The impact of corporate governance on overvaluation and its related effects

(1)

The impact of corporate governance on

overvaluation and its related effects

M

ASTER THESIS

Name Gijs Murk

Student number 10677461

Supervisor dhr. prof. dr. Vincent O’Connell Course MSc Accountancy & Control

Track Accountancy

University Amsterdam Business School

Faculty Economics and Business, University of Amsterdam

Date 22 juni 2015

(2)

Statement of Originality

This document is written by student Gijs Murk who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.

The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

(3)

Abstract

Overvaluation can result in management making poor investment decisions, fraud and earnings management. This paper investigates the influence of corporate governance on the overvaluation of a firm and the earnings management that is associated with this overvaluation. It is found that firms with a stronger board of directors are less likely to be overvalued. Even though a strong board governance leads to a reduced amount of earnings management, no evidence was found that this effect is stronger if firms are overvalued.

Implementation of SOX had a positive effect on the strength of corporate governance, which is increasing particularly fast between 2002 and 2007. Despite this positive effect, SOX did not give the board of directors more influence or incentives to prevent overvaluation.

(4)

1 Introduction

1.1 Background

In this study I examine how the strength of corporate governance affects the likelihood of a firm being or becoming overvalued. Furthermore I investigate how board governance moderates earnings manipulation that is related to overvaluation. Overvaluation is a market condition in which a firm is valued higher than its ability to generate earnings in the future (Jensen M. C., 2005). Jensen (2005) states that overvaluation leads to organizational forces that are difficult to manage and will destroy a firms long term value. He argues that overvaluation influences analyst forecasts and makes it impossible for top management to meet or beat earnings forecasts without destroying long term value. Jensen (2005) argues that these management targets create an incentive to increase short term performance at the cost of long term value creation.

Jensen & Meckling (1976) define an agency relationship as a contract between a principle and an agent, where the agent makes decisions for the principle in order to perform a service on their behalf. If both parties are trying to maximize their utility this would mean that the agent will not always act in the best interest of the principle, thus creating an agency problem and related agency costs. Jensen & Meckling (1976) separate agency costs into four categories: “contracting costs, monitoring costs, bonding costs and residual loss”. This means that the principle makes costs in selecting, monitoring and rewarding its agent and still suffers a residual loss, because it is impossible to perfectly align the interests of the agent with its own interests. Jensen & Meckling (1976) described how capital markets could potentially mitigate agency costs. However, Jensen (2005) argues that through overvaluation the markets could create or increase conflicts of interest through different value destroying forces, and thus increasing the agency costs for the firm.

Jensen (2005) describes how the causes for overvaluation are largely unknown, but it is partially caused by gamesmanship by capital markets and the management. When a firm manages to beat the earnings forecast, it’s stock prices increases approximately 5,5% more than its size matched portfolio. However, if a firm does not manage to meet its targets it will suffer a decrease in market price approximately 5,04% stronger than its size matched portfolio. According to Jensen (2005) the only way managers can repeatedly meet these targets set by analysts is to manipulate the books to mask the uncertainty in the business, which cannot be done without sacrificing firm value. Badertscher (2011) conducted empirical research on the different types of earnings management that firms use to manipulate the books, in relation to the overvaluation of these firms and the duration of the overvaluation. Badertscher (2011) found evidence that firms which are overvalued mainly use accruals based earnings

(6)

management during early stages of overvaluation and switch to real time earnings management and non-GAAP earnings management when the opportunities for accruals based earnings management are depleted.

Corporate Governance is implemented in order to prevent managers from enriching themselves at the expense of outsiders (Baber, Liang, & Zhu, 2012). Overvaluation generally has a positive effects on managers incentive payments and the value of their stock options. However, overvaluation has a negative impact on outside investors as the firm will not be able to generate earnings in the future to justify its value, which leads to reduced returns for investors in the future. This would make it a task for corporate governance to interfere in this process in order to protect its shareholders. This study investigates whether a stronger internal corporate governance system would be more capable of preventing overvaluation or the related earnings management.

1.2 Research Question

Due to the destructive effects that overvaluation has on shareholders, and the responsibility of corporate governance to protect these shareholders, the main research question for this study is: “Can overvaluation be mitigated by a stronger corporate governance?”.

1.3 Motivation

The increased interest in academic research on overvaluation, mostly caused by Jensen (2005), is the motivation for this study. Jensen (2005) requested more research on the causes and possible solutions of overvaluation. This study investigates whether a board is capable of preventing overvaluation and whether it has become more capable of doing so after the implementation of SOX. Badertscher (2011) found evidence to support that a firm increases its earnings management as overvaluation persists.

1.4 Structure

Chapter two discusses the theory on overvaluation, earnings management and corporate governance, as well as discussing the hypothesis development for this research. Chapter three describes the data and the methodology that has been used. Chapter four discusses the descriptive and empirical results found in this study. Chapter five discusses the conclusion, limitations and the implications of this research.

(7)

2 Literature review and hypothesis development

This chapter will discuss the theories used in this study and how these are related to the hypothesis and the research question. The first paragraph will discuss the theory on overvaluation. The second paragraph will discuss the theory on earnings management and the third chapter will discuss the theory on corporate governance. The fourth chapter will discuss the hypothesis used in order to answer the research question.

2.1 Overvaluation

For this paper I use the definition of overvaluation by Jensen (2005), he states: “By definition, an

overvalued equity means the company will not be able to deliver-except by pure luck-the performance to justify its value. If it could it would obviously not be overvalued.”. The ability to

deliver performance in the future is difficult to determine and estimations are often biased (Frankel & Lee, 1998). Furthermore, according to Jensen (2005) management is capable of lying about their performance through earnings management, which also influences the firms future expectations. This makes the detection of overvaluation a difficult matter as investors are not able to perceive which companies are overvalued and which are not. Jensen (2005) states that short selling could not solve these problems, however, he does not give any reasons for this conclusion. A possibility why short selling does not solve the problem is the duration of overvaluation, as has been researched by Badertscher (2011). This research concluded that overvaluation can occur for many consecutive years and can increase further over this duration, thus making it difficult to determine the timing in which the stock price will decrease. No empirical research for short selling on overvalued companies could be found. Another reason why short selling might not be profitable is due to the difficulty in detecting these overvalued firms or due to the costs and constraints of short selling. (Diamond & Verrecchia, 1987)

According to Jensen (2005) overvaluation is mostly caused by gamesmanship of management, who know that capital markets reward firms for meeting or beating analyst forecasts with a premium on the stock price. Firms which fail to meet or beat forecasts are punished in its stock prices. Dechow (2000) found that firms which are consecutively beating the benchmarks have higher levels of accruals and special items. This indicates that these firms might have used earnings management to beat their targets and to delay reporting bad news. Another possible cause for overvaluation is high amounts of speculation (Scheinkman & Xiong, 2003) (Blanchard & Watson, 1982) (Blanchard O. J., 1979). In the model by Scheinkman and Xiong (2003) the ownership of stock provides an opportunity to profit from other investments, rather than the underlying business. This means that shareholders can make a profit at the cost of other investors, by investing in a firm which is not capable of delivering

(8)

performance that would justify its value. Bubbles are usually ended by steep declines in which the stock price returns to its underlying economic values (International Monetary Fund, 2003).

