information. The law specifies what is considered as usable: information understandable for people with no accounting
knowledge.
We exclude financial companies due to their atypical behaviour in financial reporting. Our final sample includes 163
companies over 4 years. Financial data were gathered from the Thomson One Banker database. To compute readability
proxies, we first download annual reports from the Autorité des Marchés Financiers and from company websites.

3.2. Variable measurements

3.2.1. The discretionary accurals

We measure earnings management by discretionary accruals estimated using the models of Dechow et al. (1995), Raman
and Shahrur (2008) to enhance the robustness of our results.

– Jones-modified model (1995)

TAi, t/Ai, t − 1 = ˛0(1/Ai, t − 1) + ˛1[(CAi, t − CCRi, t)/Ai, t − 1] + ˛2(PPEi, t/Ai, t − 1) + εi, t

With: TAi,t: Total accrual in year t; Ai,t − 1: Total assets in year t−1; CAi,t: Change in sales; CCRi,t: change in receivables;
PPEi,t: Gross proprety plant and equipement; ␧i,t: Residuals that represent the estimation of discretionary accruals.

– Raman and Shahrur (2008) model

TAi, t/Ai, t − 1 = ˛0(1/Ai, t − 1) + ˛1[(CAi, t − CCRi, t)/Ai, t − 1]

+ ˛2(PPEi, t/Ai, t − 1) + ˛3(ROAi, t − 1) + ˛4(MTBi, t) + εi, t

With: MTBi,t: The market to book ratio.

3.2.2. Annual report readability

The reading scores, known as the “reading index”, were originally based on primary school assessments. The application
of these scores to annual reports dates back relatively far (Sydserff and Weetman, 2002) and drew quite a lot of criticism
(Jones and Shoemaker, 1994). Estimating the validity of used formulae is complex. Courtis (1997) asked the accounting
researchers in communication to make more effort to find a more formal approach for the validation of these formulae.
Consequently, these scores are still in use and continue to show their efficiency as research tools.
A reading index is a calculated score bound to standards defined by written texts classified in terms of difficulty of reading
difficulty. A reading index is a calculated score that varies between two extremes. At the first end of the scale is very basic
reading comprehension. At the other end, in contrast, reading comprehension is very advanced, corresponding essentially
to scientific documents (Courtis, 1997).
As in Li (2008), we assess the readability of annual reports using the Gunning-Fog Index derived from the computational
linguistics literature. The Fog Index has been used widely and has seen increasing use in the accounting literature (Biddle
et al., 2009; Miller, 2010; Lehavy et al., 2012; Lawrence, 2013; Ramanna and Watts, 2011).
 A. Ajina et al. / Research in International Business and Finance 38 (2016) 509–516 513

The Fog Index, developed by Robert Gunning (1952), is a well-known and simple formula for measuring readability.
Assuming that the text is well-written and logical, it captures text complexity as a function of syllables per word and words
per sentence.1 The index indicates the number of years of formal education a reader of average intelligence would need to
read the text once and understand that piece of writing with its word-sentence workload. It is calculated as follows:

Fog = (words per sentence + percent of complex words) × 0 : 4 (1)

Where complex words are defined as words with three syllables or more, the relation between the Fog and reading ease is
as follows: FOGX18 means the text is unreadable; 14–18 (difficult); 12–14 (ideal); 10–12(acceptable); and 8–10 (childish).

3.2.3. Control variables

As earnings management is not the only determining factor for annual report readability, we add control variables to our
regression models.

3.2.3.1. Ownership dispersion. Another argument that follows from agency theory is that as the ownership structure is more
dispersed, agency costs increase due to increased probability of conflicts of interest between owners (Jensen and Meckling,
1976). Therefore, to counter this effect, the firms that have high ownership dispersion will likely issue more readable annual
reports. Oliveira et al. (2006) examine the effect of ownership dispersion on the voluntary disclosures and report that firms
with higher ownership dispersion provide higher voluntary disclosures of intangible assets. Ownership dispersion is defined
as the percentage of shares held by the public.

3.2.3.2. Profitability. According to the Incomplete Revelation Hypothesis (IRH) proposed by Bloomfield (2002), managers
can reduce the market response to bad news by making bad news more costly to analyze. Li (2008) tests this hypothesis
and reports that managers make bad news costly by writing excessively long annual reports with unnecessary big words
and long sentences. An alternative explanation of this relationship is that bad and losses news are more difficult to describe.
Similarly, Bloomfield (2008) claims that the reason for which managers might write longer and complex annual reports is
to protect themselves from litigation when they report poor performance. Therefore, we measure the profitability as Net
Income scaled by Total Assets of a firm.

