The market reactions to internal control
weakness disclosures under SOX section
404
---Accelerated filers versus Non---Accelerated filers
Author: Lisa Rodenhuis Student number: 10422471
MSc Accountancy & Control Track Accountancy
Amsterdam Business School Faculty of Economics and Business University of Amsterdam
Date: August 15th 2014
Table of content
1. Background ... 4
2. Literature review and hypotheses development ... 7
2.1 Agency theory... 7
2.2 Regulation of Internal Accounting Controls ... 8
2.3 Expected benefits and costs of SOX 404 ... 9
2.4 Market reactions to ICMW disclosures under section 404 ... 10
2.5 Hypothesis development ... 11
3. Methodology ... 14
3.1 Accelerated filers ... 15
3.2 Non-accelerated filers ... 16
3.4 Cumulative Abnormal Returns ... 17
3.4 Control variables... 17
3.5 Research design ... 18
3.5 Sample selection ... 20
4. Descriptive statistics ... 22
4.1 Dependent and independent variables ... 22
4.2 Correlation matrix... 24
5. Empirical results ... 26
5.1 Hierarchical multiple regression ... 26
5.2 Tests of hypotheses ... 26
6. Conclusion ... 31
7. References ... 33
ABSTRACT:
This study examines the market reactions to the Internal Control Material Weakness (ICMW) disclosures under Section 404 of the Sarbanes-Oxley Act. A sample of 741 firm-year observations is used between 2011 and 2014. The sample consists of accelerated and non-accelerated filers who make audited and unaudited disclosures, respectively. I hypothesize that the market reactions are less for disclosures of the accelerated filers compared to the disclosures of non-accelerated filers. I find no significant relation between the disclosures of the accelerated filers and abnormal returns, neither between the non-accelerated filers and abnormal returns. This finding can be explained by the possibility that investors have difficulties with fully understanding the implications related to the disclosures. Further, I hypothesize that the abnormal returns will be positive when the auditor issues an unqualified opinion on management’s assessment on internal controls in relation to when the auditor issues an adverse opinion. I find a significant positive relationship between the auditors opinion and abnormal returns. This evidence supports my hypothesis and suggests that the market reacts positive on ICMW disclosures that include an unqualified auditor’s opinion.Keywords: SOX 404, Internal Control Material weakness disclosures, Abnormal returns, Market reactions, Auditor attestation, Accelerated filers, Non-accelerated filers.
1. Background
In 2002, Paul Sarbanes, an American Democratic Senator, proposed a new law for corporate governance. However, considering the drastic changes this new law would entail, the Senate rejected it. Shortly after, Michael Oxley, colleague of Paul Sarbanes, came with a revised version of this new law. Although this version was less drastic, the Senate rejected it as well. Then, large accounting scandals like WorldCom and Enron came to light and it changed the tone. These scandals led to the enactment of the Sarbanes-Oxley Act (SOX) to improve the quality of financial reporting. In addition, SOX has to re-establish the investors’ confidence in the reliability of financial statements. The Securities and Exchange Commission (SEC) enforced the new law. This new law exists of two provisions. The first, Section 302 of the Act, requires that chief financial officers and chief executive officers regularly evaluate the effectiveness of internal controls and must give a conclusion about their effectiveness (SEC, 2002). The second, Section 404 of the Act, requires management to disclose an internal control report in the annual report of the company that contains: (1) the overall conclusion of management on the internal control structure; and (2) an assessment made by an external auditor on management assessment of the effectiveness of the company’s internal controls over financial reporting (ICFR) (SEC, 2002).
SOX 404 of SOX raised a lot of concerns among the companies, especially the smaller public companies (non-accelerated filers). These concerns came from the notion of the high compliance costs of implementing SOX 404 which would put a heavy burden on the smaller public companies. Many studies researched the costs and benefits of SOX and these studies all show that the compliance costs of SOX 404 are high (Eldridge and Kealey, 2005; Iliev, 2010; Raghunandan et al, 2006). Since SOX was enforced in 2002, the smaller public companies were extended delays for enforcing SOX 404(b), that requires an external audit on management’s assessment. On July 21, 2010, President Barack Obama permanently exempted smaller public companies from SOX 404(b) (Tilton and Lieberman, 2011). Although the smaller public companies are exempted from SOX 404(b), these companies are still required to give an evaluation of the internal controls over financial reporting under SOX 404(a) and to disclose the conclusion of this evaluation in their annual report.
Since the compliance costs are high, it is important to know what the benefits are of the internal control disclosures. Hammersley, Myers & Shakespeare (2008) assess whether the disclosures under SOX 302 are useful to investors. They look at the stock price reaction to management’s disclosure of internal control weaknesses and to the characteristics of those 4
weaknesses. The findings suggest that the market reaction differs with the severity of the internal control weakness. They do say that the disclosures are informative and used by investors to revise their expectations about firm value. Also, regardless of the severity of the weaknesses, investors are concerned when a weakness occurs as it may impact the auditor’s ability to perform a good audit. Similarly, Francis and Ke (2006) mention that a significant stock price decline would say that characteristics of an internal control weakness disclosure would have caused investors to change their opinion about management’s oversight over financial reporting or to change their perceptions of the quality of the accounting information system. This in turn will lead to a reconsideration of expectations regarding the firm’s future probability or the risk associated with the firm.
Orenstein (2004) contradicts the results of Hammersley, Myers & Shakespeare (2006) and mention that auditors believe that investors will not fully comprehend the implications regarding the internal control weakness disclosures.
All studies, mentioned above, look at the weakness disclosures under Section 302 and Section 404 and investigate the market reactions considering the characteristics of the material weaknesses and whether the internal control over financial reporting reports are audited or unaudited. However, these studies do not reflect whether or not an audited internal control report has an adverse or unqualified auditor’s opinion regarding management’s assessment on ICFR. Ettresge et al. (2011) investigate adverse auditors’ opinions on clients’ ICFR, under SOX 404, and the impact on auditor dismissals. The authors only consider the adverse auditors’ opinions, and examine whether companies are more likely to dismiss their auditor after receiving an adverse auditors opinion on management’s assessment of ICFR. Their results show that clients are more likely to switch to higher-quality auditors when they receive an adverse auditors opinion in contrast to when they receive an unqualified auditor’s opinion.
While there are prior studies who research the market reactions to SOX 302 and SOX 404, this research will solely focus on internal controls material weakness (ICMW) disclosures under SOX 404 and what its effect is on the market. Hereby, the accelerated filers and the non-accelerated filers are considered as seperate groups. SOX 404(b) requires an auditor’s attestation report which can have either an unqualified or an adverse opinion. When there are one or more material weaknesses identified in the company’s ICFR, the external auditor is required to give an adverse opinion, otherwise an unqualified opinion is given. In this report, the auditor’s opinion will be considered when measuring the market reactions. The 5
research question addressed in this study is formulated as follows: What is the effect of internal control material weakness disclosures under SOX 404 on market reactions?
