The relationship between board Independence and accounting conservatism and the effect of the financial crisis.

Faculty of Economics & Business MSc Accountancy & Control 2013-2014

Master Thesis

The relationship between Board Independence and

Accounting Conservatism and the effect of the

Financial Crisis

Abstract

This thesis investigates the relationship between the independence of the board of directors and accounting conservatism and the moderating effect of the financial crisis. The effect of the financial crisis on the relationship was measured in three steps: (1) a change in board independence, (2) a change in accounting conservatism and (3) the impact of board

independence on accounting conservatism. To measure accounting conservatism, the C-Score method of Khan & Watts (2009) was used. The sample used for the final regression

comprised of 2.391 firm-year observations from S&P500 companies over the period 2002-2012. Results show that there is a significant positive increase in board independence, as expected in the first hypothesis. However, the change in accounting conservatism resulted in a significant negative decrease, contradicting the expectations from the second hypothesis. Therefore, this study concludes that in times of crisis companies want to restore investors’ confidence by aggressive accounting. Finally, for answering the question whether the change in accounting conservatism due to the financial crisis is associated with the change in board independence the results are not significant, due to a lack of evidence.

Table of Contents

Abstract ...2

1. Introduction ...4

2. Literature Review and Hypotheses ...6

2.1 Corporate Governance ...6

2.1.1 The Agency Theory ...7

2.1.2 The Board of Directors and Board Independence ...8

2.2 Accounting Conservatism ...9

2.2.1 Reasons for Conservatism ... 10

2.3 The Interaction between the Board Independence and Accounting Conservatism ... 11

2.4 The Financial Crisis ... 13

2.5 Hypotheses Development ... 13

3. Research Methodology ... 15

3.1 Measures of Accounting Conservatism ... 15

3.1.1 The Basu Model ... 15

3.1.2 The Khan & Watts Model ... 16

3.2 Measures of Board Independence and the Crisis ... 17

3.3 Control Variables ... 18

3.4 Empirical Model ... 19

3.5 Sample Selection ... 20

3.6 Descriptive Statistics ... 22

4. Results ... 24

4.1 Change in Board Independence ... 24

4.2 Change in Accounting Conservatism ... 27

4.3 The Impact of Board Independence on Conservatism... 29

5. Conclusion ... 32

References ... 33

Appendix A: Definition of Regression Variables ... 37

1. Introduction

One of the main internal corporate governance tools within a company is the board of directors (Hermalin & Weisbach, 1988). The most important activities of the board are monitoring as well as giving advice to management of the company and protecting

shareholder interests. Advising management includes consulting on strategic and operational decision making in the best interest of the firm. Monitoring contains keeping track of

management with regard to activities itself, as well as management acting with care to the interests of the shareholders (Hermalin & Weisbach, 1988). Moreover, the board oversees, too, legal and regulatory compliance. A very important aspect is the separation and distinction of responsibilities between the board and management. Jensen & Meckling (1976) argue that conflicts of interests can arise between managers and other parties, as managers control the firm but are not company stakeholders, resulting in managers’ acting in their own interest instead of in the interest of the shareholders. To be effective in their role, members on the board must be characterized by independence as more independent boards ascribe higher quality reporting and lower earnings management (Ahmed & Duellman, 2007).

Much research has been done between corporate governance and the relation with accounting quality. The main area of focus within this field of research has primarily been on earnings management. Klein (2002) found a negative relation between board independence and abnormal accruals. Moreover, Xie et al. (2003) argued that the composition of an audit committee within the board is related to the likelihood that a firm will engage in earnings management. The extent of earnings management represents one possible element of

accounting quality whereas another aspect of quality is the topic of earnings conservatism and timeliness. Timeliness refers to the length of time taken to reflect information in earnings (Basu, 1997). Conservatism is accountants’ tendency to require a higher degree of verification for recognizing good news than bad news in financial statements (Basu, 1997). Accounting conservatism is a fundamental characteristic of financial reporting and is considered to be the most influential principle of valuation in accounting (Watts, 2003). In the past, auditors often ignored accounting conservatism as they argued that it did not represent reliable numbers (Givoly & Hayn, 2002). Nevertheless, accounting conservatism also has benefits for the users of financial information because it enhances efficient contracting, tax optimization and the reduction of litigation risk (Watts, 2003). Other literature has investigated the link between board structure and timeliness of earnings. Ahmed & Duellman (2007) and Beekes et al. (2004) reported that the number of independent, non-executive, board members and the

number of financial experts are positively related to accounting conservatism. However, there is no explanation in the existing literature to provide an explanation on the influence of the financial crisis on the relationship between independence of the board of directors and accounting conservatism.

The current financial crisis is seen as the crisis with the most impact since the Great

Depression (Vyas, 2011). The crisis originated in the financial sector and affected economies all over the world (Brunnermeier, 2008). This thesis researches the effect of this important event on board independence and accounting conservatism, adding a new feature to the incomplete literature.

Therefore, the following research question will be tested:

‘What is the effect of the financial crisis on the relationship between board independence and accounting conservatism’

This study contributes to the existing literature in different ways. Research has been done on the relationship between board independence and accounting conservatism, however, in relation to previous studies this research did not only use a data sample from before the financial crisis, but also a combination of the years prior and during the crisis. Therefore, this study distinguishes by differentiating between the pre-crisis and during crisis periods. By answering the research question, a contribution will also be made to corporate governance literature, as board of director is an important tool of corporate governance. Thus, this study provides evidence regarding the effect of board independence on conservatism. From a societal point of view, the study will contribute in a way that firms can anticipate to the financial crisis by adapting board independence and accounting in a conservative manner. In order to answer the research question, the research is divided in three parts. First board independence will be examined by measuring independence in the pre-crisis period (2002-2007) and the within-crisis period (2008-2012). Subsequently, the C-Score method by Khan & Watts (2009) will measure accounting conservatism. This method requires a regression analysis in order to calculate conservatism. Conservatism too, will be researched in the two different crisis periods. Finally, this study will examine whether the change in board independence has impacted accounting conservatism.

The remaining structure of this paper consist of a literature review of the key literature in the area of corporate governance, the agency theory, the board of directors, accounting

conservatism and the financial crisis in chapter 2. These areas are the core of the development of the hypotheses. Subsequently, the research method will be described in chapter 3, followed by the results in chapter 4. In chapter 5 a conclusion will be provided and limitations are discussed.

2. Literature Review and Hypotheses

2.1 Corporate Governance

Corporate governance has gained a lot of interest in the beginning of this decade due to the high amount of corporate collapses and financial scandals that have taken place in big

organizations such as Enron, Arthur Andersen, WorldCom, and Parmalat (Cohen et al., 2010). It called for more transparency and accountability to help restore investor confidence in the stock market and to protect these private and institutional investors. As a result, the United States presented a new law called the Sarbanes-Oxley act of 2002.

The Sarbanes-Oxley act is a law that set standards for U.S. public company boards, management and public accounting firms, and is an example of an external corporate

governance mechanism (Cohen et al., 2010). Denis & McConnell (2003) argue that corporate governance instruments can be divided into two categories, being internal and external corporate governance mechanism. The legal system, the takeover market and market competition can be described as mechanism of external corporate governance. Whereas internal corporate governance mechanisms are instruments that companies can deploy to create strong corporate governance, examples are the board of directors, committees and ownership structure.

