Deferred tax and earnings management



2014-2015Author: Haldun Simsek: 333732Erasmus University RotterdamErasmus School of EconomicsMaster Accounting & AuditingSupervisor: Dr. C.D. KnoopsCo-reader: Dr. Y. Gan Deferred tax and earnings managementThis master’s thesis examines whether deferred tax expense is incrementally useful in detecting earnings management beyond accruals. The period from 2010 till 2013 has been investigated for an European sample. The study is conducted using two settings. The first setting exists of a sample whereby firms engaged in earnings management to avoid an earnings decline and the second setting exists of a sample whereby firms are engaged in earnings management to avoid a loss. The results indicate deferred tax expense is incrementally useful in detecting earnings management beyond discretionary accruals for the first setting. The results for the second sample indicate deferred tax expense is incrementally useful in detecting earnings management beyond total accruals, but not beyond discretionary accruals.AcknowledgementsThis thesis is the final step to finish my master education Accounting & Auditing at the Erasmus University. I am grateful to my parents for their unconditional support. Without them it would not be possible for me to graduate. I would also like to thank my friends who always motivated me through hard times. I would also like to express my grateful thanks to my supervisor Dr. C.D. Knoops for coaching me through my thesis and my thesis co-reader Dr. Y. Gan for her contribution.Contents TOC \o "1-3" \h \z \u 1. Introduction PAGEREF _Toc425767855 \h 41.1 Background PAGEREF _Toc425767856 \h 41.2 Relevance PAGEREF _Toc425767857 \h 51.3 Research question PAGEREF _Toc425767858 \h 51.4 Methodology PAGEREF _Toc425767859 \h 52. Earnings management PAGEREF _Toc425767860 \h 72.1 Definition PAGEREF _Toc425767861 \h 72.2 Actors within the subject of earnings management PAGEREF _Toc425767862 \h 82.3 Incentives for earnings management PAGEREF _Toc425767863 \h 102.3.1 Agency theory at the basis of earnings management PAGEREF _Toc425767864 \h 102.3.2 Positive accounting theory PAGEREF _Toc425767865 \h 112.3.3 Earnings management to meet thresholds PAGEREF _Toc425767866 \h 122.4 Accrual versus real earnings management PAGEREF _Toc425767867 \h 132.5 Detecting earnings management PAGEREF _Toc425767868 \h 142.6 Summary PAGEREF _Toc425767869 \h 143. Deferred taxes PAGEREF _Toc425767870 \h 163.1 Differences between book profit and taxable profit PAGEREF _Toc425767871 \h 163.2 Deferred tax assets and deferred tax liabilities PAGEREF _Toc425767872 \h 173.3 Presentation in the financial statement PAGEREF _Toc425767873 \h 173.4 Theoretical association PAGEREF _Toc425767874 \h 184. Literature review PAGEREF _Toc425767875 \h 204.1 Book-tax conformity PAGEREF _Toc425767876 \h 204.1.1 Blaylock et al. (2014) PAGEREF _Toc425767877 \h 204.1.2 Tang (2014) PAGEREF _Toc425767878 \h 234.1.3 Coppens and Peek (2005) PAGEREF _Toc425767879 \h 244.2 Deferred tax PAGEREF _Toc425767880 \h 254.2.1 Phillips et al. (2003) PAGEREF _Toc425767881 \h 264.2.2 Phillips et al. (2004) PAGEREF _Toc425767882 \h 284.2.3 Schrand and Wong (2003) PAGEREF _Toc425767883 \h 304.2.4 Christensen et al. (2008) PAGEREF _Toc425767884 \h 304.3 Summary PAGEREF _Toc425767885 \h 315. Research design PAGEREF _Toc425767886 \h 335.1 Earnings distribution PAGEREF _Toc425767887 \h 335.2 T-test and probit-regression PAGEREF _Toc425767888 \h 355.3 Measuring accruals PAGEREF _Toc425767889 \h 375.4 Measuring deferred tax expense PAGEREF _Toc425767890 \h 385.5 Long-term debt PAGEREF _Toc425767891 \h 395.6 Industry code PAGEREF _Toc425767892 \h 395.7 Sample PAGEREF _Toc425767893 \h 395.8 Differences between this thesis and Phillips et al. (2003) PAGEREF _Toc425767894 \h 396. Results PAGEREF _Toc425767895 \h 416.1 Descriptive statistics PAGEREF _Toc425767896 \h 416.2 Assumptions PAGEREF _Toc425767897 \h 426.2.1 Normal distribution PAGEREF _Toc425767898 \h 426.2.2 Outliers PAGEREF _Toc425767899 \h 426.2.3 Multicollinearity PAGEREF _Toc425767900 \h 426.2.4 Heteroscedasticity PAGEREF _Toc425767901 \h 436.3 Earnings management model PAGEREF _Toc425767902 \h 446.3.1 Earnings management to avoid an earnings decline PAGEREF _Toc425767903 \h 446.3.2 Earnings management to avoid a loss PAGEREF _Toc425767904 \h 456.4 T-test PAGEREF _Toc425767905 \h 466.4.1 T-test for the first setting PAGEREF _Toc425767906 \h 466.4.2 T-test for the second setting PAGEREF _Toc425767907 \h 466.5 Probit-Regression PAGEREF _Toc425767908 \h 476.5.1 Sample 1 PAGEREF _Toc425767909 \h 476.5.2 Sample 2 PAGEREF _Toc425767910 \h 517. Discussion PAGEREF _Toc425767911 \h 557.1 Introduction PAGEREF _Toc425767912 \h 557.2 Findings related to earnings decline avoidance PAGEREF _Toc425767913 \h 557.3 Findings related to loss avoidance PAGEREF _Toc425767914 \h 567.4 Limitations PAGEREF _Toc425767915 \h 567.5 Summary PAGEREF _Toc425767916 \h 57Literature PAGEREF _Toc425767917 \h 58Appendix PAGEREF _Toc425767918 \h 611. Introduction1.1 BackgroundThe financial report of companies exists in order to contribute to the decision making of stakeholders. Earnings management is widely used by managers, examples of scandals are, Enron, WorldCom and A-hold. Stakeholders make their decisions based on the financial statements of companies and therefore it is important that firms present a fair and true view of their financial performance. The accounting scandals proved that firms are able to manage the earnings in the way they desire to. That is the reason why the academic world investigated a lot of research with respect to earnings management. Some researchers focus on methods to detect earnings management by using accrual models. Accruals exists in order to solve the timing problem which are earned but not recorded yet and for expenses which took place, but also not recorded in the accounts yet. Examples of methods to detect earnings management are the total accrual model and the modified Jones model. The modified Jones model separates the non-discretionary accruals form discretionary accruals, which is assumed to be used for earnings management purposes. Despite researchers` investigation on accruals to detect earnings management Guay et al. (1996) argue that accrual models contain errors. As a consequence of this assumption Phillips et al. (2003) argue that the errors of the accrual models can be reduced by not only focusing on the accrual models, but also by taking the deferred taxes into consideration. Deferred taxes arise, because companies have to prepare both financial statements and tax statements. As mentioned above, financial statements are prepared to inform stakeholders about the financial performance of the company, whereas tax statements are prepared in order to show the tax authorities the profit which forms the base to levy tax on. This thesis is built on the article of Phillips et al. (2003) which assumes that deferred taxes are incrementally useful in detecting earnings management beyond accrual measurement. The first reason why Phillips et al. (2003) believe that deferred taxes are incrementally useful in detecting earnings management is that according to the Generally Accepted Accounting Principles, the management is given more discretion compared to the tax rules and regulations (Hanlon and Shevlin, 2005). The second reason why Phillips et al. (2003) believe that deferred taxes are incrementally useful in detecting earnings management is that managers may have the incentive to increase the profit of the financial statement without increasing the profit of the tax statement. Increasing the profit of the financial statement could lead to satisfied stakeholders. The probable reason why managers do not want to increase the taxable profit is because this will lead to higher tax expenses. The increasing gap between these two statements, leads to higher deferred tax amounts on the financial balance sheet. This thesis expects that managers manage their earnings in two settings, namely earnings management to avoid an earnings decline and earnings management to avoid a loss. These two settings are similar to the settings of Burgstahler and Dichev (1997) where the researchers find that firms are likely to use earnings management to avoid an earnings decline and firms are likely to use earnings management to avoid a loss. This thesis will consider the theory of Burgstahler and Dichev (1997) to investigate whether the firms in this sample are engaged within earnings management. 1.2 RelevanceAs mentioned before, this thesis is built on Phillips et al. (2003). In other words it is a replication of the used research methodology. However, the difference is that Phillips et al. (2003) focus on U.S. listed firms, whereas this thesis focuses on European listed firms. It is interesting to see whether the results of the European listed firms are consistent with the findings of Phillips et al. (2003). There exist important differences between U.S. GAAP and IFRS regarding the deferred tax policies. The standards under U.S. GAAP are more stringent when compared to the standards under IFRS. Due to this difference, the values of assets and liabilities of firms that use U.S. GAAP are more straightforward when compared to firms that use IFRS. The values of assets and liabilities of firms that use IFRS are more subject to management’s decisions. This form of principle-based regulation provides managers of firms in Europe with more discretionary room. Managers can use this discretionary room to manage the earnings upward without influencing the profits according to the tax rules and regulations. This will lead to more deferred tax items on the financial balance sheet when compared to firms in the U.S. Since managers are given more discretionary room within the IFRS, it is interesting to investigate whether this discretionary room is used to manage earnings. 1.3 Research questionThe following research question will be answered:“Are deferred taxes incrementally useful in detecting earnings management beyond total accruals and discretionary accruals?”This question will be answered in two settings. The first setting is earnings management to avoid an earnings decline and the second setting is earnings management to avoid a loss.1.4 MethodologyThis thesis consists of three parts, starting with a theoretical framework. The theoretical framework includes the definitions and the meanings of earnings management and deferred taxes. The second part includes a literature review related to earnings management and taxes. Based on the results of the literature review the third phase will be started.The third phase includes the research part which is a replication study of Phillips et al. (2003). In order to answer the research question, data is obtained from Compustat database. The sample is gathered for European listed firms for the period 2010-2013. The sample will be divided into two groups. The first group exists of firms that used earnings management to avoid an earnings decline and the second group exists of firms that used earnings management to avoid a loss. The final sample for the first group consists of 65,536 observations and the final sample for the second group consist of 55,884 observations. For each setting the firms are divided into two categories, namely firms which used earnings management and firms which did not use earnings management. Then a t-test is included in the research part in order to control for differences in means of total accruals, discretionary accruals and deferred tax expenses between the two samples. The last part within the research design is a probit-regression in order to provide an answer to the research question. 2. Earnings managementIn this chapter the theoretical background information will be provided for earnings management in order to introduce the concepts that form the basis of this study. 2.1 DefinitionMany studies investigated the existence and the consequences of earnings management. Stolowy and Breton (2004) give earnings management another term, namely accounts manipulation. They defined earnings management as follows:“accounts manipulation is defined as the use of management’s discretion to make accounting choices or to design transactions so as to affect the possibilities of wealth transfer between the company and society (political costs), funds providers (cost of capital) or managers (compensation plans)” (Stolowy and Breton, 2004, p. 6)Within this definition the manager’s discretion is an essential aspect of earnings management. Another aspect is that there are different parties involved in the process of earnings management. Furthermore this definition, defines earnings management in a neutral way by mentioning wealth transfer. The definition mentions a possible wealth transfer between the company and three possible counterparties. In the first two cases, which are the transfers between the company and the society and the transfers between the company and the funds provides, the firm benefits from the wealth transfers. In these cases wealth transfer occurs between the firm and either the society or the funds providers. The third case relates to managers and their compensation plans which are financed by the company. In this third case earnings management is meant to influence the wealth transfer between management and the firm. Managers might manage earnings in order to present a higher net income and thereby obtain a financial reward. Managing earnings is rarely in the interest of the firm. The first two cases in which management tries to influence the wealth transfer between society and funds providers, the managers are more likely to act in the interest of the firm. An example hereby is when managers inflate earnings in order to negotiate a more beneficial debt covenant. It is therefore important to know that in this definition earnings management can bring both positive and negative effects for the firm. Healy and Wahlen (1999) provide another definition of earnings management in which they depart from the neutral stand on earnings management. They define earnings management as follows:“Earnings management occurs when managers use judgment 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” (Healy and Wahlen, 1999, p. 368).In contrast to the definition of Stolowy and Breton (2004) this definition mentions “misleading stakeholders” as an objective for doing earnings management. However a similarity between this definition and the pervious one is that both mention management’s discretion in earnings management activities. The definition of earnings management that would be used in this thesis will consider aspects of both definitions. These aspects include taking into consideration that earnings management could be either positive or negative from the perspective of the firm and that managers use their discretion to make accounting choices. Eventually these aspects of earnings management are used by the management in order to mislead some stakeholders about the economic performance of the firm or to influence the contractual outcomes. In particular, different studies define earnings management diversely. This thesis, as mentioned above, defines earnings management as follows:“Earnings management occurs when managers use their discretionary room in financial reporting. The subjectivity embedded in the discretionary room could allow managers to act either positive or negative for the firm. The objective of doing earnings management is to mislead stakeholders or, in order to influence the contractual outcomes.” In this definition both a possible positive and negative outcome of earnings management are included. In this section the definitions of earnings management were discussed. The next chapter focuses on the actors that are possibly participating in earnings management or affected by it.2.2 Actors within the subject of earnings managementIn the previous section various definitions of earnings management have been discussed. This section introduces the main actors within earnings management. The main actors with respect to earnings management are; the managers, regulators, auditors, and stakeholders.For management earnings management is a way to influence the accounting numbers. Managers are hired by shareholders to run the firm. In other words, shareholders are the owners of the firm and they want managers to run the firm in the interest of the shareholders. In practice, managers use the flexibility both within and outside the regulatory framework to manage accounts in certain ways. It depends partly on the managers’ incentives in which way they use earnings management. These incentives to engage in earnings management can be various for managers. These incentives will be more thoroughly discussed in section 2.3.The second possible actor in the context of earnings management are, the regulators. Regulators propose rules in their regulatory effort to prevent or at least to minimize the use of earnings management. The purpose of rules and regulations is that they are designed in a way that they direct the financial reports to provide a true and fair view about the firm’s financial position and performance. There are several organizations that set up these rules in order to be followed by firms. An example of a regulatory board is the International Accounting Standards Board (IASB). The task of the IASB is to issue international standards for financial reporting, which is called the International Financial Reporting Standards (IFRS). These reporting standards contribute to financial reports that provide a true and fair view of the firm. However besides reporting standards, the auditing