The effects of corporate ownership on earnings’ quality：evidence from the United States

(1)

The Effects of Corporate Ownership on Earnings’ Quality：

Evidence from the United States

NAME: Yumo Yang

STUDENT NUMBER: 10605541

DATE: June 22, 2014

MSc Accountancy & Control – Track Control

Amsterdam Business School

Faculty of Economics and Business, University of Amsterdam

(2)

The Effects of Corporate Ownership on Earnings’ Quality:

Evidence from the United States

Abstract

This study provides evidence on the relationship between the ownership structure and the earnings’ quality. In this paper, the Modified Jones Model is used to measure earnings quality, together with multiple techniques as additional test. The results indicate that the institutional ownership is positively associated with the quality of earnings. In addition, it is found that a negative relationship might exist between ownership concentration and earnings quality. However, this negative relationship may be weakened by the contradictory effects of ownership concentration on earnings quality.

Key word: Earnings’ quality, Institutional ownership, Ownership concentration

1. Introduction

1.1 Background

In this study, we investigate on how corporate ownership affects the quality of earnings. When preparing the financial statements, the General accepted accounting principles (GAAP) gives some flexibility and freedom for the management to choose what accounting alternatives or policies to be applied. Using this flexibility and freedom, earnings management is the action involves in intentional influencing the process of financial reporting to achieve a predetermined target set by the management (Schipper 1989). Earnings management is conducted by the managers when they use their own judgments in financial reporting process. Financial reports

(5)

2

are altered to achieve a misleading underlying economic performance of the company. The contractual outcomes can also be influenced because these outcomes usually depend on the number of financial reporting. (Healy and Whalen 1999). Although earnings management may not necessarily mean violating the accounting rules, it is still ethically questionable. Because of the earnings management, the financial report is not presenting the truest and fairest view of the underlying economical performance (Abdelghany 2005).

As defined by Schipper and Vincent (2003), the earnings’ quality is the extent to which the reported earnings faithfully represent Hichsian economic income, which is the amount that can be consumed in forms of dividend paid out. Higher earnings’ quality provides with more information about the firm’s fundamental economical performance and gives more value for decision makers (Dechow et al. 2010). The misleading effect of earnings management mentioned above reduces the quality of earnings of the companies.

Prior literatures, like Leech (1987) and Frank and Mayer (1997), focus on corporate ownerships only and elaborate how corporate ownership impacts the organization’s control system. Based on companies in Germany, Brown et al. (2014) detect an increase in earnings’ quality after the implementation of 1998 German legislation on control and transparency (KTG), which aims at enhancing the internal control for the German firms. It is concluded in this study that effective internal control system is associated with increased earnings’ quality. Following the logic, the conclusion should be that corporate ownership may affect earnings’ quality through the organization’s control system.

In addition, Fan and Wong (2002) examine the relationship between earnings informativeness, measured by the earnings–return relation, and the ownership structure. Here, the concept of earnings informativeness is used to demonstrate the level of the quality of accounting information to investors. They argue that controlling owners are perceived to report accounting information for self-interested purposes, in this situation, the reported earnings might lose credibility to outside investors. Besides, concentrated ownership is associated with low earnings informativeness as ownership

(6)

3

concentration prevents leakage. Dempsy et al. (1993) examines how different ownership structure for U.S. firms would affect the level of earnings management through the use of extraordinary item reporting. With firms of different ownership structures show significant different level of earnings management, it provides evidence that earnings management is related to corporate ownership structure. Velury and Jenkins (2006), which focus on the effect of institutional ownership only, demonstrate significant evidence of a positive association between institutional ownership and earnings’ quality. In addition, the results indicate that concentrated institutional ownership may negatively affect earnings quality. Jiang and Anandarajan (2009) examine the effect of shareholder rights on the quality of the reported earnings. The influence of institutional investors on shareholder rights is incorporated in the research and their joint impact on earnings’ quality is studied. They find that stronger shareholder rights are positively related with higher earnings’ quality. However, when the shares of the firm are held predominantly by transient institutions, which refer to institutions with short investment horizons, the role of shareholder rights in constraining aggressive and opportunistic management of earnings is significantly diminished or turned essentially ineffective. P. Katz (2009) also provides evidence that ownership structure would affect firms’ long term performance as well as the quality of earnings. When study the effect of ownership structure on earnings’ quality, the paper mainly focuses on the role of private equities and indicates a positive influence of PE-backed ownership structure on the quality of earnings for the company. Ghabdian et al. (2012) compare the earnings management between family owned and non-family owned companies based on samples of Iran. They identify that the non-family structured companies engage more earnings management than family structured companies.

Many researches focus on corporate ownership and company performance but ignore the quality of such reported performance. Kirchmaier and Grant (2005) conduct research on corporate ownership structure and firm performance in Europe. It shows that ownership structures vary considerably across the largest European economies, and that the performance of the firm is significantly influenced by

(7)

4

ownership structure. This study found that dominant shareholders have a negative impact on long-term stock price performance of the company, while the similar studies based on US companies have contradictory results. However, as Dechow et al. (2010) point out, the “quality” of earnings is also a function of the firm’s fundamental performance. The ignorance of earnings’ quality and focusing only on the “reported” performance should be a concern. Kapopoulos and Lazaretou (2007) conduct research on the corporate ownership structure and firm performance in Greece and suggested that a more concentrated ownership structure positively relates to better firm performance. They also find that higher firm profitability is usually related with a diffused ownership. Similarly, the issue of earnings’ quality is not included in the study. As argued by Dechow et al. (2010), the firm’s fundamental performance might have an impact on earnings’ quality. When corporate ownership affects firm’s performance, the quality of earnings should also be taken into consideration.