In this research I test two hypotheses. I predict that disclosures of accelerated filers that include an auditors opinion on internal controls have less impact on the market than disclosures of non accelerated filers that do not have an auditors opinion on internal controls. I expect this because prior literature (Doyle et al, 2007b; Beneish et al, 2006) indicate that accelerated filers operate in richer information environments than non-accelerated filers and this makes it harder to detect a market reaction to disclosures. I do not find evidence for this hypothesis as the results do not show a significant relation between non-accelerated filers and abnormal returns. Further, I predict that abnormal returns are positive when an auditor issues a unqualified opinion compared to when an auditor issues an adverse opinion which results in negative abnormal returns. I find evidence that supports the second hypothesis, which implies that investors consider an auditor’s opinion on internal controls valuable and that it affects their confidence.
The most important reason for implementing SOX 404 is to enhance the investor’s confidence. Previous studies have researched whether or not SOX 404 is beneficial to the investor’s or not, since the compliance costs are very high. They have done this by examining the market reactions to the ICMW disclosures under SOX 302 and 404. Therefore, this study contributes to the questions raised around SOX 404, as it compares disclosures of non-accelerated filers that do not have an auditors opinion on management’s assessment on internal controls with disclosures of accelerated filers that do have an auditors opinion. From an academic point of view this research will add to prior literature by looking at both an adverse auditors’ opinion and an unqualified auditors’ opinion on internal controls and investigate what the impact of these two situations is on investors. The auditors’ opinion is to enhance the credibility with the investors and this research has to show whether an auditor opinion is important for the investors and if it changes their believes about a company. The remainder of this paper progresses as follows. Section two of the paper provides existing literature and describes the hypotheses development. Section three contains the methodology used in this report. It shows the relationships tested for this research, explains the research design used and the sample selected to conduct the study. Section four describes the descriptive statistics of the variables together with the correlation matrix. Section five demonstrates the empirical results which is followed by my conclusion in section six.
2. Literature review and hypotheses development
In this following section prior literature is discussed that conducted research about SOX 404 disclosures. The results of these literature is presented to see how the market reacts to these disclosures and if there are noticeable differences. Furthermore, the hypotheses are formed based on the existing literature and their results.
2.1 Agency theory
The relationship between principals (i.e. shareholders) and agents (i.e. management) is defined as agency theory (Jensen & Meckling, 1976). Jensen & Meckling (1976) describe the agency relationship as a contract under which the principal engage the agent to perform some service on their behalf which involves delegating some decision making authority to the agent. This separation of ownership and control can lead to agency problems. The agency theory explains that these problems arises when the desires or goals of the principal and agent are in conflict and the principal cannot validate what the agent is actually doing. This situation can be described as moral hazard, where separation of ownership and control makes it difficult to know it interests are aligned. Moral hazard is an information asymmetry and it occurs when one party (i.e. management) may have an information advantage over others (i.e. shareholders).
Many studies discuss the principal-agent problem and how to deal with this problem. Stiglitz and Weiss (1981) indicate that greater information flows can mitigate the principal-agent problem. In the same way, Healy and Palepu (2001), hypothesize that when relevant information is disclosed, investors get the opportunity to directly monitor a firms’ operations and in this way can see whether managers act in the best interest of the shareholders. The authors predict, with this monitoring hypothesis, that firms who are less transparent and engage in fewer disclosure practices have more severe agency problems. Although, some prior literature discusses disclosure benefits in terms of monitoring opportunities for the shareholders, it also mentions other benefits of relevant information disclosure. Barry and Brown (1986) mention in their study that investors have a better chance of estimating the parameters of an asset’s return, which lowers the cost of capital, when there is a greater transparency. Another study of Diamond and Verrecchia (1991) show that better disclosure of information decrease the information asymmetry factor of the cost of capital.
The implementation of the Sarbanes-Oxley Act (SOX) led to mandatory reporting on internal control. The adoption of this Act can be explained by the agency theory, as the 7
disclosures under SOX should show the principal (shareholders) that the agent (management) is acting in their best interests. These disclosures should also reduce the information asymmetry between management and the shareholders. Hunziker (2013), mentions that reporting on internal control can be considered as a monitoring function to reduce conflicts between shareholders and management. Similarly, Huang & Zhang (2008) find that extensive disclosure enhances external monitoring for shareholders and prevents managers to act in their own interest. The study of Qian, Strahan and Zhu (2009) investigate the economic benefits of SOX and they find in their study that SOX reduces agency problems. This finding is contradicted by the study of Hermalin and Weisbach (2010), who show that increased disclosure have two results. More information give shareholders the opportunity to make better decisions, however it can also lead to additional agency problems and other costs for shareholders.
2.2 Regulation of Internal Accounting Controls
The Internal Control Material Weakness (ICMW) disclosures under SOX serve as a better regulation of internal accounting controls. This Act has been implemented, in 2002, after numerous accounting scandals came to light. The objective of SOX is to increase the reliability and the accuracy of corporate disclosures, to improve the corporate governance and to restore the confidence of the investors in U.S. capital markets. The implementation of SOX has significantly changed the information environment for public companies as it requires companies to make disclosures about its internal controls.
The disclosures are meant to inform investors about weaknesses in the disclosing firms’ internal control systems. When the firms have weak control mechanisms in place it can increase measurement errors and managers’ ability to manage earnings. If this happens then the disclosures will classify firms with low quality financial reporting. If there is uncertainty about the financial reporting quality, investors require compensation according to previous studies. These studies mention that uninformed investors have to deal with adverse selection problems when trading securities in the market (Easley and O’Hara 2004; Francis et al. 2005; Ecker et al. 2006; Lambert et al. 2007a, Lambert et al. 2012).
When considering the weakness disclosures and the possibility that investors require compensation for when there is uncertainty about a firms financial reporting quality, then the disclosure of certain weaknesses in a firm can cause investors to change their beliefs upward 8
about firm risk. Thus, the weakness disclosures can have a significant effect on the investors’ actions and therefore the market.
The Act has two parts, SOX 302 and 404. When SOX became effective in 2002, SOX 302 was the first part that covered the act. This part requires managers to attest to the audit committee that deficiencies and weaknesses are privately reported. Furthermore, managers need to disclose the material weaknesses and material changes in the internal control to the public. SOX 302 does not need to be verified by an external auditor and it only needs to disclose the material weaknesses of which CEO and CFO know about, so it can be said that these disclosures are a task of managers’ attention to detail in identifying the material weaknesses and their discretion used for the disclosures.
SOX 404, became effective on November 15, 2004. Under SOX 404, management of companies are required to detect and disclose material weaknesses of its internal controls. These disclosure should be annually reported in an internal control report in which management makes an evaluation of the effectiveness of the internal controls. Another requirement of this section is that an external auditor gives an opinion on managements statement, which is also included in the annual internal control report.
2.3 Expected benefits and costs of SOX 404
SOX was implemented to protect investors and the main benefits from implementing Section 404 of SOX are enhanced investor confidence in financial reports, increased reliability of the financial data based on the effective internal control systems that control the data, improving corporate disclosure and the detection of financial fraud. Considering these benefits, it is difficult to measure them.
