The Effect of IAS 19R on Earnings
Management
Master thesis, 15 ECTS Summer 2014
Author: Celina Mlodzick Student number: 10607897
Abstract
This thesis researches the effect of the revised IAS 19 on earnings management. Existing literature predicts an increase in earnings management caused by the revision of the IAS 19, in particular the
abolishment of the corridor approach together with the introduction of one single interest rate. The sample for the research exists of Dutch listed
companies and is divided in three groups, each for which a different effect is expected. For the measurement of earnings management the Modified Jones Model has been used and the change in discretionary accruals in 2013 as opposed to 2012 has been researched for the whole sample as well as for the individual groups. The findings suggest no significant increase in earnings management in 2013 as opposed to 2012. However they do imply that
companies offering a defined benefit pension plan perform more earnings management in general than companies not offering this plan.
Keywords: leverage; company size; debt covenants; pension accounting; revised IAS 19; IFRS; corridor approach; OCI; earnings management; the modified Jones model; discretionary accruals;
Definitions and acronyms
EM Earnings management
H Hypothesis
IASB International Accounting Standards Board
IAS 19R International Accounting Standard 19 Revised
IFRS International Financial Reporting Standards
OCI Other Comprehensive Income
PAT Positive Accounting Theory
P&L Profit & Loss
PPE Power, Plant and Equipment
REV Revenues
REC Receivables
t Time (Year)
TAC Total Accruals
TAS Total assets
Table of Contents
Abstract ... 2
Definitions and acronyms ... 3
1. Introduction ... 6
1.1 Background ... 6
1.2 Disposition ... 10
2. Theory & literature review ... 11
2.1 Introduction ... 11
2.2 Economic consequences: Impacts of changing accounting standards ... 12
2.2.1 Economic consequences ... 12
2.2.2 Perspectives ... 14
2.3 Earnings management ... 15
2.3.1 What is Earnings Management ... 15
2.3.2 How to perform Earnings management ... 16
2.3.3 Incentives ... 17
2.4 Pension accounting ... 20
2.4.1 IAS 19 (1980-2014) ... 20
2.4.2 Defined benefit plan ... 22
Full recognition through P&L ... 25
Full recognition through equity ... 25
Corridor approach ... 25
2.4.3 Critique former IAS 19 ... 27
2.4.4 IAS 19R changes ... 28
2.4.5 Possible impact of IAS 19R ... 29
2.4.6 Development of Hypothesis ... 31
3. Research Design ... 33
3.1 Sample ... 33
3.2 Time frame ... 34
3.3 Testing for debt covenants ... 34
3.4 Testing for Earnings Management ... 35
3.4.1 Models ... 35
3.4.2 The Modified Jones Model ... 36
3.5 Control variables ... 37 3.6 Regression formula ... 38 4. Results ... 39 4.1 Descriptive statistics ... 39 4.2 Pearson Correlations ... 45 4.3 Regression Findings ... 47 5. Conclusion ... 54 5.1 Conclusion ... 54 5.2 Limitations ... 54
5.3 Future research ... 55 6. References ... 56 7. Appendix ... 63
1. Introduction
The introduction clarifies the underlying reason for this research and its contribution to the existing literature.
1.1 Background
Over the past decades the existence of private pension plans has increased significantly for cultural, economic, as well as political reasons. Not only are these arrangements the most common and desired of the assorted ‘fringe benefits’ offered by employers in many nations, they also often represent an entity’s largest single financial liability (Holt, 2012; Mackenzie, 2011)
‘Unlike for wages and other fringe benefits, the timing of the payment of cash to either the plan’s administrator or to the plan beneficiaries can vary
substantially from the underlying economic event. This creates the possibility of misleading financial statement representation of the true costs of
conducting business, unless a valid accrual method is employed. For this reason and also because of the complexity of these arrangements and the impact they have on the welfare of the workers, accounting for the cost of pension plans and similar schemes has received a great deal of attention from national and international standard setters’ (Mackenzie, 2011). This attention mainly being negative critique on the accounting regulations for pension plans stating it gives too much possibility for performing earnings management and has a lack of clarity and understandability (Bloom 2013; Bergstresser et al. 2006).
In June 2011, the International Accounting Standards Board (IASB)
announced a revision of the IAS 19, employee benefits (IAS 19R). This new standard is effective for annual periods beginning on or after January 2013
and entails several modifications with regard to the former IAS 19. The revised IAS introduces adjustments in classification, valuation, presentation and explanation that should reduce the possibility of earnings management and make reporting on employee benefits less complex and better
comparable (Böhmer et al.,2013; Bloom, 2013).
Next to having an impact on pension reporting itself, the modifications also are claimed to impact treasury, risk management and governance (Loman and Siesling, 2011). Severinson (2010) adds that the modifications also affect how companies design, finance and think about the pension promises they offer to their employees.
The revised IAS 19 clearly is expected to have great impact on companies reporting under IFRS, which makes it an often-discussed topic. The most common topics of discussion seem to be the abolishment of the corridor approach together with the introduction of a single interest rate claiming it leads to higher volatility of equity respectively lower profits and could lead to higher earnings management.
The aforementioned developments in pension reporting form the motivation to research the relationship between earnings management and pension accounting under the revised IAS 19 and result in the following research question:
What effect does the revised IAS 19 have on earnings management compared to the former IAS 19?
Since the IAS 19 revision has occurred recently and is applicable to annual periods beginning on or after January 2013, no research analysing the actual impact has been performed yet while most of the annual reports of 2013 are prepared and audited at the moment. This paper, however, will examine these recent annual reports and thus will contribute to the existing literature by being the first paper to expose some of the real effects of IAS 19R. Furthermore it will contribute to investors’, owners’, board of directors’, and regulators’ understanding of the impact of the revised regulation.
The effect of IAS 19R on earnings management is an important topic of research from a societal point of view because it affects the way the annual report is presented and despite the implications of the efficient security markets, according to the economic consequences concept, accounting policies can affect decision-making behaviour (Zeff, 1978). Decision-making behaviour of society as well as of management and employees from the organization itself is influenced. Society uses the information from annual reports mainly for making decisions concerning buying/selling shares and deciding whether they support the way of doing business of a certain firm. The preceding together with the means available for paying out dividends is also of importance for existing shareholders. Whereas for managers the financial statements are of high importance for making investment decisions and determining their bonuses and reputation.
This paper should clear up what effect the revised IAS 19 has on earnings management compared to the former IAS 19 by first explaining the meaning, relation and influence of economic consequences in the standard setting process. Next chapter 2.3 further elaborates upon one economic
chapter 2.4, the IAS 19 along with its revision and consequences is analysed and hypothesis are formed. Then in chapter 3 the research design will be explained, this includes the sample, the measurement of the variables, the control variables and the regression formula. Chapter 4 continues with the research itself consisting of the descriptive statistics, the correlations found and the regression findings. Finally a conclusion is formed on whether the revised IAS 19 has led to an increase in earnings management.
1.2 Disposition
Introduction
• The introduction clarifies the underlying reason for this research and its contribution to the existing literature.
