The Impact of Corporate Governance on Informative Earnings Management in the Chinese Market

This study investigates the association between a firm's corporate governance practices and managerial informative earnings management (IEM) in the Chinese context. The primary objectives of corporate governance are to monitor the behaviours of different interested parties and to reduce the agency costs underlying various principal–agent relations (Karpoff et al., 1996; Lemmon and Lins, 2003). In particular, corporate governance facilitates a firm's stakeholders to monitor managers’ behaviours and business operations, to safeguard shareholders’ investment, to reduce information asymmetry between managers and other stakeholders, and to ensure the reliability and informativeness of managerial performance reports (e.g., the financial statements) (Shleifer and Vishny, 1997; Denis and McConnell, 2003; Wilkinson and Clements, 2006; Kanagaretnam et al., 2007; Huson et al., 2012). However, managers often engage in earnings management when preparing and presenting financial statements to fulfil specific purposes.

There are two perspectives in earnings management. The opportunistic perspective holds that managers seek to mislead investors through manipulating periodic earnings so as to maximize their own utilities (Burgstahler and Dichev, 1997; Healy and Wahlen, 1999). On the contrary, the information perspective holds that managerial discretion is a means for managers to communicate their own expectations about the firm's future cash flows or profitability (Holthausen and Leftwich, 1983; Guay et al., 1996; Burgstahler et al., 2004). Most previous studies have predicated and drawn their conclusions on the opportunistic perspective for earnings management and have not tested the information perspective (Burgstahler and Dichev, 1997; McNichols, 2000; Baber et al., 2006; Curtis et al., 2011; Jansen et al., 2012). Ronen and Yaari (2008) point out that research is rare on factors that affect the motivations and consequences of the different types of earnings management.

In this study, we identify cases of IEM and examine whether managerial IEM is associated with corporate governance. We test whether firms with good corporate governance are more likely to signal inside information to outside stakeholders through IEM in China. The reasons for this study in the Chinese context are multifold. First, there are few alternative sources of information regarding firm performance (especially the prospect of earnings performance) to the market in China. Unlike in the US and other developed markets, financial analysts from market intermediaries in China (e.g., investment banks, securities firms, and rating agencies) are not as systematic or reliable an information channel compared to their counterparts in developed countries. Although Chinese firms may make management forecast disclosures, which is currently non-mandatory, such a practice would impose additional legal liabilities on the firm and consume additional firm resources (Liang, 2005; Chen et al., 2008). Thus, the motivation for IEM to convey inside information is more substantial in China than in other developed markets where the availability of alternative information sources mitigates managerial incentives for IEM.

Second, there are some unique corporate governance mechanisms in China, such as the split-share structure with tradable and nontradable shares, 1 high ownership concentration with strong government control, the two-board governance structure with the board of directors (BOD) and the supervisory board (SB), cross-listing in domestic and overseas markets under varied market regulations, the introduction of mandatory corporate governance codes in 2002/2003, and the weak investor protection environment. As a result, the actual function of corporate governance in the Chinese market may substantially differ from that in Western countries (Firth et al., 2007, 2012). It is therefore interesting to examine the impact of the governance mechanisms with Chinese characteristics on the quality and informativeness of accounting information (earnings).

Third, all Chinese-listed firms have the same fiscal year-end (i.e., calendar year) and the same fiscal quarter ends consistently, which is not the case in the US and other markets where firms adopt varied fiscal years for financial reporting. As we need to obtain firms’ abnormal accruals for each quarter, it is important to preclude the potential seasonality noise of reporting period end in running the Jones model (i.e., quarterly and yearly performance may be distorted for firms in the same industry with similar operational cycling when they choose varied fiscal period ends). This requirement makes China a practical setting for our study.

Fourth, there are some unique settings in the Chinese stock market that may have a specific impact on earnings management conducted by the listed firms. For instance, the Chinese market regulatory authorities (e.g., China Securities Regulatory Commission (CSRC) and the stock exchanges) have set specific earnings requirements (thresholds) for Chinese firms to pursue initial public offerings (IPOs) or seasonal equity offerings. On the basis of the fulfillment or consequential deviation from the required earnings targets, the listed firms may be subject to special trading restrictions or resumption of normal trading (i.e., special treatment—(ST)) and whether or not to be suspended from trading (or delisting) in the market (i.e., particular transfer—(PT)). 2 Obviously, Chinese firms’ motivation for earnings management (both opportunistic and informative) under such a regulatory environment is different from most other markets, which is worthy of particular empirical analysis.