When SEC proposed SOX Section 404, its implementation have become subject of discussion. According to previous studies, the costs of SOX 404 are extremely high (Foster et al. 2007; Eldridge and Kealey, 2005; Krishnan et al. 2008). These high costs are strongly criticized in terms of high audit fees and high compliance costs. Due to the concerns on SOX 404 the Security Exchange Commission (2006) gave an extension to smaller public companies (the non-accelerated filers) of compliance to section 404(b), which requires the evaluation of an external auditor, since high costs were associated with this requirement. First, the extension of compliance for non-accelerated filers was for a short period, a 5-months extension, however in 2010, President Barack Obama, signed into law the Dodd-Frank Wall 9
Street Reform and Consumer Protection act. A part of this act was that it permanently exempts small public companies with less than $75 million in market capitalization from the requirements of Section 404 (b) of SOX (Tilton et al. 2011). This was the consequence of an ongoing debate that small businesses play a critical role in the economy and that SEC had to implement Section 404 in a way that would decrease the burden that was now placed on small and mid-size public companies due to the high costs. From this data forward only accelerated filers are obliged to comply with Section 404 (b).
2.4 Market reactions to ICMW disclosures under section 404
The implementation of SOX 404 has led to many discussions as to what the effect of this act is on the capital market of disclosure firms. It is important to know what the effect is, because shareholders are important stakeholders for a firm and their actions can influence firm value. Prior literature already has a few studies that have researched the market reactions to the material weakness disclosures of SOX and the results of these studies are contradictory. Some literature suggests a positive market reaction to the disclosures under SOX, because the disclosures have a net beneficial effect on the perceived quality of financial reporting (Jain and Rezaee, 2006; Jain et al. 2008; Li et al. 2008). Botosan (1997) talks about high disclosure quality and relates it to lower costs of capital. Further, Francis et al. (2004) relates poorer earnings quality with higher costs of capital. However, prior literature gives mixed results whether prior disclosure of a material weakness gives incremental explanatory power for cost of capital levels. The results of Ogneva et al. (2007) suggest that SOX 404 disclosures give little or no explanatory power for cost of capital levels. However, the study of Ashbaugh-Skaife et al. (2007) find evidence that there is a positive relation between weakness status and cost of capital levels obtained from the disclosures of SOX 302 and 404. Although, some literature discusses an association between weakness status and cost of capital levels, this does not immediately imply that there is a market reaction to the disclosure of internal control weaknesses.
In contrary, Orenstein (2004) states that auditors believe that investors will not fully comprehend the implications regarding the internal control weakness disclosures. Additionally, it is argued by regulators that the disclosures will not necessarily induce reactions by investors (Nicolaisen, 2004).
Beneish et al (2006) investigates market reactions to SOX 302 and 404 disclosures to 10
assess the effects of these disclosures on risk. The authors look at unaudited material weakness disclosures under SOX 302 and audited material weaknesses disclosures under SOX 404. Although firms that disclose under SOX 302 are unaudited, they are required to hire an auditor to audit the financial statements. SOX 404 explains that larger public companies need to obtain a separate auditor opinion on managements assessment of their internal controls. When a high quality auditor is used to perform the audit it may mitigate the market reaction. The results of this study indicate that there are no abnormal returns and that there is no evidence that the cost of capital changed when the firm made SOX 404 disclosures. These results are in line with Doyle et al. (2007b) who discusses that audited internal control disclosures mirror a lower materiality threshold for disclosure.
2.5 Hypothesis development
The goal of SOX 404 is to improve investors confidence in the financial statements and to inform them about internal control weaknesses. Firms with a weak internal control systems have an increase in measurement errors or an increase in managers’ ability to manage earnings. Thus, the weakness disclosures may indicate that these firms experience low quality financial reporting. Recent studies show that when a firm is experiencing uncertainty about its financial reporting quality investors may require compensation. Further, these studies discuss that when investors are uninformed they face an adverse selection problem in trading securities in the capital market (Easley and O’Hara 2004; Francis et al. 2005; Ecker et al. 2006; Lambert et al. 2007a, 2007b). If uncertainty about a firms financial reporting quality ensures that investors require compensation, then it is likely that the weakness disclosures made by firms have the potential to cause investors to change their believes about firm risk upwards. Thus, expecting that disclosing firms will experience both an increase in their cost of capital and negative abnormal returns in response to lower perceived reporting quality. When investors change their believes this is either because the disclosures convey new information to the market or that these disclosures make sure that investors place more weight on value-relevant information previously overlooked in valuation (e.g., Ou and Penman 1989; Sloan 1996).
Non-accelerated filers (smaller public companies) are exempted from the auditor attestation requirement under Section 404(b) as the costs outweigh the benefits. Studies show that disclosures made by non-accelerated filers have a different effect on the market and 11
subsequently the investors than disclosures made by accelerated filers. Doyle et al. (2007b) supports the studies of OU and Penman (1989) and Sloan (1996) and suggest that the market reactions may be greater for the non-accelerated filers because these companies experience greater uncertainty in their information environments. The reason for this is because these companies are smaller and it is presumed that smaller companies have less complicated information systems and less analyst following. Thus, when non-accelerated filers make the disclosures, these disclosures will convey new information to the investors which can impact the investors believes and in turn change the returns of a company significantly. This is also supported by Beneish et al (2006) who show that the disclosures made by non-accelerated filers, which do not include the auditors opinion on internal controls, are more informative as these filers experience higher pre-disclosure information uncertainty. Their results show that non-accelerated filers experience significantly negative abnormal returns when companies make the SOX 404 disclosures.
SOX 404 requires accelerated filers (larger public companies) to include the auditor attestation requirement. Recent studies argue that the market reactions to weakness disclosures under SOX 404 are different for accelerated filers compared to non-accelerated filers (smaller public companies). Doyle et al. (2007b) argues that the auditor attestation requirement results in a lower threshold for disclosure under Section 404 than Section 302. The reason behind this, is that the accelerated filers know that the internal controls and the disclosed weaknesses under Section 404 are audited. Furthermore, Beneish et al (2006) show that accelerated filers (the larger firms that are required to file under Section 404) operate in richer information environments, which makes it harder to detect a market response when disclosures are made. This means that when the accelerated filers make their disclosures, these disclosures will not be highly informative to the investors. Therefore, the following hypothesis is formulated:
H1: ICMW disclosures of accelerated filers that include an auditors opinion on internal controls have less impact on the market than ICMW disclosures of non accelerated filers that do not have an auditors opinion on internal controls.
The auditor attestation requirement under SOX 404(b), which applies only to accelerated filers can either be an adverse auditors opinion, which signifies that internal control problems
continue or an unqualified auditors opinion, which indicates no prior disclosure of internal control problems. This separate auditors opinion on internal controls over financial reporting (ICFR) includes evaluating and testing the controls. When the auditors identify even a single material weakness in the company’s ICFR, they are required to give an adverse opinion. Ashbaugh-Skaife et al (2009) assess how changes in internal control quality affects firm risk and cost of equity. To evaluate this effect they use unaudited pre-SOX 404 disclosures and SOX 404 audit opinions. In their test they find that firms that subsequently receive an adverse auditors opinion have a modest but insignificant increase in cost of equity when the SOX 404 opinion is released. This result can be explained by the fact that an adverse auditors opinion may indicate that a firm is unable or unwilling to remediate its internal control problems or it may appear that the auditor has found new internal control deficiencies.