Corporate Governance can be interpreted in many ways and several economists have been given several definitions over the time. These definitions range from purely economic definitions to more global, all-encompassing definitions. For example, Shleifer & Vishny (1997) argue that corporate governance deals with the ways in which suppliers of finance to corporations assure themselves of getting a return on their investment. Another definition for corporate governance is: “Corporate governance is the system by which business corporations

rights and responsibilities among different participants in the corporation, such as, the board, managers, shareholders and other stakeholders, and spells out the rules and procedures for making decisions on corporate affairs. By doing this, it also provides the structure through which the company objectives are set, and the means of attaining those objectives and monitoring performance” (OECD, 2004, p. 11). Finally, Merchant & Van Der Stede (2007)

define corporate governance as the set of mechanism and processes that help ensure that companies are directed and managed to create value for their owners while concurrently fulfilling responsibilities to other stakeholders.

These different definitions and interpretations can lead to a general view that outlines that corporate governance describes relationships between managers, the board, shareholders and other stakeholders. This ensures an adequate systems of controls, encourages transparency and accountability and aims to manage the company in the best interests of shareholders and stakeholders.

2.1.1 The Agency Theory

Corporate governance is a mechanism used to control the agency problem, and therefore the agency problem is one of the most important theories explaining corporate governance. The agency theory describes the problem that can occur in the principal-agent relationship between the company’s management and its shareholders (Jensen & Meckling, 1976). In most large public companies there is usually a separation between ownership and control. Within these companies the ownership is divided between different shareholders, who all have a say in the company. This theory defines shareholders as the principle. The actual control of the company is in the hand of the management, who are in charge of the day-to-day business decisions. The company’s management is called the agent. The agency theory defines the relationship between the principle and the agent and describes how to resolve conflicts between those parties (Jensen & Meckling, 1976). There are two main assumptions within the theory. The first one argues a conflict of interest between the principle and the agent. This is caused by the fact that managers and shareholders have different interests and different goals, and that the agent is risk-averse. These are reasons that the agent makes decisions based on self-interest, and those decisions are not the same as the principle would make, based on the goals and interests of the shareholder. This is due to the theory that every party will act in their own self-interest first (Jensen & Meckling, 1976). The second

of the agent and monitoring the actions of the agent is hard and costly. This leads to

information asymmetry because the principle does not have enough information, as he does not know the goals of the agent (Jensen & Meckling, 1976).

To conclude, corporate governance is an instrument to mitigate the conflicting interests in the relationship between the principal and the agent and to monitor the actions of the agent, thereby reducing agency problems. Moreover, Eisenhardt (1989) acknowledges this fact and argues that accounting standards and good corporate governance mechanisms are instruments to moderate agency problems.

2.1.2 The Board of Directors and Board Independence

Shareholders of companies do not have the knowledge and the time needed to control, monitor and guide the actions of the management, the company has a mechanism to perform these tasks, hence is done by the board of directors. As mentioned before, the board of directors in an internal corporate governance instrument to solve problematic issues between management and stakeholders (Fama & Jensen, 1983; Baysinger & Butler, 1985). The board of directors has the power to determine the structure of the management team as well was their remuneration. The board of directors has two main roles, the first role is monitoring and advise the management and the second is to protect the interest of the shareholders (Hermalin & Weisbach, 1988). Furthermore, the board supervises legal matters and regulatory

compliance.

There are a lot of characteristics that can be allocated to a board of directors. In order to be effective in their tasks, members of the board must have financial knowledge, do not have extensive board tenure, have an independent CEO and board size must not be too big

(Carcello et al., 2011; Dechow et al., 1996; Bowen et al., 2008; DeFond et al., 2005). Another very important characteristic of the board is to be independent (Klein, 2002), an independent board is characterized by more outside, non-executive, directors on the board compared to inside, executive, directors. Jensen & Meckling (1976) describe that boards with a high rate of outside directors help reduce agency problems due to the control of outside directors on opportunistic behavior of managers.

Additional research done on an independent board of directors shows that more independent boards attribute to lower earnings management and higher quality of information (Klein, 2002; Xie et al., 2003). Furthermore, the association between board independence and

fraudulent reporting has been researched by Dechow et al. (1996) who have investigated board independence and analysts’ rating of financial statements quality. An independent board with a high number of outside directors decreases the possibility of financial fraud, as Beasly (1996) found a negative relation between independent board member and financial fraud. Finally, Ahmed & Duellman (2007) found a positive relation outside directors and accounting conservatism, suggesting that a more independent board will use more

conservatism compared to a board that has more inside board members. 2.2 Accounting Conservatism

Accounting conservatism is a phenomenon, which was firstly described by Basu (1997). In his paper, Basu (1997) researched return data from firms in the period 1963-1990 by using a reverse regression method and found negative returns had a much higher impact on earnings than positive returns. From these results, Basu (1997) concluded that conservatism exists. Basu (1997) defines conservatism as “accountants tendency to require a higher degree of

verification to recognize good news as gains than to recognize bad news as losses” (p. 7).

This implies a systematic difference in timeliness of earning in a way that earnings are more timely and sensitive in reflecting bad news than good news. Givoly & Hayn (2000) also acknowledges conservatism within firms in the USA in the period 1950-1998, using several methods. Similarly to the study of Basu (1997), they found evidence of conservatism, and this conservatism increased over the researched years.

Studies have shown a number of ways in which conservatism can be useful. Conservatism is found to improve debt efficiency, leads to better managerial decisions on investments, and reduces conflicts between agents and principles as described in the agency theory (Watts, 2003; Ball & Shivakumar, 2005; Ahmed & Duellman, 2007). Also, Watts (2003) and Basu (1997) state that when earnings reflect bad news more quickly than good news, more useful information is generated leading to an increase in decision usefulness of financial reporting. For a long time auditors did not do anything with conservatism because they believed that it did not represented reliable financial statements (Givoly & Hayn, 2000). However, research showed that conservatism did increase over the years (Watts, 2003). Givoly & Hayn (2000) also recognized that companies were engaging in conservative reporting and found an increase of conservatism. Also, Ball & Shivakumar (2005) acknowledged the increasing importance of conservatism by showing a growth in conservatism in 2002, after the accounting scandals and corporate collapses of Enron, Arthur Anderson etc..

2.2.1 Reasons for Conservatism

The article of Watts (2003) explains the increase in use and the importance of conservatism, as found by several researchers (Lobo & Zhou, 2006; LaFond & Watts, 2008; Givoly & Hayn, 2000).This paper is the main research in the area of conservatism that explains the reasons for firms to report conservative. Those explanations are: contracting, litigation, taxation and accounting regulation (Watts, 2003).

The contracting explanation by Watts (2003) can be divided into compensation contracts and debt covenants. In general, contracting describes that managers have incentives to influence results in a positive way, but conservatism limits this because gains needs a higher degree of verification. This is positive for shareholders, because conservatism decreases their risks and this goes together with managers receiving less bonus rewards. This is called compensation contracts (Watts, 2003). Debt contracting is described by Watts (2003) as reducing risks for debt holders, because debt providers want to reduces the threats that come with lending money to companies.

Litigation is an explanation for conservatism as understated asset and earnings are much less subjected to litigation as the overstatement of assets and earnings (Watts, 2003). When a company does overstate its assets, the market will drop for the company when the

overstatement is found out. This drop results in losses for shareholders of the company, whom are harmed financially and can sue the company. This litigation threat is an incentive for auditors and managers to report conservatively, because otherwise they face high litigation costs (Watts, 2003).

The reason that regulations are related to conservatism is because when firms overstate their assets due to rules this could lead to losses for shareholders. Those shareholders can hold standard setters responsible which could result into political consequences for the regulators (Watts, 2003). This leads to regulators creating conservative rules, to protect themselves and protect shareholders for losses due to overstatements.

