1
Master Thesis:
Institutional investors and earnings quality:
Evidence from pre- and post-China SOX
period
Student name: Ningbo Li Student number: 10839151 Date: 22 June, 2015 Word count: 12011
Supervisor: Dr. Réka Felleg
MSc Accountancy & Control, specialization Accountancy Faculty of Economics and Business, University of Amsterdam
2
Abstract
This study examines the effects of institutional investors on earnings quality from pre- and post-China SOX period. Since Chinese capital market and regulations are still developing nowadays, these differences cause unique situation in china compared to the developed countries. Moreover, it is a heated debate between corporate governance and institutional investors. As earnings quality is a key part of corporate governance, the impact of institutional investors on earnings quality has been investigated but the results are not sufficient to explain some aspects. Earnings quality is typical to represent accounting information quality, and it is the important aspect that investors concern the most.
This study uses earnings informativeness to measure earnings quality and select 2695 firm-years over the period from year 2009 to 2014, which is three years before and three years after the effectiveness of China SOX. And the results of this paper suggest that (1) percentage of institutional ownership is positive related to the earnings quality, and (2) China SOX plays an important role on the earnings quality. Contrast to the pre-China SOX period, earnings quality increases in the post-China SOX and (3) I also find that the China SOX helps to strengthen the relation between institutional investors and earnings quality.
3
Statement of Originality
This document is written by Ningbo Li, 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.
Acknowledgements
Firstly, I wish to express my sincere thanks to my supervisor Ms Réka Felleg for providing me with all the necessary helps for the research and for her constant encouragement and guidance.
Secondly, I would like to express my gratitude to my colleagues and friends who gives me helps during this Master program.
Lastly, I would like to thank all of the Faculty of Economics and Business members for all works you have done.
4
Table of content
1. Introduction ... 5
2. Literature review and hypotheses ... 9
2.1 Institutional investor ... 9
2.2 The Interaction between Institutional investor and corporate governance. ... 10
2.3 Earnings quality ... 13
2.4 Sarbanes–Oxley Act ... 15
2.5 Current situation in China ... 16
2.5.1 Institutional investors in China ... 16
2.5.2 China Sarbanes-Oxley Act ... 18
2.6 Hypotheses development ... 19
2.6.1 Institutional investors and earnings quality ... 19
2.6.2 China SOX and earnings quality ... 22
2.6.3 Institutional ownership and earnings quality in pre-China SOX and Post-China SOX period ... 22
3. Research design ... 23
4 Evidence ... 28
4.1 Sample... 28
4.2 Descriptive Statistics ... 28
5. Tests of hypotheses ... 33
5.1. Institutional ownership and earnings quality ... 33
5.2. China SOX and earnings quality ... 35
5.3. Institutional ownership and earnings quality in pre-China SOX and Post-China SOX period... 36
6. Conclusion ... 38
5
1. Introduction
Nowadays, as China‟s economy is developing fast, it has become the world's second largest economy valued by nominal GDP and has and the world's largest economy valued by purchasing power parity (Geng et al., 2013).
However, the capital market is not mature. Not well-developed and single-layer capital market system increases information asymmetry. And information asymmetry causes agency problems which include adverse selection and moral hazard. Therefore, adverse selection and moral hazard problems in China are more serious than the markets in western countries (Jensen and Meckling, 1976).
Knowing the current situation of capital market in China, the Ministry of Finance of P.R China and other legislatures announces a regulation named “Basic Norms of Internal Control”. “Basic Norms of Internal Control”, which also called China SOX, was effective in 2012 and aims at enhancing internal control of firms.
For the investors, the most common way for investor to gain information is from the information on financial reports. The information in financial report is deemed to be reliable and is able to directly influence the decision-making of many investors due to it is a easy way to provide information of firms. Earnings quality is one of the most important factors to represent the quality of financial reports.
From the 1980s, the increasing number of Institutional investors has played an important role in the financial market. Compared to small investors who are likely to invest for speculation, institutional investors are knowledgeable and sophisticated in capital market. Institutional investors are most likely to have influence on management and inside private information and they are able to conduct direct monitoring (Carleton et al., 1998). Due to the fact that institutional investors usually hold a lot of share on firms, it is reasonable to infer that institutional investors can vote or propose
6
suggestions on the board of director and the power of institutional investors cannot be ignored. Because of the heavy information asymmetry in China‟ capital market, I anticipate institutional investors play an important role on the earnings quality.
However, regarding to former researches, under the circumstance that the strength of financial oversight and regulations are not the same, it is unclear that whether the researches in US would be able to explain the institutional investors‟ behaviors in China. Specially, it is unclear whether institutional investors may help to enhance earnings quality or may reduce earnings quality in China.
Given that research about institutional investors‟ behavior in China has not been investigated deeply, I am motivated by this to add new insights into how institutional investors in China affect earnings quality. Therefore, my motivations are the following aspects. First, the former studies focus on earnings management research, while using accounting accruals as proxy to test the relation between earnings management and institutional investor. In the meantime, although some other research directly tests earnings quality, they use earnings persistent as the proxy to test this relation. My study aims at using ERC as the proxy to test earnings informativeness, aiming to enhancing the understanding of the relationship between earnings quality and institutional investor and contributing to the former research which lacks of using ERC a the proxy of earnings quality, because ERC is the proxy which has been commonly used to analyze the earnings quality (Wan &Xiao 2012). It could be that my results are different from others‟, consequently helping to better understand the function of institutional investors, and add a debate on the relationship between institutional investor and earnings quality.
Secondly, former research investigates the relationship between institutional investors and earnings quality during the period when the China capital market was just beginning to develop. Because the China capital market has been developing for another decade, the influence of institutional investors is unclear nowadays and the situation might have changed. Therefore, the results may vary from different time periods and my results may be able to more representative to demonstrate the relation
7
nowadays.
Thirdly, because of lacking of data at early time, the samples in former researches were neither enough nor representative. My samples would be sufficient to test the real relationship for this reason that data is more available and more easily collect than before.
Fourthly, China SOX is such an important regulation of the whole Chinese market. However, there is no research to test the impacts of China SOX on the earnings quality and institutional investors. I believe my work would contribute to this gap.
Specially, I compare the earnings quality of different firms whose ownerships are held by different institutional investors.
Based on the research question “To what extent does Institutional investor have impacts on the earnings quality of companies?”, I investigate the relationship between institutional investor and earnings quality. Moreover, the regulation “Basic Norms of Internal Control” in China, normally called China SOX or C-SOX, was applied to the firms listed in the main part of China stock market. The China SOX mandates to report internal control annually for these firms. Does China SOX enhance earnings quality? Furthermore, relating to the research question that I propose, I wonder if China SOX have an impact on this relationship.
Based upon prior literature, I expect a positive relationship between institutional investor ownership and earnings quality. Additionally, I predict that the China SOX help to promote the earnings quality. Finally, I suggest that positive relationship between institutional investor ownership and earnings quality is stronger in the post-SOX period than in the pre-SOX period.