profession is well-regulated in order to have them contribute to achieving true and fair financial reports. This audit regulation comes from auditing standard boards of which the International Auditing and Assurance Standards Board (IAASB) is one of them. The auditing standards contribute in the process of true and fair reporting. Finally besides these auditing standards also auditing principles exist for the auditor, these are: integrity, objectivity, professional competence and due care, confidentiality and professional behaviour (Hayes et al. 2014). An auditor is required to remain objective in the audit engagement. Also integrity is essential for the auditor. Integrity relates to the auditor being honest and straightforward in his relation to the audit client. Confidentiality is also of importance for the auditor since the auditor is closely involved with his audit clients. This involvement allows the auditor to become knowledgeable about the firm and its operations. The confidentiality principle requires the auditor to treat the acquired knowledge of a client firm with a high level of confidentiality (Hayes et al., 2014).Professional competence and due care puts a requirement on the level competences of the auditor. In case of the auditor lacks professional competence, he is not able to audit a certain firm. Therefore an auditor is required to possess professional competences in order to successfully audit a financial report and optimally contribute in representing the firm in a true and fair view. Finally the principle of professional behaviour requires the auditor to perform his job in accordance with regulation and should avoid bringing discredit to the auditing profession (Hayes et al. 2014).These six auditing principles are essential for the auditor in his role as an auditor. This role is to provide reasonable assurance that the financial statements provide a true and fair view. A financial statement that obtains reasonable assurance from the auditor is assumed to better aid financial statement users in their decision making. Therefore an auditor makes it a priority to present a financial statement which is free from inflated earnings, for example, which could be due to earnings management.Finally the last group of actors which will be discussed in relation to earnings management are the stakeholders. The stakeholder group consists of a wide variety of stakeholders which all have a stake in specific firms. These can be debt holders, shareholders, employees, the government et cetera. Each of these stakeholders make different decisions in relation to the firm and have different interests, such as employees are best served with an expansion in their compensation plans while debt holders assign great importance to the firm’s liquidity and solvency. Although interests vary among the different stakeholder groups the information need is identical. This information need is mainly met with the financial statements. Stakeholders derive part of their information on the firm, from the financial statements. Therefore it is highly required that these financial statements are free from any bias due to earnings management and represents the firm in a true and fair view.This section discussed the different actors that are either participating or affected in relation to earnings management. The next section focuses specifically on one actor which has the ability and possibly the incentive to do earnings management, which is the management itself.2.3 Incentives for earnings managementFirms are expected to present good performance. Management is charged with the responsibility to lead the firm towards high performance and meeting stakeholders’ expectations, especially shareholders’ expectations. Unfortunately, not always is the case that firms perform in line with expectations, therefore managers might try to change the numbers in a way that makes firm performance meet expectations. 2.3.1 Agency theory at the basis of earnings managementThe agency relationship can be described as one whereby one group engages another to serve on their behalf. In this relation the first group are the principals and the second are the agents. In section 2.2 it was discussed that managers, regulators, auditors and shareholders are the main groups that are either participating or affected by the presence of earnings management. The agency theory describes in more detail the conflicting interests of managers and shareholders. The agency theory is based upon an engagement between agent and principal. Within this engagement the agent receives, to some extent, decision making authority from the principal (Hill and Jones, 2001). When reflecting the setting of the Agency Theory to a firm, one could see that management takes the role of the agent and shareholders take the role of the principal. The cornerstone within this theory is the assumption that interests of principals and agents differ. In fact there is an agreement between shareholders and the management whereby the management should run the company in the interest of the shareholders. Despite this agreement, the risk of management to focus solely on its own interest remains. In order for the shareholders to reduce this risk they can invest in controlling and monitoring activities. Although management is hired by shareholders they might become more knowledgeable about the firm than the shareholders themselves. This is due to the fact that management is closely involved with the firm while shareholders are less closely involved. Because of this asymmetry in involvement an asymmetry in information can evolve. This is qualified as information asymmetry between shareholders and management. Before, the conflicting interests between management and shareholders were discussed as a reason for why management could deviate from serving shareholders’ interests. Information asymmetry could support management in their act of deviation since this asymmetry helps management hide their actions. The two main types of problems that arise from information asymmetry are the moral hazard problem and the adverse selection problem. First a moral hazard problem arises when external investors are not able to see and to observe the actions and the choices that managers have made (Walker, 2013). Second, since managers have access to superior information an adverse selection problem arises. Shareholders do not have access to this information and therefore are not able to judge with the same judgement as management does. Moral hazard relates mainly to hidden actions while adverse selection is mainly related to hidden information between two parties (Eisenhardt, 1989).The agency theory describes the setting that possibly gives rise to earnings management. The Positive Accounting Theory focuses more deeply on this earnings management by discussing possible motives for management to do earnings management. In the next section these motives will be discussed.2.3.2 Positive accounting theoryThe Positive Accounting Theory (PAT) assumes the firm to be a nexus of contracts in which the contract between management and shareholders is one of them. Also contracts with debt holders, external suppliers, customers et cetera are included in PAT’s concept of the firm. Furthermore it assumes that management acts opportunistically or in order to allow for efficient contracting. These different ways for the management to act are included by either the efficiency perspective or the opportunistic perspective. In the latter one, managers are assumed to strive for presenting the true performance of the firm. While in the opportunistic perspective, managers are assumed to focus mainly on maximizing their own utility (Scott, 2012). Dechow (1994) tested the two versions of the Positive Accounting Theory. Her reasoning was that if accruals are the results of opportunistic manipulation of reported earnings, the market will “punish” them in favour of cash flows. This means that cash flows should be more strongly associated with share returns than net income. The other case is that when accruals reflect efficient contracting, net income should be more strongly associated with share returns compared to the cash flows. She finds that the net income is more strongly associated with share returns compared to cash flows. In relation to this opportunistic perspective the Positive Accounting Theory developed three hypotheses, which are: the bonus plan hypothesis, the debt/equity hypothesis, and the political cost hypothesis (Watts and Zimmerman, 1986). The bonus plan hypothesis discusses the use of earnings management when bonuses are implemented in the salary payments of the managers. It assumes that managers of firms with bonus systems have incentives to maximize earnings in case bonuses are related to the level of earnings, this is in order to maximize bonus payoff for themselves (Watts and Zimmerman, 1986).The debt/equity hypothesis assumes that managers attempt to choose accounting methods that increase the earnings in cases when the debt/equity ratio is high. The reason behind this is that creditors of firms impose restrictions for dividend payments, share repurchases and the issuance of additional debtors. Therefor managers have incentives to engage in earnings management in order to satisfy the conditions or keep within the debt covenants (Watts and Zimmerman, 1986). The political cost hypothesis examines the role of accounting choices with respect to political issues. Since larger firms attract more political attention than smaller firms managers choose to reduce the earnings. The three hypotheses that are discussed above are stated in opportunistic form. That means that Watts and Zimmerman (1986) assume that managers choose accounting policies in such a way that they maximize their own expected utility.2.3.3 Earnings management to meet thresholdsIn the previous section various reasons for management to engage in earnings management are mentioned. This thesis however focuses on avoiding a possible earnings decline or loss as a reason for management to perform earnings management. This is based on the study of Burgstahler and Dichev (1997). According to Burgstahler and Dichev (1997) managers have strong incentives to engage in earnings management in order to avoid reporting an earnings decline, to avoid reporting a loss or to meet or beat the earnings forecasts of analysts. Managers want to avoid reporting an earnings decline, because they believe that when firms have a decline in earnings compared to the previous year it could attract a lot of negative public attention. Which could lead to a decrease of investor appreciation for the firm (Burgstahler and Dichev, 1997).The reason why managers want to meet or beat analyst forecasts is that firms will receive a market premium. Share prices rise for firms that meet or exceed the earnings expectations of analysts (Bartov et al., 2002). Since managers` bonuses are tied to stock price performance, managers are even more interested to meet or beat to analysts’ forecasts (Lee, 2007).Finally managers are in general also eager to avoid a loss. The theory behind this managerial behaviour may be explained by the Prospect theory of Tversky and Kahneman (1991). This theory predicts that individuals are risk seeking in a domain of losses whereas individuals are risk averse in a domain of gains. This means that people have different values for gains and losses. Therefore individuals base their decisions on gaining or losing. This concept, also known as the concept of loss aversion, refers to the tendency of people to strongly prefer avoiding losses over obtaining gains. The prospect theory has a relation with earnings management such that managers could have incentives to avoid losses. It is therefore possible that earnings management is used in order to avoid losses. In case the firm does not report a loss this may be perceived as good performance and managers may be awarded. Another reason to avoid losses for managers is that stock prices could react negatively to losses. The prospect theory therefore could be an interesting theory in explaining why managers engage in earnings management to avoid losses. This section provided an additional reason for managers to do earnings management which is based on the study of Burgstahler and Dichev (1997). The next section discusses the ways for management to actually do earnings management.2.4 Accrual versus real earnings managementManagers can manage earnings by involving accrual based and real transaction based earnings management. The earnings of a firm are determined by the following equation: “Earnings = Total Accruals + Cash Flow”This equation shows that earnings exist of two components. The first component in the equation is total accruals. Managers are able to do earnings management through accruals by manipulating the accruals whereas real earnings management is used to manipulate the second component in the equation namely the cash flows. Accruals are the difference between the net income and cash flows and can be divided into discretionary accruals and non-discretionary accruals. Since earnings management can be done by using both total accruals and discretionary accruals these accounts are often used as a proxy for earnings management.Despite the fact that accruals are susceptible to earnings management, the main objective of accruals is to show the true economic performance of the firm through recognizing revenues and expenses for the period in which they are incurred instead of only presenting the cash inflows and cash outflows. The non-discretionary accruals arise from normal transactions and cannot be manipulated by the management. However, as mentioned before, with discretionary accruals management is provided some discretionary room. The regulatory framework provides this discretion in order to allow management to share their inside-information with stakeholders. This allows management to reduce the information-asymmetry between the owners and management (Palepu, Healy, Bernard & Peek, 2010).The second way of earnings management is real earnings management. Managers also have the ability to manipulate the earnings through operating, investing and financing activities (Xu et al., 2007). Li describes real earnings management as structuring the actual business activities to achieve a desired earnings result. An example is that the management sells equipment on a specific date that will translate in a gain within a period in which the earnings are needed. Managers can choose between accrual earnings management and real earnings management, but it is also possible for a manager to do both at the same time. However besides multiple ways to do earnings management also multiple ways to detect earnings management exist. The next section discusses these ways individually.2.5 Detecting earnings managementMost of the time earnings management is detected through accruals. Within the accruals approach to detect earnings management, there are two options available. The first approach is the single accrual approach which measures earnings management by focusing on one accounting item. The second approach is the total accrual model approach which focusses on multiple accounting items. This thesis will not consider the single accrual approach, but will focus on the total accruals approach. As mentioned in section 2.4 total accruals can be divided into discretionary accruals and non-discretionary accruals. In accounting research often discretionary accruals is used as a proxy for earnings management. This is done because discretionary accruals are influenced by managers. The estimation of discretionary accruals consists of three steps. The first step involves capturing the total accruals. The second step is to calculate the non-discretionary accruals by time-series or a cross-sectional analysis. Finally the non-discretionary accruals need to be subtracted from the total accruals so that the discretionary accruals are calculated (Dechow et al., 1995).There are a lot of models that try to capture the non-discretionary part of the accruals. These models include the following models: Healy model, DeAngelo model, Jones model, Modified jones model, Industry model, the Forward-Looking model and the Performance-Matching model (Ronen and Yaari, 2008) This thesis uses total accruals and discretionary accruals as a proxy for earnings management. In estimating discretionary accruals the modified jones model is used. This model will be further discussed in chapter five.2.6 SummaryIn the first section the definition of earnings management is discussed. Earnings management occurs when managers use their discretionary room in financial reporting. It was found that subjectivity is critical in order to perform earnings management. The managers hope by engaging in earnings management to mislead stakeholders or they hope that they can influence the contractual outcomes.The actors that are involved with earnings management are managers, regulators, auditors and stakeholders. Next, various incentives for engaging in earnings management have been discussed. These are: Agency theory and the Positive accounting theory. The positive accounting theory could exist in two ways; the opportunistic and the efficient contracting. The agency theory is about the fact that managers run the company on behalf of the shareholders, but each group has different incentives. According to Burgstahler and Dichev (1997) managers engage in earnings management to avoid an earnings decrease, to meet the expectations of the analysts and to avoid losses. Finally, it is mentioned that earnings consists of two components, namely cash flow and