The effects of IFRS adoption on earnings management : evidence from the Dutch public sector

1

MSc Accountancy & Control, variant Accountancy Faculty of Economic and Business, University of Amsterdam

Master Thesis:

The effects of IFRS adoption on

Earnings Management: Evidence

from the Dutch Public sector

Final Version

Name: Aisseta Harlequin Student number: 10603379

Date: 14th of August 2014 Supervisor: dr. A. Sikalidis

2 Contents Abstract ... 3 1. Introduction ... 4 1.1 Background theory ... 4 1.2 Research question ... 5

1.3 Motivation and contribution... 5

1.4 Main Analysis and Findings ... 6

1.5 Thesis outline ... 7

2. Literature review and hypothesis ... 7

2.1 IFRS ... 7

2.1.1 IFRS and Public Firms ... 8

2.2 Accounting quality ... 9

2.3 Earnings management , Discretionary accruals & hypothesis development ... 10

2.4 Real activities manipulation & hypothesis development ... 12

3. Methodology ... 14

3.1 Research method of accounting quality , discretionary accruals, real activities manipulation ... 14

Real activities manipulation ... 15

3.2 Empirical model & Variables ... 16

3.2.1 Variables ... 18

3.2.2 Empirical Model ... 23

3.3 data collection ... 25

4. Results ... 27

4.1 Descriptive statistics ... 27

4.2 Test of Baseline Predictions ... 31

5. Conclusion and limitation ... 36

Refrences... 38

3

Abstract

This thesis investigates the effect of IFRS on accounting quality. In particular I examined the effect of IFRS adoption on real earnings management and the use of discretionary accruals in earnings management for Dutch public firms. In previous research Discretionary accruals and real earnings manipulation are used as common proxies for accounting quality. Since 2005 public firms in Europe where obligated to adopt these international accounting standards (IFRS). Prior literature shows some contradiction in the adoption of these international accounting standards. Some studies show that the adoption of the international standard IFRS improves accounting quality but other studies also show that chancing certain accounting standards can cause debatable accounting choices from managers which means that accounting quality does not particularly improves. In order to measure the effect of IFRS on these specific proxies for accounting quality, I conduct an empirical investigations of the effects of IFRS adoption on the use of discretionary accruals in earnings management and real earnings management.

In this paper the research question is as follow “What is the effect of IFRS adoption on real activities manipulation and the use of discretionary accruals in Earnings management for Dutch Public firms”. I conducted an archival research in which I collected available data from Compustat between the period of 2002 till 2011. Due to some financial data missing I performed the regression for 343 Dutch companies. The results of the regression analysis show no effect of IFRS on the use of discretionary accruals. The regression result shows an insignificant coefficient. Thus there is no relation between IFRS and discretionary accruals in earnings management. But the results do show an effect of IFRS on real activities manipulation. The regression results show a significant coefficient. Thus IFRS and real activities manipulation are negatively associated.

Keywords: IFRS, Accounting quality, Earnings management, Discretionary accruals, Real activities manipulation, Dutch public firms, Netherland

4

1. Introduction

In this chapter I will describe the research topic and the purpose of my thesis. The chapter will consist of background information, the research question and contribution. There will also be an outline provided of the thesis and the main findings will be discussed.

1.1 Background theory

With the globalization of international financial markets, the idea of adopting common rules for financial reporting to develop international comparability has become widespread. More than 100 countries have agreed to require or allow adoption of IFRS, or have established timelines for the adoption of IFRS. In recent years, Brazil, Canada, China, and India have all committed to formal timelines for adoption of IFRS (Stolowy, 2008). In 2002, the European Union adopted IFRS as the required financial reporting standards for the consolidated financial statements of all European companies whose debt or equity securities trade in a regulated market in Europe, effective in 2005. Meaning that in the period of 2005 the listed companies in the Netherland were mandatory to implement the IFRS standards for the preparation and presentation of Financial statements (IFRS, 2014). The International Accounting Standards Board (IASB) is the standard setter of these accounting standard. The standards were developed with the aim of eliminating differences in accounting standards across countries and to improve the information in financial statements which would lead to an increased reliance of financial reporting, a reduction in barriers for investors and better performance on the capital markets. But the main goal of the IASB is to improve the accounting quality with the development of an internationally acceptable set of high quality financial reporting standards (Barth 2007). There has been much research done on accounting quality being reduced due to earnings manipulations by managers. An explanation for the relationship between managers discretionary reporting decisions and firms' current (pre-managed) earnings levels is based on the assertion that managers manipulate earnings when their compensation depends on the accounting earnings (Healy, 1985; Holthausen, Lacker,Sloan., 1995).

According to Burgstahler, Hail, Leuz (2007) a number of countries have adopted IFRS in an attempt to improve accounting quality. Accounting quality depends heavily on judgments and estimates. Earnings management is highly susceptible to judgment, because of the incentives or motives of managers to present the figures in a specific way.

5

The research of Cai, Rahman & Courtenay (2008) and Barth, Landsman & Lang (2008) shows that earnings management in countries that adopted IFRS has been decreased in the years after adoption. Therefore, IFRS contributed to an improvement of accounting quality. However other research also shows that changing accounting standards is not sufficient to decrease the level of earnings management, because high quality standards alone, harmonization and changes in accounting standards are not sufficient to improve accounting quality (Tendeloo & Vanstraelen 2005; Ding, Hope, Jeanhean,Stolowy 2007; Leuz et al 2007). Therefore it is necessary to research whether the adoption of IFRS increases accounting quality, measured by the variety of earnings management.

1.2 Research question

This research will meanly focus on the relationship between discretionally accruals, earnings management and real activities manipulation. I will examine how the IFRS adoption in the Netherland has affected the use of discretionally accruals in earnings management and real activities manipulation. This examination will be based on Dutch Public companies specifically on how the consequences of IFRS adoption has affected the accounting quality. On the basis of the aforementioned, the research question is as follows:

What is the effect of IFRS adoption on real activities manipulation and the use of discretionary accruals in Earnings management for Dutch Public firms?

1.3 Motivation and contribution

Prior literature has mainly focused on the widespread adoption of IFRS that resulted in a fundamental change in the business environment, prior to 2005, companies followed a variety of country-specific Generally Accepted Accounting Principles (Soderstrom & Sun 2007). The International Accounting Standards Board (IASB) is the standard setter for the International Financial Accounting Standards. The general goal of financial reporting is to provide financial information about the reporting entity that is useful to capital providers. The IFRS adoption is based on trying to ensure a high degree of transparency and comparability to financial statements and also to improve the efficient functioning of the Capital market (EC, 2002).