In order to determine overvaluation there are different methods used in prior literature. A common way to determine overvaluation is by using expected earnings to calculate the net present value of future dividends (Edwards & Bell, 1965) (Ohlson, 1995). Comparing these future dividends to the current market value can be used to determine whether or not a stock is overvalued. This view is often biased as earnings forecasts can be overly optimistic and based on past performance, which could have been manipulated (Frankel & Lee, 1998).

A different way of determining overvaluation is by comparing the market value to the book value of equity. By dividing the market value with the book value of equity it can show how much of a firms value is not reflected in its equity. According to Rhodes-Kropf et al. (2005) the market-to-book ratio can be a determinant of overvaluation. Firms can create value that is not reflected on the balance sheet through certain forms of less tangible assets, such as reputation, know-how, customer lists and brands (Cole, 2012). The amount of value creation that is not reflected on the balance sheet is similar among each industry, according to Rhodes-Kropf et al. (2005). Comparing firms market-to-book ratio to the normal levels of the industry would control for the relative assets that are not reflected in the balance sheet, and would allow determination of which firms are overvalued. Besides market-to-book ratio, there is also a ratio that takes the full assets and liabilities of a firm into account, called ‘Tobin’s Q’ (Tobin, 1969). This ratio is calculated by dividing the total market value of equity and liabilities by the total assets of the firm. It calculates how much more (or less) a firms value is compared to the replacement costs of its assets. As Tobin’s Q takes the full market value of liabilities into account and not just the market value of equity it is less sensitive for firms with high levels of financial leverage than market-to-book ratio is.

Jensen (2005) theorizes that overvaluation has a harmful impact on long term firm performance. As overvaluation starts occurring the manager has two options. The manager can either communicate to the market that he cannot meet the expectations set by the market, either by telling the market directly or by not meeting the next forecast. The second option is to inflate the reported performance in order to justify the inflate stock price (Chi & Gupta, 2009). Considering that the first option has a negative effect on the managers compensation and career, the manager is incentivized to use the second option and lying to the market. In order to maintain this overvaluation for longer periods the manager would have to overinvest in acquisitions or expansions (Shleifer & Vishny., 2003), commit fraud (Jensen M. C., 2004) and/or manage earnings (Badertscher, 2011) (Chi & Gupta, 2009).

Jensen (2004) discusses how the major frauds of 2001, such as WorldCom, Enron and Xerox where all related to these firms trying to maintain their overvaluation. Jensen (2005) uses

(9)

Enron as an example, where he states that Enron was valued at $70 billion, while its underlying economic value was only $30 billion. Jensen (2005) believes that the senior management of Enron destroyed the core value of $30 billion, because of their efforts to defend the $40 billion of overvaluation. Jensen (2005) believes that if they had talked down their value, they would have been able to defend their $30 billion of core value, instead of the bankruptcy that occurred in 2001.

Shleifer & Vishny (2003) investigate the relationship between the misvaluation of a company and its behavior in mergers and acquisitions. They note that firms which are overvalued are more likely to make acquisitions, while undervalued firms are more likely to be the target of acquisitions.

2.2 Earnings Management

Schipper (1989) defines earnings management as “a purposeful intervention in the external

financial reporting process, with the intent of obtaining some private gain”. In this definition

managers use earnings management to mislead stakeholders about the firm’s performance. Ali Shah et al. (2011) document six reasons for managers to manage earnings: to meet internal targets, meet external expectations, provide income smoothing, window dressing, taxation and change in management. These different motivations show that some types of earnings management are for the benefit of the firm, potentially at the cost of its stakeholders. An example for this would be to understate provisions in order to show better debt/equity ratios and gain favourable terms on a loan or in order to stay within debt covenants. Other types of earnings management are for the benefit of the manager, at the cost of the stakeholder. An example for this would be to overstate earnings in order to increase compensation.

Earnings management can be divided into accruals based earnings management and real time earnings management. According to Healy (1985), accruals based earnings management is when the manager modifies the timing of reported earnings to enable managers to transfer earnings between years. This is done mostly through discretionary accruals, which managers can modify through accounting without influencing the underlying business. Healy and Wahlen (1999) claim that subjectivity in reporting has certain costs and benefits. Managers using earnings management and potential misallocation of recourses that results are costs of having discretion in accruals. However, management could also use this subjectivity to improve credibility of information that is communicated to stakeholders and improve the decision usefulness of this information. The purpose of accruals is to account revenues and expenses to the period in which these occurred (FASB, 1985). Healy and Wahlen (1999) request further research to identify to which extend these accruals are used to improve communication vs

(10)

earnings management. Accruals are also subject to auditor scrutiny and legislation, such as GAAP and SOX (U.S Securities and Exchange Commission, 2002).

Managers also have the opportunities to manipulate earnings through real operating activities during the year. Through the use of investment activities, sales manipulation or overproduction the managers are able to transfer earnings between years (Roychowdhury, 2006). This form of earnings management has an influence on the actual processes and cash flows of the organisation. Real time earnings management is considered costly and has a negative impact on firm value (Roychowdhury, 2006). Due to its nature, real time earnings management is less likely to be subjected to auditor and regulation scrutiny (Cohen, Dey, & Lys, 2008). Changes in regulations due to the implementation of SOX have resulted in a shift from accruals based earnings management towards real earnings management (Cohen, Dey, & Lys, 2008).

Many different researchers have generated models in order to identify the levels and directions of earnings management that is being used in firms, for both accruals based earnings management (McNichols, 1988) (Healy P. M., 1985) (DeAngelo, 1986) (Jones, 1991) (Kothari, 2005) (Dechow & Sloan, 1991) and real time earnings management (Roychowdhury, 2006). A study by Dechow et al. (1995) compared different models for detecting accruals based earnings management and found that the modified Jones model had the most power in detecting accruals based earnings management. The model by Roychowdhury (2006) is used in different studies in order to detect real time earnings management (Zang, 2006) (Gunny, 2005) (Cohen, Dey, & Lys, 2008) (Badertscher, 2011).

Even though identification of earnings management has received much attention from the research, a study by Sloan (1996) concluded that investors do not use the information in accruals properly. This could potentially be caused by investors being unable to identify the magnitude of these accruals properly and are unable to make proper estimates for the reversal of these accruals. Lo (2008) theorized that despite it being known that earnings management exists, it is still difficult for researchers to identify whether abnormal changes are caused by earnings management or natural causes within the firm.

Earnings management has several effects on firm performance. As stated earlier, earnings management could be used to benefit the firm and its shareholders. Fukui (2000) theorizes that investors have the most interest in a firms permanent income. Since managers are more capable of making estimates of a firms permanent income, it would be in the shareholders best interest if the managers would report smoothed earnings. This means that managers should use earnings management to smooth earnings, if these are declines caused by transient shocks which do not influence a firms permanent income. Arya et al. (2003) compare income smoothing to a car ride: “A smooth car ride is not only comfortable; it also reassures the

(11)

passenger about the driver’s expertise”. Furthermore, as examined by Cohen et al. (2008), if

managers are limited in their ability to use accrual based earnings management, they will switch to real time earnings management. This form of earnings management is less likely to be scrutinized by auditors or regulators according to Roychowdhury (2006).