Theory and Literature review
• In this section the existing literature and theory with regard to economic consequences, earnings management and pension accounting is discussed and hypotheses are formed.
Research Design
• In this paragraph a method is developed for testing the relationship between the revised IAS 19 and earnings management.
Results
• This chapter presents results of research on the collected data.
Conclusion
• The final part draws a conclusion, based on the performed research, with regard to the relationship between the revised IAS 19 and earnings management
2. Theory & literature review
In this section the existing literature and theory with regard to economic consequences, earnings management and pension accounting is discussed and hypotheses are formed.
2.1 Introduction
Several authors argue that the implementation of new accounting standards not only influence the accounting numbers in the financial statements but can also impact the behaviour of different market participants like the society as a whole, investors, employees, managers and auditors (Zeff, 1978; Rappaport, 1977; Blake 1992). These possible economic consequences of changes in accounting standards can be analysed from three different approaches that will be further described in paragraph 2.2.
One of these economic consequences, namely management behaviour, in specific earnings management, is of special interest in this research and will be elaborated upon in paragraph 2.3. The reasons for performing earnings management as well as its methods are described.
Next, pension accounting under IFRS is explained. The former IAS 19, its changes as well as its consequences will be included in paragraph 2.4.
2.2 Economic consequences: Impacts of changing accounting standards
2.2.1 Economic consequences
Before the 1960’s accounting policy-making was seen as merely a technical process assumed neutral in its effect. This view changed as several individuals and groups became aware of its impact on decision-making behaviour of business, government, unions, investors and creditors, also referred to as ‘economic consequences’. As a result these parties tried to influence the standard-setting process by lobbying and standard setting evolved into a complex process with many outside forces (Zeff, 1978).
Blake (1992) classifies the economic consequences in three groups namely compliance/analysis costs, mechanistic consequences and judgemental consequences.
• Compliance/analysis costs refer to the impact of a change in
accounting requirements on the cost of compliance for companies. For analysts, an increased required disclosure may reduce costs of obtaining that information from other sources or bearing risks of ignorance, while at the same time leading to increased processing costs for companies. Reduced disclosure requirements may have the converse effect.
• Mechanistic consequences are considered as those consequences that arise because the reported numbers ‘trigger off’ a mechanism that affects the economic position of the reporting entity. Two types of mechanisms can be distinguished namely:
Regulatory. In this case the mechanism is imposed by some regulatory body, think of tax assessments on reported figures by governments.
Contractual. Here the form of the mechanism is defined in a contract between the company and some other party. A frequently cited example are the limitations on a company’s borrowing powers found in the articles of association.
• Judgemental consequences refer to those economic consequences that arise because of decisions taken by some readers of accounts in response to the information provided. A distinction is made between two categories;
At the micro level individual users of accounts may change their personal conduct towards the company in some way. Think of the decision to buy/hold/sell shares.
At the macro level the published figures can influence the individuals in a way that affects the political, economic and social climate.
2.2.2 Perspectives
The economic consequences can be analysed from three different
perspectives namely the positive accounting-, behavioural- and market-based approach.
Positive accounting
This theory tries to predict a firms’ choices of accounting policies and their response to new accounting standards. Positive accounting theory is organized around three hypotheses formulated by Watts and Zimmerman (1978), all three assuming that managers choose accounting policies in their own best interest.
• Bonus plan hypothesis states that, all other things being equal, managers of firms with bonus plans are more likely to choose
accounting procedures that shift reported earnings from future periods to the current period as this should enlarge their remuneration.
• Debt covenant hypothesis predicts that, ceteris paribus, the closer a firm is to violation of accounting based debt-covenants, the more likely the firm manager is to select accounting procedures that shift reported earnings from future periods to the current period.
• Political cost hypothesis states that, ceteris paribus, the greater the political costs faced by a firm, the more likely the manager is to choose accounting procedures that defer reported earnings from current to future periods. (Scott, 2012)
Behavioural approach
This approach answers how managers, users and investors respond to changing accounting standards. It pays attention to the psychological aspects of decision-making. Reactions, being judgments and decisions of
market participants, following a change in accounting standards are
researched (Aslan, 2009). This could be, for example, a survey based study. Market-based approach
This last approach studies the relation between accounting numbers and the capital market. Lev and Ohlson (1982) explain market based accounting research as ‘research into the relationship between publicly disclosed
accounting information and the consequences of the use of this information by the equity investors – as such consequences are reflected in characteristics of common stocks traded in major exchanges’.
2.3 Earnings management
2.3.1 What is Earnings Management
As mentioned earlier, management can choose amongst several accounting procedures in presenting the financial statements. This is possible due to the principle-based character IFRS has. The freedom it consequently offers to companies is meant to be used to adapt the IFRS to their particular situation, leading to more understandable and useful statements (Mohr). There is, however, a downside to the flexibility IFRS allows. Companies namely get the opportunity to let the financial statements show merely desired results by only utilizing the method they wish. This can lead to revenue or profit manipulation and can even encourage fraud (Mohr). This process where managers use judgement in financial reporting and in structuring transactions to alter 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 is called earnings management (Healy and Wahlen 1999).
2.3.2 How to perform Earnings management
Since earnings are composed of cash flow from operations and accruals, firms have two options to manage earnings. First, firms can manage earnings
through deviation from the normal business operations, so that the cash flow from operations will be affected. Deviating from normal business practices to manipulate reported income is defined as real earnings management
(Roychowdhury, 2006). Real earnings management includes actions as altering the level of discretionary expenditures, such as research and development expenditures and selling, general and administrative expenditures, overproduction and provision of price reductions to boost sales volume (Roychowdhury, 2006).
Second, a firm can alter the level of accruals to obtain the desired level of earnings. This is defined as accrual-based earnings management (Healy and Wahlen, 1999). The primary objective of accruals is to demonstrate the true performance of the firm by recording revenues and expenses to the period in which they are incurred, rather than presenting the cash in- and outflows (FASB 1985, SFAC no.6). However, as stated, accruals can also be used for manipulation of earnings since managers themselves have to use judgment in deciding on the value of these. Examples of accruals are cash payable, cash receivable, cash received in advance and deferred tax liabilities.
Past research either used total accruals or merely discretionary accruals as a measure of earnings management. In this last method non-discretionary accruals are excluded for the reason that they reflect business conditions that a manager has no control over, like sales growth, while discretionary accruals identify management choices (Christensen et al., 2013).
2.3.3 Incentives
Earnings management is a managerial activity and consequently driven by managers’ incentives (Stolowy and Breton, 2004). Healy and Wahlen (1999) distinguish three categories of motivation that will be presented in this paragraph supplemented by the signalling theory.
2.3.3.1 Capital Market motivation
According to Healy and Wahlen (1999) ‘the widespread use of accounting information by investors and financial analysts to help value stocks, can create an incentive for managers to manipulate earnings in an attempt to influence short-term stock price performance.’ Usually a company will try to satisfy the expectations of investors and financial analysts by showing a steady growth in their earnings. So management will attempt to prevent large fluctuations in their stock prices by means of earnings management. This is confirmed by Gaver et al. (1995) who claim that companies will try to decrease the earnings in years with higher than average earnings and increase the earnings in years with lower than average earnings.