6

This research will contribute to prior literature by specifically examining the relationship between accounting quality and the use of discretionary accruals in earnings management and real activities manipulation. Currently, there is no available research on the effects of the IFRS adoption on the use of discretionary accruals, real activities manipulation for public companies in the Netherlands. According to prior literature firms that apply IFRS engage in less earnings management. They also find an improvement in accounting quality (Barth et al, 2008). However as I mentioned before it is also shown that changing accounting standards are not sufficient to decrease the level of earnings management behavior, this is because high quality standards alone are not sufficient in improving accounting quality. Therefore this research will provide insight in the use of discretionary accruals , real accounting activities in relation with earnings management. And whether IFRS has increased or decreased the association between the use of discretionally accruals in earnings management and real activities manipulation for public firms in the Netherlands. In conclusion this research will provide insight in how the consequences of IFRS adoption has affected the accounting quality.

1.4 Main Analysis and Findings

The regression were performed on a relatively small amount of companies. The initial sample consisted of 1362 firms. Due to removal of missing data from the statement of financial position and profit and loss the remaining final sample was 343 during the period 2002-2011. Descriptive statistics show an increase in firms that adopted IFRS after the mandatory IFRS adoption in 2005. However descriptive statistics also show that overall a bigger percentage did not adopted IFRS of the total sample after the mandatory period . In accordance with previous research, which shows that IFRS does not have to increase accounting quality or that a switch to higher standards does not have to mean an improvement in accounting quality. I find that for the regression analyses for the first hypothesis “ the adoption of IFRS effected the use of discretionary accruals in earnings management for Dutch companies” , IFRS does not affect the use of discretionary accruals. The results of the second hypothesis which is “ the adoption of IFRS effected the use of real activities manipulation in Dutch public companies” ,are also in accordance with prior literature, which suggest that IFRS adoption does improve accounting quality and leads to less earnings management. I find that IFRS does effect real activities manipulation. The regression results show a significant association between IFRS and real earnings manipulation.

7

1.5 Thesis outline

The remaining part of the thesis will be structured as follows. In Chapter 2, the literature review, I will examine available literature regarding the theories of accounting quality, IFRS and earnings management. Furthermore, the hypothesis will be developed and explained. In Chapter 3, I will discuss the research method and the model that is used for the research. In addition the sample selection, variables and empirical model will be explained. This Chapter is followed by Chapter 4 which consist of the empirical results and main analysis. Chapter 5 will provide the conclusion and limitation of the research.

2. Literature review and hypothesis

This chapter will provide a literature review of the various topics: IFRS standards, accounting quality and earnings management. In this chapter the theoretical concepts will be explained based on the findings of prior literature and in addition the hypothesis will be developed.

2.1 IFRS

In 1973 International Accounting Standards Committee (IASC) was established. The IASC published international rules which were named International Accounting Standards (IAS) in that period. In 2001 the IASC was taken over by a newly – reconstituted IASB and they describe these standards under a new name International Financial Reporting Standards (IFRS). The IASB is a private sector that aimed to develop an internationally acceptable set of high quality financial reporting standards. The adoption of these International reporting standards in Europe was to achieve a convergence in reporting between Europe and the rest of the world. IFRS are accounting standards that should be applied equally to financial reporting by public companies worldwide (Ball 2013). In 2005 the IFRS became mandatory for all listed companies in all 15 members (at that time) of the European Union. The IFRS is implemented with the idea to reduce information asymmetry and information risk and to increase the relevance and reliability of financial reporting (www.iasplus.com). Accounting information would then be more transparent and the comparability would increase.

8

Furthermore IFRS reduces the investment risks and cost of capital worldwide because investors are able to take better decisions, it lowers the costs arising from multiple reporting methods, it encourages international investments and it improves the allocation of savings worldwide (Street, Gray & Bryant. 1999). IFRS represented an impact in financial reporting for European firms because many requirements in IFRS are different from those local general accepted accounting principles (GAAP) standards of European countries. These International standards are endorsed by the European Commission for which the primary requirements are relevance, reliability and comparability.

2.1.1 IFRS and Public Firms

This research will focus on the impact of financial reporting for public companies in the Netherlands. There has been much prior literature about the impact of financial reporting in public companies. In 2005 after the adoption of IFRS Listed companies were required to use these International standards in their Financial Statement. The change in the local GAAP to IFRS led to improvement the information that was provided. According to prior literature of Barth et al (2006) listed companies that adopted IFRS have less earnings management , more timely loss recognition and more value relevance of earnings which all together means higher accounting quality. Burgstahler et al (2006) also shows that if the companies use new disclosures and accrual measures this leads to significantly lower earnings management. Cai et al (2008) also concluded that earnings management did actually decline in the year after the adoption of IFRS and that IFRS improved accounting quality.

On the other hand, other studies also show evidence that IFRS does not improve accounting quality. The change in accounting standards does not decrease earnings management. For example Burgstahler (2006) also shows that change in the standards could play a role but there should be a proper reporting incentives and legal enforcement. The research of Ding et al (2007) and Leuz et al (2003) showed that changing accounting standards is not sufficient to decrease the level of earnings management and improve the accounting quality. The harmonization of and changes in accounting standards are not enough to influence earnings management. In addition Goncharov and Zimmerman (2006) also showed that adoption of IAS ( the standard of IASB before IFRS) does not make a change compared to the situation before IAS.

9

2.2 Accounting quality

Accounting quality can be defined as the extent to which the financial statement information reflects the underlying economic situation (Chen et al, 2010). There is no precise definition of accounting quality. Most of prior literature argue that the measures are related to faithful representation of the underlying economics which is broadly accepted by standard-setters, regulators, practitioners, and users, as well as by academics, as an important feature of high-quality accounting (FASB,2008; SEC 2000; Healy and Wahlen, 1999; Schipper and Vincent ,2003; Ball , 2006 and Dechow, Ge, & Schrand 2010). Ahmed , Neel and Wang 2013 argue that accounting choices that result in greater management of earnings to meet a target, and overstatement of earnings (or delayed recognition of losses) leads to compromising faithful representation of the underlying economics and therefore reducing accounting quality.

Measures for Accounting Quality

Quality of accounting can be measured by earnings management, timely loss recognition and value relevance, which are considered the most commonly measures to determine accounting quality (Ahmed, Neel & Wang 2013; Barth et al, 2007; Chen et al, 2010). In this research accounting quality will be measured with the variety of earnings management methods. In particular the use of discretionary accruals in earnings management and real manipulation of accounting will be examined.