These literatures provide with good basis for my research. By answering the research question of this study, contribution might be made to these literatures and it may help to give a more complete and holistic knowledge to the field. The contribution of my research to these prior literatures and the motivation for the research will be elaborated more specifically in the following part.

1.2 Research question

The research question of this paper is how the corporate ownership (type of shareholders and ownership concentration) would affect the earnings’ quality: evidence from the United States. Here, two types of shareholders will be considered, that is, the institutional shareholders and individual shareholders.

1.3 Motivation of the study

As mentioned above, in previous literatures, relationship between corporate ownership and the reported company performance is intensively studied (Kirchmaier & Grant, 2005; Kapopoulos & Lazaretou, 2007 etc.). However, few researches have

(8)

5

looked into whether the quality of such reported performance is influenced by the corporate ownership. By answering the research question of this research, a more comprehensive knowledge might be built on the effects of corporate ownership, considering both the “quantity” and “quality” of company performance. When corporate ownership affects company performance, whether the quality of such reported performance is also affected? As argued by Dechow et al. (2010), the including of the quality of earnings provides a more complete view of firms’ fundamental performance. As mentioned above, some papers have provided insights into the effects of corporate ownership on the earnings informativeness (Fan & Wong 2002) and earnings management (Dempsy et al. 1993). It is noted that the higher level of earnings management is associated with lower level of earnings’ quality, and earnings informativeness can be used as one of the measures of earnings’ quality (Bryan et al. 2013). And Dempsy et al (1993) focus more on the impacts of owner managers and non-owner managers on earnings management, while Velury and Jenkins (2006) only focus on the effect of institutional ownership on earnings quality, none of them provide a broad view of the overall corporate ownership. More importantly, none of the studies mentioned above has explicitly looked into earnings’ quality. Studying the relationship between earnings’ quality and corporate ownership directly might help to provide with a more comprehensive knowledge on this issue.

Corporate ownership has great impacts on organizations; it is not enough to study only one side of the effects of corporate ownership. Studying the relationship between corporate ownership and earnings’ quality is helpful to increase the completeness and coherency of this issue. Previous papers have provided evidence of the effects of ownership structure and the quality of earnings mainly from one side. As mentioned above, Ghabdian et al. (2012) only identify the relation between earnings management and family and non-family ownership. Velury and Jenkins (2006) research focuses more on the influence of institutional ownership on earnings’ quality while P. Katz (2009) focuses on the role of private equity for corporate ownership. In my study, a more complete and comprehensive view of corporate ownership is given, considering both the type and concentration of the shareholders,

(9)

6

not only the effects of institutional ownership. In addition, in some studies, evidence can be found on the effect of shareholder concentration on earnings’ quality (Fan and Wong 2002; Ali et al. 2007), however, mixed results are yielded. Thus, conducting research on this issue using different samples or method is still needed. Jiang and Anandarajan (2009) focus on shareholder rights and earnings’ quality, it is noted in this study that shareholder rights are influenced by corporate ownership. By affecting shareholder rights, earnings’ quality is influenced, this study provides insight on one side of the effects of corporate ownership on earnings’ quality, a more in-depth study directly on the corporate ownership and the quality of earnings is needed in order to give a more holistic view.

Furthermore, by answering my research question, a broader impact of the overall corporate ownership on the earnings’ quality will be examined and this is helpful to provide insight on whether the quality of earnings could be improved if the corporate ownership changes in a certain way. As mentioned by Schipper and Vincent (2003), the quality of earnings is of great interest to those who use the financial report for decision makings as well as for contracting purposes. Low quality of earnings would have impact on variety of stakeholders. Given the importance of earnings’ quality, as well as the concerns on the relevance and reliability of accounting information, this study may be helpful for economic reformers and regulators who are trying to improve the earnings’ quality and transparency of accounting information in the United States. In addition, as noted by Schipper and Vincent (2003), earnings’ quality is also measure for accounting standard setters to receive feedback on whether the accounting standards they set are functioning effectively. Since the quality of earnings is also of great interest for the accounting standard setters, the impact of corporate ownership on earnings’ quality might also have some implications for them.

2. Literature review and hypothesis

(10)

7

quality is defined as the overall reasonableness of reported earnings (Knechel et al. 2000). It is the degree to which earnings reflect underlying economic effects, the better estimate of cash flows, are conservative, or are predictable. (Financial Education. n.d. What is Earnings Quality? Retrieved June 29, 2008) In order to give a more comprehensive view when studying the effect of corporate ownership on the earnings’ quality, two perspectives of corporate ownership are considered, the type of shareholders and shareholder concentration.

2.1 The type of shareholders and earnings’ quality

In this study, shareholders are categorized into two types, the institutional shareholders and individual shareholders. The separation of ownership and operating rights for modern corporations increases agency costs. These costs can become very high when the firms are owned by a large amount of diffused outside shareholdings, because it becomes difficult for the shareholders to have incentive or power to monitor the management of the firm. (Koh 2003)

The presence of institutional shareholders, which have larger shareholdings, helps to bear this agency cost. In addition, the institutional shareholders have the economies of scale for gathering information (Koh 2003). Monks and Minow (1995) point out that with more stakes holding, institutional investors are more likely to engage in the monitoring activities because the benefits from monitoring can cover the related monitoring cost. However, individual shareholders, which usually have the smaller stake, have less incentive to undertake the monitoring activities because the related cost is too high compare to the increased returns from the monitoring activity.