accruals. Real earnings management is done through cash flow whereas accrual earnings management is done through accruals. In the next chapter theoretical information will be provided about deferred tax assets and deferred tax liabilities3. Deferred taxesTaxes are levied by governments, for this net income acts as the base of calculation. Each country has its own rules and regulations with respect to the percentage payable tax and the definition of taxable income; however the idea of collecting taxes is conceptually the same across different countries. In this chapter the different accounting and tax concept of profits are discussed. After the deferred tax item both the deferred tax asset and the deferred tax liability are discussed. This chapter concludes with the research article of Hanlon and Shevlin (2005) in which the theoretical association between deferred tax items and earnings management is presented.3.1 Differences between book profit and taxable profitAn important element of firm performance is reported profit. In general the book profit is reported more extensively. However another way of measuring profit exists. This is the tax profit. With book profit, profit is measured following accounting methods which includes the method of accrual accounting. Accrual accounting allows management, when certain requirements are met, to recognize revenues for which a cash flow is not yet received. These revenues become accruals in the financial statement. A large part of the accruals however are not part of the tax profit since these revenues are not received by the firm yet and therefore do not provide a tax base (Picker et al. 2013). With tax profit, the entire amount of the profit forms the tax base. Therefore International Accounting Standard (IAS) 12 paragraph 5 defines tax profit as the profit or loss which is obtained over a certain period, whereby the net income forms the tax base. As is discussed due to the accounting policy, differences may arise between the tax profit and the book profit. These differences however are of temporary kind. In the long run a cash flow related to a specific accrual is received and the tax profit is corrected, which leads to the difference to disappear (Picker et al. 2013). Besides temporary differences also permanent differences can exist between the book profit and tax profit. A difference between book and tax profit becomes permanent when such a difference does not disappear in the long run. This can be due to a transaction which is recognized in the accounting system but remains unrecognized in the tax system of profit calculation. Permanent differences can arise from prohibited deductions. A prohibited deduction is, for example, a fine which the company is obligated to pay. This fine can be presented as an expense in the accounting method of profit calculation, in the calculation of taxable income however, this fine remains out of the calculation (Picker et al., 2013).For differences between book and taxable profit a deferred tax asset or liability arises in the financial statements. This deferred tax item is formed exclusively for temporary differences since these differences will reverse in the long run. A deferred tax item informs financial statement users about a tax payment or receipt in the future which is related to a current difference between accounting and taxable profit. In this section the deferred tax item is introduced, the next section will discuss the two forms of the deferred tax items which are the deferred tax asset and the deferred tax liability.3.2 Deferred tax assets and deferred tax liabilitiesDue to differences between book and taxable profit deferred tax items can arise. This item can be either an asset or a liability. IAS 12 defines deferred tax assets as “the amounts of income taxes recoverable in future periods”. A deferred tax asset in general arises when taxable profit is higher than book profit, additionally such a deferred tax asset can also be the consequence of unused tax losses or tax credits which are carried forward to future periods. Also according to IAS 12, Deferred tax liabilities, in contrast to deferred tax asset, reflect the amount of income taxes that is payable in future periods. A deferred tax liability can arise when an instalment sale receivable is obtained by the firm. An instalment sale receivable allows the creditor to divide payment over multiple years. Due to this payment division also the tax payable is divided over multiple years. In this case a deferred tax liability is formed in order to present the fact that the firm is obliged to pay more tax in the future. Figure 1 provides an overview of the concept and occurrence of the deferred tax item.Figure1: Scheme of deferred taxes, Picker et al. (2013)3.3 Presentation in the financial statementIn relation to deferred tax items IAS 12.74 determines in which cases a firm is allowed to recognize a deferred tax item. According to IAS 12.74 a firm is allowed to recognize a deferred tax asset or a deferred tax liability in its financial statement when two requirements are met. The first is about whether the firm is allowed to set off current tax assets against current tax liabilities. The second requirement is that the both tax items are payable to the same tax authority. If both requirements are met as a net position, i.e. the deferred tax liability is subtracted from the deferred tax asset, the net deferred tax asset or liability remains.3.4 Theoretical associationIn chapter one it was mentioned that a company is able to manage their earnings by using accruals. In this chapter it is argued that for determining taxable profit and book profit different methods are used. The tax law is generally close to cash base accounting since tax authorities are more concerned with cash, whereas the accrual method is used for determining the book profit. Managing of earnings through accruals therefore has limited effect on the determination of taxable profit since this is generally more based on cash flow. By using accruals in order to manage earnings, temporary differences arise between the book profit and the taxable profit which leads to deferred tax assets or deferred tax liabilities. Therefore the presence of a deferred tax item and the divergence of book profit and taxable profit can indicate the presence of earnings management (Hanlon and Shevlin, 2005).Hanlon and Shevlin (2005) discuss this relation between deferred tax and earnings management. They discuss the issues related to the proposal to conform the book and taxable income. Hanlon and Shevlin (2005) provide a review of the recent literature with respect to the issues of conforming the financial book incomes and taxable incomes. The first argument for studying this topic is that there is evidence of the expanding divergence between these two incomes. According to Hanlon and Shevlin (2005), the problem is not that book and tax incomes are different per se, but that the cause of the divergence is an unanswered question. One possible explanation for the expanding divergence is that management does misleading activities or even fraudulent activity of which earnings management possibly is a part.In Hanlon and Shevlin (2005) there are proposals taken into consideration to conform both incomes. The conforming of book- and taxable incomes could be positive or negative. The potential benefits of conforming book- and taxable income are that the incentive of executives to mislead shareholders about their economic performance would be significantly reduced or even eliminated (Yin, 2001; Desai & Darmaphala, 2004). Book-tax conformity means the level of conformation between the tax reporting and the financial reporting. The reasoning behind this is that executives will not overstate the book income since a higher book profit immediately leads to a higher taxable income and consequently higher taxes payable since taxes are levied on this tax profit. Furthermore executives are also assumed not to understate the book income since this would mean lower results reported to creditors and shareholders. The second argument in favour of conforming the book income and the taxable income is that it will decrease the compliance cost. Under the proposed system firms will report the same measure of income to both the tax authorities and the shareholders. However there are also potential costs of conforming both incomes. The most important cost is the loss of information to the financial markets. In the literature there are a lot of studies that prove the usefulness of financial accounting earnings and earnings announcements (Ball and Brown, 1968). This means that the financial report adds value-relevant information which is received by the capital market. In this chapter the occurrence of deferred tax items and its relation to earnings management is discussed. The next chapter contains an empirical literature review about the association between deferred tax and earnings management.3.5 SummaryThis chapter discussed the underlying theory of the deferred tax assets and the deferred tax liabilities. A deferred tax asset arises when taxable profit is higher than book profit, whereas a deferred tax liability arises when taxable profit is lower than book profit. The differences in tax profit and book profit could be temporary or permanent. Temporary differences are shown on the financial statements of firms. Finally, the theoretical association between book profit and tax profit have been discussed. It was found that the tax law is generally close to cash base accounting since tax authorities are more concerned with cash, whereas the accrual method is used for determining the book profit. In the following chapter, the existing literature will be analysed. 4. Literature reviewThere is much written on earnings management and its possible association with taxes. In this section, previous research on this is discussed. The literature in this chapter can be divided into two main groups. The groups exist of different proxies for the alignment between the book-tax conformity. The first group relates to possible ways to calculate the ratio of the alignment between the book-tax conformity, whereas the second group takes deferred tax as a proxy for book-tax conformity instead of a ratio, which is stated on the financial balance sheet. Deferred taxes arise due to different rules and regulations for the book profit and for the tax profit. The differences in rules and regulations lead to different presented profits for both book profit and taxable profit and this difference is expressed as a deferred tax item on the financial balance sheet. Figure 2 below shows the subdivision of the proxies in this chapter. This thesis focuses solely on deferred taxes as a proxy for the book-tax conformity. In this chapter also the hypotheses development will be included. The hypotheses will be presented with the related literature component.Figure 2: Subdivision of the proxies4.1 Book-tax conformityThis section discusses the papers which focus on calculating the ratio of the alignment between the book-tax profits. The articles find different results for the association between earnings management and book-tax conformity. Below these studies are discussed.4.1.1 Blaylock et al. (2014)Blaylock (2014) performs an archival study in order to investigate the association between book-tax conformity and earnings management. Blaylock et al. (2014) mentions that prior literature reports a negative association between earnings informativeness and book-tax conformity (Hanlon et al., 2005; Atwood et al., 2010) and they mention that none of these prior articles attributes this negative relation to earnings management. Blaylock et al. (2014) proceed to research a possible attribution of this negative relation to earnings management.They state that there is a discussion going on whether it is required to increase the book-tax conformity in order to reduce earnings management. Blaylock et al. (2014) evaluated probable arguments for lower ratio of earnings management by higher book-tax conformity and probable arguments for higher ratio of earnings management when the conformity between the book profit and taxable profit is increased. Arguments in favour of increasing the book-tax conformity are that it probably leads to less earnings management, because the increased conformity will lower the managerial discretion over financial reporting. However, Blaylock et al. (2014) assume that the tax rules are used as the basis for reported income, because of political pressure. This is important to know, because tax rules are assumed to allow less managerial discretion. The second reason that there is a chance of mitigating incentives to manage earnings is as follows. In chapter 2.4 it is mentioned that earnings management is executed through accruals while tax rules generally focuses on cash. Blaylock et al. (2014) mention that under accounting reporting, revenues are recognized when earned and expenses are matched to the revenues to solve for timing problems. In contradiction to the tax law which is generally close to cash base accounting since tax authorities are more concerned with cash. This means that high alignment between book-tax profit, reduces the incentives of managers to engage in earnings management since management becomes conscious of the fact that every dollar of inflated earnings contributes to the tax base on which the tax payable is calculated.Opponents of high book-tax conformity argue that conformity will lead to a significant loss of financial information, since the users of financial statements differ from the users of the tax report, which are often tax authorities. It is an information loss for the users of financial statements, because they lose an additional measure of performance, namely the taxable income, to detect earnings management. Multiple reasons are stated by Blaylock in support of both a positive and a negative relation between book-tax conformity and earnings management. In order to perform their research objective Blaylock et al. (2014) develop the following hypothesis: “H1: Book-tax conformity is associated with earnings management’’ (Blaylock et al., 2014, p. 150). The researchers state their hypothesis in such a way to reflect their expectations on the relation between book-tax conformity and earnings management.The sample exists of U.S. listed firms for the period 1996 to 2007 whereby Compustat is used to gather the data from.Four different measures of earnings management are used in order to widely capture earnings management.The first measure is the extent to which earnings are stable, in other words the extent to which earnings smoothing is present. The second measure also focuses on earnings smoothing by assessing changes in accruals and changes in operating cash flows. The third measure relates to the magnitude of accruals which is a measure for the use of accounting discretion. The fourth measure focusses on measuring earnings management by assessing the degree to which firms avoid reporting small financial losses.The results show that higher book-tax conformity is associated with a smoother earnings management (measurement 1). There is a positive relation between book-tax conformity and a smoother earnings management (measurement 2). The results also show a positive correlation between the magnitude of accruals and book-tax conformity (measurement 3). This finding suggests that conformity leads to more use of accounting discretion. For the last measurement of earnings management it is found that there is a lower probability of reporting a small loss in high conformity regimes. This finding suggests that there is more earnings management in high conformity regimes, which is consistent with the notion that when there is higher book-tax conformity there is a greater use of discretion in financial reporting. The researchers find that the correlation between book-tax conformity and the different measures of earnings management is positive and significant for all measures at a 0.01 level. Which indicates higher levels of earnings management in situations of higher book-tax conformity.Concluding it can be said, based on the results that a positive association exists between earnings management and book-tax conformity. Wherein earnings management is measured on four dimensions. In order to assess the robustness of these findings control variables need to be added. These control variables are known to be influencing the magnitude of earnings management. They consist of the following factors: “outside investor rights, legal enforcement, ownership concentration, legal tradition, firm size, the percentage of manufacturing firms, gross domestic product (GDP), and inflation” (Blaylock et al., 2014, p. 143). After including these factors the results are found to be persisting. These results are a positive association between earnings management and book-tax conformity.4.1.2 Tang (2014)Similar to Blaylock et al. (2014) Tang (2014) focusses on the relation between the book-tax conformity and opportunistically reporting of taxable income and book profits. This archival study specifically investigates whether higher levels of book-tax conformity are associated with lower levels of earnings management and tax avoidance activities.Proponents of higher book-tax conformity argue that book-tax divergence creates increased opportunities for managers to aggressively report financial statements and taxable statements simultaneously. The reasons for less earnings management provided by Tang is that when book and tax profits were to converge, overstating of book profits would mean that the firm has to pay more