2.3.3.2 Contracting motivation
This motivation entails two of the PAT hypothesis namely the bonus plan and debt covenant hypothesis. Both can be seen as a contract either to align the incentives of management and external stakeholders or to limit managers’ actions that benefit the firm’s stockholders at the expense of its creditors These contracts create incentives for management to manipulate earnings in order to comply with the conditions included in them (Healy and Wahlen, 1999).
• Bonus plan
The bonus plan hypothesis simply entails that managers tend to choose accounting procedures that shift reported earnings from future periods to the current period too enlarge their bonus if this (partly) depends on reported net income.
• Debt covenant
Most business borrow money, either they use it to expand their business or simply to receive tax benefits. The banks and other organisations lending this money face the risk that the company borrowing it, won’t be able to pay the interest and principal amount back. To constrain this possibility, the banks include a few conditions in their lending agreement. One of these conditions often being the solvability ratio, also called the debt/equity ratio, which makes sure that the company won’t have too much debt as opposed to equity. When the debt/equity ratio stated in the loan agreement is violated there often is a decent fine attached to it, that companies most certainly want to avoid. To avoid, or at least postpone the violation, management tries to influence the ratio through shifting future revenues/profits to the present and hereby increasing the equity (Sweeney, 1994)
2.3.3.3 Regulatory motivations
A third incentive for performing earnings management are industry-specific and anti-trust regulations. These regulations are put into place to, amongst others, prevent earnings management but can also have the opposite effect. Industry-specific regulations are sometimes explicitly tied to accounting data, like capital adequacy for banks. Previous literature finds that such regulations create incentives to use accounting discretion to get around industry-specific regulatory constraints. Whereas for the anti-trust regulations, it seems that
companies vulnerable to anti-trust investigation or other political
consequences have incentives to perform earnings management to appear less profitable (Watts and Zimmerman, 1978). This refers to the political costs hypothesis discussed in paragraph 2.2.2. ‘Political costs can, for example, be imposed by high profitability, which may attract media and consumer
attention. Such attention can quickly translate into political ‘heat’ on the firm, and politicians may respond with new taxes or other regulations. This has happened to oil companies during periods of restricted crude oil supply and rising gasoline prices. Often sheer size can lead to political costs as very large firms may be held to higher performance standards, for example with respect to environmental responsibility, simply because they are large and powerful. If the large firms are also highly profitable, such political costs will be
magnified’ (Scott, 2012).
2.3.3.4 Signalling motivations
The last frequently discussed incentive that can be distinguished is the signalling motivation, which entails the attempt by management to signal a certain image of the company using earnings management. This is in line with Iatridis and Blanas (2011) who claim that by reporting lower profitability the firm sends negative signals to their stakeholders about the firm’s current financial performance and future potential and vice versa. Subramanyam (1996) adds an advantage to managing earnings namely that by using estimations to disclose inside information managers provide shareholders with blocked information, increasing the relevance of the financial
2.4 Pension accounting 2.4.1 IAS 19 (1980-2014)
The objective of IAS 19 is to prescribe the accounting and disclosure for employee benefits, requiring an entity to recognize a liability where an employee has provided service and an expense when the entity consumes the economic benefits of employee service. The principle underlying all of the detailed requirements of the Standard is that the cost of providing
employee benefits should be recognized in the period in which the benefit is earned by the employee, rather than when it is paid or payable [IAS
19(2011).2].
The regulation has a long history. Being first introduced in 1980, effective in 1983, it has already been modified several times. In the early days, almost any approach to pension accounting was acceptable (Peters, 2010).
Nowadays the IASB has limited these options but the rules are still far from ideal, issues keep arising. The latest attempt to make the reporting on employee benefits less complex and better comparable is introduced in 2011. The exact content of these latest amendments will be described later in this paragraph. First the foundation of accounting for employee benefits will be shortly elaborated upon.
The standard IAS 19 identifies four categories of employee benefits namely: • Short term employee benefits such as wages, salaries and social
security contributions;
• Post-employment benefits such as retirement benefits and post-employment medical care;
• Other long term employee benefits such as long-service leave and long-term disability benefits and
• Termination benefits
This paragraph will continue with a focus on the category post-employment benefits, as this thesis intends to specifically research the effect of the revision on earnings management for this category.
Post-employment benefit plans are informal or formal arrangements where an entity provides post-employment benefits to one or more employees. The accounting treatment for a post-employment benefit plan depends on the economic substance of the plan and results in the plan being classified as either a defined contribution plan or a defined benefit plan (iasplus.com). Under defined contribution plans, an entity pays fixed contributions into a separate entity (a fund) and will have no legal or constructive obligation to pay further contributions if the fund does not hold sufficient assets to pay all employee benefits relating to employee service, in the current and prior periods (iasplus.com). The accounting for this plan is straightforward; The pension expense equals the required contributions and in case of
prepayments or payments due, these are recorded on the balance sheet. All other post-employment benefit plans are defined benefit plans. These are generally retirement benefit plans under which amounts to be paid as
retirement benefits are determinable, usually by reference to employees’ earnings and/or years of service. The fund (and/or employer) is obligated either legally or constructively to pay the full amount of promised benefits whether or not sufficient assets are held in the fund (IAS 19). The next paragraph is dedicated to the accounting for this plan since it rather complex.
A special common pension plan is a multi-employer plan, which is a
collectively bargained plan maintained by more than one employer, usually within the same or related industries. Like other plans, multi-employer plans can either be defined benefit or a defined contribution plans (PBGC, 2014). Most employers, however, don’t have access to enough information to account for this plan as a defined benefit plan thus tend to process it as a defined contribution plan.
2.4.2 Defined benefit plan
The accounting for defined benefit plans is comparatively more complex and remains a controversial subject because of the heavy impact that various management assumptions have on expense determination (Mackenzie et al.
2013). The complexity along with its increasing costs has diminished the
plans’ popularity. Most major employers allow new hires to sign up for defined contribution retirement plans instead (Bloom, 2013).
2.4.2.1 Measurement
In this paragraph the accounting for the defined benefit plan is presented. The measurement in the balance sheet and in the P&L is discussed
separately.
Pension Asset/Liability
The net pension asset/liability reported in the balance consists of the following components:
• Present value of the pension obligation • - Fair value of Pension Asset,
• +/- Unrecognized Actuarial Gain/Loss that can be processed either through the P&L, equity or corridor approach, see next paragraph • - Past Service cost not yet recognized
Present value of the pension obligation
This is calculated as the present value of expected future payments required to settle the obligation resulting from employee service in the current and prior periods. The present value of the defined benefit obligation should be determined using the Projected Unit Credit Method. [IAS 19(1998).64] The rate used to discount estimated cash flows is determined by reference to market yields at the end of the reporting period on high quality corporate bonds, or where there is no deep market in such bonds, by reference to market yields on government bonds. [IAS 19(1998).78]
Fair value of pension assets
This is the investment fund. During the year investments are made to
increase the size of the fund. This is the return on plan assets. Also, employer contributions, cash the company gives from its own bank account, will
increase the fund. Finally, benefits paid to current retirees will reduce the plan assets (investopedia.com).
