Effect of Audit Quality on the relation between Internal Control and Earnings Management

Transcription

1 Effect of Audit Quality on the relation between Internal Control and Earnings Management Master thesis Judith van Ravenstein MSc Accountancy & Control Variant Accountancy Amsterdam Business School Faculty of Economics and Business University of Amsterdam First draft: May 27, 2012 Final draft: June 6, 2012 First supervisor: Dr. Georgios Georgakopoulos Second supervisor: Dr. Sander van Triest

2 Abstract Material weaknesses in the internal control system of a company creates more opportunities for managers to engage in opportunistic earnings management. This thesis investigates the relation between earnings management and disclosed material weaknesses in the internal control, both under SOX 302 and SOX 404, and looks at whether audit quality, measured as being audited by a Big Four auditor, has an effect on that relation. The results suggest that material weakness firms have more absolute discretionary accruals and greater incomedecreasing discretionary accruals. So evidence is provided that material weakness firms engage in more earnings management, however not in opportunistic income-increasing earnings management. When audit quality is high, measured as being audited by a Big Four auditor, the disclosed material weaknesses are lower just as total and absolute discretionary accruals. Interesting is the finding that when material weakness firms are audited by a Big Four auditor this will lead to positive discretionary accruals, suggesting that when a firm is audited by a Big Four auditor, material weaknesses in the internal control will lead to opportunistic earnings management.

3 Acknowledgement First of all, I would like to thank PricewaterhouseCoopers Accountants N.V. for giving me the opportunity to write my thesis at their office in Amsterdam during my internship. Specifically, I would like to thank my internship supervisor Jens de Waardt for giving me advice and suggestions on my thesis. Further, I would like to thank my Master thesis supervisor dr. Georgios Georgakopoulos from the University of Amsterdam for his comments and suggestions on my thesis. Also, I would like to thank dr. Peter Kroos for his feedback on the concept of my thesis. Finally, I would like to thank my parents for making my study possible, my fellow students and professors from the University of Amsterdam who have helped and supported me throughout my education. Specifically, I would like to thank Shivesh Changoer for sharing his knowledge of Stata, Bora Turktas for reviewing my thesis and Priscilla den Besten for moral support.

4 Table of contents I. Introduction 1 II. Literature Review 3 1. Capital Markets Agency problems Financial Accounting & Auditing 4 2. Earnings Management 6 3. Internal Control Material Weaknesses 8 4. Sarbanes-Oxley Act Auditor Audit quality Audit size Deep pocket theory Other factors 16 III. Research Methodology 16 Sample selection 16 Earning Management 16 Audit Quality & Internal Control 18 Empirical Specification 21 Auditor choice model 21 Accrual choice model 22 Material weakness model 23 IV. Empirical Results 24 Descriptive statistics & univariate comparisons 24 Internal Control 24 Audit Quality 27 Discretionary Accruals 29 Correlation 30 First-stage probit regression 35 Second-stage regression (OLS) 36 Logit regression 37 Sensitivity analysis 39 V. Conclusion 40 Appendix 43 Sarbanes-Oxley Act Section 302 and Explanation of variables 45 References 47

5 I. Introduction This paper will investigate the effect of audit quality, with auditor size as a proxy for quality, on the relation between internal control and earnings management. More specifically, it will examine to what extend the relationship between material weaknesses in the internal control and earnings management is affected by high audit quality. Several United States (U.S.) financial reporting scandals jeopardized the public s faith in the U.S. capital markets, with Enron as most familiar (Bédard, 2006). Already in the 1980s the existence of fraud and unexpected business failures led to questions among the adequacy of the financial reporting system and especially the internal controls of public companies (Ge and McVay, 2005). Corporate controls are the first line of defense against misstatements in the financial statements. Auditors provide another layer of investor protection by reducing the risk of misstatements (Hoitash et al. 2008). The public is concerned that earnings management erodes the quality of financial reporting and the Securities and Exchange Commission (SEC) questions the role of the auditor in this issue (Heninger, 2001). A higher audit quality usually will lead to less earnings management, since it limits the discretion that management can use to manage earnings in the financial statements (Becker et al. 1998). Yet Beneish et al. (2005) state that the direction of this causality between audit quality and earnings management is not clear, since firms with high quality financial reporting may choose for high quality auditors. On the other hand, high quality auditors are better able to detect manipulations and their actions result in higher earnings quality. Internal control weaknesses increase the risk of intended and unintended misstatements in the financial reports. Therefore it is assumed that this will increase the possibility of earnings management. For example, when there is an inadequate separation of duties, there is less control over the actions of employees and managers, so management has more opportunities to engage in earnings management. Auditing can be helpful in mitigating this problem, depending on the quality of the audit. First of all, high audit quality is expected to decrease earnings management, since high quality auditors are more likely to detect and reveal earnings management. This is due to their expertise, strictness and the fact that they have more to lose, such as reputation, when there are misstatements in the financial reports. Secondly, high quality auditors provide a better opinion on internal control, since it is expected that they are more likely to detect weaknesses in internal controls due to sufficient substantive testing. An auditor can provide an adverse opinion on the internal controls, yet provide an unqualified opinion on the financial statements as a whole. So this means that an auditor provides a reasonable assurance to the public that the financial statements are 1