Because the China SOX imposed mandatory disclosure of self-assessment of internal control in annual financial reports for the firms which are listed in Shenzhen Stock exchange (SZSE) and Shanghai Stock Exchange (SSE), my firms sample come from these two stock exchange markets. After analyzing the results of my regression models, I find that as the institutional investors ownership increases, earnings quality increase. Besides, the result also indicates China SOX helps to enhance earnings quality.
8
Finally, consistent with my prediction, China SOX strengthen the interaction between earnings quality and institutional investors.
This study has contributions to former researches. To begin with, my study provides empirical evidence to the earnings quality literatures. My results suggest that institution investor enhance earnings informativeness, and high earnings informativeness can be described as high earnings quality. Although there are plenty of studies analyze the relation between institution investors and earnings quality in China, seldom studies use earnings informativeness as the measure of earnings quality. What is more, this study contributes to the corporate governance literature. The involvement of institution investors not only changes the structure of board of directors but also give pressures to the decision-making process of managers. In light of researches (Monks and Minow,1995; Gillna and Sartks 2000), institution investors would be able to more or less monitor and regulate firms. At last, my study also contributes to the literatures about SOX in an aspect, pointing out some similarity between America SOX and China SOX, which improve earnings quality. Because it is not a long time since CHINA SOX became effective, researches about impacts and functions of China SOX are limited.
In the next section prior literature will be discussed as well as the part about how hypotheses have been developed. And then in the third section, I will discuss the methodology and research design. After this section, the results are presented in the tables and I will analyze the result, finding out whether the results are supporting my hypotheses or reject hypotheses. Finally, in the conclusion part, I give the overall results, opinions about to who this research contributes and the limitations of my research.
9
2. Literature review and hypotheses
2.1 Institutional investor
Institutional investors are financial institutions which receive funds including pension funds, mutual funds and insurance companies from third parties. Institutional investors are also organizations gathering money from small investors and investing in stocks, real property and other assets of companies. They hire more professional and influential financial experts to operate the fund they received.
Callen and Fang (2013) describe that as for institutional investors, there are two opposing views: 1. Short-termism party which focuses on short-term value of their investments. 2. Monitoring party which focuses on long-term value of their investments.
Allegedly placed excessive focus on short-term developments, short-termism institutional investors make managers fear that an earnings disappointment will trigger large-scale institutional investor to sell stocks, and meanwhile results in a temporary undervaluation of the firm's stock price (Graves& Waddock, 1990). The results also imply that transient institutional investor ownership has an adverse impact on public firms, ultimately increasing the risk of future stock price crash. Thus, they results cast strong doubt on any strategy aiming at attracting institutional investors without considering the investment focus and preferences of these institutions. Moreover, Coffee (1991) documents short-termism institutional investors concentrate on near term earnings, overlooking the long-term firm value.
However, this view of institutional investor behavior is not universal. Others argue that the large stockholdings and sophistication institutional investors allow them to monitor and discipline managers, ensuring that managers may choose investment levels to maximize long-run value rather than to meet the goals of short-term earnings (Monks and Minow 1995, chap. 2; Dobrzynski, 1993).
10
investments. This is because of the magnitude of their equity stakes and the penalties that they bear. For example, under the premise that institutional investors hold a large share of a company, when the company face a penalty, the amount of cost that institutional investors bear is the penalty multiplies the percentage of share that held by the institutional investors. Therefore, the larger the ownership, the more cost institutional investors bear. (Johnson & Greening, 1999; Ciccotello & Grant, 1999). Similarly, Maug (1998) suggest that the activism of institutional investor improve firms‟performance. Chaganti et al. (1995) use ROA to evaluate performances of firms, concluding that the ownership of institutional investor is strong related to firms‟ performance.
2.2 The Interaction between Institutional investor and corporate governance.
As the separation of control and ownership in the firms, agency problem and information asymmetry arise and are the main risks for a company. Agency problem is what the managements are acting for their own best interest which is inconsistent with the goals of the company and all shareholders.
Agency problem and information asymmetry create moral hazard and adverse selection. Managements take actions, such as using earnings management to increase small earnings to meet the benchmark and reducing R&D to increase the profit for the current year, to make their performance good in the financial report, which actually reduces the value of firm in a long term (Ruxandra-Adriana, 2015).
Douma and Schreuder (2013) argue that good corporate governance mechanisms and controls are able to lower the negative impacts which arise from moral hazard and adverse selection. The corporate governance mechanisms and controls not only come from internal control but also from external monitoring systems. And, by disclosing information in a timely manner, governance mechanisms can help reduce information asymmetry (Sanjeev, 2003).
11
what it is. Feleaga (2010) argues that corporate governance can be viewed by different value related groups, for instance,, shareholder group and stakeholders group. For shareholder group, corporate governance is the way how funds suppliers ensure that they will receive proper reward for their investments. And as for stakeholder views, corporate governance stands for a set of mechanisms to segregate duties and rights, to influence management‟ decisions, and to reduce managers‟ discretionary space.
The study on corporate governance includes a lot of theories and themes. And the influence of institutional investor on the corporate governance is an important branch. The Board of Directors can be classified into three broad components; executive, monitoring and instrumental. And the role that institutional investors participate is the monitoring components.
When institutional investors make an investment in a certain company, they have the fiduciary obligation of closely monitoring the companies which they invest in. Therefore, institutional investors would take necessary actions to protect them from the loss, thus ensuring the value of this firm following an upward trend (Mallin et al., 2005). Because institutional investors are most likely having easier access to management and inside private information, they are able to conduct direct monitoring (Carleton et al., 1998).
However, there are two contradict views about relation between Institutional investor and corporate governance. Monks and Minow (1995) believes that institutional investors are capable of monitoring, regulating and reducing earnings management. But the effectiveness of monitoring on the director board varies from different types of institutional investors. Some types of institutional investors acting as supervisors are likely to challenge the board of directors (Brickiye, 1987). Gillna and Sartks (2000) examine 2042 shareholders‟ proposals which contain the opinions and suggestions of shareholders, finding that the proposals provided by institutional investors receive more supports than those provided by normal shareholders. I believe that the reason why institutional investors‟ proposals get more support is that institutional investors are more sophisticated and are capable of preventing their
12
investments from loss in a more professional way. Therefore, the other shareholders tend to believe that the suggestions made by institutional investors are more effective in enhancing the performance of the firm.
Besides the direct impact on corporate governance, institutional investors also affect the corporate governance of firm in an indirect way. For reason that institutional investors hold a lot of capital, firms want to attract institutional investors to raise money. Therefore, firms would increase the effectiveness of corporate governance and improve the financial reports‟ quality, which are some of the main factors valued by the institutional investors to decide whether they should invest, or attract the institutional investors (Ertuna& Tukel, 2013).
Porter (1992) also found another indirect way. Since selling a lot of shares will lead to a sudden and dramatic drop of stock price, firms are afraid that, when institutional investors find the financial reports look bad, they will sell a large number of shares. As a result, the management board of firms may choose to manipulate the financial report.