Unrecognized Actuarial Gain/Loss
On an on-going basis, actuarial gains and losses arise that comprise experience adjustments (the effects of differences between the previous actuarial assumptions and what has actually occurred) and the effects of
changes in actuarial assumptions. Under the former IAS 19 the unrecognized
gains or losses could be processed in three ways; full recognition through P&L, full recognition through equity or by using the corridor approach
Past service costs
Past service costs consist of the change in the obligation for employee
service in prior periods, arising as a result of changes to plan arrangements in the current period. Past service costs are recognised immediately to the extent that it relates to former employees or to active employees already vested. Otherwise, it is amortised on a straight-line basis over the average period until the amended benefits become vested. [IAS 19(1998).96]
Pension Expense
The amount recognised in the P&L is made up of the following components: [IAS 19(1998).61]
• current service cost (the actuarial estimate of benefits earned by
employee service in the period)
• interest cost (the increase in the present value of the obligation as a
result of moving one period closer to settlement)
• expected return on plan assets and on any reimbursement rights
• actuarial gains and losses
Expected return on plan assets
The return on plan assets is interest, dividends and other revenue derived from the plan assets, together with realised and unrealised gains or losses on the plan assets, less any costs of administering the plan (other than those included in the actuarial assumptions used to measure the defined benefit obligation) and less any tax payable by the plan itself. [IAS 19(1998).7] (iasplus.com)
2.4.2.2 Three approaches of pension benefits recognition IAS 19
As mentioned, the former IAS 19 gave three possibilities for processing actuarial profits and losses in the annual report. In this paragraph the methods will be discussed more thoroughly.
Full recognition through P&L
The first method entails processing the actuarial results directly in the profit-and-loss account. The advantage of this method is that no profits and/or losses are parked at the balance sheet. A big disadvantage is that it leads to fluctuations in the results that can’t be influenced by the management.
Full recognition through equity
A second possibility under the IAS 19 is to process actuarial results directly into the equity. This prevents inclusion in the balance sheet and big
fluctuations in the result, which can both be seen as advantages. The disadvantage it brings along is that, in case of big actuarial results, it can cause severe fluctuations in equity affecting the solvability ratio.
Corridor approach
The third and last possibility and also the most criticized method is the
corridor-approach. It prescribes that the cumulated actuarial gains and losses that exceed a predetermined level (corridor) have to be recognized in profit or loss. The corridor of actuarial gains and losses that does not have to be recognised is limited, in IAS 19.92, to 10% of the higher of the present value of the defined benefit obligations and the fair value of plan assets. According
to IAS 19.93, the excess should be recognized during the average remaining working lives of the participating employees (Amen, 2007).
The big advantage of this method is that actuarial results, which management can’t really influence, almost don’t impact the result and equity. The
disadvantage, and also the most criticized effect of this method, is that the profits and losses aren’t directly visible. This is contradictory with every general accepted accounting standard that prescribes to take losses
immediately when they are known. IAS 8, for example, states that changes in estimations and assumptions should be processed in the period in which they are known. IAS 37 indicates that modifications in obligations (also long-term obligations like pension obligations) should be processed in the year these changes occur (iasplus.com). Peters (2010) agrees with this stating that under the former IAS 19 companies have the option of ignoring almost all of the volatility and risks arising from pension schemes indefinitely.
2.4.2.3 Usage of IAS 19 methods
Research by Fasshauer et al. (2008) shows that the choice between the three options, as mentioned in the previous paragraph, differs per country. Where countries in the United Kingdom, Ireland and Germany mainly choose to process the actuarial results in the equity, firms in France, Finland,
Switzerland, Sweden and the Netherlands seem to choose mainly for the corridor method. The firms previously using the corridor approach are likely to suffer the greatest consequences from the IAS 19R.
(Fasshauer, Glaum en Street, 2008)
2.4.3 Critique former IAS 19
Not only the corridor method was heavily criticized, other aspects of the former IAS 19 were also questioned.
The interest rate used to calculate the pension asset/liability for example; Under the former IAS 19 the pension funds and organizations with a pension scheme had to valuate their obligation based on actual market interest, as a result of fair value accounting. The purpose of the implementation of the fair value accounting was to eliminate subjective and complex valuation methods (Swinkels, 2009). However, in practice, the opposite was proven; According to Vergoossen (2007) the subjectivity while valuating and as a result of that, the possibility to influence the profit numbers has increased significantly. This is confirmed by Brouwer and Siesling (2009) who claim that the attractiveness of earnings management regarding defined-benefit plans and the space for creativity and subjectivity has increased as a result of implementation of fair
value accounting in pension reporting. Bergstesser et al. (2006) also agree with this statement by saying that IAS 19 gives too much possibility in performing earnings management through overestimating the expected proceeds on investments.
Another often heard point of critique is that the IAS 19 has a lack of clarity, comparability and understandability (Bloom, 2013;Fasshauer et al.,
2008;Peters, 2010). One aspect contributing to this lack is that there are too many methods of processing unrecognized gains and losses. Fasshauer et al. (2008) add that the explanation on defined benefit pension schemes,
especially regarding the actuarial assumptions, is on an insufficient level. This results in users of the annual reports, having a hard time distinguishing
between the various pension schemes, making the comparability of annual reports harder. Furthermore this lack of explanation makes it harder for users to estimate the financial impact of a defined benefit pension.
2.4.4 IAS 19R changes
In March 2008, the IASB published a discussion paper regarding the former IAS 19 to communicate the desired modifications, most of them as a
response to the existing critique. This project formed part of the Memorandum of Understanding between the IASB and the Financial Accounting Standards Board, by further aligning IFRS and US GAAP (Holt, 2012)
The revised IAS 19 should make reporting on employee benefits less complex better comparable and should reduce the possibility of earnings management (Böhmer et al.,2013; Bloom, 2013). By making the reporting less complex it should also raise awareness among users of the financial
statements of the risks associated with the pension commitments (Holt, 2012).
The key changes introduced by IAS 19 (2011) as compared to IAS 19(1998) were the following:
• Expected return on assets and interest costs are replaced with an item equal to net interest, which is the interest rate associated with high-quality corporate or government bonds. This change to a single interest rate is implemented to prevent companies from overstating the expected return on plan assets and thereby lower their defined-benefit plan expense thus raising their net income (Bloom, 2013) • Introducing a requirement to fully recognize changes in the net
defined benefit liability/asset including immediate recognition of defined benefit costs and requiring the recognition of remeasurements in other comprehensive income (eliminating the 'corridor' approach) • Introducing enhanced disclosure requirements for defined benefit
plans. The characteristics of benefit plans, the amounts recognised in the financial statements and the risk arising from defined benefit plans should now be presented. Which not only leads to more extensive disclosure but also more subjectivity in determining that disclosure (Holt, 2012).
