Earnings management and corporate governance in UK:

The role of board of directors and audit committee

Thesis Proposal by

Kang Lei

Faculty of Business Administration National University of Singapore

CHAPTER 1 INTRODUCTION
Financial reported earnings have powerful influence on a full range of business activities of a firm and its management decisions. The earnings could either affect investors evaluation of the firm or impact contractual outcomes which are related to financial leverage or compensation of mangers. Therefore managers have strong intentions to adjust earnings numbers to the desirable level. The flexibility of current accounting principles also provides managers with considerable ability to adjust accounting earnings. The practice that the management uses judgment in financial reporting and in structuring transaction to alter financial earnings is called earnings management (Healy and Wahlen (1998)).

Although earnings management problem is not new, the recent popularity of earnings manipulation has drawn serious attention from regulators, financial press and academic research. For example, in a speech at NYU Center of law and business in 1998, Author Levitt, the Chairman of U.S Securities and Exchange Commissions (SEC), said earnings management impaired the quality of financial reporting. His comments showed serious concern of the adverse consequence of earnings management to the US capital market. Earnings management could also be undesirable to shareholders. When the interests of shareholders and mangers diverge, mangers can manipulate earnings for their own purposes at the cost of shareholders interest. Hence how to improve the reliability and integrity of financial reporting is an

interesting topic for research.

Corporate governance mechanisms, board of directors and audit committee in particular, are responsible for monitoring managers on behalf of shareholders and overseeing financial reporting process by company law. Therefore the board of directors and audit committee should play a role in retaining earnings management. However, the boards of public firms are generally considered as passive entities which are controlled by management. In an effort to improve the effectiveness of the board and audit committee, many corporate governance reports, including Cadbury report in UK of 1992, Toronto Stock Exchange Corporate Governance Guidelines of 1994, and Blue Ribbon Committee Report in US of 1999, propose best practice recommendations on various aspects of corporate governance. The objective of this thesis is to empirically examine the effects of some of the best governance practices by the board of directors and audit committees on the level of earnings management.

I study the relation between, on the one hand, the attributes of the board and audit committee and, on the other hand, earnings management measured as discretionary accruals. I am particularly interested in the role of non-executive directors. I examine the effects of their proportion on board/audit committee, their competence, their compensation scheme and the activities of board and audit committee. Discretionary accruals are estimated using modified Jones Model. I conduct the analysis in a sample of large publicly traded UK firms, because UK data have more variances than those of

U.S, and the corporate governance environment of UK is different from that of U.S in several ways, therefore this UK study may provide some interesting results for issues of board/audit committee monitoring.

This study could enrich the literature of relationship between board monitoring and financial reporting by selecting UK firms for analysis. By now, most previous studies in this field are U.S based, and a few have provided empirical evidence for UK cases. Peasnell et al (2000a) and Peasnell et al (2000b) are two exceptions. However they only focus on the effects of the proportion of outside directors and the existence of audit committee on the level of earnings management. This thesis does a more comprehensive study on various characteristics of board/audit committee, and thus may provide more valuable information to UK accounting regulators in making recommendations for corporate governance practice.

This study also extends the research of board effectiveness by including the compensation of the directors as a determinant. It is possible that the directors make different performance under different remuneration scheme, but few previous studies have taken this financial motivation of the directors into consideration. Therefore, the results of this study might be useful for companies to design more effective compensation package.

Another potential contribution of this thesis is that I make additional estimation of the

discretionary accruals after considering some differences in accounting standards between UK and US. Wilkins (2002) suggests that some particular country differences in accounting environment may results in significant difference in some of the variables in modified Jones model, however most previous UK studies ignore this potential problem. For this thesis, one important difference for this UK study is that revaluation of non-current assets has been banned in US since 1933, but is permitted in UK (Company Act (CA) 85 4.31, Statement of Standard Accounting Practice (SSAP) 19, and Financial Reporting Standard (FRS) 15). In additional test, I will adjust modified Jones model to examine whether this difference could alter the results.

The remainder of this paper is organized as follows: Chapter Two reviews literature of earnings management and corporate governance, and develops hypotheses for test. Chapter Three gives an overview of corporate governance in UK, and Chapter Four describes the data sources and research methodology.

CHAPTER 2 LITERATURE AND HYPOTHESIS DEVELOPMENT 2.1 Earnings management

This is generally believed by the regulators and the public that managers manipulate reported earnings (Levitt (1998); Loomis (1999)). A large body of academic research has examined the existence of earnings management, in particular, around specific corporate events in which agency problem is most likely to occur. Perry and Williams (1994) provides evidence of managers manipulation of earnings in the predicted direction in the year preceding the public announcement of management's buyout intention. Erickson and Wang (1999) find that acquiring firms manipulate accounting earnings upward prior to stock for stock corporate mergers. Teoh et al (1998a and 1998b) find that managers raise reported earnings before initial public offerings and seasoned equity offerings.

Managers have various incentives to manipulate earnings. Some incentives are provided by contractual arrangements based on accounting earnings such as bonus plan, debt and dividend covenants, etc. For example, DeFond and Jiambalvo (1994) find that sample firms accelerate earnings prior to lending covenants, and Holthausen, Larcker and Sloan (1995) find that managers manipulate earnings downwards when their bonus are at their maximum. In some cases, earnings management is motivated by regulatory reasons. Previous studies find that managers would manipulate earnings

to circumvent industry regulations and reduce the risk of investigation by anti-trust regulators (Collins et al (1995); Cahan (1992)). However, recent research has been focus more on incentives provided by the capital market. Dechow and Skinner (2000) suggest that accounting information such as earnings is important for capital market to value the firm, and the increased stock market valuations and stock-based compensation during 1990s make managers have more incentives to manage earnings. The results of empirical researches are generally supportive to this assertion. Some recent studies show that firms overstate earnings prior to seasoned equity offering (SEO), initial public offering (IPO) and stock for stock mergers (Teoh et al (1998a, b); Erickson and Wang (1999)) in order to get favorable valuation by capital markets. Moreover Perry and Williams (1994) find evidence of earnings understatement problem prior to a management buyout.

