Accounting Variables Between Updating And Manipulation Of Earnings Accounting Essay

Published:

In this article we look at the possible relation between the IAS-IFRS, which promotes a present value, and manipulation of net income enhanced by the application of these standards. We are expanding, to the French and Tunisian context, the problem of information content of the discretionary component of accounting income in connection with the application of IAS-IFRS.

On the basis of a sample of 120 firms-year observations, chosen from Tunisia, which uses national standards based on the international one, and a sample of 120 firms-year observations selected from France, that adopted the IAS-IFRS since 2004, we studied the relationship of association between accounting variables and dividend yields on one side and the impact of the use of IAS-IFRS on discretionary accruals in on another side. In other words, we tried to analyze the advantages and disadvantages of the adoption of IAS-IFRS.

The study results show that the application of IAS-IFRS accounting standards increases the information content of accounting numbers. The published accounting variables explain at 33% the dividend yield. However, the adoption of IAS-IFRS allows managers to exercise their discretion to adjust the financial statements based on circumstances and certain requirements. Managers believe that their intervention in earning management increases the credibility of accounting information but also increases the level of discretionary accruals in the company. Among the ten accounting variables selected in the model, the results show that five variables are manipulated and undergo a change.

Keywords. IAS-IFRS - ratios - dividend yield - accruals

Introduction

Lady using a tablet
Lady using a tablet

Professional

Essay Writers

Lady Using Tablet

Get your grade
or your money back

using our Essay Writing Service!

Essay Writing Service

The financial statements published by companies to investors remain a source of insider information. Several empirical studies have shown the existence of an association between accounting variables and stock returns such as those of Daniel, Collins & Kothari (1989), Lev & Thiagarajan (1993), Beaver et al. (1997), Kothari et al. (2003) and Bao (2004). In addition, some studies have shown the superiority of the explanatory power of accounting income on other accounting figures (Beaver et al. (1997), Kothari et al. (2003) and Bao (2004)). Other studies are based on the usefulness and necessity of all the variables accounting standards issued by the firms. In fact, each variable has its own effect explanatory Performance (Ali & Zarowin (1992), Lev & Thiagarajan (1993), Chan & Seow (1996), Kothari (2001), Hung & Subramanyam (2004)).

However, all these studies have led to what the difference between the economic value and the book exists and is real. Reducing this gap is only possible if all firms use the same accounting system and that system is based on the fair value disclosure of financial statements quality.

The IASB has established international standards (IAS-IFRS) to ensure, on one side, more comparability and transparency and the other, a better quality of financial statements. The establishment of such standards by the IASB aims at achieving harmonization of accounting practices between countries. According Nobe (1998), IAS-IFRS can publish financial statements that reflect the true picture of the firm, to protect the interests of investors by carefully calculating distributable profits and predict future cash flows.

Some studies have shown that differences between national accounting systems affect the importance and credibility of accounting information (Hung (2001), Ali & Hwang (2000), Ball et al. (2000), Pope & Walker (1999), Joos & Lang (1994) and Alford et al. (1993)). According to Gelard (1989), it is necessary to adopt an effective procedure. This should not be a compromise between systems often contradictory among themselves, but by a rational choice for the most effective way to reach the goal. He noted also that we must not ignore the international character of the IAS-IFRS that may negatively affect some environments.

For other researchers, the global application of IAS-IFRS accounting standards as admitted it would be difficult to achieve despite the benefits it can bring about the publication of financial statements with all the qualities required and sought by investors (Chamber (1976), Nobes (1995) and Hoarau (1995)).

In addition, the IAS-IFRS are known by their flexibility. In fact, they enable managers to exercise their judgments to publish results based on discretionary targets (Watts & Zimmerman (1986, 1990)). They seek their intervention to strengthen the information content of accounting figures in order to better inform the market about the prospects of the firm. According to Healy and Wahen (1999), the application of IAS-IFRS allows managers to exercise their discretion to adjust the financial statements based on circumstances and certain requirements. They believe that their intervention in the management of results increases the credibility of accounting information but also increases the level of discretionary accruals in the firm.

Lady using a tablet
Lady using a tablet

Comprehensive

Writing Services

Lady Using Tablet

Plagiarism-free
Always on Time

Marked to Standard

Order Now

Contrary to most published studies on this subject in this article we look at the possible relation between the IAS-IFRS, which promotes a present value, and manipulation of accounting results enhanced by the application of these standards. In fact, in this paper, we extend, to the French and Tunisian companies, the problem of information content of the discretionary component of accounting income in connection with the application of IAS-IFRS.

A comparison of this context for Anglo-Saxon work is particularly interesting given the specificities French and Tunisian context.

The aim of this paper is to examine the merits of these changes in accounting systems in Tunisia and France and to analyze the impact of these decisions on the information content of financial statements on the one hand and on the variation manipulation of accounting results under IAS-IFRS, on the other. We chose to compare the practices of performance management in different contexts, two countries with different accounting standards, a developed country such as France and developing countries such as Tunisia. Both countries have chosen different contexts of accounting standards and financial market. France offers a framework based on international standards and Tunisia provides a framework for appropriate standards at national level.

Considering the environmental nature of the two countries, our empirical approach is to measure the effect of the IAS-IFRS and local GAAP on the information content of the discretionary component of the result in France and Tunisia. This is done the two followings steps:

The first is based on the study of the association relationship between stock returns and book values in two different financial structures.

- The second is the effect of the use of international standards on the handling of the outcome of the exercise.

We used a sample of 120 Tunisian firms-year observations (2004-2007), which uses national standards based on international standards. The Tunisian accounting system, which is mandatory since 1997, is adapted to its economic, legal and tax environment.

In France, the transition to IAS has been spread over time. The French companies have gradually applied these financial standards, until 2004, financial publications have made only under French GAAP in 2005, consolidation of accounts according to standards IFRS is optional, in 2006, companies must publish their financial statements to IFRS in 2007, the international benchmark is the repository of applicable law.