income taxes whereas understating of taxable profits would imply that a lower book profit is reported to stakeholders. Another argument used by Tang (2014) is similar to Blaylock et al. (2014), namely increasing conformity will probably limit the managerial discretion over the financial reporting and as a consequence of less managerial discretion earnings management will be lower. Because this article focusses on both earnings management and tax avoidance, two hypotheses are developed. These hypotheses are: “H1: There is no relation between earnings management and book-tax conformity” and “H2: There is no relation between tax avoidance and book-tax conformity” (Tang, 2014, p. 6).In order to investigate book-tax conformity Tang deploys a proxy which has not been used previously. This proxy is based on book-tax differences, since book-tax differences can indicate the level of conformity. Wherein high book-tax differences indicates a lower level of conformity. However book-tax differences are not driven solely by legitimate differences between accounting and tax rules. Manipulations made in book and tax income also drives book-tax difference, these manipulations can be a consequence of opportunistic managerial behaviour. Hence, the book-tax difference which is not explained by legitimate difference in book and tax income should be filtered out. In order to do so, the author regresses total book-tax difference on two forms of possible opportunistic behaviour. These are earnings management and tax avoidance. Earnings management is measured as the level of discretionary accruals and tax avoidance by assessing the difference between the effective tax rate and the related statutory tax rate.The proxies used to measure earnings management consist of discretionary accruals, discretionary revenue, and discretionary current accruals.In order to investigate the previously mentioned hypotheses, a regression analysis is constructed based on a sample which is obtained from the Compustat Global Industrial/Commercial files. The sample is taken from the year 1995 till the year 2007 and yields 372 country-year observations across 32 countries. The main results obtained in this analysis suggest that conformity deters earnings management and tax avoidance. In contrast to Blaylock et al. (2014) the regression results present a negative relation between conformity and earnings management which is significant at the 0.01 level. It also presents a negative association between book-tax conformity and tax avoidance. Based on these findings it can be concluded that higher levels of conformity do deter managers from engaging in earnings management and tax avoidance. The reasons why managers will decrease the engagement in earnings management relies on the argument that the incentives of managers will reduce to act in this way and the second reason is that higher book-tax conformity will reduce the discretion of managers over financial reporting. The reason why firms engage less in tax avoidance is that high conformity between book profit and tax profit provides fewer opportunities to avoid taxes and the second reason is that tax authority’s act as an extra monitor to control reported earnings. The results from the study of Blaylock et al. (2014) clearly contradict the results obtained in this research. Since no major differences exist in the sample collection and the sample period, it might be more relevant to focus on the research design. When observing the research design, one point of difference clearly steps out. This is the measurement of the book-tax conformity. Blaylock et al. (2014) use a model wherein book-tax conformity is estimated by looking at the amount of variation in current tax expense that is not explained by pre-tax earnings.Tang constructed a new conformity index to measure conformity by adjusting for the opportunistic managerial effect. Tang argues that this index better reflects the extent to which the book income differs from taxable income. The conclusion is that this paper provides evidence when there is high book-tax conformity it deters management from earnings management and tax avoidance. This difference between both measurement proxies might be a very important one which explains the difference in results obtained in both studies. There are also studies available whereby the researchers use deferred tax as a proxy for book-tax differences in order to investigate the association between book-tax differences and earnings management (Philips et al., 2003; Phillips et al. 2004; Schrand and Wong, 2003; Christensen et al. 2008). These studies will be discussed in the section 4.2. In chapter three is mentioned that temporary differences between book profit and tax profit leads to deferred taxes on the commercial balance sheet.4.1.3 Coppens and Peek (2005)Coppens and Peek (2005) investigate whether private firms engage in earnings management in eight European countries and whether tax incentives affect this behaviour. This study examines the shape of the earnings distribution of private firms by comparing this with the distribution of public firms. In this study private firms are firms which are not listed on a public stock exchange, public firms are firms which are represented on a public stock exchange. They examine large private firms in Belgium, Denmark, France, Germany, Italy, the Netherlands, Spain and the U.K. The researchers expect in countries where financial accounting and tax accounting practices are strongly aligned, that firms will not be able to avoid reporting losses and earnings decreases. In other words, this study assumes that situations of high levels of book-tax conformity will be associated with lower levels of earnings management. Coppens and Peek (2005) test their hypotheses for high alignment countries and low alignment countries separately. The sample for private firms comes from Amadeus database which contains data for EU member states. In this database data availability is highest for Belgium, Denmark, France, Germany, Italy, the Netherlands, Spain, and the U.K. for the time period between 1993 and 1999. The sample for public firms, against which the researchers benchmark the private firms, is from Worldscope database. The research design of this study is centred around the earning distributions, because the researchers believe that analysing the distributional properties of earnings is a strong way to detect earnings management. Coppens and Peek (2005) do not use the accruals method, since they believe that this method is not feasible in their research setting since not all data is available. Two hypotheses are developed in order to study whether private firms engage in earnings management. These are: H1: private firms do not manage earnings to avoid reporting losses. H2: Private firms do not manage earnings to avoid reporting earnings decreases. The results show that private firms manage earnings more often to avoid losses in both high alignment countries and in low alignment countries in comparison to public firms. However they fail to reject the second hypothesis, namely that private firms manage earnings to avoid earnings decreases. The difference in results between private and public firms can be attributed to the fact that public firms are possibly more exposed to analyst following. This difference in exposure could lead private firms to engage in more earnings management.4.2 Deferred taxThis section will discusses the articles which use the deferred tax item as a proxy for the alignment between the book-tax profits. The first article to be discussed is Phillips et al. (2003) which finds that deferred tax expense is incrementally useful in detecting earnings management. This article is followed by Phillips et al. (2004) which decomposes deferred taxes into several components in order to see which of these components are significantly associated with earnings management. The articles after these, focus in more detail to the deferred tax assets.4.2.1 Phillips et al. (2003)In this study Phillips et al. (2003) investigate the usefulness of deferred tax expense in detecting earnings management. This article examines the usefulness of deferred tax expense for the use of earnings management in order to meet three earnings objectives. The first objective is that managers want to avoid reporting an earnings decline, the second objective is to avoid reporting a loss and the third objective is to meet the analysts` forecasts. Phillips et al. (2003) argue that the measurement error in accrual methods can be reduced by focusing on deferred tax expense instead of attempting to decompose accruals into normal and abnormal components. “The deferred tax expense is a component of a firm`s total income tax expense and reflects the tax effects of temporary differences between book income and taxable income that arise primarily from accruals for revenue and expense items that affect both book and taxable income, but in different periods” (Phillips et al., 2003, p. 492). Deferred tax expense is a better measure for the managers` discretionary choices under GAAP, since the tax law allows less discretion when compared to GAAP. Furthermore managers want to use earnings management to achieve specific targets, they prefer to manage book profit upward without increasing the tax profit. Researchers often conducted research on accruals to detect earnings management compared to cash flow since management discretion over accruals generally is less observable. Phillips et al. (2003) argue that book-tax differences will aid in preventing earnings management by distinguishing between managers` discretionary choices from non-discretionary choices. In this article the proxy which is used for the book-tax difference is the deferred tax expense. Temporary differences create deferred tax assets or deferred tax liabilities whereby an increase in the deferred tax liability is consistent with an increase in the deferred tax expense. In other words, in this article with deferred tax expense it is referred to the part in which the deferred tax liability is increased. This paper focuses on deferred tax expense, because it reflects temporary book-tax differences associated with the income statement. This paper involved the study of Burgstahler and Dichev (1997) which hypothesizes that managers manage earnings because they do not want to report an earnings decline or report a loss. Burgstahler and Dichev (1997) find a higher frequency of zero or small increases in earnings than expected in cross-sectional distributions of annual scaled earnings changes. Phillips et al. (2003) showed a replication of this study in their own sample. The results are consistent with the findings of Burgstahler and Dichev (1997). The first hypothesis which they examine is: “H1: Deferred tax expense is incrementally useful to accrual measures in detecting earnings management to avoid an earnings decline” (Phillips et al., 2003, p. 496). The second hypothesis is: “H2: Deferred tax expense is incrementally useful to accrual measures in detecting earnings management to avoid a loss” (Phillips et al., 2003, p. 499). Finally the third hypothesis is: “H3: Deferred tax expense is incrementally useful to accrual measures in detecting earnings management to avoid failing to meet or beat analysts` forecasts” (Phillips et al., 2003, p. 499). In order to investigate whether managers use earnings management to avoid an earnings decline, Phillips et al. (2003) use a probit regression, which is as follows:For earnings management (hereafter: EM) Phillips et al. (2003) follow Burgstahler and Dichev (1997) so that EM equals 1 if the change in a firms net income from yeart-1 to yeart divided by the market value of equity at the end of yeart-2 is within the boundary of 0-0,01 and -0,01-0. DTE in the probit regression refers to the deferred tax expense while accruals are measured by three accrual measurement methods, which are total accruals, discretionary accruals, and the forward looking abnormal accruals. The two control variables they include in this regression are the change in a firm’s cash flow from continuing operations from yeart-1 to yeart and a dummy variable whether a firm is an industrial operating firm or not. Phillips et al. (2003) uses three accrual models in order to test the incrementally usefulness of deferred tax expense beyond these accrual models. With regard to the first accrual model the coefficient of deferred tax expense is significant for earnings decline and loss avoidance, which means that deferred tax expense is incrementally useful in detecting earnings management to avoid an earnings decline and to avoid a loss. The coefficient of total accruals is also found to be significant which means that total accruals are also useful in detecting earnings management in this case. For earnings management to avoid loss the results are similar. The results for the deferred tax expense are similar when discretionary accruals are used as a proxy for accruals. This means that deferred tax expense is incrementally useful to detect earnings management when managers avoid reporting a decrease or when managers avoid reporting a loss. But in case when managers use earnings management to meet the forecasts of analysts deferred tax expense is not incrementally useful to detect earnings management.The results are consistent with H1 and H2, but they are not consistent with the third hypothesis, namely that deferred tax expense is incrementally useful to accrual measures in detecting earnings management to avoid failing to meet the forecasts of analysts. 4.2.2 Phillips et al. (2004)In this paper Phillips et al. (2004) investigate the relation between changes in annual earnings and changes in the components of deferred tax assets and deferred tax liabilities inferred from firms` annual income tax footnote disclosure.Phillips et al. (2004) assume that managers typically have more discretion in financial reporting compared to tax reporting and they assume that managers are able to use such discretion to manage the book profit upward without increasing current taxable profit. This will increase the net deferred tax liability, deferred tax assets – deferred tax liability. Phillips et al. (2003) find that deferred tax expense is incrementally useful beyond total accruals and two abnormal accrual measures to detect earnings management. However firms are not obligated to disclose the components of deferred tax assets and deferred tax expense. According to Statement of Financial Accounting Standards No. 109 firms must report the significant components of deferred tax assets and deferred tax liabilities in the note of the financial statements. Phillips et al. (2004) used this information to investigate which components of the net deferred tax liability (DTA-DTL) reflect earnings management to avoid an earnings decline.They focus on the change in net deferred tax liabilities and changes in its components to examine the relation between book-tax differences and the likelihood of earnings management to avoid a decline in the earnings. The researchers used hand collected data for the change in net deferred tax liability and after having the data available the researchers utilized the components of firms` deferred tax assets and deferred tax liabilities to decompose the total change in a firm`s net deferred tax liability into changes in the next components; “(1) revenue and expense accruals and reserves, (2) compensation, (3) depreciation of tangible assets, (4) other asset valuation (e.g., expenses related to intangible assets, inventory, and leases), (5) miscellaneous items, (6) tax carry forwards, (7) unrealized gains and losses from securities, and (8) the deferred tax asset valuation allowance account” (Philips et al., 2004, p. 45).In this paper the change in a firm`s net deferred tax liability is used as the proxy for book-tax differences. Three hypotheses are developed in order to study the relation between the different components of the deferred tax liability and the deferred tax asset and earnings management. These are: “HI: The change in the net deferred tax liability is incrementally useful to accrual measures in detecting earnings management to avoid an earnings decline. H2: Changes in the net deferred tax liability components relating to accruals and reserves, compensation, depreciation, other asset valuation, and/or miscellaneous items are useful in detecting earnings management to avoid an earnings decline. H3: The change in the deferred tax asset valuation allowance is useful in detecting earnings management to avoid an earnings decline” (Philips et al., 2004, p. 46). The first and second hypotheses are related to the different components of the deferred tax liability and the third hypothesis is related to the Valuation Allowance Asset (VAA) which is a deferred tax asset. The sample is gathered from the Compustat database and the sample period consists of U.S. firms for the years 1994 till 2000. The final sample consists of 396 firm-year observations. The researchers find initially that the total change in the net deferred tax liability is not incrementally useful to accrual based measures by detecting earnings management. But, they find evidence on the second hypothesis namely changes in the components of the net deferred tax liability provide evidence in explaining the probability of earnings management in order to avoid a decline in earnings. These components are revenue, expense accruals and reserves. With these findings hypothesis 2 is confirmed. The other hypotheses remain unconfirmed due to insignificant relations.4.2.3 Schrand and Wong (2003)The researchers investigate whether banks