Johnson and Greening (1999) find that larger shareholders have more incentive to monitor their sizeable investment because of the magnitude of their equity stakes. In addition, the related loss of market exit is more likely to hold onto the shares of larger shareholders. Koh (2003) also documents that when the size of investment is large, the market exit becomes unattractive for institutional shareholders. In this situation, they are more likely to engage actively in the monitoring process of their

(11)

8

portfolio firms. Consistence with this, Ramsay et al. (2000) point out that the institutional shareholders in Australia communicate frequently with the senior management of the firm they invest. They find that because of the large amount of money they invested into the companies, they engage more in the proactive monitoring process. The institutional activism is growing in Australia both in awareness and practice. Chung et al. (2002) find that the presence of institutional ownership helps to prevent the managers from manipulating the profit to their own desired level. It is documented that institutional shareholders have more concerns on the long term profitability of the company and are more vigilant about the use of discretionary accruals. Institutional shareholders focusing on firm’s long-term performance is also evidenced by Bushee (1998), which find out that the institutional ownership is positively associated with the develop and research investment of the company. Arguably, because of the incentives of large institutional shareholders to monitor the earnings management, the quality of earnings might increase when management reduce their opportunistic earnings management behavior. This effect turns out to be more significant when more shares are held by institutional investors. It is concluded that institutional shareholders have more incentives to monitor the quality of earnings compare to individual shareholders.

Besides, previous literatures provide evidence that institutional shareholders have more power to monitor the earnings’ quality and have the ability to analyze the financial statements more thoroughly than individual shareholders. Edmans (2009) point out that larger owners are more able to detect accounting manipulation and to prevent it from happening. In addition, shareholders activism becomes more effective when there is a higher level of institutional ownership. On one hand, monitoring activities are conducted explicitly in the form of directly engaging in the corporate governance practices, such as submitting a shareholder proposal and directly negotiating with the management etc. On the other hand, it can also be conducted in an implicit way through information gathering and correctly pricing the impact of managerial decisions (Bushee, 1998; Gillan and Starks, 1998).

(12)

9

conduct collective actions as share ownership is controlled by a smaller number of shareholders. It becomes easier to form relatively collective group which facilitate the monitoring process.Rajgopal and Venkatachalam (1998) find out that the presence of institutional shareholders is helpful to reduce the income increasing discretionary accruals. Bushee (1998) also provides evidence that because of the investment size of institutional shareholders, they can serve better to reduce the earnings management as a concentrated group. In addition, Koh (2003) also point out that when institutional shareholders present, the monitoring costs for the homogeneous shareholder group can be shared. This further helps to improve the effectiveness of the monitoring process. Chung et al. (2002) argue that the institutional shareholders are more able to monitor and constrain the self-serving behaviors of the corporate managers. When institutional managers effectively monitor the corporate managers, their ability for opportunism earnings management will be reduced. However, the ability for individual shareholders to monitor the corporate managers effective is limited. (Monks and Minow 1995, Chung et al. 2002). These stronger monitoring abilities arguably increase the earnings’ quality. In comparison, the monitoring ability of individual shareholders is much weaker.

In summary, it appears that institutional owners have more incentives to monitor earnings and also have more monitoring abilities, whereas individual owners have fewer incentive and also fewer abilities. These incentives and power of institutional owners to monitor earnings might result in higher earnings quality. Meanwhile, the lack of incentive and monitoring ability for individual ownership might lead to lower quality of earnings. Therefore, the first hypothesis is given as follows:

H1: The higher level of institutional ownership is associated with higher earnings’

quality.

2.2 The ownership concentration and earnings’ quality

In this study, the ownership concentration is defined as the quantitative indicator that demonstrates to what extent the shareholder ownership is centralized or dispersed,

(13)

10

due to the different shareholding ratio of each shareholder. (Ding and Zhang 2004) The degree of ownership concentration affects the nature of contracting, creating agency problems between managers and outside shareholders. When the level of ownership concentration is low, typically like the companies in the U.S. and the U.K. Agency problems may arise because the managers only own an insignificant amount of shares and may act in their own interest instead of acting in the best interests of the shareholders. On the other hand, when the ownership structure is highly concentrated, as many companies in East Asia, agency problem reduces, however, the conflicts between controlling shareholders and minority shareholders arise. In terms of the different impacts of ownership concentration on earnings’ quality, two competing theories are used in this study to explain them more clearly – The entrenchment theory and the alignment theory.

2.2.1 The entrenchment theory

Entrenchment theory predicts that controlling shareholders have incentives to engage in opportunistic earnings management. Increasing managerial ownership may entrench managers, as they are increasingly less subject to governance by board of directors. Also because of the large portion of shares owned by these controlling shareholders, they are less subject to discipline by the market for corporate control as well. This is evidenced by Gul et al. (2010), based on companies from China, they find out that the large shareholder concentration is associated with decreasing stock price synchronicity. Fan and Wong (2002) note that the controlling shareholders may have the incentives and power to deprive the rights of minority shareholders in order to pursue personal interests. They can use control-enhancing mechanisms such as dual class shares and stock pyramids to maintain control over the firm with a small fraction of equity ownership to gain private benefit (Bebchuck at al 2000). The entrenchment effects may cause them to reduce the quality of accounting information in order to cover the wealth effects of their expropriation activities.