3.2.3.3. Foreign sales. we use foreign sales percentage as a proxy for complexity of operations. In general, higher foreign
sales will lead to longer annual reports and difficult to read annual reports. In addition, this variable acts as a control variable
for the domestic culture as it is expected that increased foreign sales will mitigate the effect of domestic culture (secrecy
in the home country) since this means that the firm is more exposed to foreign culture norms and behaviors (Hope et al.,
2011).

3.2.3.4. Leverage. Debt ratio is measured as total liabilities divided by total assets. In general, companies with a higher
proportion of debt are expected to provide less readable annual statements. This measure was also adopted by DeFond and
Jiambalvo (1994), Ben Othman and Zéghal (2006) and Zéghal et al. (2011).

3.2.3.5. Firm size. Various studies have shown firm size as a determinant for accounting disclosures and size is also used as
a proxy variable for omitted variables (Botosan, 1997; Hossain et al., 1995). This paper uses the natural log of total assets
measured in millions of euros as Firm size to explain the readability of annual reports. In general, it is expected that a larger
firm will have more complex operations and will provide longer and more complex annual reports. The use of the logarithm
has the advantage of circumventing the problem of scale arising from the small measurements of other variables in the
model. This measure was used by Ben Othman and Zéghal (2006), and Zéghal et al. (2011) (Table 1).

3.3. The models

We use a panel regression analysis for a sample of 163 firms listed on the French CACALL index for four years (2010–2013).
Panel data analyses include two special dimensions: an individual dimension, as indicated by the i index, standing for the
company, and a t index standing for the period dimension (Gujarati, 2004). The Hausman test is used to choose between
fixed effect and random effect models.
The Hausman test compares the variance—covariance matrix of the two estimators: W = (␤f − ␤a) [var
(␤f − ␤a)] − 1(␤f − ␤a) with ␤f, fixed effects estimator and ␤a, random effects estimator. The result follows a chi-2

1
There are two other popular measures of readability: the Kincaid index and the Flesch Reading Ease Index. The Kincaid Index, also referred to as the
Flesch–Kincaid formula and calculated as (11.8 syllables per word) + (0.39 words per sentence) 15:59, rates text according to U.S. grade school level.
Therefore, a score of 8.0 means that the document could be understood by an average eighth grader. The Flesch Reading Ease rates text on a 100-point scale
and is calculated as 206:835 ð1:015 words per sentenceÞ ð84:6 syllables per wordÞ. The higher the Flesch Reading Ease index, the easier the text
will be. The empirical results based on the Kincaid Index and the Flesch Index are similar to those based on the Fog Index and are therefore unreported.
For more information about these readability measures, see http://www.plainlanguage.com/Resources/readability.html.
514 A. Ajina et al. / Research in International Business and Finance 38 (2016) 509–516

Table 1
Variable definition.

Variable Definition Measure

Dependent variable
Fog-Index Index of readability, denotes Equal to 0.4*(average number of words
proxies for annual reports’ per sentence +percent of complex
readability level words). High values of the Fog Index
imply less readable text
Independent variables
DA 1 Discretionary accruals using Absolute value of residuals estimated
Jones modified (1995) model using the Jones modified (1995) model
DA 2 Discretionary accruals using Absolute value of residuals estimated
Raman and Shahrur (2008) using Raman and Shahrur (2008)
Control variables
Float Ownership dispersion The percentage of shares held by the
public
Profit Profitability Net Income of the company divided by
Total Assets
FoSales Foreign Sales Foreign sales divided by total sales
Lev Leverage The ratio of total debt to total assets
FirmSize Firm size The natural logarithm of total assets

Table 2
Descriptive statistics.

Mean Std deviation Min Max

Fog-Index 20.83 2.105 17.081 26.612

DA 1 0.795 3.593 0 115.64
DA 2 0.531 2.965 0 59.902
Float 0.048 0.189 0 0.29
profit −0.599 4.78 −9.458 4.592
FoSales 0.35 0.41 0 0.98
Lev 0.219 0.232 0 0.831
FirmSize 5.968 2.296 2.263 11.922

with a K–1 ◦ of freedom. If we cannot reject the null hypothesis, that is to say if the p-value is greater than the level of
confidence, then we use the random effects that are effective
The results of the Hausman test, not reported here, show that the fixed effect model is preferable to the random effect.
This method allows us to focus on changes in the variables over time to estimate the effects of the independent variables on
dependent variable.
Our model is as follows:

(Fog − Index) = ␣1 + ␣2DA + ␣3 (Float) + ␣4 (Profit) + ␣5 (FoSales) + ␣6 (Lev) + ␣7 (FirmSize) + ␧

With: DA: discretionary accruals estimated using three models: Jones modified (1995) and Raman and Shahrur (2008).