2.4.5 Possible impact of IAS 19R
Some of the objectives of this revision seem accomplished. The standard is indeed expected to lead to greater transparency, uniformity and
comparability in financial statements. (Holt, 2012; Bloom, 2013; Hartwell, 2012)
However the fact that the revised IAS 19 is mainly seen as a quick fix by individual researchers, as well as by the big four and the authors themselves already leads us to expect that there will be quite a few points of critique left regarding this accounting standard (Peters, 2010; comments on discussion paper by BDO, KPMG, EY, PWC and Deloitte).
This indeed seems the case; the first point of critique is on the introduction of one single discount rate. It will namely not reflect the benefits from a higher return on riskier investments anymore. This will not only generally lead to lower net profit but also to firm becoming more conservative in its
investment strategies for example through moving out of equity into bonds This de-risking in pension plans could in turn lead to lower returns and lower interest on the plan assets for members, which may result in increases in contributions to compensate for this (Holt, 2012).
Another often-discussed consequence is the volatility of the Other
Comprehensive Income (OCI) that is present now that the actuarial gains and losses cannot be deferred anymore. These actuarial results can vary
significantly from period to period as they include not only changes in estimates regarding employee turnover and life expectancy but also investment gains and losses and changes in the discount rates (Bloom,
2013;Holt, 2012;Deloitte, 2012). According to Severinson (2010) this volatility will lead to even more company’s offering defined contribution plans instead of defined benefit plans. Peters (2010) also expect the revised IAS 19 to lead companies to end their defined benefit schemes. But he expects this from another perspective, namely that a transparent reflection of pension risks on the balance sheet might provoke shareholders to pressurise companies to end their defined benefit schemes.
Furthermore the implementation of the revised IAS 19, more specifically the abandonment of the corridor approach, can significantly harm the solvability ratio, which may also have consequences on compliance for the finance arrangements with the bank (Deloitte, 2012; Arnold and van Santen, 2012). A considerable amount of previous literature (paragraph 2.3.3.2) exists stating that the near violation of debt covenants, of which the most common one is the debt/equity ratio, will incite companies to manage their earnings.
2.4.6 Development of Hypothesis
Thus it seems like earnings management hasn’t decreased as a result of the revision, its opportunities simply have changed in incentives. Where the previous IAS 19 gave opportunities for performing earnings management by overstating the expected returns from investment, the current IAS 19 has limited this opportunity through the use of a single discount rate. This single discount rate generally leads to a decrease in net profit and thus an incentive to use earnings management to increase this profit for bonus plan reasons. Furthermore, through introducing more volatility in the equity, and thereby increasing the chance on debt covenant violation, another incentive for earnings management is developed.
So the revised IAS is expected to lead to higher earnings management. The impact is expected to differ amongst several types of companies. The first distinctive category consists of companies using the corridor approach in 2012, for these the increase is expected to be highest since they will experience both the impact of the higher volatility in equity and thus
debt/equity ratio as well as lower profits due to the introduction of the single interest rate. The second category includes companies already using full
to have less impact on earnings management than on the first category. Yet the increase in EM is still expected to be present due to decline in net profit caused by the replacement of the interest rate used. The last category consists of companies on which the revised IAS 19 is expected to have no impact, because they either only offer defined contribution plans, multi-employer plans accounted for as defined contribution plans or no pension plans at all.
From the previous stated, the following hypotheses can be developed to help answer the research question.
1. The revised IAS 19 leads to a significant increase in earnings
management for companies previously using the corridor approach 2. The revised IAS 19 leads to a increase in earnings management for
companies already using full recognition, yet this increase will be lower than that of companies previously using the corridor
approach
Furthermore, firms with a high debt/equity ratio are more likely approaching violation thus more manipulation of earnings is expected for this category. This leads us to the last hypothesis namely:
3. The higher the debt/equity ratio the higher the earnings management
3. Research Design
In this paragraph a method is created for testing the relationship between the revised IAS 19 and earnings management.
3.1 Sample
This paper is meant to research the effect of the revised IAS 19 on earnings management and related debt covenant violation. For this research a distinction is made between companies previously using the corridor approach, already using full recognition and companies with no/limited defined benefit plans, because the effect of IAS 19 is expected to be
different for each of these categories. To find a clear difference between the effect on these categories, it is important that all are sufficiently represented in the sample. However, it is of highest significance that there are enough companies previously using the corridor approach included in the sample, since they are expected to suffer the greatest consequences.
From the research of Fasshauer et al. (2008) we can derive that, amongst others, companies in the Netherlands frequently use the corridor approach. For this reason and due to the possible language barrier of some annual reports of other countries, Netherlands has been chosen as research environment. In specific, all firms noted on the Euronext Amsterdam are included in the sample. The organisations noted on the Euronext are chosen for the sample since these firms are listed and thus have to comply with IFRS and apply IAS 19. The data for the research are collected from Compustat Global. Except for using the databases, data with respect to the pension plan and method used is collected manually.
Consistent with prior research (Burgstahler and Dichev, 1997) Financial services (GIC codes 4010-4040) and Utilities (5510) are excluded from the
sample. This leads the initial sample to consist of 200 firm years. Missing values further reduce the data to 184 firm years / 92 firms.
3.2 Time frame
As the IAS 19R is effective as per the 1st of January 2013, the first annual reports in which the use of this standard is obliged are those of 2013. The earnings management in this year is going to be compared to 2012, the year where the former IAS 19 was still valid. Next to needing data of 2012 and 2013, data of 2011 is also of importance to calculate the accruals in 2012.
3.3 Testing for debt covenants
Due to the cost of accessing actual debt covenant information, most researchers have generally used a proxy for the existence and tightness of accounting-based covenants. This is in accordance with Watts and
Zimmerman (1986) who suggest that the testing of a simple hypothesis (e.g. the debt-equity hypothesis) rather than the more theoretically correct
hypothesis (e.g. the covenant-based hypothesis) is justified in the early stages of theory development because ’ debt covenant details are costly to gather, and simple hypotheses… are one way to see whether incurring that cost is likely to pay off. ’
The most frequently used proxy is the debt-equity ratio (Deloitte, 2012). Duke and Herbert (1990) have tested the validity of the debt-equity ratio to measure closeness to debt covenant restrictions. For this they examined its relation to actual debt covenant restrictions for a random sample of U.S. firms and concluded that several versions of the debt-equity ratio capture the existence and tightness of retained earnings restrictions and the existence of
net tangible asset and working capital restrictions. This research uses total debt/total assets (TD/TA) to capture the leverage ratio. The ratios will be measured for both 2012 and 2013 to see whether this ratio has significantly increased and whether a higher ratio comes with higher earnings
management.
3.4 Testing for Earnings Management 3.4.1 Models
The most popular models for measuring earnings management by means of accruals are the DeAngelo (1986) Model, Healy (1985) Model, the Jones (1991) Model, the Industry Model (Dechow and Sloan 1991) and the Modified Jones Model (Dechow et al. 1995).