Accounting quality and IFRS

The quality of the accounting standards determine accounting quality. Soderstorm et al (2007) argues that if the IASB continues to improve the quality of IFRS, there could be expected financial reporting under IFRS to become increasingly value relevant and reliable. Barth et al. (2008) provides different reasons why the adoption of IFRS may improve accounting quality. First IFRS eliminates certain accounting alternatives by reducing managerial discretion. This means that the extent of opportunistic earnings management can be reduced and improve accounting quality. Also IFRS is viewed as principles-based standards and are therefore potentially more difficult to avoid. For example, under a principles-based standard it should be more problematic to not recognize a liability through transaction structuring. Furthermore IFRS allows measurements, such as use of fair value accounting, that may better reflect the underlying economics than domestic standards.

10

On the other hand Barth et al. (2008) also argues that adopting IFRS may also reduce accounting quality. For example IFRS could eliminate accounting alternatives that are most appropriate for communicating the underlying economics of a business which could lead to forcing managers of these firms to use less appropriate alternatives, resulting in a reduction in accounting quality. Also because IFRS are principles-based, they do not provide detailed implementation guidance which creates greater flexibility for managers to exploit accounting discretion to their advantage.

2.3 Earnings management , Discretionary accruals & hypothesis development

Earnings management

Healy and Wahlen (1999) argue that earnings management occurs when managers use judgment in financial reporting and in structuring transactions to change financial reports to either mislead some stakeholders about the underlying economic performance of the company or to influence contractual outcomes that depend on reported accounting numbers. According to them there are five ways in which management is able to use their judgment. These are: to estimate future economic events, to choose accounting methods, to manage working capital, to make or defer expenditures and to structure corporate transactions. Management can use these judgments for two purposes. The first is to use judgment to make financial reports more informative for users. The second is to use judgment to make financial reports less informative for users or to mislead stakeholders about the underlying economic performance of the firm. So therefore Healy and Wahlen (1999) argue that earnings management is seen through the second perspective.

Gunny (2005) and Nelson, Elliot & Tarpley (2002) argue that earnings management is a purposeful intervention in the external financial reporting process, with the intention of obtaining some private gain. Burgstahler et al (2006)also argues that earnings management happens, because management has the possibility and ability to do so. They can manage earnings by using their judgment in making decisions. Earnings management is seen as an important dimension of accounting quality. Accounting standards provide substantial flexibility to firms because measurements are based on private information and the application of standards involves judgment from management.

11 Discretionary accruals

Healey and Palepu (1993) define discretionary accruals as more subjective. These authors argue that discretionary accruals reflect higher degrees of managerial judgment. Discretionary accruals often contain two elements. The first element is noise caused by managers aggressive and opportunistic reporting and the second element is based on an information component that allows managers to communicate their private information. Subramanyam (1996) argues that discretionary accruals are associated with several performance measures, and concludes that managers' accrual choices increase the in formativeness of accounting earnings. He finds that discretionary accruals are associated with contemporaneous stock prices and future earnings and cash flows in which there is a lot of managerial judgment involved. Bernard and skinner (1996) look at two respect regarding the use of discretionary accruals. First financial statement users are interested in how discretionary accruals should be interpreted. Meaning in what settings do these numbers increase and decrease the in formativeness of reported earnings. The Second point of view is standard setters and how they tend to act to reduce managers' ability to exercise discretion in the reporting process if managers exercise their accounting discretion opportunistically.

Hypothesis development

Based on the above mentioned theories, earnings management means manipulating reported earnings so that they do not accurately represent economic earnings at every point in time. For example managers could be showing a stable growth to stakeholders or risk adverse managers with bonus contract would try to prevent showing large fluctuations. The net income will therefore show a stable line. Numerous studies have used discretionary accruals as a method to measure earnings management. According to Subramanyam (1996) discretionary accruals are associated with contemporaneous stock prices and future earnings and cash flows and concludes that managers choose accruals to enhance the in formativeness of accountings. Meaning that managers can use this information opportunistically that can lead to reducing the faithful representation of the underlying economics.

As mentioned before according to prior literature earnings management has decreased over the years in countries that implemented IFRS Standards. Because IFRS provides higher quality standards and also eliminates certain accounting alternatives there by reducing managerial discretion.

12

So IFRS contributed to an improvement of accounting quality. However eliminating certain accounting standards can cause controversial accounting choice for managers to use less appropriate standards. Furthermore using higher accounting standards does not have to mean that accounting quality will be improved. In addition Both studies of Ball & Shivakumar (2005) and Burghstahler et al. (2006) examined public and private firms within the same country. These studies provide evidence that earnings quality is lower among private than public firms despite these firms facing the same regulations on auditing and accounting standards. These results are consistent with incentives and market forces dominating accounting standards as the main determinants of accounting quality. So therefore based on these above contradiction in prior literature the following hypothesis is formulated:

Hypothesis 1

The adoption of IFRS affected the use of discretionary accruals in earnings management for Dutch Public companies.

2.4 Real activities manipulation & hypothesis development

Earnings management can be classified into two categories. The first category is accruals management and the second is real activities manipulation. Real activities manipulation occurs when managers undertake actions that change the timing or structuring of an operation, investment, and ⁄ or financing transaction in an effort to influence the output of the accounting system (Gunny 2010). Roychowdhury (2006) defines real activities manipulation as management actions that deviate from normal business practices, undertaken with the primary objective of meeting certain earnings thresholds. He also argues that real activities manipulation as departures from normal operational practices are motivated by managers desire to mislead at least some stakeholders. Manager want to mislead these stakeholders into believing certain financial reporting goals have been met in the normal course of operations. Other prior literature as for example Healy and Wahlen (1999), Fudenberg and Tirole (1995),and Dechow and Skinner (2000) argue that acceleration of sales, alterations in shipment schedules, and delaying of research and development (R&D) and maintenance expenditures can be considered earnings management methods available to managers. Overall can be concluded that Managers have incentives to manipulate real activities during the year to meet certain earnings targets.

13

So therefore Real activities manipulation affects cash flows and in some cases, accruals. Meaning that the accounting quality will be effected and the faithful presentation of the underlying economics will be reduced.

Hypothesis development

Based on the above mentioned theories real activities manipulation can reduce firm value because actions taken in the current period to increase earnings can have a negative effect on cash flows in future periods. For example, aggressive price discounts to increase sales volumes and meet some short-term earnings target can lead customers to expect such discounts in future periods as well. As mentioned before real earnings management is motivated by managers’ desire to mislead at least some stakeholders into believing certain financial reporting goals have been met in the normal course of operations.

