مطالعه تجربی از آشکار سازیهای داوطلبانه قیمت گذاری انتقال در چین

This paper empirically investigates the factors that affect the management’s voluntary disclosures of the transfer pricing details of related-party transactions. Using Chinese data from 2004 and 2005, we hypothesize and find that firms that make voluntary disclosures of the pricing methods of related-party transactions are negatively associated with (i) a higher level of earnings management (as captured by abnormal related-party transactions) and (ii) its underlying incentives (as captured by the management’s performance-linked bonuses and the firm’s incentives to achieve earnings targets); further, they are positively associated with (i) a higher percentage of independent directors and (ii) a higher percentage of government ownership. Overall, our findings suggest that earnings management and its incentives, board composition, and ownership structure significantly influence the voluntary disclosure decisions of managers.

Previous studies suggest that managers voluntarily disclose information to reduce information asymmetry and to minimize investor-driven adverse selection effects (Wagenhofer, 1990 and Lundholm and Van Winkle, 2006). Healy et al. (1999) find that an increase in disclosure ratings is associated with an increase in firms’ stock returns, institutional ownership, analyst forecasts, and stock liquidity. Cheung et al. (2010) reveal a positive relation between firms’ voluntary disclosures and their market valuation. Other studies show that greater levels of disclosure are associated with a lower cost of equity and debt capital (Botosan, 1997, Sengupta, 1998, Richardson and Welker, 2001 and Francis et al., 2005). Despite these many benefits, in practice, managers may elect to not voluntarily disclose information because of the costs of disclosure (Verrecchia, 1983, Graham et al., 2005 and Suijs, 2005).1 Therefore, the identification of factors that affect the voluntary disclosure decision is an important research topic that has implications for policy-makers, auditors, investors, and other stakeholders. Voluntary disclosures examined by other studies include management earnings forecasts (Nagar et al., 2003, Ajinkya et al., 2005 and Karamanou and Vafeas, 2005), personnel announcements (Miller, 2002), and environmental disclosures (De Villiers and Van Staden, 2006 and Clarkson et al., 2008), among others. One neglected area of research concerns the disclosure of related-party transactions. Previous studies provide empirical evidence that related-party transactions are widely used to shift income and manage earnings for financial reporting and tax purposes (Harris, 1993, Jacob, 1996, Wang et al., 2008, Aharony et al., 2010, Jian and Wong, 2010, Lo et al., 2010a and Lo et al., 2010b). However, to the best of our knowledge, no preceding study has empirically examined the economic determinants of related-party transaction disclosures. We contribute to the current literature by examining the voluntary disclosure of transfer pricing methods used for related-party transactions in China. The Accounting Standard for Business Enterprises (MOF, 1997) specifies that Chinese-listed companies are mandated to disclose various details of their related enterprises as well as the types and amounts of all their related-party transactions in the notes to financial statements. However, it does not explicitly require listed companies to disclose the transfer pricing policies and methods of their related-party transactions. Therefore, the disclosure of transfer pricing methods is a voluntary management decision in China; managers can choose to disclose or not disclose transfer pricing methods in the notes to financial statements.2 This offers us an excellent setting to investigate managers’ decisions on transfer pricing disclosures. In particular, we aim to examine whether the voluntary disclosure of transfer pricing policies is associated with earnings management incentives and governance structure. We argue that independent directors and institutional shareholders can help to monitor the management’s decisions and lead to more voluntary disclosures of transfer pricing methods. Conversely, management is less likely to voluntarily disclose transfer pricing information if it manages earnings by manipulating transfer prices of related-party transactions and anticipates negative consequences (costs) of disclosing such information to stakeholders. The findings support our hypotheses and suggest that firms with a higher percentage of independent directors are more likely to voluntarily disclose the pricing details of their related-party transactions. Moreover, we find that firms with a higher percentage of government ownership are more likely to voluntarily disclose transfer pricing methods. This finding is consistent with the view that government-owned companies are more likely to make voluntary disclosures to reduce adverse