Earnings management is different from accounting frauds which violate Generally Accepted Accounting Principles, because the opportunities of earnings management are inherent in the current financial reporting system. Xie et al (2003) argue that the nature of accrual accounting gives managers considerable discretion in determining the earnings in any given period. According to Teoh et al (1998a), within the boundary of GAAP, managers have several sources to manipulate earnings. They can choose an accounting method to advance or delay the recognition of revenues and expenses, use discretionary aspects of the application of the chosen accounting method, or adjust the timing of asset acquisitions and dispositions to alter reported

earnings.

Although Schipper (1989) suggests that earnings management could possibly be beneficial by providing a means for management to convey their private information on firm performance, there is a potential danger of wealth loss for shareholders when the interests of mangers and shareholders conflict. Since the managers are compensated explicitly (salary, bonus, stock option, etc) and implicitly (job security, reputation, etc) on the firms earnings. Managers may conceal the true performance by earnings management to get a higher compensation or keep their jobs at the cost of shareholders interest. Moreover, the earnings management increases the information asymmetry between managers and shareholders, thus investors may make wrong decisions based on misleading earnings information. For example, Teoh et al (1998) find that IPO issuers who manage earnings aggressively perform relatively bad after IPO compared to those manage earnings conservatively and demonstrate the wealth of outside shareholders can be harmed by earnings management.

A number of prior studies examine the existence of earnings management by identifying a situation where earnings management is likely to occur and estimating discretionary accruals. However, some recent papers test earnings management by examining the distribution of reported earnings (Burgstahler and Dichev (1997); Degeorge et al (1999); Brown (2001)). These studies find that the frequency of firms with small positive earnings (positive earnings changes or earnings surprise) is higher

than expected, while the frequency of firms with small negative earnings (negative earnings changes or earnings surprises) is less than expected. These results are explained as the evidence of income-increasing earnings management and support the hypothesis that managers have incentives to avoid reporting loss, earnings declines, and earnings missing analysts forecasts. The reason why meeting such simple benchmarks is so important to managers is probably due to the capital market reaction. According to Barth et al (1999) and Skinner and Sloan (2000), failure to meet these earnings benchmarks will cause a dramatic drop of stock price. Since the personal wealth of top managers is tied more closely to their firms stock price in form of stock-based compensation plan in recent years, it is reasonable to argue that managers have strong incentives to manipulate earnings to avoid missing earnings benchmarks. Thus, in this thesis, I identify the firms which are in danger of missing some earnings benchmarks and test whether the board of directors and audit committee could constrain earnings management behavior when managers have strong incentives to do so. However, managers may also adjust earnings downward in some situations. Degeorge et al (1999) argue that earnings far beyond the thresholds will be reined in to make future earning thresholds more attainable. Therefore I will also examine the ability of board and audit committee in reducing income-decreasing earnings management.

2.2 The role of the board of directors

The separation of ownership and control which is inherent in the modern corporate form of organization causes the agency problem between shareholders (the principals)

and management (the agent). This is because the ownership structure of a company is highly dispersed, and shareholders generally hold more than one kind of security to diversify their risks, therefore, no individual shareholder has enough incentives and resources to ensure that management is acting for the shareholders interest. To control this agency problem, corporate governance, according to Denis (2001), encompasses the set of institutional and market mechanisms that induce selfinterested managers to maximize the value of the residual cash flows of the firm on behalf of its shareholders.

Among the set of corporate governance mechanisms, the board of directors is often considered the primary internal control mechanism to monitor top management, and protect the shareholder interest. For example, Fama (1980) argues that board of directors is a market-induced institution, the ultimate internal monitor of the set of contracts called a firm, whose most important role is to scrutinize the highest decision makers within the firm.

There is a large literature examining the relationship between board monitoring and firm performance on various aspects such as CEO turnover, stock return, operating performance and financial reporting quality (Weisbach (1988); Brickley (1994); Vefeas (1999); Dechow et al (1996); Beasley (1996)). These previous papers not only confirm that the board of directors does affect firm performance, but also find some characteristics of the board are related to the effectiveness of the board, especially in

10

monitoring top managers. These characteristics are the independence of the board, the competence of outside directors, outside directors ownership, and the activity of the board.

Based on previous literature, Earnings management can be seen as a potential agency cost since managers manipulate earnings to mislead shareholders and fulfill their own interests. Therefore the board of directors which is in charge of solving the agency conflicts between mangers and shareholders should play a role in constraining the level of earnings management. Further more, prior similar researches of financial reporting fraud suggest that effective board monitoring helps to maintain the credibility of financial reports. Thus it is reasonable to hypothesize that an effective board of directors will help to limit the earnings management. In the following sections, I will bring forward several testable hypotheses on the relationship of the board characteristics and earnings management.

2.2.1 Board independence from management Fama and Jensen (1983) recognize the control function of board as the most critical role of directors and argue that the board is not an effective device for decision control unless it limits the decision discretion of individual top managers. Moreover, Cadbury Report suggests that an important aspect of effective corporate governance is the recognition that the specific interests of the executive management and the wider interests of the company may at times diverge. Therefore board independence from

11

management is one of the important factors determining the board effectiveness and monitoring ability. Hence, we expect to see the board independence has a positive relation with the board effectiveness in limiting earnings management. However the independence is fundamentally unobservable, I have to measure it by some proxies developed from previous literature. One proxy is the board composition, and another is whether the CEO is also the Chairman of the board, and the last one is the financial dependence of outside directors.