6 presented fairly, even though there are material weaknesses in internal control. Since these internal control weaknesses can lead to opportunistic earnings management, it is interesting to investigate the role of audit quality on this relation. The research question of this thesis is therefore: To what extend does audit quality influences the relation between internal control weaknesses and earnings management in the post-sox period? Answering this research question has a contribution to society. Earnings management remains a popular topic of debate and discussion among investors, regulators, analysts and the public (Krishnan, 2003). Researchers try to understand why managers manipulate earnings, how they do so, and what the consequences are (McNichols, 2000). As Scott states (1997) an understanding of earnings management is important to auditors, because it enables an improved understanding of the usefulness of net income, both for reporting to investors and for contracting. This understanding of usefulness of net income is also of importance for investors, in order to know how much they can rely on net income to base their investment decisions on. Further it may assist auditors to avoid some of the serious legal and reputational consequences that arise when firms become financially distressed. Such distress is often preceded by serious abuse of earnings management, as we have seen at Enron. Since SOX was introduced to increase the independence and quality of auditors and internal control, it is relevant to investigate the relation between audit quality, internal control and earnings management after SOX was implemented. This has a societal contribution to regulators, as it shows whether the implementation of SOX is effective, and whether it nowadays still is efficient. Also, the examination of material weaknesses in internal control is informative for managers and auditors. This is because managers must identify material weaknesses within their firm and auditors must attest to the manager s report of internal control under Section 404 of SOX (Ge and McVay, 2005). Furthermore, the paper contributes to a broad research area, as the research question is related to three different research area s: financial accounting, internal control and auditing. Also this question has not been examined before in this type of setting. The effect of audit quality on earnings management has been examined before (Becker et al. 1998; Tendeloo and Vanstraelen 2005). This research will however use a different time area, use a performance adjusted Jones (1991) model to measure earnings management, and also takes internal control in consideration. Ge and McVay (2005) state that future research could help explore links between the disclosure of material weaknesses and fraud, earnings management or restatements, as this could provide insights to the benefits of SOX and Section 404 in particular. This paper will contribute to their research, since it specifically investigates 2

7 disclosures of material weaknesses under SOX 302 and SOX 404 and the relation with earnings management. To investigate the effect of audit quality will also contribute prior literature, since mixed results are found with relation to the disclosure of material weaknesses and audit quality. The remainder of the paper is outlined as follows. In the next section a literature review will be presented from which three hypotheses are developed. Section III provides the research design and describes the sample selection and empirical specification. Section VI describes the empirical results. Concluding comments are given in Section V. Finally, in the appendix explanatory information can be found. II. Literature Review This section provides a background of prior literature among the research topic. First the problems that arise in capital markets will be pronounced, how these problems could lead to earnings management and how financial accounting and auditing can prevent these problems. Than a brief explanation of earnings management will be given, including the different types of earnings management. Furthermore, the effect of internal control weaknesses on earnings management is pronounced, including a reference to the Sarbanes Oxley Act of Finally, audit quality is defined and it is shown how audit quality can reduce earnings management and affect internal control. 1. Capital markets 1.1. Agency problems In capital markets, principal-agency problems arise when there is a separation between ownership and control. The problem that arises is that the owner (principal) cannot monitor the actions of the manager (agent), but can only see the results of these actions namely the profit of the firm. The owner is therefore unable to know whether this profit is reliable, or whether the manager has for example engaged in earnings management in order to be able to report the profit (Grossman and Hart, 1983). Eisenhardt (1989) states that these agency problems arise when (1) the principal and agent have different types of interest, and (2) it is hard or expensive for the principal to monitor the actions of the agent, because of information asymmetry. Conflicts of interest arise when managers prefer higher rather than lower reported earnings, while stakeholders prefer a timely indication of potential problems (Heninger, 2001). Information asymmetry exists when a principal and an agent do not have the same 3