Despite the positive effects that earnings management could have on a firms performance, managers could also use it to enrich themselves at the cost of its shareholders (Watts, 1977). According to Healy (1985), managers use earnings management to change the contracting outcomes of their compensation plans and to increase the value of their stock options. Managers use discretion in accounting to maximize their own utility in accordance with their bonus plans and stock options. Efendi et al. (2007) concluded that managers with ‘in the money’ stock options are more likely to issue restatements to the financial reports. Jensen (2005) expands on this by theorizing that managers of overvalued firms use earnings management in order to justify the high valuation of their firms. Previous earnings management inflates the firms expected permanent income to levels it will not be able to maintain in the long term. Jensen (2005) believes that real operating decisions that would maximize firm value are compromised to meet short term market expectations, thus destroying long term value. Jensen (2005) describes how managers get stuck in situations where the market will punish them for not meeting their targets, but their short term targets cannot be met without sacrificing long term firm value, thus postponing and increasing the problem. This means managers will have to increase their accounting manipulation or turn to fraud in order to continue the appearances of growth and value creation. These ideas where researched by Badertscher (2011) who found empirical evidence that overvalued firms use increased amounts of earnings management and switch to real time and non-GAAP earnings management after several years of overvaluation.

2.3 Corporate Governance

Fama and Jensen (1983) describe problems in a firm that arise when a firm separates its ownership from control. This separation causes an agency problem as described by Jensen & Meckling (1976) due to the managers having different incentives than the owners of the firm. Both parties are trying to maximize their own utility, which can cause managers to act in their own best interest rather than that of the firm. Jensen and Meckling (1976) describe that the owners of the firm incur costs in the contracting, monitoring and bonding of their managers and suffer a loss from unaligned interests, referred to as residual loss. Fama and Jensen (1983) describe a board of directors as an important monitoring tool. The board of directors has the power to hire, fire and compensate top level managers and to ratify and monitor important decisions. Fama and Jensen (1983) theorize that decision making as a group helps align the interests of the owners and its management. Most organisation in which ownership and control

(12)

is separated have some form of a board of directors. Ruigrok et al. (2006) concluded that the board of directors is also involved in setting a firms strategy.

A board of directors consists of inside directors and outside directors and is led by a chairman. The inside directors serve a meaningful function within the organisation in addition to serving the board, while an outside director merely serves the board. In Europe most firms have their outside directors in a supervisory board that runs a company alongside the management, also referred to as a two tier board (Higgs, 2003). In the United States there is generally one board that consist of both inside and outside directors, also referred to as a one tier board (Higgs, 2003). Besides this board of directors the corporate governance also consists of separate committees with specific functions. These committees consist of several board members which meet separately from the full board. Every firm in the S&P 500 has at least one committee, but Klein (1998) found that some firms utilize up to eight different committees that meet separately from the board. Some example of separate committees are the audit committee, investment committee, nominating committee and the remuneration committee.

Baysinger and Butler (1985) describe that board differ considerably among firms in different attributes, such as the ratio of insiders to outsiders, the number and functions of committees and the activism of the board. Research has been dedicated to identifying how changes in board or committee characteristics could influence firm performance. Considering the roles of the board of directors in monitoring and setting the firms strategy, O’Connell and Cramer (2010) found evidence that the composition of a board of directors plays an important role in firm performance. The study by O’Connell and Cramer (2010) concluded that firms with higher percentages of independent directors have better financial performance, which could indicate that these independent directors increase the utility for firm shareholders. Dechow et al. (1996) found a correlation between earnings management and some board characteristics. Dechow et al. (1996) conclude that weak boards with poor oversight have higher degrees of earnings management.

Baber et al. (2012) designed a method for determining a boards strength, referred to as ‘board index’. This board index is based on governance characteristics and is designed to be a proxy for board strength. Considering the measurement and the weights of these characteristics, it would be impossible to identify a score system that would accurately show how much influence a board of directors really has over the decisions of the executives. However, the study by Baber et al. (2012) provides a method to estimate governance strength based on critical governance characteristics. The board index takes the following characteristics into account: Board independence, audit committee independence, nominating committee independence, compensation committee independence, CEO is not chairman of the board and board size. How these are measured and calculated is described in chapter three.

(13)

Rules and regulation also play an important role in the functioning and design of corporate governance. Particularly, the implementation of the Sarbanes-Oxley act (SOX) in 2002 had a large impact on the boards of directors in the U.S (U.S Securities and Exchange Commission, 2002). SOX requires all audit committee members to be fully independent and increased the responsibilities for the board of directors. Besides the implementation of SOX the Organisation for Economic Co-operation and Development (OECD) also published corporate governance principles in 1999 (OECD, 1999). These principles contain a set of standards and guidelines on how firms should improve their corporate governance regime.

2.4 Hypothesis Development

The paper by Jensen (2005) created an interest in overvaluation in the accounting field, particularly in the destructive effects on long term firm performance that are associated with this overvaluation. Jensen (2005) theorizes how overvaluation is caused and maintained due to overinvestment in acquisitions or expansions, fraud or earnings management. Considering the responsibilities of corporate governance to monitor and ratify the decisions made by the CEO, the board of directors might be able to mitigate these effects thus preventing overvaluation from occurring or reverting the share price back to the true underlying value of the organisation. I predict that overvaluation is more likely to occur if a firm has a weak corporate governance.

H1: Firms with a weaker corporate governance are more likely to be overvalued

Badertscher (2011) investigated the relationship between overvaluation and earnings management. He noted that firms that are overvalued have increased amounts of earnings management, particularly when the duration of overvaluation continues. A study by Dechow et al. (1996) concluded that earnings management is more common among firms with weaker corporate governance. I predict that a strong corporate governance can mitigate earnings management for overvalued firms.

H2: Overvalued firms have reduced earnings management if Corporate Governance is stronger

The implementation of SOX has had a large impact on corporate governance and the business world. Baber et al. (2012) concluded that firms significantly improved their corporate governance characteristics between 1997 and 2005. Besides the increase in board index I also expect corporate governance to be able to better mitigate overvaluation after the passage of SOX.

H3: Implementation of SOX resulted in Corporate Governance being able to better reduce overvaluation

(14)

3 Data and research methodology

This section will expand on the process of data collection and the research methodology. The first paragraph will describe the data collection process. The second paragraph will describe how this paper determines whether or not a firm is overvalued. The third paragraph will show the measurement of corporate governance strength as a board index score. The fourth paragraph will show the calculations for determining the direction and the magnitude of earnings management, both real time and accruals based.