Although the specific knowledge about the organization that the inside directors have can make valuable contribution to decision control function of the board, the domination by managers on the board can lead to the collusion and transfer of stockholder wealth (Fama (1980)). Outside directors are generally considered independent of management and more effective in protecting the interests of shareholders when there is an agency problem. Therefore it is necessary to adding outside directors to keep the independence of board. Moreover, Fama and Jensen (1983) observe that outside directors have incentives to develop reputations as experts in decision control and monitoring because the outside directors labor market will price their services according to their performance.

The percentage of outside directors in the boards is increasing in practice these years and many corporate governance reports recommend for adding outside directors ( for example, Blue Ribbon Committee Report), and the Previous empirical studies

12

generally demonstrate the association between the proportion of outside directors and the board effectiveness in monitoring management. Weisbach (1988) find stronger association between performance and CEO turnover in outside-dominate boards than inside-dominate boards. Beasley (1996) and Dechow et al (1996) document negative relationship between outside directors and incidence of financial fraud. More specifically, some studies find evidence that the proportion of outside directors is negatively related to the level of earnings management (Peasnell et al (2000), Klein (2002) and B.Xie et al (2003)). Based on theory of Fama and Jensen (1983) regarding board composition and the results of prior studies, I make the following hypothesis: H1: there is a negative relationship between the proportion of outside directors on the board and the level of earnings management.

Besides the composition of outside directors on the board, the separation of roles of the Chairman of the board and the CEO can also affect the independence of the board. The role of the Chairman is pivotal to securing good corporate governance. Jensen (1993) defines the function of the Chairman of the board as to run the board meetings, oversee the process of hiring, firing, evaluating and compensating the CEO. Therefore when the Chairman of the board and the CEO is the same person, there is a very real danger that firm is controlled by one man, and the board is not independent from the management. Therefore Cadbury report recommends the roles of the Chairman and the CEO should be separate. Some empirical researches have demonstrated the combination may affect the board effectiveness of monitoring management. For

13

example, Dechow et al (1996) find firms are more likely to be subject to accounting enforcement actions by SEC for alleged violations of GAAP if they have the CEO simultaneously serves as the Chair of the board. Thus I hypothesize that: H2: the existence of combination of the CEO and the Chairman of the board is positively related to the level of earnings management.

It is usual for outside directors to receive a fixed annual fee for their services. However, they may also receive other forms of remuneration or reward from the company. One remuneration form which might hurt the independence of nonexecutive directors is stock option. When the directors are rewarded by large blocks of stock option, the temptation for them is to focus on ensuring that company price is as high as possible when the time comes for exercising the options. If the earning figure does not come out right, and managers have to adjust it, the directors may not have incentives to prevent this practice. Therefore, Cadbury report recommends nonexecutive directors should not participate in share option schemes since the independence of non-executive directors might be undesirably. Hence, I expect that: H3: the compensation for outside directors as stock option is positively related to the level of earnings management.

2.2.2 Competence of outside directors 2.2.2.1 Tenure of outside directors Increasing the proportion of outside directors can not guarantee the effectiveness of

14

the board monitoring. Outside directors have to possess the necessary competences to carry out their control and overseeing duties, among which the knowledge of company specific affairs is particularly essential (Chtourou et al (2001)). The longer the experience of outside directors on the board, the better knowledge of company and their executives they will get. Therefore, outside directors may be more capable of monitoring managers and financial reporting process if they have served the board for long time. This assertion is supported by many previous studies. For example, Beasley (1996) finds the likelihood of financial reporting fraud is negatively related to the average tenure of non-executive directors. And Chtourou et al (2001) find that average tenure of outside directors is negatively associated with level of earnings management.

However, there is another possibility that the outside directors with longer tenure are more likely to be entrenched with managers and thus become less effective monitors. This speculation is consistent with National Association of Corporate Directors (NACD) Board Guidelines 1999 which states outside directors may lose some of their independent edge if they stay on the board too long. Xie et al (2003) also find a positive association between average tenure of outside directors and level of earnings management. Although the Combined Code 1998 says that a reasonably long period on the board can give directors a deeper understanding of the companys business, the revised Combined Code recommends outside directors who have served more than 9 years must be re-elected at next Annual General Meeting. I empirically test the

15

following hypothesis using UK samples. H4: The relationship between the average tenure of outside directors and the level of earnings management is a converted U shape.

2.2.2.2 Directorships of outside directors The mixed results of tenure of outside directors show that tenure may not be a perfect proxy for directors competence. Another possible measure is the outside directorship. As Fama and Jensen (1983) indicate that outside directors have incentive to monitor firms effectively to seek director position in labor market, we can consider the directorships hold by outside directors as a signal of their ability as monitors. Empirical researches have demonstrated the positive relation between outside directorships and quality of financial reporting. Chtourou et al (2001) find the number of outside directorship is negatively related to the level of earnings management. Xie et al (2003) get similar results. However, if the outside directors sit on too many boards, they may not have enough time to perform their duties effectively. In 1995, Authur Levitt, the Chair of SEC, said the commitment of adequate time is an essential requirement for directors. NACD 1999 also suggests retired executives or professional directors should serve on no more than six boards. However some previous surveys suggest that the average directorship hold by UK outside directors is relatively low (Peanell et al (1999); Cook and Leissle (2002)). Based on prior literature, I will test the following hypothesis: H5: The number of directorship of outside directors is negatively related to the level

16

of earnings management.

2.2.3 Ownership of outside directors It is generally believed that the directors who hold substantial stock ownership are more likely to fight against management to protect shareholders interest, because they have their own wealth involved. For non-executive directors who hold no position in the firm other than serving on board, Jensen (1993) asserts that holding sizable stock ownership will provide them with better incentives to monitor management closely. Many empirical studies lend support to this assertion. For example, Beasley (1996) finds that the likelihood of accounting fraud is negatively related to non-executive ownership. Consistent with these evidence, Combined Code 1998 recommends that payment of part of a non-executive directors remuneration in shares can be a useful and legitimate way of aligning the directors interest with those of the shareholders. Therefore I expect that: H6: The stock ownership hold by outside directors is negatively related to the level of earnings management.