Furthermore, the study of Hammersley et al (2008) posit that the disclosure of a material weakness may cause investors to raise concerns about the cost of remediating the weakness, the possibility that the weakness will remain unremediated and require auditors to give an adverse opinion on internal controls under SOX 404, and the chance for uncorrected errors to stay in the financial statements. They find that audit discovery of a material weakness in the internal controls has no effect on returns, however they do find some evidence that returns are associated with their proxy for audit quality. Particularly, they show that returns are less negative when a Big Four auditor is used instead of a smaller auditor. Overall, the results show that the material weakness disclosures are informative and cause investors to change their expectations about firm value. The study also shows that investors appear to be concerned about the presence of any weaknesses that could impact the auditors ability to conduct a successful audit.
Gupta and Nayar (2007) support the previous discussed studies (Ashbaugh-Skaife et al, 2009; Hammersley et al, 2008) and show that ICMW disclosures that include an adverse auditors opinion will generate negative market reactions and the cost of equity capital increases. This reaction can be mitigated to some extent, if firms use a Big Four auditor. Prior literature also shows that companies with an adverse SOX 404 opinion are associated with lower earnings quality (Ashbaugh-Skaife et al. 2008; Chan et al. 2008) and larger management forecast errors (Feng et al. 2009). Also, Shelton and Whittington (2008) find that adverse audit opinions on ICFR result in investment analysts making a higher assessment of company risk, a lower assessment of the strength of ICFR and a marginally significant
difference in the likelihood of recommending stock to their client. These findings provide evidence that the assessment of internal control made by auditors gives information to investment analysts. In addition, Lopez et al (2009) discuss the perceptions of investors towards adverse auditors opinions and show that the returns of companies with an adverse auditor opinion on ICFR are negative and when an unqualified opinion is given the returns are positive. Basically, their evidence suggest that investors value the auditor’s opinion on ICFR. Also, it indicates that an adverse auditor opinion on internal controls provides incremental value-relevant information to investors relative to an unqualified opinion. This result is also supported by the study of Schneider and Church (2008), which investigates the auditor’s internal control opinion over financial reporting on bank loan officers’ loan decisions. Based on these studies, the following hypothesis is formulated:
H2: Abnormal returns are positive when an auditor issues a unqualified opinion compared to when an auditor issues an adverse opinion which results in negative abnormal returns.
3. Methodology
This study will use an archival research to test the relation between Internal Control Material Weakness (ICMW) disclosures under SOX 404 of accelerated and non-accelerated filers and the abnormal returns. Accelerated filers are obligated to comply to the auditor attestation requirement under SOX 404(b) where auditors give their opinion about the internal controls, contrary to the non-accelerated filers that only have to comply with SOX 404(a). Therefore, this study distinguish itself from other studies by also looking at accelerated filers and focus on how the moderating variable, namely auditor opinion on a company’s Internal Control over Financial Reporting (ICFR), has an effect on the previous mentioned relation. Figure 1 provides a schematic overview of the relationships tested.
Figure 1: Associations tested
3.1 Accelerated filers
To test the hypotheses I have to define an ‘accelerated filer’ and I have to find out which forms are used for the ICMW disclosures under SOX 404. The implementation of SOX 404 is determined by the Securities Exchange Commission (SEC) and is based on a company’s classification. SEC has made up the so-called accelerated filer rules, and the SEC wanted the larger companies to comply first to SOX 404. The larger companies were classified as the accelerated filers. The definition of ‘accelerated filers’ (Protiviti, 2007) according to the Exchange Act Rule 12b-2 is that a company is an accelerated filer when (a) it has an equity market capitalization over $75 million, (b) it has been subject to the requirements of Section 13(a) or 15(d) of the Exchange Act for at least 12 months, (c) it has filed an annual and quarterly report with the Commission, and (d) it is not eligible to use forms KSB or 10-QSB for its annual and quarterly reports. These companies have to disclose the ICMW under SOX 404 in the forms 10-Ks and 10-Qs, and have to follow the commission’s accelerated filing requirements to do this. This makes the distinction of being an ‘accelerated filer’. SOX 404(a), which is the requirement for management to issue an internal control report, must be included in the form 10-K. Further, the form 10-K should include SOX 404(b), which is the auditor attestation requirement. When there is any change in a company’s ICFR it is disclosed in the form 10-Q, which is a quarterly report. The U.S domestic accelerated filers were required to comply with SOX 404 beginning in fiscal years ending on or after November 15,
2004. Further, Large foreign private issuers were to comply with SOX 404 for their fiscal year ending on or after July 15, 2006 (see table 1).
A small moderation has been made by the SEC for the accelerated filers group. A new classification was created, the so-called ‘large accelerated filer’, where a company has a market capitalization of $700 million or more (Protiviti, 2007). This change had no effect on the U.S companies, however the foreign private issuers that have a market capitalization between $75 or $700 were given an extra year (i.e., fiscal years ending on or after July 15, 2007) to comply with SOX 404(b), the auditor attestation requirement (see table 1). Thus, there are two different groups that fall under accelerated filer. When I will test the first hypothesis, all accelerated filers (market cap $75 - $700 million and market cap of $700 million or more) will be compared to the non-accelerated filers. For the second hypothesis, the same group of accelerated filers will be used. Besides, taking the accelerated filers into consideration, I will specifically look at the auditor’s opinion on management’s report on internal controls. The auditor can either give an unqualified or adverse opinion. I will investigate the 10-K filings, since this is where management includes the report on internal controls and it includes the auditor attestation requirement. Moreover, it is not required to include a management report on internal control in the 10-Qs (Protiviti, 2007), so therefore I will not take this form in my test.
3.2 Non-accelerated filers
A non-accelerated filer is a company that has a market capitalization lower than $75 million (Protiviti, 2007). These companies will be used for the first hypothesis and compared to the accelerated filers. All of these companies have to comply with SOX 404(a) and the compliance started for fiscal years ending on or after December 15, 2007 (see table 1). The compliance to SOX 404(b) of the smaller companies was first delayed multiple times, however in 2010 these companies were completely exempted from this requirement due to the fact that the costs outweighed the benefits. Calendar-year reporting companies had to file their first internal control report in the form 10-K for the calendar year 2007 (see table 1). Both forms, 10-Ks and 10-Qs are filed as these forms are subject to SOX 404. Again, I will only look at the 10-Ks when I test my hypotheses, as I explained with the accelerated filers.
TABLE 1
Compliance dates public reporting companies regarding ICFR
Accelerated filer status Management’s report Auditor’s attestation
U.S. Issuer
Large Accelerated Filer or Accelerated Filer
($75 million or more)
Annual reports for fiscal years ending on or after November 15, 2004
Annual reports for fiscal years ending on or after November 15, 2004
Non –Accelerated Filer Annual reports for Exempted since July (less than $75 million) fiscal years ending on
or after December 15, 2007
21, 2010
3.4 Cumulative Abnormal Returns
The dependent variable that will be tested in this study is the Cumulative Abnormal Returns (CAR). To measure the reactions of investors to SOX 404 disclosures, an event study will be conducted over the three-day window -1 to +1 surrounding the disclosure date of SOX 404. Where 0 is the day the SOX 404 disclosure is made. This three-day window is to isolate the reactions of investors to the disclosures of SOX 404 and from the reactions to other news. By doing this, the tested association is measured more reliably.