With regard to taxation, conservatism companies recognize all their costs made in the current fiscal year, but only recognize a part of the gains. This reduces their profits and reported income and leads to a lower tax payable in the current year (Watts, 2003).

2.3 The Interaction between the Board Independence and Accounting Conservatism Accounting conservatism and corporate governance, specifically the board of directors are connected to each other in several ways. Watts (2003) describes that conservatism helps to solve a bias that is created by information asymmetry. The information generated through accounting systems is used by directors on the board to monitor and advise managers. Managers also have information for the board of directors, however they can hold back important information, or information for the board can be biased in favor of the managers (Ahmed & Duellman, 2007). Since conservatism needs a higher degree of verification for good news, gains for example, the chance of managers overstating earnings decreases (Basu, 1997). So, accounting conservatism is a tool that averts managers from producing biased information, which leads to bonuses and overcompensation (Ahmed & Duellman, 2007; Watts, 2003).

Ball (2001) states that conservatism, with timely recognition of losses can help the board to recognize negative net present value (NPV) investments. Managers could try to postpone or even elude reporting losses negative NPV projects because this could cause negative

performance evaluation for the managers, timely recognition could help to prevent this (Ahmed & Duellman, 2007). So accounting conservatism can serve as a mechanism for directors on the board to identify negative NPV projects, and thus helping to identify and investigate bad investments by top management.

The probability of corporate collapses can be reduced through accounting conservatism, and shareholders could boost this accounting mechanism by communicating with the board of directors and their connection with top management (Lim, 2011). Reporting conservatively leads to a more timely disclosure of losses compared to gains. This way financial issues will surface in early stages (Garcia Lara, 2009). So through conservatism, the directors on the board notice losses and can intervene to prevent the company from collapsing (Lim, 2011). Fama & Jensen (1983) describe that agency problems cannot be solved only through a good board of directors, nor only with accounting conservatism. However, the connection between these instruments does lead to a reduction in agency problems and also to a fair representation of accounting information by managers. As previously described, this connection between accounting conservatism and the board of directors can lead to better monitoring of management, a decrease in biased information, the investigation of managers’ poor

2011). Garcia Lara (2009) also states that an effective board leads to more transparency, a decrease in earnings management and more independence between top managers in the company and directors on the board. An effective board can be defined as an independent board, which is a board with a high percentage of outside directors on the board.

Conservatism turns out to be a very suitable mechanism for the board to monitoring,

managing and advising management. Therefore, it can be assumed that boards that comprise of a high number outside directors would want more accounting conservatism. On the other hand, a board that has a high number of inside directors probably would want less accounting conservatism. To conclude, an independent board of directors will be positively related to accounting conservatism.

Beekes et al (2004) researches the connection between earnings timeliness and conservatism, and the composition of the board of directors. They argue whether different characteristics of the board lead to changes in conservatism and timeliness of earnings. The sample they use for their study consists of non-financial companies from the United Kingdom from the years 1993 to 1995. Beekes et al. (2009) used the Basu (1997) model for the reverse regression to

measure the timeliness of the timeliness of earnings for both good and bad news. The authors predict that firms with a high proportion of outside board members are timelier in the

recognition of bad news and less timely in the recognition of good news scenarios. The predictions are validated by their findings, which point out that firms with boards dominated with outside directors have increased earnings timeliness when it comes to bad news (Beekes et al., 2004). They conclude their research by stating that the composition of the board of directors is an important factor with respect to the timeliness of bad news in earnings and conservatism.

Ahmed & Duellman (2007) extended on the study by Beekes et al (2004). They also

researched the connection between conservative accounting and independence of the board, which are proxied by board of director characteristics. The sample they use consist of firms from the United States from the years 1999 to 2001. The research of Ahmed & Duellman (2007) differs in three ways of the research by Beekes et al. (2004). Firstly, as described, Ahmed & Duellman (2007) use a sample with firms from the US, they state that the results found by Beekes et al. (2004) might not hold in the US due to differences in legislation between the UK and US GAAP. Second, Beekes et al. (2004) use only one measure for accounting conservatism whereas Ahmed & Duellman (2007), besides from the Basu (1997) model, add two additional measures, being a market-value proxy and an approach by

Roychowdhury & Watts (2007) using backward-cumulation. Lastly, instead of only researching board independence, Ahmed & Duellman (2007) use a broader set board characteristics. Therefore, compared to Beekes et al. (2004), Ahmed & Duellman (2007) preform a more extensive research on a different sample. The findings by Ahmed & Duellman (2007) are the same as the findings from Beekes et al. (2004), which prove that the percentage of outside members on the board is positively related to conservatism. Also they find that the percentage of inside members on the board negatively related is to conservatism. Their results hold after controlling for variables like industry, firm size, leverage etc. Ahmed & Duellman (2007) conclude by stating that for the reduction of agency costs in the firm, accounting conservatism supports the board of directors.

Overall, taking the findings of Beekes et al. (2004) and Ahmed & Duellman (2007) into account, it can be argued that firms that comprise of a strong governance system, as an independent board of directors, are more likely to use conservative accounting.

2.4 The Financial Crisis

The collapse of the housing market in the United States in the end of 2006 was an important aspect in the start of a huge recession within the financial systems (Vyas, 2011).

Brunnermeier (2009) states that the stock market in the US lost around eight trillion dollar of its value between 2007 and 2008. Also, the recession led to a series of bankruptcies in

financial institutions like Lehman Brothers and Bear Stearns. The downfall of the US markets led to financial decline and affected markets worldwide due to economic globalization. Banks all over the world had to write down bad loans for hundreds of billions of dollars. The crisis not only affected banks and financial institutions, but it also led to declines in customer wealth, and downturns in economic activities. All these aspects let to the consideration that this 2007 financial crisis is the most severe financial crisis since the Great Depression of 1930 (Vyas, 2011; Brunnermeier, 2009). Governments, academics and other people are currently trying to figure out all the factors that triggered this financial crisis, and how can it be prevented in the future.

2.5 Hypotheses Development

This thesis researches whether the latest financial crisis had an effect on the relationship between the independence of the board of directors and accounting conservatism. To research this, three hypotheses were developed. Starting with the effect of the crisis on the variables

board independence and accounting conservatism. The relation between these two together with the moderating effect of the crisis will tested through hypothesis 3.

Weak corporate governance has been cited as one of the causes of the East Asian financial crisis (Mitton, 2002). Corporate governance could have made countries more vulnerable to a financial crisis. Today research again questions corporate governance, as the Organization of Economic Cooperation and Development (OECD) blames weak corporate governance as one of the main causes of the current financial crisis (Kirkpatrick, 2009).

As described by Jensen & Meckling (1976) (see paragraph 2.1) a good corporate governance system and boards with a high rate of non-executive directors help reduce agency problems due to the control of outside directors on opportunistic behavior of managers. Knowing this regulators want to improve corporate governance standards and companies will try take better care of stakeholders and investors. Corporate governance in the form of an independent board of directors is a mechanism to protect these stakeholders and investors, for their protection board independence should increase during the crisis. This leads to the following hypothesis:

H1: Relative to the pre-crisis period, board independence has increased during the financial crisis.

Research has been conducted on the effects of crises on conservative accounting, as this is not the first crisis companies had to withstand. Vichitsarawong et al. (2010) examined whether the Asian Crisis impacted accounting conservatism. Their results showed that during the crisis firms used less conservatism. A possible explanation for the decrease of accounting

conservatism is, to reduce the negative impact on firms, the need of these firms to release good news to rebuild investors’ confidence (Vichitsarawong et al., 2010). Gul et al. (2002) investigated the same crisis as Vichitsarawong et al. (2010) and measured the effects on conservatism focusing on Hong Kong. They acknowledge the fact that during the crisis period, firms were less conservative. Gul et al. (2002) give the explanation that companies operating in a financial crisis can lead to more aggressive accounting by managers compared to a period of prosperity.