Both the DeAngelo Model and the Healy Model treat the levels or changes in working capital accruals as discretionary accrual proxies. However the
assumption implicit in these models (that non-discretionary accruals are constant) is unlikely to be empirically descriptive, because nondiscretionary accruals are expected to change with firms’ underlying business activities (Kaplan, 1985; McNichols and Wilson, 2000). The industry model and Jones model first introduce the thought that non-discretionary accruals might be variable. The industry model assumes that the changes in non-discretionary accruals are the same for companies in similar industries and the Jones model uses previous years as a benchmark for the non-discretionary accruals. The modified Jones model as introduced by Dechow, Sloan and Sweeney (1995) is derived from the standard Jones (1991) model by adjusting changes in sales for changes in receivables and is designed to reduce the
sale. Dechow et al .(1995) prove its strength by evaluating the relative performance of the previous mentioned models in detecting earnings
management. Islam et al. (2011) confirm that this last model is considered the most powerful tool in detecting earning management and thus will be used in this research.
3.4.2 The Modified Jones Model
The modified Jones model is used to estimate the discretionary accruals. This is done in 4 steps. Firstly the Total Accruals are determined by the following formula :
TACt= Δ Current Assets - Δ Cash - Δ Current Liabilities – Depreciation and Amortization Expense, where the change (Δ) is computed between time t and time t-1, in this case 2013 and 2012 (Jones, 1991).
When the total accruals are calculated, the discretionary accruals can be calculated. In order to do this, specific parameters are generated using a multiple regression:
TAC= a1 (1/ TASt-1) + a2 (Δ REV) + a3 (PPEt)+ ε, where TASt-1 stands for
total assets in year t-1, Δ REV stands for revenues in year t0 (2013) less
revenues in year t-1 (2012) scaled by total assets at t-1, PPEt stands for Gross
Property plant and equipment in year t0 scaled by total assets at t-1 and ε is the residual.
The third step is to calculate the nondiscretionary accruals, which is done using the following formula where the parameters calculated in the previous step should be entered:
NDA = a1 (1/TASt-1) + a2 (Δ REVt- ΔRECt) + a3 (PPEt), where Δ REC stands for
assets at t-1.
The final step is determining the discretionary accruals by subtracting the calculated non-discretionary accruals from the calculated total accruals.
3.5 Control variables Size
The first variable added to the regression formula is the size. This factor is added as according to, amongst others, Davidson et al. (2005) a positive relationship exists between the size of a firm and the use of earnings
management. This can be explained by the fact that most large firms receive more political attention than smaller firms, which in turn could motivate managers to adopt certain accounting techniques that reduce the reported earnings. This is in accordance with the PAT. The company size will be measured by the natural log of total assets of a firm.
Big4
From previous research can be derived that earnings management is lower in a company audited by a big 4-audit firm in comparison to a company audited by a non-big 4 audit firm; Becker et al. (1988) find that clients of non-big 4 firms report discretionary accruals that increase relatively more than the discretionary accruals reported by clients of big 4 audit firms. Francis et al. (1999) confirm this by finding that, even though Big 4 audited firms have higher levels of total accruals, these firms have a lower amount of estimated discretionary accruals. For this purpose PWC, EY, KPMG and Deloitte are called big-4 auditors. Most of the companies in the dataset are audited by these firms. Only 10 companies from the sample are audited by a
non-big-4-companies not audited by the big four, and the fact that these non-big-4 companies still are reasonably big, no significant difference in accruals is expected for the data set.
3.6 Regression formula
The previously presented variables result in the following regression formula: DAt = β0 + β1*YEAR + β2*OCIt + β3*CORRIDORt + β4*YEAR*OCI+
β5*YEAR*CORRIDOR+ β6*LEVt + β7*SIZEt + β8*BIG4 εt
Where:
DAt = absolute value of the proxy for discretionary accruals standardized by lagged total assets;
YEAR = dummy variable (YEAR=1 is 2013, YEAR=0 is 2012)
OCI = dummy variable (OCI=1 companies using full recognition
(either through OCI or P&L))
CORRIDOR = dummy variable (CORRIDOR=1 companies previously
using corridor approach)
LEVt = the ratio of the total debt at the end of the year by the total assets at the end of the year
SIZEt = LN total assets in year t;
BIG4t = dummy variable (BIG4=1 companies audited by a Big-4
auditor) ε = error, and β0, β1, β2, β3... = parameters.
The regression will also be performed on the OCI and Corridor group separately. Furthermore the hypothesis will also be tested using an adapted regression formula, namely:
DAt = β0 + β1*YEAR + β2*OCIt + β3*CORRIDORt + β4*LEVt + β5*SIZEt +
β6*BIG4 εt
4. Results
This chapter presents results of research on the collected data. 4.1 Descriptive statistics
The sample consists of 184 firm years of 92 firms listed on the Euronext Amsterdam. The firms are spread across several industries and sectors as can be seen in the following overview:
Table 1: Observations per Industry
Industry SIC Number of observations
Mining 1000-1400 8
Construction 1500-1700 10
Manufacturing 2000-3900 86
Transportation & Public Utilities 4000-4900 14 Wholesale Trade 5000-5100 8 Retail Trade 5200-5900 10 Services 7000-8900 46 Public Administration 9100-9900 2
expected to experience no effect. Of the 21 companies already applying a full recognition method, 2 processed the actuarial results through the Profit & Loss and 19 organisations recognized these results through the Other
Comprehensive Income.
The mean size, measured in total assets, of the firms included in the sample is 5131 million. According to the criteria set by the European commission (Appendix 1) this can be labelled as a large company as the assets total is higher than 43 million. However the sizes in the sample are highly spread from that mean as the standard deviation is 12762 million and the minimum and maximum size of the selected companies are 0,17 respectively 93311 million.
The mean discretionary accruals are negative (-1,125) as are both the maximum and the minimum discretionary accruals (-3,4 respectively -0,245). So in sum only income decreasing earnings management seems to be performed. The mean leverage amounts to 0,51 where the minimum and maximum seem to deviate significantly (the minimum being 0,07 and maximum 1,08) the related standard deviation is 0,18.
Table 2: Total Sample
Descriptive Statistics
N Minimum Maximum Mean Std. Deviation
LEVERAGE 184 ,06546338 1,08047422 ,5064635109 ,17656341653
ACCRUALS 184 -3,40044693 -,24494963 -1,1254849522 ,41653655736
SIZE 184 ,17 93311,00 5130,7759 12762,38430
Valid N
The difference between the mean sizes of the three distinguishable groups is also worth mentioning; Where OCI has the highest mean size of 13511 million, the mean of corridor amounts to 4447 million and the group with no effect has a petty mean of 907 million, yet still falls under the category large firms as according to the European Commission. OCI’s standard deviation is quite large (22462 million) however even the minimum size included in this group falls in the category large. This is different for the corridor as well as the no effect group where minimum included in the corridor firms can be categorized as Medium-sized and the minimum of no effect even as micro. Furthermore the mean leverage of the corridor group (0,55) stands out for being higher than that of the OCI (0,49) and no effect group (0,47). The related standard deviation is similar for all three groups as is the big spread between the minimum and maximum leverage ratio. For all groups the highest accrual still is negative. Striking is that the highest mean accrual belongs to the no-effect group (-1,28).