4. Results and discussion

4.1. Descriptive analysis

Table 2 shows the sample statistics. The mean value of the dependent variable, measured by Fog-Index, is around 20.83.
This value is similar to that reported in Li (2008), for a sample of 55,719 firm-years with annual report filing dates between
1994 and 2004, where the mean and median were 18.23 and 17.98, respectively and suggest that firms’ annual reports are,
on average, difficult to read. Moreover, FOG displays a substantial cross sectional variation as evidenced in the standard
deviation of 2.24.
The variable, measured by the absolute value of discretionary accruals, displays a level of 0.795 according to the Jones-
modified model (1995). This is comparable to the level achieved by the Raman and Shahrur (2008) model, which establishes
a level of 0.531, and lower than that found by Mard and Marsat (2012) on a sample of listed companies of the SBF index 250
over the period 2004–2008.
The percentage of shares held by the public is 44.1%. This shows that the French listed companies have a relatively
concentrated ownership structure. Net Income of the company divided by Total Assets represents performance measure in
the paper. Table 2 depicts that profitability is −0.599. The mean value of total debt on total assets ratio for the sample is
0.219.
 A. Ajina et al. / Research in International Business and Finance 38 (2016) 509–516 515

Table 3
Panel regression, Fog-Index.

Variable Fog-Index Fog-Index

Coef t student Coef t student

DA1 0,125** 2,48

DA2 0,252** 2,97
Float −2324** −1,99 −2185** −2,20
profit −0,856* −1,86 −0,328** −2,15
FoSales −0,325 −1,21 −0,096 −1,24
Lev 0,241** 1,99 0,190*** 2,92
FirmSize 0,236** 2,19 0,219** 2,52
Constant 1602 0,65 −0,128 −0,46
R2 0,421 0,472
Fisher 22,29*** 27,98**

Standard errors are given in parentheses; *, **, and *** indicate significance at the 10%, 5%, and 1% levels, respectively.

4.2. Multivariate analysis

Table 3 presents the multiple regressions results aiming to determine the effect of earnings management on the readability
of annual reports. The results show the existence of a positive and significant relation between the level of discretionary
accruals measured by models of Dechow et al. (1995) and Fog Index. This confirms our hypothesis according to which firms
that have managed earnings have annual report disclosures that are less readable. The firms that manage their earnings
publish reports with the highest Fog Index (the least readable reports).
This result confirms that, when managers notice the drop in earnings, they may exploit some discretion left regarding
accounting choices or regarding operations structuring. This has the aim of generating modification of the risk of transferring
company wealth, such as this risk is perceived in practice by the market. In this context, they have to try to hide this behavior
by increasing the complexity of their financial statements. This, the financial information quality will be indirectly affected
by this new complexity. This report indicates that some managers try to conceal managed and manipulated earnings by
publishing less readable reports. This result supports the predictions of Li et al. (2011).
The author was able to confirm the hypothesis that earnings management incites the managers to publish difficult to
read and less readable annual reports. Our empirical results are also in accordance with the theoretical report stipulating
that to tell the truth is easier than to tell lies (Hancock et al., 2007). Lying is difficult if it is to be convincing because the
communicator has to ensure the consistency of the supposed “facts.” While earnings management in many cases does not
fall into the lying category outright, the activity does involve some active efforts on the part of management to bias the
financial statements through accruals or other means.
This result is similar when using a Raman and Shahrur model (2008) as an alternative measure for earnings management.
The link between ownership dispersion and annual report readability is positive and significant. In companies with high
dispersal, the costs of agency increase because of the greater probability of conflicts of interest between owners (Jensen and
Meckling, 1976). Consequently, to counter this effect, the companies with more dispersed ownership publish more readable
annual reports. Oliveira et al. (2006) have confirmed this relation.
As regards the profitability variable, we find a positive and significant impact on the readability of annual reports. Thus,
we confirm the ontological explanation proposed by Bloomfield (2008) i.e., that good performance is inherently easier to
communicate than bad performance. Bad and losses news is more difficult to describe. Managers can reduce market response
to bad news by making bad news more costly to analyze. Li (2008) tests this hypothesis and reports that managers make
bad news costly by writing excessively long annual reports with unnecessary big words and long sentences.
Debt ratio is positively correlated with Fog Index, indicating that companies with more debt produce less readable annual
reports. Managers attempt to make their annual reports difficult to read to disguise the reasons behind their high debt ratio.
Firm size is also positively correlated with Fog Index, indicating that bigger companies provide difficult to read annual
reports. A larger firm will have more complex operations and will provide longer and more complex annual reports. Ajina
et al. (2015) and Botosan (1997) confirm this relationship.