3.1 Sample

Data has been collected from the Compustat database between 1996 and 2013, resulting in 230,130 unique firm observations. First the firms with missing market values have been removed, as the valuation of these firms cannot be determined. This resulted in a loss of 66,138 observations. Second, the financial industry sector (2 digit SIC codes 60-67) is excluded because this is a specific industry in which book value of equity is likely not to be the most important driver for valuation, due to changes in the regulatory environment creating large shifts in firm value (Damodaran, 2009). This means that the technique I use to calculate whether overvaluation is occurring would not be suitable for this industry, resulting in a deletion of 30,366 observations. Third, the observations with missing data from ISS (RiskMetrics) or other relevant Compustat items are excluded, resulting in a loss of 115,709 and 3,049 observations respectively. Finally, the observations from the agriculture group (2 digit SIC codes 01-09) and the observations from the public administration (2 digit SIC codes 91-99) have been removed, as these groups contain too few observations to compare them accordingly. This resulted in a loss of 48 and 36 observations respectively. The final sample contains 14,784 unique year observations among 2,001 unique firms. The summary of data collection is provided in table 3.1. The variables gathered from Compustat and ISS can be found in appendix A. The distribution of this sample among the different industry groups and the averages in Compustat adjusted by the removed SIC groups is included in table 3.2. This summary shows that there are no notable differences from the Compustat averages.

(15)

Table 3.1: Sample selection

Table 3.2: Distribution of sample by Industry

SIC

code Industry Group

Sample firms Sample percentage Compustat percentage adjusted for removal 10-14 Mining 698 4.72% 5.18% 7.64% 15-17 Construction 140 0.95% 0.83% 1.22% 20-39 Manufacturing 7,970 53.91% 32.41% 47.80%

40-49 Transportation & Public Utilities 1,884 12.74% 9.94% 14.66%

50-51 Wholesale Trade 623 4.21% 2.46% 3.63%

52-59 Retail Trade 1,428 9.66% 4.18% 6.17%

70-89 Services 2,041 13.81% 12.80% 18.88%

Total 14,784 68% 100%

3.2 Measurement of overvaluation

In order to calculate whether a firm is overvalued I used three different methods, which are explained in detail below. All methods are calculated empirically by using firm fundamentals data from Compustat. No analyst forecast data has been collected from IBES as these forecasts tend to be subjective and vary with respect to many different factors, such as the number of analysts following the firm (Frankel & Lee, 1998). Furthermore, according to Frankel and Lee (1998) analyst forecasts can be biased and overly optimistic, particularly for firms with higher growth rates and higher price to book ratios. The combination of subjectivity and bias would make it difficult to determine which firms are actually overvalued.

The first method used to calculate overvaluation is by using the Price-to-Book ratio (P/B) of the firm-year observation. This is calculated by dividing the market value of the firm by its total equity. Market value is extracted from Compustat or calculated by multiplying the number of shares outstanding with the close price at the end of the fiscal year if the market value is missing. These ratios are ordered and divided into quantiles by grouping them on their three digit industry codes. If the sample is smaller than 20, the observations are moved and

Unique Compustat database 230,130

Less observations with missing market values 66,138 Less observations from financial services (sic #60-67) 30,366 Less observations not recorded in ISS (riskmetrics) 115,709 Less observations with missing data 3,049 Less observations from Agriculture (sic #01-09) 48 Less observations from public administration (sic 90-99) 36

Total removed observations 215,346

Total number of observations in final sample 14,784

(16)

compared to their two digit industry code. Groups smaller than 20 on two digits are compared on their single digit industry code. Firms with a quantile value of five are marked as overvalued in P/B ratio.

𝑇𝑜𝑏𝑖𝑛

′

_{𝑠𝑄 =}

𝑀𝑎𝑟𝑘𝑒𝑡 𝑉𝑎𝑙𝑢𝑒+𝐿𝑖𝑎𝑏𝑖𝑙𝑖𝑡𝑖𝑒𝑠

𝐸𝑞𝑢𝑖𝑡𝑦+𝐿𝑖𝑎𝑏𝑖𝑙𝑖𝑡𝑖𝑒𝑠 (Equation 1)

The second and third method to calculate overvaluation are based on the Tobin’s Q of the firm-year observation. The Tobin’s Q is calculated using equation 1. In the second method this Tobin’s Q is ordered and divided into quantiles using the same method as was discussed previously with P/B ratios. The third method divides the Tobin’s Q into first level industry groups. These groups are placed into quantiles based on their firm size. New groups are generated for each first level industry group and size quantile and the Tobin’s Q values are divided into quantiles compared to the group they are in. For every group the top quantile is marked as overvalued. The breakdown for the three methods and how overvaluation was determined based on the three previous methods is specified in table 3.3.

The results of the three methods and the differences between them are tabulated in table 3.3. The observations which are marked in all three methods are identified as overvalued for further regressions, reducing the amount of overvalued firms from approximately 2,800 per method to 1,599.

Table 3.3: Overvaluation on three methods

The method for determining overvaluation was drawn from a study by Rhodes-Kropf et al. (2005). This study used Market-to-Book ratios as a method for determining misvaluation and how this affected mergers and acquisitions. In addition I also use Tobin’s Q, as market-to-book ratio only calculates the market value of equity, while Tobin’s Q calculates the market value of all assets and compares a firm value to its replacement value (Tobin, 1969). In addition, I controlled for firm size, but in order to do so I had to broaden the industry groups to inhibit too many groups with small amounts of observations. I decided to combine both methods to display firms which have increased Tobin’s Q compared to their three digit industry codes and compared to their size and first digit industry code.

Firms identified as overvalued by MTB ratio (method 1) 2,809

Total observations that are not overvalued 1,210

Final number of overvalued observations 1,599

Less observations which are not overvalued by Tobin's Q,

controlled for first digit industry code and firm size (Method 3) 402 Less observations which are not overvalued by Tobin's Q,

(17)

I divided the results into quantiles, in which firms with a 5 have the highest valuation and firms with a 1 have the lowest valuation. I used quantiles in order to make the three methods comparable, which is also used by Beck and Mauldin (2014). Finally, I created two new variables; ‘overvalued’ and ‘undervalued’. I marked an observation as overvalued if it scored a 5 in all three methods, while I marked it as undervalued if it scored a 1 in all three methods. This resulted in 1,599 overvalued firms and 2,171 undervalued firms in the sample. By comparing the three methods I strive to increase the robustness in identifying overvaluation.

3.3 Measurement of internal governance

After determining the overvaluation I calculated the strength of the internal corporate governance. In line with Baber et al. (2012) I use six board attributes to measure the strength of board governance, expressed in a score of 0 to 6, referred to as ‘Board Index Score’, or B-score for short. B-score increases by one if each of the following conditions is met: (1) More than 2 out of 3 board members are independent. (2) There is an independent audit committee. (3) There is an independent compensation committee. (4) There is an independent nominating committee. (5) The CEO is not the chairman of the board. (6) The board size is greater than the median, adjusted for firm size and the year. Variables used to compute B-score and overvaluation are summarized in exhibit 3.1. The CEO not being a chairman improves internal governance, because according to Jensen (1993) it is much more difficult for a board to perform its critical function without the direction of an independent leader. Jensen (1993) also states that the chairman should be given the right to initiate board meetings, determine committee assignments and set the board’s agenda (along with the CEO). The actual tasks and rights of the chairman cannot be measured empirically and will not be part of this study. The independence of the directors and the different committees plays an important role for reflecting the prominence of the internal governance in academic studies (Jensen M. C., The modern industrial revolution, exit, and the failure of internal control systems, 1993), and is also part of proposed rules and regulation on the governance of public corporations (Klein, 2003).