To calculate the Cumulative Abnormal Returns, the market-adjusted model is used based on a value weighted index. I use the market-adjusted model as it is typically used in prior research when computing the abnormal returns around SOX 404 disclosures (Palmrose et al., 2004; Beneish et al., 2006; Hammersley et al., 2008).
3.4 Control variables
When I test the aforementioned hypotheses, some control variables are included. The characteristics firm size and leverage have been frequently used as they are related to the market reactions to disclosure (Bhojraj et al., 2004; Collins et al., 1987; Palmrose et al., 2004). According to Palmrose et al. (2004) the market of larger firms already incorporate information through other sources and that is why the market reactions to disclosures will be less severe for larger firms. Furthermore, they explain that higher risk is associated with leverage and this tends to make the market reactions stronger. Thus, I add these two control variables, namely SIZE and DEBT into my model.
The ratio of book value to market value of equity is another characteristic known to affect returns. According to Fama and French (1995) book value to market value of equity is associated with persistent properties of earnings. They find, like Penman (1991) that firms who have low book-to-market equity have high average returns whereas firms with high book-to-market equity are quite distressed. Based on their findings, I include the control variable BOOK_TO_MARKET, as it could influence the abnormal returns.
Finally, auditor quality might change the responses of investors. Previous studies (Titman and Trueman, 1986; Nichols and Smith, 1983) show that the difference in auditor quality does affect firm value and market response. Furthermore, Ge and McVay (2005) argue that firms who are audited by large auditors are more likely to be large firms as well who have well-established internal controls and these controls are tested even before SOX. Besides, Becker et al. (1989) find that larger audit firms are associated with higher quality audits. Therefore, the difference in auditor used by a firm can affect the returns when this firm makes an ICMW disclosure. Thus, I include a dummy variable, namely BIG4, where BIG4 is equal to one if the firm is audited by one of the largest four audit firms (KPMG, PricewaterhouseCoopers, Ernst and Young, Deloitte) and zero when the firm is audited by other audit firms.
3.5 Research design
To test the previous shown associations I use the following regression model:
CAR = β0 + β1DEBT + β2BOOK_TO_MARKET + β3SIZE + β4BIG4 +
β5ACCELERATED + β6NON-ACCELERATED + β7OPINION + ε
(1)
This model will test the effect of SOX 404 disclosures of accelerated and non-accelerated filers on abnormal returns. Included are control variables to make sure the association tested is really a reaction on SOX 404 disclosures and not from other factors.
For the first hypothesis, I explore if the market reacts less on SOX 404 disclosures of accelerated filers compared to disclosures of non-accelerated filers, as accelerated filers include an auditor’s opinion on Internal Control over Financial Reporting (ICFR). Thus, I will investigate if there are indeed hardly any market reactions when accelerated filers disclose 18
under SOX 404 as existing literature state and whether or not there are strong negative market reactions to the disclosures of non-accelerated filers. A lower coefficient is expected for
ACCELERATED (β1 ) than for NON-ACCELERATED (β2 ).
As for the second hypothesis, an auditor’s opinion on ICFR is included and I will test the market reactions to the disclosures of the accelerated filers that have either an adverse or unqualified auditor opinion. I will explore if abnormal returns are positive when an auditor issues an unqualified opinion on ICFR in relation to an issued adverse opinion, for which I expect negative abnormal returns. The dummy variable OPINION(β7), where 1 is an
unqualified opinion and 0 is an adverse opinion, is expected to have a positive coefficient. This is consistent with the second hypothesis. It investigates whether or not investors value an independent auditors opinion on management’s assessment on internal controls.
The variables used in this model are as follows:
TABLE 2
Definition and measurement of variables
Type of variable Variables Definition
Dependent variable CAR Three day Cumulative Abnormal Return for firm j over event i
Independent variables
ACCELERATED
NON-ACCELERATED
OPINION
Public reporting companies that have a market capitalization of $75 million or more
Public reporting companies that have a market capitalization < $75 million
1 if the auditor opinion is adverse, 0 if auditor opinion is unqualified Control variables SIZE DEBT BOOK_TO_MARKET BIG4
The natural log of total assets Total debt divided by total assets
Book value of equity divided by market value of equity
1 if the firm is audited by a BIG4 audit firm, 0 otherwise
3.5 Sample selection
To construct my sample I retrieve data from AuditAnalytics. The data that is retrieved is based on the years 2011 till 2014. The motivation behind this choice is as follows. Although large en normal accelerated filers started filing a management report and auditors attestation report in 2004 regarding ICFR, non-accelerated filers started filing a management’s report in 2007. Due to the costs of an auditors attestation report, these non-accelerated filers which only have a market cap below $75 million, were not obliged to immediately file an auditors attestation report in 2007. The Securities Exchange Commission (SEC) kept delaying this obligation to file an auditors attestation report for the non-accelerated filers until they decided in 2010 that these filers should be exempted permanently from filing an auditors attestation report, since the costs were outweighing the benefits. Therefore I start retrieving data from 2011, as it is absolutely sure the non-accelerated filers only file a managements report and the accelerated filers file both reports, since I need to compare these two groups to test my first hypothesis.
AuditAnalytics mentions five specific variables that are required to constitute a ‘business key’ to uniquely identify an observation. These variables are COMPANY_FKEY,
AUDITOR FKEY, FISCAL_YEAR_IC_OPINION, IC_OPINION_TYPE_A_OR_M, and NTH_TIME_RESTATED. Besides these mandatory variables I have included the variables
needed for my research. For my first hypothesis, the accelerated filers and non-accelerated filers need to be compared with each other, therefore I selected the variables
IS_ACCEL_FILER and IS_SMALL_REPORT. The first two stands for accelerated filer and
non-accelerated filer. The variables indicate how the registrant identified its filer status. They are coded with a 0 and 1, where 0=No and 1=Yes. These variables are based on the research of Beneish et al (2006) which also include the variables that indicate a firms’ filing status. For my second hypothesis, the adverse and unqualified auditors opinion is compared with each other for the accelerated filers and therefore the variable OPINION is included. Where 0=adverse opinion and 1=unqualified opinion. This variable is based on the research of Lopez (2009) which investigates whether an adverse auditors opinion provides value-relevant information to investors. Furthermore, the variables ACCEL_FILER_KEY, FILE_DATE and
FISCAL_YEAR_ENDED_IC_OPINION are included. The variable ACCEL_FILER_KEY
indicates in which report the disclosure is made, the variable FILE_DATE is needed to see when the disclosure is made and the last variable is needed to generate the input since I selected the time period 2011-2014. It indicates the fiscal year ended date of internal control 20
assessment.
The sample consists of only SEC registered companies from the United States as these companies are required to disclose under SOX 404. SOX 404 disclosures are made in a 10-K and a 10-Q report. As I mentioned earlier, I only consider the 10-K reports, since this is where management includes the report on internal controls and it includes the auditor attestation requirement. Moreover, it is not required to include a management report on internal control in the 10-Qs (Protiviti, 2007). When I selected the necessary variables in AuditAnalytics I retrieved 37419 disclosures made between 2011-2014. Next, I excluded all other reports besides the 10-K report which brought the sample down to 7101 disclosures. Further, I excluded 833 observations that did not give information regarding the accelerated filer status, which gave a total sample of 6268 firm-year observations.