Conversely, in the study of the United States, Wahal (1995) finds that large institutional investors have passive stance on corporate governance issues. Institutional investors activism may also be hindered by lack of accountability, absence of appropriate incentives and free-rider problem due to the fact that any one of the institutional investors holds one or two percent of a corporation. Additionally, others have argued that institutional investors have limited incentives to monitor management actions. And this could be because that free-riding among institutional investors makes it difficult for them to take collective action (Black 1990; Admati, Pfleiderer, and Zechner, 1994). Due to a free rider problem, it has been argued that only a large shareholder has incentive to undertake monitoring or other costly control activities. However, all shareholders may benefit from such activities even if they do not bear the cost of the process. And the investor with a larger stake in the firm has stronger incentives to undertake monitoring activities, as it is more likely that the large shareholder‟s increased return from monitoring is sufficient to cover the associated
13
monitoring costs (Hart, 1980).
Further, for the studies in the concentration of institutional investor ownership, one hypothesis suggests that, to secure benefits that are to the detriment of other providers of capital (shareholders), concentrated ownership allows the block-holder to exercise undue influence over the management. This is called the “private benefits hypothesis”.
Comparatively, a conflicting hypothesis (“shared benefits hypothesis”) suggests that block-holding leads to more efficient monitoring by the management and that the benefits from such monitoring are shared by all stock holders.
Since monitoring is costly, former researches also suggest that only large share of shareholder are likely to gain the large return which can cover the monitoring cost. (Gillan and Starks, 2007; Shleifer and Vishny, 1986). From this interpretation, Schnatterly et al. (2008) argue that only large ownership institutional investors can cover monitoring cost, supporting the opinion of Hart (1980).
2.3 Earnings quality
According to the interpretation of the literature, earnings quality is based on the term “earnings”. Earnings are the net benefit of a company (Robert G. Eccles, 2011). Earnings are commonly used to contract agreements. If compensation agreements and debt agreements are based on low-quality or defective earnings, it will induce unnecessary wealth transfers (Schipper and Vincent, 2003).
For example, as earnings are commonly used as a main indicator to performance of managers, managers tend to overstate earnings by earnings management, containing both real earnings management and accrual earnings management. In this way, managers receive more compensation than the amount that they are supposed to receive. In consistent with agency theory, the actions of overstating earnings are decreasing the firm long-term value, deteriorating solvency, leading lenders mistakenly to continue lending and causing some other negative effects. The quantity is good in paper but is bad indeed. As earnings quality is a term which is full of study, this is only a small part
14
of the discussions of the earnings quality.
The FASB's (2006) Conceptual Framework points to decision usefulness, which fundamentally based on relevance and faithfulness, as the benchmark for assessing earnings quality. They believe the earnings are in high quality when it is benefit to improve decision making for all stakeholders. Likewise, Kormendi and Schipper (1987) suggest that sustainable earnings, which are also named “earnings persistence”, stand for high earnings quality. Beaver et al. (1979) select the 276 firms from year 1965 to 1974, divide them into 25 portfolios, and analyze the relation between change of earnings and the change of stock price. They find out that the as the change of earnings increase, the change of stock price also increase. There is a positive relation between change of earnings and the change of stock price. This means that the earning is important factor to influence stock price. Therefore, this result implies earning quality is really important for investors.
Just because of the widespread study and use of earnings quality, however, there is no consensus on this term‟s definition.
In this paper, the informativeness of accounting earnings is applied to investors as a measure of the quality of accounting information (Fan and Wong 2002). And earnings response coefficient (ERC) is adopted as a proxy of earnings informativeness. It is significant to apply ERC to measure informativeness, as distinct studies have provided more direct evidences on ERCs as a proxy for earnings quality or earnings informativeness (Dechow, 2010).
From the beginning, Ball and Brown (1967) have used equity market responses of earnings to infer earnings quality. They find that the information in equity market is corresponding to the earnings information. In the late 1980s, researchers started to investigate a new area – the earnings response coefficient (ERC) that is theoretically defied as “a change in the price induced by a one-dollar change in current earnings” (Collins and Kothari, 1989). ERC is included in time series properties of annual earnings, which also includes persistence and variability (Schipper and Vincent, 2003).
15
able to is can enhance relation between prices (or returns) and earnings because low variability and high persistence. While the studies on the average price-to-earnings ratio (Stankevičienė, Gembickaja, 2012) are concentrating on market reactions to earnings announcements, the studies on the earnings response coefficient are more interested in the nature of information about reported earnings and how these are linked to firm valuation (Kormendi and Lipe, 1987). Therefore, I use ERC as a proxy of earning quality.
2.4 Sarbanes–Oxley Act
After the collapse of Enron and Worldcom, American government realized strict regulations are needed to deal with accounting scandals and reinforce accounting information. The Sarbanes–Oxley Act of 2002 (SOX) is a law which aims to enhancing standards for all U.S. public company boards, management and public accounting firms.
Sarbanes-Oxley (2002) reported that auditors should discuss the quality of the financial reporting methods to increase the quality of financial reporting, thus subsequently having a great impact on accounting quality and accounting conservatism. The results of Martijn Verleun (2011) provid evidence that the enactment of SOX has a positive effect on accounting quality.
Cohen (2008) find an increase in earnings management at the period preceding the passage of SOX. And evidence suggests that subsequent to the passage of SOX, the level of earnings management returned to the pre-SOX trend. In terms of SOX, there was a dramatic drop in earnings management. But just a period of time after SOX, firms shifted from using accrual-based to real earnings management.
As the implementation of SOX in America has increased earnings quality (Peter, 2010), it is rational to expect that the China SOX will also increase earnings quality because both SOXs are focusing on the effectiveness of internal control.
16
2.5 Current situation in China
2.5.1 Institutional investors in China
At present China's institutional investors includes the various types, mutual funds position of absolute dominance, and follows by brokers, trust funds and financial companies holding and some other type of institutional investors. (Wang, 2008)
The real development of China's institutional investors should be from the 1997 "Interim Measures on Securities Investment Fund Management", which set up the definitions and requirements of establishment, fund-raising and trade of securities investment funds and define the rights and duties of fund trustees, fund managers, and fund holders.
In particular, February 2nd, 2004 the ministry of finance in China promulgated the regulation named "The suggestions of promotion of the capital market reforming and of stable developments", which promotes the effectiveness of capital market, helps capital market to play an main role in the Chinese economy, and provides a lot of supports for the developments of institutional investors.
Before 2006, the proportion of small investors holding mutual funds is about 50%. Since 2006, mutual funds became more popular in the small investors. At the end of 2009, the number of mutual fund accounts reached 31.66 million and the proportion of individual investors holding the fund reaches 89 percent, marking the structure of China's securities investment fund investor change significantly.