Table 3A: OCI Group
Descriptive Statistics
N Minimum Maximum Mean Std. Deviation
LEVERAGE 42 ,06546338 ,97056983 ,4936435501 ,18618776207
ACCRUALS 42 -1,57166437 -,31008689 -,9745453352 ,31409667865
SIZE 42 239,15 93311,00 13511,3304 22461,50572
Valid N
Table 3B: Corridor Group
Descriptive Statistics
N Minimum Maximum Mean Std. Deviation
LEVERAGE 70 ,30421556 ,97037338 ,5511964732 ,13972101083
ACCRUALS 70 -1,59243896 -,42259926 -1,0613071435 ,25493401705
SIZE 70 11,95 40941,00 4447,2045 8098,34895
Valid N
(listwise) 70
Table 3C: No Effect Group
Descriptive Statistics
N Minimum Maximum Mean Std. Deviation
LEVERAGE 72 ,09589476 1,08047422 ,4704514412 ,19487419725
ACCRUALS 72 -3,40044693 -,24494963 -1,2759281541 ,53532781583
SIZE 72 ,17 12203,37 906,7024 2152,68596
Valid N
(listwise) 72
When we compare the accruals means of 2013 as opposed to 2012 we find that this has increased (From -1,10 to -1,15). No striking changes can be noticed with regard to the leverage ratio and size.
Table 4A: 2012 Total
Descriptive Statistics
N Minimum Maximum Mean Std. Deviation
LEVERAGE 92 ,09589476 1,08047422 ,5128830341 ,17383393127 ACCRUALS 92 -3,40044693 -,31008689 -1,1014974030 ,42933419677 SIZE 92 1,00 92102,00 5160,0116 12792,06851 Valid N (listwise) 92 Table 4B: 2013 Total Descriptive Statistics
N Minimum Maximum Mean Std. Deviation
LEVERAGE 92 ,06546338 ,97056983 ,5000439877 ,17997304420
ACCRUALS 92 -2,65820119 -,24494963 -1,1494725014 ,40425329509
SIZE 92 ,17 93311,00 5101,5402 12802,65730
Valid N
(listwise) 92
If we look further into the changes we find that the accruals means as well as the leverage means is higher for the corridor group as for the OCI group in both years. For OCI the accruals have increased in 2013 as compared to 2012 whereas the leverage ratio shows a small decrease. The corridor group shows an increase in both the accruals and the leverage mean, where the lowest number is in both cases higher in 2013 as in 2012.
Table 5A: OCI Group 2012
Descriptive Statistics
N Minimum Maximum Mean Std. Deviation
LEVERAGE 21 ,21572534 ,94392649 ,5001289547 ,17419341512
ACCRUALS 21 -1,50461872 -,31008689 -,9470248667 ,33395615796
SIZE 21 5,74 11,43 8,2144 1,74094
Valid N
(listwise) 21
Table 5B: Corridor Group 2012
Descriptive Statistics
N Minimum Maximum Mean Std. Deviation
LEVERAGE 35 ,30421556 ,81341186 ,5485943659 ,13569098955
ACCRUALS 35 -1,52483908 -,42259926 -1,0349823306 ,26859003176
SIZE 35 2,58 10,57 7,0933 1,89732
Valid N
(listwise) 35
Table 5C: OCI Group 2013
Descriptive Statistics
N Minimum Maximum Mean Std. Deviation
LEVERAGE 21 ,06546338 ,97056983 ,4871581456 ,20157724764
ACCRUALS 21 -1,57166437 -,45945580 -1,0020658037 ,29854493434
SIZE 21 5,48 11,44 8,2043 1,74391
Valid N
Table 5D: Corridor Group 2013
Descriptive Statistics
N Minimum Maximum Mean Std. Deviation
LEVERAGE 35 ,31125914 ,97037338 ,5537985806 ,14557510843 ACCRUALS 35 -1,59243896 -,62315500 -1,0876319563 ,24150953465 SIZE 35 2,48 10,62 7,0265 1,86105 Valid N (listwise) 35
Next the data has been tested for potential outliers that could disturb the result. For both accruals and leverage the highest and lowest outlier has been deleted. These being three companies in the manufacturing industry, so in sum 6 firm years are removed from the sample. This leaves 89 firms, 178 firm years for the regression. A histogram in which the spread of the variables is visible is included in appendix 2.
4.2 Pearson Correlations
Table 6: Pearson Correlation Matrix Correlations
SIZE LEVERAGE ACCRUALS
SIZE Pearson Correlation 1 ,268 ** ,311** Sig. (2-tailed) ,000 ,000 N 178 178 178 LEVERAG E Pearson Correlation ,268 ** 1 -,174* Sig. (2-tailed) ,000 ,020 N 178 178 178 ACCRUA LS Pearson Correlation ,311 ** -,174* 1 Sig. (2-tailed) ,000 ,020 N 178 178 178
**. Correlation is significant at the 0.01 level (2-tailed). *. Correlation is significant at the 0.05 level (2-tailed).
The Pearson correlations matrix that is derived from SPSS clarifies that there is a significant correlation between the variables size, leverage and accruals. This is in line with previous research. There is an on-going discussion in the literature about the effect of size on leverage; some claim its negative and some positive. Empirical research can’t give a decisive answer either; Rajan and Zingales (1995) find a positive relationship for the United States, United Kingdom, Japan and Canada, but report a negative impact for Germany. The Pecking Order Theory (Myers and Majluf, 1984) support the found significant
positive relationship and explains that information asymmetries are smaller for large companies so that issuance costs of debt financing decrease making this mode of financing more attractive.
The positive significant correlation between size and accruals can be explained by political costs. As explained in paragraph 2.3.3.4 large highly profitable firms attract political attention, which can lead new taxes and other regulations. To prevent this large companies are prone to use earnings
management to lower their profits.
The last existing significant correlation is the negative relation between accruals and leverage. This can be explained by the debt covenant
hypothesis discussed in paragraph 2.3.3.2. It equates to companies using earnings management to prevent from getting fined as they get near to debt covenant violation.
4.3 Regression Findings
This paragraph shows and discusses the results of the regressions. The results regarding original and adjusted regression formula are presented
consecutively. Total
The model has an adjusted R square of 0,178, which means that the 17,8% of the variance of the dependent variable can be explained by the independent variables. Furthermore the ANOVA table tells us that the model is significant thus usable. The coefficient table implies that there is no significant
insignificance of the years can be attributed to the fact that this total sample includes a large amount of companies for which no difference is expected and the result of the big-4 variable can possibly be attributed to the lack of representation of the companies not audited by the big-4. The results with regard to the corridor and OCI companies, however, are not in line with the expectations. They namely present no significant relationship between accruals and companies using OCI and Corridor in general as well as no significant difference in 2013 as compared to 2012 for both of the categories. As a result of this H1 as well as H2 cannot be accepted.