Given the requirement to present a comparative exercise in the same accounting basis, the actual date of transition to IFRS is January 1st 2004, which is the first day of the year provided for comparison. We selected a sample of 120 French firms-year observations (2004-2007), which adopted the IAS-IFRS with everything that they allow flexibility and multiple-choice left to businesses and their managers.

Our study has three sections: The first development literature and research hypotheses. The second presents the methodological aspects. The third discusses the empirical results. In conclusion, we summarize the main contributions of this work.

1. Literature review and hypotheses:

1.1. Accounting data and stock returns:

The issue of the usefulness of accounting information in decision-making in financial markets as a topic that has been widely debated in the larger Anglo-Saxon academic journals since the late sixties. The first study in this direction was that of Ball & Brown (1968), it was mainly intended to verify the information contained in the annual net profits of a business is useful to the stock market. Net income was selected as the research data of Ball & Brown, accounting information most relevant to investors and financial analysts on that day and it provided a relatively reliable information about the firm's future prospects.

This research is based on accounting information contained in the reports certified by qualified auditors (Daniel, Collins & Kothari (1989), Beaver et al. (1997), Kothari et al. (2003), Bao (2004), Lev & Thiagarajan (1993)). These researchers based in their analysis on two important assumptions:

The accounting system in effect provides accounting and financial data to meet the multiple needs of investors and the full range of users (Ball 2006).

To assess the value of the company if it allows a recovery closely related to that which appears in the reading courses in the stock market. This quality information helps bridge the financial reality of the business, as reflected in the accounts of the reality perceived by investors (Herrman & Wayne (1995)).

Lady using a tablet
Lady using a tablet

This Essay is

a Student's Work

Lady Using Tablet

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

Examples of our work

The majority of these studies were based on accounting data published by companies that use the international model in keeping their accounts. Indeed, the IAS-IFRS are considered by some researchers as the most reliable accounting model (Bernhen (1997), Gelardi (1997)). The IASB is considered a source of inspiration and a benchmark for national standard setters, and also holds a very important role, which alone justifies its existence because it is the only global organization of accounting where many cultures rub. He added that the IASB is a body of principles for large multinational publicly traded and that its standards should not be applied directly in the countries having no national standards body.

Guerreiro & Rodrigues (2008) studied the effect of the application of IAS-IFRS in the Portuguese context, it is clear from their analysis that countries planning to adopt the IAS-IFRS must prepare and take all legal precautions, tax, commercial, auditing and accounting. Indeed, the adoption of these international standards will change not only the system of standardization but also all the related disciplines.

According to Yalkin et al. (2005), the adoption of IAS-IFRS by banks and listed companies in Turkey, since January 2005, provides information comparable with international firms. The authors add that application of IAS-IFRS can be spread in the future, to other institutions.

Unlike Bernhen (1997) and Yalkin et al. (2005), Bhimani (2008) shows, in a study that the solution to the crisis lies not only in the adoption of IAS-IFRS. The adoption of these standards is not the logical solution to meet the needs of operators and to face the challenges posed by the crisis. Bhimani (2008) explains that international relations, globalization, trade and national economic growth should form infrastructure for legitimate international accounting standards.

In the same current of ideas, Nobe (2009) shows that there is a significant difference between national accounting standards and international ones. The bases and measurement principles of accounting variables in the two systems differ, which makes their applicability difficult in the rough.

For Glaum (2000), Leuz & Verrecchia (2000) and Dumontier & Raffournier (1998), the adoption of IAS-IFRS significantly improves the information content of accounting and financial information published by companies.

According to Aboody et al. (1999), the effect of applying the "fair value" is substantially higher than historical cost because the capital and other financial instruments are measured at market value, which increases the correlation between accounting income and stock returns.

For Cormier & Magnan (2005), the accountant must acquire knowledge and superior skills and this, to understand the business context and grasp the future after the transition from national standards to the international ones. Similarly, Carmona & Trombetta (2008) consider the IAS-IFRS as the basis for the success of the process of accounting harmonization, but it still seems far from achieving the level of comparability of measures of financial and accounting information between countries (Tort 2005). Ball (2006) considers, for its part, the economic value or market value is infinitely more informative than the book value.

For Boukari & Richard (2007), the transition to new accounting standards lead to lower equity and higher financial indebtedness. This is due to the revaluation of the repurchase of shares of the minority who are considered by French standards as liabilities. Boukari & Richard (2007) add that the net result has been more strongly influenced by the transition to IAS-IFRS and other accounting systems reviewed by the study. The same study shows that despite positive changes in results of companies listed after the introduction of IAS-IFRS, a weak reaction of French financial markets is felt following the publication of the accounting figures of the transition.

Heverals (2007) presented a study on the impact of the adoption of IAS-IFRS on the tax base of Belgian companies. Indeed, among the objectives of the adoption of these standards by the EEC, it is necessary to have a common tax base for each sector between the member countries. His study shows that the impact of the use of these standards in calculating the tax base is very wide and varies from one sector to another. Some sectors, like construction and automobiles, pay much higher taxes than others. In general, the introduction of IAS-IFRS in the preparation of financial statements increases significantly the tax base for all sectors.

Milot (2005) has suggested the use of IAS-IFRS in the public sector. Although, commercial and for-profit companies used these standards, they can be adapted to the public sector as they can give a clearer picture on costs and revenues of these entities. They also control their social obligations and their dealings with other public and private entities (Toms 2002)).

According to Blanco et al. (2004), the adoption of IAS-IFRS by the US companies will not exclude GAAP (General Accepted Principal Accountant). In fact, his comparative study between the two poles of accounting rules shows that the IAS-IFRS and GAAP converge and are even complementary (Colasse (1997)). According to the above-mentioned studies, we can formulate the following assumptions:

H1a: The adoption and / or adaptation of IAS-IFRS will increase the information content of dividends.