However, the current situation of institutional investors still face serious problem. First, the structure of capital market lacks of diversity. The current development of China's securities market is mainly concentrated in the main board market and other markets are not mature. The disadvantage of single-layer capital market system is that the risks of futures, options and other financial derivatives cannot transfer effectively. Because risk is the main factor influencing the price, I expect the capital market in China is still not able to cultivate stable and rational institutional Investors.
17
market in China is still developing. The market lacks of the price discovery mechanism, increasing the Systematic risk and price risk of investment. Therefore, institutional investors are easy to manipulate the capital market by releasing unreliable news, which is unfavorable for the formation of concept of long-term investment.
In the case of Chinese stock markets, there is widespread popular belief that the tremendous growth in recent years has attracted sentiment-driven investors. A large number of literatures suggest that domestic Chinese investors indeed engage in short-term trading. Chinese investors are eager to earn in a short period instead in a long period. Mei et al. (2005) analyze domestic investors‟ trading activity, building on a model where heterogeneous beliefs and short-sale constraints imply a positive relationship between speculative stock prices and trading volume. Lei et al. (2007) point out that the scale of institutional investors is not large enough. The relatively small scale of institutional investors limits the ability of institutional investors to influence the corporate governance. Institutional investors do not have enough power to intervene the decision-making procedure. Sometimes they just behave like a individual investor.
Conversely, Lou Wei (2002) selects the samples from year 1998-2000 and justifies the positive relation between institutional investor ownership and the firms‟ performance. She finds that the institutional investor ownership is positively significantly related to Tobin Q and predicts that this positive relation is due to participation of institutional investor in the corporate governance.
However, the problem that institutional investors focus on the short-term performance and ignore the long-term firm value might be more serious in China than in developed countries. Although institutional investors have a large share of firms, the closed relation between other large shareholder and firms might diminish the impact of institutional investor on the board of director. Moreover, the less-developed oversight regulation in the capital markets may increase the possibility of another worse problem. The problem is that the listed firms and institutional investors join together to manipulate stock prices. In this way, both firms and institutional investors can earn a
18
large amount. Therefore, it is necessary to find out the real behavior of institutional investors.
2.5.2 China Sarbanes-Oxley Act
The regulation “Basic Norms of Internal Control”, which is viewed as“China Sarbanes-Oxley Act”, formally came into effect at the beginning of fiscal year 2012 to all qualified companies, including the companies listed on the main boards of Shanghai and Shenzhen stock exchange markets in China a. China SOX will apply to over 2000 companies in China.
All qualified companies reported their assessments of the design and operating effectiveness of internal controls over financial reporting, together with the external auditors‟ opinion on the effectiveness of these controls before the deadline, April 31, 2013. It is motivated by the similar regulatory purpose as SOX to improve the internal control quality for more reliable financial report and less restatement.
China SOX requires management to undertake an annual self-assessment of internal control effectiveness and disclose the conclusion in annual reports. China SOX aims to provide clear guidelines for Chinese firms to enhance their internal control sysrem to provide reasonable assurance on such objectives as legal compliance, asset safety, reporting accurancy and completeness.
Because the concept of internal control was new at that time, seldom companies had effective internal control standards and regulations of internal control. In order to comply with the China SOX, they have to invest a large amount on the testing and upgrading of internal control.
According to the result of the document named “internal control reports analysis of listed firm of shanghai stock exchange in 2007”, 147 firms out of 862 listed firms in shanghai stock exchange market provided self-assessment report of internal control. From this document, the defects of internal control heavily impacts on the effectiveness of daily operation. The main defects are that 1.Lack of regulation of internal control; 2. It is hard to implement because of the unreasonable design of internal control regulation,
19
3.Lacking of pre-training leads to interpretation difficulties in employees.
Thereupon I would anticipate the China SOX do impose a strong enforcement on internal control of the Chinese firms, thus improving the earnings quality of these firms.
2.6 Hypotheses development
2.6.1 Institutional investors and earnings quality
Monitoring earnings quality is necessary for institutional investors because of several incentives. First, high earnings quality is a main indicator of high quality financial statements, which are important source of information about the firm. Because institutional investors are professional investors, they are presumably interested in using all types of relevant information to make investment decision. In this reason, institutional investors prefer the firms with high earnings quality.
Due to the fact that institutional investors have impacts on corporate governance, a lot of scholars expect that there would be a relation between earnings quality and institutional investors.
However, this relationship is not clear in the former studies. On the one hand, Short-termism party is the party focuses on short-term value of their investments and has negative impacts on earnings quality. In support of the “passive monitoring hypothesis,” Karpoff et al. (1996) failed to confirm the positive effect of institutional activism on shareholder value. They even give pressure on the management board when the financial reports look not good. Similar results have been found by Lipoton (1991) Institutional investors are not likely to conduct effective measure to improve decision-making. Moreover, even when they want to take some actions, institutional investors face a lot of interference. In the light of “passive monitoring hypothesis”, fiduciary reason leads some institutions to rely on current earnings as a value proxy, leading to a potentially shortsighted focus on near-term earnings (Porter 1992). Lang and McNichols (1997) point out that trade of stock by institutional investor is increase (decrease) to earnings good (bad)news. What‟s more, the volume of trade is increasing when institutional ownership increase, leading to the huge fluctuation of stock
20
price.The ownership of short investment horizons institutions and stringent fiduciary standards institutions are associated with short-term earnings, while short investment horizons institutions tie with significantly mispricing of short term earnings (Bushee, 2001) Since some studies reveal that the concentration institutional investors adversely affect earnings quality. Without large institutional ownership, the concentration of institutional investors means nothing. Therefore, in some case, large institutional ownership is negatively related to earning quality. Sanjeev and Partha S, 2003)
On the other hand, some papers argue that institutional investor can be defined as monitoring party, which focus on long-term value of their investments and have a positive impact on earnings quality. In support of active monitoring hypothesis Brickley et al. (1988) documented that institutional shareholders are more likely to vote against harmful amendments which reduce shareholder wealth and firm value. In the research in Australia, an interesting result is that the graph of relationship between institutional ownership and levels of earnings management is a concave instead of a line. This non-liner relation reveal that the discretionary accruals is low when the institutional ownership is too high or too low while the discretionary accruals is at the highest point when the institutional ownership is around 54.3% (Koh, 2003). Warfield et al. (1995) find that earnings quality is positively related to the managerial ownership. Compared to small investor, institutional investors are sophisticated investors with plenty of knowledge to conduct monitoring function. As for the aspect of R&D investment, Bushee (1998) find that the institutional investors can reduce the possibility that managers using real earnings management, which mostly is cutting R&D expense, to increase short term earnings. It is just because institutional investors are sophisticated investors, they can collect more information to prevent managers from reducing R&D investments. This implies that institutional investors helps to enhance earnings quality.
As for the study of institutional owners in China, Cheng (2006) selects China's firms listed A-share market from year 2000 to 2003 and analyzes the role of
21
institutional investors in corporate governance. The results show that: institutional investors hold the proportion of shares positively related to the timeliness of earnings information, and negatively related to earnings management.