2.2.2 The alignment theory

(14)

11

amount of shares the controlling shareholders have might give them incentives to protect these shares. Fan and Wong (2002) argue that high ownership concentration can also create alignment effects: when an owner’s share ownership increases beyond the minimum level needed for effective control, the alignment of interests between the controlling owner and the minority shareholders improves; and at the same time, the effects of entrenchment reduces. In case that the entrenched controlling shareholders further increase their portion of ownership or even let the company go privacy, the problem exposed in entrenchment theory may be significantly reduced. Wong (2006) shows that ownership concentration can provide incentives for controlling shareholders to closely monitor the management to produce high-quality earnings. In addition, large shareholders are more likely to focus on the long-run of the firm. Ownership concentration may help to mitigate the myopia problem. The managers of these firms are usually less likely to focus on short-term earnings than managers of widely held firms and are consequently less prone to using earning management as a device to boost short-term earnings, leading to the increase of earnings’ quality. (Villalonga and Amit, 2006; Ali et al. 2007)

In summary, when the controlling shareholders achieve effective control, two contrary incentive effects can be identified. When the ownership concentration is below a certain level when effective control achieved, there might be entrenchment effect that is harmful to earnings’ quality; however, when the level of ownership concentration goes up beyond a certain level, this entrenchment effect may be mitigated to become an alignment effect, which arguably increase the quality of earnings.

Mixed evidence on how ownership concentration affects earnings’ quality is provided by previous literatures. For instance, Fan and Wong (2002) find a negative relationship between ownership concentration and earnings’ quality. In contrast, Ali et al. (2007) note that the concentrate ownership structure of family controlled firms helps them to provide financial information that is more in align with the future cash

(15)

12

flow. In addition, the level of positive discretionary accrual is also lower in family controlled firms than non-family controlled firms.

Although contradictory effects of ownership concentration are demonstrated by the two theories, one of the two effects might have a stronger impact in certain sample firms. For instance, in this study, the samples are chosen from the companies listing on U.S. stock market, where ownership concentration is relatively dispersed, it is less likely that the ownership concentration would achieve a level that is high enough for alignment effect. This assumption is consistence with Shuto and Takada (2010), which point out that the entrenchment effect has no influence only if the shareholders have sufficient large portion of shareholdings. For the firms in U.S., this is less likely to be the situation. For this reason, the entrenchment theory is the one that is more applicable for this study.

In addition to the entrenchment effect applied here, it is noted that higher ownership concentration also allows firms to limit their disclosure of public information; it is easier for companies with concentrated ownership to have a tight control over their information flow (Fan and Wong 2002). This information impact arguably has a negative influence on the transparency of accounting information as well as the quality of earnings. Together with the theories mentioned above, the second hypothesis is given as follows:

H2: The higher level of ownership concentration is associated with lower earnings’

quality.

3. Research methodology

3.1 Overview

The paper will use quantitative research method. The target sample companies are chosen from the listed companies in the United States, using Wharton Research Data Service and Datastream. Regression analysis is used to examine the two hypothesizes. From the available companies in NASDAQ, I excluded the companies that do not

(16)

13

have all the data available and then excluded the companies with extremely value (the firms with the value of the assets closed to 0). This yields a total sample of 484 observations per year.

With regard to corporate ownership, two perspectives will be considered, the first perspective is the type of the shareholders (Institutional shareholders and individual shareholders), which can be measured by the percentage of institutional ownership. And the second one is how concentrate is the ownership structure, which is also measured quantitatively. Then the earnings’ quality will be measured using Modified Jones Model, earnings surprise indicator and modified Penman 2001 approach, all of which will be explained below. Together with other control variables (the company growth, the total accruals and the long-term debt / assets ratio, which are explained below) the relationship between the corporate ownership and the earnings’ quality would be examined. All the data needed are retrieved from Datasteam and Wharton Research Data Services databases.

3.2 Dependent variable: Measurement of earnings’ quality

Quality of earnings is an important theoretical concept; however, no uniform method is identified by researchers to measure it. The variety of methods for measuring the earnings’ quality involve in broad dimension, for instance, earnings persistence, smooth earnings, magnitude of accruals, income increasing accruals and the closeness between earnings and cash flows etc (Dichev et al. 2013). Among the different dimensions to measure the earnings’ quality, the Modified Jones Model is believed to be powerful as well as reliable in detecting earnings management (Hadani et al. 2010). This can be evidenced by intensive researches that using the Modified Jones Model to measure the earnings’ quality. And as suggested by Dichev et al. (2013), discretionary factors are one of the most important determinants of earnings’ quality. Consistence with previous research, I will use Modified Jones Model as a main method to measure the quality of earnings. However, different from some of the researches that only use one model as a measurement, this study also adopts two other methods, the Barton

(17)

14

and Simko (2002) approach and the approach modified on Penman (2001), which are explained below, to measure the quality of earnings in multiple dimensions. I believe this is also a contribution to add credibility of the results and to study earnings’ quality more comprehensively.

Accounting accruals can be categorised into normal accruals and abnormal accruals. The normal accruals refer to the adjustment that aims at reflecting the fundamental performance of the firm, while the abnormal accruals involves in the distortion of accounting rules or using earnings management (Dechow et al. 2010). Then basic assumption here is that normal accruals are the proper using of the flexibility of accounting rules, while the abnormal accruals are distortion of accounting information and reduce the quality of earnings. For the Modified Jones Model, I examine the magnitude of abnormal increasing accruals, we can detect whether the management of the company report their performance neutrally. If there is intention for the companies to attain a predetermined result and to report their performance unfaithfully, the quality of earnings is compromised. Larger abnormal accruals indicate a lower quality of earnings while smaller abnormal accruals are indications for higher earnings’ quality.

To calculate normal accruals, the following cross-sectional model is used to generate coefficient estimates for group of firms in the same calendar year:

where, for sample firm i at time t:

TAi, t total accruals, defined as net income before extraordinary items less operating

cash flows

Ai, t−1 total assets at time t−1

ΔREVi, t change in revenue from year t−1 to year t

PPEi, t gross property plant and equipment

(18)

15

The coefficient estimates generated by the above model are used in the following model (ai, b1i and b2i) to generate expected accruals of each firm for the same calendar

year. Thus, expected accruals for firm i at time t are calculated as:

where: E(TAi,t/Ai,t-1) represents expected total accruals scaled by total assets of year

t−1. ΔRECi,t represents change in accounts receivable from year t-1 to year t.