H1b: The adoption and / or adaptation of IAS-IFRS will increase the information content of fixed assets, turnover and liabilities.

H1c: The adoption and / or adaptation of IAS-IFRS will increase the information content of the earnings.

H2: The accounting variables set according to local GAAP have an informative effect below those established by international standards.

1.2. Accounting information and earnings manipulation:

According to the studies of association, earning is considered as the most used accounting variable in studying the relationship between book value and economic value. This result is composed of two parts:

Some reliable part corresponding to incoming and outgoing financial flows documented.

And another part called "accruals", which characterizes the accrual accounting.

These accruals allow the principle of matching costs with revenues in calculating the earning of the current year. However, these accruals may become subject to manipulation and adjustment of net income by the management. Part of manipulated accruals is called the discretionary accruals.

These adjustments can be made using different kinds of processes, referred to as natural (Blanco et al. (2004), Heverals (2007)) and institutional (Ashbaugh & Pincus (2001), Santella & Turrini (2008)).

Taking into account the discretionary component in the relationship between stock returns and accounting data in an environment using the IAS-IFRS is a relatively new problem and has become the subject of some new research, especially after the last economic crisis.

The reasons for earning management are many and studies have addressed this subject are important. However, we focus on studies linking the impact of these manipulations on the relationship performance - accounting variables.

Ball et al. (2003) and Tendeloo & Vanstraelen (2005) showed in their studies that the informational quality of accounting information depends on the degree of adaptation of accounting standards to the country's economic environment.

Ashbaugh & Pincus (2001) went even further and concluded that the use of IFRS may increase or decline to publish the earnings, altering the information content of accounting and financial data.

In the same current of ideas, Tort (2006.a, 2006.b, 2006.c and 2007) confirmed the existence of a contingency of the accounting policy of listed companies studied in the context of transition to IFRS consolidated accounts. He added that these companies adopt different trading strategies to IFRS while integrating structural and external constraints in these strategies in order to adjust their accounting policy disclosure. He also confirmed the results of Tendeloo & Vanstraelen (2005), Stolowy & Breton (2003) who considered that the transition to IFRS helps managing accounting data by corporations and more specifically gives access to the information of consolidated balance sheet.

For Santella & Turrini (2008), the distribution of dividends computed on the basis of IAS-IFRS may lead to unrealistic profits and unjust enrichment. Using these standards as the basis for accounting generates changes in the traditional accounting methods and mechanisms.

Aussenegg et al. (2009) chose to study the level of earnings management by replacing the local GAAP with IAS-IFRS, for 15 European countries using a sample of over 18,000 firms-years observations. The results of their study show that countries belonging to the German legal system and some countries belonging to the French legal system experience a significant decline in performance management of their businesses compared to their counterparts that use local GAAP. These results are not confirmed in the country following a statutory English and other Northern European countries. In these two regions, the authors found that companies handling more than their results after the adoption of IAS-IFRS. Aussenegg et al. (2009) add that the "time factor" and "size of the firm" had no significant effect on the level of earnings management. However, growing companies promote a high level of manipulation of results.

In their study, Aussenegg et al. (2009) tested four hypotheses:

Companies that adopted IAS-IFRS are not engaged in earning management more than those using local GAAP.

Growing companies tend to engage more earning management.

Companies with significant cash flows have fewer earning management.

Companies with significant financial power tend to have more earning management.

Based on the use of discretionary accruals to stabilize the value of the outcome of the exercise, LaPointe et al. (2006) studied in the context of Switzerland, the effect of the use of accruals on the result set on the basis of IAS-IFRS, US GAAP and Swiss standards. Their study shows that Swiss firms use discretionary accruals to stabilize results. They add that the value of these results decreases when calculated on the basis of IAS-IFRS or U.S. GAAP. This shows that the adoption of IAS-IFRS accounting treatment negatively affects the handling of the results of exercises.

According to Lapointe et al. (2006), the structure of the accounting system of Switzerland is favorable to engage companies in earning management. This system gives the Swiss companies the choice of measurement method and does not impose many rules and standards of substance (Achleitner (1995) and Raffournier (1995)). International standards allow companies to present users with more detailed information. To take advantage of the transparency and quality of this information, the Swiss financial market demand, therefore, to listed companies to publish financial statements prepared on the basis of IAS-IFRS. The purpose of this recommendation is to maintain the confidence of investors and shareholders who find it difficult to have information that assist them in making their decisions, when financial statements prepared on the basis of Swiss GAAP.

Using the approach of the information content, Wang & Williams (1994), Michelson et al. (1995) and Subramanyam (1996) showed that there is a relationship of positive association between income stability measured and stock returns. Using the approach of the relevance of accounting results, Habib (2004) found that the stability of income contributes to the decline in relevance. From this literature we can formulate the following assumptions:

H3a: The use of discretionary accruals is positively associated with the value of shareholders' equity and fixed assets in companies that adopt IAS-IFRS.

H3b: The use of discretionary accruals is negatively associated with the value of liabilities in firms that adopt IAS-IFRS.

H4: Both firms that adopt GAAP and IAS-IFRS manage their earnings.

2. Methodological aspects:

2.1. Sample:

This study is based on a sample of 120 Tunisian firms-year observations, which uses national standards and a sample of 120 French firms-year observations, which adopted the IAS-IFRS since 2004. We eliminated from the two by financial firms and other companies with sector based accounting rules. Our choice of the period from 2004 to 2007 is justified by the entry into application of IAS-IFRS by French companies.

2.2. Accounting and Financial Variables:

Following the interviews conducted and the accounting and financial information collected annually from Tunisian listed firms, we can divide the operations in the past four financial years (2004-2007), object of study in two sets. The first set, related to the cyclical or current operation, can isolate single concepts related to business activity: Turnover, value added, operating surplus, operating resources. The second, which considers aggregate behavior, takes into account not only the operating functions, but other functions related to external growth, portfolio management, debts and cash management, etc. This allows the study of different ratios related to the financial structure, margins and overall income and financing.