Contrary to research that shows a drop in accounting conservatism, there are studies that show an increase in conservative reporting. Walsh et al. (1991) describe that in times of crisis companies can engage in big bath accounting. The crisis is a period were companies can make a clean sheet, as the stock market is already very low and reporting all losses does not affect

the stocks more. So, in periods of crisis there is an increasing level of conservatism (Walsh et al., 1991). Research done by Vyas (2011) on the 2007 financial crisis showed an increase in the number of lawsuits during this crisis and a wave of shareholder class action litigation. Conservative accounting is an instrument to reduce litigation risk (Watts, 2003). Not only the number of lawsuits increased during the financial crisis, Vyas (2011) found that contracting conflicts, and therefore contracting costs increased during the financial crisis. Both litigation and contracting are described by Watts (2003) as the most important reasons for conservative accounting. Accounting conservatism is a useful tool to reduce these litigations costs and contracting conflicts. Francis et al. (2013) describes that use of more conservative accounting is a way to control for agency costs.

Based on prior studies of accounting conservatism in crisis periods, it could be argued that conservatism decreases, and on the other hand, there is research suggesting an increase in conservative reporting. The research on the increase of conservatism outweighs the findings on a decrease. This leads to the following hypothesis:

H2: Relative to the pre-crisis period, accounting conservatism has increased during the financial crisis.

Conservatism is an important mechanism for accounting, and it also has positive effects on corporate governance. As described in paragraph 2.3 there is a relation between board of directors independence and accounting conservatism. In this thesis, the objective is to examine the impact of the change in independence of the board of directors on accounting conservatism, moderated by the financial crisis. This analysis leads to the third hypothesis:

H3: The change in accounting conservatism due to the financial crisis is associated with the change in board independence.

3. Research Methodology

3.1 Measures of Accounting Conservatism 3.1.1 The Basu Model

Basu (1997) describes accounting conservatism as the tendency to require a higher degree of verification to recognize good news as a gain than to recognize bad news as a loss. He created a regression model of asymmetric timeliness based on the differential reaction of good and

bad news in earnings. To measure conservatism, Basu (1997) uses stock prices as a proxy for good and bad news. These stock prices can be positive and negative, where negative stock returns indicate conservatism because losses are recognized earlier than profits.

The model of Basu is specified as:

Xi= β1 + β2Di + β3Ri + β4DiRi + εi (1)

Where i indicates the company, X is earnings, R is returns and D is a dummy variable that is used to show the effects for negative returns with 1 if return (R) is negative, and 0 otherwise and ε is the residual error.

Basu’s (1997) expectation is that earnings are more sensitive for bad news than for good news. β4 coefficient shows whether a company is conservative, β4 is higher when a company is more conservative. This shows that earnings are more sensitive for bad news than for good news. β3coefficient measures the sensitivity of good news, which is positive returns on earnings, because when annual returns are positive the dummy variable is 0.

The Basu model has certain limitations, of which one is by Givoly et al. (2007). They found that the level of conservatism is influenced by other factors like time periods, countries and reporting regimes, which are generally related to conservatism and therefore influencing the measurement method of Basu. Ball & Shivakumar (2005) state that the model of Basu (1997) does not make a distinction between profit or loss components into earnings that arise as a result of earnings management and random errors. Therefore, the model only identifies the existence of gains and losses, not the timeliness. Furthermore, Givoly & Hayn (2000) describe a limitation of the method of Basu, being that the method depends on the stock price

movements for bad news and good news. This leads to the fact that bad news is only reflected by negative stock returns, therefore assuming that every negative stock return is bad news and therefore a signal for conservatism.

Due to these limitations, and the fact that the Basu (1997) model does not take into account firms-specific measures, the C-Score method designed by Khan & Watts (2009) is used, as this method incorporates these measures for accounting conservatism.

3.1.2 The Khan & Watts Model

The model by Khan & Watts (2009) is based on the model by Basu (1997) and is designed to calculate the level of conservatism with the use of firm-year measurements. Khan & Watts

(2009) believe that firms within an industry react in a different way to conservatism and that these reactions are influenced by firm-specific characteristics. This differs from the Basu (1997) model of asymmetric timeliness, which measures conservatism in an industry. These firm-specific characteristics added by Kahn & Watts (2009) are market-to-book ratio, size and leverage.

To estimate timeliness reflecting both good news and bad news, Khan & Watts (2009) created the G-Score for the timeliness of good news, and the C-Score for the timeliness of bad news. These scores are created as a linear function of the added firm-specific characteristics, resulting in the following equations:

G-Score = β3= μ1 + μ2Sizei+ μ3M/Bi+ μ4Levi (2)

C-Score = β4= λ1 + λ2Sizei+ λ3M/Bi+ λ4Levi (3)

Where Size is the natural log of market value of equity, M/B is the market-to-book ratio, and Lev stands for leverage which is the debt-to-equity ratio.

As previously stated, the above equations are linear functions, by substituting β3 (G-Score) and β4 (C-Score) by respectively equations (2) and (3) in equation (1) and implementing additional terms to control for the firm-specific characteristics separately, Khan & Watts (2009) created the following empirical regression model:

Xi= β1 + β2Di + Ri(μ1 + μ2Sizei+ μ3M/Bi+ μ4Levi) + DiRi(λ1 + λ2Sizei+ λ3M/Bi+

λ4Levi) +( 1Sizei + 2M/Bi + 3Leveragei + 4DiSizei + 5DiM/Bi + 6DiLeveragei) +

εi (4)

Being only interested in the timeliness of bad news, the C-Score, the G-Score will not be determined. To estimate the C-Score for each firm-year and thus get to the measure of

conservatism of Khan & Watts (2009), a few steps have to be followed. Firstly, it is necessary to run the cross-section regression as stated as equation (4), where coefficients λ1 to λ4 are to be found. These coefficients helps calculating the C-Score as stated as equation (3). The higher the C-Score is, the greater the degree of conservatism.

3.2 Measures of Board Independence and the Crisis

This study researches the relationship between board of directors’ independence and conservatism while taking the financial crisis as a moderating variable into account. As

previously stated, conservatism is measured by the C-Score and calculated using the Khan & Watts (2009) method. Conservatism is the dependent variable in the second regression. The used independent variables are described with the following proxies:

Percentage of Non-Executive Board Members (%NEBM) is the proportion non-executive

members of the board of directors. Non-executive board members are seen as independent board members. Fama & Jensen (1983) describe a positive relationship between independent members on a board and accounting conservatism.

Crisis is a dummy variable, which takes the value of 0 when firm-year observations are in the

pre-crisis period (2002-2007), and 1 when the firm-year observations are in the within-crisis period (2008-2012).

Percentage of Non-Executive Board Members * Crisis (%NEBM*Crisis) is an interaction

variable showing the proportion of non-executive members on the board divided in the pre-crisis period and within-pre-crisis period.

3.3 Control Variables

In this study eight control variables are used to control for conservatism, because this is the dependent variable:

Size is a control variable for conservatism because larger firms have higher political costs

related to smaller firms. Higher political costs lead to more accounting conservatism as described by Watts (2003). Furthermore Ahmed & Duellman (2007) argue that large companies have more symmetric information and are timelier in disclosing information compared to smaller firms.