Overall misleading financial reporting goals, greater management of earnings to meet a target, and overstatement of earnings (or delayed recognition of losses) leads to compromising faithful representation of the underlying economics and therefore reducing accounting quality. The paper of Barth, Landsman and Lang (2008) concludes that indeed firms applying IFRS exhibit more value relevant accounting figures than other companies. IFRS leads to more reliable information because IFRS goal is to reduce managerial discretion. However other researches also shows that changing accounting standards more particular high quality standards alone, harmonization and changes in accounting standards are not sufficient in improving accounting quality ( tendeloo, 2005, ding 2007& leuz 2007). Based on the contradictions in the above theories the following hypothesis can be formulated:

hypothesis 2a

The adoption of IFRS affected the use of real activities manipulation in Dutch public companies.

14

3. Methodology

In this chapter the research method of the thesis will be discussed. This research will examine how IFRS adoption in the Netherland has affected the use of discretionally accruals in earnings management and real activities manipulation for public firms. In chapter 3.1 the measurement for the use of discretionary accruals and the measurement for real earnings manipulation will be provided. Furthermore this chapter will consist of 3.2 the empirical model used in this study and in addition 3.3 the data collection used for the research.

3.1 Research method of accounting quality , discretionary accruals, real activities manipulation

As mentioned in the introduction, This examination will be based on how the consequences of IFRS adoption has affected the accounting quality for Dutch public companies. The research will mainly focus on how the IFRS adoption in the Netherland has affected the use of discretionally accruals in earnings management and real activities manipulation.

the research question is as follow ; What is the effect of IFRS adoption on real activities manipulation and the use of discretionary accruals in Earnings management for Dutch Public firms?

The research evidence is an empirical archival examination. Data of Dutch listed companies will be included for the years 2002 till 2011. This period is chosen because it provides available information before and after the IFRS adoption period in order to examine the effect of the implementation of these mandatory international standards on accounting quality. The measurement of the theoretical concepts is done by utilizing previous research methods and models. These are described below.

Accounting quality

To examine whether the introduction of IFRS effected the quality of accounting for Dutch public firms in the Netherlands. I will particularly focus on the use of discretionary accruals and real activities manipulation as proxies for earnings management. One of the most common measurement of accounting quality is earnings management. Most of prior literature used income smoothing or earnings management as measurement for accounting quality.

15

For instance Ahmed, Neel & Wang ( 2013) and Barth et al 2008 included earnings management as a measurement for accounting quality specifically they used meeting or beating a benchmark as a proxy for earnings management. The research model used for the examination in my thesis is based on the research methods used in prior research Kim et al (2012). In this research the use of discretionary accruals and real activities manipulation are used as surrogates for earnings management.

Discretionary accruals

The use of discretionary accrual will be used as a surrogate for earnings management. I will specifically use the discretionary model in Kim, Park & Wier (2012). The model used in the Kim research method is based on a cross sectional version of the jones modified model. The Jones model correctly decomposes accruals into its discretionary and nondiscretionary components. In prior literature the Jones model and its variations have been used extensively to detect earnings management. Prior literature Dechow et al (1995) clarifies the importance of a careful consideration of the hypotheses being tested, because firms with extreme performance are also likely to engage in earnings management. According to Dechow et al (1995) and Kothari et al (2005) discretionary accrual models are used to correctly test such manipulation. The result from Kothari analysis show that discretionary accruals estimated with using the Jones or the modified-Jones model, and adjusted for a performance-matched firm’s discretionary accrual , tend to be the best specified measures of discretionary accruals across a wide variety of simulated event conditions. The reason the cross sectional model is used in the research model is because it involves superior specification and less restrictive requirements. Furthermore I will specifically focus on the absolute value of discretionary accruals (ABS_DA) for the main analyses as used in the Kim paper because earnings management can involve either income-increasing or income-decreasing accruals.

Real activities manipulation

Real activities manipulation (Ram) or Real earnings management (Rem) is defined as management actions that deviate from normal Business. Prior literature considers Rem as a measurement for earnings management. As mentioned before real earnings management will be used as a proxy for earnings management. Prior literature Cohen et al (2008) shows evidence that managers take real economic actions to maintain accounting appearances.

16

In particular they would decrease discretionary spending on R&D, advertising, and maintenance to meet an earnings target. Royschowdgury (2006) used empirical models to detect real earnings management by examining cash flow from operation, production cost and discretionary expenses. My research method will be based on Kim et al 2012 research model which used the following four measures to detect real activities manipulation: (1) abnormal levels of operating cash flows (AB_CFO), (2) abnormal production costs (AB_PROD), (3) abnormal discretionary expenses (AB_EXP), and (4) a combined measure of real activities manipulation. Kim et al 2012 model is based on Cohen et al. (2008) paper which uses three individual proxies as well as a combined proxy (COMBINED_RAM) in order to detect management action that deviate from normal business.

3.2 Empirical model & Variables

The key in this research is to capture the relation between the adoption of IFRS and on the other hand the use of discretionary accruals and real activities manipulation as proxies for earnings management for public firms in the Netherland. As stated earlier the research model that is used for the examination is based on the research model used in Kim et al 2012. I will specifically examine the effect of IFRS adoption on the use of discretionary accrual and real earnings manipulation. Below in Paragraph 3.2.1 & 3.2.2 a more detailed explanation will be given about the variables and the two models. furthermore a detailed explanation will be provided for the regressions that will be done for assessing the performance of both dependent variable ABS_DA and RAM_PROXY.

So therefore to answer the research question, the empirical regression model for the first hypothesis on whether IFRS adoption has affected the use of discretionary accruals in earnings management. the model is as follow;

ABS_DAt (or DAt)=

α

α

α

α

α

ε

17 Dependent variable:

ABS_DA= Absolute value of discretionary accruals (signed discretionary accruals)

Independent variable

- Dummy variable IFRS= 1 If it is in the Post IFRS (after adoption) and 0 if it is in the Pre IFRS period

In addition control variables:

- COMBINED_RAM= AB_CFO - AB_PROD + AB_EXP;

- AB_CFO= The level of abnormal cash flows from operations, - AB_PROD= The level of abnormal production costs,

- AB_EXP=The level of abnormal discretionary expenses, - SIZE= Natural logarithm of Total Asset

- TA = Total Asset

- ROA= Return On Asset

- LEV = Leverage

- RD_INT = R&D intensity

The empirical regression model for the second hypothesis on whether IFRS adoption has effected real earnings manipulation. The model is as follow;

RAM_PROXY =

α

α

α

α

α

ε

Dependent variable:

RAM_PROXY= AB_CFO, AB_PROD, AB_EXP, or COMBINED_RAM

- COMBINED_RAM= AB_CFO - AB_PROD + AB_EXP;

- AB_CFO= The level of abnormal cash flows from operations, - AB_PROD= The level of abnormal production,

18 Independent variable

- Dummy variable IFRS= 1 If it is in the Post IFRS (after adoption) and 0 if it is in the Pre IFRS period

In addition control variables;

- ABS_DA= Absolute value of discretionary accruals (signed discretionary accruals) - SIZE= Natural logarithm of Total asset

- TA= Total Asset

- ROA= Return On Asset

- LEV = Leverage

- RD_INT = R&D intensity (R&D expense/net sales) for the year;

3.2.1 Variables

The dependent variables are the absolute value of discretionary accruals (ABS_DA) and the Real Activities Manipulation proxy (RAM_proxy). The ABS_DA measures the response to earnings or whether earnings are being manipulated through the use of discretionary accrual. The RAM_proxy also measures the response of earnings through real business activities performed by the management.