(18)

Exhibit 3.1

3.4 Measurement of Earnings Management

According to Badertscher (2011) there are many different methods which firms can use to manage their earnings and to disguise their true economic performance, some within GAAP and others outside of GAAP. Badertscher (2011) separates earnings management into three categories: Accruals based, Real Time and Non-GAAP earnings management. Due to the difficulty in identifying non-GAAP earnings management and the limited frequency of these occurrences, the measurement of non-GAAP earnings management has been left out of this study. In this study I identify accruals based earnings management and three types of real time earnings management. All methods and the methods used to identify and calculate them are described below.

First I calculate accruals based earnings management using the modified Jones model (Kothari, 2005). Dechow et al. (1995) examine several different measures of calculating accrual based earnings management. This study compares the models by McNichols and Wilson (1988), Healy (1985), DeAngelo (1986), Jones (1991), the modified Jones model (Kothari, 2005) and the Industry model by Dechow and Sloan (1991). Dechow et al. (1995)conclude that the modified Jones model had the most power in detecting accruals based earnings management, as this model exhibits the most power in the regressions. I follow Dechow et al. (1995) and use the

Variable Description

Internal Governance

B-Index Board Index Score Variables Used to Compute B-Index

Independent Number of independent board members in the firm

Board Size Number of total board members in the firm

Audit_Ind Percentage of audit committee that is independent (B2 is 1 when 100%)

Com_Ind Percentage of compensation committee that is independent (B3 is 1 when 100%)

Nom_Ind Percentage of audit nominating that is independent (B4 is 1 when 100%)

CEO_Chairman 1 if CEO is chairman of the board (B5 is 1 when 0) Control variables

asset Total Assets (Firm size)

SIC Standard Industry Classification Code Overvaluation

MTB_over 1 if firm scores in the top quantile of MTB, adjusted for 3 digit SIC code

Tobin1_over 1 if firm scores in the top quantile of Tobin's Q, adjusted for 3 digit SIC code

Tobin2_over 1 if firm scores in the top quantile of Tobin's Q, adjusted for 1 digit SIC code and firm size

Overvalued 1 if firm scores a 1 on all previous methods of determining overvaluation Variables used to compute overvaluation

mkvalt Total Market Value

CSHO Common Shares Outstanding

prrc_f Closing price fiscal year (annual)

SEQ Stockholders Equity (total)

liability Total Liabilities Control variables

asset Total Assets (Firm size)

(19)

modified Jones model for detecting accruals based earnings management in my sample. Equation 2 shows the equation used to calculate the estimated amount of accruals based earnings management among the different firms, as a regression within a three digit industry code by year. 𝑇𝐴𝐶𝐶𝑗,𝑡 𝐴𝑗,𝑡−1

= λ1

1 Aj,t−1

+ λ2

△Sj,t−△ARj,t Aj,t−1

λ3

PPEj,t Aj,t−1

+ ℇ𝑗, 𝑡

(Equation 2) Where:

TACCj,t = Total Accruals at the beginning of the year Aj,t-1 = Total assets at the beginning of the year

△Sj,t = Change in net sales (sales current year – sales prior year) △ARj,t = Change in accounts receivable (AR current year – AR prior year) PPEj,t = Property, Plant and Equipment

ℇ = Error term

After this I calculate real time earnings management. In line with the research by Roychowdhury (2006) I separate real time earnings management(RTM) into three categories; sales manipulation, reduction of discretionary expenditures and overproduction. The first method of RTM is calculated through the cash flow from operations. Abnormal cash flows are calculated and identified as a potential effect of sales manipulation. Equation 3 shows the method used to calculate the abnormal operational cash flows, and is used as an estimation of the real time earnings management activities for each firm year observation.

𝐶𝐹𝑂𝑗,𝑡 𝐴𝑗,𝑡−1

= ∝ 0

1 Aj,t−1

+ ∝ 1

Sj,t Aj,t−1

∝ 2

△Sj,t Aj,t−1

+ ℇ𝑗, 𝑡

(Equation 3) Where:

CFOj,t = Cash flow from operations Sj,t = Net sales

All other variables are as defined previously.

Reduction of discretionary expenditures is measured by calculating the abnormal discretionary expenses and comparing this to the normal discretionary expenses. The total discretionary expenses are the sum of all advertising expenses, R&D expense and selling, general and administrative expenses. The normal discretionary expenses are calculated as a regression within three-digit industry by year, using equation 4.

𝐷𝐼𝑆𝐸𝑋𝑃𝑗,𝑡 𝐴𝑗,𝑡−1

= ∝ 0

1 Aj,t−1

+ ∝ 1

Sj,t Aj,t−1

∝ 2

△Sj,t Aj,t−1

+ ℇ𝑗, 𝑡

(Equation 4) Where:

DISEXP j,t = Discretionary expenses All other variables are as defined previously.

(20)

Overproduction is measured by calculating the abnormal production, minus the normal production within three-digit industry by year, using equation 5. Managers can produce more inventory in order to reduce the costs of goods sold for that year.

𝐶𝑂𝐺𝑆𝑗,𝑡+△inv 𝐴𝑗,𝑡−1

= ∝ 0

1 Aj,t−1

+ ∝ 1

Sj,t Aj,t−1

∝ 2

△Sj,t Aj,t−1

+∝ 3

△Sj,t−1 Aj,t−1

ℇ𝑗, 𝑡

(Equation 5) Where:

COGSj,t = Costs of Goods Sold

△inv = Change in inventory (Inventory current year – Inventory at the beginning of the year) All other variables are as defined previously.

Total RTM is calculated as the sum of the three types of RTM previously defined. Prior to summing the abnormal CFO and the abnormal discretionary expenses are multiplied by -1, as lower amounts of cash flows and discretionary expenses would indicate a higher degree of earnings management. Higher abnormal productions are an indicator of a higher degree of earnings management. Therefore, abnormal productions are not multiplied by -1.

4 Empirical Results

This chapter will describe all the empirical results from this study, starting with descriptive statistics. This is followed up with the empirical results from the regressions. All regressions are tested for multicollineairity using the variance inflation factor. No issues have been detected as all untabulated VIF’s are lower than three, except for the interactions. Affected regressions have been rerun prior to adding these interaction effects, noting no differences and reducing the VIF below three.

For the regressions I report the coefficient with the t-statistic below it. The P-value is expressed in asterisks, where one, two and three asterisks represent a significance level of respectively 90, 95 and 99 percent. The variables in the regressions and a list of all control variables used in the following chapter are mentioned and explained in Appendix B.