5. Conclusion

The objective of this article is to give a perspective on an issue that is seldom researched in France and relative to the
impact of earnings management on annual report readability. The extant literature is not convergent. The results obtained
in various contexts or in the United States are not similar, given the differences in institutional settings among countries.
The obtained result shows the existence of a positive and significant link between the level of discretionary account-
ing adjustments and annual report readability. This result is confirmed by using both measures of earnings management.
Companies using earnings management techniques have the highest (least readable) Fog Index. This result confirms that,
when managers notice a drop in results, they attempt to hide future profits by increasing the complexity of their financial
statements. Their aim is to modify the wealth transfer risk of the company.
516 A. Ajina et al. / Research in International Business and Finance 38 (2016) 509–516

In this context, they have to attempt to hide this behavior by increasing the complexity of their annual report. Lying is
difficult if it is to be convincing because the communicator has to ensure the consistency of the supposed “facts.” While
earnings management in many cases does not fall into the lying category outright, the activity does involve some active
efforts on the part of management to bias financial statements through accruals or other means.
This confirms the importance of annual report readability. The authorities must increase the regulations and recom-
mendations in favor of readability in financial documents. In defining quality of information, readability must be taken into
account. Readability may establish some confidence on stock markets and attract investors. Besides, an interesting area of
research may focus on the effects of annual report readability on stock market liquidity.
Our research has a limitation. In fact, annual report complexity is studied as a whole and does not take into account the
nature of the information. Further research could allow us to examine and assign one particular Fog Index to each category
of information.