2.2.4 Board Activity It is generally believed that a more active board is better for shareholders interest, because directors have to spend more time and energy on the company affairs in an active board. Recently, many financial and academic publications have criticized that directors have too little time to attend meeting regularly and this will limit their ability

17

to monitor management well. Conger et al (1998) also suggest that board meeting time is an important resource for improving the effectiveness of board. Vafeas (1999) empirically tests the relation between board activity, which is measured by board meeting frequency, and the firm performance. He finds that the increase in board meetings leads to improved firm performance, and this result suggests the frequent board meeting can help to make up the limited director interaction time.

According to above articles, it is reasonable to get the implication that the frequent board meetings can help to improve the board effectiveness in monitoring, and thus have some effects in constraining earnings management. Therefore, I will test the following hypothesis: H7: The frequency of board meetings is negatively related to the level of earnings management.

2.3 The role of audit committee

Since the board of directors bears diverse responsibilities to manage the business and affairs of the corporation, it usually delegates some authorities and specific functions to several committees which consist of subsets of board members. Therefore the effectiveness of board monitoring is also related to the structure of the board. For example, Klein (1998) finds that overall board composition is unrelated to firm performance, but the composition of accounting and finance committee does impact the performance.

18

As a part of the board, audit committee takes over the board function to oversee the firms financial reporting process. Klein (2002) comments that audit committee meet regularly with firms outside auditors and internal financial managers to review the corporations financial statements, audit process and internal accounting controls. Thus audit committee helps to alleviate the agency conflicts between the top management and the shareholders by improving the quality of financial reporting and reducing the information asymmetry between inside managers and outsider shareholders.

Previous literature of audit committee and the quality of financial reporting has tended to focus on the existence of audit committee.Wild (1996) shows that the informativeness of a firms earnings report increases after the formation of an audit committee. McMullen (1996) and Dechow et al (1996) both find that firms committing financial fraud are less likely to have audit committees. Although these studies suggest the existence of audit committee has a positive impact on financial reporting, they do not investigate if the characteristics of an audit committee would affect the quality of financial reporting.

Some recent papers explore whether the audit committee characteristics could affect the different aspects of financial reporting process. Abbott and Parker (2000) find that active and independent audit committees are more likely to hire an industry specialist

19

external auditor. Abbott et al (2002) show that financial misstatements are less likely to occur in firms whose audit committees are independent and have at least one financial expert. More closer to the issues of this thesis, some studies investigate the relationship between audit characteristics and earnings management. Klein (2002) finds that the independence of audit committee is negatively related to abnormal accruals. Xie et al (2003) find that audit committee which are more independent, meeting more frequently, and have members with corporate or financial backgrounds are less likely to engage in earnings management. As a summary of above evidence, the independence, member financial expertise and activity are three important characteristics which could affect the effectiveness of audit committee in monitoring financial reporting.

In the following sections, I will develop testable hypothesis to examine how the characteristics of audit committee affect the level of earnings management.

2.3.1 Independence of audit committee It is generally believed that members who are independent from management are better monitors. Previous studies have provide plenty of evidence that an independent audit committee is better at monitoring financial reporting and auditing process of the firm ( e.g., Abbott and Parker (2000); Beasley et al (2000); Carcello and Neal (2000); McMullen and Raghunandan (1996)). Moreover, some studies have examined the negative link between the independence of audit committee and earnings management

20

using sample of United States (e.g., Chtourou et al (2001); Klein (2002); Xie et al (2003)). However, the independence research of audit committee has gone beyond the simple classification of outside and inside directors.

Researchers generally classify outside directors into one of the two categories: independent directors and grey directors. Grey directors include former officers or employees of the company or a related entity, relatives of management and professional advisors to the company (Beasley (1996); Carcello and Neal (2000)), while independent directors have no affiliation with the firm other than being on the board. Previous studies have shown that the personal or economic affiliation that grey directors have with the corporate management may impair their independence. For example, Carcello and Neal (2000) find that the percentage of inside and grey directors has a negative relationship with the probability that the auditor will issue a going-concern report when the firm is experiencing financial distress. Therefore the presence of those grey members on audit committee may impair the monitoring effectiveness of audit committee. Aware of this problem, BRC (1999) recommends that audit committee should be comprised only of independent directors.

The Cadbury report does not require a totally independent audit committee, but just suggests that the membership of audit committee should be confined to non-executive directors and majority of outside directors serving on the committee should be independent. However the prior empirical studies have suggest possible different

21

performance of independent directors and grey directors, so I expect to find the audit committees which have no insider or grey directors would be more able to restrain earnings management than those have. Hence, I propose the following hypothesis: H8: The presence of audit committee comprised solely of independent directors is negatively related to the level of earnings management.

2.3.2 Member financial expertise Another important variable which could affect the effectiveness of audit committee is the member competence. Independent directors may have intentions to curb earnings management for shareholders, but they may not be able to do so without certain level of financial knowledge. BRC 1999 recommends that each member of audit committee should become financially literate and at least one should have accounting or related financial management expertise. The positive effect of audit committee members who have financial expertise is supported by a number of empirical studies. Agrawal and Chadha (2003) which find the incidence of independent directors with accounting or finance background on audit committee is negatively related to the probability of earnings restatement. McMullen and Raghunandan (1996) also find that companies with financial reporting problems are less likely to have CPAs on the audit committee.