The control variales are obtained from several databases. The variable SIZE,
BOOK_TO_MARKET and DEBT are retrieved from Compustat, whereas the control variable BIG4 comes from AuditAnalytics. In addition, I extracted the dependent variable CAR from
Eventus. It appears that Eventus stores data until the end of 2013, so my observations from 2014 were dropped. After deleting the observations with incomplete data, my final sample consists of 741 firm-year observations which are observations from 396 unique firms.
TABLE 3
Sample selection firms disclosing under SOX 404
No of observations No of unique firms
Initial sample: SOX 404 internal controls disclosures during the period 2011 – 2014
Less: All other reports except 10-K filings
Less: Firms with incomplete data regarding
accelerated filing status
Less: Firms with incomplete data regarding
DEBT, SIZE, BOOK_TO_MARKET and CAR
Full sample selection
37419 30318 9729 7438 7101 833 2291 293 6268 5527 1998 1602 741 396
4. Descriptive statistics
The following section describes the descriptive statistics on my dependent and independent variables. The statistics are arranged in a comprehensible table. Next, the correlation is presented and explained.
4.1 Dependent and independent variables
In table 4 the descriptive statistics are described for the full sample, which is calculated in table 3. The total sample consists of 741 observations and from these observations 49,8% (N=369) are accelerated filers and 50,2% (N=372) are non-accelerated filers.
TABLE 4
Descriptive statistics on variables used for sample of 741 observations
No of observations in sample Auditor agrees with mgt. assessment on internal controls Auditor disagrees with mgt. assessment on internal controls Audited by Big Four
Audited by other
ACCELERATED NON-ACCELERATED Total sample
369 (49,8%) 364 (98,6%) 5 (1,4%) 326 (88,3%) 43 (11,7%) 372 (50,2%) 372 (100%) 0 (0%) 57 (15,3%) 315 (84,7%) 741 (100%) 736 (99,3%) 5 (0,7%) 383 (51,7%) 358 (48,3%) DEBT (mln) Mean Median Min. Max. Std. Dev. 0,5038 0,4680 0,0249 1,3513 0,2378 0,5507 0,5164 0,0010 4,1319 0,3746 0,5273 0,4919 0,0010 4,1319 0,3147 BOOK_TO_MARKET (mln) Mean Median Min. Max. Std. Dev. 0,5056 0,4323 -1,7575 4,6920 0,5335 1,7192 1,1424 -4,8653 16,8599 2,2064 1,1149 0,6227 -4,8653 16,8599 1,7178 SIZE (mln) Mean Median Min. Max. Std. Dev 7,1589 7,0470 2,0342 12,2215 1,5926 4,4127 4,4113 0,1424 8,6973 1,5609 5,7802 5,9288 0,1424 12,2215 2,0907 CAR (mln) Mean Median Min. Max. Std. Dev. -0,0017 -0,0031 -0,2192 0,3288 0,0464 0,0040 -0,0038 -0,4780 0,6839 0,1044 0,0012 -0,0032 -0,4780 0,6839 0,0809 23
As for the variable OPINION, which looks at whether an auditor agrees with management’s assessment on internal controls, I find only five disagreements in my total sample and 736 agreements. These five disagreements are for the accelerated filers, as these firms are required to include an auditors attestation report. Furthermore, from the total sample 383 firms are audited by a Big Four audit firm (Deloitte, PricewaterhouseCooper, Ernst & Young, KPMG), whereby 88,3% of these firms are in the accelerated filers group and 15,3% is in the non-accelerated filers group. Thus, it seems that larger firms are mostly audited by Big Four audit firms, which is consistent with the suggestions of Ge and McVay (2005). Additionally, 358 firms are audited by other audit firms, whereby 11,7% of these firms are in the accelerated filers group and 84,7% is in the non-accelerated filers group.
The variable DEBT has an average mean of 0,52 million for the total sample, whereby the accelerated filers group has a mean of 0,50 million and the non-accelerated filers group has a mean of 0,55 million. It seems that the non-accelerated filers group has a slightly higher total debt than the accelerated filers group. As for the BOOK_TO_MARKET variable, the total sample has an average mean of 1,11 million, the accelerated filers has a mean of 0,50 million and the non-accelerated filers have a mean of 1,72 million. These numbers show a large difference between the larger and smaller firms. Next, the variable SIZE gives an average mean of 5,78 million for the total sample, whereby the accelerated filers have a mean of 7,16 million and the non-accelerated filers have a mean of 4,41 million. The accelerated filers are larger in terms of total assets than the non-accelerated filers. This was to be expected as these firms are per definition already the larger firms. Finally, the independent variable CAR has an average mean of 0,0012 million for the total sample, whereby the accelerated filers have a mean of -0,0017 million and the non-accelerated filers have a mean of 0,0040 million.
4.2 Correlation matrix
The correlations between the variables are presented in table 5. This table shows that
ACCELERATED FILERS are negatively correlated with CAR (-0,035), and NON-ACCELERATED FILERS are positively correlated with CAR (0,036). However, the
correlation matrix shows that these correlations are not significant. For my first hypothesis, I investigate if the market reacts less on SOX 404 disclosures of accelerated filers compared to disclosures of non-accelerated filers, as accelerated filers include an auditor’s opinion on Internal Control over Financial Reporting (ICFR). These correlations indicate that there is no 24
relationship between these variables, consequently my first hypothesis is probably not supported.
In addition, table 5 shows that the dummy variable OPINION is positively correlated with CAR at a .01 significance level. As for the second hypothesis, I hypothesize that abnormal returns are lower when the auditor issues an adverse opinion on the internal controls over financial reporting than when he/she issues an unqualified opinion. For this reason, I expected the positive correlation between OPINION and CAR, indicated in the correlation matrix, which means that when an auditor issues an unqualified opinion there is an increase in the abnormal returns. Considering this significant association in the correlation matrix, my second hypothesis is probably supported.
TABLE 5 Correlation Matrix
CAR = β0 + β1DEBT + β2BOOK_TO_MARKET + β3SIZE + β4BIG4 + β5ACCELERATED + β6NON-ACCELERATED + β7OPINION + ε
CAR DEBT BOOK_ TO_ MARKET
SIZE. BIG4 ACCEL NON-ACCEL OPINION CAR(-1,+1) 1 DEBT 0,034 1 BOOK_TO_ MARKET 0,091* -0,013 1 SIZE -0,001 0,249** -0,077* 1 BIG4 -0,007 -0,015 -0,313** 0,552** 1 ACCEL -0,035 -0,074* -0,353** 0,657** 0,731** 1 NON-ACCEL 0,036 0,077* 0,352** -0,658** -0,733** -0,997** 1 OPINION 0,177** -0,004 -0,024 -0,032 -0,019 -0,016 0,016 1
* Correlation is significant at the 0,05 level (2-tailed). **
Correlation is significant at the 0,01 level (2-tailed).
Variable definition
CAR = Cumulative Abnormal Return DEBT = total debt divided by total assets
BOOK_TO_MARKET = Book value of equity divided by market value of equity SIZE = the natural log of total assets
BIG4 = 1 if the firm is audited by a Big4 audit firm, 0 otherwise
ACCELERATED = public reporting companies that have a market capitalization of $75 million or more NON-ACCELERATED = public reporting companies that have a market capitalization of <$75 million OPINION = 1 if the auditor opinion is adverse, 0 if auditor opinion is unqualified
5. Empirical results
This section presents the empirical results of this research. In order to test my hypotheses I used a hierarchical multiple regression that has four models. First, the multiple regression is explained. Then, the results are discussed to see whether my hypotheses are supported or can be rejected.