engage in earnings management by setting a high valuation allowance associated with deferred tax assets and whether banks adjust the valuation allowance in the subsequent periods. According to SFAS 109 companies are allowed to create a valuation allowance against deferred tax assets. Valuation allowance for deferred tax assets is a balance sheet item that shows the part of the deferred tax asset, the company does not expect it will be able to realize this amount in future periods. Schrand and Wong (2003) argue that firms could overestimate the valuation allowance and strategically write off this amount in order to increase earnings in future years. The study focuses on a sample of publicly trading banks. The reason why Schrand and Wong (2003) choose for banks is, because banks have large deferred tax assets and as a consequence they have the potential to create such valuation allowances. The final sample consists of 235 banks that adopted SFAS 109 in 1993. The article argues that banks, which have incentives to use the valuation allowances as a tool for earnings management, will overestimate the valuation allowance against deferred tax assets. Schrand and Wong (2003) perform a regression in order to test whether banks use the deferred tax assets as a tool for earnings management. And indeed Schrand and Wong (2003) find that banks use the valuation allowance against deferred tax assets as a tool to do earnings management. This paper is interesting, because it examines the relation of deferred tax assets as a tool to engage in earnings management while the papers above investigated whether deferred taxes could be incrementally useful beyond accruals to detect earnings management. Another study by Christensen et al. (2008) investigated whether valuation allowances against deferred tax assets by taking a final sample of 444 ‘bath’ firms and 444 control firms.4.2.4 Christensen et al. (2008)Christensen et al. (2008) investigate whether firms which make large write-offs or were undertaking restructuring projects use the deferred tax valuation allowance in order to present better results related to a big bath. Big bath accounting is a form of earnings management whereby management is able to overestimate the expenses in the current period, since the firm is undertaking restructuring projects which is related to expenses in this period. If the expenses are overestimated, the profit in the future will be higher since there are few expenses left to show. In other words, the researchers examine whether firms choose the level of the valuation allowance against deferred tax assets when they incur special charges like restructuring costs or write-offs. SFAS No. 109 allows firms to use their discretion to determine the level needed for the valuation allowance. The paper argues that managers may use this discretion to set the allowance too high thereby allowing the later reversal of these cookie jar reserves in order to manage the earnings (Christensen et al., 2008). A valuation allowance could be recognized because there are concerns about future profitably calls into question. Sometimes managers recognize a larger than necessary valuation allowance to create a “cookie jar reserve” that can be reversed in subsequent periods. Sometimes managers recognize a smaller than necessary valuation allowance to minimize the earnings impact. Two hypotheses are developed in order to study whether which firms use the deferred tax valuation allowance to make the current “big bath” even larger. “H1A: Bigger bath firms are more likely to subsequently decrease the deferred tax asset valuation allowance in the year following the bath than are smaller bath firms, and H1B: Bigger bath firms have worse operating performance in Year + 1 (relative to the bath year) compared to smaller bath firms” (Christensen et al., 2008, p. 606). The reason for the first hypothesis is because the researchers argue that if there are firms which create valuation allowance, then it is more likely that these are the bigger bath firms. The sample exists of 444 bath firms and 444 control firms. The researchers expect that firms increase their deferred tax asset valuation allowances, when there is a large special charge, compared to firms of similar size within the same industry which not recognized a special charge. The results are consistent with their expectations. The results for the first hypothesis are that there is evidence on average that bigger bath firms establish larger than necessary valuation allowances in order to reverse those and increasing earnings in year +1. The results for hypothesis H1B are that the magnitude of a valuation allowance provides information about how it will perform in the future. The difference with the paper of Schrand and Wong (2008) is that this article did their research on firms and not on banks. Schrand and Wong find that deferred tax assets are used as an earnings management tool by banks and this paper finds evidence of firms using the valuation allowances to create even a bigger “big bath”. 4.3 SummaryThis chapter provided insight in the literature of tax associated with earnings management. Blaylock et al. (2014) conclude that there is a positive association between earnings management and book-tax conformity. Tang (2014) concludes that when there is high book-tax conformity it deters management from performing earnings management and tax avoidance activities. This contradicts the conclusion found by Blaylock et al. (2014), but an explanation for the different results can be due to the fact that Tang constructed a new conformity index to measure conformity by adjusting for the opportunistic managerial effect. Tang (2014) showed that when there is high book-tax conformity less earnings management occurs. Coppens and Peek (2005) is included in this review for their investigation on whether firms manage earnings to avoid losses and to avoid decreases in countries where there is high alignment between book profit and taxable profit and low alignment between book profit and taxable profit. The results show that private firms manage earnings to avoid losses in both high alignment countries and in low alignment countries. However they fail to reject the second hypothesis, namely that private firms manage earnings to avoid earnings decreases.Phillips et al. (2003) find that deferred tax expense is incrementally useful in detecting earnings management beyond accrual models in which earnings management was measured as loss avoidance and earnings decline avoidance. Phillips et al. (2004) expand the previous mentioned study by decomposing deferred tax item in different components. They focus on the change in net deferred tax liabilities and changes in its components to examine the relation between book-tax differences and the likelihood of earnings management to avoid a decline in the earnings. Phillips et al. (2004) find evidence that some components of the net deferred tax liability provide evidence in explaining the probability of earnings management in order to avoid a decline in earnings. These components are revenue, expense accruals and reserves. Two articles investigate the relationship of tax and earnings management by looking at the valuation allowances which is a component of the deferred tax asset. Schrand and Wong (2005) find that valuation allowances against deferred tax assets are an earnings management tool used by banks. Christensen et al. (2008) investigate whether firms which make large write-offs or were undertaking restructuring projects use the deferred tax valuation allowance in order to present better results related to a big bath. Schrand and Wong (2005) find that deferred tax assets are used as an earnings management tool by banks and Christensen et al. (2008) find evidence of firms using the valuation allowances to do more “big bath accounting”. 5. Hypothesis developmentFrom the agency theory it is known that the management can act out of self-interest. In other words, managers can act in their own interest while harming the needs of the stakeholders. Following from the positive accounting theory it is known that there are three hypotheses which describe why managers engage in earnings management. The paper of Burgstahler and Dichev (1997), which will be discussed more thoroughly in chapter five, argues that managers use earnings management to (1) avoid decreases, and (2) to avoid losses which can be due to management striving to maximize their own benefits. The Generally Accepted Accounting Principles (GAAP) provides management with discretionary room to achieve this, while tax accounting restricts this discretionary room. Due to this regime difference, timing differences arise between accrual accounting and cash based tax accounting, which is discussed in chapter three.This thesis follows the studies of Burgstahler and Dichev (1997) and Phillips et al. (2003). The article of Burgstahler and Dichev (1997) will be explained in detail in the next chapter. This chapter introduced the article of Phillips et al. (2003) in which it is investigated whether deferred tax expense is incrementally useful in detecting earnings management in three settings which are: earnings decline avoidance, loss avoidance and meeting analysts forecast. This thesis will not perform the third setting, since necessary data is not available for the European setting. Hence, this thesis will test the following hypotheses:H1: Deferred tax expense is incrementally useful to total accruals in detecting earnings management to avoid an earnings decline.H2: Deferred tax expense is incrementally useful to discretionary accruals in detecting earnings management to avoid an earnings decline.H3: Deferred tax expense is incrementally useful to total accruals in detecting earnings management to avoid a loss.H4: Deferred tax expense is incrementally useful to discretionary accruals in detecting earnings management to avoid a loss.This thesis expects that deferred tax expense is incrementally useful in detecting earnings management beyond accruals for both settings. The next chapter will discuss the research design that will be used in order to accept or reject the four hypotheses.6. Research designThis thesis uses a probit-regression in order to test the formulated hypotheses. To detect firms engaged in earnings management, the study of Burgstahler and Dichev (1997) is followed by using the earnings distribution approach. The method of earnings distribution to detect earnings management divides firms into two categories: firms which are assumed to engage in earnings management and firms which are not. This thesis will include a t-test in order to investigate whether these two groups differ significantly from each other. Also this thesis uses a probit-regression in order to assess the incremental usefulness of deferred tax expense over total accruals and discretionary accruals in detecting earnings management. This chapter therefore includes information about what the t-test is and it will include information about a probit-regression. With this information also the variables that are included in the probit-regression are discussed. 6.1 Earnings distributionBurgstahler and Dichev (1997) investigate whether firms manage earnings to avoid an earnings decline and to avoid losses. In their research they use earnings distribution as a proxy measure for earnings management, whereas this thesis also includes deferred tax expense as a proxy measure. Burgstahler and Dichev (1997) collect all available observations of the annual industrial and research Compustat databases for the year 1976-1994 which meet minimal data requirements. Banks and financial institutions are deleted. In this study there are two types of evidence to determine whether earnings management to avoid earnings decreases and losses are present. The first way is by plotting graphical evidence in the form of histograms of the pooled cross-sectional empirical distributions of scaled earnings changes and levels of earnings. The second way is by presenting statistical tests of two hypotheses, these hypotheses are. “H1: Earnings are managed to avoid earnings decreases. H2: Earnings are managed to avoid losses” (Burgstahler and Dichev, 1997, p. 102). The assumption they have is that under the null hypothesis of no earnings management, the distributions of earnings changes and earnings levels are relatively smooth. In the figure below there is a histogram which shows a single peaked, bell shaped distribution. Also the figure shows that there is an irregularity near zero which indicates that firms might be engaging in earnings management to avoid earnings decreases. The irregularity near zero is also confirmed by the statistical test. Figure 3 shows a distribution that departs from a normal distribution. It can be seen that especially near zero there are too few firms which report a small earnings decline and too many which reports a small earnings increase or steady earnings, compared to a normal distribution.Figure 3: Burgstahler and Dichev, earnings management to avoid an earnings declineFigure 4 shows that there is a single-peaked, bell-shaped distribution which is not smooth in the area of zero earnings. Earnings that are just less than zero occur much less than expected and earnings just more than zero occur much more than expected. The irregularity near zero is also significant for the statistical test.Figure 4: Burgstahler and Dichev, earnings management to avoid a loss.Burgstahler and Dichev (1997) motivated two theories for using earnings management to avoid earnings decreases and losses. The first theory involves the transaction cost and the second theory involves the prospect theory. The transactions cost theory implies that when firms report an earnings decrease or loss the costs in transactions with stakeholders are higher compared to firms which reported an earnings increase or profit.However, other articles argue that the irregularity near zero is not due to earnings management per se. Durtschi and Easton (2005) provide evidence that the irregularity near zero of the earnings distribution is affected by deflation and by sample selection. Beaver et al. (2007) showed that income taxes and special items for profit and loss lead to a discontinuity at zero in the earnings distribution. The reason for this fact is that due to income taxes the reported profit moves near to zero. Which might indicate that income taxes contribute to the discontinuity of the earnings distribution. 6.2 T-test and probit-regressionThis thesis uses a t-test in order to assess whether there are differences in accruals and deferred tax expenses of firms which engage in earnings management and firms which did not engage in earnings management. Whether a firm is assumed to be engaged in earnings management, in this thesis, is proxied by assessing the earnings distribution. In this distribution, firms that are positioned near zero earnings or show a zero or slightly positive increase in earnings are assumed to be earnings management firms. A t-test is a statistical comparison of two means between two groups therefore this thesis classified the sample into two dummy groups; EM=1 and EM=0. As this thesis is a replication study of Phillips et al. (2003) a probit regression will also be included in order to examine whether deferred tax expense is incrementally useful in detecting earnings management beyond total accruals and discretionary accruals. A probit regression is a regression whereby the dependent variable is an indicator variable. In this thesis, for the variables deferred tax expenses, accruals, long-term debt and cash flow, the natural logarithm will be taken in order to correct for high differences between observed data values. To test the hypotheses this thesis takes an archival research approach. This paper will use data included in Wharton Research Data Services (WRDS) in order to test the hypotheses. Data is extracted from Compustat which is a part of the database of WRDS. The hypotheses will be tested for the years between 2010-2013, however necessary data is extracted for the years between 2008-2013; this is done to be able to calculate variables like total accruals and non-discretionary accruals. This thesis analyses the incremental ability of deferred tax expense over accrual measurement methods to detect earnings management. In this thesis two situations are considered in which earnings management is likely to be present. Firms that engage in earnings management to avoid a decrease in the earnings and firms that avoid a loss using earnings management. The following probit regression is used to examine whether deferred tax expense is incrementally useful beyond the accruals in detecting earnings management:EM= ∝ + β1DTE+β2AC+ β3?CFO+β4LTDEBT+ β5IND+ εEM=1 if (Net Income t-Net Income t-1)/ MVEquity t-2 is between 0 and < 0.01 and 0 if (Net Income t- NetIncomet-1)/MVEquitet-2 is <0 and > -0.01 for Sample 1; For sample 2: 1 if (Net Income / Equity t-1) =is between 0 and < 0.01 and 0 if (Net Income t / Equity t-1) is <0 and >-0.01 for Sample 2DTE= Deferred tax expense AC= a measure of firm i's accruals in year t;ACFO= ? in the firms cash flow from continuing operations from year t-1 to tLTdebt= the long term debt of a companyIND= 1,2,3,4,5,6,7,9,10 industry codes based on SIC codesFollowing Burgstahler and Dichev (1997), EM holds 1 if a firm reports a scaled earnings change in yeart equal to zero or greater than zero and less than 0.01 of the firms market value at yeart-2. EM holds 0 if the scaled earnings change in year t is equal to -0.01 or greater and less than 0. The deferred tax expense is the deferred component of the firms` total