selection problems. However, we find no significant association between institutional ownership and voluntarily disclosures of transfer pricing methods. Regarding the earnings management incentives, we find that firms with a higher level of earnings management resulting from related-party transactions (measured by abnormal related-party transactions) are less likely to voluntarily disclose transfer pricing methods. Our empirical results further suggest that managers who boost earnings to increase their bonuses or to meet earnings thresholds are less likely to engage in voluntary disclosures of transfer pricing methods. These results are robust in controlling for the type of goods being traded, tax effects, audit quality, and various firm characteristics (e.g., profitability, leverage, volume of related-party sales, capital needs, size, and growth) as well as the industry and year effects. In addition to investigating voluntary disclosure decisions, we conduct supplementary analyses to test the firms that are more likely to be disciplined by the Shanghai Stock Exchange or the China Securities Regulatory Commission (CSRC) for noncompliance with applicable mandatory disclosure requirements for related-party transactions (RPT sanctions). The results of our supplementary analyses show that firms with a higher percentage of independent directors, institutional investors, or government ownership are less likely to be sanctioned for noncompliance with mandatory disclosure requirements. However, we find no association between earnings management incentives and RPT sanctions. These findings imply that the decision to not comply with mandatory disclosure requirements is different from the decision to make additional voluntary disclosures; therefore, the same variables may not necessarily determine both decisions.3 This paper contributes to the existing literature by addressing an unexplored research issue: the voluntary disclosure of transfer pricing methods. The transfer pricing manipulation of companies can have far-reaching implications for the economies and societies in which they operate. In particular, outward income shifting by companies adversely affects government plans and policies with respect to tax revenue collection, foreign exchange, and market development (Lo et al., 2010a). Understanding managers’ motivations regarding the voluntary disclosure of transfer pricing methods should help policy makers to obtain more accurate information for improving economic, social, and financial policies. We believe that the findings of this paper may be generalizable to other Asian and developed countries. OECD (2009) suggests that the abusive manipulations of related-party transactions are one of the most significant corporate governance challenges faced by enterprises in China and other Asian countries, because many Asian enterprises are family-owned or state-owned. Given the similar institutional environments across the region, we believe that our findings are generalizable beyond the Chinese context.4 In addition, the empirical results of this study indicate that earnings management and governance structures influence management’s voluntary disclosure decisions and a firm’s risk of being sanctioned. The findings on RPT sanctions also provide the first form of evidence on regulatory disciplinary actions for noncompliance with mandatory disclosure regulations in China. Our data may help auditors to better evaluate risk when auditing related-party transactions and may provide insights for policy-makers seeking to prevent future financial scandals. Our findings are particularly important in countries where the monitoring and auditing processes for related-party transactions are extremely difficult (Levine et al., 1997 and AICPA, 2001) or where the presence of high-quality auditors may not effectively prevent or alleviate transfer-pricing manipulations (Lo et al., 2010a). The remainder of this paper is organized as follows. The next section discusses the institutional background. The third section formulates our research hypotheses, while the fourth section discusses the research methodology. The fifth section explains our results, and the sixth section presents our supplementary analyses on RPT sanctions. We provide concluding remarks in the final section. 