Among these ratios, 10 were selected in our study. They are the key points of the analysis made by financial analysts (Baccouche & Nafti (2007)), when the introduction of the company stock, or, after its introduction.

Table 1: Selected accounting variables

Variables

Nature of Each Variable

V1

Dividend per Share

V2

Customers Turnover: Turnover / Customers

V3

Assets Turnover: Turnover / Fixed Assets

V4

Equity Turnover: Turnover / Equity

V5

Non-Current Liabilities / Equity

V6

Non-Current Liabilities / Earnings

V7

Equity / Net Fixed Assets

V8

Performance: Earnings / Turnover

V9

Earnings / Equity

V10

Gross profit / Turnover

The yield per share for a period of time t can be expressed as follows:

Pit is the value of investment at beginning of the period

P it+1 is the value of this investment at end of the period

Dit is dividends paid during the period

datt represents the rights of attribution

dsous represents the rights of subscription.

2.3. Models of the study:

2.3.1. Model of the relationship of yield-accounting variables association:

Empirical validation of the relationship dividend yield-accounting variables requires to measure at a given date the relationship between the Yit value or dividend yield Rit for the firm i and one or more accounting variables intended to reflect the ability of the firm to create wealth and pay dividends:

Yit = 0 + 1 VC1it + 2 VC2it + 3 VC3it + 4 VC4it + 5 VC5it + 6 VC6it + 7

VC7it + 8 VC8it + 9 VC9it + 10 VC10it + i + t

With:

Yit represents the dividend yield

VCkit represents the accounting variable k published by the company (i) at the date (t)

i represents the companies specific effects

t is the error term

The R2 of the regressions provide evidence of the intensity of the relationship between the level or changes in share price and accounting numbers. Most studies have chosen the only book income as a predictor of profitability or market value of the company, though it shows rather poor performance and that its coefficient of determination is low. The model developed and tested in the study of Beaver et al. (1997) aims to measure the relationship between changes in share price and those of book earnings. The advantages of such a study can be summarized as follows: all the alternative estimation procedures, which are aimed at fundamental issues of identifying variables, are developed and tested. In addition, the comments about the weak relationship (R2) found by previous studies are placed in a different perspective. According to this study, the benefits may change for reasons that have no relation to the price change, and the share price may also change due to the existence of events, which have no relation with accounting profits. Finally, prices and profits are two endogenous variables.

Kothari et al. (2003) used, as the case of several other researches, earnings as a benchmark in relation to dividend yield - profits. The specificity of their study lies in the fact that it is based on the aggregated earnings (aggregate earnings news). According to this study, the relationship dividend yield - earnings is significantly different in the case of aggregate data.

2.3.2. Models of accruals and accounting variables:

2.3.2.1. Measuring discretionary accruals:

The measure of the degree of performance management used in our study is the absolute value of discretionary accruals (AbDA). The estimation of discretionary accruals will be realized based on the model of Jones (1991) and on the model of Subramanyam (1996). However, to estimate discretionary accruals, it must first determine the total amount of accruals for each firm and each year. Due to lack of financial data for the application of the approach based on the balance sheet, we will use the cash flow approach to calculate total accruals in accordance with JeanJean (2001) research.

The total accruals (TAit) are calculated as the difference between the net result of a year (NRit) and cash flows generated by the activity of the company during the same period (CFEit).

TAit = RNit - CFEit

With

TAit: The total accruals of the firm i in year t

NRit: The earnings of the firm i in year t

CFEit: The operations cash flow of the firm i in year t

TAit = DAit + NDAit

With

TAit: The total accruals for the firm i in year t

DAit: The discretionary accruals of the firm i in year t

NDAit: The non-discretionary accruals of the firm i in year t

The model is as follows:

TAit / Ait-1 = α1 (1 / Ait-1) + α 2 ((ΔCAit - ΔCREit) / Ait-1) + α 3 (IMMOit / Ait-1) + εit

With

TAit: total accruals in period t for firm i

ΔCAit: Change in turnover between t and t-1 for firm i

ΔCREit: Change in receivables between t and t-1 for firm i

IMMOit: gross non-financial fixed assets in year t for firm i

Ait-1: Total Assets at end of period t-1 for firm i

2.3.2.2. Model of the association relationship accruals-accounting variables:

The multivariate analysis is used to estimate the coefficients of the regression model. First, this model allows us to demonstrate the association relationship between accounting variables (based on the international model) and discretionary accruals. Second, it can analyze the same relationship between the accounting variables (based on local GAAP) and discretionary accruals.

DAit = β0 + β1V1it + β2V2it + β3V3it + β4V4it + β5V5it + β6V6it

+ β7V7it + β8V8it + β9V9it + β10V10it + eit

With

DAit: discretionary accruals (in absolute value).

VC k it represents the variable amount k (number or ratio) published by the company (j) the date (t).

it represents the error term.

3. Results of panel data:

3.1. The information content of accounting data:

The independent variables were introduced to the model one by one. We eliminated non-significant variables and having no information content on performance. In fact, for the results of the regression between stock returns and accounting variables to be valid and interpretable, it must not introduce any variables into the model simultaneously, and this, to ensure that the introduction of a variable has no effect on the relationship of another variable with performance. The introduction of all variables at the same time to the model may conceal a relationship of some variables with the performance because of the greater effect of the first variable.

3.1.1. The relationship between dividends, performance and yield:

3.1.1.1. Case of Tunisia:

The adoption by Tunisia, since 1997, of an accounting system based on the International one is a first step towards effective adoption of international standards. The conceptual framework of the Tunisian accounting system includes concepts that are very close to that of the IASB.

The production quality of accounting information is regarded as one of the important issues in accounting and is appreciated for its ability to meet the needs of users.