Xia and Li Gang (2008) use the samples of non-financial firms listed in A-share stock exchange market from year 2000 to 2004 in China to study the effect of institutional investors on accounting earnings quality. They compare the firms with and without involvement of institutional investors, finding that earnings managements are lower in the firm with institutional investors participation. And they also find that earnings managements increase as the proportion of institutional ownership decrease. Their results indicate that institutional investors can significantly improve earnings quality.
Institutional investors holding the Company's accounting conservatism also significantly higher than other company, and the robustness of accounting earnings stake with institutional investors to improve and enhance the timeliness attract institutional investment motive.
Based on three main factors, institutional ownership has impact on quality of earnings. First, in the motivation aspect, institutional investors which gather a lot of money from small investors are regulated and monitored by governments and some organizations. And they face the pressure to make sure their investment is a less risk project. These factors force institutional investors tend to invest on the companies which earnings quality is high because high quality of earnings means less evaluation risk. And the institutional investor will invest more on the company which has high earnings quality.
Second, in the ability aspect, the investments give them rights to participate the operation and finance policy which small investors have no right to do. In the theory part that I mentioned above, as the institutional investors who have large ownership of the company, they would be easier to discover private information and participate the decision-making process.
22
companies as his own companies and focus on the long-term running of these companies. In consistent to the results above, large ownership institutional investors have higher incentives to vote against the actions which reduce the firm value and to provide good suggestions for the long-term value of firms. The sustainability of the companies requires disclose of earnings is reliable, which is high earnings quality. Therefore, I develop the first hypothesis:
H1: Ceteris paribus, the percentage of institutional ownership is positive related to the earnings quality.
2.6.2 China SOX and earnings quality
The main objective of the China SOX act was to enhance the financial reporting quality and to restore the confidence of the investors regarding the financial numbers. Henceforth I would anticipate the China SOX does impose a strong enforcement on internal control of the Chinese firms, thus improving the earnings quality of these firms. Therefore, I form the second hypothesis:
H2: Ceteris paribus, China SOX has positive impacts on the earnings quality.
2.6.3 Institutional ownership and earnings quality in pre-China SOX and Post-China SOX period
Because China SOX require that companies reported their assessments of the design and operating effectiveness of internal controls over financial reporting, the private information in companies are easier to release to the public than before.
China SOX helps to reveal information and reduce the level of information asymmetry in the post-China SOX period.
I anticipate the relation between earnings quality and institutional investors to be strengthened in the post-China SOX period for two reasons: First, the lower information asymmetry allows institutional investors to monitor management actions at
23
a lower cost. It is easy and logical to expect that lower cost leads to the more willingness for institutional investors to conduct monitoring function.
2 The lower information asymmetry means the higher transparency. So the institutional investors would be capable of discovering some information that is unknown to them, thus helping to have better monitoring function.
Taking all of the above arguments together, I form the following hypothesis:
H3: Ceteris paribus, the percentage of institutional ownership is more positively related to the earnings quality in the post-China SOX period than in the pre-China SOX period.
3. Research design
Measures of earnings quality
ERC measures informativeness is an important and distinct study that provides more direct evidence on ERCs as a proxy for earnings quality or earnings informativeness (Dechow, 2010).
I use the model of Easton and Harris (1991) which propose that the ERC model can explain earnings quality well and use ERC to measure earnings quality. ERC is the proxy which has been commonly used to analyze the earnings quality (Wan& Xiao, 2012).
I employ the earnings level deflated by price as the explanatory variable in the price earnings regression. According to the prior research of Easton and Harris (1991), the based model of Easton and Harris about earnings informativeness is:
Rt+1 = β + β1ESP t+1/prit + ε (5.1)
Rt+1 = Return of the t period. Equal to the stock price at the end of t+1 period divided by the stock price at the end of t period
24
PRIt = Stock price at the end of t period
ESP t+1/Prit = Earnings per share of t period divided by Stock price at the end of t period
Measures of China SOX and institutional ownership
Variables Description
ESP t+1/Prit*INSt
The influence of Institutional ownership on earnings response coefficient
ESPt+1/Prit* China SOXt
The earnings response coefficient divided in the pre-China SOX period and post-China SOX period
ESPt+1/Prit* China SOXt*INS
The influence of Institutional ownership on earnings response coefficient divided in the pre-SOX period and post-SOX period
Control variables
In this study, I use four control variables.
First, I included a dummy variable to control for the influence of SOX on earnings quality (Chang et al., 2012). The dummy variable (China SOXt) is 1 for the post-SOX period and 0 for the pre-SOX period.
Second, I include firm size (SIZE), which measure size =log (asset of each firm), to control for institutions‟ preferences for larger firms due to concerns over prudent person standards, the desire for liquidity, and the presence of lower information asymmetries (Cready, 1994).
Third, I use debt to asset ratio, Because firms with a high level of leverage tend to have greater debt-holder and shareholder conflicts (Watts, 2003).
Fourth, I use Tobin‟s Q. The Tobin‟s Q usually is used as an indicator of a company performance (Joseph, 2003).
25
Empirical models
Following the objective of this paper, I add the Tobin‟s Q, liability/asset ratio, currency/asset Ratio, industries and year as controlled variables, and then form the new regression model as followed:
Rt+1 = β + β1ESPt+1/prit + β2SIZEt + β3TOBINQt + β4DTARt + ε (5.2)
SIZEt = Scale of firm, while size =log (asset of each firm) TOBINQt = Market value of firm to book value of firm. DTARt = Debt to asset ratio
For the first hypothesis H1, I focus on the percentage of institutional ownership. So I will add the percentage of institutional ownership to the model 5.2. According to the hypothesis, I expect it a positive number. And then I form a new model:
Rt+1 = β + β1ESPt+1/prit + β2ESPt+1/PRIt*INS + β3INSt
+ β4SIZEt + β5TOBINQt + β6DTARt + ε (5.3)
INSt = Percentage of shares held by Institutional investor
ESPt+1/Prit*INSt = ESPt+1/Prit. multiplied by percentage of shares held by institutional investor
The coefficient for multiply percentage of shares held by Institutional investor by ESPt+1/PRIt stands for the influence of Institutional ownership on the earnings quality. I mainly focus on the coefficient β2 and the hypothesis 1 would be supported if β2 is positive.
For the second hypothesis H2, I focus on the influence of China SOX on earnings quality. So I will add the Dummy variable China SOX to the model 5.3. According to the hypothesis, I expect it a positive number. And then I form a new model:
26
Rt+1 = β + β1ESPt+1/prit + β2ESPt+1/Prit*ChinaSOXt
+ β3ChinaSOXt + β4SIZEt + β5TOBINQt + β6DTARt + ε (5.4)
ChinaSOXt =Dummy variable; 0 is the pre China SOX period and 1 is the post China SOX period
ESPt+1/Prit*ChinaSOXt = Dummy variable of ChinaSOXt multiplied by ESPt+1/Prit.