2. Institutional background 2.1. Reform of state-owned enterprises and related-party transactions in China Because of its unique institutional environment, Chinese listed firms are likely to have significant related-party transactions with their unlisted parent enterprises and with other related companies. Since the 1990s, state-owned enterprises (SOEs) in China have been experiencing reform aimed at improving their efficiency and profitability. As part of this reform, profitable units of SOEs are separated from the original enterprises and then listed on the stock exchange (Green, 2003 and Chow, 2007). These SOEs, which are wholly (100%) owned by agencies of the state or the local municipality, retain the unprofitable units of the business as well as engage in social activities (e.g., hospitals and schools). Even after separation, the SOEs usually retain significant ownership in the listed spin-off firms and exert substantial influence over them (Ding et al., 2007). One of the main influences is on the decision regarding related-party transactions (Lo et al., 2010b). Related-party transactions are a normal part of business, and many firms report a high volume of such transactions without committing accounting and financial fraud (Gordon et al., 2007). However, under certain conditions, related-party transactions can allow controlling shareholders or executives of a company to earn personal benefits and to expropriate wealth away from the minority shareholders of the listed firm5 (OECD, 2009). For example, Jian and Wong (2010) find evidence that Chinese-listed companies use related-party sales to achieve certain earnings targets. Lo et al. (2010b) find that firms with higher levels of government ownership are more likely to shift profits to their parent SOEs via transfer pricing. This transfer of profits to the parent SOEs allows them to subsidize their loss-making operations. 2.2. Disclosure requirements for related-party transactions in China Because of the significant volume of related-party transactions and their potential implications for profits and cash flows, adequate disclosure of related-party transactions and governance in approving such transactions are important for minority investors. In China, the disclosure requirements for related-party transactions are generally governed by the Accounting Standards for Business Enterprises (MOF, 1997) and the Shanghai Stock Exchange Listing Rules (Shanghai Stock Exchange, 2004). According to the Accounting Standards for Business Enterprises (MOF, 1997), listed companies must disclose information on any related enterprises in the notes to financial statements, including (i) the name, (ii) type of business entity, (iii) legal representative, (iv) place of registration, (iv) amount of registered capital and changes therein, (v) principal business, and (vi) proportion of shareholding and changes therein. The regulations require the disclosure of such information even though there may be no transactions between the listed companies and their related enterprises in a given reporting year. In cases where related-party transactions have occurred, the listed companies should disclose the nature of the relationships, types of transactions, and essential elements of the transactions (usually the amounts of related-party transactions) in the notes to financial statements. In addition to making regular disclosures of related-party transactions in the notes to financial statements, the Shanghai Stock Exchange Listing Rules require listed companies to follow specific procedures in approving related-party transactions and to provide timely disclosures of significant related-party transactions (Shanghai Stock Exchange, 2004). For example, if a director is involved in a related-party transaction or has controlling power over the transaction, s/he is prohibited from voting to approve or disapprove that transaction.6 If the related-party transaction is material (i.e., if the amount of a related-party transaction exceeds RMB 3 million and accounts for more than 0.5% of the absolute value of the listed company’s latest audited net assets), the listed company must announce to the public the details of the related-party transaction. The listed company must also submit relevant information to the stock exchange. Examples of relevant information are draft announcements, agreements or letters of intent relating to the transaction, board resolutions, competence reports, independent directors’ prior written approval for the transaction, independent directors’ opinions, and other related documents. It is important to note that there is no regulation requiring listed companies to disclose the transfer pricing policies and methods for related-party transactions in the notes to financial statements. During our sample period (from 2004 to 2005), the disclosures of pricing policies were conducted on a voluntary basis. Thus, the regulations encourage but do not require listed companies to disclose pricing policies and related details. This offers an excellent setting for investigating managers’ decisions on transfer pricing disclosures before such a disclosure policy becomes mandatory.7