The results of the panel data are obtained on the basis of estimates based on the fixed model (fixed effect) and based on the Hausman test which has a chi2 value of 4,25 with prob> chi2 equal to 0.043. This means that the hypothesis is accepted and corr(ui,V) = 0. This is a prerequisite for the GLS (Generalized Least Squares) is BLUE (Best Linear Unbiased Estimator) and more convergent (if N tends to ∞, â tends to a).

Table 2: Regression Results: informational content of the dividend per share and performance

Model A: Yieldit = a0 + a1 divit + i + εi

Parameter estimate

t

Probability> | T |

Constant

0.58

2.54

0.013*

V1

-0.36

-2.06

0.043**

Adjusted R²: 6%

Model B: Yieldit = a0 + a1 divit + a2 earnings/turnoverit + i + εi

Parameter estimate

t

Probability> | T |

Constant

0.516

2.22

0.03**

V1

-0.36

-2.11

0.039**

V8

0.74

1.64

0.1***

Adjusted R²: 10%

Significant at: *1%, **5% and ***10% (***).i represents the specific effects.

These results show that the variable chosen, V1 (dividend per share) explains the market return at 6%. It is a fairly low rate, but considered as interesting in a regression performed on a number of observations of 120 (30 companies observed over 4 years).

Most empirical studies in this direction have proved the reality of the existence of a correlation between these two types of variables (accounting and economic) with percentages that differ from one study to another (Ball & Brown (1968), Toms (2002), Cheng (2005) and Bao (2004)). The dividend is for the Tunisian investor an important goal. It is the result of its investment in companies and a valuable part of the earnings.

Some studies have shown that earnings, dividends and equity are the most useful accounting information to the end users (Cheng (2005), Toms (1998), (2002), William et al. (1997)).

The introduction of the variable V8 (earnings/turnover) the model increases the adjusted R2 from 6% to 10%. The rate remains quite low but the growth shows the importance of the V8 variable in explaining the variation in performance (Lev & Thiagarajan (1993)).

The importance of performance ratio (Haller & Schlobgang (2005)), in making investment decisions, is the importance of its components namely, earnings and turnover, which are essential in the understanding of the internal management of corporate resources.

The empirical analysis of Ou & Penman (1989a and b), Lev & Thiagarajan (1993) and Abarbanell & Bushee (1997 and 1998) show that the variation of turnover, changes in equity and change in gross margin are strongly associated and correlated to the performance of market returns and assist in the evaluation of future profits.

3.1.1.2. Case of France:

The European regulation of July 19, 2002, by requiring firms doing initial public offering to present their consolidated accounts under IFRS, mainly aims at ensuring greater transparency in accounting. Indeed, the presentation of the accounts of these companies according to harmonized standards facilitates the understanding and especially the comparison to the European level.

The adoption of IFRS will require profound computing adjustments, both in large groups and in SMEs. The change in accounting standards that all companies, will experience will transform the operating of financial markets of companies and economies on the one hand, and the preparation of accounts, on the other hand. Less publicized to the general public than the transition to the euro, the change to IAS/IFRS standards will have much impact. The main objective is to restore the clarity of the companies accounts, investors confidence in capital markets, undermined in the past five years by repeated scandals and the explosion of corporate debt.

Two new accounting principles are added:

Primacy of the economics over the legal: the accounts must give a true and fair view of the company and its assets

Relative importance: information should be included in the notes (often called notes to the financial statements or foot-notes) where it can influence the future choices of users.

The review of financial statements of listed French companies has allowed us to calculate a number of ratios and test the explanatory power of these variables. Since the introduction of IAS-IFRS aims to recover the confidence of investors and provide a firm basis for their decisions, the explanatory power of accounting information should increase and should reach the appropriate level (over 30% of dividend yield).

Table 3: Regression Results: informational content of dividends per share and performance

Model C: Yieldit = a0 + a1 divit + i + εi

Parameter estimate

t

Probability> | T |

Constant

0.04

8.94

0.000*

V1

0.01

2.41

0.016*

Adjusted R²: 4.1%;

Model D: Yieldit = a0 + a1 divit + a2 earnings/turnoverit + i + εi

Parameter estimate

t

Probability> | T |

Constant

0.04

-0.87

0.382

V1

0.01

2.43

0.015*

V8

-0.036

7.66

0.000*

Adjusted R ² = 5.6%;

Significant at: *1%, **5% and ***10% (***).i represents the specific effects.

The results in Table 3 show that the dividend per share variable explains at 4% THE dividend yield in the French context. This shows that investors do not rely in their decisions on THE distributed dividends. They focus more on the future prospects of the firm and its development projects.

The selected variables (V1 (dividends per share) and V8 (earnings/turnover)) explain the dividend yield at 5.6%. The adjusted R2 calculated is different from zero, which proves the existence of a relationship between these variables and the performance (Kam, Chan and Seow (1996), William et al. (1997), Toms (1998) and (2002), Bao (2004) and Cheng (2005)). Dividend per share (V1) is a decisive accounting variable in explaining the variation in performance. The variable V8 has a market return of explanatory content, although its coefficient is negative (Nafti & Baccouche (2007)). These results show that the assumption H1a is not verified.

3.1.1. Relationship between the turnover, assets, debts and performance:

3.1.1.1. Case of Tunisia:

The addition of the variable V10 (Gross profit/turnover) had no effect on the information content of accounting data in the context of Tunisia (The adjusted R2 is 10%, Table 4). Tunisians Investors are more interested in knowing their dividends; they are not getting enough information about the internal operation of the company.

Although the variable V3 (assets turnover) has a negative coefficient and insignificant, it has helped to increase the information content of accounting figures from 10% to 12%. Fixed assets are an important topic in the financial statements of Tunisian companies. These are required to keep pace with technology that creates a partnership relation between assets turnover and market return. This relationship is hidden by other qualitative or quantitative factors related to economic conditions that characterized the Tunisian financial market. The introduction of the variable V6 (not current liabilities / earnings) has no significant effect on the relationship of association between accounting data and stock returns.