5.1 Hierarchical multiple regression
A hierarchical multiple regression that contains four models was conducted to test the hypotheses with CAR as the dependent variable. The control variables were entered into model one, and then the independent variables ACCELERATED, NON-ACCELERATED and
OPINION were entered into model two, three and four, respectively. The variables were
entered in this order because for my first hypothesis I look at the accelerated and non-accelerated filers, and it is not until my second hypothesis that I look at the auditors opinion on management’s assessment on internal controls. The regression statistics are shown in table 6 and 7.
5.2 Tests of hypotheses
For my first hypothesis I wanted to compare the market reactions on SOX 404 disclosures of accelerated filers to those of non-accelerated filers. My expectations for the accelerated filers is that the market will hardly react to the SOX 404 disclosures on ICFR and that in turn the disclosures of non-accelerated filers induces great market reactions, resulting in negative
abnormal returns. I expect a lower negative coefficient between CAR and ACCELERATED than between CAR and NON-ACCELERATED, since prior literature argues that it is hard to detect a market reaction with disclosures of accelerated filers. The multiple regression shows a negative coefficient (β = -0,026) for ACCELERATED, which is insignificant. Further it shows a positive coefficient (β = 0,188) for NON-ACCELERATED, which is not consistent with my hypothesis. However, there is no evidence that NON-ACCELERATED is actually positive, since this result is not significant as well. Thus, there are no abnormal returns incident to SOX 404 disclosures of the accelerated and non-accelerated filers. The inability to detect a market response to SOX 404 disclosures made by accelerated filers is consistent with prior literature and can be explained by two reasons. First of all, accelerated filers have lower thresholds for disclosure, because an auditor gives a separate opinion about their internal controls. This is supported by Doyle et al. (2007b). Second, when accelerated filers make their SOX 404 disclosures, the market’s reaction may already be satisfied by certain firm attributes as higher quality auditors and larger size. Further, the lack of market response to SOX 404 disclosures made by accelerated filers is consistent with the results of Beneish et al (2006) who say that accelerated filers operate in richer information environment and therefore the disclosures do not induce any market reactions. Next, the inability to detect a significant relationship between non-accelerated filers and cumulative abnormal returns means that the disclosures of non-accelerated filers do not have a significant impact on the cumulative abnormal returns. This result is not in line with Doyle et al (2007b) who supports the studies of OU and Penman (1989) and Sloan (1996), because they state that the market reactions are greater for non-accelerated filers, because these companies experience greater uncertainty in their information environment. A possible explanation for this insignificant relationship between SOX 404 disclosures and the abnormal returns is that investors may not fully understand the SOX 404 disclosures about the internal control weaknesses and its implications which is supported by Orenstein (2004). Furthermore, Nicolaisen (2004) states that regulators argue that disclosures will not necessarily induce reactions by investors. In summary, the results indicate that the disclosures of accelerated filers do not influence the abnormal returns as I predicted and they also point out that there are no market reactions with the disclosures of non-accelerated filers which I did not predicted. Therefore, my first hypothesis is not supported and can be rejected.
TABLE 6
Hierarchical multiple Regression of a Three-Day Cumulative Abnormal Returns during SOX 404 disclosures
CAR = β0 + β1DEBT + β2BOOK_TO_MARKET + β3SIZE + β4BIG4 + β5ACCELERATED + β6NON-ACCELERATED + β7OPINION + ε Model 1 Model 2 B Beta Sig. Std. Error B Beta Sig. Std. Error (CONSTANT) -0,008 0,429 0,010 -0,008 0,402 0,010 DEBT -0,011 0,042 0,273 0,010 0,010 0,038 0,353 0,010 BOOK_TO_MARKET 0,005 0,102** 0,009 0,002 0,005 0,097* 0,018 0,002 SIZE -0,001 -0,025 0,589 0,002 -0,001 -0,014 0,808 0,002 BIG4 0,006 0,039 0,410 0,008 0,008 0,050 0,364 0,009 ACCELERATED -0,004 -0,026 0,697 0,011 NON-ACCELERATED OPINION Adjusted R Square R Square F Change 0,005 0,010 1,951 0,004 0,011 0,152 a. Dependent Variable: CAR
b. Predictors: (Constant), BIG4, total debt, book_to_market, size c. Predictors: (Constant), BIG4, total debt, book_to_market, size, opinion
d. Predictors: (Constant), BIG4, total debt, book_to_market, size, opinion, accelerated
e. Predictors: (Constant), BIG4, total debt, book_to_market, size, opinion, accelerated, non-accelerated
*
Correlation is significant at the 0,05 level (2-tailed)
** Correlation is significant at the 0,01 level (2-tailed)
Variable definition
CAR = Cumulative Abnormal Return DEBT = total debt divided by total assets
BOOK_TO_MARKET = Book value of equity divided by market value of equity SIZE = the natural log of total assets
BIG4 = 1 if the firm is audited by a Big4 audit firm, 0 otherwise
ACCELERATED = public reporting companies that have a market capitalization of $75 million or more NON-ACCELERATED = public reporting companies that have a market capitalization of <$75 million OPINION = 1 if the auditor opinion is adverse, 0 if auditor opinion is unqualified
Further, I hypothesize that abnormal returns are positive when the auditor issues an unqualified opinion on internal controls over financial reporting compared to when the auditor issues an adverse opinion. For this reason, I expect a positive coefficient for the dummy variable OPINION, where 0 is unqualified opinion and 1 is an adverse opinion. The multiple regression results indicate a positive coefficient of 0,180 at a .001 significance level, which indicates that there is a positive relationship between OPINION and CAR. Thus, when the auditor gives an unqualified opinion on internal control over financial reporting it induces positive market reactions which as a result will give positive abnormal returns. This means that there is evidence to support my hypothesis as I predict positive abnormal returns for when an unqualified opinion is issued compared to when an adverse opinion is issues. This result is consistent with previous studies (Gupta and Nayar, 2007; Ashbaugh-Skaife et al, 2009) that show that ICMW disclosures that include an adverse auditors opinion will generate negative market reactions and the cost of equity capital increases, and when an unqualified opinion is given the market will react positively which increases abnormal returns. Lopez et al (2009) supports my results as well, as their research shows that the abnormal returns of companies with an unqualified opinion on ICFR are positive and with an adverse opinion the abnormal returns are negative. To conclude, my second hypothesis is supported by the regression results and is consistent with prior literature.