The difference between actual accruals and expected accruals is attributed to abnormal accruals. The value of total accruals minus expected accruals (ABNAC) is used as a measure of earnings’ quality.

In addition, as mentioned above, simpler measurements will be used as supplementation to examine the earnings’ quality in multiple dimensions. This is consistence with the belief in Abdelghany (2005), which argues that since there is no agreed-upon definition or technique to measure the earnings’ quality, it is inappropriate to conclude that the quality of earnings for a certain company is high or low. Researchers should apply more than one measure for the quality of earning in order to have strong evidence about the level of quality. We can judge the quality of earnings of a company only when consistent results are found using different approach or technique for measurement. According to the approaches discussed in Abdelghany (2005), I will also use Barton and Simko (2002) approach as a triangulation of measuring the earnings’ quality. It assesses earnings’ quality by focusing on the earning persistence; high quality earnings are more persistent and useful in the process of decision making. In this approach, earnings’ quality is measured by the earning surprise indicator, which is the ratio of the beginning balance of net operating assets relative to sales. The smaller ratio indicates the higher quality of earnings.

Another method used is the approach based on Penman (2001), where the quality of earnings is measured by the ratio of cash flow from operation divided by the net income. It is believed that the purpose of accounting quality analysis is to distinguish

(19)

16

between the “hard” numbers resulting from cash flows and the “soft” numbers resulting from accrual accounting. The more the earnings are closed to operating cash flow, the higher is the earnings’ quality. (Penman 2001, Abdelghany 2005). Consistent with this, Dichev et al. (2013) find that the inconsistence with the trends in earnings and the trends in cash flows is regarded as an important “red flag” for detecting earnings management. Thus, in the initial Penman (2001) approach, the smaller ratio indicates the higher earnings’ quality. However, using this ratio has its limit; the assumption here is that the operating cash flow is bigger than the net income, and it also neglects the effect of negative cash flow or negative net income. In this study, the basic idea of Penman (2001) approach is followed; however, a modified ratio is adopted in order to overcome the limit of the original approach. To examine the closeness of the “hard” number of cash flow and the “soft” number of net income, the ratio is changed into an absolute value, which is then scaled by total assets to mitigate the effect of company size. That is, PEN is first calculated by using net income minus operating cash flow, the absolute value of this is then scaled by total assets. The smaller ratio of PEN indicates the higher quality of earnings.

3.3 Independent variables

Two independent (explanatory) variables are used in the regression analysis. 3.3.1 Institutional shareholders

Consistence with Bushee (1998), here the institutional shareholders are defined as large investors such as bank trusts, insurance companies, mutual funds and pension funds that invest on behalf of others. These firms are legally required to report their equity holdings. I use INS as the percentage of common shares owned by these institutional shareholders. The individual shareholders are the shareholders that do not meet the criteria of institutional shareholders above, minus government held. In this study, all the sample firms have no government held, so all the rest of the shareholders are included as individual shareholders

(20)

17

3.3.2 Ownership concentration

Ownership concentration (CON) is measured by the percentage of company's stock held by the major shareholders of the firm. In accordance with the definition of Datastream database, the major owners are defined as any individual or company that owns more than the local legal disclosure requirement of the outstanding shares of a company.

3.4 Regression analysis

To examine the relationship between corporate ownership and earnings’ quality, the following model is used:

ABNAC=

α0+α1INSit+α2CONit

+α3TACCRit+α4GROWTHit+α5DEBTit+εit

Despite the dependent variable and independent variables explained above, the control variables and dummy variable are as follow:

TACCR: total accruals of the company in period t

GROWTH: percent change in total assets from prior year DEBT: long term debt scaled by total assets

εit: error term

Here, the quantity of total accruals of the company (TACCR) is used as one of the control variable. This is consistence with Velury and Jenkins (2006). It is argued that a positive relationship between the level of abnormal accruals and total accruals exists. It is explained that when total accruals increase, it will be more difficult for the shareholders to detect and distinguish discretionary portion of accruals from non-discretionary portion. As argued by Dechev at el. (2013), 60.1% of the executives they interviewed feel that higher level of earnings management will be conducted if they find the misrepresenting will go undetected. Thus, the management might get the opportunity to increase some level of earnings management. For this reason, and consistence to prior literature, TACCR is included as one of the control variable.

(21)

18

Lee et al. (2006) predict that the quality of earnings decrease with earnings growth. Although the mixed empirical results for this prediction are yielded, they note that firms with higher growth over-report earnings by a larger amount. It is noted that although the influence of growth on earnings’ quality may not be significant, the effects of this variable (GROWTH) cannot be neglected.

The DEBT is included as one of the variable because it is believed that debt can have a positive influence on earnings quality because managers are likely to use their accounting discretion to provide private information about the future prospects of the firm, trying to increasing firm value and to lower financing costs. Here the debt could act as a disciplinary instrument for the managers and helps to increase the quality of earnings. (Ghosh and Moon 2010) On the other hand, for high debt, it can also have a negative influence on earnings quality. When leverage goes up, the level of debt constraint is likely to increase (Ghosh and Moon 2010). Managers are more intend to use accruals aggressively to manage earnings to avoid covenant violations. (Betty and Weber 2003) This is consistent with Dichev et al. (2013), they document that avoiding covenant violations is one of the most important motivation for earnings management

A negative (positive) and significant α1 and α2 would indicate a higher (lower) earnings’ quality when the percentage of institutional (individual) ownership is high. A negative (positive) and significant α3 indicates a higher (lower) earnings quality when ownership concentration is high.