3.1.1.1. Case of France:

First, the coefficient of the variable V10 (Gross profit/turnover) is negative and significant; it contributes to increase the information content of accounting figures from 5.6% (Table 3) to 15% (Table 5). The introduction of IAS-IFRS allows investors to get information closer to economic reality. Investors are interested in information related to the internal operations of the firm and its development.

Second, the introduction of the variable V3 (assets turnover) in the model increases the adjusted correlation R² from 15% to 25%. This growth demonstrates the importance of the V3 variable in explaining the variation in yield. It is a derived variable based on values very close to those of the market, since the assets are presented in financial statements to their true values. This result proves the superiority of historical evidence about those in decision-making. (Maghraoui & Dumontier (2006), Glaum (2000), Leuz & Verrecchia (2000), Dumontier & Raffournier (1998) and Aboody et al. (1999)).

Finally, the introduction of the variable V6 (not current liabilities / earnings) to the model increases the coefficient of correlation of 25% to 33%. This result shows that the actual level of debt may have information content and can therefore assist decision-making and understanding of the debt policy followed by the company (Boukari & Richard (2007)).

These results show that the assumption H1b is verified in the French context but not in the Tunisian one.

Table 4: Regression Results: informational content of operating income/sales, property and debts

Model E: Yieldit = a0 + a1 divit + a2 earnings/turnoverit + a3 Gross profit/turnoverit + i + εi

Parameter estimate

t

Probability> | T |

Constant

0.49

2.06

0.043

V1

-0.36

-2.07

0.042**

V8

0.66

1.4

0.167

V10

0.14

0.59

0.55

Adjusted R²: 10%;

Model F: Yieldit = a0 + a1 divit + a2 earnings/turnoverit + a3 Gross profit/turnoverit + a4 assets turnoverit + i + εi

Parameter estimate

t

Probability> | T |

Constant

0.52

2.22

0.03

V1

-0.33

-1.93

0.058**

V8

0.7

1.47

0.14

V10

0.15

0.61

0.54

V3

-0.004

-1.3

0.2

Adjusted R²: 12%;

Model G: Yieldit = a0 + a1 divit + a2 earnings/turnoverit + a3 Gross profit/turnoverit + a4 assets turnoverit + a5 not current liabilities / earningsit + i + εi

Parameter estimate

t

Probability> | T |

Constant

0.53

2.22

0.03

V1

-0.33

-1.93

0.05**

V8

0.6

1.17

0.24

V10

0.14

0.55

0.58

V3

-0.004

-1.28

0.2

V6

0.004

0.44

0.66

Adjusted R ²: 12%;

Significant at: *1%, **5% and ***10% (***).i represents the specific effects.

3.1.3. The relationship between net income and yield:

3.1.3.1. Case of Tunisia:

The results of the model K show that the introduction of other variables in the model increased their information content although some variables have no significant coefficients. However, they may have a hidden behind the other variables whose coefficients are more significant. From the results in Table 6, it can be concluded that the accounting data of Tunisian companies explain at 16% the yield. The variables whose coefficients are the most significant are dividends per share (V1), the assets turnover (V3) and equity/net fixed assets (V7).

The study of ratios, which include fixed assets (V3 and V7), allows the investor to see whether the company has new or old fixed assets and if it is possible to keep up with the high evolving technology and find the necessary funds to finance them.

O 'Connor (1973) regressed the dividend yield on several accounting variables. He found five ratios that can explain more than 32% of the market value of the firms in his sample.

The accounting data in financial statements of listed Tunisian companies, prepared in accordance with the accounting system of 1997, inspired by international standards, have low information content and their association relationship with the yield is quite low. However, we cannot ignore the existence of this relationship.

According, Ball et al. (2003) and Tendeloo & Vanstraelen (2005), the informational quality of accounting information depends on the degree of adaptation of accounting standards to the economic environment of the country. For Ashbaugh & Pincus (2001), IFRS may enhance management to publish the results, altering the information content of accounting and financial data.

Table 5: Regression Results: informational content of operating income/sales, property and debts

Model H: Yieldit = a0 + a1 divit + a2 earnings/turnoverit + a3 Gross profit/turnoverit + i + εi

Parameter estimate

t

Probability> | T |

Constant

0.06

8.4

0.000*

V1

0.006

1. 71

0.094**

V8

0.06

1

0.323

V10

-0.2

-2.49

0.016*

Adjusted R²: 15%;

Model I: Yieldit = a0 + a1 divit + a2 earnings/turnoverit + a3 Gross profit/turnoverit + a4 assets turnoverit + i + εi

Parameter estimate

t

Probability> | T |

Constant

0.078

8.32

0.000*

V1

0.006

1.9

0.062**

V8

0.07

1.13

0.262

V10

-0.27

-2.8

0.007*

V3

-0.005

-2.66

0.01***

Adjusted R²: 12%;

Model J: Yieldit = a0 + a1 divit + a2 earnings/turnoverit + a3 Gross profit/turnoverit + a4 assets turnoverit + a5 not current liabilities / earningsit + i + εi

Parameter estimate

t

Probability> | T |

Constant

0.07

8.07

0.000*

V1

0.006

1.95

0.056**

V8

0.06

1

0.3

V10

-0.17

-2.25

0.029**

V3

-0.004

-2.41

0.019**

V6

-0.004

-2.58

0.013**

Adjusted R ²: 12%;

Significant at: *1%, **5% and ***10% (***).i represents the specific effects.

Table 6: Regression Results: information content of all model variables

Model K: Yieldit = a0 + a1 V1 + a2 V2 + a3 V3 + a4 V4 + a5 V5 + a6

V6 + V7+ a8 V8+ a9 V9 + a10 V10+ i + εi

Parameter estimate

t

Probability> | T |

Constant

0.74

2.53

0.014*

V1

-0.34

-1.91

0.061**

V2

0.004

1.07

0.288

V3

-0.02

-1.6

0.11***

V4

-0.024

-0.27

0.788

V5

0.067

0.3

0.76

V6

0.0035

0.38

0.7

V7

0.0014

1.33

0.18***

V8

0.64

1.17

0.246

V9

0.06

0.36

0.7

V10

0.14

0.56

0.59

Adjusted R ²: 16,2%;

Significant at: *1%, **5% and ***10% (***).i represents the specific effects.