8 information, which leads to contracting problems such as moral hazard and adverse selection 1 (Scott, 1997). Several analytical models demonstrate that the extent of earnings management increases with the level of information asymmetry (Zhou and Elder, 2004). This is because shareholders cannot observe a firm s performance and prospects in an environment in which they have less information than management. In this condition management can use flexibility to manage the reported earnings and opportunistically manipulate accounting numbers (Piot and Janin, 2005). The Positive Accounting Theory (PAT) 2 predicts three opportunistic form hypotheses, indicating that managers choose accounting policies in their own best interests, which not necessarily be in the firm s best interests: 1.The bonus plan hypothesis. Managers of firms with bonus plans are more likely to choose accounting procedures that shift reported earnings from future periods to the current period, in order to maximize their remuneration. 2.The debt covenant hypothesis. The closer a firm is to violation of accounting-based debt covenants, the more likely the firm manager is to shift reported earnings from future periods to the current period. 3.The political cost hypothesis. The greater the political costs faced by a firm, the more likely the manager is to choose accounting procedures that defer reported earnings from current to future periods (Scott, 1997) Financial Accounting & Auditing Financial accounting can reduce information asymmetries that exist between managers and firm stakeholders by providing financial statements. Financial statements are a primary source of information in capital markets, which depend on high-quality information to function properly (Behn et al. 2007). Yet, these statements do not provide reliable information if management has engaged in opportunistic earnings management. Auditing can reduce this problem by providing an independent assessment of the accuracy and fairness with which financial statements represent the results of operations, financial position, and cash flows in conformity with Generally Accepted Accounting Principles (GAAP). Besides that, auditing also reduces information risk (Chen et al. 2010). High audit quality can mitigate negative 1 Moral hazard is a type of information asymmetry whereby one or more parties to a certain transaction can observe actions in fulfillment of the transaction but other parties cannot. Adverse selection is a type of information asymmetry whereby one or more parties to a certain transaction have an information advantage over other parties (Scott, 1997). 2 This theory is concerned with predicting such actions as the choices of accounting policies by firm managers and how managers will respond to proposed new accounting standards (Scott, 1997). 4

9 effects of information uncertainty, because audits increase precision and credibility of financial disclosures and well-capitalized audit firms provide potential indemnification against losses suffered due to potential misstatements. The role of auditing is to enforce the application of proper accounting policies. Managers however prefer discretion in the reporting process and auditors could have incentives to go along with earnings management behavior and the reporting of low quality earnings in order to avoid dismissal by clients (Francis and Wang, 2006). The effectiveness of auditing, and its incentives to constrain or go along with earnings management, is expected to vary with the quality of the auditor. High quality audits improve the reliability of financial statement information and allow investors to make a more precise estimation of the firm s value. In comparison to low-quality auditors, high-quality auditors are more likely to detect questionable accounting practices and, when detected, to restrict their use and/or to qualify the audit report. Thus, high-quality auditing can be an effective deterrent to earnings management because management's reputation is likely to be damaged and firm value is reduced if misreporting is detected and revealed (Becker et al. 1998). From this the first hypothesis is developed: Hypothesis 1: High audit quality will lead to less earnings management. Gul et al. (1999) support this hypothesis by concluding that a higher audit quality reduces insiders incentives to exploit accounting-based contractual constraints and manage earnings as a result of separation of ownership and control, which will increase earnings quality. A large separation of ownership and control, thus a low management ownership, will increase the discretion that management uses and lower earnings informativeness, because there is being more relied on earnings-based compensation for managers. A high management ownership, so a smaller separation of ownership and control, on the other hand will reduce principal-agency conflicts, since the interests between owners and managers are more aligned. When firms have sufficient incentives and opportunities to manage earnings, the value of auditing increases as it reduces managerial opportunities to use discretion in financial statements. The effect of audit quality on earnings management varies with a firm s incentive to manage accounting performance. So the more incentives, the higher the value of auditing. 5

10 2. Earnings management Earnings management is the direct intervention in the external financial reporting process, with the intention to obtain a private gain. Managers can intervene by modifying how they interpret financial accounting standards and accounting data, or by timing or structuring real transactions, which affect earnings in order to achieve some specific reported earnings objective (Scott, 1997). From this, a distinction can be made between accrual earnings management and real earnings management. Real earnings management refers to the purposeful altering of reported earnings in a particular direction by changing the timing or structuring of an operating, investing, or financing decision (Badertscher, 2011). Real operational activities to manipulate earnings are for example the decrease of discretionary spending on research and development, advertising and maintenance, or delay starting a new project to meet earnings target, even if this delay entailed a small sacrifice in value (Cohen et al. 2007). Accrual earnings management refers to the purposeful altering of accruals in a particular direction, either within-gaap or outside the boundaries of GAAP, that is achieved when managers adjust revenue or expense accruals to alter financial reports (Badertscher, 2011). This thesis will focus on accrual-based earnings management. Reported earnings are an interaction of income-reporting incentives faced by two issuers of financial statements, namely corporate managers and external auditors (Kim et al. 2003). Conflict or convergence of reporting incentives between these two issuers is an important factor in determining the effectiveness of external auditing for deterring opportunistic earnings management. Reported earnings consist of a cash flow part and an accrual part. Accrual-based earnings are superior to cash flows, because they overcome the timing and matching problems that cash flows face (Krishnan, 2002). This increases the ability of earnings to reflect underlying economic value. However, the accrual part of earnings requires managers to report the effects of economic transactions based on expected cash realizations. This gives managers some flexibility by using discretion in making numerous judgments and assumptions in reporting firm performance (Barton and Simko, 2002). Managers could abuse the flexibility permitted by GAAP by engaging in aggressive reporting of accruals that undermines the informativeness of reported earnings, which is misleading for investors (Krishnan, 2002). On the other hand, management can use this flexibility to communicate private and inside information to investors when contracts are rigor, which is informative for investors. Accruals can be divided in normal and abnormal accruals. Normal accruals arise in the ordinary course of a business and are unlikely to reflect managerial manipulations. Abnormal accruals are a manipulation of accounting numbers, which are also 6