To detect whether ABNAC fairly represent the quality of earnings, I will examine whether the level of ABANC is consistent with the value yielded by earnings surprise indicator and the PEN modified from Penman 2001 approach. This can be conducted by examining the correlations of the three measures respectively. Only when the three approaches yielded highly consistent results of earnings’ quality, it can be confirmed the techniques used in this study fairly measured the quality of earnings. Then I will replace ABNAC with earnings surprise indicator and PEN respectively and run the model again to test the results.

(22)

19 4. Findings

4.1 Descriptive statistics

Table 1 presents the selected descriptive statistics of all the variables for the sample firms. The mean and median of institutional ownership is around 12.12% and 11% respectively, and the average percentage of company stock held by major shareholders is around 28.6%, with the first quartile of 14.43%, the median of 27.03% and the third quartile of 40.5%. The median of INS and CON are 11% and 27.03% respectively. The mean of total accruals is about -230878 and the average growth rate and debt level are 9.78% and 12.23% respectively. It is noted that the institutional ownership varies from 0% to 83% and the ownership concentration ranges from 0% to 99.5%, which demonstrates that the sample firms give a good coverage of the different level of ownership structure.

Table1

Descriptive statistics

ABNAC = E(TAi,t/Ai,t-1)- TAi,t/Ai,t-1, representing the total abnormal accruals

INS = The percentage of shares owned by the institutional shareholders CON = The percentage of shares owned by the major shareholders of the firm

TACCR = net income before extraordinary items less operating cash flows, representing total accruals

GROWTH = Percentage change of total assets from previous year DEBT = Long term debt / total assets

Table 2 presents the correlation coefficients of all variables. The results indicate that the correlation coefficients between independence variables as well as between

N minimum Maximum Mean 25% Median 75% Std. Deviation

ABNAC 484 -.4064 .9805178 -.0276 -.0744 -.0292 .0077 .1059 INS 484 .0 83.0 12.1180 .0 11.0 19.0 11.6944 CON 484 .0 .99500 .2860 .1443 .2703 .4050 .1909 TACCR 484 -21979558.00 1320886.00 -230878.2541 -101929.25 -30576.00 -5946.00 1357688.5946 GROWTH 484 -.7823 2.7030 .09778 -.0148 .0677 .1484 .2773 DEBT 484 .0 1.1013 .1223 .0 .0489 .1877 .1740 Valid N (listwise) 484

(23)

20

control variables are very low, with most of the correlation coefficients below 0.1. Little multicollinearity is present in the variables. With a positive correlation coefficient between INS and ABNAC and a negative correlation coefficient between CON and ABNAC, preliminary evidence for the hypothesis can be identified.

Table 2 Correlations

INS CON TACCR GROWTH DEBT ABNAC

INS Pearson Correlation 1 Sig. (2-tailed) N 484 CON Pearson Correlation .209 1 Sig. (2-tailed) .000 N 484 484 TACCR Pearson Correlation .098 .092 1 Sig. (2-tailed) .031 .042 N 484 484 484 GROWTH Pearson Correlation .026 .041 .003 1 Sig. (2-tailed) .575 .369 .944 N 484 484 484 484 DEBT Pearson Correlation .092 .225 -.043 .054 1 Sig. (2-tailed) .044 .000 .340 .239 N 484 484 484 484 484 ABNAC Pearson Correlation -.083 .070 .102 .064 -.028 1 Sig. (2-tailed) .067 .127 .024 .161 .540 N 484 484 484 484 484 484

(24)

21 accruals

4.2 Test of hypothesis

The results of regression which show the relationship between earnings’ quality and ownership structure are shown in table 3 and the model summary is shown in table 4. It can be found that the INS is negatively and significantly associated with ABNAC. (With a t-value of -2.383 and a Sig. of 0.018) That is to say, the higher level of institutional ownership is associated with the lower abnormal accruals. This is consistent with hypothesis 1. That is to say, the institutional ownership positively associated with the quality of earnings while the higher level of individual ownership is associated with lower earnings’ quality.

A positive coefficient is determined between ABNAC and CON, however, the t-test shows that this positive relationship is not significant (With a Sig. level of 0.061). This might be explained by the fact that not only entrenchment theory has an impact on earnings’ quality, but also it is influenced by the alignment effect. When the ownership concentration goes beyond a certain level, the alignment effect might present. The two effects work together and reduce the significance of this positive relationship between ownership concentration and earnings’ quality. For different companies, when ownership concentration goes up, it might be hard to determine upon which level the alignment effect will become stronger. And it might be impractical to expect only one of the effect presents solely in a group of firms. However, the positive relationship still support our theory in the second hypothesis, that is, for samples with a relatively low ownership concentration, fewer firms are able to achieve an ownership concentration that is high enough to generate alignment effect. The entrenchment theory would have a stronger effect on earnings’ quality, resulting in a positive instead of a negative relationship between abnormal accruals and ownership concentration. In practice, it is more likely to be the case that this contradictory effects of entrenchment and alignment would exists at the same time for

(25)

22

most types of sample firms, making it hard to generate a significant relationship. Future research might contribute on upon which level that we will observe the stronger impact of alignment theory on the earnings’ quality for certain companies.