Cadbury report 1992 does not mention the importance of the members financial expertise, but the Audit Committees Combined Code Guidance 2003 recommends that at least one member of the audit committee should have significant, recent and

22

relevant financial experience, for example as an auditor or a finance director of a listed company. It is highly desirable for this member to have a professional qualification from one of the professional accountancy bodies. This change reveals that the qualification of audit committee members has been realized as an important issue for the corporate governance of UK firms.

According to US experience, member expertise can increase financial reporting and auditing quality, decrease the probability of accounting fraud, so it is probable that the financial expertise of audit committee members can increase their ability to detect and constrain earnings management. Therefore I will test the following hypothesis to see if this relationship still holds for UK cases. H9: The financial expertise of audit committee members is negatively related to the level of earnings management.

2.3.3 Meeting of audit committee A common proxy for committee diligence, meeting frequency has been generally considered as an essential component of audit committee effectiveness. Menon and Williams (1994) note that audit committees that do not meet, or meet only a small number of times, are unlikely to be effective monitors. In line with this argument, NACD 2000 also recommends that The audit committee should meet as frequently as necessary to perform its role. Some studies find the negative relationship between meeting frequency and the occurrence fraudulent financial reporting (e.g., Abbott et al

23

(2000), Beasley et al (2000)). Some other studies, such as Abbott and Parker (2000), link the meeting numbers with higher audit quality. In conclusion, a plenty of empirical results support the assertion that the meeting frequency of audit committee is positively associated with financial reporting quality. Since earnings management practice impairs the integrity and transparency of financial reports by adjusting real earnings, I can similarly propose this hypothesis: H10: The frequency of audit committee meeting is negatively related to the level of earnings management.

24

CHAPTER 3 CORPORATE GOVERNANCE IN UK

Similar to what has happened in US, the corporate governance issue has drawn increasing attention from the public and regulators in UK. Although company law holds that the board of directors is responsible for financial reporting process, the boards of UK firms were generally considered passive entities which were controlled by management several decades ago. A series of unexpected business failures and high profile accounting scandals which occurred in late 1980s and early 1990s reduced the reliability of financial reporting results and thus exposed the corporate governance weakness of UK firms.

As a response to the weak governance of UK firms, a series of corporate governance codes were developed through the 1990s. The Cadbury report was issued by the committee on the Financial Aspects of Corporate Governance in 1992. The Cadbury report contained the code of Best Practice as the criteria of good governance and focused on the board monitoring responsibilities and highlighted the role of Nonexecutive directors. Following Cadbury report, Greenbury Code was issued by the Committee of Executive pay in 1995, and the Combined Code was released in 1998. On 23 July 2003, the Financial Reporting Council published the final text of the revised Combined Code which will apply to reporting years commencing on or after 1 November 2003. While companies are not under obligation to comply with the

25

recommendation of codes, London Stock Exchange requires all the UK-incorporated listed firms to include a statement of compliance with the code in their annual report and clearly explain identify and explain areas of non-compliance, thereby making non-compliance a potentially costly action.

As an essential part of board, audit committee provides critical oversight on financial reporting process. The history of audit committee is quite long in US. Since 1978, NYSE has required all listed companies to have audit committees comprised solely of independent directors. While in UK, the development of audit committee is relatively slow. A report of the Accountants International Study Group in 1977 indicates that the practice of Audit committee in UK was unusual, and the concept of audit committee had not been generally accepted. Only 38 percent of the companies had audit committee in 1988 according a survey by the bank of England. However the Cadbury report (1992) highlights the importance of audit committee and recommends this practice to all the companies as one way to improve the quality of financial reporting. Collier (1996) shows that the audit committees had generally become more widespread among large firms after the issue of Cadbury report 1992. By 1995, almost 92% companies have established audit committees (Cadbury compliance report 1995).

Although the regulators and companies of UK have made obvious effort to improve the level of corporate governance, very limited empirical researches have been done

26

to examine the association between corporate governance and firm performance in UK market. Peasnell et al (2000a and 2000b) find that board monitoring helps to reduce the level of income-increasing earnings management. However they only examine two aspects of board monitoring: the proportion of Non-executive directors on the board and the existence of an audit committee, and thus leave the effects of many other aspects of corporate governance unclear. Therefore the main motivation of this paper is to fill up the lack of comprehensive examination of relation between corporate governance and earnings management of UK companies.

Moreover, the board characteristics of UK firms have more variances compared to US companies, and this provides a unique opportunity for research. This is probably because that the compliance with the UK corporate governance code is voluntary, so the companies are free to choose their own governance policy. Another possible reason is the current Combined Code is more flexible than the requirements in US. For example, although the Combined Code has acknowledged the importance of Nonexecutive directors, it only suggests that they should constitute at least one third of the board members, instead of majority of the board. In conclusion, the variety of UK sample has the potential to generate some interesting results which different from those US studies.

27

CHAPTER4 METHODOLOGY AND DATA SOURCE 4.1 Measure of earnings management

Although there is no perfect proxy for earnings management, most current studies focus on managers use of discretionary accruals. Consistent with previous literature, I estimated discretionary accruals using cross-sectional version of modified Jones Model (e.g., Dechow et al (1995); Teoh et al (1998a and 1998b)). More specifically, I focus on the discretionary current accruals (DCA).

There are several reasons for detecting earnings management by DCA instead of total discretionary accruals. First, current accruals are more easily for mangers to manipulate (Teoh et al (1998b)). Second, this measure is widely used by previous similar researches, so it is easier for me to compare my results with theirs (e.g., Peasnell et al (2000b); Xie et al (2003); Park and Shin (2003)). Third, the accounting principles for treatment of tangible fixed assets of UK are different from those of U.S, so if I use the original modified Jones Model, the relationship between gross property plant and equipment (PPE) and depreciation expenses may not be hold. I will address this issue in detail in the sensitivity tests.

The model for estimating DCA is as following. First, I use the following regression

28

model to get the estimates of 1 , 2 .