income expense. It reflects the amount that is added to the deferred tax amount on the balance sheet.AC reflects one of the two accrual variables which are used to detect earnings management. The measurement approaches, which are used in this thesis, are the total accruals and the Modified jones model. Also included is ACFO to control for the change in cash flow from continuing operations. Because this could affect the current performance of the firm and it may reduce engaging in earnings management to achieve a zero or slightly positive earnings change. Also included are the SIC industry dummy variables to control for the differences in the incentives to engage in earnings management across industries. Finally, this thesis includes the long-term debt variable to control for the incentives for engaging in earnings management. Based on the theory of the debt covenant hypothesis, firms are likely to engage in earnings management while not exceeding contractual norms that have been set by banks with an increasing cost of capital as a consequence (Watts and Zimmerman, 1986).6.3 Measuring accrualsAs mentioned in the second chapter there are different methods to measure the accruals (Healy Model, Jones Model etc.). This thesis however uses only the total accruals and the modified jones model in an attempt to detect earnings management. By doing this the study of Phillips et al. (2003) is followed. Phillips et al. (2003) use total accruals and the Modified jones model, both will also be included in this thesis.For the measurement of the accruals, this paper uses the total accruals and the Modified jones model (Dechow et al., 1995) as proxies for accruals that should reflect the earnings management. The total accruals are as mentioned in the chapter of earnings management, the earnings from continuing operations minus cash flow from continuing operations. TA=EBEI-(CFO-EIDO)TA= Total Accruals in year tEBEI= Income before extraordinary items in year tCFO= Cash flow from operations in year tEIDO= Extraordinary items and discontinued operations in year tThe total accruals are scaled by the assets of the previous year in order to control for the size of the firms. The Modified jones model tries to estimate the non-discretionary part of the accruals. This is done by firstly calculating total accruals, where after the non-discretionary accruals are calculated.Below the Modified jones models is presented:Non-discretionary accruals= ∝1(1/A_(t-1) )+∝2(?Rev_t-?Rec_t )+∝3(PPE_t)This model scales the variables revenue, receivables and property plant and equipment by the total assets of the previous period. The first variable At-1 is the value of the assets available in the previous period. The second variable is calculated by taking the difference of the current and the previous year for revenues and receivables. The third variable is the amount of the property plant and equipment. The Modified jones model can be calculated by regressing the data for many firms for the same period (cross-sectional) or by regressing the data for the same firm during a time period (time-series). This thesis focuses on the cross-sectional method to calculate the discretionary accruals.The values for the alpha`s are determined by an OLS estimation of a1, a2 and a3 due to using the following formula:TA_t= a1(1/A_(t-1) )+a2(?Rev_t )+a3(PPE_t )+eIn this formula all the variables are scaled by the total assets of the previous period. The total accruals consist of two elements, namely the non-discretionary accruals and the discretionary accruals. This means that in the last step for calculating the discretionary accruals, the non-discretionary accruals are subtracted from the total accruals. Discretionary accruals = Total accruals – Non-discretionary accruals6.4 Measuring deferred tax expenseAs mentioned in chapter 4 this thesis will use deferred taxes as a proxy for the book-tax conformity. In chapter 3 it is mentioned that companies are obligated to prepare two statements, namely one whereby the book profit is shown and one which is prepared for the tax authorities in order to determine the taxable income. Accounting rules and regulation allow firms to include accruals in their net income, whereas tax authorities focus more on cash flow since they want to calculate the taxable profit. This profit forms the basis to levy taxes for by governments and therefore tax rules and regulations allow less discretion to managers. This leads to differences in the profit between the accounting statement and the tax statement. This difference is expressed as a deferred tax item in the financial statement. The proxy for book-tax conformity in this thesis is the increase or decrease of deferred tax expense in a year. It captures the increasing or decreasing amount of deferred tax liabilities or deferred tax assets. In chapter 3 I explained that deferred tax assets are created since taxes are paid but not yet recognized in the income statement, whereas a deferred tax liability is created where tax relief is provided in advance or when income is accrued and not taxed yet. As mentioned the proxy for book-tax conformity in this thesis is the increase or decrease of the deferred tax assets or liabilities in year t. Since this thesis analyses the firms in the period between 2010-2013 data for deferred tax assets and deferred tax liabilities is gathered for the year 2009 in order to calculate the deferred tax expenses in 2010 until 2013. Including both the accruals and the deferred tax expense in the probit regression allows this thesis to determine whether deferred tax expense is incrementally useful in detecting earnings management. Also included is the difference of cash flow from operations. An increase in operating cash flows means an increase in the current performance of the firm and this reduces the need for managers to manage earnings to achieve earnings objectives. 6.5 Long-term debtAdditional to the studies as mentioned above, this study incorporates the control variable: long term debt. Based on the theory of the debt covenant hypothesis, firms are likely to engage with earnings management while not exceeding contractual norms that has been set by banks with an increasing cost of capital as a consequence (Watts and Zimmerman, 1986). 6.6 Industry codeTo control for industries this thesis divides the firms according to the SIC codes into dummy variables according to the distribution as mentioned in table 1. The industry dummy variables are included in order to control for possible differences in the incentives to avoid an earnings decrease or to avoid loses across these industries. Table 1: Assembling of SIC-codesSic CodesIndustryDummy variable01-09Agriculture, Forestry, Fishing110-14Mining215-17Construction320-39Manufacturing440-49Transportation and Public Utilities550-51Wholesale Trade652-59Retail Trade760-67Finance, Insurance, Real estate870-89Services991-99Public Administration106.7 Sample The sample will consist of European firms. The reason why this thesis focuses on European firms is because Phillips et al. (2003) did research on US firms and they find significant evidence that deferred tax expense is incrementally useful in detecting earnings management. This thesis tries to examine whether these results are consistent within the European firms. Also it is interesting to see whether the results of the European listed firms are consistent with the findings of Phillips et al. (2003). There exist important differences between U.S. GAAP and IFRS regarding the deferred tax policies. The standards under U.S. GAAP are more stringent when compared to the standards under IFRS. Due to this difference, the values of assets and liabilities of firms that use U.S. GAAP are more straightforward when compared to firms that use IFRS.For the sample the WRDS database is used which includes several separate databases. These databases can be used for research. For this thesis Compustat, which is a separate database in WRDS, is used to gather the data needed for examining whether deferred tax expense is incrementally useful in detecting earnings management beyond accrual measurements. The time period for which data is obtained includes fiscal year 2008 until the fiscal year 2013. 6.8 Differences between this thesis and Phillips et al. (2003)-Control variablesAdditional to the studies as mentioned above, this study incorporates the control variable: long term debt. Based on the theory of the debt covenant hypothesis, firms are likely to engage with earnings management while not exceeding contractual norms that has been set by banks with an increasing cost of capital as a consequence (Watts and Zimmerman, 1986). -PopulationThis paper examines whether deferred tax expense is incrementally useful in detecting earnings management in European context whether Phillips et al. (2003) examined this research question for US firms. 6.9 SummaryThis chapter describes the research method that will be used in order to answer the research question. Also the variables are discussed and the differences between the study of Phillips et al. (2003) have been discussed. The next chapter will analyse the two samples in order to accept or reject the hypotheses.7. ResultsThe previous chapter described the research design. It was mentioned that this thesis will be a replication study of Phillips et al. (2003) with implementing of the study of Burgstahler and Dichev (1997). This thesis has the same settings as in Phillips et al. (2003), except for the third setting. In Phillips et al (2003) the third setting is related to earnings management in order to meet the analysts forecast. In this thesis there will be two samples whereby one sample considers earnings management to avoid an earnings decrease whereas the other sample considers earnings management to avoid a loss.First of all this chapter will provide insights in the necessary assumptions for regression models. After discussing the assumptions, first the results of the earnings management distributions will be presented following the results that are obtained from the probit-regression.Table 2: Sample 1 and 2 per year and sector before logarithm.SectorYear12345678910TotalSample 120101487475568.1781.5148147253.5002.07819318.45320111568145838.9631.5728847673.6662.16621119.78220121457215408.4851.4758297343.5052.08519418.71320131467835399.0531.5218507513.6502.15419119.638Sample 220101005403955.5601.0156164292.7131.48313212.98320111055604426.0441.0536124342.9301.55714613.88320121196214426.7551.0826404652.8221.56215014.65820131205634156.6131.0916544782.6901.54416014.338Table 2 shows the distribution of the sample per year and sector. For both sample 1 and 2 the distribution of the observation per year is relatively constant. The amount of observations in this research is relatively high, as all European listed firms are included. This thesis does not suggest any difference between the countries as all European countries apply IFRS. The table shows that sector 4 (manufacturing) is by far the biggest sector in this research. However in this research for the estimation of all regressions a control variable is added for sector. Therefore this should not lead to any statistical associations caused by an over- or under-representation of a certain sector. For all regressions data per sample, as numerically displayed above, is included. Due to the lack of data availability, as a consequence of missing financials in the WRDS-database, certain observations might be omitted by Stata in order to be able to estimate the applicable regression. As each hypothesis might request different data, the omitted observations might differ per hypothesis. In order to exclude any possible effect of contaminated observations, such as outliers, the 1st and the 99th percentile of the independent variables in the data is removed by means of winsorization. The number of observations which are used for the final regressions can be found in the tables containing the estimations for each regression. 7.1 Descriptive statisticsIn this thesis, multiple explanatory and control variables are used. This section provides an overview of these variables and their descriptive statistics. The descriptive statistics are shown in table 3. Table 3 shows the statistics related to the first data sample and to the second data sample. In this presentation of the descriptive statistics both the explanatory and the control variables will be discussed. These are emdummy, dummyloss, logdte, logta, logda, logcfo and logdebt. Table 3 descriptive statistics of sample 1 and 2 after winsorization and logarithm,.VariableObsMeanStd. Dev.MinMaxSample 1emdummy76.6070,600,4901logdte22.2292,173,24-5,3010,26logta42.120-3,201,09-7,02-0,96logda24.65-2,442,43-7,255,98logcfo31.5074,463,00-2,4212,12logdebt54.6045,753,54-2,9414,37Sample 2dummyloss55.8830,320,4701logdte22.1011,553,72-5,4710,75logta31.562-3,251,11-7,24-0,94logda17.384-2,372,67-7,576,53logcfo15.9882,843,47-5,1211,21logdebt39.7005,603,55-2,9814,16VariableMeaningemdummy (only sample 1)Dummy variable that takes value 1 for firms that engaged in earnings management in order to avoid an earnings decline.dummyloss (only sample 2)Dummy variable that takes value 1 for firms that engaged in earnings management in order to avoid a loss.logdteThe logarithm of the deferred tax expense, which presents the difference in deferred tax assets or deferred tax liabilities compared to previous year.LogtaThe logarithm of the total accruals.logdaThe logarithm of the discretionary accruals.logcfoThe logarithm of the operating Cash Flow which presents the difference compared to previous year.logdebtThe logarithm of the long-term debt.Table 4: Legend The overview shows logdte ranging between the minimum value of -5,30 and the maximum value of 10,26, with a mean value of 2,17 and a standard deviation of 3,24 for sample one. For sample two, the overview shows logdte ranging between the minimum value of -5,47 and the maximum value of 10,75, with a mean value of 1,55 and a standard deviation of 3,72. The mean presents the average value of the variables and the standard deviation presents the deviation distance from the average observed value. 7.2 AssumptionsThis thesis uses a t-test and a probit-regression in order to statistically analyse the data. The data is gathered for a European sample for the year 2010 until and including 2013. In order to optimally perform the statistical analysis, the data first needs to meet certain requirements. These requirements are: the data is normal distributed, possible outliers in the data are removed, the data is assessed for multicollinearity, and finally, the data needs to be assessed for heteroscedasticity.7.2.1 Normal distributionFirst the data will be assessed for normal distribution by means of plotting al residuals into a histogram. A normal distribution shows the data concentrated around a central value. In order to assess the distribution of the data, a histogram is used. First a regression is estimated in which all variables are included with exception of the dummy variables. The data which is used for this analysis contains the logarithm of the independent variables and is winsorized. Based on this regression, the residuals are estimated in Stata. To assess the data for normal distribution means to assess the residuals for normal distribution. Figure 5 and figure 6 shows this distribution of the residuals and does not indicate that the sample is not normally distributed for both sample 1 and 2. All analysis with independent variables which are used in this thesis concern the logarithm of the initial value. Figure 5: histogram sample 1Figuur 6: histogram sample 27.2.2 OutliersThe second assessment of the data is on possible outliers. This thesis uses a t-test and a probit regression. The results that are obtained from both tests can be severely distorted when the data contains outliers. In order to clear the data from potential outliers this thesis uses the method of winsorization. With winsorization the 1st and the 99th percentile of the data is removed. This process helps removing extreme data observations that might distort a statistical analysis.7.2.3 MulticollinearityThe third requirement that needs to be satisfied is the absence of multicollinearity. With multicollinearity, variables are highly correlated. This is troublesome since due to this high correlation between variables, a possible effect on the dependent variable is impossible to relate to an individual explanatory variable. Therefore before running statistical tests on the data, it is important to first assess the data for possible multicollinearity. For this a variance inflation factor (VIF) is calculated. Table 5 shows the results of this VIF-test. The value of the mean VIF is 2.0 for sample 1 and 1.57 for sample 2, which indicates that the data does not suffer from multicollinearity. In the VIF-test the mean VIF value should be near 1 in order to conclude the absence of multicollinearity. Considering the relative low VIF-value no indication of multicollinearity exists. A value higher than 5 might indicate multicollinearity, this is not the case (O’ Brien, 2007). Table 5: Multicollineairity testVariableVIFSample 1logcfo2,73logdte2,56logdebt2,43logta1,15logda1,15Mean VIF2,00Sample 2logcfo1,46logdte1,49logdebt1,67logta1,65logda1,58Mean VIF1,577.2.4 HeteroscedasticityThe final requirement is the requirement of the absence of heteroscedasticity. With heteroscedasticity the variance of one variable does not remain equal to the variance of another variable in the analysis. If heteroscedasticity is present, the outcome of the analysis can be greatly affected. In order to ensure the results are not driven by