Table 3

Regression results, dependence variable: ABNAC

Model Unstandardized Coefficients Standardized Coefficients t Sig. B Std. Error Beta 1 (Constant) -.027 .010 -2.845 .005 INS -.001 .000 -.110 -2.383 .018 CON .049 .026 .089 1.876 .061 TACCR .000 .000 .103 2.266 .024 GROWTH .025 .017 .065 1.433 .152 DEBT -.022 .028 -.037 -.793 .428 Table 4 Model summary R R Square Adjusted R Square Std. Error of the Estimate .176 .031 .021 .1047765024

4.3 Additional analysis

As mentioned above, multiple techniques should be used in order to measure the quality of earnings. In order to examine whether the different approaches yield consistent results of earnings’ quality, I first examine whether significant and positive relationship exists between ABNAC, earnings surprise indicator and the value yielded from the modified Penman (2001) approach. It is found that both earnings surprise

(26)

23

indicator and the value from modified Penman (2001) approach are significantly and positively related with ABNAC. Thus, the three different approaches for measuring earnings’ quality yielded similar results, we can judge the quality of earnings based on the approach used in this study.

Then, the ABNAC is replaced by earnings surprise indicator and PEN respectively, and the model is run again. Similarly, positive and significant relationship between institutional ownership and earnings’ quality is determined after replacing ABANC with PEN (as shown in table 5). When replacing ABNAC with earnings surprise indicator, a positive relationship, although not significant, between institutional ownership and earnings’ quality is also determined. In summary, all the three approaches adopted show positive relationships between institutional ownership and the quality of earnings, and two of them are significant. Because of this consistency among the results of more than approach, the hypothesis 1 is further supported.

Similarly, for ownership concentration and earnings’ quality, negative relationship is determined after replacing ABNAC with earnings surprise indicator and PEN respectively. And as shown in table 5, the positive relationship between PEN and CON is significant (with a t-value of 1.975 and a Sig. of 0.049), which gives stronger support to hypothesis 2 than the original model.

Table 5

Regression results, dependence variable: PEN

Model Unstandardized Coefficients Standardized Coefficients t Sig. B Std. Error Beta 1 (Constant) .071 .008 9.007 .000 INS -.001 .000 -.090 -1.979 .048 CON .042 .021 .092 1.975 .049 TACCR .000 .000 -.024 -.538 .591 GROWTH .069 .014 .218 4.915 .000 DEBT .023 .023 .045 .992 .322

PEN = Absolute value of net income minus operating cash flow, scaled by total assets of year t-1 INS = The percentage of shares owned by the institutional shareholders

(27)

24

CON = The percentage of shares owned by the major shareholders of the firm

5. Conclusion

This paper contributes on the issue of how corporate ownership would affect the quality of earnings. It brings together both the type and concentration of ownership and gives a more complete and comprehensive view of corporate ownership. The modified Jones Model is used as one of the most powerful model to measure earnings’ quality. In addition, additional measurements of earnings’ quality are adopted, namely, the earnings surprise indicator and modified Penman 2001 approach. This arguably increases the reliability of the measuring results of this study. The impacts of corporate ownership on earnings’ quality are examined based on 484 companies listed on NASDAQ. The percentage of institutional ownership and the percentage of shares owned by major shareholders of the company are used as the independence variables to illustrate ownership structure.

Empirical findings indicate that the institutional ownership is associated with higher earnings’ quality while the individual ownership is negatively related with the quality of earnings. This is consistence with the theories that the institutional shareholders have more incentive and ability to monitor the quality of earnings compared to the individual shareholders. Negative relationship between ownership concentration and earnings’ quality is detected in this study. However, only one approach used yielded significant relation. This can be explained by the contradictory effects of entrenchment theory and alignment theory when ownership concentration goes up. A negative instead of positive relationship between ownership concentration and earnings’ quality is found because that the entrenchment theory is more applicable for U.S. listing companies. It might be explained that although entrenchment effect

(28)

25

has a stronger impact, the alignment effect still exists and might to some extent mitigate this negative effect on earnings’ quality.

Future research could focus on at what level the alignment effect could be in place to replace the entrenchment effect when ownership concentration goes up. In addition, this study might have its limit because of the samples that are chosen. Companies from different countries might have different corporate ownership structures and it may have different impacts on the quality of earnings. For instance, future study can focus on the East Asian companies which have a higher level of ownership concentration, different level of entrenchment effect and alignment effect might be detected. Thus, it will be valuable to examine the impacts of ownership concentration on earnings’ quality based this different type of sample firms. Moreover, because of the availability of the databases used in this study, how the institutional ownership and ownership concentration are measured might not be the most accurate way. For instance, in this study, the institutional ownership is defined as the total share owned by pension funds, investment companies and mutual funds etc. However, different from Bushee (1998), which only includes the institutional owners with an equity holding of more than 100 million, no threshold value of equity owned is established in this study because of the limited availability of the database. This may arguably reduces the accuracy of the level of institutional ownership. Future studies could adopt a more accurate way of measuring the institutional ownership as well as the ownership concentration, for example, measuring the percentage of shares owned by five or three largest shareholders of the company. It is valuable to examine whether the result would change after different methods of calculating the independence variables are adopted. Furthermore, future research could use a larger sample size to allow more control variables to be included and increases the comprehensiveness of the model.

(29)

26 References

Abdelghany, K. E. (2005). Measuring the quality of earnings. Managerial Auditing

Journal, 20(9), 1001-1015.

Ali, A., Chen, T. Y., &Radhakrishnan, S. (2007). Corporate disclosures by family firms. Journal

of Accounting and Economics, 44(1), 238-286.

Beatty, A., & Weber, J. (2003). The effects of debt contracting on voluntary accounting method changes. The Accounting Review, 78(1), 119-142.

Bebchuk, L. A., Kraakman, R., & Triantis, G. (2000). Stock pyramids, cross-ownership, and dual class equity: the mechanisms and agency costs of separating control from cash-flow rights. In Concentrated corporate ownership (pp. 295-318). University of Chicago Press. Brown, N. C., Pott, C., & Wömpener, A. (2014). The effect of internal control and risk

management regulation on earnings quality: Evidence from Germany. Journal of

Accounting and Public Policy, 33(1), 1-31.