Market-to-book ratio complies as a control variable, as Khan & Watts (2009) describe that

accounting conservatism requires a higher degree of verification for losses compared to gains. This indicates a cumulative understatement of the book value of assets compared to the market value.

Leverage is a suitable control variable, as Watts (2003) describe that firms with a lot of

leverage are more likely to account in a conservative manner due to higher shareholder and debt-holder conflicts.

Volatility is used as a control variable because companies who have significant growth

options relative to assets, have more volatile stock returns, which is used as a proxy for firm-specific uncertainty (Khan & Watts, 2009). Volatility is likely to be higher in times of bad news compared to times of good news, so conservatism is expected to be positively related to volatility.

Cycle is short for investment cycle and is used as a control variable. Based on Khan & Watts

(2009) there is an expectation that the investment cycle is positively associated with

conservatism due to the increase of agency costs in this variable. This suggests the longer the investment cycle and the greater the firm-specific uncertainty, the more conservative the firm’s accounting.

Big 4 is used to control for companies that might act more or less conservative when having

an auditor from a Big 4 audit firm. The Big 4 auditors are Ernst & Young, Deloitte

PriceWaterhouseCoopers, and KPMG. Herrmann et al (2008) researched this connection and argue that companies engaged in more conservative accounting when audited by a Big 4 auditor. This acknowledges the research by Michas (2011), who found earnings reported in a more conservative manner by companies with a Big 4 auditor. Big 4 is a dummy variable which takes the value of 0 when a company is audited by a non-Big 4 auditor, and 1 when a company is audited by a Big 4 auditor.

Growth is the percentage of annual sales growth. Roychowdhury & Watts (2007) describe

that firms with high growth rates use less conservative accounting. Therefore, evidently growth and accounting conservatism are negatively related.

Profitability is another variable stated by Yunos et al. (2010) to be a common control for

conservatism. Yunos et al. (2010) argue there is a positive relationship between profitability and conservatism whereas companies with higher profitability apply more conservatism in accounting.

3.4 Empirical Model

With the independent, dependent and control variables the following empirical model is estimated:

Con = α0 + α1%NEBM + α2Crisis + α3%NEBM*Crisis + α4Size + α5M/B + α6Lev + α7Vol + α8Cycle + α9Big4 + α10Growth + α11Prof + ε (5)

Where:

Con Conservatism measured by the C-Score and calculated by the method of Khan & Watts (2009).

%NEBM The proportion of non-executive board members as described by the company.

Crisis Dummy variable; 0 for firm-year observations is the pre-crisis period (2002-2007) and 1 for firm-year observations in the within-crisis period (2008-2012).

%NEBM*Crisis The proportion of non-executive members on the board divided in the pre-crisis period and within-pre-crisis period.

Size Size of the company.

M/B Market-to-book ratio, market value of equity divided by book value of equity.

Lev Leverage, debt divided by equity.

Vol Volatility, the price movement of the company stock. Cycle Investment cycle of the company.

Big4 Dummy variable for Big4 auditor; 0 for companies audited by a non-Big4 auditor and 1 for companies audited by a Big4 auditor.

Growth Growth in annual sales of the company. Prof Profitability of the company.

3.5 Sample Selection

An important factor of this research is the financial crisis, therefore, companies from the United States are used as it is suggested that the crisis started in this country and that their stock markets were directly influenced and impacted early (Erkens et al., 2012). The sample consists of the firms that are included in the S&P 500 index. The sample comprises of the all the companies from the index excluding the financial and assurance firms as these companies are eliminated due to the fact that this sector had an important impact on the start of the crisis (Erkens et al., 2012). Moreover, also due to their circumstances of the companies in this period and because these companies’ accounting characteristics differ from other companies, making it hard to compare.

The data collected consists of firms between the years 2002 and 2012. This sample can be divided into two periods. Firstly the pre-crisis period, which is from 2002 to 2007. Secondly, the period is the within-crisis period, which is from 2008 to 2012.

The data is extracted from the Datastream database. Table 1 shows how the sample for the final regression is obtained. The total sample consists of firm-year observation from the companies in the S&P 500 for 11 book years. Financial and assurance firms, with SIC codes of 6000 to 6999, are excluded. Subsequently, firm-year observations with missing data are deleted from the sample. Lastly, the top and bottom 1% of the data of the variables used is trimmed using STATA. This is done in two steps, first the top and bottom 1% from the variables needed for the C-Score calculation are excluded, next the same is done for the second regression. This is to for control outliers, as described by Khan & Watts (2009). Therefore, the final sample is 2.391 firm-year observations.

Table 1 Sample Sampling procedure Number of observations Total sample 2002-2012 5513

After excluding financial institutions, real estate and assurance (SIC 6000-6999) 4391 After excluding firms with missing data for variables required

for regression 3119

After excluding firms with data in top and bottom 1% of variables required for

regression C-Score 2690

After excluding firms with data in top and bottom 1% of variables required for

regression 2391

Table 2 shows the firm-year observations divided in different industries, the classification of the firm-year observations by industry was made based on the SIC codes of the different companies. Manufacturing represents the largest part of the sample with 41,57%.

3.6 Descriptive Statistics

Table 3 defines the descriptive statistics of the independent variable, the dependent variables and the control variables. This table includes the observation, the mean, standard deviation and the minimum and maximum per variable.

Table 3 Descriptive statistics

Variable Observations Mean Std. Dev. Min Max

C-Score1 2391 -0,061 0,778 -5,860 5,834 NEBM 2391 68,711 10,505 27,190 80,150 Crisis2 2391 0,588 0,492 0,000 1,000 NEBMCrisis3 2391 41,247 35,310 0,000 80,150 Size 2391 16,355 1,024 14,147 19,226 M/B 2391 3,363 2,558 -10,392 19,227 Lev 2391 0,314 0,341 0,000 2,176 Big4 2391 0,998 0,041 0,000 1,000 Growth 2391 9,110 15,382 -35,330 79,150 Prof 2391 0,122 0,056 0,004 0,296 Cycle 2391 0,036 0,022 0,001 0,131 Vol 2391 26,020 7,882 12,740 52,260