The dummy variable used to measure the effect of IFRS adoption on the use of discretionary accruals in earnings management and real activities manipulation is the independent variable. In this research I use six control variables for the first model. Which consist of the RAM_PROXY (or Combined_RAM,), the firm Size, Total assets (TA), the return on assets (ROA), LEVERAGE of the firm and R&D intensity. For the second model I also include six control variables. Which consist of ABS_DA, Total Asset, firm Size, ROA, LEVERAGE and R&D intensity. The variable definitions are listed in panel A of Table 1.

19 TABLE 1

Panel A: Variable definitions

Variable Definition

Dependent Variables ABS_DA Variable which shows the absolute discretionary accruals computed using the modified Jones model (examining the residual of the Total assets regression formula). RAM_PROXY Variable which shows the earnings that are

managed through real activities. Calculated by the abnormal cash flows – abnormal production costs + abnormal expenses.

Independent Variables IFRS DUMMY Dummy which has a value of 1 if the entity employs IFRS or has a value of zero when the it did not adopt the IFRS.

Control Variables ABS_DA Variable which shows the absolute discretionary accruals computed using the modified Jones model (examining the residual of the Total assets regression formula). RAM_PROXY

SIZE

Variable which shows the earnings that are managed through real activities. Calculated by the abnormal cash flows – abnormal production costs + abnormal expenses.

An control variable which indicates the firms size. Calculated by the natural logarithm of total assets.

TOTAL ASSETS An indicator of the total assets of the entity for the years in the period.

ROA Shows the return on assets. Calculated as the net income of year x divided by total assets in

20

year x.

LEVERAGE An indicator that shows the debt scaled by total assets for the year.

RD_INT An indicator that reflects the R&D intensity. Calculated by the R&D expenses scaled by Net sales for the year.

Additional explanations AB_CFO Shows the level of abnormal cash flows from

operations. Calculated through the residual to estimate the normal level of operating cash flows following the Roychowdhury’s, 2006 model.

AB_PROD An indicator of the abnormal production costs from operations. Calculated through the residual to estimate the normal productions costs following the Kim et al., 2012 model.

AB_EXP Shows the abnormal discretionary

expenditures. Calculated through the residual to estimate the normal level of discretionary expenses following the Kim et al., 2012 model.

Note: this table shows the variable definitions used in the continuation of the paper.

Based on prior literature Kothari 2005 the absolute value of discretionary accruals is estimated by the use of residuals from the annual cross- sectional industry regression model. The regression model that I used is based on the Kim et al (2012) model which used the modified jones model by including Return On Asset (ROA here after). The reason for including ROA is to avoid potential misspecification in order to increase the reliability of discretionary accrual estimates. Following Hirshleifer, Hou & Teoh ( 2009) analysis (operating) accruals can be divided in normal accruals which reflects business conditions and discretionary accruals which most likely characterize managers manipulating earnings.

21

So therefore to determine the accruals which were used for manipulating earnings or deviate from the normal accrual as mentioned earlier I followed the Kim et al (2012) regression model. The residuals from the regression of the following model determine the absolute value of discretionary accrual; TAit/Ait-1=α0(1/Ait-1)+α1(⌂REVit-⌂RECit)/Ait-1+α2 PPEit/Ait-1 + α3

IBXIit-1/ Ait-1+εit. The model consist of the depend variable, Total accrual for a firm i at year t (TA) divided by the lagged total assets (Ait-1). And the independent variable consist of change

in net revenues ( ⌂REVit) , change in net receivables (⌂RECit) , gross property, plant and

equipment (PPEit), income before extraordinary items (IBXIit-1) and lagged total assets( Ait-1).

The real activities manipulation can be determined by the Abnormal Cash flow from Operations (hereafter AB_CFO) or the Abnormal Production cost ( hereafter AB_PROD) or the Abnormal Discretionary Expenses (hereafter AB_EXP). For the dependent variable Real activities Manipulation ( hereafter RAM_PROXY) I also followed the Kim et al paper (2012). Kim (2012) calculated Combined measures (COMBINED_RAM) from the three individual real activities manipulation AB_CFO , AB_PROD and AB_EXP. The RAM_PROXY is calculated by the following formula (AB_CFO - AB_PROD + AB_EXP). The AB_CFO is based on sales manipulation which are expected to lead to a lower current period operating cash flow. These sales manipulation are the residual of the regression from the following model; CFOt/At-1=α0+α1(1/At-1)+β1(St/At-1)+β2(⌂St/At-1)+εt. The model is used to estimate the normal level of operation cash flow which consist of the dependent variable cash flow from operation divided by total lagged assets and the independent variables total assets and net sales.

The other measure of real activity manipulation AB_PROD is the residual of the regression from the following model; PRODt/At-1=α0+α1(1/At-1)+β1(St/At-1)+β3(⌂St-1/At-1)+εt. The model is used to estimate the normal production cost from. According to prior literature Roychowdhury (2006) and Cohen (2008) define production cost as the sum of cost of goods sold (cogs) and the change in inventory during the year. The model consist of dependent variables these production cost divided by total lagged asset and the independent variables consist of total assets and net sales.

The third individual measure for real activities manipulation AB_EXP is the residual from the following model; DISEXPt/At-1=α0+α1(1/At-1)+β(St-1/At-1)+εt. The model is used to estimate the normal level of discretionary expenses. These expenses are defined as the sum of R&D, advertising and SG&A expenses. In my research these discretionary expenses will only consist of R&D and SG&A expenses due to insufficient data for advertising expenses.

22

In conclusion the dependent variable for the second model RAM_proxy of my research is these three individual real activities manipulation combined. These combined measures will determine real earnings management as a proxy for earnings management.

The independent variable IFRS (dummy) for both models shows the public companies in Netherland that adopted IFRS in the period between 2002 and 2011. The reason for this (as mentioned before ) is to see whether public firms that adopted IFRS have affected the use of discretionary accrual in earnings management and real earnings management. Therefor I can determine whether firms that adopted IFRS has influenced managers manipulating earnings.