4.1 Descriptive statistics

Board Governance

Table 4.1 documents the distribution of internal corporate governance strength in B-score by the fiscal years. These results are also charted in graph 4.1 to provide a visual image of the increases that have occurred in since 1996. The distribution has changed significantly over the years towards an increased strength in corporate governance. This increase is in line with the research by Baber et al. (2012), who noted a significant increase between 1997 and 2005. This increase is partially caused by the implementation of SOX and public debate. SOX act 301

(21)

requires all audit committee members to be independent from the firm, other than through board membership (U.S Securities and Exchange Commission, 2002). Furthermore, the US Corporate Governance Principles (1998) state that certain committees should consist entirely of independent directors, including the audit, nominating and the compensation committee. Table 4.2 documents the percentages of firms that have their respective committees fully independent in 1998 and 2013.

Table 4.1: Distribution of B-score by Fiscal Year Fiscal Year B-score 0 1 2 3 4 5 6 Total Average 1997 94 235 188 26 0 0 0 543 1.27 1998 57 130 149 134 126 58 3 657 2.50 1999 60 155 180 157 120 77 5 754 2.49 2000 57 133 161 194 154 80 3 782 2.65 2001 60 138 164 209 203 94 5 873 2.75 2002 40 106 147 181 201 98 4 777 2.91 2003 40 71 117 155 279 147 18 827 3.30 2004 25 52 95 158 292 211 30 863 3.61 2005 17 45 88 140 356 215 36 897 3.74 2006 10 49 86 133 331 239 43 891 3.81 2007 0 7 31 68 285 364 111 866 4.50 2008 0 9 30 73 381 352 89 934 4.40 2009 6 14 32 82 412 353 95 994 4.33 2010 1 6 36 71 413 391 96 1014 4.41 2011 0 5 25 65 350 404 150 999 4.57 2012 3 1 23 63 347 463 159 1059 4.62 2013 3 1 19 59 348 463 161 1054 4.64 Total 473 1157 1571 1968 4598 4009 1008 14784 3.70

likelihood-ratio chi2(96) = 6.6e+03 Pr = 0.000

Pearson chi2(96) = 6.4e+03 Pr = 0.000

Graph 4.1: Distribution of B-score by fiscal year

1995 2000 2005 2010 2015

Fiscal Year

B1 - B6

(22)

Table 4.2: Independence of committees

Overvaluation and earnings management

As stated in the previous chapter, overvaluation is measured as a combination of three different methods. A firm is overvalued if it scored in the top quantile of all methods based on market-to-book ratios and values in Tobin’s Q. Firms that are identified as overvalued in this study have different characteristics than firms that are not overvalued. Some of these differences have been summarized into means in table 4.3. Overvaluation is indicated as a 1 if a firm is overvalued and 0 if it is not.

Table 4.3 shows the difference in Market-to-Book ratio and Tobin’s Q between firms which appear overvalued the other firms. It also shows differences in firm characteristics and earnings management between overvalued firms and the rest. Contrary to the findings of Badertscher (2011), I found that firms which are overvalued have higher amounts of cash flows from operations, higher discretionary expenses and reduced earnings management or reversal of previous methods of earnings management. It should be noted that the methods of detecting overvaluation is different from that of Badertscher (2011), who uses analyst earnings forecast and market values as an indicator of overvaluation. Both this research and the research by Badertscher (2011) used the method of Roychowdhury (2006) for calculating real time earnings management and the modified jones model for calculating accruals based earnings management. (Kothari, 2005)

This study merely looked at market value compared to the intrinsic value of the firm for the determination of overvaluation. It is possible that firms which have higher market-to-book ratios and Tobin’s Q have higher market value because of their investment in intellectual capital, or their ability to produce more positive cash flows. Considering the methods used to calculate earnings management, this would make them less likely to use earnings management to increase periodical earnings. Furthermore, Badertscher (2011) concluded that earnings management, particularly real time earnings management, increases as the duration of overvaluation increases. This study does not look at the duration of overvaluation.

Year Audit_Ind Com_Ind Nom_Ind

1998 50,23% 65,30% 27,40%

(23)

Table 4.3: characteristics of overvalued firm year observations

4.2 Empirical results

Board Governance

Table 4.4 shows the regressions for the strength of internal corporate governance and the effects this has on different aspects of the firm. The table shows a significant negative correlation with both overvaluation and undervaluation. This indicates that firms with stronger corporate governance are less likely to be either overvalued or undervalued, thus resulting in a more balanced firm value. The regression also shows that faster growing firms have significantly reduced B-scores and also have significantly lower increases in their B-scores. An untabulated regression on the interaction between SOX and overvaluation shows that this interaction is also positively significant up to a 95% with the B-score of the current year. This indicates that overvalued firms were more likely to increase the strength of their internal corporate governance after the implementation of SOX. Consistent with prior research the regression indicates a negative correlation between earnings management and B-score (Xie, Davidson, & DaDalt, 2003) (Dechow, Sloan, & Sweeney, 1996). This indicates that firms with a stronger internal corporate governance are less likely to conduct earnings management.

0 1 Total diff. P-value t

Market Value MTB 2.808 9.922 3.577 7.113 *** 0.0001 5.5504 tobin 1.722 4.421 2.014 2.699 *** 0.0001 85.8871 Firm characteristics Massets 8.225 6.191 8.005 -2.034 *** 0.0007 -3.3723 growth 0.102 0.183 0.111 0.081 *** 0.0001 9.0899 Msale 6.943 6.836 6.932 -0.107 0.8505 -0.1885 leverage 0.521 0.448 0.513 -0.073 *** 0.0001 -12.465 CFO 0.110 0.207 0.121 0.097 *** 0.0001 45.6553 disexp 0.273 0.446 0.292 0.172 *** 0.0001 27.4783 prod 0.774 0.816 0.778 0.042 ** 0.0202 2.322 Earnings Management AM 0.001 -0.007 0.000 -0.008 *** 0.0001 -5.7809 RTM 0.031 -0.254 0.000 -0.285 *** 0.0001 -33.746 EM 0.033 -0.270 0.000 -0.303 *** 0.0001 -34.581 Observations 13185 1599 14784 *** p<0.01, ** p<0.05, * p<0.1 Overvaluation

(24)

Table 4.4: Regressions of Board Governance statistics

Overvaluation

Table 4.5 shows the regression for the overvaluation dummies and the effects this has on the different aspects of the firm. The table shows the results for the regression on the different methods of determining overvaluation. Even though ‘overvalued’ is the main determinant for overvaluation in this study, the other types are added to the regression to test for differences in different methods of determining overvaluation. Overvaluation is significantly less likely to occur when a firms B-score increases. This supports the first hypothesis: “Firms with weaker corporate governance are more likely to be overvalued”. Even though this effect is negative and significant, it still does not completely disprove Jensen (2005) who claimed that corporate governance is ineffective at preventing overvaluation. Even though there is a statistically significant correlation between a firm’s B-score and its chance of being overvalued, the coefficient is fairly low which indicates that the effect might not be strong enough to efficiently