Note: see Appendix 1 for variable definitions

1 Independent variable 2 Moderating variable 3 Interaction variable

Table 2 Observations per industry

Industry SIC code

Number of observations

Percentage of sample Agricultural, Production-Livestock and Animal

Specialties 1000-1999 182 7,61

Manufacturing 2000-3999 994 41,57

Transportation, Communication, Electric, Gas and

Sanitary 4000-4999 384 16,06

Wholesale and Retail Estate 5000-5999 305 12,76

Services 7000-8999 387 16,19

Not Available NA 139 5,81

Tabel 4 Correlation matrix

C-Score NEBM Crisis NEBMCrisis Size M/B Lev Big 4 Growth Prof Cycle Vol

C-Score 1,0000 NEBM -0,0267 1,0000 0,1918 Crisis -0,2092 0,1629 1,0000 0,0000* 0,0000* NEBMCrisis -0,2033 0,3060 0,9779 1,0000 0,0000* 0,0000* 0,0000* Size -0,1360 0,0501 -0,0048 0,0030 1,0000 0,0000* 0,0143** 0,8148 0,8851 M/B -0,1365 -0,0661 -0,0928 -0,1024 0,1642 1,0000 0,0000* 0,0012* 0,0000* 0,0000* 0,0000* Lev 0,1085 0,0719 0,0988 0,1082 -0,2618 -0,3058 1,0000 0,0000* 0,0004* 0,0000* 0,0000* 0,0000* 0,0000* Big4 0,0035 0,0026 0,0073 0,0036 0,0391 0,0088 0,0214 1,0000 0,8643 0,9006 0,7204 0,8585 0,0562*** 0,6654 0,2947 Growth 0,0756 -0,0573 -0,1690 -0,1678 0,0404 0,0573 -0,0950 0,0057 1,0000 0,0002* 0,0051* 0,0000* 0,0000* 0,0482** 0,0051* 0,0000* 0,7812 Prof -0,0430 -0,0934 0,0003 -0,0215 0,0887 0,2664 -0,2867 0,0050 0,1702 1,0000 0,0353** 0,0000* 0,9879 0,2943 0,0000* 0,0000* 0,0000* 0,8058 0,0000* Cycle 0,0202 -0,0552 -0,0933 -0,0935 -0,0415 -0,0083 0,0360 -0,0259 0,1367 0,3635 1,0000 0,3228 0,0069* 0,0000* 0,0000* 0,0426** 0,6861 0,0783*** 0,2058 0,0000* 0,0000* Vol -0,0148 -0,1090 0,0699 0,0497 -0,1488 0,0459 -0,0364 -0,0014 0,1516 0,0660 0,1149 1,0000 0,4708 0,0000* 0,0006* 0,0152** 0,0000* 0,0249** 0,0751 0,9468 0,0000* 0,0012* 0,0000*

Note: see Appendix A for variable definitions * significant at 0,01 level

** significant at 0,05 level *** significant at 0,10 level

Table 4 shows the correlation matrix for the variables used in the regression. From the matrix there can be concluded that NEBM (board independence) is a significant positive correlated with the financial crisis, 0,1629, which is described in the first hypothesis. The second hypothesis describes the connection between the crisis and the C-Score (accounting

conservatism). From the correlation matrix can be concluded that this is a significant negative correlation. However, hypothesis 2 expects a positive relationship between the crisis and the C-Score. Another relation worth taking a look at is the of the board independence and

accounting conservatism, the correlation matrix shows a value of -0,0267, there is a negative correlation between these two variables which means that when board independence

increases, conservatism decreases, which is opposite to the expectations.

4. Results

This chapter will answer the hypothesis using tests and regressions conducted. First, the change in board independence due to the financial crisis is tested. Second, the change in accounting conservatism due to the financial crisis is tested. Lastly, a regression analysis will show whether the change in accounting conservatism due to the financial crisis is associated with the change in board independence.

4.1 Change in Board Independence

H1: Relative to the pre-crisis period, board independence has increased during the financial crisis.

Figure 1 illustrates the change in board independence, with board independence measured by non-executive members on the board of directors from 2002 to 2012. Note, that in this paper the assumptions is made that the crisis started in 2008, as it emerged in the end of 2007.

The plotline in the figure shows that the percentage of non-executive members on the board is constantly changing, but generally increases over the years. Board independence fluctuates in the years before the crisis and slowly increases every year from 2007 to 2011. In 2011 board independence is at its highest point with a mean of 70,91.

Table 5 illustrates the descriptive statistics of board independence broken down by firm year and shows the mean, standard deviation, median, minimum and maximum of the percentage of non-executive board members. The total mean over the 11 year period of the analysis is 67,97. The mean for the pre-crisis period is 66,67, this is relatively lower from the mean within the crisis period which is 70,14.

56,0 58,0 60,0 62,0 64,0 66,0 68,0 70,0 72,0 74,0 76,0 78,0 20022003200420052006200720082009201020112012 Firm year

Figure 1 Change in Non-Executive Board Members in %

Mean Median Linear (Mean)

Table 5 Descriptive statistics Non-Executive Board Members by firm year

Obs Mean Std. Dev. Median Min Max

2002 91 63,9470 13,5312 68,69 28,6 78,28 2003 123 63,0269 13,0129 66,74 27,66 78,99 2004 116 67,6619 11,0776 71,59 30,88 79,91 2005 171 67,2030 11,2658 71,59 30,88 79,91 2006 257 66,7360 10,9382 71,41 28,79 80,04 2007 227 68,7366 9,3739 72,45 27,19 78,84 2008 239 68,8642 9,2398 73,25 28,82 78,29 2009 273 69,3890 9,4286 72,69 31,35 80,15 2010 292 70,4695 10,0154 74,11 29,3 80,15 2011 305 70,9053 9,6819 74,11 29,3 80,15 2012 297 70,7620 9,5372 75,17 30,81 79,83 Total 2391 67,9728 10,5160

Note: crisis dummy where '0' stands for the pre-crisis period (2002-2007) and '1' stands for crisis period (2008-2012)

From the means of the pre-crisis and within-crisis period there can already be seen an positive differences between these means. To test whether this increase is significant, and thus to test hypothesis 1, an independent T-test of a two-sample mean with unequal variances is

performed. The results of this T-test are shown in table 6, this table divides the data for an independent board in the pre-crisis period and within-crisis period, and shows the means of these periods, which are already discussed. The T-test gives a T-value of -7,8, meeting the 99% significance level and with a p-value of 0,000 it can stated that the increase in means is significant at a 1% level.

Since the positive change in the means of non-executive board members between the pre-crisis period and the pre-crisis period is significant, it can be concluded that hypothesis 1 is

accepted. Therefore, relative to the pre-crisis period, board independence has increased during the financial crisis.

Table 6 T-test1 of change in Non-Executive Board Members

Crisis Obs Mean Std. Err. Std. Dev. (99% Confidence Interval)

0 985 66,66632 0,3616295 11,34964 65,73302 67,59963

1 1406 70,14314 0,256509 9,61823 69,48151 70,80476

Combined 2391 68,71082 0,2148289 10,50467 68,15702 69,26463

Difference -3,476812 0,4433652 -4,62 -2,333623

T-value -7,8419

Satterthwaite's degrees of freedom 1888,45

Note: crisis dummy where '0' stands for the pre-crisis period (2002-2007) and '1' stands for crisis period (2008-2012) 1 Independent samples T-test of two-sample mean with unequal variances

4.2 Change in Accounting Conservatism

H2: Relative to the pre-crisis period, accounting conservatism has increased during the financial crisis.

Figure 2 shows the change in accounting conservatism, which is obtained by calculating the C-Score from 2002 to 2012. The descriptive statistics and the results from the regression analysis for calculating the C-Score can be found in Appendix B.

This figure shows a lot of fluctuation during the firm years, with an overall trendline of a decreasing mean for accounting conservatism. The peak in 2004 where the C-Score has a

-0,6 -0,4 -0,2 0,0 0,2 0,4 0,6 0,8 1,0 1,2 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 Firm year

Figure 2 Change in C-Score

Mean Median Linear (Mean)

mean of 0,7528 and the trough in this figure was in 2009 with a mean of -0,4196. The descriptive statistics of the C-Score, broken down by firm year are displayed in table 7. This table describes the mean, standard deviation, median, minimum and maximum values of accounting conservatism per year.