The coefficient for multiply ChinaSOXt by ESPt+1/PRIt stands for the influence of China SOX on the earnings quality. I mainly focus on the coefficient β2 and the hypothesis 2 would be supported if β2 is positive.
For the third hypothesis H3, I focus on the influence of China SOX on the relation between institutional ownership and earnings quality. According to the hypothesis that China SOX strengthen this relation, I expect it a positive number. And then I form a new model:
Rt+1 = β + β1ESPt+1/prit + β2ESPt+1/Prit*ChinaSOXt*INS + β3ESPt+1/Prit*INSt
+ β4ESPt+1/Prit*ChinaSOXt + β5ChinaSOXt*INSt + β6ChinaSOXt
+ β7INSt + β7SIZEt + β8TOBINQt + β9DTARt + ε (5.5)
ESPt+1/Prit*INSt = ESPt+1/Prit. multiplied percentage of sharesheld by institutional investor
ESPt+1/Prit*ChinaSOXt = Dummy variable of ChinaSOXt multiplied ESPt+1/Prit. ESPt+1/Prit*ChinaSOXt*INSt= Dummy variable of ChinaSOXt multiplied ESPt+1/Prit
and percentage of shares held by Institutional investor ChinaSOXt*INSt = Dummy variable of ChinaSOXt multiplied percentage
of shares held by Institutional investor
The coefficient β2 for multiply ChinaSOXt by ESPt+1/PRIt and institutional ownership stands for the influence of ChinaSOXt on this relation. I mainly focus on the coefficient β2 and the hypothesis 3 would be supported if β2 is positive.
27
Summary the variables
Rt+1 = Return of the t period. Equal to the stock price at the end of t+1 period divided by the stock price at the end of t period
ESPt+1 = Earnings per share of t period PRIt = Stock price at the end of t period
ESPt+1/Prit = Earnings per share of t period divided by Stock price at the end of t period
INSt = Percentage of shares held by Institutional investor SIZEt = Scale of firm, while SIZE =log (asset of each firm) TOBINQt = Market value of firm to book value of firm.
DTARt = Debt to asset ratio
ChinaSOXt = Dummy variable; 0 is the pre China SOX period and 1 is the post China SOX period
ESPt+1/Prit*INSt = ESPt+1/Prit. multiplied percentage of shares held by institutional investor
ESPt+1/Prit*ChinaSOXt = Dummy variable of ChinaSOXt multiplied ESPt+1/Prit. ESPt+1/Prit*ChinaSOXt*INSt = Dummy variable of ChinaSOXt multiplied ESPt+1/Prit and percentage of shares held by Institutional investor China SOXt*INSt = Dummy variable of ChinaSOXt multiplied percentage
28
4 Evidence
4.1 Sample
This paper selects the firms are all included in Shenzhen Stock exchange (SZSE) and Shanghai Stock Exchange (SSE), two stock exchanges in the People's Republic of China. Because the China SOX was effective in 1st January 2012, I choose the firm year from 2009 to 2014, which are 3 years before and 3 year after effective date of China SOX. Because of the way I choose is to measure the earnings response coefficient, independent variables and control variables are all 1 year behind the dependent variable. Therefore, the years I measure are actually from 2009-2013. Except for the data of institutional investors which is collected from Resset database, all the data that I collect are from CSMAR database.
First I collect 2699 firms with 13495 observation years. After deleting the firms with missing data, 553 firms remain with 2765 observation years. For 3 firms that is listed in the B share stock market of SZSE and SSE, which is a special stock market only for the foreign companies, and 11 firms with ST or *ST, which represent the firm with negative income for 2 or 3 years and facing the delisting risk, I neglect these firms. Finally, I have totally 539 firms with 2695 observation years as my samples to measure earnings quality.
4.2 Descriptive Statistics
Panel A of Table 1 illustrates the industry distribution across my sample firms and indicates that the sample firms are from a wide range of industries. Among these industries, The majority type of industry is manufacturing for the measure of earnings quality, which has 300 firms and occupies 55.66% of the total samples.
29
Table 1Sample reconciliation and sample distribution
This table provides the sample reconciliation (Panel A) and the sample distribution according to industry (Panel B).
Panel A: Sample reconciliation
Total number of observations in Resset and CSMAR (2009-2014) 13495
less: Observations with missing data to calculate earning quality (10730)
(mainly due to missing data on institutional investors) less: Observations of ST or *ST companies (55)
less: Observations of B share market companies. (15)
Final sample of firm observations from 2007 to 2010 2695
Numbers of firms in the final sample 539
Panel B: Sample Industry Distribution Industry
Industry Observati
ons
Perce nt
Agriculture, Forestry and Fishing A 5 0.93%
Mining B 19 3.53%
Manufacturing C 300 55.6%
Electric, Gas and Water D 23 4.27%
Construction E 12 2.23%
Wholesale and Retail Trade F 40 7.42%
Transportation G 34 6.31% Services H 5 0.93% Software I 16 2.97% Finance J 18 3.34% Real Estate K 48 8.91% Rental service L 4 0.74% Sanitary N 8 1.48%
Culture, Sport and Entertainment R 1 0.19%
General S 6 1.11%
Total 539 100%
Panel B of Table 1 illustrates the industry distribution across my sample firms and indicates that the sample firms are from a wide range of industries. Among these industries, the majority type of industry is manufacturing for the measure of earnings quality, which has 300 firms and occupies 55.66% of the total samples.
30
Table 2.Institutional investors ownership
Panel A: average proportion of share held by institutional investors for the sample firms.
Fiscal year Percent
2009 18.45%
2010 18.21% 2011 20.01%
2012 21.31%
2013 21.20%
As can be seen on the panel A of table 2, the proportion of institutional investors ownership has been increasing over time. However in 2012, which is the first year after the implementation of China SOX, I see the strong increase of the number of institutional investors ownership. And the compared to fiscal year 2012, 2013 see a little drop on the number of institutional investors ownership. Combining these two changes, I conclude it could be an indication of the relationship between the China SOX and institutional investors. And the impact of China SOX on institutional investors was larger in the first year after implementation than the second year.
TABLE 3.Descriptive statistics
Variable Mean Median Standard
Deviation First Quartile Third Quartile Rt+1 1.022 0.935 0.404 0.758 1.193 ESPt+1/Prit 0.0307 0.0273 0.0452 0.0115 0.0476 INSt 0.198 0.141 0.193 0.042 0.298 SIZEt 9.746 9.604 0.760 9.249 10.076 TOBINQt 1.980 1.575 1.350 1.189 2.289 DTARt 0.513 0.517 0.202 0.373 0.652 ChinaSOXt 0.4 0 0.489 0 1
ESPt+1/Prit*INSt 0.0076 0.0031 0.0154 0.0005 0.0101
ESPt+1/Prit*ChinaSOXt 0.0138 0 0.0407 0 0.0207
ESPt+1/Prit*ChinaSOXt*INSt 0.0037 0 0.0134 0 0.0017
31
Variable definitions:
Rt+1= return of the t period. Equal to the stock price at the end of t+1 period divided by the
stock price at the end of t period ESPt+1= Earnings per share of t period
PRIt= Stock price at the end of t period
ESPt+1/Prit= Earnings per share of t period divided by Stock price at the end of t period
INSt= percentage of shares held by Institutional investor
SIZEt= Scale of firm, while SIZE =log (asset of each firm)
TOBINQt= Market value of firm to book value of firm.