Bscore lag_score scorechange

overvalued -0.156*** -0.167*** 0.0192 (-4.425) (-4.272) (0.621) undervalued -0.1000*** -0.117*** -0.00575 (-3.317) (-3.485) (-0.217) SOX 1.713*** 1.636*** -0.147*** (74.74) (61.18) (-6.963) AM -0.0923 0.0346 -0.195 (-0.456) (0.149) (-1.064) RTM -0.0771** -0.0959** 0.0248 (-2.220) (-2.449) (0.803) big4 0.0662 0.0354 0.0429 (1.279) (0.609) (0.933) Massets 0.00146* 0.00147* 3.63e-05 (1.888) (1.800) (0.0563) Msale 0.00130 0.00138 -0.000402 (1.587) (1.609) (-0.592) growth -0.166*** -0.123*** -0.132*** (-5.367) (-3.123) (-4.244) leverage 0.479*** 0.538*** -0.0172 (9.958) (9.955) (-0.404) ROE 0.00194 0.00122 0.000827 (1.278) (0.768) (0.661) LROE 0.00106 0.00220 -0.00115 (0.693) (1.386) (-0.913) Observations 14,784 12,814 12,814 R-squared 0.286 0.236 0.005 t-statistics in parentheses *** p<0.01, ** p<0.05, * p<0.1

(25)

prevent overvaluation and the related destructive effects. Another interesting finding from the regression is the negatively significant result in earnings management for overvalued companies. This is consistent with the findings from the descriptive statistics.

An untabulated result was the calculation of the effect in the interaction between B-score and SOX, thus attempting to prove that the strong boards where more capable of preventing overvaluation after the passage of SOX. This effect was not significant for any of the methods used. Therefor no evidence is found to support the third hypothesis: “Implementation of SOX resulted in Corporate Governance being able to better reduce earnings management”. The implementation did have a positive effect on the B-score of the firms, which in turn reduced overvaluation. However, the interaction effect between the two did not provide significant extra opportunities for strong boards to reduce overvaluation.

Table 4.5: Regression of overvaluation statistics

overvalued over_mtb overx overz

Bscore -0.00825*** -0.00487 -0.00249 -0.00858** (-2.689) (-1.256) (-0.650) (-2.212) SOX 0.0233* 0.00738 0.0113 -0.00453 (1.672) (0.418) (0.648) (-0.257) AM 0.0516 0.175*** 0.132** 0.105* (1.086) (2.908) (2.217) (1.752) RTM -0.259*** -0.331*** -0.378*** -0.333*** (-33.31) (-33.67) (-38.88) (-33.93) big4 0.0122 0.00842 -0.0346** -0.0173 (1.005) (0.548) (-2.280) (-1.128) Massets -0.00112*** -0.00129*** -0.00140*** -0.00103*** (-6.181) (-5.605) (-6.176) (-4.486) Msale 0.00128*** 0.00182*** 0.00163*** 0.00100*** (6.655) (7.486) (6.797) (4.125) growth 0.0523*** 0.0555*** 0.0545*** 0.0920*** (7.207) (6.056) (6.018) (10.05) leverage -0.105*** 0.211*** -0.108*** -0.216*** (-9.239) (14.73) (-7.612) (-15.13) ROE 0.000398 -0.000596 0.000416 0.000325 (1.116) (-1.321) (0.932) (0.721) LROE 0.000277 -1.28e-05 0.000275 0.000302 (0.767) (-0.0281) (0.610) (0.663) Bscore * SOX 0.000481 0.00216 -0.00188 0.00384 (0.122) (0.435) (-0.383) (0.774) Observations 14,784 14,784 14,784 14,784 R-squared 0.089 0.087 0.108 0.104 t-statistics in parentheses *** p<0.01, ** p<0.05, * p<0.1

(26)

Earnings Management

The regressions for Earnings Management have been included in table 4.6. The regressions show that overvalued firms are significantly more likely to reduce or reverse their earnings management, while undervalued firms are using this to manage their earnings upwards. The interaction between overvaluation and B-score is positive and not significant. This means that no support is found for the second hypothesis: “Overvalued firms have reduced earnings management if corporate governance is stronger”. The expectation was that overvalued firms would have positive earnings management, and the interaction with internal corporate governance would reduce this earnings management. However, firms which are overvalued have negative earnings management and the interaction has no significant effects on earnings management.

Table 4.6: Regressions of Earnings Management statistics

EM RTM AM overvalued -0.302*** -0.285*** -0.00414 (-13.25) (-13.67) (-1.156) undervalued 0.133*** 0.133*** -0.00282** (16.92) (18.46) (-2.278) Bscore -0.00397** -0.00361** -0.000202 (-2.017) (-1.999) (-0.653) big4 -0.0521*** -0.0419*** -0.00280 (-3.811) (-3.343) (-1.303) Massets -0.00104*** -0.00112*** 8.27e-05** (-5.087) (-5.950) (2.567) Msale 0.00160*** 0.00159*** -2.12e-05 (7.369) (8.004) (-0.622) growth -0.0212*** -0.0192** -0.00455*** (-2.584) (-2.561) (-3.530) leverage 0.0603*** 0.0681*** -0.0120*** (4.750) (5.849) (-6.021) ROE -0.000305 -0.000295 -4.18e-06 (-0.759) (-0.799) (-0.0661)

LROE 1.23e-06 -5.09e-05 5.82e-05

(0.00304) (-0.137) (0.912) overvalued * Bscore 0.00695 0.00742 -0.00133 (1.201) (1.398) (-1.459) Observations 14,784 14,784 14,784 R-squared 0.096 0.104 0.007 t-statistics in parentheses *** p<0.01, ** p<0.05, * p<0.1

(27)

5 Conclusion

5.1 Core findings

This paper has found evidence that board governance is related to the valuation of a firm. Firms with stronger board governance are less likely to be overvalued. This means that stronger board governance prevents firms from being overvalued due to them being better able to execute their monitoring function, thus passively preventing overvaluation. Another cause for this finding could be that boards are observing stock prices and are taking actions when these prices rise too high, thus actively preventing overvaluation. This study also concludes that board index is negatively associated with undervaluation, which shows that firms with stronger board governance generally have a more balanced firm value.

The findings when investigating earnings management are surprising. I have noted negative earnings management for firms which are overvalued, which is inconsistent with the findings by Badertscher (2011) and the theory by Jensen (2004). This difference in outcomes could be explained by a difference in calculating overvaluation. Badertscher (2011) used the residual income approach by Frankel and Lee (1998) while I used the market-to-book ratio as was done by Rhodes-Kropf et al. (2005). Badertscher also noted that earnings management increased as the duration of overvaluation continued, while this study does not take the duration of overvaluation into account. The association between the board index and earnings management was negatively significant up to the 95% interval, which is consistent with prior research (Dechow, Sloan, & Sweeney, 1996) (Xie, Davidson, & DaDalt, 2003).

I have noted a large increase in board index, particularly between 2002 and 2007. The passage of SOX caused firm to revise the composition of their board of directors and the separate committees, which gave corporate governance more power to monitor the management according to Baber et al. (2012). There is no significant results for the interaction between SOX and overvaluation, which indicates that SOX did not give the board of directors more influence or incentives in preventing overvaluation, either directly through communication with the market or indirectly through its monitoring function.