3.1.3.2. Case of France:

According to the results obtained by the regression model of accounting variables to dividend yield of French companies, which have adopted the IAS-IFRS, the information content of accounting variables is significant (Boukari & Richard (2007)).

The results of model L show that the coefficients of variable V8 (performance), which measures the performance of the company, and variable V9 (earnings/equity), which measures the return of the company's own resources are significant. Based on these findings the hypothesis H1c is verified in the French context and not in the Tunisian.

Table 7: Regression Results: information content of all model variables

Model L: Yieldit = a0 + a1 V1 + a2 V2 + a3 V3 + a4 V4 + a5 V5 + a6

V6 + V7+ a8 V8+ a9 V9 + a10 V10+ i + εi

Parameter estimate

t

Probability> | T |

Constant

0.08

6.2

0.000*

V1

0.004

1.54

0.1***

V2

-0.002

-0.99

0. 327

V3

-0.005

-1.66

0. 1***

V4

-0.005

-0.12

0.9

V5

-0.004

-1.24

0.22

V6

-0.001

-2.75

0.008*

V7

-0.004

-0.79

0.433

V8

-0.11

-1.85

0.07**

V9

0.04

1.78

0.07**

V10

-0.02

-0.48

0.6

Adjusted R ²: 16,2%;

Significant at: *1%, **5% and ***10% (***).i represents the specific effects.

Table 8: summary of the relationship between accounting and economic variables

French Model

Tunisian Model

V1 Dividend per share

X

X

V2 Rotate customers: turnover / customers

-

-

V3 Rotation assets: turnover / fixed assets

X

X

V4 rotation equity: turnover / equity

-

-

V5 Non-current liabilities / Equity

-

-

V6 Non-current liabilities / earnings

X

-

V7 Equity / Net fixed Assets

-

X

V8 earnings/ turnovers: Performance

X

-

V9 earnings / Equity

X

-

V10 Gross profit / turnover

-

-

Adjusted R²

30%

16%

In conclusion, the accounting variables, in the French context, established in accordance with IAS-IFRS, explain at 33% of the variation in dividend yield (Emstberger & Vogler (2008)). These results clearly show that the association relationship between dividend yield and accounting data is more readable in the financial statements of French companies (Cormier & Magnan (2005)). The accounting variables in the Tunisian context, explain the variation in performance with 16%. We can conclude that the hypothesis H2 is verified. Thus, the accounting information prepared under local GAAP have less information content than those prepared under IFRS.

3.2. Impact of IAS-IFRS on the manipulation of accounting information:

The model coefficients of total accruals (TA) are estimated using the modified Jones model. The estimation of these coefficients enabled us to calculate the non-discretionary accruals (NDA). By reducing the NDAs of TA we obtain discretionary accruals (DA).

DAit = TAit - NDAit

With:

TAit: The total accruals for firm i in year t

DAit: The discretionary accruals of firm i in year t

NDAit: The non-discretionary accruals of firm i in year t

3.2.2. The case of France: Regression results of discretionary accruals to accounting variables:

The examination of the correlation matrix (Appendix 1) allows us to detect a significant correlation relationship between variables V10 (Gross profit/turnover) and V8 (performance). This result leads us to eliminate these two variables from the model to avoid having a biased model.

Recall that the objective of our study is to examine the adequacy of accounting system change in Tunisia and France and to analyze the impact of these decisions on the financial statements information content firstly, and secondly the change of earnings manipulation under IAS-IFRS. To analyze this impact, we have tried to find an association relationship between discretionary accruals and accounting variables in the context of Tunisia and France. This reflects this discretionary component in the relationship between stock returns and accounting data in an environment using the IAS-IFRS.

Table 10: Regression results of DA to accounting variables

Model M: DAit = a0 + a1 divit + a2 turnover/customersit + a3 rotation assetsit + a4 rotation equityit + a5 non current liabilities/equityit + a6 not current liabilities / earningsit + a7 equity/net fixet assetsit + a8 earnings/equityit +  i + εi

Parameter estimate

t

Probability> | T |

Constant

0.131

3.89

0.000

V1

0.004

-0.44

0.66

V2

-0.003

-0.55

0.58

V3

0.012

1.83

0.067 **

V4

-0.014

-1.57

0.1 ***

V5

0.029

3,05 3.05

0.002 *

V6

-0.001

-1.69

0.09 *

V7

-0.044

-2.61

0.009 *

V9

-0.04

-0.46

0.64

Adjusted R²: 38%

Significant at: *1%, **5% and ***10% (***).i represents the specific effects.

The results presented in Table 10 were obtained on the basis of estimates using the random effect model and the Hausman test which has a χ2 of 31,88 with prob>χ2 equal to 0.00.

The V3 variable coefficients (assets turnover), V4 (equity turnover), V5 (non-current liabilities / equity), V6 (non-current liabilities / earnings) and V7 (equity / fixed assets) are significant and explain changes in discretionary accruals in absolute value at 38%. These variables are of balance and profitability ratios. They allow investors to have a clear idea about assets and equity turnover, the percentage of debt that the company can honor and the percentage of assets investment compared to equity. This result shows that firms that adopt IAS-IFRS manipulate upward the value of fixed assets and equity (V3, V4 and V7) and lower the value of long term debt (V5 and V6). The adjusted R2 is 38% in this case. The flexibility of the IAS-IFRS allows managers to exercise their decision to publish earnings based on discretionary targets (Watts & Zimmerman (1986, 1990)).