TABLE 7
Hierarchical multiple Regression of a Three-Day Cumulative Abnormal Returns during SOX 404 disclosures
CAR = β0 + β1DEBT + β2BOOK_TO_MARKET + β3SIZE + β4BIG4 + β5ACCELERATED + β6NON-ACCELERATED + β7OPINION + ε Model 3 Model 4 B Beta Sig. Std. Error B Beta Sig. Std. Error (CONSTANT) -0,039 0,639 0,082 -0,043 0,594 0,081 DEBT 0,009 0,037 0,365 0,010 0,009 0,036 0,365 0,010 BOOK_TO_MARKET 0,005 0,097* 0,018 0,002 0,005 0,102* 0,011 0,002 SIZE 0,000 -0,012 0,822 0,002 0,000 -0,007 0,897 0,002 BIG4 0,008 0,052 0,348 0,009 0,009 0,054 0,315 0,009 ACCELERATED 0,026 0,160 0,750 0,081 0,028 0,170 0,730 0,080 NON-ACCELERATED 0,030 0,188 0,710 0,081 0,032 0,199 0,688 0,080 OPINION 0,177 0,180** 0,000 0,036 Adjusted R Square R Square F Change 0,003 0,011 0,139 0,034 0,043 24,688 a. Dependent Variable: CAR
b. Predictors: (Constant), BIG4, total debt, book_to_market, size c. Predictors: (Constant), BIG4, total debt, book_to_market, size, opinion
d. Predictors: (Constant), BIG4, total debt, book_to_market, size, opinion, accelerated
e. Predictors: (Constant), BIG4, total debt, book_to_market, size, opinion, accelerated, non-accelerated
*
Correlation is significant at the 0,05 level (2-tailed).
**
Correlation is significant at the 0,01 level (2-tailed).
Variable definition
CAR = Cumulative Abnormal Return DEBT = total debt divided by total assets
BOOK_TO_MARKET = Book value of equity divided by market value of equity SIZE = the natural log of total assets
BIG4 = 1 if the firm is audited by a Big4 audit firm, 0 otherwise
ACCELERATED = public reporting companies that have a market capitalization of $75 million or more NON-ACCELERATED = public reporting companies that have a market capitalization of <$75 million OPINION = 1 if the auditor opinion is adverse, 0 if auditor opinion is unqualified
6. Conclusion
In 2002, the Sarbanes-Oxley Act (SOX) was enacted after numerous of scandals like WorldCom and Enron. One of the acts is §404. This act raised a lot of concerns among the companies as the costs for the second part of the act, that requires an auditors opinion on internal controls, were considered very high for the smaller public companies. Therefore, smaller public companies (non-accelerated) were exempted from this requirement in 2010. Since the compliance costs were very high, more questions were raised about whether the internal control disclosures were beneficial to investors. Prior research investigates disclosures under SOX 404 and suggest that disclosures are informative and used by investors to change their expectations about a company, however prior literature also suggests that investors will have a hard time to fully understand the internal control weakness disclosures. Most studies look at the disclosures under SOX 404 and what their impact on the market is, considering audited or unaudited internal control reports for accelerated and non-accelerated filers, respectively. However, little research takes into account that an audited internal control report can have an adverse or unqualified opinion. Therefore, I look at accelerated and non-accelerated filers and consider both type of opinions.
For my research, I examine 741 firm-year observation from SEC registered companies from the United States between 2011 and 2014. In my research I test two hypotheses. To test these hypotheses I perform a hierarchical multiple regression. For my first hypothesis, I predict that accelerated filers have less impact on the market than non-accelerated filers, because accelerated filers operate in richer information environments than the non-accelerated filers. I find no market response to the disclosures of accelerated filers made under SOX 404. In addition, I do not find evidence that there is a relation between non-accelerated filers and abnormal returns either. This is not consistent with my hypothesis. The lack of evidence of finding a significant relation between the non-accelerated filers and abnormal returns means that there is no market response when the non-accelerated filers make their disclosures. This is inconsistent with prior literature (Doyle et al., 2007b; Beneish et al., 2006) as they suggest that non-accelerated filers have greater uncertainty in their information environments and when they make internal control disclosures it is considered more useful to the investors. For the second hypothesis, I predict that abnormal returns are positive when an unqualified auditor’s opinion is given instead of an adverse opinion. I find evidence that the market reacts positive when accelerated filers make their internal control disclosures in which the auditor gives an unqualified opinion compared to the adverse opinion. This evidence is consistent 31
with my second hypothesis. Further, based on this evidence, it can be said that the inclusion of an auditor’s opinion on internal control over financial reporting influences investors’ valuation judgements which is supported by Lopez (2009).
One possible limitation of this study is the time frame over which I gathered my data and conducted my tests. It may be possible that the market reacts differently to the internal control material weakness disclosures under SOX 404 in the earlier years than in the future years as in the beginning years of the new regime the auditors and management were not so familiar with the process of implementing, evaluating and reporting on the internal controls compared to now. A second limitation is that the analyses performed in this research used a sample which was limited by data availability. The limited data availability may have caused my tests to lack sufficient statistical power to discover relations in the data. This research considers unqualified and adverse auditors opinion on management’s assessment on internal control. It does not consider whether a Big Four auditor gave that opinion or a different audit firm. Future research could investigate this situation to see whether there is a difference in abnormal returns if a high quality auditor (Big Four audit firm) gives the opinion about internal controls compared to a low quality auditor (Non-Big Four audit firm).
7. References
Ashbaugh-Skaife, Hollis, Daniel W. Collins, and William R. Kinney Jr. "The discovery and reporting of internal control deficiencies prior to SOX-mandated audits." Journal of
Accounting and Economics 44.1 (2007): 166-192.
Ashbaugh-Skaife, Hollis, et al. "The effect of SOX internal control deficiencies and their remediation on accrual quality." The Accounting Review 83.1 (2008): 217-250. Ashbaugh-Skaife, H., Daniel W. Collins, and Ryan Lafond. "The effect of SOX internal control deficiencies on firm risk and cost of equity." Journal of Accounting
Research 47.1 (2009): 1-43.
Barry, Christopher B., and Stephen J. Brown. "Limited information as a source of risk." The
Journal of Portfolio Management 12.2 (1986): 66-72.
Beneish, Messod Daniel, Mary Brooke Billings, and Leslie D. Hodder. "Internal control weaknesses and information uncertainty." The Accounting Review 83.3 (2008): 665-703.
Bhojraj, Sanjeev, Walter G. Blacconiere, and Julia D. D'Souza. "Voluntary disclosure in a multi-audience setting: An empirical investigation." The Accounting Review 79.4 (2004): 921-947.
Botosan, Christine A. "Disclosure level and the cost of equity capital." Accounting review (1997): 323-349.
Chan, Kam C., Picheng Lee, and Gim S. Seow. "Why did management and auditors fail to identify ineffective internal controls in their initial SOX 404 reviews?." Review of
Accounting and Finance 7.4 (2008): 338-354.
Collins, Daniel W., Stephen P. Kothari, and Judy Dawson Rayburn. "Firm size and the information content of prices with respect to earnings." Journal of Accounting and
Economics 9.2 (1987): 111-138.
De Franco, Gus, Yuyan Guan, and Hai Lu. "The wealth change and redistribution effects of Sarbanes-Oxley internal control disclosures." Available at SSRN 706701 (2005). Diamond, Douglas W., and Robert E. Verrecchia. "Disclosure, liquidity, and the cost of capital." The journal of Finance 46.4 (1991): 1325-1359.
Doyle, Jeffrey, Weili Ge, and Sarah McVay. "Determinants of weaknesses in internal control over financial reporting." Journal of Accounting and Economics 44.1 (2007): 193-223.
Francis, Jennifer, et al. "Costs of equity and earnings attributes." The Accounting Review 79.4 33