Following Kim et al (2012) I included RAM_PROXY as a control variable for the regression of the First hypothesis model ABS_DA. The proxy ABS_DA is used as a control for the regression of the second hypothesis model RAM_PROXY. The reason these proxies were also included as control variables is to control for the substitutive nature of these two earnings management measures. Prior research Zang( 2011) and Cohen, Dey, Lys (2008) states that firms use a mix of discretionary accrual and real activities manipulation as tool for managers to manipulate earnings. So therefore firms choose a measure that is less costly to them.

The variable Total assets show total value of assets for the companies, during the period 2002-2011. The other control variable SIZE shows the firms size calculated by an the natural logarithm of total assets. These variables are used as control variable, because the firm’s assets and size could potentially explain significant variation in earnings management. An firm size or assets value could affect the financial reporting behavior. Therefore assets and firm size could be associated with the managers wanting to mislead financial reporting goals in order to meet a target, and overstatement of earnings.

The variable Return on assets represents, the net income divided by the amount of total assets. The ROA is viewed as an indicator for firm performance. A variable that may influence investors actions or beliefs when deciding to invest or not to invest in a potential firm. Therefore, this variable may also be interestingly associated with variation in earnings management and the firms reporting financial goals.

23

Leverage is calculated by dividing debt with total scaled assets . According to prior literature Kim et al (2012) leverage is viewed as an incentives for earnings management. Leverage is also seen as an indicator to cause variation in earnings management.

In conclusion I included total assets, ROA and leverage to control for variations in earnings management because the empirical proxies for earnings management might include measurement errors that are correlated with these firm characteristics.

The variable RD_INTS is used to control a firms R&D expenses. This variable is Calculated by the R&D expenses scaled by Net sales for the year. The control variable should give the precise R&D intensity of a firm. These independent and control variables do not exclude for any other effects or competing explanations on the absolute value of discretionary and real earnings manipulation as proxies for earnings management, but there concluded that they are very important variables to use in researching the relationship between earnings management and the effect of the IFRS adoption.

The choice of which control variables to use was dependent on the papers by Kim et al. (2012), Cohen et al (2008) and Burgstahler et al (2006). Only those variables that could reasonably be expected to have an effect on the earnings management more specific the proxies ABS_DA and RAM_PROXY were included.

3.2.2 Empirical Model

The research consist of two research models. As mentioned before the research models are based on the models used in Kim et al (2012) paper. The models will examine the following relationships (hypothesis); the first relationship is whether the adoption of IFRS did or did not effected the use of discretionary accruals in earnings management for Dutch Public companies. And the second relationship is whether the adoption of IFRS did or did not effected the use of real activities manipulation in Dutch public companies.

24

Figure 1 Emperical research design

Independent Variable Dependent Variable Concept level

Operational level

Figure 1 summarizes how the hypothesis (conceptual level) will be examined by the two research models (operational level). The use of discretionary accruals in earnings management and real activities manipulation are used as proxy for accounting quality. Accounting quality can be reduced due to managers using accounting choices to manipulate earnings in order to beat targets. In result this could lead to unfaithful presentation of financial reporting (Ahmed, Neel and Wang 2013).

Accounting Quality

IFRS Dummy (1,0) IFRS

Earnings Management ( proxy of Acc. Quality)

AEM -> discretionary acrruals (ABS_DA proxy) REM-> Real activities manipulation ( RAM_PROXY) H1: (+/- ) H2: (+/-)

25

As mentioned before Prior research shows that the adoption of the international standard IFRS improves accounting quality but other studies also show that chancing certain accounting standards can cause debatable accounting choices from managers which means that accounting quality does not particularly improves. Due to these contradictions in prior literature this research will examine the effect of implementing IFRS standards on Dutch public firms accounting quality. This empirical model will provide evidence in the Netherlands whether IFRS adoption has an effect on these specific proxies for accounting quality, the use of discretionary accruals and real earnings manipulation. The models will be examined by comparing companies that have adopted or have not adopted IFRS between the period of 2002 and 2011. So therefore a dummy variable is introduced. If the dummy variable is 1 it is considered the Post IFRS period (after adoption) and if it is 0 it is considered in the Pre IFRS period (before IFRS adoption). The first hypotheses will reflect the use of discretionary accrual in earnings management by Dutch public companies that did adopt IFRS standards compared to companies that did not adopted IFRS. In similarity the second hypothesis will reflect real earnings manipulation by also comparing firms that adopted IFRS and did not adopted IFRS.

Based on prior research if adopting the IFRS standard does improves accounting quality or on the other hand does not increase accounting quality I expect the first hypothesis to have a positive effect , meaning that the use of discretionary accruals will be reduced, or to have no effect at all. Similarly the same prediction stands for the second hypothesis, if adopting IFRS does improve accounting quality and provides a faithful presentation or on the other hand does not improve accounting quality. I expect the second hypothesis to have a positive effect, meaning that real earnings manipulation will decrease or on the other hand to have no effect at all.

3.3 data collection

The examinations will be based on the usage of existing data. The empirical approach can be considered archival study. The sample size will be based on all public companies in the Netherlands. I will focus on firm year observation between the period of 2002 till 2011. This period is chosen because the available information needs to be before and after the mandatory implementation period by public firms in Europe. The necessary data will be gathered via the WRDS integrated database system, using the Compustat database.

26 TABLE 2

Sample composition and Variable definitions Panel A: Sample derivation

Description N N(%)

Initial sample from Compustat 1362 100

Removal of firms with no data from the Statement of Financial Position

Assets 121 8.9

Liabilities 136 10.0

Equity 220 16.2

Removal of firms with no data from the Profit & Loss account

Revenue/Income 96 7.0

Expenses 446 32.7

Sample of Dutch Public firms from 2002 to 2011 343 25.2

Note: this table shows the sample composition using the Compustat database extracting data from Dutch public companies from 2002-2011.

The sample consist of 343 companies which are active listed companies in the Netherlands. I searched the fundamental annuals for the Dutch public firms in the entire Compustat database between the period 2002 and 2011. Table 2 shows that I initially obtained data of 1362 companies. After excluding firms with no data from the statement of financial position, specially firms missing asset , liability and equity data. In addition after excluding firms with no data from the profit and loss account , specifically firms missing revenue and expenses. I was overall left with 343 companies. Table 2 shows that a big part of the sample was excluded due to equity and expenses being the biggest data missing, a total of 16.2%+32.7%=48.9%.

27

4. Results

In this chapter, the results of the data analyses will be presented and discussed. This research examines two specific measures of earnings management. The research is based on the effect of IFRS adoption on earnings management. The two proxies that are used to measure earnings management are discretionary accruals and real activities manipulation. The sample consist of public firms in the Netherland between the period 2002 till 2011. Due to missing data firms with no information statement data and firms with no data from the profit and loss account were excluded. The final sample size consist of 343 companies. This chapter consist of chapter 4.1 were the descriptive statistics will be explained. Furthermore The results of the data analyses will be presented and discussed in chapter 4.2.