11 called discretionary accruals (Heninger, 2001). Outsiders cannot directly observe earnings, so high accrual firms face greater agency costs, relative to low-accrual firms (Krishnan, 2002). Because of this investors and creditors demand a higher cost of capital in order to compensate the risk they face. Incentives for managers to engage in earnings management are for example a need to meet analysts estimates and influence stock markets to reach targets set by compensation contracts or debt covenants, as explained previous by PAT, to communicate economic information to stakeholders, and to smooth income or improve future income (Nelson et al. 2002). Current-income-increasing attempts are typically intended to benefit the current period, rather than enabling future-income-decreasing attempts. Current-income-decreasing attempts are typically intended to avoid political pressures or build reserves to use more future-income-increasing earnings management, the so called cookie-jar reserves (Nelson et al. 2002). Conservative auditors will be less tolerant in case of income-increasing earnings management and less strict toward income-decreasing accruals, because of the litigation risk that auditors face when a firm is overvalued (Piot and Janin 2005; Kim et al. 2003). 3. Internal Control The SEC defines internal control as: a process, effected by an entity s board of directors, management and other personnel, designed to provide reasonable assurance regarding the reliability of financial reporting (Doyle et al. 2006). For parties that have contractual relations with a company, reliable reporting is extremely important and effective internal controls are critical in achieving reliable reporting. Auditing Standard No. 5 3 states that effective internal controls are designed to provide reasonable assurance regarding the reliability of financial reporting and the preparation of financial statements for external purposes in conformance with GAAP (PCAOB, 2007, paragraph 2). Effective internal controls are a fundamental driver of earnings quality as shown by prior literature, since they curtail the intentional manipulation of information reported to outsiders, reduce the risk of random procedural and estimation errors in reporting, and mitigate the inherent risks of business operations and strategies that may affect the quality of reported information (Brown et al. 2008). Besides that, high-quality internal controls and have implications for management s oversight of the financial reporting process (Schneider and Church, 2008). A control environment of high quality establishes the importance of internal control throughout 3 Auditing Standard No. 5 replaced Auditing Standards No. 2 in

12 the firm. Potential causes of financial misstatements are assessed as to risk and control activities are designed and put in place to reduce important risks. Exceptions that arise are communicated to parties who can take corrective action. The internal control process is monitored for exception by management and an internal audit function (Ashbaugh-Skaife et al. 2008). When internal controls are not effective, we can speak of internal control weaknesses or deficiencies. There are two different internal control weaknesses as stated by Auditing Standard No. 5: 1. Significant deficiencies: control deficiency with more than a remote likelihood of not preventing or detecting a future misstatement that is more than inconsequential. 2. Material weaknesses: more than a remote likelihood of not preventing or detecting a material misstatement (Hoitash et al. 2008). The focus of this thesis will be on material weaknesses Material weaknesses A material weakness is defined as a deficiency, or a combination of deficiencies, in internal control over financial reporting, such that there is a reasonable possibility that a material misstatement of the company s annual or interim financial statements will not be prevented or detected on a timely basis (PCAOB, 2007, paragraph A7). Moody s, the bond rating company, divide material weaknesses into two categories: account-specific material weaknesses, which relate to control over specific account balances or transaction-level processes, and company-level weaknesses such as an insufficient segregation of duties. Most common account-specific material weaknesses occur in current accruals accounts, such as accounts receivable and inventory accounts (Ge and McVay, 2005). These types of material weaknesses are identifiable by auditors through substantive testing and thus do not represent as a serious concern regarding the reliability of the financial statements (Doyle et al. 2006). While account-specific weaknesses are auditable, company-level weaknesses are more difficult to audit around and call into question not only management s ability to prepare accurate financial reporting, but also its ability to control the business (Doyle et al. 2006). Also, not all possible misstatements can be uncovered, because fraud can be perpetrated through management collusion. Higher billing rates on risky engagements compensate firms for higher litigation insurance costs and the creation of sufficient financial reserves (Hoitash et al. 2008). 8