References

Adelberg, A., 1979. Narrative disclosures contained in financial reports: means of communication or manipulation? Account. Bus. Res. 9, 179–189.
Ajina, A., Danielle, S., Lakhal, F., 2015. Corporate disclosures, information asymmetry and stock-market liquidity in France. J. Appl. Bus. Res. 31, 223–238.
Antweiler, W., Frank, 2004. Is all that talk just noise? the information content of internet stock message boards. J. Finance 59, 1259–1294.
Ben Othman, H., Zéghal, D., 2006. A study of earnings-management motives in the Anglo-American and Euro-Continental accounting models: the
Canadian and French cases. Int. J. Account. 41, 406–435.
Berle, A., Means, G., 1932. The Modern Corporation and Private Property. MacMillan, New York.
Biddle, C., Hilary, G., Verdi, S., 2009. How does financial reporting quality relate to investment efficiency? J. Account. Econ. 48, 112–131.
Bloomfield, R.J., 2002. The “incomplete revelation hypothesis” and financial reporting. Account. Horiz. 16 (3), 233–243.
Bloomfield, J., 2008. Discussion of Annual report readability, current earnings, and earnings persistence. J. Account. Econ. 45, 248–252.
Botosan, C., 1997. Disclosure level and the cost of equity capital. Account. Rev. 72, 323–349.
Coase, H., 1937. The nature of the firm. New Ser. 4 (16), 386–405.
Copeland, T., Galai, D., 1983. Information effects on the bid-ask spread. J. Finance 38, 1457–1469.
Courtis, K., 1997. Corporate annual report graphical communication in Hong Kong: effective or misleading? J. Bus. Commun. 34, 269–288.
Courtis, J.K., 2004. ‘Corporate report obfuscation: artefact or phenomenon?’. British Account. Rev. 36 (3), 291–312.
De Franco, G., Hope, K., Vyas, G., Zhou, Y., 2013. Analyst report readability. Contemp. Account. Res., Institutional Investor. 2010 (October) (Forthcoming).
DeFond, L., Jiambalvo, J., 1994. Debt covenant violation and manipulation of accruals. J. Account. Econ. 17, 145–176.
Dechow, M., Sloan, G., Sweeney, P., 1995. Detecting earnings management. Account. Rev. 70, 193–225.
Fama, E.F., Jensen, M.C., 1983. S̈eparation of Ownership and Control¨. Journal of Law and Economics 26, 301–325.
Godfrey, J., Mather, P., Ramsay, A., 2003. Earnings and impression management in financial reports: the case of CEO changes. Abacus 39 (1), 95–123.
Gujarati, D.N., 2004. Basic Econometrics. McGraw Hill, USA.
Hancock, J.T., Curry, L.E., Goorha, S., Woodworth, M., 2007. On lying and being lied to: a linguistic analysis of deception in computer-mediated
communication. Discourse Processes 45, 1–23.
Holmstrom, B., 1979. Moral hazard and observability. Bell J. Econ. 10 (1), 74–91.
Hope, O.-K., Thomas, W., Vyas, D., 2011. The cost of pride: why do firms from developing countries bid higher? J. Int. Bus. Stud. 42 (1), 128–151.
Hossain, M., Perera, M.H.B., Rahman, A.R., 1995. Voluntary disclosure in the annual reports of New Zealand companies. J. Int. Financ. Manage. Account. 6
(1), 69–87.
Jensen, M.C., Meckling, W.H., 1976. Theory of the firm: managerial behaviour, agency costs and ownership structure. J. Financ. Econ. 3, 305–360.
Jones, M.J., Shoemaker, P.A., 1994. Accounting narratives: a review of empirical studies of content and readability. J. Account. Lit. 13, 142–184.
Keasey, K., Short, H., 1997. Equity retention and initial public offerings: the influence of signalling and entrenchment effects. Appl. Financ. Econ. 7, 75–85.
Lawrence, A., 2013. Individual investors and financial disclosure. J. Account. Econ. 56, 130–147.
Lehavy, R., Li, F., Merkley, K., 2012. The effect of annual report readability on analyst following and the properties of their earnings forecasts. Account. Rev.
86, 1087–1115.
Li, F., 2008. Annual report readability, current earnings, and persistence. J. Account. Econ. 45, 221–247.
Li, F., Abeysekera, I., Ma, S., 2011. “Earnings management and the effect of earnings quality in relation to stress level and bankruptcy level of Chinese listed
firms”. Corp. Ownersh. Control 9 (1), 366–391.
Loughran, T., McDonald, B., 2011. When is a liability not a liability? Textual analysis, dictionaries, and 10-Ks. J. Finance 66, 35–65.
Machuga, S., Teitel, K., 2007. The effects of the Mexican corporate governance code on quality of earnings and its components. J. Int. Account. Res. 6, 37–55.
Mard, Y., Marsat, S., 2012. Gestion des résultats et structure de l’actionnariat : le cas français. Comptabilité-Contrôle-Audit 18 (3), 11–42.
Merkl-Davies, D.M., Brennan, N.M., 2007. Discretionary disclosure strategies in corporate narratives: incremental information or impression
management? J. Account. Lit. 26, 116–196.
Miller, B., 2010. The effects of reporting complexity on small and large investor trading. Account. Rev. 85, 2107–2143.
Morris, R., 1987. Signalling, agency theory and accounting policy choice. Account. Bus. Res. 18, 47–56.
Oliveira, L., Rodrigues, L.L., Craig, R., 2006. Firm-specific determinants of intangibles reporting: evidence from the Portuguese stock market. J. Hum.
Resour. Cost. Account. 10, 11–33.
Raman, K., Shahrur, H., 2008. Relationship-specific investments and earnings management: evidence on corporate suppliers and customers. Account. Rev.
83, 1041–1081.
Ramanna, K., Watts, R.L., 2011. Evidence on the use of unverifiable estimates in required goodwill impairment. Rev. Account. Stud., Forthcoming.
Gunning, R., 1952. The Technique of Clear Writing. McGraw-Hill International Book Co., New York, NY.
Ross, S., 1977. The determination of financial structure: the incentive—signalling approach. Bell J. Econ. 8, 23–40.
Rutherford, B.A., 2003. Obfuscation, textual complexity and the role of regulated narrative disclosure in corporate governance. J. Manage. Governance 7,
187–210.
Sydserff, R., Weetman, P., 2002. Developments in content analysis: a transitivity index and DICTION scores. Account. Audit. Account. J. 12, 1459–1488.
Tetlock, J., Paul, C., 2007. Giving content to investor sentiment: the role of media in the stock market. J. Finance 62, 1139–1168.
Tomoni, B., 2012. Conceptual metaphors in financial discourse: the ‘Romanian Way’ of thinking, speaking and writing finances. Selected Papers from
UK-CLA, vol. 1., pp. 146–163.
Zéghal, D., Chtourou, S., Sellami, Y.M., 2011. An analysis of the effect of mandatory adoption of IAS/IFRS on earnings management. J. Int. Account. Audit.
Tax. 20, 61–72.