Hypothesis H3 is then verified. Consequently, the use of discretionary accruals is positively associated with the value of equity and capital in companies that adopt IAS-IFRS (H3a). In addition, the use of discretionary accruals is negatively associated with the value of liabilities in firms that adopt IAS-IFRS (H3b).

3.2.1. Case of Tunisia: Regression results of discretionary accruals to accounting

The examination of the correlation matrix (Appendix 2) allows us to detect a significant correlation relationship between variables V3 (assets turnover) and V7 (equity / fixed assets). This result brings us to eliminate both variables from the model to avoid having a biased model.

Table 9: Regression results of DA to accounting variables

Model M: DAit = a0 + a1 divit + a2 turnover/customersit + a3 rotation assetsit + a4 rotation equityit + a5 non current liabilities/equityit + a6 not current liabilities / earningsit + a7 equity/net fixet assetsit + a8 earnings/equityit +  i + εi

Parameter estimate

t

Probability> | T |

Constant

0.055

3.93

0.000

V1

0.004

0.75

0.453

V2

-0.0005

-0.88

0.378

V4

0.007

2.14

0.033**

V5

-0.02

-1.72

0.085*

V6

0.005

0.51

0.28

V8

-0.043

-1.08

0.28

V9

-0.022

-1.55

0.12

V10

0.022

0.77

0.44

Adjusted R2: 27%

Significant at: *1%, **5% and ***10% (***).i represents the specific effects.

Table 9 shows that the coefficients of V4 (equity turnover) and V5 (non current liabilities / Equity) are significant. This result supports the conclusion that companies that adopt the local GAAP manipulate toward the increase of their turnover and toward lowering their long-term debt. D'après les entretiens informels effectués auprès des entreprises objet de l'étude, nous avons constaté qu'elles n'ont pas de règles rigoureuses et précises concernant la mesure des provisions pour risques et charges comptabilisées, selon le système comptable tunisien, dans la rubrique : passif non courant. Based on informal interviews conducted with the businesses purpose of the study, we found that they do not have strict and specific rules regarding determining the provisions for risks and charges, classified by the Tunisian accounting system as non-current liabilities. Leaders have a flexible margin in determining the value of these reserves each year. We conclude that H4 is verified and consequently, firms that adopt GAAP and firms that adopt IAS-IFRS handle their bottom line.

Conclusion

The IASB has established international standards (IAS-IFRS) to ensure, on one side, more comparability and transparency and on the other side to ensure better quality of financial statements. The establishment of such standards by the IASB aims at achieving harmonization of accounting practices between countries.

Some studies have shown that differences between national accounting systems affect the importance and credibility of accounting information: Hung (2001), Ali & Hwang (2000), Ball et al. (2000), Pope & Walker (1999), Joos & Lang (1994) and Alford et al. (1993).

This flexibility allows managers to exercise their decision to publish earnings based on discretionary targets (Watts & Zimmerman (1986, 1990)). They seek their intervention to strengthen the information content of accounting figures in order to better inform the market about the prospects of the firm.

In this article we are interested in the relationship that may exist between the IAS-IFRS, which promotes a present value, and the manipulation of earnings enhanced by the application of these standards. We are expanding, to the French and Tunisian context, the problem of information content of the discretionary component of accounting income in connection with the application of IAS-IFRS.

On the basis of a sample of 120 firms-year observations from Tunisia, which uses national standards based on those international, and a sample of 120 firms-year observations from France, which adopts the IAS-IFRS since 2004, we studied the relationship of association between accounting variables and dividend yield on one side and the impact of the use of IAS-IFRS on discretionary accruals in this relationship on the other side. In other words, we tried to analyze the advantages and disadvantages of the adoption of IAS-IFRS. The study results show that the application of IAS-IFRS accounting standards in France increases the information content of accounting numbers. The accounting variables explain at 33% the dividend yield. However, the adoption of IAS-IFRS allows managers to exercise their discretion to adjust the financial statements based on circumstances and with certain requirements. They believe that their intervention in the management of earnings increases the credibility of accounting information. This action increases as the level of discretionary accruals in the firm. Among the ten accounting variables selected in the model, five variables are manipulated and undergo a change.

This paper has certain limits, which deal with the samples size (120 firm's observation in each context). We can study the impact of the adoption of the IAS-IFRS by the Tunisian firms since 2014 is the starting date for adapting IAS-IFRS by Tunisian companies (We already studied the impact of the adoption of the IAS-IFRS by the French firms).

Appendix 1: Correlation matrix of variables: Pearson correlation

V1

V2

V3

V4

V5

V6

V7

V8

V9

V10

V1

1

V2

-0015

1

V3

-0020

0.5282

1

V4

0082

0.6068

0.55

1

V5

0058

-0.05

-0.22

0.22

1

V6

0.092

-0.027

-0.05

.12

-0.08

1

V7

-0095

0.0006

0.5

-0.23

-0.5

-0.19

1

V8

-0.195

-0.041

-0.1

-0.33

-0.29

0.08

0.09

1

V9

-0.065

0.533

0.49

0.37

-0.15

0.1

0.04

0.54

1

V10

-0.16

-0.08

-0.2

-0.3

0.14

0.15

-0.1

0.7

0.31

1

Appendix 2: Correlation matrix of variables: Pearson correlation

V1

V2

V3

V4

V5

V6

V7

V8

V9

V10

V1

1

V2

-0.015

1

V3

0.031

-0.028

1

V4

-0.065

-0.029

-0.1

1

V5

-0.045

-0.039

-0.076

0.74

1

V6

-0.003

-0.02

-0.053

-0.076

-0.035

1

V7

0.0347

-0.022

0.94

-0.11

-0.061

-0.047

1

V8

0.26

0.01

0.63

-0.26

-0.3

0.17

0.56

1

V9

0.12

0.013

0.066

-0.59

-0.09

0.075

0.028

0.32

1

V10

0.04

-0.026

0.45

-0.31

-0.21

0.1

0.41

0.58

0.22

1