13 When there are material weaknesses it is likely that the quality of internal data that management uses to make decisions is low (Bryan and Lilien, 2005). This indicates that material weaknesses in internal control cast a bad light on the integrity of the financial reporting process and also on management s ability to run the business. However, it is likely that material weaknesses are disclosed in management reports, since it is a criminal offense for managers to conclude that controls are effective when they have knowledge of material weaknesses. As Section 906 of SOX states: criminal penalties can be imposed on managers who knowingly certify a periodic report that does not comport with the requirements. Allowable penalties include a fine of up to $5,000,000 and up to 20 years imprisonment (Ge and McVay, 2005). Material weaknesses in internal controls must also be disclosed in the auditor reports. If an auditor provides an adverse opinion on internal control, the reliability of financial data decreases and uncertainty increases, which impacts the creditworthiness of the firm (Schneider and Church, 2008). An adverse opinion means that at least one material weakness exists, whereas an unqualified opinion indicates that internal controls are effective. The disclosure of material weaknesses suggests that management s oversight may be lacking and, in turn, the risk of wrongly stated financial data will increase (Schneider and Church, 2008). Managers of firms with weak internal controls can more easily override the controls and intentionally prepare biased accrual estimations that facilitate meeting their opportunistic financial reporting objectives (Ashbaugh-Skaife et al. 2008). The quality of accruals will likely decrease, due to both intentional and unintentional misstatements. Unintentional errors due to lack of adequate policies, training, or diligence by company employees, such as pricing errors. Intentional misrepresentations or omissions by employees or management by engaging in earnings management by over- or understating earnings, such as cookie jar reserves. So, internal control weaknesses reduce the reliability of financial reporting, which increases information risk for investors and therefore the cost of equity (Ashbaugh-Skaife et al. 2009). Poor internal controls usually relate to insufficient commitment of resources for accounting controls. Ge and McVay (2005) categorize disclosed internal control problems into nine major deficiencies: account-specific, training (inadequate qualified staffing and resources), period-end reporting/ accounting policies, revenue recognition, segregation of duties, account reconciliation, subsidiary-specific, senior management and technology issues. Each material weakness firm, can have multiple deficiencies and thus multiple deficiency types. Overstating revenues is the most common method of earnings management. The presence of weak internal controls over the revenue recognition process provides managers 9

14 with more flexibility to manage earnings in this fashion (Ge and McVay, 2005). Revenuerecognition deficiencies are usually related to the timing of revenue recognition and other contracting practices. Deficiencies in revenue-recognition policies might facilitate earnings management through improper revenue recognition, such as channel stuffing 4. Channel stuffing at period-end is often motivated by incentives to meet or exceed analysts earnings expectations. Chan et al. (2007) found that firms reporting material internal control weaknesses under Section 404 of SOX have more earnings management compared to other firms. They provide mild evidence that there are more positive and absolute discretionary accruals for firms reporting material internal control weaknesses than for other firms. Since this paper investigates material weaknesses disclosed botch under SOX 302 and SOX 404, this leads to the second hypothesis: Hypothesis 2: Firms that disclose material weaknesses under SOX 302 or SOX 404 engage in more earnings management than non-material weakness firms. 4. Sarbanes-Oxley Act Corporate fraud, such as the Enron case and other accounting scandals, have jeopardized the auditing profession, which has shown the importance of auditing and internal control (DeFond and Francis, 2005; Cohen et al. 2007). This was the main reason why the SEC implemented the Sarbanes-Oxley Act in The goal of SOX is to improve the accuracy and reliability of corporate disclosure in order to restore investors confidence in the financial statements and the auditing profession. Improved disclosures of financial information and better corporate governance, including internal control over accounting measurement, recognition and disclosure processes are assumed to improve investor confidence (Bryan and Lilien, 2005). SOX mandates management evaluation and independent audits of internal control effectiveness. This is costly for firms but may yield benefits through lower information risk that translates into a lower cost of equity (Ashbaugh-Skaife et al. 2009). SOX makes internal control weaknesses regarding financial reporting of public companies available for the first time. According to Hoitash et al. (2008) these material weakness disclosures are associated with lower earnings quality, abnormally negative returns and higher risk. Two specific sections of SOX deal with internal control, namely Section 302 and 404. Section 302 focuses on the disclosure of controls and procedures and Section 404 focuses on internal control over 4 Sales are boosted as a result of shipping more products to subsidiaries or vendors than are needed. These excess products are often returned in subsequent quarters. This will inflate the distributor s accounts receivable balance. 10

15 financial reporting (Ge and McVay, 2005). SOX 302 requires management to certify in quarterly reports if the company has maintained effective disclosure controls and procedures. An auditor s evaluation of disclosure controls and procedures and attestation of the management s report is not required. Disclosure controls and procedures have a broader scope than internal controls over financial reporting under SOX 404, since disclosure controls and procedures include company disclosures not directly related to financial statements 5 (Chan et al. 2007). 5. Auditor Hogan and Wilkens (2008) state that earnings management may seem possible when there are internal control deficiencies. Yet they contend that for financial statements to be in compliance with GAAP, strong internal controls are not necessary. This is because financial reports are a joint product of management and an independent auditor, so audit quality also should be taken in consideration. For example, even when internal control weaknesses are identified, auditors can still provide an unqualified opinion by increasing their substantive testing. More specifically, when either inherent risk or control risk increases, auditors can reduce detection risk by increasing substantive testing in order to maintain a desired level of overall audit risk 6. So, the awareness of a material weakness in internal control will increase auditors effort by adjusting the audit program to perform additional substantive procedures in area s with weaknesses which will increase the probability that they will detect misstatements in the financial statements (Bédard, 2006). The need of reporting on internal control effectiveness, attested by the auditor is debatable. Generally Accepted Auditing Standards (GAAS) require the auditor to obtain a sufficient understanding of internal control, which is necessary to properly plan and conduct an audit. For example, the mix of tests of controls and substantive tests. Besides that, the auditor can issue an unqualified opinion on financial statements even if there are material weaknesses in internal control, which provides reasonable assurance, irrespective of the effectiveness of a company s internal control (Schneider and Church, 2008).Users of financial statements however may view an adverse opinion on internal controls as bad news. Schneider and Church (2008) believe that because dominant audit suppliers provide high quality services, society views Big auditors as better able to adapt to factors that increase 5 See Appendix for a full reference of Section 302 and 404 of Sarbanes-Oxley Act. 6 This is the risk the auditor faces that he provides an unqualified opinion on the financial statements even though there are misstatements in the financial statements. 11