4.1 Descriptive statistics

TABLE 3

Panel A: Summary Statistics

Frequency of IFRS adoption between 2002-2011

Description N N(%)

Firms that implemented IFRS prior to 2005 49 14.3

Firms that did not implement IFRS prior to 2005 42 12.2

Firms that implemented IFRS after 2005 97 28.3

Firms that did not implement IFRS after 2005 155 45.2

Total Companies 343 100

Note: this table shows the summary statistics for firms that implemented IFRS prior to and after the year 2005.

Table 3 shows the descriptive statistics. The numbers of firms that have implemented IFRS prior to 2005 is 49. The reason I chose to compare before and after 2005 is because these international standards became mandatory in 2005. The tables shows how many firms switched to IFRS after the mandatory IFRS adoption period in 2005.The total sample consist of 343 companies. After 2006 table 3 shows that 97 companies implemented IFRS. So therefore more companies had implemented IFRS after the international standard became mandatory. Prior to 2006 42 firms did not implemented the IFRS standards.

28

The table also shows that after 2006 a big percentage of the total sample 343 companies did not implement the standard. Prior literature Hans, Edward & Martin 2008 examined the effect of IFRS adoption on accounting quality in Germany. Their sample also views an increase in sample of firms that adopted IFRS after the standards were mandatory. The switch sample consisted of 177 companies.

TABLE 3

Panel B: Descriptive statistics

Variables

Non-IFRS IFRS

(N=197) (N =146)

Mean Median Mean Median

Difference in Mean P-value T-test (A) (B) (C) (D) (C) - (A) Total Assets (x1000) 5.95 0.69 10.76 3.76 4.81*** 0.000 ROA 0.01 0.03 0.04 0.04 0.03*** 0.002 Leverage 0.15 0.14 0.20 0.01 0.05*** 0.000 RD_INT 0.08 0.02 0.06 0.06 -0.02 0.263 Size 6.57 6.54 8.39 8.23 1.82*** 0.000 ABS_DA -0.21 -0.64 0.28 -0.56 0.49 0.440 AB_CFO -0.02 0.06 0.02 0.03 0.04 0.885 AB_PROD 0.25 1.33 -0.34 1.05 -0.59 0.683 AB_EXP -0.11 -1.97 0.14 -1.72 0.25 0.808 RAM Proxy -0.37 -3.40 0.50 -2.76 0.87 0.669

Note: ***, **, * indicates that the variables for the group of firms that implemented IFRS during the period, as well as for firms that did not employ IFRS during the 2002-2011 period, at 10%, 5% and 1% sign. level respectively, based on a two-tailed t-test for the mean with unequal variances. All the variables are defined in Tablel 1.

29

Table 3 Panel B show the descriptive statistics of the sample groups. The firms did not adopted IFRS are compared to the control group that did adopted the IFRS standards. In order to obtain results from the sample that are representative for the population, I winsorized all continuous variables at the top and bottom 1% of the data. To exclude these ‘outliers’ I replaced the over extreme values with the most extreme, but acceptable, value at the top and bottom. In This table 3 panel B, I examine the characteristics of IFRS and non-IFRS. Characteristic IFRS consist of firms that implemented IFRS as a standard for their annual report. The other characteristic non- IFRS consist of firms that do not require IFRS as an standard for their annual report. The mean in the table represents the average of the numbers, while the median is the middle value in a list of numbers.

I calculated the mean and median for IFRS and non-IFRS for multiple characteristics. These characteristics are Total asset, Roa, Leverage, Rd_int, Size, ABS_DA, AB_CFO, AB_PROD, AB_EXP, RAM_PROXY. After calculating the mean and median for these multiple characteristics I compared the means and conducted a two-tailed T-test for comparing two means using independent samples when the populations have unequal variances. Firms not using IFRS have mean total asset of 5.95 and firms using IFRS have mean total asset of 10.76. IFRS firms have a mean ROA of 0.04 and non IFRS firms have a mean ROA of 0.01. Firm acquiring IFRS have e mean leverage of 0.20 and firms not requiring IFRS have a mean leverage of 0.15. Firms that obtain IFRS mean size is 8.39 and non IFRS firms mean size is 6.57. Firms that acquire IFRS standards, the means for ABS_DA, AB_CFO, AB_PROD & AB_EXP are respectively 0.28; 0.02; -0.034 & 0.14. For the non IFRS firms the mean for the variables ABS_DA, AB_CFO, AB_PROD & AB_EXP are respectively -0.21; -0.02, 0.25; & -0.11. The table also shows that firms IFRS companies have a mean RD_INT of 0.06 and on the other hand firms not using IFRS have a mean RD_INT of 0.08. Furthermore IFRS firms have mean RAM PROXY of 0.50 and the non IFRS firms have a mean RAM PROXY of -0.37.

This analysis shows that the difference between means is significant for the total assets, the return on assets, firm size as well as the Leverage. Based on these results I can conclude that firms acquiring IFRS have significantly higher total assets and significant return on assets compared to non-IFRS firms. The leverage and firms size are also significantly higher for IFRS firms, compared to non-IFRS firms. Furthermore the two dependent variables both have insignificant difference between the means of IFRS firms and non IFRS firms. This could mean that the use of discretionary accruals and real earnings manipulation is the same for both IFRS firms and non IFRS firms.

30 Table 3

Panel C :Correlation matrix

Note: this table shows the Pearson’s correlation between all of the variables. All the variables are defined in Table 1 panel A. Each of the continuous variable is winsorized at 1% and 99% to mitigate outliers.

The purpose of the above Pearson’s correlation table is to see whether there is a relationship between variables. For instance the strength between the dependent variables can be found. The correlation coefficient measures the strength of the relationship between the variables. This measures varies from 0 (no relationship) to +1 and to -1 (perfect relationship). If the coefficient varies from 0.1 till 0.3 there can be concluded that the relationship between the variables is weak. The relationship between the variables is moderate when the coefficient varies between 0.4 and 0.6. Also there can be concluded that the relationship is strong if the coefficients vary from 0.7 till 0.9.

The results show that Total assets and Size are strongly associated. The reason for this is because Total assets is used to calculate the companies Size. Furthermore AB_PROD and AB_EXP are strongly associated with COMBINED_RAM. This is due to the fact that COMBINED_RAM consist of AB_PROD, AB_EXP and AB_CFO. Overall the results show weak relationships between the variables. Thus the independent variables are not associated with each other.