1 REV i ,t CA i ,t + 2 = 1 TA TA TAi ,t 1 i ,t 1 i ,t 1
C i ,t A

+ i ,t

(1)

is the current accruals of firm i defined as the change of non-cash current assets

RV less the change of current liabilities. E

i ,t

is the change of revenue between year

A t and t-1, and T i ,t is the book value of total asset of year t-1. The regression is 1

carried out for each industry-year combination.

Second, non-discretionary current accruals are estimated as:

1 REV i ,t REC i ,t +2 NDCA i ,t = 1 TA TAi ,t 1 i ,t 1

(2)

1 Where and 2 are OLS estimates for the coefficients in equation (1) and

E RC

i ,t

is the change of net receivables.

Finally, we obtain the discretionary accruals as the remaining portion of current accruals:

29

DCA i ,t =

CA i ,t TA i ,t 1

NDCA

i ,t

(3)

4.2 Earnings benchmarks

Several papers document managers have incentives to meet simple earnings benchmarks, which include avoiding losses and earnings declines, and beating analysts forecasts. This argument can be supported by Burgstahler and Dichev (1997) and Degeorge et al (1999) which both find that small reported losses (small profit declines) are rare, while small reported profits (small profit increases) are common, and imply that earnings management is more pronounced when earnings are below certain benchmarks. Thus, I examine the role of the board and audit committee in constraining earnings management around the earnings benchmarks.

Following Park and Shin (2003) and Peasnell et al (2000b), I use two earnings benchmarks: zero earnings and last years earnings. I will split the sample based on if the pre-managed earnings (actual earnings minus the discretionary accruals) meet or miss the target, and expect to find earnings management around targets. I expect to find income-increasing accruals when the pre-managed earnings are below the targets, and find income-decreasing accruals when pre-managed earnings are above the targets. Next, I will examine the relationship between the board of directors, audit committee and earnings management to see if they could help to decreasing the income-increasing (income-decreasing) accruals.

30

4.3 Regression Analysis

To analysis the effect of the board characteristics, audit committee on earnings management, I first construct following regression models: Model 1: for board of directors
DCA = 0 + 1OUT ( IND ) + 2 DUAL + 3OPTION + 4TENURE + 5 (TENURE ) 2 + 6 DIRSHIP

+ 7 STKOWN + 8 MEET + i controls +

Model 2: for audit committee

DCA = 0 + 1INDAUD + 2 FIN + 3 MEETAUD + i controls +

According to the pre-managed earnings, I split the sample into two sub-samples, and then run regression Model 1 and 2 in each sub-sample. Therefore, for every model, there are four regressions. Regression a (b) is for firm-year observations that premanaged earnings below (above) zero, and Regression c (d) is for firm-year observations that pre-managed earnings less than (more than) last year earnings.

4.3.1 Variables for board characteristics and audit committee For the independent variables in Model 1, OUT is the percentage of the number of outside directors or so called non-executive directors on the board, and IND is the proportion of independent outside directors on the board. DUAL is a dummy variable with value 1 if the CEO and chairman of the board is the same person and zero otherwise. OPTION is a dummy variable with value 1 if non-executive directors receive stock option compensation and zero otherwise. TENURE is the average years of board services of non-executive directors. DIRSHIP is the average number of

31

directorships held by non-executive directors in unaffiliated firms. STKOWN is the cumulative percentage of shares held by non-executive directors. MEET is the number of board meetings per year. I expect OUT, IND, TENURE, DIRSHIP, MEET and STKOWN to have negative (positive) coefficients in the regression of sub-sample that pre-managed earnings below (above) earnings targets, while DUAL and OPTION have positive (negative) coefficients.

For MODEL 2, INDAUD is a dummy variable with value 1 if the audit committee is composed solely of independent directors. FIN is the proxy for members financial expertise and also an indicator variable with value 1 if there is at least one member has past employment experience in finance or accounting. This definition is more restrictive than that of BRC 1999, and consists with revised combined code which recommends one member must possess recent and relevant experience in finance. MEETAUD is the number of meetings that the audit committee have in a year. I expect that the INDAUD, FIN and MEETAUD will have negative (positive) coefficients when the pre-managed earnings are below (above) earnings targets.

4.3.2 Control Variables The above regression models control for some other dimensions of the corporate governance and incentives for earnings management which could affect the level of discretionary accruals. BRDSIZE is the number of directors on the board. I control this factor because Jensen (1993) finds that a larger board is less effective and more

32

easily to be control by CEO. Klein (2000) also finds that board size is a determinant of the independence of audit committee. NOM is a dummy variable with value 1 if the nominating committee is composed of majority of independent directors. Previous studies demonstrate this variable reflect the management power on the board, and thus is related to the level of discretionary accruals (e.g., Chtourou et al (2001) and Peasnell et al (2000)).

BLOCK is the cumulative percentage of outstanding shares hold by outside block holders who own at least 5% of firm shares. This variable controls for the potential monitoring from block holders on earnings management. INSTOWN is the total percentage of outstanding shares held by institutional investors. Many papers find that institutional investors monitor and constrain the self-serving behavior of managers. For example, Chung et al (2003) find that large institutional shareholdings inhibit managers from using discretionary accruals to manipulate earnings. MANOWN is the faction of outstanding shares holding by managers. This measure is related to discretionary accruals because it reflects the extent to which the interests of managers are aligned with those of shareholders. As Warfield et al (1995) say, managerial ownership is inversely related to magnitude of accounting accrual adjustments. BIG4 is indicator variable with value 1 if the external auditor of the firm is BIG 4 to control of the audit quality effects. Previous studies suggest that Big 5 auditors are generally more effective in deterring earnings management than other auditors (e.g., Becker et al 1998, Kim et al (2003)).