16 the risk of conducting an audit. Client-specific characteristics, such as financial distress, managers incentives to engage in opportunistic behavior and internal control weaknesses, may create concerns about the reliability of financial data. These concerns can be taken away by a Big auditor as they provide greater monitoring strength, due to their competence and independence. Krishnan (2005) agrees on this by saying that Big auditors may do better than non-big auditors in detecting and reporting internal control issues. Failure by the auditor to identify an existing internal control problem relating to financial reporting would likely reflect the quality of the audit. So, Big auditors are more likely to identify internal control weaknesses than non-big auditors. Ge and McVay (2005) find that being audited by a large audit firm is positively associated with the disclosure of a material weakness. This could be due to the fact that large auditors are exposed to a greater legal liability and they might be more diligent about searching for and reporting material weaknesses. Besides that, they have greater resources to identify material weaknesses and to react to recent changes in the regulatory environment, such as being subject to annual reviews by the PCAOB. If larger audit firms historically imposed stronger internal control standards for their clients, Ge and McVay (2005) expect to see fewer weaknesses disclosed under SOX Section 302 and 404. From this the final hypothesis is developed: Hypothesis 3: High audit quality will decrease material weaknesses in the internal control, both under SOX 302 and SOX Audit quality Piot and Janin (2005) define audit quality as the joint probability that the external auditor detects an anomaly in the financial statements, and then reveals it to the users of these statements. Wherein the probability of detection deals with the matter of competence and the probability of revelation depends on the independence of the auditor. This shows the willingness of the auditor to face the pressure exerted by the producers of the financial statements. Schwartz (1997) mentions that because a higher audit quality provides better information to investors, the legal regime that induces the highest audit quality also generates the most efficient investment. This is because an independent auditor reduces uncertainty about future cash flows. Audit effectiveness depends on the accuracy of auditors non-error frequency knowledge. From this Krishnan (2003) makes the assumption that auditors industry-specific knowledge is associated with audit effectiveness. Besides having the resources and expertise to detect earnings management, specialist auditors who enjoy a brand- 12

17 name reputation have particular incentives to deter and report questionable or aggressive accounting practices. While auditing is valuable in controlling managerial discretion, its value is expected to vary with the quality of the audit firm according to Becker et al. (1998). Higher-quality auditors are expected to be less willing to accept questionable accounting methods and are more likely to detect and report errors and irregularities. Behn et al. (2007) assume that audit quality is positively related to unobservable financial reporting reliability, which is congruent with Davidson and Neu (1993) who state that higher audit quality increases the credibility of financial information. They provide evidence that analysts earnings forecast increases and forecast dispersion decreases for firms that are audited with high quality, indicating that earnings are more predictable. They define high audit quality as big auditors and industry specialist non-big auditors. Forecasts are used as a proxy for capital markets earnings expectation when examining the information content of accounting information (Behn et al. 2007). Kim et al. (2009) state that analysts are less likely to follow firms with weak internal controls, since this causes analysts forecast error and dispersion. But, high quality audits increase reporting reliability by reducing both intentional and unintentional measurement errors in historical earnings (Becker et al. 1998). Therefore it is likely that analysts forecasts are more accurate and less dispersed when audit quality is high (Behn et al. 2007) Audit size Several researches have been done to the relation between auditor size and audit quality and they all determine Big audit firms as high audit quality and non-big audit firms as lower audit quality (Becker et al. 1998; Behn et al. 2007; Chen et al. 2010; Zhou and Elder 2004; Lennox 1999; Deis and Giroux 1992; Chen et al. 2005; Davidson and Neu 1993; DeFond and Francis 2005; Francis and Yu 2007). Brand name auditors are viewed as providing higher quality audits based on their perceived competence and independence. Big audit firms spend heavy on auditor training facilities and programs, which increases their competence. Besides that, their size and large portfolio of clients, gives them presumably the financial strength to stand up to, or walk away from, a client if this is necessary to stay independent and protect their reputation from a possible loss (Behn et al. 2007; Krishnan, 2003). DeFond and Francis (2005) however say that Big as well as Small audit firms perform a potential competent audit in accordance with GAAS. All public accounting firms must comply with minimal professional standards, so size does not matter according to them. Yet, Big audit firms voluntarily invest in higher levels of expertise and have more incentives to provide higher 13