Bryan, D., Liu, M. C., Tiras, S. L., & Zhuang, Z. (2013). Optimal versus suboptimal choices of accounting expertise on audit committees and earnings quality. Review of Accounting

Studies, 18(4), 1123-1158.

Bushee, B. J. (1998).Investors on Myopic R&D Investment Behavior. The accounting review. Chung, R., Firth, M., & Kim, J. B. (2002).Institutional monitoring and opportunistic earnings

management. Journal of Corporate Finance, 8(1), 29-48.

Dechow, P., Ge, W., &Schrand, C. (2010). Understanding earnings quality: A review of the proxies, their determinants and their consequences. Journal of Accounting and

Economics, 50(2), 344-401.

Dempsey, S. J., Hunt, H. G., & Schroeder, N. W. (1993). Earnings management and corporate ownership structure: An examination of extraordinary item reporting. Journal of Business

Finance & Accounting, 20(4), 479-500.

Dichev, I., Graham, J., Harvey, C. R., & Rajgopal, S. (2013). Earnings quality: Evidence from the field. Journal of Accounting and Economics, 56(2-3), 1-33.

Ding, R., & Zhang, H. (2004). International comparison of ownership concentration in China.

(30)

27

Edmans, A. (2009). Blockholder trading, market efficiency, and managerial myopia. The

Journal of Finance, 64(6), 2481-2513.

Fan, J. P., & Wong, T. J. (2002).Corporate ownership structure and the informativeness of accounting earnings in East Asia. Journal of accounting and economics, 33(3), 401-425. Franks, J., & Mayer, C. (1997). Corporate ownership and control in the UK, Germany, and

France. Journal of Applied Corporate Finance, 9(4), 30-45.

Gillan, S., & Starks, L. (1998). A survey of shareholder activism: Motivation and empirical evidence. Contemporary Finance Digest, 2(3), 10-34.

Ghabdian, B., Attaran, N., & Froutan, O. (2012). Ownership Structure and Earnings Management: Evidence from Iran. International Journal of Business and

Management, 7(15), p88.

Ghosh, A. A., & Moon, D. (2010). Corporate debt financing and earnings quality. Journal of

Business Finance & Accounting, 37(5‐6), 538-559.

Gul, F. A., Kim, J. B., & Qiu, A. A. (2010). Ownership concentration, foreign shareholding, audit quality, and stock price synchronicity: Evidence from China. Journal of Financial Jiang, W., & Anandarajan, A. (2009). Shareholder rights, corporate governance and earnings quality: The influence of institutional investors. Managerial Auditing Journal of Financial

Economics, 95(3), 425-442.,

24(8), 767-791.

Hadani, M., Goranova, M., & Khan, R. (2011).Institutional investors, shareholder activism, and earnings management. Journal of business research,64(12), 1352-1360.

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.

Johnson, R. A., & Greening, D. W. (1999). The effects of corporate governance and

institutional ownership types on corporate social performance. Academy of Management

Journal, 42(5), 564-576.

Kapopoulos, P., & Lazaretou, S. (2007). Corporate ownership structure and firm performance: evidence from Greek firms. Corporate Governance: An International Review, 15(2), 144-158.

(31)

28

sponsors. The Accounting Review, 84(3), 623-658.

Knechel, W. R. (2000). Auditing: assurance & risk. South-Western Pub.

Koh, P. S. (2003). On the association between institutional ownership and aggressive corporate earnings management in Australia. The British Accounting Review, 35(2), 105-128.

Kirchmaier, T., & Grant, J. (2005).Corporate ownership structure and performance in Europe. European management review, 2(3), 231-245.

Leech, D. (1987). Corporate ownership and control: a new look at the evidence of Berle and Means. Oxford Economic Papers, 39(3), 534-551.

Lee, C. W. J., Li, L. Y., & Yue, H. (2006). Performance, growth and earnings management. Review of Accounting Studies, 11(2-3), 305-334.

Monks, R. and Minow, N. (1995), Corporate Governance, Blackwell, Cambridge, MA.

Penman, S. H., & Penman, S. H. (2007). Financial statement analysis and security valuation. Rajgopal, S., & Venkatachalam, M. (1998). The role of institutional investors in corporate

governance: An empirical investigation. Working paper, University of Washington/Stanford

University.

Ramsay, I.M., Stapledon, G.P., Fong, K., 2000. Corporate governance: the perspective of Australian institutional shareholders. Company and Securities Law Journal 18 (2), 110–142.

Schipper, K. (1989). Commentary on earnings management. Accounting horizons, 3(4), 91-102.

Schipper, K. and Vincent, L. (2003), Earnings quality, Accounting Horizons, Annual, pp. 97-111 Shuto, A., & Takada, T. (2010). Managerial ownership and accounting conservatism in Japan:

A test of management entrenchment effect. Journal of Business Finance &

Accounting, 37(7‐8), 815-840.

Velury, U., & Jenkins, D. S. (2006).Institutional ownership and the quality of earnings. Journal

of Business Research, 59(9), 1043-1051.

Villalonga, B., & Amit, R. (2006). How do family ownership, control and management affect firm value?. Journal of Financial Economics, 80(2), 385-417.

(32)

29 Research, 44(3), 619-656.

The effects of corporate ownership on earnings’ quality：evidence from the United States

The Effects of Corporate Ownership on Earnings’ Quality：

Evidence from the United States

NAME: Yumo Yang

STUDENT NUMBER: 10605541

DATE: June 22, 2014

MSc Accountancy & Control – Track Control

Amsterdam Business School

Faculty of Economics and Business, University of Amsterdam

Contents

The Effects of Corporate Ownership on Earnings’ Quality:

Evidence from the United States