33

BONUS is the average weight of bonus to total remuneration for all executives. Maximization of bonus is a potential incentive of earnings management, and many studies, such as Healey (1985) and Holthausen et al (1995), can provide consistent empirical evidence. LEV is the financial leverage measured as the ratio of total debt to total assets. On one hand, firms with high leverage ratio may have incentives to adjust earnings upward to avoid debt-covenant violation. On the other hand, such firms may under close scrutiny of lenders and are less able to do so. Therefore, the relationship between LEV and discretionary accruals is undetermined. MBRATIO is the market to book ratio of asset which proxy for growth opportunity of the firm. Matsumoto (2002) argues that firms with high growth prospects have greater incentives to manipulate earnings to avoid unfavorable market reaction to negative earnings news. SIZE is the natural log of sales. Following previous studies, this factor is controlled as a possible determinant of choice of discretionary accruals (e.g., Beck et al (1998), Park et al (2003), and Peasnell et al (2000)). LOSS is a dummy variable with value 1 if the firm has two or more previous consecutive years of losses. Matsumoto (2002) finds that the earnings of loss firms are less value relevant and thus managers are less likely to adjust earnings to meet targets. CFO is the cash flow from operation and this will control for its relationship with discretionary accruals documented by Dechow et al (1995). Finally, Year is an indicator variable control for time effects.

34

4.4 Sensitivity Tests

4.4.1 Alternative measure of earnings management Some studies measure the level of earnings management by total discretionary accruals which include the long-term accruals. As a supplementary test, I will use Modified Jones Model to estimate total discretionary accruals. The procedures are as follows;
TAC i ,t TAi ,t 1 1 REV i ,t + 2 = 1 TA TA i ,t 1 i ,t 1 PPE i ,t +3 TA i ,t 1 + i ,t

1.

TC A

i ,t

is the total accrual of firm i calculated as net income minus the cash flow

P from operations. P E

i ,t

is gross property plant and equipment. The regression is

carried out for each industry-year combination to estimate 1 , 2 and 3 . 2. Non-discretionary accruals are estimated as:
1 REV i ,t REC i ,t +2 NDA i ,t = 1 TA TAi ,t 1 i ,t 1 PPE i ,t +3 TA i ,t 1

3. The total discretionary accruals are DA i ,t =

TAC i ,t TA i ,t 1

NDA i ,t

The above method is commonly used by previous studies. However few of them consider the potential problem that the Jones model is developed based on U.S accounting system, so it may not work well in UK research as it does in U.S studies without taking UK financial reporting environment into consideration. In the

35

discussion of Yoon and Miller (2002), Wilkins (2002) suggests that their ignorance of some particular differences in accounting standards between Korean and U.S that would likely result in significant differences in some of the model variables may be the reason why their results of earnings management of SOE firms are different from prior U.S studies.

Since the Modified Jones Model considers the effects of fixed asset on discretionary accruals, the difference between the treatment of tangible fixed asset in UK and U.S may affect the power of this model in UK studies. In U.S, the revaluation of property, plant and equipment is not allowed and they should be reported as historical cost (opinion of the Accounting Principles Board (APB) 6.17), while the revaluation of tangible fixed assets is permitted in UK (CA 85, SSAP19, FRS 15). Whittred and Chan (1992) find that the majority of asset revaluations were taken at the last month of the financial year. Since the gross value of fixed assets is recorded as the revalued amount at the end of the year, while the depreciation expense for that year is unchanged. Thus asset revaluation could possibly disrupt the relation between gross property plant and equipment (PPE) and depreciation expenses of the period in Modified Jones Model.

Concerning the potential problem caused by revaluation, I will do supplementary tests to re-estimate the discretionary accruals with an alternative measure of PPE. For those firms have revaluations of depreciable assets at the fiscal year end, I adjust the value

36

of PPE to the historical cost equivalents to match with the depreciation expenses. The values of historical cost equivalents are available because CA 85 and FRS 15 require that the historical cost equivalents (cost, cumulative provision for depreciation and diminution in value and carrying amount) of each class of revalued assets should be disclosed. After getting new discretionary accruals, all the multivariate regressions examining the board and audit committee characteristics will be re-run to see if the results are consistent.

4.4. 2 Lack of Independence Since some firms may have two years observations in the sample, it is possible that pooled OLS is biased because of unable to control the firm-fixed effect. I will control this problem by re-run the regressions in sub-samples which constitute only one observation per firm. However this method will suffer the reduction of sample size.

4.5 Data Sources

The empirical tests are conducted using data of U.K listed companies in fiscal year 1999 and 2002. My primary sample is the constituents of FTSE 350 index at the end of every year. I then exclude all the financial firms (SIC codes 60-69), since it is difficult to define accruals and discretionary accruals for financial firms. I also exclude all regulated utilities firms (SIC codes 40-44, 46.48-49), because they have different earnings management incentives. Accounting data will be collected from

37

Datastream, supplemented where necessary by annual reports. Other data including board and audit committee characteristics, managerial ownership, block holder ownership, and outside auditors will be hand-collected from annual reports.