DTARt= Debt to asset ratio
ChinaSOXt= Dummy variable; 0 is the pre China SOX period and 1 is the post China SOX period
ESPt+1/Prit*INSt = ESPt+1/Prit. multiplied percentage of shares held by Institutional investor
ESPt+1/Prit*China SOXt= Dummy variable of ChinaSOXt multiplied ESPt+1/Prit.
ESPt+1/Prit*ChinaSOXt*INSt= Dummy variable of ChinaSOXt multiplied ESPt+1/Prit and percentage of
shares held by Institutional investor
ChinaSOXt*INSt= Dummy variable of ChinaSOXt multiplied percentage of sharesheld by
Institutional investor
Descriptive statistics for the pooled samples are in Table 3. Compared to average institutional investors ownership of the research of Uma &David(2006) from 1992–1999 in US, which is 45.4984%, the average institutional investors ownership (19.8%) in China from 2009-2013 is pretty low. It is consistent with the current situation of institutional investors in China, which is just beginning to develop. The average Rt+1 is 1.022, meaning that the average stock price is growing gradually.
32
TABLE 4.Correlation matrix between independent and dependent variables
Rt+1 ESPt+1/Prit INSt CSOXt ESPt+1/Prit
*CSOXt
ESPt+1/Prit*INSt CSOXt * INSt ESPt+1/Prit ChinaSOXt*INSt Rt+1 1.0000 ESPt/Prit 0.0894 <0.0001 1.0000 INSt -0.0187 0.3324 0.1754 < 0.0001 1.0000 CSOXt 0.0699 0.0003 -0.072 0.0002 -0.0600 0.0018 1.0000
ESPt+1/Prit*ChinaSOXt
0.0828 <0.0001 0.4695 <0.0001 0.0895 < 0.0001 0.4710 <0.0001 1.0000
ESPt+1/Prit*INSt
0.0413 0.0319 0.6388 < 0.0001 0.5841 < 0.0001 -0.0960 <0.0001 0.2729 < 0.0001 1.0000 ChinaSOXt* INSt 0.0137 0.4780 0.0480 0.0128 0.5822 < 0.0001 0.5464 <0.0001 0.3917 < 0.0001 0.2498 < 0.0001 1.0000
ESPt+1/Prit*ChinaSOXt
*INSt 0.0512 <0.0001 0.3183 <0.0001 0.4209* < 0.0001 0.3388 <0.0001 0.6892 < 0.0001 0.4991 <0.0001 0.6884 <0.0001 1.0000 Variable definitions:
Rt+1 = Return of the t period. Equal to the stock price at the end of t+1 period
divided by the stock price at the end of t period ESPt+1 = Earnings per share of t period
PRIt = Stock price at the end of t period
ESPt+1/Prit =Earnings per share of t period divided by Stock price at the end of t period
ChinaSOXt = Dummy variable; 0 is the pre China SOX period and 1 is the post China
SOX period
ESPt+1/Prit*INSt = ESPt+1/Prit. multiplied percentage of sharesheld by Institutional investor
ESPt+1/Prit*ChinaSOXt = Dummy variable of ChinaSOXt multiplied ESPt+1/Prit.
ESPt+1/Prit*ChinaSOXt*INSt = Dummy variable of ChinaSOXt multiplied ESPt+1/Prit and percentage of
shares held by Institutional investor
ChinaSOXt*INSt = Dummy variable of ChinaSOXt multiplied percentage of shares held by
33
Table 4 reports the Pearson and Spearman correlations among 8 variables. Institutional investors ownership (INSt) is significantly positively correlated with earnings per share divided by stock price (ESPt+1/Prit). This means that the high institutional investors ownership is linked with the high return rate. Moreover, ESPt+1/Prit*ChinaSOXt, ESPt+1/Prit*INSt, and ESPt+1/Prit*ChinaSOXt*INSt are all positively correlated to Rt+1. These correlations are in line with my hypotheses. However, the correlation between INSt and Rt+1 is negative and not significant.
5. Tests of hypotheses
5.1. Institutional ownership and earnings quality
The results of the three regression model are showed on the table 5- table 7. Table 5.Cross-sectional association between institutional ownership and earnings quality
Variable Coefficient t-statistic
Intercept 1.962239 15.43
ESPt+1/Prit 1.427514*** 6.25
INSt -.0494085 -0.98
ESPt+1/Prit*INSt 1.86516** 2.26
SIZEt -.1267455*** -9.46
DTARt .2539305*** 5.93
TOBINQt .0590424*** 9.35
Adj R-squared 0.0988
Note: ***, **, * denote significance at 0.001, 0.05, and 0.10 levels, respectively, based on t-tests (two-tail)
Model (1): Rt+1 = β + β1ESPt+1/prit + β2ESPt+1/PRIt*INS + β3INSt
+ β4SIZEt + β5TOBINQt + β6DTARt + ε (5.3)
34
Rt+1 = Return of the t period. Equal to the stock price at the end of t+1 period divided by the
stock price at the end of t period
ESPt+1/Prit =Earnings per share of t period divided by Stock price at the end of t period
INS = Percentage of shares held by Institutional investor SIZE = Scale of firm, while size =log (asset of each firm) TOBINQ = Market value of firm to book value of firm. DTAR = Debt to asset ratio
ESPt+1/Prit*INS = ESPt+1/Prit. multiplied by percentage of shares held by Institutional investor
Table 5 reports the results for hypothesis 1, which is the percentage of institutional ownership is positive related to the earnings quality. In order words, I examine whether, as institutional ownership increase, institutional investors have enhance earnings quality. As shown, the adjusted R-Square of this model is 0.0988, which means 9.88% of total variation can be explain by independent variables and control variables. As can be seen from the table, the coefficient for ESPt+1/Prit is positive (1.427514) and is significant at 0.01 level. This implies there is positive relation between ESPt+1/Prit and Rt+1. And as ESPt+1/Prit increase by 1, Rt+1 will increase by 1.427514. This relation shows that the t+1 period‟s earnings are strong related to the t+1 period‟s stock return. Moreover, the coefficient for ESPt+1/Prit*INS is positive (1.86516) and significant at 0.05 level. The positive coefficient supports the Hypothesis 1. This indicates that the institutional ownership positive influence the relation between ESPt+1/Prit and Rt+1. Institutional ownership strengths the impact of ESPt+1/Prit on Rt+1. Under the involvement of institutional investors, as the institutional ownership increases, t+1 period‟s stock return can more be predicted by the t+1 periods earnings, indicating high earnings quality. This means that there is positive relationship between institutional ownership and earnings quality, in consistent with findings of Xia and Gang (2008)and Cheng (2006) in China.