Fifth, the Chinese government has promoted international accounting convergence and market regulation reforms toward international conventions in recent years. Thus, a study in the Chinese context may generate interesting evidence on the particular incentives and consequences of managers’ IEM and the impact of corporate governance on such managerial behaviour with broad implications.

There are both internal and external governance mechanisms to help bring the interests of managers in line with that of shareholders (Walsh and Seward, 1990). We focus on three specific aspects of Chinese firms’ internal corporate governance mechanisms as measured by: (1) the equity ownership structure (e.g., ownership concentration, blockholding, split-share system with large portion of nontradable shares, and cross-listing status); (2) the characteristics of oversight/monitoring establishments (e.g., the independence of the BOD and the SB); and (3) the activeness (effectiveness) of monitoring functions (e.g., by the BOD and the SB, and shareholder activism). We hypothesize that those internal governance devices or arrangements that promote good corporate governance should induce managers to pursue IEM and vice versa.

This study is motivated by three factors. First, an important function of financial reporting is to provide shareholders with a means of evaluating and monitoring managerial activities and business operations, and good corporate governance reduces information asymmetry between managers and external stakeholders and contributes to more comprehensive and reliable information disclosures while informative earnings can positively affect the usefulness of accounting information (Healy and Palepu, 2001; Fan and Wong, 2002; Bedard and Johnstone, 2004; Cornett et al., 2009; Jansen et al., 2012). A study of managerial IEM with respect to the function of internal governance mechanisms should expand our knowledge of the impact of corporate governance on managerial disclosures and information quality, especially the communication of managers’ private information to external stakeholders. Second, as pointed out by Dechow and Skinner (2000), because managerial discretions are unobservable from the financial statements, most definitions of earnings management have rested on the opportunistic perspective. As a result, the majority of previous research was concluded on the basis of the opportunistic perspective rather than on the informative perspective (Dechow et al., 1995; Holthausen et al., 1995; Teoh et al., 1998; Baber et al., 2006; Lento and Cotter, 2011), which is an incomplete picture in regard to managerial motivations and behaviours for earnings management. Third, although it is recognized in the literature that earnings management can be informative or ‘signalling’ instead of opportunistic in nature, this role has long been understated or neglected because of the difficulty in classifying IEM ex ante. This study intends to develop a practical approach to identifying IEM in the context of the relationship between quarterly managerial discretionary accruals and yearly earnings reported to facilitate further studies on the varied managerial motivations and consequences of earnings management and their impact on the quality of financial reports and informativeness of earnings. Overall, this study aims to contribute to the earnings management literature by analyzing the association of corporate governance with IEM in an effort to fill the gap in the extant literature.

An essential issue in this study is to identify IEM from opportunistic earnings management (OEM), for which there is no explicit benchmark/model to differentiate it in the literature. We use quarterly financial data to run the Jones (1991) model and modified Jones model (Dechow, 1994; Kothari et al., 2005) and estimate discretionary accruals in the quarterly reports disclosed by the Chinese listed firms to identify IEM. 3 When yearly firm performance (e.g., annual earnings) is expected to be lower (higher) than the previous year, managers’ private prediction of current annual earnings can be signalled out through negative (positive) quarterly discretionary accruals in the case of IEM. 4 Assuming that managers have inside information that enables them to predict that the current year's performance is worse than that of last year, if managers would like to convey this bad performance to external stakeholders, they can choose to lower the quarterly earnings (using negative discretionary accruals) so that external stakeholders will not hold an unreasonably optimistic view. On the other hand, managers would use positive discretionary accruals to send out the IEM signal in their quarterly reports if they expect a better yearly performance to come.

Subramanyam (1996) finds that discretionary accruals are useful in predicting future operating cash flows, and accounting earnings are basically more useful than current cash flows in predicting future cash flows and profitability. According to FASB/IASB (2010), an important objective of financial reporting is to provide useful information that will assist financial statement users (e.g., investors, creditors, regulators, etc.) in predicting a firm's future cash flows. We therefore argue that it is reasonable to view the quarterly earnings information as informative if it can help investors predict a firm's annual accounting performance. We thus define the IEM, which is original and simultaneously intuitively understandable. Because there is rare research on the measurement of IEM and whether corporate governance has an impact on IEM, our study should have profound theoretical and practical implications.

We further identify IEM with three different measures in terms of increasing rigorousness in specification. We then apply regression analyses of the relationship between the proxies for IEM and corporate governance mechanisms. Our regression results show that, in general, firms with weak corporate governance are less likely to conduct IEM. The findings indicate that although firms’ managers have incentives to engage in IEM, such managerial behaviour is positively related to the quality of corporate governance.