5.2 Implications

My findings that overvaluation is less likely to occur if the firm has a stronger board governance has implications for researchers. This shows that board governance is capable of preventing overvaluation and incentivized to do so. Researches can use these findings to identify the catalysts for overvaluation and the mechanisms that effectively prevent this from occurring.

(28)

5.3 Future research

Future research would be required in order to create a more fitting model for calculating overvaluation. Currently the two leading models are the residual income approach by Frankel and Lee (1998) and the market-to-book ratio by Rhodes-Kropf et al. (2005). The first has a severe weakness due to the bias in analyst forecasts and the availability of information. Some firms might have less or different analysts following the firm causing differences in opinions, thus reducing the validity of these variables. Furthermore, if a manager is able to convince the analysts following the firm that the future earnings are higher than what the firm can deliver, then the firm would not be identified as overvalued by the residual income approach even though it might be severely overvalued.

The method by Rhodes-Kropf (2005) merely looks at market-to-book ratios compared among industries. This method is trying to identify misevaluation by comparing the market-to-book ratio of a firm to its industry. This method does not accurately take into account the assets that a firm has which are not reflected on the balance sheet, such as reputation, know-how, customer lists and brands. A study that compares different methods with actual stock returns could provide an answer which model would be best for determining overvaluation. If overvalued companies could be more accurately identified then we could more accurately determine the effects that are caused by overvaluation for firms, investors and managers.

5.4 Limitations

The major limitations from this study are from the operationalisation of the empirical models. The measurement and calculation of proxies for overvaluation, earnings management and corporate governance are difficult matters. Much prior research has been dedicated to detecting and measuring earnings management, however, due to management effort into hiding this and the difficulty of separating normal variations in accruals or operations from managed accruals or operations these models will only have moderate explaining power. The systems of corporate governance is so complex and its power and incentives to monitor the executives depend on many different factors, some which cannot be measured empirically. An example of this would be the relationship between the chairman and the CEO. The model used in this study is simplistic and will only provide an indication for the actual strength of corporate governance.

The identification of overvaluation is the biggest limitation of this research. Despite the combination of several methods, I cannot guarantee that the firms in this sample are in fact overvalued according to my previously described definition. I merely compared certain balance sheet items with its market value. The change in market value could be caused by firms in my sample having a heightened amount of assets not reflected in the balance sheet.

(29)

References

Ali Shah, S. Z., Butt, S. A., & Tariq, Y. B. (2011). Use or abuse of creative accounting techniques.

International Journal of Trade, Economics and Finance, 2(6), 1-6.

Arya, A., Glover, J. C., & Sunder, S. (2003). Are unmanaged earnings always better for shareholders? Accounting Horizons, 17, 111-116.

Baber, W. R. (2012). Associations between internal and external corporate governance characteristics: Implications for investigating financial accounting restatements.

Accounting Horizons, 26(2), 219-237.

Baber, W. R., Liang, L., & Zhu, Z. (2012). Associations between internal and external corporate governance characteristics: Implications for investigating financial accounting restatements. Accounting Horizons, 26(2), 219-237.

Badertscher, B. A. (2011). Overvaluation and the choice of alternative earnings management mechanisms. The Accounting Review, 86(5), 1491-1518.

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

Organization, 101-124.

Beck, M. J., & Mauldin, E. G. (2014). Who's really in charge? Audit committee versus CFO power and audit fees. The Accounting Review, 89(6), 2057-2085.

Blanchard, O. J. (1979). Speculative bubbles, crashes and rational expectations. Economics

letters, 3(4), 387-389.

Blanchard, O. J., & Watson, M. W. (1982). Bubbles, rational expectations and financial markets. Chi, J. D., & Gupta, M. (2009). Overvaluation and earnings management. Journal of Banking &

Finance, 33(9), 1652-1663.

Cohen, D. A., Dey, A., & Lys, T. Z. (2008). Real and accrual-based earnings management in the pre-and post-Sarbanes-Oxley periods. The accounting review, 83(3), 757-787.

Cole, S. (2012). The Impact of Reputation on Market Value. World Economics-Abingdon, 13(3), 47-68.

Damodaran, A. (2009). Valuing financial service firms. Stern Business.

DeAngelo, L. E. (1986). Accounting numbers as market valuation substitutes: A study of management buyouts of public stockholders. Accounting Review, 400-420.

Dechow, P. M., & Sloan, R. G. (1991). Executive incentives and the horizon problem: An empirical investigation. Journal of accounting and Economics, 14(1), 51-89.

Dechow, P. M., Richardson, S. A., & Tuna, A. (2000). Are benchmark beaters doing anything wrong? Are benchmark beaters doing anything wrong?

Dechow, P. M., Sloan, R. G., & Sweeney, A. P. (1995). Detecting earnings management. Accounting

(30)

Dechow, P. M., Sloan, R. G., & 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.

Diamond, D. W., & Verrecchia, R. E. (1987). Constraints on short-selling and asset price adjustment to private information. Journal of Financial Economics, 18(2), 277-311. Edwards, E. O., & Bell, P. W. (1965). The theory and measurement of business income. Univ of

California Press.

Efendi, J., Srivastava, A., & Swanson, E. P. (2007). Why do corporate managers misstate financial statements? The role of option compensation and other factors. Journal of Financial

Economics, 85(3), 667-708.

Fama, E. F., & Jensen, M. C. (1983). Separation of ownership and control. Journal of law and

economics, 301-325.

FASB. (1985). Elements of Financial Statements Statement of Accounting Concepts No. 6. Norwalk, CT: FASB.

Frankel, R., & Lee, C. M. (1998). Accounting valuation, market expectation, and cross-sectional stock returns. Journal of Accounting and economics, 25(3), 283-319.

Fukui, Y. (2000). Earnings management with the help of historical cost accounting: Not for

managers but for investors. Manuscript in Preparation: Tohoku University.

Gunny, K. A. (2005). What are the consequences of real earnings management?

Healy, P. M. (1985). The effect of bonus schemes on accounting decisions. Journal of accounting

and economics, 7(1), 85-107.

Healy, P. M., & Wahlen, J. M. (1999). A review of the earnings management literature and its implications for standard setting. Accounting horizons, 13(4), 365-383.

Higgs, D. (2003). review of the role and effectiveness of non-executive directors. London: Department of Trade and Industry.

International Monetary Fund. (2003). World Economic Outlook.

Jensen, M. C. (1993). The modern industrial revolution, exit, and the failure of internal control systems. the Journal of Finance, 48(3), 831-880.

Jensen, M. C. (2004). The agency cost of overvalued equity and the current state of corporate finance. European Financial Management, 10(4), 549-565.

Jensen, M. C. (2005). Agency costs of overvalued equity. Financial management, 5-19.

Jensen, M. C., & Meckling, a. W. (1976). Theory of the firm: Managerial behavior, agency costs and ownership structure. Journal of financial economics, 3(4), 305-360.

Jones, J. J. (1991). Earnings management during import relief investigations. Journal of

The impact of corporate governance on overvaluation and its related effects