31

4.2 Test of Baseline Predictions

In order to examine the baseline predictions (hypothesis) as stated earlier in the paper, I use a regression analysis. The results from the regression analysis are shown in table 4 below. The actual regression outputs are displayed in the appendix.

TABLE 4: Model 1

Panel A: The determinants of the ABS_DA coefficient

Total Sample IFRS - only non-IFRS

Independent Variables Dependent Variable

ABS_DA (1) ABS_DA (2) ABS_DA (3)

IFRS -1.00 TOTAL ASSET 0.00*** 0.00 0.00*** ROA 1.16 -9.67 4.88 LEVERAGE 9.32*** 4.48 12.72*** RD_INTS 1.24 -2.77 2.11 SIZE 0.96*** 0.99* 0.77** COMBINED_RAM 0.03** -0.04 0.05** N 342 146 628 R Square 0.14 0.06 0.31 Adj R Square 0.12 0.02 0.30

This table reports the regression summary output for the sample. ABS_DA (1) shows the regression output for the total sample to be able to interpret the results. The independent variable (IFRS dummy) is coded 1 if the listed company adopted IFRS and 0 if the listed company did not adopted IFRS. ABS_DA (2) shows the output only for the IFRS firms and ABS_DA (3) for the firms that did not adopted IFRS during the period 2002-2011. *,**,*** denote two-tailed statistical significance at 10%, 5% and 1% respectively. All variables are explained in Table 1 Panel A.

32

The regression analysis as conducted in table 4 Panel A for ABS_DA is as follows. ABS_DAt (or DAt)=

α

α

α

α

+

α

ε

A regression analysis shows how one or multiple independent variables affect the dependent variable based on a significance interval (Alfa). This regression function shows how the independent (IFRS dummy) and control variables affect the ABS_DA. The first variables is the dummy used to indicate whether a firm has adopted IFRS or has not adopted IFRS or switched to IFRS. The additional variables are the control variables included. The results show that the independent IFRS (dummy) variables has an insignificant coefficient.

According to the literature review, that if I expected an significant coefficients for the IFRS variables. I could reasonably expect that when these international standards do provide higher financial reporting or present a more faithful economic presentation, the use of discretionary accruals in earning management could be affected. In this case there could be expected for the ABS_DA to decrease.

When these international standard or higher standards do not necessary improve accounting quality or do not increase financial reporting. I could reasonably expect for the IFRS variable to be insignificant. Thus not effecting the discretionary accruals. So therefore this regression analysis shows that the value of the coefficients is negative and the related P-Values is insignificant. There could be concluded that the IFRS variable has no effect on ABS_DA.

In table 4( Panel A )ABS_DA ( 2), I examine the regression of a sample, including only IFRS firms. These are the firms that had adopted or switched to IFRS during the period 2002-2011. Table 4 ABS_DA (3) shows the regression result for the sample of firms that did not adopted or switched to IFRS during the 2002-2011 period. This sample could be named the control group. A comparison of the results from the regression analysis could explain more on the differences between the firms in the two sample groups. Table 4 ABS_DA 2 shows that only the variable Size, is significantly associated with the ABS_DA. So therefore Size has a positive relationship with the use of discretionary accruals in earnings management for the regression of a sample only including IFRS firms. Table 4 ABS_DA 3 shows multiple variables that are significantly associated to the ABS_DA. Thus Total asset, Leverage, Size and Combined_Ram have a positive relationship with the use of discretionary accruals in earnings management.

33

The regression analysis as conducted in table 4 Panel B for ABS_DA is as follows. RAM_PROXY =

α

α

α

α

+

α

ε

This regression analysis shows how one independent variables and multiple control variables affect the dependent variable based on a significance interval (Alfa).

TABLE 4: Model 2

Panel B: The determinants of the RAM_PROXY coefficient

Total Sample IFRS-only non-IFRS

Independent Variables Dependent Variable

RAM_PROXY (1) RAM_PROXY (2) RAM_PROXY (3) IFRS -4.32* TOTAL ASSET 0.00 0.00 0.00 ROA 23.28* 2.54 26.60 LEVERAGE -1.30 -2.29 -1.64 RD_INTS 11.17 -1.19 14.37 SIZE 2.92*** 0.67 3.15** ABS_DA 0.38** -0.30 0.58** N 342 146 196 R Square 0.07 0.09 0.10 Adj R Square 0.05 0.05 0.07

Note: this table reports the regression summary output for the sample. RAM_PROXY (1) shows the regression output for the total sample. The independent variable (IFRS dummy) is coded 1 if the listed company adopted IFRS and 0 if the listed company did not adopted IFRS. RAM_PROXY (2) shows the output only for the IFRS firms and RAM_PROXY (3) for the firms that did not adopt IFRS during the period 2002-2011. *,**,*** denote two-tailed statistical significance at 10%, 5% and 1% respectively. All variables are explained in Table 1 Panel A.

34

The above regression formula shows how the independent (IFRS dummy) variable and the control variables affect the RAM_PROXY. Similar with the first regression model the first variables is the dummy used to indicate whether a firm has adopted IFRS or has not adopted IFRS or switched to IFRS. The additional variables are the control variables included. The results show that the independent IFRS (dummy) variables has an significant coefficient. Based on the before mentioned literature review, if there is an significant coefficients for the IFRS variables. I could reasonably expect that when the adoption of these international standards do provide an higher financial reporting or present a more faithful or reliable economic presentation, the use of real earning management could be affected. Thus this could mean that there could be expected for RAM_Proxy to decline. On the other hand When these international standard IFRS or high qualified standards do not exactly increase financial reporting or on the other hand improve accounting quality. I could reasonably expect for the IFRS variable to be insignificant. This regression analysis shows that the value of the coefficients is negative and the related P-Values is significant. So therefore could be concluded that the IFRS has an significant effect and negative relation with real earnings manipulation.

In table 4( Panel B ) RAM_Proxy ( 2), similary as in Tabel 4 Panel A , I examine the regression of a sample including only IFRS firms. These are the firms that had adopted or switched to IFRS during the period 2002-2011. Table 4 RAM_Proxy (3) shows the regression result for the sample of firms that did not adopted or switched to IFRS during the 2002-2011 period. A comparison of these results from the regression analysis for RAM_Proxy could also explain more on the differences between the firms in the two sample groups. Table 4 RAM_Proxy 2 shows that not any of the variables are significantly associated with the RAM_proxy. Thus ROA, Leverage, Size, RD_INTS nor ABS_DA have a relationship with the RAM proxy. Table 4 ABS_DA 3 shows that the variables size and ABS_DA are significantly associated to RAM_proxy. Thus exanimating the regression of the sample including non IFRS FIRMS , Size and ABS_DA have a positive relationship with the use of real earnings manipulation.