Looking further into the results we find that they imply a significant
relationship between size and accruals and leverage and accruals as already predicted by the Pearson correlation matrix. This leads to H3 being
accepted.
Table 7: Original Regression Results Total
Model Summary Model R R Square Adjusted R Square Std. Error of the Estimate 1 ,464a ,215 ,178 ,328164670 68 a. Predictors: (Constant), SIZE, YEAR, CORRIDOR, LEVERAGE, BIG4, YEAROCI, OCI,
YEARCORRIDOR ANOVAa Model Sum of Squares df Mean Square F Sig. 1 Regressio n 4,984 8 ,623 5,785 ,000 b Residual 18,200 169 ,108 Total 23,184 177
b. Predictors: (Constant), SIZE, YEAR, CORRIDOR, LEVERAGE, BIG4, YEAROCI, OCI, YEARCORRIDOR
Coefficientsa Model Unstandardized Coefficients Standardize d Coefficients t Sig. B Std. Error Beta 1 (Constant) -1,170 ,113 -10,391 ,000 YEAR -,084 ,080 -,116 -1,056 ,292 OCI ,126 ,211 ,146 ,598 ,551 CORRIDOR ,087 ,179 ,118 ,487 ,627 YEARCORRID OR ,037 ,112 ,081 ,329 ,742 YEAROCI ,030 ,131 ,056 ,230 ,818 LEVERAGE -,626 ,157 -,285 -3,973 ,000 BIG4 ,134 ,093 ,112 1,436 ,153 SIZE ,032 ,014 ,214 2,287 ,023
a. Dependent Variable: ACCRUALS
Total
As can be seen in the results 18,7% of the variance of the dependent variable can be explained by the independent variables. Furthermore the ANOVA table tells us that this adjusted model is also significant. Compared to the results of the original regression formula there is only one striking difference, namely the positive significance of OCI and Corridor at the 0,05 level. As the constant is a negative, this result indicates that companies using defined benefit plans apply less income decreasing earnings management than companies only offering a defined contribution plan, a multi-employer plan or no pension plan at all. A possible reason for this is to compensate for the higher costs/expenses caused by a defined benefit plan. This can either be done through using less income decreasing accruals or more income
increasing earnings management. Either way it seems that the excess on rules on defined benefit plans not only make it less popular it also seems to have an impact on earnings management used.
Table 8: Adapted Regression Results Total Model Summary Mode l R R Square Adjusted R Square Std. Error of the Estimate 1 ,463a ,214 ,187 ,326353738 02 a. Predictors: (Constant), SIZE, YEAR, CORRIDOR, LEVERAGE, BIG4, OCI
ANOVAa Model Sum of Squares df Mean Square F Sig. 1 Regressio n 4,971 6 ,828 7,779 ,000 b Residual 18,213 171 ,107 Total 23,184 177
a. Dependent Variable: ACCRUALS
Coefficientsa Model Unstandardized Coefficients Standardize d Coefficients t Sig. B Std. Error Beta 1 (Constant) -1,181 ,107 -11,027 ,000 YEAR -,063 ,049 -,087 -1,284 ,201 OCI ,172 ,078 ,198 2,189 ,030 CORRIDO R ,142 ,063 ,193 2,270 ,024 LEVERAG E -,624 ,157 -,284 -3,988 ,000 BIG4 ,134 ,093 ,112 1,444 ,151 SIZE ,032 ,014 ,214 2,298 ,023
a. Dependent Variable: ACCRUALS Corridor
The regression results with regard to companies previously using the corridor approach are significant and explain the variance of the dependent variable for 15,7 %. The results suggest no significance other than the leverage ratio. This confirms that H1 cannot be accepted.
Table 9: Regression results Corridor Model Summary Mode l R R Square Adjusted R Square Std. Error of the Estimate 1 ,454a ,206 ,157 ,234095798 33 a. Predictors: (Constant), SIZE, YEAR, LEVERAGE, BIG4 ANOVAa Model Sum of Squares df Mean Square F Sig. 1 Regressio n ,922 4 ,231 4,208 ,004 b Residual 3,562 65 ,055 Total 4,484 69
a. Dependent Variable: ACCRUALS
Coefficientsa Model Unstandardized Coefficients Standardize d Coefficients t Sig. B Std. Error Beta 1 (Constant) -,821 ,157 -5,231 ,000 YEAR -,047 ,056 -,092 -,832 ,408 LEVERAG E -,839 ,213 -,460 -3,941 ,000 BIG4 ,073 ,132 ,067 ,553 ,582 SIZE ,025 ,017 ,184 1,473 ,146
a. Dependent Variable: ACCRUALS OCI
The regression result for OCI suggests that in this case the model doesn’t explain the variance of the dependent variable at all. So no further
conclusions about the effect of the revised IAS 19 on earnings management, of companies previously using the full recognition method, can be drawn. This means that the previously drawn conclusion that H2 cannot be accepted still applies.
Table 10: Regression results OCI
Model Summary Model R R Square Adjusted R Square Std. Error of the Estimate 1 ,179a ,032 -,044 ,320979735 75 a. Predictors: (Constant), SIZE, YEAR, LEVERAGE
5. Conclusion
The final part draws a conclusion, based on the performed research, with regard to the relationship between the revised IAS 19 and earnings management
5.1 Conclusion
This thesis researches the impact of the revised IAS on earnings management for companies previously using the corridor and companies already using full recognition of actuarial results. The effect on the full recognition group is expected to be lower than that of the corridor group, yet still existent. For the research data of listed Dutch companies are used. The results indicate no significant increase in earnings management for neither the corridor nor the OCI group. So H1 as well as H2 cannot be accepted and the research question cannot be answered. The results do confirm the correlation between leverage and accruals and size and accruals as found by prior research. The results also imply a rather interesting difference in earnings management between companies using defined benefit pension plans and companies using a defined contribution plan, a multi-employer plan or no pension plan at al. They namely indicate that a company offering defined benefit plans uses less income decreasing accruals in sum. Probably to compensate for the higher costs/expenses related to this plan.
5.2 Limitations
The research has some limitations that may have disrupted the outcome. Firstly the research is performed in a Dutch environment, the results might have been different if it were performed for another country. Secondly it may also be possible that the organizations decided to take a big bath in 2013/ make the earnings lower and blame it on the change of rules, and perform
income increasing earnings management as per 2014. Thirdly only accrual-based earnings management is measured; tests on real-accrual-based earnings management might have given different results. Fourthly if the sample sizes would have been larger, insignificant results could have been significant. The limitations could, however, be possible extensions for future research.
5.3 Future research
The same research could be performed in other countries and with a bigger sample to check whether the results would differ. Furthermore the earnings management in 2014 could be researched to check if companies generally have taken a big bath in 2013. Moreover, the earnings management could be checked using calculations for real earnings management. Also the effect for companies previously using full recognition through the P&L could be
researched separately. Furthermore a more profound research could be done regarding the difference between earnings management for firms using a defined benefit plan and for organizations offering a defined contribution plan, a multiple-employer plan or no pension plan at al.