REFERENCE
Arthur Levitt, The Numbers Game, remarks at New York University Center for Law and Business, September 28, 1998 Barth,M.E., J.A.Elliott, and M.W.Finn. 1999. Market awards associated with patterns of increasing earnings. Journal of Accounting Research 37 (Autumn), 387-413. Beasley, M.S. 1996, An empirical analysis of the relation between the board of director composition and financial statement fraud. The Accounting Review 71 (October): 433-465. Beasley, M.S., J.V. Carcello, D.R. Hermanson, and P.D. Lapides,2000, Fraudulent financial reporting: Consideration of industry traits and corporate governance mechanisms. Accounting Horizons 14, 441454

38

Blue Ribbon Committee (BRC). 1999. Report and recommendations of the Blue Ribbon Committee on improving the effectiveness of corporate audit committees. New York Stock Exchange and National Association of Securities Dealers (www.nyse.com). Burgstahler, D., Dichev, I., 1997. Earnings management to avoid earnings decreases and losses. Journal of Accounting and Economics 24, 99-126. Cadbury Committee, 1992. Report of the committee on the financial aspects of corporate governance. Gee,London. Carcello, J.V. and T.L. Neal, 2000, Audit committee composition and auditor reporting, The Accounting Review 75, 453-467 Chtourou, S., Bedard, J., Courteau, L., 2001, Corporate governance and earnings management, Working paper, Universite Laval, Canada. Collier,P.A., 1996. The rise of audit committee in UK quoted companies: a curious phenomenon. Accounting, Business and Financial History 6 (2), 121-140. Conger, Lawler and Finegold, 2001, Corporate Boards: strategies for adding value at the to Dalton, D.R., Daily, C.M., Johnson, J.L., Ellstrand, A.E., 1999. Number of directors and financial performance: a meta-analysis. Academy of Management Journal 42, 674686. DeAngelo, L. 1986. Accounting Numbers as Market Valuation Substitutes: A Study of Management Buyouts of Public Shareholders. The Accounting Review 61, 400-420. DeAngelo L. 1988, Managerial Competition, information cost and Corporate Governance: the use of accounting performance measures in proxy contests, Journal of Accounting and Economics 10, 3-36 Dechow, P., Sloan, R., Sweeney, A., 1995. Detecting earnings management. The Accounting Review 70, 193-225. Dechow, Patricia M, Sloan, Richard G, Sweeney, Amy P, Causes and consequences of earnings manipulations: An analysis of firms subject to enforcement actions by the SEC, Contemporary Accounting Research. Toronto: Spring 1996. Vol. 13, Iss. 1; p. 1 (36 pages) DeZoort .F. T and Steven E. Salterio 2001, The Effects of Corporate Governance Experience and Financial-Reporting and Audit Knowledge on Audit Committee Members Judgments, AUDITING:A Journal of Practice& Theory Vol. 20, No. 2 Degeorge, F., Patel, J., Zeckhauser, R., 1999. Earnings management to exceed thresholds. Journal of Business 27, 1-33.

39

Fama, E.F., and M.C. Jensen. 1983. The separation of ownership and control. The Journal of Law and Economics 26 (June): 301-325. Healy, P. M. 1985. The Effect of Bonus Schemes on Accounting Decisions. Journal of Accounting and Economics 7, 85-107. Healy, P.M., Wahlen, J.M., 1998. A review of the earnings management literature and its implication for standard setting. Harvard Working Paper. Holthausen,R,, David F. Larcker and Richard G. Sloan,1995, Annual bonus schemes and the manipulation of earnings, Journal of Accounting and Economics 19, 29-74 Jensen. M. C 1993, The modern industry revolution, exit, and the failure of internal control system, Journal of Finance 48,831-880 Jensen, Michael C., and Murphy, Kevin J, 1990, Performance Pay and Top-Management Incentives, Journal of Political Economy 98,22564. Jeong-Bon Kim, Richard Chung, Michael Firth, Auditor Conservatism, Asymmetric Monitoring and Earnings Management, Contemporary Accounting Research, Vol 20, No 2. Jones, J.J., 1991, Earnings management during impact relief investigations. Journal of Accounting Research 29,193228 Klein A ,2002, Audit committee, board of director characteristics, and earnings management, Journal of Accounting and Economics 33, 375400. Matsumoto D.A. 2002, Managements Incentives to avoid negative earnings surprises, The Accounting Review 77, No 3, 483-514. McMullen, Dorothy A and Raghunandan, K,1996, Enhancing audit committee effectiveness, Journal of Accountancy (August), 79-81 McDaniel, L., R.D. Martin and L.A. Maines, 2002, Evaluating Financial Reporting Quality: The Effects of Financial Expertise vs. Financial Literacy, The Accounting Review vol 77, 139-167 Merle Erickson and Shiing-wu Wang, 1999, Earnings management by acquiring firms in stock for stock mergers, Journal of Accounting and Economics 27, 149-176 National Association of Corporate Directors 1999. The NACD Board Guidelines NYSE Listed Company Manual. Updated. New York, NY: New York Stock Exchange.

40

Park Y.W and Hyun-Han Shin, 2003, Board composition and earnings management in Canada, Journal of Corporate Finance, article in press Peasnell, K.V., Pope, P.F., Young, S., 2000a, Accrual Management to meet earnings targets: UK evidence pre- and post-Cadbury, British Accounting Review (2000) 32,415-445 Peasnell, K.V., Pope, P.F., Young, S., 2000b, Board monitoring and earnings management: Do outside directors influence abnormal accruals? Working paper Perry, S and T. Williams, 1994, Earnings Management preceding management buyout offers, Journal of Accounting and Economics 18,157-179 Steven R. Matsunaga and Chul W. Park, 2001, The effect of missing a quarterly earnings benchmark on the CEOs Annual Bonus, The Accounting Review Vol 76, No 3, 313-332 Teoh, S.H., Welch, I., Wong, T.J., 1998a. Earnings management and the underperformance of seasoned equity offerings. Journal of Financial Economics 50, 6399. Teoh, S.H., Welch, I., Wong, T.J., 1998b. Earnings management and the underperformance of initial public offerings. Journal of Finance 53, 19351974. Vafeas, N., 1999, Board meeting frequency and firm performance, Journal of Financial Economics 53, 113 142 Vafeas, N., 2000, Board structure and the informativeness of Earnings, Journal of Accounting and Public Policy 19, 139-160 Xie, Wallace N. Davidson III and Peter J. DaDalt, 2003, Earnings management and corporate governance: the role of the board and the audit committee, Journal of Corporate Finance 9 (2003) 295 316

41