18 quality audits to protect their reputation (Behn et al. 2007). Schneider and Church (2008) on the other hand argue that Arthur Anderson has contaminated the reputation of the remaining Big Four audit firms. Chen et al. (2005) agree on this by saying that audit quality is of higher concern after the fraud at Enron and the collapse of Arthur Anderson that followed. Large audit firms have brand-name reputation, charge higher audit fees and/or behave qualitatively differently from smaller audit firms, since smaller audit firms have less litigation risk (Behn et al. 2007). Francis and Yu (2007) furthermore state that larger audit firms are more likely to detect material problems in client financial statements, which results in a higher audit quality. According to Becker et al. (1998) auditors are more likely to object to management's accounting choices that increase earnings (as opposed to decrease earnings) and that auditors are more likely to be sued when they are associated with financial statements that overstate earnings (as compared to understate earnings). Becker et al. (1998) states that Big auditors limit the discretion of management to manage earnings in the financial report in the pre-sox period. They treat Big auditors as a homogeneous group in terms of audit quality, while the findings of Krishnan (2003) demonstrate that audit quality varies even among Big auditors. Furthermore, Big auditors more likely have higher audit quality since it is expected that earnings quality increases when a firm is audited by a Big auditor. Teoh and Wong (1993) provide evidence for this by showing that firms audited by a large auditor have a significant higher earnings response coefficient (ERC), which is a measure of the extent to which new earnings information is capitalized in stock price. A higher ERC means that earnings have a greater effect on investors valuations of the firm when numbers are perceived as more accurately reflecting true economic value. ERC can be seen as a proxy for earnings quality (Ghosh and Moon, 2005). Finally, Deis and Giroux (1992) explain the fact that larger auditors provide higher audit quality with reputation and power conflict. When a client exerts pressure on the auditor to violate professional standards a power conflict exists. Large, financially healthy clients can exert greater pressure with a threat of replacing the auditor. The client can do this since it is easier and less costly for him to replace the auditor, than it is for the auditor to replace lost business. Big auditors have higher audit quality that provide assurance of a greater conformance with GAAP. Likewise, failure to report a problem when identified would depend on auditor quality and independence, and on the auditor's concern with potential litigation and loss of reputation. This litigation and reputation risk is higher for Big auditors, since they have more to lose because of their size (Lennox, 1999). This will be further explained in the next section by using the deep pockets theory. Taken the 14

19 aforementioned all together, this paper will use Big Four auditors 7 as high audit quality and non-big Four auditors as low audit quality Deep pocket theory Deep pockets theory suggests that investors will sue auditors in case of client business failures to recover losses. Auditors may incur reputation losses in these cases. Higher litigation risk and reputation costs may drive auditors to avoid auditing specific clients or to change a risk premium. Auditors are only sued for issuing reports that are insufficiently conservative (type I errors), they are never sued for being too conservative (type II errors). Gul et al. (1999), just as Kim et al. (2003) mention that Big auditors are more conservative. Auditor conservatism means that auditors prefer income-decreasing accounting choices instead of incomeincreasing accounting choices (Kim et al. 2003). So Big auditors allow less flexibility in income-increasing accruals than non-big auditors. This makes sense since Nelson et al. (2002) found that auditors risk of litigation is highest for overstatements of current income and equity. Besides that, large auditors have more incentives to be accurate than small auditors, since they are more likely to be sued when a type I error occurs since they are more prone to deep pockets court actions (Lennox, 1999). Heninger (2001) found that the risk of auditor litigation is positively associated with a sharper measure of earnings management and abnormal accruals. This means that the risk of auditor litigation increases as clients report more positive (income-increasing) abnormal accruals. External stakeholders expect from the independent auditor to decrease earnings management and ensure a fair financial reporting. Stakeholders are more likely to sue the auditor if they perceive a failure in the financial reporting. Auditors face a higher litigation risk if they fail to attenuate managers propensity to manipulate abnormal accruals to portray an overly favorable impression of the firm s financial health (Heninger, 2001). When the allegation of audit failure arises, Big auditors are likely to face greater publicity in the financial media and thus bear greater reputation loss. Besides that they have deeper pockets, they also have higher insurance coverage, which means they have more resources to recompensate the plaintiffs through out-of-court settlements or through court-awarded damages or costs. For this reason, it is less likely that non-big auditors are being sued, since the expected benefits from the settlement of lawsuits may be insufficient to cover associated costs (Kim et al. 2003). A lower litigation risk causes a higher tolerance of auditors on 7 Big Four auditors: Deloitte, Ernst&Young, KPMG and PricewaterhouseCoopers. 15