Table 7 Descriptive statistics C-Score1 by firm year

Obs Mean Std. Dev. Median Min Max

2002 91 0,0560 0,0209 0,0504 0,0239 0,1097 2003 123 0,1546 0,1664 0,1865 -0,5875 0,5200 2004 116 0,7258 2,8463 0,9690 -5,8596 5,8340 2005 171 0,0029 0,1388 -0,0250 -0,2473 0,4975 2006 257 0,1385 0,1535 0,0913 -0,0174 1,0237 2007 227 -0,0572 0,1273 -0,0453 -0,4195 0,1597 2008 239 -0,0410 0,1123 -0,0301 -0,5535 0,1946 2009 273 -0,4196 0,5289 -0,4972 -2,0626 2,5110 2010 292 -0,0404 0,8211 -0,4041 -4,5782 2,3997 2011 305 -0,1130 0,2691 -0,0392 -1,5261 0,3417 2012 297 0,0979 0,1745 0,1248 -0,5463 0,5219 Total 2391 -0,0608 0,7775

Note: crisis dummy where '0' stands for the pre-crisis period (2002-2007) and '1' stands for crisis period (2008-2012)

1

C-Score calculated with: Xi = β1+ β2Di+ Ri(μ1+ μ2Sizei+ μ3M/Bi+ μ4Levi) + DiRi(λ1+ λ2Sizei+ λ3M/Bi+

λ4Levi) +(δ1Sizei + δ2M/Bi +δ3Levi + δ4DiSizei + δ5DiM/Bi + δ6DiLevi) + εi

In the years before the crisis the means of the C-Score were constantly positive whereas in the years of the crisis, the means of the C-Score are negative, except for 2012. This increase in accounting conservatism can be interpreted as a sign that the financial crisis was coming to an end.

The decrease in accounting conservatism during the financial crisis is conflicting with the increase as stated in the hypothesis and therefore also contradicts the reasons for accounting conservatism as described by Watts (2003). It can already be concluded that hypothesis 2 is rejected due the fact that there is a decrease instead of an increase. However, to test whether this decrease in accounting conservatism over the years is significant, an independent T-test of two-sample mean with unequal variances is performed. The results from this test are displayed in table 8. According to this table the mean for the C-Score in the pre-crisis period is 0,1334 and the mean for the C-Score within the crisis period is -0,1970. The T-test results in a T-value of -9,4 of meeting a 99% significance level. Combined with a p-value of 0,000 it

can be established that the decrease in the means of accounting conservatism is significant at a 1% level.

Table 8 T-test1 of change in C-Score2

Crisis Obs Mean Std. Err. Std. Dev. (99% Confidence Interval)

0 985 0,1334212 0,0321293 1,008369 0,0505008 0,2163416

1 1406 -0,1969695 0,0138911 0,5208714 -0,2327994 -0,1611397

Combined 2391 -0,0608613 0,0159015 0,7775487 -0,1018535 -0,019869

Difference 0,3303907 0,0350036 0,2400999 0,4206815

T-value 9,4388

Satterthwaite's degrees of freedom 1353,14

Note: crisis dummy where '0' stands for the pre-crisis period (2002-2007) and '1' stands for crisis period (2008-2012)

1

Independent samples T-test of two-sample mean with unequal variances

2

C-Score calculated with: Xi = β1+ β2Di+ Ri(μ1+ μ2Sizei+ μ3M/Bi+ μ4Levi) + DiRi(λ1+ λ2Sizei+ λ3M/Bi+ λ4Levi)

+(δ1Sizei + δ2M/Bi +δ3Levi + δ4DiSizei + δ5DiM/Bi + δ6DiLevi) + εi

The result is contradictory of the expectation and hypothesis and can be explained with the article from Erkens et al. (2012). Erkens et al. (2012) researched the effect of independent boards on firm performance, and argue that companies with more independent boards have worse stock returns during the crisis period. The regression results for the C-Score calculation in Appendix B, indicate a significant role of returns in calculating the C-Score.

A decrease in accounting conservatism means that companies are accounting less conservative and more aggressive. More aggressive accounting can also found in the

researches of Vichitsarawong et al. (2010) and Gul et al. (2002) on accounting conservatism during the Asian crisis. The decrease in conservatism can be explained by companies having need to issue positive accounting numbers, good news, to restore investors’ confidence. To summarize, as the hypothesis argues that there is an increase in accounting conservatism expected in the crisis period, hypothesis 2 can be rejected as of the decrease in the C-Score instead of an increase. This decrease in accounting conservatism is tested for significance and resulted in a significant change. Therefore, it can be concluded that relative to the pre-crisis period, accounting conservatism has decreased during the financial crisis.

4.3 The Impact of Board Independence on Conservatism

H3:The change in accounting conservatism due to the financial crisis is associated with the change in board independence.

Previous paragraphs conclude the significant increase in board independence and a significant decrease in accounting conservatism. As a result, hypothesis 1 is accepted and hypothesis 2 was rejected. The last question to answer is whether the decrease of conservatism occurs due to the increase in the independence of the board of directors during the crisis. This paragraph shows the results from the regression analysis to answer hypothesis 3. To perceive the moderating effect of the financial crisis, on the relationship between board independence and accounting conservatism an interaction variable, %NEBM*Crisis, was added.

Before running the regression, the White-test for heteroscedasticity was performed, of which the outcome is summarized in table 9. White (1980) describes heteroscedasticity as a

condition where the variability of a variable is not equal across the values of another variable that predicts it. From the table it can be concluded that the p-value is very small, resulting in the rejection of the null-hypothesis for homoscedasticity and hence proof for

heteroscedasticity. To control for the heteroscedasticity, a regression with robust standard errors will be performed.

Table 9 White's test for Heteroscedasticity

Chi2(65) 127,92

Prob > Chi2 0,0000

White's test for H0: homoscedasticity H1: unrestricted heteroscedasticity

Table 10 shows the coefficients of the variables, the t-statistics and the p-value as obtained through the regression analysis. The influence of board independence on accounting

conservatism is very small positive, 0,0004536, a positive influence is also found by Ahmed & Duellman (2007) and Beekes et al. (2004). However, this influence is not significant, so this research lacks evidence to conclude this positive relationship between board

independence and accounting conservatism. The variable of crisis has a coefficient of -0,3198478 and is significant at a 10% level. This implies that accounting conservatism is indeed lower in the crisis period. This result also supports the conclusion on the T-test for hypothesis 2 (see paragraph 4.2) that accounting conservatism significantly decreased in the crisis period. The coefficient of the interaction variable of board independence and crisis is negative, -0,0004162, which could suggest that board independence had less impact on accounting conservatism and firms used less accounting conservatism because of the crisis itself. However the coefficient is insignificant, implying not enough evidence for this research

to conclude that the relation between board independence and accounting conservatism has changed during the financial crisis.

Table 10 Regression analysis1

Variables Coefficient t-statistics2 p-value

NEBM 0,0004536 0,20 0,841 Crisis -0,3198478 -1,80 0,071** NEBMCrisis -0,0004162 -0,15 0,878 Size -0,0791684 -4,98 0,000* M/B -0,0379265 -3,91 0,000* Lev 0,1737222 3,68 0,000* Big4 0,1437534 0,89 0,372 Growth 0,0029704 2,95 0,003* Prof 0,2821112 0,87 0,382 Cycle -0,785977 -1,13 0,259 Vol -0,0013562 -0,73 0,464 Constant 1,339229 3,92 0,000* Number of observations 2391 Prob> F 0,0000* R-squared 0,0889 Adjusted R-squared 0,0847

Note: see Appendix 1 for variable definitions

1

Conservatism = α0+ α1%NEBM + α2Crisis + α3%NEBM*Crisis + α4Size + α5M/B + α6Lev + α7Vol + α8Cycle +

α9Big4 + α10Growth + α11Prof + ε 2

The t-statistics are corrected for heteroscedasticity * significant at 0,01 level

** significant at 0,10 level

When going over the control variables, we can conclude that size and market-to-book ratio have a significant negative impact on the use of accounting conservatism, where the variables leverage and growth influence conservative accounting in a significant positive way. The positive effect from leverage on accounting conservatism is consistent with Watts (2003), as he argues that firms high leverage are more likely use accounting conservatism due to higher shareholder and debt-holder conflicts.