This study contributes to the literature in several ways. First, we examine earnings management from the informative or signalling perspective. Most previous studies on earnings management have focused on the opportunistic perspective and conclude that earnings management has a negative impact on the quality of financial reporting. In particular, some recent China-related studies (Xie et al., 2003; Firth et al., 2007; Gulzar and Wang, 2011) report that earnings management dampens the informativeness of earnings in the Chinese market, and corporate governance mechanisms should be improved to reduce earnings management and to enhance the informativeness of earnings information. The underlying assumption in these studies is that earnings management is detrimental to earnings informativeness, and it should be curbed or eliminated (optimally be reduced to zero) for the sake of improving financial reporting quality and earnings informativeness. The assumption is, however, superficial because earnings management could contribute positively to earnings informativeness by conveying managerial private information to external investors and the market, or IEM is beneficial and should be encouraged instead of constrained or eliminated to enhance the informativeness of earnings information. Thus, our findings on IEM could go beyond the conclusions of mainstream studies and provide a new insight into the role of earnings management in quality financial reporting and the information relevance of earnings to users’ decision making.

Second, we develop a measurable proxy for IEM by comparing the correlation between the managerial discretionary accruals in quarterly reports and the yearly earnings in annual financial statements. Such a proxy is not only theoretically appealing but also measurable in testing. 5 Although the view of IEM has been recognized in the literature, the empirical analysis of its effect is rare because of a lack of appropriate measurement to differentiate IEM and OEM or because of the difficulty in attributing discretionary accruals to opportunistic or informative prospective ex ante. As a result, most previous research on earnings management ends up with the opportunistic perspective and reveals its negative impact on investors’ decisions (Healy and Wahlen, 1999; Dechow and Skinner, 2000; Ronen and Yaari, 2008; Cohen et al., 2011). However, we find that some managers in our sample firms have in fact conducted IEM through their use of quarterly discretionary accruals. Our IEM proxy should facilitate the study of informative earnings management and enrich the related literature.

Third, our study findings reveal that managers have a motivation to pursue IEM, or they will apply discretionary accruals in quarterly reporting and convey their private information about the firm's future cash flows and earnings potential to the market. Unlike OEM, IEM has a positive impact on the usefulness or predictability of earnings information and can enhance the relevance of financial statements to investors’ decision making. This finding should shed a new light on the motivations and consequences of earnings management, supplementing the prevailing views derived from an opportunistic perspective on earnings management. In particular, previous studies have documented a rampant earnings management phenomenon in the Chinese market because of specific institutional or regulatory requirements (Chen and Yuan, 2004; Firth et al., 2007). However, those studies are mainly based on the opportunistic perspective of earnings management, that is, they assume that managers use judgement in financial reporting or structure transactions to misrepresent the underlying economic performance or to influence contractual outcomes that depend on reported accounting numbers. We do not deny that Chinese firms, compared with their counterparts in most developed markets, have even stronger motivations or incentives to engage in earnings management for opportunistic purposes (such as competition for IPO quotas, seasonal rights offers, maintaining listing status or avoiding special trading treatments or delisting, tunnelling to parent companies through related-party transactions, etc.), but we do find that some Chinese listed firms (27%–47% of our observations in terms of the varied specifications of IEM) have used quarterly discretionary accruals for IEM, which is a significant earnings management motivation, and the consequence cannot simply be ignored. As there is a lack of reliable alternative information about prospective firm performance to the market in China, managers of some Chinese firms have to rely on quarterly IEM to convey private information about future firm performance. Thus, our study findings can supplement other China studies to provide a more complete picture of the motivations and implications of earnings management in the Chinese context.