20 opportunistic accounting practices by management. Lennox (1999) explains the deep pocket model as auditors having different wealth levels. A higher wealth causes more incentives to issue accurate reports, because they will suffer more from litigation penalties. He also makes the assumption that effort is privately observable and costly to the auditor, due to the moral hazard problem. Therefore he assumes that the threat of litigation, rather than the loss of reputation, deters an auditor from shirking Other factors Besides the size of the audit firm, there are other factors that affect audit quality. Piot and Janin (2005) state that the tenure of the auditor-client relationship and the quality of an audit committee also influence audit quality, because this threats the independence due to personal ties and familiarity which reduces the vigilance of the auditor. Besides that, the audit engagement may become routine over time and the auditor will devote less effort to identify weaknesses of internal control and risk sources. Ghosh and Moon (2005) agree on this by saying that audit quality decreases as the tenure extends, because the longer the tenure, client firms have greater reporting flexibility and earnings forecast errors decline. Audited financial statements are less reliable for investment and credit decisions if users of financial statements view lengthy tenure as having an adverse effect on auditor independence and audit quality. Hoitash (2007) has the opinion that audit fees also can affect audit quality. On one hand large fees paid to auditors increases the effort of the auditor, which will lead to a higher audit quality. On the other hand, large fees particularly provided for non-audit services will lower the independence of the auditor, which causes audit quality to decrease. Nelson et al. (2002) state that audit fees increase with client size, so Big auditors overall will receive higher audit fees. III. Research methodology Sample selection Earnings Management To investigate the research question, an archival research method will be done. A sample will be selected of U.S. firms in the period by using Compustat 8. Information about the financial statements is gathered from the fundamentals annually database. From the Segments (Non-Historical) database information about firm complexity is gathered, such as the number 8 See Appendix for a full reference of data collected from Compustat, including short notes used by Compustat, an explanation of variables and remaining calculations. 16

21 of operating segments and the report of a foreign currency translation. Data is being processed by using Stata Firms that do not have their fiscal year end at December 31 are being excluded from the sample. Also firms with missing data and non-positive assets are being deleted. Just like prior research, companies from the financial industry are being deleted because discretionary accruals estimation is problematic for these firms. Financial industries are companies with a SIC 9 of (DeFond and Subramanyam 1998; Krishnan 2002). To mitigate the problem of outliers, outliers are being filtered out of the data by winsorizing the variables to the 1 st and 99 th percentiles of their distributions. After that, SIC two-digit industries with less than 20 observations each year are also deleted to obtain meaningful cross-sectional estimates of the parameters. This leads to a final 36,935 firm year observations. The focus here is on accrual-based earnings management, so first total accruals are being calculated. Following prior literature (Kim et al. 2003; Tendeloo and Vanstrealen 2005), total accruals are being calculated by:. Where, for firm j in year t: = total accruals; = change in current assets; = change in cash and cash equivalents; = change in current liabilities; = change in long-term debt included in current liabilities; = change in taxes payable; and = depreciation and amortization expenses. To separate total accruals into non-discretionary and discretionary accruals the performanceadjusted Modified-Jones (1991) model is used: Where, for firm j in year t: 9 SIC is a firm s four digit Standard Industrial Classification code. 17

22 = total accruals; = lagged total assets; = change in sales revenue; = change in receivables; = gross property plant and equipment; = income before extraordinary items; = unspecified random factors. This model is designed to eliminate the conjectured tendency of the Jones Model to measure discretionary accruals with error when discretion is exercised over revenues. This modified version of the Jones Model implicitly assumes that all changes in credit sales in the event period result from earnings management. The reason behind this is that it is easier to manage earnings by exercising discretion over the recognition of revenue on credit sales than it is to manage earnings by exercising discretion over the recognition of revenue on cash sales. This paper will use the performance-adjusted Jones (1991) since, it is important to control for the impact of financial performance on accruals as evidenced by prior research (Bédard 2006; Cahan et al. 2011; Cahan and Zhang 2006; Chen et al. 2010; Kim et al. 2003; Kothari et al. 2001). These researchers state that the regular modified Jones model is misspecified for firms with extreme performance so prior research is followed by using an alternative model which includes return on assets (ROA) as an extra independent variable in the modified Jones (1991) model for discretionary accruals. The discretionary accrual measure is adjusted for the accrual performance of a matched firm where matching is on the basis of return on assets and industry. Kothari et al. (2001) suggest that performance matching is crucial to the design of well-specified tests based on discretionary accruals, since it is of critical importance to control for the effect of past performance in tests of earnings management. By using a performancematched discretionary accrual measure, more reliable inferences will be able to draw. Audit Quality & Internal Control Information about the auditor and internal controls of firms are gathered from AuditAnalytics. The specific databases used are: SOX 302 Disclosure Controls and SOX 404 Internal Control. SOX 302 requires management to certify in quarterly reports if the company has maintained effective disclosure controls and procedures. SOX 404 requires issuers to publish information in their annual reports concerning the scope and adequacy of the internal control structure and 18