35
5.2. China SOX and earnings quality
Table 6.Cross-sectional association between China SOX and earnings quality
Variable Coefficient t-statistics
Intercept 1.962239 7.30
ESPt+1/Prit 1.506023 *** 0.11
ChinaSOXt .0021609 2.49
ESPt+1/Prit* China SOXt .8812212** -9.69
SIZEt -.1225603 *** 6.15 DTARt .2620103*** 9.44 TOBINQt .059381*** 15.75 Adj R-squared 0.1003
Note:***, **, * denote significance at 0.001, 0.05, and 0.10 levels, respectively, based on t-tests (two-tail)
Model (2): Rt+1 = β + β1ESPt+1/prit + β2ESPt+1/Prit*ChinaSOXt
+ β3ChinaSOXt + β4SIZEt + β5TOBINQt + β6DTARt + ε (5.4)
Variable definitions:
Rt+1 = Return of the t period. Equal to the stock price at the end of t+1 period divided
by the stock price at the end of t period
ESPt+1/Prit = Earnings per share of t period divided by Stock price at the end of t period
SIZE = Scale of firm, while size =log (asset of each firm) TOBINQ = Market value of firm to book value of firm. DTAR = Debt to asset ratio
ChinaSOXt = Dummy variable; 0 is the pre China SOX period and 1 is the post China SOX
period
ESPt+1/Prit*ChinaSOXt = Dummy variable of ChinaSOXt multiplied ESPt+1/Prit.
For the hypotheses 2, table 6 describes the results for the effect of China SOX on the earnings quality. From this table, you can see that the variable ChinaSOXt is not
36
significant and the coefficient of ChinaSOXt is and positive but relatively small. This result might indicate ChinaSOXt has little impact on the stock return. And small coefficient implies that influence of China SOX on stock return is influenced by randomness.
However, the interpretation of the positive and significant coefficient (p-value < 0,05) of the 2-way interaction of ChinaSOXt and ESPt+1/Prit ( variable ESPt+1/Prit* ChinaSOXt) in model 2 indicates that ChinaSOXt strengths the impact of ESPt+1/Prit on Rt+1. Relation between earnings and returns has increased in the post-China SOX period compared to in the pre-China SOX period. China SOX reinforces the earnings informativeness, specifying that China SOX plays a prominent role on the earnings quality. Meanwhile, my prediction of the influence of China SOX is accurate and Hypothesis 2 is supported by the result.
5.3. Institutional ownership and earnings quality in pre-China SOX and Post- China SOX period
Table 7.Cross-sectional association between institutional investor and earnings quality in the pre-China SOX and post-China SOX period
Variable Coefficient t-statistic
Intercept 1.998684 15.45
ESPt+1/Prit 1.300243 *** 4.78
INSt= .0409588 0.57
ChinaSOXt .0369637** 1.45
ESPt+1/Prit*ChinaSOXt .3525425 -2.10
ESPt+1/Prit*INSt 1.018397 0.72
ChinaSOXt*INSt -.2152829 ** 1.06
ESPt+1/Prit*China SOXt*INSt 3.56212** 1.99
SIZEt -.1326052 *** -9.81 DTARt .2523553 *** 5.89
37
TOBINQt .0572586*** 8.99
Adj R-squared 0.1029
Note: ***, **, * denote significance at 0.001, 0.05, and 0.10 levels, respectively, based on t-tests (two-tail)
Model (3): Rt+1 = β + β1ESPt+1/prit + β2ESPt+1/Prit*ChinaSOXt*INS + β3ESPt+1/Prit*INSt
+ β4ESPt+1/Prit*ChinaSOXt + β5ChinaSOXt*INSt + β6ChinaSOXt + β7INSt + β7SIZEt + β8TOBINQt + β9DTARt + ε (5.5)
Variable definitions:
ESPt+1/Prit =Earnings per share of t period divided by Stock price at the end of t period
INSt = Percentage of shares held by Institutional investor
SIZEt = Scale of firm, while size =log (asset of each firm)
TOBINQt = Market value of firm to book value of firm.
DTARt = Debt to asset ratio
ChinaSOXt = Dummy variable; 0 is the pre China SOX period and 1 is the post China
SOX period
ESPt+1/Prit*INSt = ESPt+1/Prit. multiplied percentage of sharesheld by Institutional investor
ESPt+1/Prit*ChinaSOXt = Dummy variable of ChinaSOXt multiplied ESPt+1/Prit.
ESPt+1/Prit*ChinaSOXt*INSt = Dummy variable of ChinaSOXt multiplied ESPt+1/Prit and percentage of
shares held by institutional investor
ChinaSOXt*INSt = Dummy variable of ChinaSOXt multiplied percentage of shares held by
institutional investor
Result of relation between Institutional ownership and earnings quality in pre-China SOX and Post- China SOX period has been show on the table 7. From this table, I can find out that the coefficients for ESPt+1/Prit*ChinaSOXt and ESPt+1/Prit*INSt is not significant. It may be explained by the influence of three of the two-ways interactions and by the three-ways interaction. At least, the coefficients for ESPt+1/Prit*ChinaSOXt and ESPt+1/Prit*INSt are not negative which indicates opposite
38
relation. Nevertheless, the coefficients for ESPt+1/Prit*ChinaSOXt*INSt is 3.56212, which indicates really strong positive relation, and significant at 0.05 level, suggesting China SOX largely positively influences the other 2 –way interaction (ESPt+1/Prit *INSt). This result also supports my hypothesis 3.
To summary this section, from table 5-7, the control variables SIZEt, TOBINQt and DTARt are all significant at 0.01 level. The coefficient of TOBINQt and DTARt are all positive, meaning both the ratio of market value of firm to book value of firm and debt to asset ratio are all have positively related to stock return in the coming year. However, the negative coefficient for variables SIZEt indicates that the larger the firm, the less the stock return. In the hypothesis 1 the coefficient for the ESPt+1/Prit*INStand in
hypothesis 2 the coefficient for ESPt+1/Prit*ChinaSOXt,,these 2 coefficient of two-ways
interaction are both positive and significant, supporting the hypothesis 1 and hypothesis 2 separately. And the positive and significant coefficient for three-ways interaction in hypothesis 3 also supports my hypothesis 3. Overall, the results all support my expectations.
6. Conclusion
“Basic Norms of Internal Control”, which is deemed as China-SOX, was effective in 1st January 2012 for all listed firm in the China stock market, aiming at improving the effectiveness of internal control of firms. Moreover, institutional investors ownership has been increasing over years and the participation of institutional investors on the board of directors in the firm is changing the corporate governance more or less. Do China-SOX and institutional investors influence the earnings quality of firms? It is meaningful to know the result of whether China-SOX and institutional investors play roles on earnings quality as well as how do they impact.
In my research, I collect data to analyze the relation between the institutional investors ownership and earnings quality in the pre-China SOX and post-China SOX.