heteroscedasticity the robust function in Stata is used for all regresions. This function is an alternative for the OLS-regression and estimates each regression with robust standard errors. In case of absence of heteroscedasticity and in the case of presence of homoscedasticity, this function can still be useful (Yamano, 2009).7.3 Earnings management modelBurgstahler and Dichev (1997) detect earnings management through an earnings management distribution approach. The study assumes two situations whereby earnings management is likely to be present. The first situation is that firms engage in earnings management to avoid an earnings decline which can be described as firm-years with zero or slightly (positive) changes in earnings. The second situation is that firms engage in earnings management to avoid a loss which can be described as firm-years with zero or positive (slightly) earnings level. Earnings management equals 1 if a firm has a scaled earnings change between 0 and 0.01. Alternatively earnings management equals 0 if a firm has scaled earnings change less than 0 and greater or equal to -0.01.This thesis focuses on earnings management within two settings; (1) earnings management to avoid a decrease and (2) earnings management to avoid a loss. 7.3.1 Earnings management to avoid an earnings declineFigure 7 shows a replication of the study of Burgstahler and Dichev (1997) with respect to scaled earnings changes (net income t – net income t-1 / equity t-2). The figure shows an irregularity near zero which means that there is earnings management to avoid earnings decrease. It shows that there is a high frequency of firms that reported a slightly earnings increase. It can be said that next to the zero a lot of firms reported a slightly positive earnings increase which means that firms prefer to choose to engage in earnings management in order to prevent an earnings decline. The same results are consistent with the study of Burgstahler and Dichev (1997) as they find a histogram which showed a single peaked, bell shaped distribution and an irregularity near zero which means that firms engaged in earnings management to avoid earnings decreases.As mentioned in chapter 2 a decline in earnings management could attract a lot of public attention. Managers don’t like public attention since they have to take responsibility over their performance. Shareholders for example can ask the manager to explain why earnings are declined. However except for the irregularity near zero, the histogram shows a normal distribution. Figure 7: earnings management distribution sample 131750106045The histogram in figure 7 shows the distribution of earnings changes relative to prior year earnings. The distribution clearly shows a spike in earnings changes which is concentrated just above zero. 7.3.2 Earnings management to avoid a lossThis section will focus on earnings management to avoid a loss (net income t / equity t-1). Burghstahler and Dichev (1997) find that earnings just less than zero occurs much less than expected and earnings just more than zero occur much more than expected. These results are consistent for the setting of European listed firms with respect to earnings management to avoid loss. The figure 8 shows that there is a single-peaked distribution. A lot of firms reported a slightly positive earnings. Furthermore a lot of firms reported earnings just around zero. In other words earnings that just are less than zero occur much less than expected whereas earnings just more than zero occur much more than expected which indicates that firms likely use earnings management to avoid a loss. Figure8: earnings management distribution sample 226670266707.4 T-testIn this section an analysis will be performed in order to assess whether the irregular distribution in earnings is a significant one. In order to do this, two statistical methods/instruments will be deployed. The first one is a t-test in which a possible difference between the EM=0 sample and the EM=1 sample will be analyzed. In this test the focus is on whether the mean value of the deferred tax expense, total accruals, and discretionary accruals is significantly different between the two groups. The number of observations in both groups per year can be found in the underlying table 6: Table 6: Observations EMVariableDummy01TotalSample 120106.26912.19018.45920118.39811.38919.78720127.97010.74718.71720137.82011.82419.644Sample 220109.0333.95712.99020118.2465.64313.889201210.2554.43814.633201310.5373.80414.3417.4.1 T-test for the first settingIn this section a t-test will be performed in order to investigate whether the means differs significantly from each other for the following variables: total accruals, discretionary accruals and deferred tax expenses. In this sample earnings management is assumed to be used in order to avoid reporting an earnings decline, where after in section 7.4.2 the t-test will be discussed for the second setting.From table 7, it can be said that for the first sample, significant differences do exist among the two groups in this sample (with exception of the total accruals). The two groups exist of firms which are engaged in earnings management and of firms which are not engaged in earnings management. These significant differences relate to the deferred tax expense and discretionary accruals. This finding indicates that for the group where earnings management is assumed to exist, the items of deferred tax expense and discretionary accruals show a significantly different mean value when compared to the group where earnings management is not assumed to be present. This finding could make the items of: deferred tax expense and discretionary accruals useful in detecting firms that have a higher likelihood of engaging in earnings management. Whether these items ultimately are useful will be investigated using a probit-regression in section 7.5.7.4.2 T-test for the second settingThis section is dedicated to perform a t-test for sample two. In this sample earnings management is assumed to be used in order to avoid reporting a loss. Similarly to sample one, in this sample also two groups are distinguished. Again, the first group it is assumed to have a lower likelihood of earnings management compared to the second group in which earnings management is assumed to be more likely. First the deferred tax expense is assessed where after total accruals and finally discretionary accruals are examined for potential significant differences among the groups in sample two. Table 7 shows that the mean for total accruals, discretionary accruals and deferred tax expenses are significantly different.The second statistical test is a probit-regression, with this probit-regression the effectiveness of accruals models and deferred tax expense to detect earnings management will be assessed. In this test the incremental usefulness of deferred tax expense beyond the accrual models to detect earnings management will be assessed.Table 7: t-test sample 2VariableMean Dummy01ProbabilitySample 1log Total Accruals-3,20-3,210,37log Discr Accruals-2,34-2,490,00log Deffered tax expense2,122,210,05Sample 2log Total Accruals-3,30-3,150,00log Discr Accruals-2,64-1,800,00log Deffered tax expense1,641,350,007.5 Probit RegressionThis thesis investigates the incremental usefulness of deferred tax expense beyond accrual models to detect earnings management. In this section a probit regression is used to assess this potential incremental usefulness. The data upon which this analysis is based on, is divided into two samples. The first sample focuses on avoiding an income decrease as a reason for earnings management. The second sample relates to loss avoidance as a reason for firms to engage in earnings management.7.5.1 Sample 1This section is dedicated to the first two hypotheses which are dedicated to earnings decline avoidance. The hypotheses are: H1: Deferred tax expense is incrementally useful to total accruals in detecting earnings management to avoid an earnings decline.H2: Deferred tax expense is incrementally useful to discretionary accruals in detecting earnings management to avoid an earnings decline.In the same order as in the hypothesis development this statistical analysis will first focus on the incremental usefulness of deferred tax expense beyond total accruals. After which a second probit-regression will be estimated to assess the incremental usefulness of deferred tax expense beyond discretionary accruals. The included tables only show the outcome with regard to the independent variables, the outcome for the year- and industry dummy takes many values and is therefore not included. These variables are included to exclude any possible sector- and/or year-effects. The outcome of the sector- and year-dummy can be found in the appendix. When the results in table 8 are analysed with regard to the first hypothesis, the outcome of the probit regression shows that deferred tax expense is positively associated with the earnings management dummy, according to the expectation. However this effect does not seem to be statistically significant. This finding is not consistent with the finding in Philips et al. (2003); this means that the deferred tax expense is not useful to total accruals when detecting earnings management to avoid an earnings decline. This outcome however only applies for the total accruals. This result is achieved after controlling for a possible effect of total accruals, long-term debt, industry effects, and cash flow mutations. The estimation further shows that the log ta, log debt and log CFO are effective and significant variables in detecting earnings management to avoid an earnings decline, this is consistent with the expectation in this thesis. However the outcome shows an opposite sign for the log debt and log ta. For the outcome of the log debt variable this indicates that the probability of earnings management decreases in case of higher debts. This could be caused by the fact that a company has an incentive to perform earnings management in case of debt, but the accountant anticipates that this could be the case thus considering this in the risk assessment. This will lead to (possible) extra attention with regard to the analysis of earnings management. The sign for the total accruals also shows that the result is in contrast with the expectation. This however seems to be caused by the non-discretionary accruals. The total accruals consist of the non-discretionary accruals and discretionary accruals. It is not possible for management to manipulate the non-discretionary accruals; however the discretionary accruals might be manipulated. When analysing the outcome of the second hypothesis, where only the discretionary accruals are included, it can be concluded that the discretionary accruals are positively and significantly associated with earnings management (as expected) on a 10% significance level. The discretionary accruals therefore seem to be the mediating variable, as the variable turns the effect of the deferred tax expense on earnings management significant. The effect of the deferred tax expense is also positively and significantly associated, as expected. The control variables for debt and cash flow from operations are not significant any more. This indicates that the discretionary accruals are also a moderating variable with regard to the cash flow- and debt variable. This can be derived from the fact that the significance of these two variables is disappeared with the replacement of the total accruals by the discretionary accruals. Another possible explanation with regard to the disappearance of the significance might be the reduction in sample size. The sample size for the discretionary accruals is much smaller than the sample for the total accruals. This is caused by the fact that much more data is required in order to calculate the discretionary accruals. The lack of data availability might have caused the disappearance of the earlier results.For the estimation of the probit-model with regard to the second hypothesis sector-code 8 is dropped by Stata. The reason for this is the fact that only observations in the ‘1’ group are present, therefore sector-code 8 predicts the chance of earnings management perfectly, no counter-group is present. The analysis is based on a sample which contains data for the years 2010 until and including 2013. Controlling for possible year effects is done by including these years as a multiple indicator variable. In short DTE is found not to be useful in detecting earnings management. However total accrual is also found to be positive (-0,11) and significant (0,00). This indicates that TA is an effective variable in detecting earnings management to avoid an earnings decline. The results of this probit regression are presented in table 8 and provide reason to reject hypothesis 1. The results show that deferred tax expense is incrementally useful in detecting earnings management beyond discretionary accruals, in the setting whereby firms are engaged in earnings management to avoid an earnings decline. Therefore the first hypothesis is rejected, whereas only the second hypothesis is accepted so far. Hypothesis 3 and 4 are related to the second setting, namely earnings management to avoid a loss. These hypotheses are tested in the following subsection. Table 8: Probit-regression hypothesis 1+2Hypothesis 1Hypothesis 2VariableExpected SignCoefficientStd. ErrorPCoefficientStd. ErrorPlog dte(+)0,0110,0080,1630,0480,0210,025**log ta(+)-0,1130,0170,000***log da(+)0,0310,1800,085*log debt(+)-0,0320,0080,000***-0,0320,0210,130log CFO(+)0,0450,0090,000***0,0310,0250,210Year-dummySector-dummyN79951132Pseudo R-Squared0,0170,037*= On 10% level significant, **= On 5% level significant, *** = On 1% level significant.7.5.2 Sample 2This section is dedicated to the third and the fourth hypotheses which are dedicated to loss avoidance. The hypotheses are: H3: Deferred tax expense is incrementally useful to total accruals in detecting earnings management to avoid a loss.H4: Deferred tax expense is incrementally useful to discretionary accruals in detecting earnings management to avoid a loss.In the same order as in the first setting this statistical analysis first focuses on the incremental usefulness of deferred tax expense beyond total accruals. After which a second probit regression will be estimated to assess the incremental usefulness of deferred tax expense beyond discretionary accruals. The results of the probit regression in table 9 show that total accruals are significantly useful in detecting earnings management. The results obtained in relation to sample 2 show that the coefficient on deferred tax expense is significant at a 1% significance level (p= 0,00). This finding is also consistent with the results of Phillips et al. (2003). Deferred tax expense being significant suggests that this variable is incrementally useful in detecting earnings management beyond total accruals. This thesis expected the sign of the coefficient to be positive. However, the results show that the coefficient of the deferred tax expense is negative. This could be caused by the fact that in the sample which is used for this study the earnings management takes place in the opposite direction compared to the sample of Philips et al. (2003). This means that earnings could be manipulated by management in order to reduce earnings in the current year when it is unavoidable that a loss will be made. This will give management the possibility to boost the earnings in subsequent years.The results also show that total accruals are significantly related in detecting earnings management. This result is achieved after controlling for a possible effect of long-term debt, industry effects and cash flow mutations. The effect of the debt variable was significant for detecting earnings management in case of a loss compared to the prior year, but this is not the case anymore for earnings management to avoid a loss. The reason for this could be that the supplier of debt is more interested in whether the recipient of this money makes a profit or loss, the magnitude of the loss is less relevant. Therefore when it is unavoidable a loss will be made, there is no incentive to manage the earnings. The analysis is based on a sample which contains data for the years 2010 until and including 2013. Controlling for possible year effects is done by including these years as a multiple indicator variable. The results of this probit regression are presented in table 9 and provide reason to accept hypothesis 3. In some cases Stata dropped certain sector-code groups as these groups do not have a counter-sample due to the lack of data-availability. The results of the fourth hypothesis can also be found in table 9. The results obtained in relation to sample 2 and hypothesis 4 show that the discretionary accruals are significant on a 10% level, thus discretionary accruals are significant in detecting earnings management to avoid a loss. Deferred tax expense is not significant in detecting earnings management beyond deferred tax expense whereby firms use earnings management as a tool to avoid a loss. The long-term debt and the cash flow variables seem also not to be significant in detecting earnings management. This result partly contradicts to the findings of Phillips et al. (2003), namely that deferred tax expense is incrementally useful in detecting earnings management to avoid a loss. The difference in relation to the outcome of Philips et al. (2003) could be the consequence of different fiscal legislation in the United States. In the next chapter a conclusion will be given to the results found for the hypothesis and a discussion