Thus, the very small negative coefficient of the interaction variable %NEBM*Crisis, describes the relation between board independence and accounting conservatism with the moderating effect of the crisis. This is insignificant and statistically meaningless due to lack of evidence, therefore rejecting hypothesis 3. Hence the change in accounting conservatism due to the financial crisis is not associated with the change in board independence.

5. Conclusion

This thesis investigated the relationship between the independence of the board of directors and accounting conservatism with the financial crisis as a moderating effect. This analysis used a comprised sample of 2.391 firm-year observations from S&P500 companies over the period 2002-2012. In order to examine the connection and the relation with the financial crisis in a sensible way, the research was broken down into three different hypotheses.

The first hypothesis deals with the change in board independence. Board independence was examined by calculating the means of board independence from the sample and broken down into the pre-crisis period (2002-2007) and the within-crisis period (2008-2012). This indicated an increase in board independence, as confirmed by T-test pointing out the significance. The second hypothesis covered the change in accounting conservatism. To estimate

accounting conservatism, C-Score method of Khan & Watts (2009) was used. This consisted of a regression and a formula to calculate the C-Score. Based on the reasons for accounting conservatism by Watts (2003), the expectation was that conservatism would increase in the financial crisis. However, the T-test performed showed a significant negative result. This was explained arguing that in times of crisis companies want to restore investors’ confidence by aggressive accounting (Vichitsarawong et al. (2010).

The final and third hypothesis discusses whether the change in accounting conservatism is related to the change in board independence. The performed regression analysis resulted in being not significant, due to lacking evidence, and therefore rejecting the third hypothesis. It was concluded that the change in accounting conservatism due to the financial crisis is not associated with the change in board independence.

Due to this thesis’ limitations, the results must be interpreted carefully. Two variables were excluded from the final regression, due to data limitations, the inclusion of these variables could have led to different results. Moreover, Khan & Watts (2009) argue that the C-Score does not optimally measures accounting conservatism. Additional research could resolve these limitations.

References

Ahmed S. and S. Duellman (2007). Accounting conservatism and board of director

characteristics: An empirical analysis. Journal of Accounting and Economics, 43: 411-413 Ball, R. (2001). Infrastructure requirements for an economically efficient system of public financial reporting and disclosure. Brookings-Wharton papers on financial services, (1): 127-169.

Ball, R., and L. Shivakumar (2005). Earnings quality in UK private firms: comparative loss recognition timeliness. Journal of accounting and economics. 39 (1): 83-128.

Basu, S. (1997). The Conservatism Principle and the Asymmetric Timing of Earnings,

Journal of Accounting and Economics, 24: 3-37.

Baysinger, B.D, and H.N. Butler (1985) Corporate governance and the board of directors: Performance effects of changes in board composition. Journal of Law, Economics, &

Organization. 1 (1): 101-124.

Beasley, M. S. (1996). An empirical analysis of the relation between the board of director composition and financial statement fraud. Accounting Review, 443-465.

Beekes, W., P. Pope and S. Young (2004). The Link Between Earnings Timeliness, Earnings Conservatism and Board Composition: evidence from the UK. Corporate Governance: An

International Review, 12: 47-59.

Bowen, R. M., Rajgopal, S., and Venkatachalam, M. (2008). Accounting Discretion,

Corporate Governance, and Firm Performance. Contemporary Accounting Research, 25 (2): 351-405.

Brunnermeier, M. K. (2008). Deciphering the liquidity and credit crunch 2007-08 (No. w14612). National Bureau of Economic Research.

Carcello, J. V., Neal, T. L., Palmrose, Z. V., and Scholz, S. (2011). CEO Involvement in Selecting Board Members, Audit Committee Effectiveness, and Restatements. Contemporary

Cohen, J., Krishnamoorthy, G., and Wright, A. (2010). Corporate Governance in the Post -Sarbanes-Oxley Era: Auditors’ Experiences. Contemporary Accounting Research, 27 (3): 751-786.

Dechow, P. M., Sloan, R. G., and Sweeney, A. P. (1996). Causes and consequences of earnings manipulation: An analysis of firms subject to enforcement actions by the SEC.

Contemporary accounting research, 13 (1): 1-36.

DeFond, M. L., Hann, R. N., and Hu, X. (2005). Does the market value financial expertise on audit committees of boards of directors?. Journal of accounting research, 43 (2): 153-193. Denis, D. K., and McConnell, J. J. (2003). International corporate governance. Journal of

Financial and Quantitative Analysis, 38 (1): 1-36.

Eisenhardt, K. M. (1989). Making fast strategic decisions in high-velocity environments.

Academy of Management Journal. 32 (3): 543-576.

Erkens, D. H., Hung, M., and Matos, P. (2012). Corporate governance in the 2007–2008 financial crisis: Evidence from financial institutions worldwide. Journal of Corporate

Finance. 18 (2), 389-411.

Fama, E. F., and M. C. Jensen (1983). Separation of ownership and control. Journal of law and economics. 26 (2): 301-325.

Francis, B., Hasan, I., & Wu, Q. (2013). The benefits of conservative accounting to shareholders: Evidence from the financial crisis. Accounting Horizons, 27(2): 319-346. Garcia Lara, J.M. (2009). Accounting conservatism and corporate governance. Review of

Accounting studies, 14 (1): 161-201.

Givoly, D., and Hayn, C. (2000). The changing time-series properties of earnings, cash flows and accruals: Has financial reporting become more conservative?. Journal of Accounting and

Economics. 29 (3): 287-320.

Gul, F., Srinidhi, B., and Shieh, T. (2002). The Asian financial crisis, accounting

conservatism and audit fees: Evidence from Hong Kong. Accounting Conservatism and Audit

Fees: Evidence from Hong Kong (Undated).

incentives on firm performance. Financial Management, 20: 101–112.

Herrmann, D. R., Pornupatham, S., and Vichitsarawong, T. (2008). The impact of the Asian financial crisis on auditors' conservatism. Journal of international accounting research, 7 (2): 43-63.

Jensen, M.C. and W.H. Meckling (1976). Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure. Journal of Financial Economics. 3, (4): 305-360.

Khan, M. and R.L. Watts (2009). Estimation and empirical properties of a firm-year measure of accounting conservatism. Journal of Accounting and Economics. 48 (2/3): 132-150. Kirkpatrick, G. (2009). Corporate Governance Lessons from the Financial Crisis. OECD

Journal: Financial Market Trends, 2009 (1).

Klein, A. (2002). Audit committee, board of director characteristics, and earnings management. Journal of Accounting and Economics, 33, (3): 375-400.

LaFond, R., and Watts, R. L. (2008). The information role of conservatism. The Accounting

Review, 83 (2): 447-478.

Lim, R. (2011) Are corporate governance attributes associated with accounting conservatism?

Accounting & Finance, 51 (4): 1007-1030.

Lobo, G. J., and Zhou, J. (2006). Did conservatism in financial reporting increase after the Sarbanes-Oxley Act? Initial evidence. Accounting Horizons. 20 (1): 57-73.

Merchant, K. A., and Van der Stede, W. A. (2007). Management control systems:

performance measurement, evaluation and incentives. Pearson Education.

Michas, P. N. (2011). The importance of audit profession development in emerging market countries. The Accounting Review. 86 (5): 1731-1764.

Mitton, T. (2002). A cross-firm analysis of the impact of corporate governance on the East Asian financial crisis. Journal of financial economics. 64 (2): 215-241.

OECD. (2004). OECD Principles of Corporate Governance. Last accessed 06 09, 2014, at http://www.oecd.org/corporate/ca/corporategovernanceprinciples/31557724.pdf.