Fourth, there is a void in the extant literature about the relationship between corporate governance and IEM, although many studies contend that corporate governance can improve the quality and usefulness of accounting information disclosed by managers (Karamanou and Vafeas, 2005; Kalelkar and Nwaeze, 2011). This study is, to our knowledge, the first to directly examine the impact of corporate governance on firms’ IEM. Previous studies report that good corporate governance is associated with less earnings management (Klein, 2002; Xie et al., 2003; Gulzar and Wang, 2011) but fail to differentiate the possible existence of both opportunistic and informative earnings management. Also, Firth et al. (2007) investigate the association between some corporate governance mechanisms of Chinese-listed firms (e.g., ownership and board structure) and informativeness of earnings in terms of the earnings response coefficient and discretionary accruals, and conclude that ‘discretionary accruals are opportunistic then firms with a very dominant shareholder have lower accounting quality’ (p. 487). Although they suggest that good corporate governance can improve the trustworthiness of earnings information, their contention is confined to the traditional assumption of OEM that the use of discretionary accruals is opportunistic or negative to the informativeness of earnings. However, we posit and find evidence that good corporate governance practices actually motivate or facilitate managers to engage in IEM by communicating inside information about the earnings prospect, which can assist external investors in assessing the amount and timing of future cash flows and enhance the effectiveness of their evaluation of management performance and related decisions. Our findings extend beyond the extant studies and demonstrate that good corporate governance can facilitate managerial IEM and enhance the predictability of accounting information, which further testifies to the value of corporate governance. The positive relation between corporate governance and IEM revealed by this study should have important policy implications for market regulators, standard setters, and investors.

This is a study in the Chinese context in respect of the association between corporate governance and IEM, but the underlying principles and research design could have general applicability. Although the Chinese stock market regulation has some unique characteristics, managerial motivations or behaviours for IEM should also exist in other markets, which has been recognized or hypothesized in the literature. Thus, our study based on Chinese evidence should not only help readers understand the development of financial reporting and corporate governance in the largest emerging capital market in the world but also facilitate other researchers to carry out studies of IEM under different market settings.

Literature Review

Hypotheses Development

Financial statements are a crucial part of the information pipeline that facilitates the execution of various contracts and internal monitoring and control. Sound corporate governance practices reduce information asymmetry by prompting managers to provide investors and outsiders with reliable and complete information, thereby increasing corporate transparency and reducing monitoring costs in contract execution. Thus, the quality of information available to investors is influenced by the way a firm is governed (Healy and Palepu, 2001). We posit that corporate governance should induce or constrain managers’ motivations and behaviours regarding earnings management. In light of the positive impact of corporate governance on reducing information asymmetry and enhancing financial reporting quality, we establish our overall proposition as follows:

Because a firm's internal corporate governance mechanisms consist of various oversight or monitoring arrangements or devices (Denis and McConnell, 2003), we further specify the overall proposition with three hypotheses in respect of three major characteristics of corporate governance under the existing regulatory environment in China: (1) the equity capital structure arrangement (i.e., ownership concentration, dominance of nontradable shares, blockholding, and cross-listing status) 8; (2) the independence of the oversight/monitoring establishments (i.e., the duality of BOD chairman and CEO positions, the proportion of independent directors, and the SB size); and (3) the activeness (effectiveness) of oversight/monitoring by the BOD, the SB, and the shareholders. We offer the following hypothesis in respect of equity structure arrangement:

In countries with a relatively weak corporate governance environment such as China, ownership concentration is frequently used to moderate the lack of investor protection. Concentrated ownership structure leads to two obvious consequences for corporate governance, as surveyed in Morck et al. (2005). On one hand, dominant shareholders have both the incentive and the power to discipline management. On the other hand, because the interests of controlling and minority shareholders are not fully aligned, concentrated ownership will increase the entrenchment of large shareholders and create a new (Type II) agency problem (La Porta et al., 1999; Mayhew et al., 2003). In China, firm ownership is highly concentrated to state agents, and the legal protection of minority shareholders is insufficient (Firth et al., 2007; Ding et al., 2010); thus, ownership concentration may lead to severe entrenchment problems and information opacity.

A Chinese listed firm usually has a large controlling shareholder, often the government or a parent state-owned enterprise (SOE). Johnson et al. (2000) argue that more narrowly held firms may face greater agency conflicts because the controlling shareholders have a dominant influence on corporate affairs and they can easily bypass monitoring by other shareholders. In China, controlling shareholders frequently engage in benefit transfers through the misappropriation of funds and party-related transactions to expropriate the interests of minority shareholders (Lin et al., 2007). Therefore, listed firms with high ownership concentration are more likely to engage in OEM for the sake of exploiting minority shareholders (Firth et al., 2007), so they are less likely to convey inside information to the public. In other words, ceteris paribus, the more shares held by the largest shareholder, the less likely there will be managerial IEM. Therefore, we further specify the first sub-hypothesis as follows:

The finance literature contends that equity blockholders and institutional investors tend to promote shareholder-driven corporate strategies, thus increasing monitoring over large shareholders and managers and improving corporate governance. Several empirical studies have concluded that both large equity blockholders and institutional investors can improve firm performance, suggesting that there is a positive association between large noncontrolling shareholders and firm value (Holderness and Sheehan, 1985; Mikkelson and Ruback, 1985; Lemmon and Lins, 2003). This is because the large noncontrolling shareholders (blockholders) are, out of self-interest, willing and able to monitor the controlling shareholder, preventing him or her from taking too much (Pagano and Roel, 1998; La Porta et al., 1999; Claessens et al., 2002). This is particularly true for Chinese listed firms, as equity blockholders—other than the largest shareholder (e.g., a government agency or parent SOE)—hold relatively fewer shares with a vulnerable position and have a great incentive to stay alert regarding potential exploitation by the largest shareholder (Wei et al., 2005). We expect that, ceteris paribus, more shares held by blockholders (e.g., the second to the fifth largest shareholders) can improve corporate governance and can prompt firm managers to pursue IEM:

Before China's split-share structure reform (‘circulation of all shares’) in 2005, domestic A-shares were divided into nontradable and tradable shares. Nontradable shares represent the interests of governments and related parent entities and have a dominant influence on a firm's operating decisions, whereas holders of tradable shares can exercise little influence on the decisions made by the nontradable shareholders and management. This is a typical structure with severe agency conflicts (Ding et al., 2010). The split-share structure reform introduced by the government in 2005 requires that nontradable shareholders have to bargain with tradable shareholders to gain liquidity. 9 Kwan (2006) explains that the main aim of the reform is to improve corporate governance in listed firms and to enhance the intermediary function of the stock market by giving equal rights to tradable and nontradable shareholders. Firth et al. (2007) find that Chinese firms publish higher quality accounting information when they have more tradable shares. Therefore, we argue that, with tradable shares, information asymmetry decreases while earnings informativeness increases, and, ceteris paribus, a firm with a higher percentage of nontradable shares will have poor corporate governance and information disclosure environment and is less likely to engage in IEM:

Some Chinese companies choose to list their shares simultaneously in the domestic market and overseas stock exchanges such as Hong Kong, London, New York, and Singapore. The most frequently cited reasons or benefits for a firm to cross-list in foreign markets are improved access to capital, greater liquidity, lower capital costs, heightened corporate prestige, and better investor protection (Karolyi, 1998). Bushman et al. (2004) document that financial transparency is higher in countries with higher levels of judicial efficiency and a common law background, or in countries where stock markets are more developed and better regulated. Thus, cross-listing in different stock markets, especially in the more developed capital markets, will expose listed firms to more stringent or transparent listing rules and disclosure requirements, which should improve the firm's corporate governance and information quality (Karolyi, 1998; Lang et al., 2003b). Cabán–García (2009) compares the quality of the reported earnings of cross-listed European firms with that of a control group of non-cross-listed domestic firms and finds that the variance in earnings quality is positively associated with the difference in the strictness of regulatory/disclosure requirements between foreign and domestic exchanges. Therefore, cross-listing can also act as an ownership structure device to carry out governance monitoring over accounting and reporting for the Chinese listed firms. We posit that, ceteris paribus, a firm with a higher percentage of shares cross-listed in overseas stock exchanges is more likely to engage in IEM.

We present our second hypothesis in light of the independence of monitoring/oversight governance establishment:

The BOD is a vital device for overseeing management and business operations as it is responsible for appointing, removing, and remunerating senior managers to ensure that managers act in the best interests of the shareholders. The independence of the BOD assures its effective oversight (Denis and McConnell, 2003). However, there is increasing criticism that the BOD's oversight function is weakened if a BOD chairman is the same person as the CEO (Shivdasani and Yermack, 1999) because this practice results in low transparency of the CEO's activities, and his or her actions can go unmonitored, which paves the way for scandals and corruption. 10 On the other hand, the separation of the two positions allows the BOD chairman, on behalf of stockholders, to be more impartial in overseeing the CEO's work and overall management performance (La Porta et al., 2002; Yeh and Woidtke, 2005). Splitting the two positions should therefore enhance the independence of the BOD and its oversight function, which will not only improve corporate governance but also lead to better communication of inside information to the BOD and shareholders at large. Several studies have confirmed this proposition with convincing empirical evidence (Chang and Sun, 2009). Therefore, we hypothesize that separating the positions of BOD chairman and CEO enhances the chance for managerial IEM in China:

Most stock markets require the appointment of independent directors as a means of strengthening the BOD function (Denis and McConnell, 2003; Gordon, 2007). Independent directors can deter CEO entrenchment and enhance the BOD's oversight regarding the CEO and other senior managers and their operating decisions. Independent directors can also be more readily mobilized by legal standards to help provide more accurate disclosures, enhancing the informativeness of the financial statements and improving compliance with laws (Sale, 2006). In addition, independent directors are less wedded to inside accounts of the firm's prospects but more sensitive to external assessments of their performance. Weisbach's (1988) widely cited study of CEO terminations in New York Stock Exhange listed firms over the 1974–1983 period concludes that ‘when boards are dominated by outside directors, CEO turnover is more sensitive to firm performance than it is in firms with insider-dominated boards’ (pp. 33–34). Uzun et al. (2004) show a negative relationship between business frauds and board independence. Firth et al. (2007) report that Chinese listed firms with more independent directors sitting on the BOD have greater earnings informativeness and are more likely to receive clean audit opinions. Thus, we propose that, ceteris paribus, a firm with a large proportion of independent directors in the BOD is more likely to engage in IEM.

Chinese firms adopt a two-tier board structure, that is, the BOD and the SB under company law. The SB is composed of the shareholders’ representatives and an appropriate proportion of employee representatives who are nominated by the firm's employee union. The Standard Code of Corporate Governance for Listed Companies in China, issued by the CSRC and the State Economic and Trade Commission in 2002 further requires that the SB members should have some professional knowledge or work experience in the fields of law and accounting. As per company law, the SB has a clear monitoring mandate to carry out a series of responsibilities, including monitoring the performance of directors and senior managers for compliance with the laws and articles of incorporation, reviewing the financial affairs of the firm, and submitting an SB report to shareholders’ annual general meeting (Lin and Chen, 2005).

Under the two-tier board system, the BOD and the SB play a supplementary oversight role in regard to internal governance. Normally, expertise and monitoring capability increase with the size of the SB, that is, a large SB should have more members equipped with accounting and law knowledge and carry out a more effective SB monitoring function (Xiao et al., 2004; Firth et al., 2007; Ding et al., 2010). Lin and Liu (2009) empirically find that firms with smaller SBs are less likely to hire a Top 10 (high-quality) auditor in China, and their financial statements are of relatively low quality. Previous studies (Krishnan, 2003) indicate that lower quality auditors are less likely to deter and detect questionable accounting practices and irregularities in firms, which provides a greater chance for opportunistic reporting. We therefore posit that SB size has a bearing on governance monitoring and on managerial engagement in earnings management. Ceteris paribus, a firm with a large SB is more likely to engage in IEM:

Our last hypothesis on the association between corporate governance and IEM concerns the activeness (effectiveness) of the governance oversight or monitoring function:

The effectiveness of internal governance mechanisms relies mainly on their actual implementation/operation. Previous research shows a negative relationship between the level of board activeness and the likelihood of a firm's opportunistic earnings restatements (O'Connor et al., 2006). Vafeas (1999) reports that the activity level of directors is positively associated with firm performance and suggests that the number of board meetings per year may indicate the activity level of the directors. It is generally recognized that high activity leads to better BOD oversight on accounting and reporting matters and may thus enhance the informativeness of earnings (Vafeas, 1999; Karamanou and Vafeas, 2005). Therefore, we posit that, ceteris paribus, a more active BOD increases the chance of managerial IEM. Similarly, Firth et al. (2007) report a direct relationship between the activeness of the SB and the informativeness of earnings data. Their empirical results indicate that a larger and more active SB can improve the earnings–returns association and financial reporting quality. Thus, we expect that an active SB is able to play an effective governance role and to improve the firm's information disclosure. Ceteris paribus, a more active SB more likely leads to managerial IEM:

Along with increasing attention to corporate governance, the active role of shareholders in monitoring and controlling the BOD and the management is becoming more important (Choi and Cho, 2003). Shareholder activism has emerged as a significant factor in corporate governance. Proponents of shareholder activism argue that firms with active and engaged shareholders are more likely to succeed in the long term (La Porta et al., 2002). Vigilant shareholders act as the arbitrators in internal governance matters. Shareholders with significant ownership positions have both the incentive to monitor the executives and the influence to bring about changes they feel are beneficial (Holderness and Sheehan, 1985; Claessens et al., 2002). The active participation of shareholders in the annual general meeting will put great pressure on the BOD and the managers, which can strengthen internal governance, reduce information asymmetry, and improve financial reporting quality. For this study, we use the registered shares attending annual general meetings over total shares issued to measure shareholder activism and assume that, ceteris paribus, a firm with more active shareholders is more likely to engage in IEM:

Research Methodology

Because accounting accruals are commonly used by managers to pursue earnings management (Dechow and Dichev, 2002; Gonchrov, 2005), we apply the Jones (1991) model to separate normal and abnormal accruals in the main test, which is expressed as follows:

where ACCR is the total accrual, ∆REV is the change in revenues, and PPE is the gross amount of property, plant, and equipment, all deflated by total assets. In this model, the fitted values are taken as normal accruals and the residuals as abnormal accruals. We estimate the Jones model cross-sectionally for each combination of a reporting quarter and a two-digit SIC code with a minimum of 10 observations based on the population in the China Securities Market and Accounting Research (CSMAR) database. Normal accruals are estimated as follows:

where aˆ₁, aˆ₂, and aˆ₃ are the coefficient estimates from the Jones model, and abnormal accruals are determined as follows:

To identify IEM, we firstly use quarterly data for the first three quarters in a year to run the Jones model and to estimate discretionary accruals for the quarterly reports of all sample firms. By definition, managers should use negative (positive) quarterly discretionary accruals to signal poorer (better) yearly performance compared with the previous year in regard to the perspective of IEM. For example, if managers predict that the current year's earnings performance is worse than the previous year's and would like to signal a bad yearly performance to external investors, they would choose to lower the quarterly earnings (by using negative discretionary accruals) so that external investors will not expect a good year ahead. On the contrary, if managers predict better performance for the current year, they should signal to the external investors by using positive discretionary accruals in quarterly reporting. As shown in Figure 1, there are four possible combinations for yearly earnings (YE) and quarterly discretionary accruals (QDA). In the shaded areas, QDA(+) and QDA(−) are useful for investors to predict YE changes (i.e., YE(+) and YE(−)) before the end of the fiscal year, and these cases are identified as IEM. 11 However, in the areas with white background, QDA is more likely to lead investors to predict a wrong sign (inconsistent changing direction) of YE changes and hence is identified as non-IEM.

After identifying the observations with IEM, we then examine the relationship between the IEM and the corporate governance proxies for the first, second, and third reporting quarter, respectively, for each sample firm over the test period. We construct the following model to test our hypotheses using the following logit regression:

Variables are defined in Table 1. In order to control for the mechanical relationships between discretionary accruals and earnings that might interfere with the identification of managerial behaviour (opportunistic vs. informative), we apply three specifications to identify IEM. As indicated, IEM₁ is defined as the use of discretionary accruals in quarterly reporting to foretell yearly earnings (consistency in the sign of QDA and YE). IEM₂ has an additional condition, as the magnitude of quarterly discretionary accruals is larger than that in yearly discretionary accruals to mitigate the potential managerial manipulation of quarterly accruals for OEM, which addresses the issue that discretionary accruals may not be truly ‘discretionary’, as pointed out in previous studies (e.g., Siregar and Utama, 2008). IEM₃ is specified as the use of quarterly accruals to foretell net-of-discretionary-accruals yearly earnings to moderate the potential distortion effect of discretionary accruals in the yearly earnings per se. As quarterly discretionary accruals are included in the annual earnings and may lead to a mechanical relationship, IEM₃ can help tackle such a concern. 12 For corporate governance variables, we assume that the equity capital structure of a listed firm, including ownership concentration (shares held by the largest owner (LSH) and by other blockholders, the second to the fifth largest shareholders (LSH_2–5)), percentage of nontradable shares (NONTRA), and cross-listing shares (CROLIS), will have an impact on the strength of corporate governance monitoring, which will either dampen (e.g., LSH and NONTRA) or stimulate (e.g., LSH_2–5 and CROSLIS) managerial motivations for IEM. For the characteristics of oversight or monitoring establishment, a higher proportion of independent directors (INDDIR) and a large SB (SUPER) will enhance the independence of oversight/monitoring establishment, whereas the duality of the CEO and the BOD chairman (CEOCHR) will hamper the BOD's oversight role. Nonetheless, the activeness (measured by the number of meetings held each year) of the BOD (BODMEET) and the SB (SBMEET) as well as shareholder activism (percentage of registered shares participating in annual general meetings (SAGM)) should strengthen the effectiveness of internal oversight/monitoring over management. Intuitively, variables that contribute to the effective function of corporate governance mechanisms should induce managers to pursue IEM, whereas those representing poor corporate governance should expand managerial incentives for opportunistic rather than informative earnings management. In summary, those governance mechanisms/arrangements that strengthen (weaken) corporate governance functions should positively (negatively) contribute to the IEM conducted by management.