is outlined to explain the findings.Table 9: Probit-regression hypothesis 3+4Hypothesis 3Hypothesis 4VariableExpected SignCoefficientStd. ErrorPCoefficientStd. ErrorPlog dte(+)-0,0290,0100,003***-0,0680,0490,166log ta(+)0,1860,0240,000***log da(+)0,1160,0630,063*log debt(+)-0,00020,0100,9780,0240,0440,584log CFO(+)-0,0230,0100,017**-0,0270,0450,550Year-dummySector-dummyN5382235Pseudo R-Squared0,0450,095*= On 10% level significant, **= On 5% level significant, *** = On 1% level significant.8. Discussion8.1 IntroductionThis chapter provides a conclusion of the obtained results in this thesis and a discussion with respect to the association of deferred tax expenses and earnings management will be given. A conclusion and a discussion will be given for each earnings management setting; (1) earnings management to avoid an earnings decline and (2) earnings management to avoid a loss.8.2 Findings related to earnings decline avoidancePrior literature presents conflicting result on the association between book-tax conformity and earnings management to avoid an earnings decline. Blaylock et al. (2014) shows that a positive association exists between book-tax conformity and earnings management in which they use four different measures of earnings management. These measures are discussed in chapter four and of these measures one measure is directly related to the first two hypotheses in this thesis, which is the extent to which earnings remain stable over the years. A higher book-tax conformity indicates a lower deferred tax expense. Therefore the results of Blaylock et al. (2014) suggest that deferred tax expense is negatively associated with earnings management. In contrast to the findings of Blaylock et al. (2014), Tang (2014) finds a negative relation between book-tax conformity and earnings management. Related to this thesis, Tang (2014) assumed that higher book-tax conformity is associated with less earnings management. This assumption is based on the notion that higher conformity decreases discretionary room for the management. In chapter 2 I mentioned that the discretionary room is one of the aspects of the definition of earnings management. With higher book-tax conformity, deferred taxes are lower. The notion of Tang therefore is related to the assumption in this thesis; this thesis assumes lower levels of earnings management when deferred taxes are low; this is due to the fact that lower deferred taxes indicate higher book-tax conformity which eventually decreases discretionary room for management to do earnings management. Consequently this thesis assumes that deferred taxes are a useful tool in detecting earnings management. Directly related to this thesis, is the study of Phillips et al (2003). They use the deferred tax expense to detect earnings management, in the setting of earnings decline avoidance and loss avoidance. They find that deferred tax expense is incrementally useful in detecting earnings management beyond accrual models. The accruals models consist of total accruals and discretionary accruals. The hypotheses and findings of Phillips et al. (2003) are related to this thesis. In this thesis deferred tax expense is not incrementally useful in detecting earnings management in an European setting for the years 2010-2013 in avoiding an earnings decline to total accruals. When analysing the outcome of the second hypothesis, where only the discretionary accruals are included, it can be concluded that the discretionary accruals are positively and significantly associated with earnings management. The discretionary accruals therefore seem to be the mediating variable, as the variable turns the effect of the deferred tax expense on earnings management significant. The effect of the deferred tax expense is also positively and significantly associated, as expected. In the next section, the results will be discussed in the setting whereby earnings management is likely used to avoid a loss.8.3 Findings related to loss avoidanceIn the previous section I mentioned that conflicting results were found with respect to the association between book-tax conformity and earnings management. Phillips et al. (2003) showed that deferred tax expense is incrementally useful in detecting earnings management to avoid a loss beyond the accrual models, which are the total accruals and the discretionary accruals. This thesis shows that deferred tax expense is incrementally useful in detecting earnings management beyond total accruals. Deferred tax expense being significant suggests that this variable is incrementally useful in detecting earnings management beyond total accruals. But when discretionary accruals are used as a proxy for earnings management, deferred tax expense is not useful in detecting earnings management, whereas discretionary accruals are useful in detecting earnings management. The difference in relation to the outcome of Philips et al. (2003) could be the consequence of different fiscal legislation in the United States.8.4 LimitationsThis part of the thesis discusses the limitations with respect to the research design. The first limitation of this thesis is related to the selected sample. The selected sample exists of European listed firms. In other words, private companies are not included in the sample, because these were not available. In chapter four Coppens and Peek (2005) discussed whether private firms engage in earnings management and whether tax incentives affect this behaviour. The researchers expected that in countries where financial accounting and tax accounting practices are strongly aligned, firms will have less incentives to use earnings management as a tool to avoid reporting losses and earnings decreases. They found that private firms manage earnings more often to avoid losses in both high alignment countries and in low alignment countries in comparison to public firms. This could be due to the fact that private firms do not attract public attention as much as public firms do. And therefore private firms have more flexibility to use earnings management. Including private firms in this thesis could add value since it would provide the opportunity to do earnings management. Another limitation of this thesis is that it does not make a distinction between countries in Europe with high book-tax conformity and low book-tax conformity. The expectation is that countries within high alignment of book profit and tax profit, there is less earnings management. This is due to the fact that when there is high alignment between accounting rules and tax rules, the discretionary room of the management reduces. Hanlon (2005) mentioned that accounting rules will be given up for tax rules, and these provide less discretion to management. It could be attractive for some to investigate whether there are differences between high alignment countries and low alignment countries in a European setting and see whether there are differences for the incrementally usefulness of deferred tax expenses to detect earnings management beyond accrual measurement. The last limitation in this thesis is the time period of the sample. The sample exists of European firms for the period 2010-213. This period is associated with the financial crisis whereby a lot of companies had a tough time. This may have pushed the managers to use earnings management in order to report good results.8.5 SummaryThis chapter discussed the results which are obtained in chapter six. The results were consistent with Phillips et al. (2003) for two hypotheses, namely deferred tax expense is incrementally useful in detecting earnings management beyond discretionary accruals when firms are engaged in earnings management to avoid an earnings decrease and deferred tax expense is incrementally useful in detecting earnings management beyond total accruals when firms engage in earnings management to avoid loss. LiteratureAtwood, T. J., Drake, M. S., & Myers, L. A. (2010). Book-tax conformity, earnings persistence and the association between earnings and future cash flows.?Journal of Accounting and Economics,?50(1), 111-125.Ball, R., & Brown, P. (1968). An empirical evaluation of accounting income numbers.?Journal of Accounting Research, 6(2) 159-178.Bartov, E., Givoly, D., & Hayn, C. (2002). The rewards to meeting or beating earnings expectations.?Journal of Accounting and Economics,?33(2), 173-204.Blaylock, B., Gaertner, F., & Shevlin, T. (2012). The association between book-tax conformity and earnings management. Review of Accounting Studies, 20(1), 141-172.Burgstahler, D., & Dichev, I. (1997). Earnings management to avoid earnings decreases and losses.?Journal of Accounting and Economics,?24(1), 99-126.Cairns, D., 1946, Creighton, B., & Daniels, A. (2002). Applying international accounting standards (3rd ed.). London, etc.: LexisNexis Butterworths Tolley.Chludek, A. K. (2011). The Impact of Deferred Taxes on Firm Value. Three Empirical Studies on the Cash Flow and Value Relevance of Deferred Taxes and Related Disclosures (Doctoral dissertation, Universit?t zu K?ln).Christensen, T. E., Paik, G. H., & Stice, E. K. (2008). Creating a bigger bath using the deferred tax valuation allowance.?Journal of Business Finance & Accounting,?35(5‐6), 601-625.Christie, A. A., & Zimmerman, J. L. (1994). Efficient and opportunistic choices of accounting procedures: Corporate control contests.?Accounting Review, 539-566.Coppens, L., & Peek, E. (2005). An analysis of earnings management by European private firms.?Journal of International Accounting, Auditing and Taxation,?14(1), 1-17.Dechow, P. M. (1994). Accounting earnings and cash flows as measures of firm performance: The role of accounting accruals.?Journal of Accounting and Economics,?18(1), 3-42.Dechow, P. M., Sloan, R. G., & Sweeney, A. P. (1995). Detecting earnings management.?Accounting Review, 70(2), 193-225.Desai, M. A., & Dharmapala, D. (2006). Corporate tax avoidance and high-powered incentives.?Journal of Financial Economics,?79(1), 145-179.Epe, P., & Koetzier, W. (2011). Jaarverslaggeving [theorieboek] Wolters-Noordhoff.Guay, W. R., Kothari, S. P., & Watts, R. L. (1996). A market-based evaluation of discretionary accrual models. Journal of Accounting Research, 34(2), 83-105.Hanlon, M., & Shevlin, T. (2005). Book-tax conformity for corporate income: An introduction to the issues. In?Tax Policy and the Economy, 19(2), 101-134. Hayes, R. S. 1., Gortemaker, J. C. A., & Wallage, P., (2014). Principles of auditing: An introduction to international standards on auditing (Thirdition. ed.). Amsterdam; Harlow, England: Prentice Hall, Financial Times.Healy, P. M., & Wahlen, J. M. (1999). A review of the earnings management literature and its implications for standard setting.?Accounting Horizons,?13(4), 365-383.Hill, C. W., & Jones, T. M. (1992). Stakeholder-agency theory.?Journal of Management Studies,?29(2), 131-154.Hribar, P., & Collins, D. W. (2002). Errors in estimating accruals: Implications for empirical research.?Journal of Accounting Research, 40(1), 105-134.Kahneman, D., & Tversky, A. (1979). Prospect theory: An analysis of decision under risk.?Econometrica: Journal of the Econometric Society, 47(2), 263-291.O’brien, R. M. (2007). A caution regarding rules of thumb for variance inflation factors. Quality & Quantity, 41(5), 673-690.Palepu, K., Healy, P., & Peek, E. (2010). Business analysis and valuation; IFRS-edition. South-Western Cengage Learning.Phillips, J. D., Pincus, M., Rego, S. O., & Wan, H. (2004). Decomposing changes in deferred tax assets and liabilities to isolate earnings management activities.?Journal of the American Taxation Association,?26(1), 43-66.Phillips, J., Pincus, M., & Rego, S. O. (2003). Earnings management: New evidence based on deferred tax expense. The Accounting Review, 78(2), 491-521.Picker, R., Leo, K., Loftus, J., Wise, V., Clark, K., & Alfredson, K. (2013). Applying international financial reporting standards (3rdition. ed.). Milton, Queensland: Wiley.Ronen, J., & Yaari, V. (2008). Earnings management. Springer US.Scott, W.R. (2012). Financial accounting theory, 6th edition. Toronto: Pearson Prentice Hall.Schrand, C. M., & Wong, M. H. (2003). Earnings management using the valuation allowance for deferred tax assets under SFAS No. 109*.Contemporary Accounting Research,?20(3), 579-611.Stolowy, H., & Breton, G. (2004). Accounts manipulation: A literature review and proposed conceptual framework.?Review of Accounting and Finance,?3(1), 5-92.Tang, T. Y. (2014). Does Book-Tax Conformity Deter Opportunistic Book and Tax Reporting? An International Analysis.?European Accounting Review, 1-ersky, A., & Kahneman, D. (1991). Loss aversion in riskless choice: A reference-dependent model.?The Quarterly Journal of Economics, 106(4), 1039-ersky, A., & Kahneman, D. (1992). Advances in prospect theory: Cumulative representation of uncertainty.?Journal of Risk and Uncertainty,?5(4), 297-323.Walker, M. (2013). How far can we trust earnings numbers? What research tells us about earnings management.?Accounting and Business Research,?43(4), 445-481.Watts, R. L., & Zimmerman, J. L. (1986). Positive accounting theory, The Accounting Review, 65(1), 131-156.Xu, R. Z., Taylor, G. K., & Dugan, M. T. (2007). Review of real earnings management literature.?Journal of Accounting Literature,?26(3), 195-228.Yamano, T., Lectures notes on Advanced Econometrics. Retrieved 2013.Yin, G. K. (2001). Getting serious about corporate tax shelters: Taking a lesson from history.?SMUL Rev.,?54, 209.AppendixAuthor(s)Object of the studySampleMethodologyOutcomeBlaylock, Gaertner, Shevlin (2014)Investigate the association between book-tax conformity and earnings management.Firm-year observations in Compustat’s Global Vantage files from1996 to 2007 with necessary data to compute our four earnings management variables.Correlation tests.Positive association exists between earnings management and book-tax conformity.Tang(2014)This article focusses on the relation between the book-tax conformity and opportunistically reporting of taxable income and financial profits.The sample is taken from the year 1995 till the year 2007 and yields 372 country-year observations across 32 countries.Regression. The main results obtained in this analysis suggest that conformity deters earnings management and tax avoidance.Phillips, Pincus, Rego(2003)The deferred tax expense in examining a potential relation with earnings management besides this objective the researchers also compare the usefulness of deferred tax expense over three specific accrual models in detecting. earnings management.A sample period which starts in 1994 and ends in 2000.Probit regression. They conclude that the deferred tax expense is more useful in detecting earnings management than accrual models are.Phillips, Pincus, Rego, Wan(2004)to find out whether certain deferred tax asset and deferred tax liability components are associated with earnings management.The sample is gathered from the Compustat database and the sample period consists of the years 1994 till 2000. The final sample consists of 396 firm-year observations.Probit regression. The authors find that changes in specific components of the deferred tax liability are significantly useful in detecting earnings management, whereas the total deferred tax liability is not. These specific components are the revenue and expense accruals and reserves. With these findings hypothesis 2 is confirmed. The other hypotheses remain unconfirmed due to insignificant relations.Coppens & Peek (2005)This study examines the shapes of and discontinuities in the earnings distributions of private firms by comparing this with the public firms as a benchmark.Large private firms in Belgium, Denmark, France, Germany, Italy, the Netherlands, Spain and the U.K. between 1993 and 1999Earnings distributionsThe results show that private firms manage earnings to avoid losses in both high alignment countries and in low alignment countries. However they fail to reject the second hypothesis, namely that private firms manage earnings to avoid earnings decreases.Schrand & Wong (2003)Whether banks do earnings management by setting a high valuation allowance associated with deferred tax assets and if banks adjust the valuation allowance in the subsequent periods. The final sample consists of 235 banks that adopted SFAS 109 in 1993.RegressionThe results show that valuation allowances against deferred tax assets are an earnings management tool used by banks. The sample is restricted to banks and it does not capture the behavior of other companies.Christensen et al. (2008)The researchers examine whether firms choose the level of the valuation allowance against deferred tax assets when they incur special charges like restructuring costs or write-offs.The sample exists of 444 bath firms and 444 control firms.The results for the first hypothesis are that there is no evidence on average that bigger bath firms establish larger than necessary valuation allowances in order to reverse those and increasing earnings in year +1. The results for hypothesis H1B are that the magnitude of a valuation allowance provides information about how it will perform in the future.Burgstahler & DichevIn this study researchers examine whether there is that firms do earnings management to avoid earnings decreases and losses.For the study Burgstahler and Dichev used all available observations of the annual industrial and research Compustat databases for the year 1976-1994 which meet minimal data requirements.Earnings distribution and statically test.Burgstahler and Dichev (1997) found that firms use earnings management to avoid earnings decreases and losses. ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download