In addition, firm size (LNASSET), profitability (ROA), and leverage (LEV) are controlled, as some previous studies report that they are influential factors to the quality of corporate governance and financial reporting (Healy and Palepu, 2001; Huson et al., 2012). We also control for managerial shareholding (MSH), 13 seasonal equity offering (SEO), audit quality (BigN), and the listing status of special trading treatments (ST/PT), as these variables may directly influence management incentives or the capability for earnings management (Mayhew et al., 2003; Kanagaretnam et al., 2007; Chen et al., 2008; Firth et al., 2012). Because earnings management may be affected by a change in real economic activity, we also include the regional gross domestic product (GDP) as a control (Cahan et al., 2008). In addition, industrial and yearly dummies are included in the regressions.

Empirical Results

Concluding Remarks

This paper investigates the relationship between corporate governance and IEM for Chinese listed firms. We use three measures to specify managerial IEM and define corporate governance according to three groups of variables in terms of internal governance mechanisms, that is, the equity capital structure arrangement, the independence of the oversight/monitoring establishment, and the activeness (effectiveness) of internal monitoring functions. We outline an overall proposition on the positive relationship between corporate governance and IEM and develop three hypotheses with respect to the main characteristics of corporate governance in China for empirical tests. The hypotheses are supported by empirical results, indicating that better corporate governance, which can reduce agency problems and information asymmetry between firm managers and external stakeholders, prompts managers to pursue IEM (by using quarterly earnings to signal yearly earnings changes), and therefore leads to more transparent financial reporting and better informativeness of reported earnings.

This study contributes to the literature on earnings management research. First, although previous studies have examined the impact of earnings management on a firm's financial transparency and reporting quality, most of these studies do not differentiate the two types of earnings management driven by different managerial motivations or behaviours, namely, opportunistic and informative earnings management. We distinguish IEM from OEM in this study and posit that IEM is actually beneficial to external investors who cannot have direct access to private managerial information regarding the underlying fundamentals of the firm. We specifically construct three measures to proxy for, which are not only innovative but also measurable and testable. Our findings indicate that in earnings management conducted by Chinese firms, on average, approximately 27% to 47% (based on varied IEM specifications) are informative rather than opportunistic, sharply contrary to the general view of earnings management being opportunistic and detrimental to the quality of financial reporting. Our results provide convincing evidence that managers may engage in IEM, which is beneficial to external stakeholders. The study findings shed new light on earnings management research and will possibly stimulate more studies regarding IEM in the future.

We particularly investigate the firm's IEM in regard to corporate governance mechanisms. Although previous research has examined the positive impact of corporate governance on the quality of accounting information, there is, to our knowledge, no study that directly tests the association between corporate governance and IEM in the extant literature. By regressing IEM to a series of corporate governance mechanisms that are widely viewed as having a role to play in overseeing or monitoring a firm's accounting and financial reporting processes, we find that governance mechanisms that promote good corporate governance can actually motivate managers to pursue IEM, whereas those that represent poor governance practices will discourage managers from conveying private information about the future prospect of earnings or cash flows to external investors. This study, about the association of corporate governance functions and managerial reporting incentives, should lead to more research in this underexplored area.

This study will also have important policy implications. Market regulators (or standard setters) and investors need to be aware of the different types of earnings management. Although OEM tends to mislead investors, IEM will, in fact, provide useful information through signalling out managers’ private information about the firm's future cash flows and earnings potential. Rational investors should refer to different types of earnings management behaviours and adjust their decisions accordingly. To bolster the confidence of market participants, regulatory bodies should promote IEM and constrain OEM. Cahan et al. (2008) provide evidence that country-level investor protection helps improve earnings informativeness. Our study findings imply that firm-specific corporate governance mechanisms can enhance financial reporting quality and accounting information relevance to external users through the positive impact of corporate governance on managerial motivations for IEM.

There are some limitations in this study. In particular, some corporate governance-related issues are unaddressed, such as the actual composition and expertise of the BOD and the SB, the involvement of the government in the BOD and the SB, the interaction of the BOD and the SB, the protection of minority interests (although it is relatively weak in the Chinese market at present), and the potential market reaction toward managerial IEM. As better corporate governance is associated with lower information asymmetry (Chung et al., 2010), it will also be interesting to examine the association between information asymmetry and IEM and how corporate governance mediates such an association. In addition, we did not examine the correlation between IEM and alternative sources of disclosures (i.e., management forecasts and analyst coverage) because of the unavailability of data in China. Future studies on these issues will help us form a more solid understanding of the impact of corporate governance on informative earnings management.