Table 1 lists descriptive statistics on all 996 firm-year observations. These results show that 67.8% of our sample firms voluntarily disclose details on the transfer pricing methods adopted in their related-party transactions. The mean abnormal related-party transactions value (ABN_RPT) is 0.007. Table 1 also indicates that 20.7% of the sample firms have high abnormal related-party transactions and use a management remuneration package that contains performance-linked bonus elements (HRPT_BONUS) and 18.3 percent of the sample firms have high abnormal related-party transactions and strong incentives to boost profits upward to issue new shares or to avoid losses (HRPT_ROE). The mean percentage of independent directors on the board (IND_DIR) is 34.5%, the mean percentage of shares owned by institutional shareholders (INSTITUTE) is 9.9%, and the mean percentage of shares owned by the Chinese government (GOV) is 34.3%.14 The mean values for debt-equity ratio, market-to-book equity ratio, return on equity, volume of related-party sales, and firms’ total assets are 1.230, 2.254, 0.050, RMB 602 million and RMB 3.526 billion, respectively.Journal of Accounting and Public Policy Volume 30, Issue 6, November–December 2011, Pages 607–628 Cover image An empirical study of voluntary transfer pricing disclosures in China * Agnes W.Y. Loa, Corresponding author contact information, E-mail the corresponding author, * Raymond M.K. Wongb * a Department of Accountancy, Lingnan University, Tuen Mun, Hong Kong * b Department of Accountancy, City University of Hong Kong, Kowloon Tong, Hong Kong Available online 24 August 2011 * http://dx.doi.org/10.1016/j.jaccpubpol.2011.08.005, How to Cite or Link Using DOI * Permissions & Reprints Abstract This paper empirically investigates the factors that affect the management’s voluntary disclosures of the transfer pricing details of related-party transactions. Using Chinese data from 2004 and 2005, we hypothesize and find that firms that make voluntary disclosures of the pricing methods of related-party transactions are negatively associated with (i) a higher level of earnings management (as captured by abnormal related-party transactions) and (ii) its underlying incentives (as captured by the management’s performance-linked bonuses and the firm’s incentives to achieve earnings targets); further, they are positively associated with (i) a higher percentage of independent directors and (ii) a higher percentage of government ownership. Overall, our findings suggest that earnings management and its incentives, board composition, and ownership structure significantly influence the voluntary disclosure decisions of managers. 1. Introduction Previous studies suggest that managers voluntarily disclose information to reduce information asymmetry and to minimize investor-driven adverse selection effects (Wagenhofer, 1990 and Lundholm and Van Winkle, 2006). Healy et al. (1999) find that an increase in disclosure ratings is associated with an increase in firms’ stock returns, institutional ownership, analyst forecasts, and stock liquidity. Cheung et al. (2010) reveal a positive relation between firms’ voluntary disclosures and their market valuation. Other studies show that greater levels of disclosure are associated with a lower cost of equity and debt capital (Botosan, 1997, Sengupta, 1998, Richardson and Welker, 2001 and Francis et al., 2005). Despite these many benefits, in practice, managers may elect to not voluntarily disclose information because of the costs of disclosure (Verrecchia, 1983, Graham et al., 2005 and Suijs, 2005).1 Therefore, the identification of factors that affect the voluntary disclosure decision is an important research topic that has implications for policy-makers, auditors, investors, and other stakeholders. Voluntary disclosures examined by other studies include management earnings forecasts (Nagar et al., 2003, Ajinkya et al., 2005 and Karamanou and Vafeas, 2005), personnel announcements (Miller, 2002), and environmental disclosures (De Villiers and Van Staden, 2006 and Clarkson et al., 2008), among others. One neglected area of research concerns the disclosure of related-party transactions. Previous studies provide empirical evidence that related-party transactions are widely used to shift income and manage earnings for financial reporting and tax purposes (Harris, 1993, Jacob, 1996, Wang et al., 2008, Aharony et al., 2010, Jian and Wong, 2010, Lo et al., 2010a and Lo et al., 2010b). However, to the best of our knowledge, no preceding study has empirically examined the economic determinants of related-party transaction disclosures. We contribute to the current literature by examining the voluntary disclosure of transfer pricing methods used for related-party transactions in China. The Accounting Standard for Business Enterprises (MOF, 1997) specifies that Chinese-listed companies are mandated to disclose various details of their related enterprises as well as the types and amounts of all their related-party transactions in the notes to financial statements. However, it does not explicitly require listed companies to disclose the transfer pricing policies and methods of their related-party transactions. Therefore, the disclosure of transfer pricing methods is a voluntary management decision in China; managers can choose to disclose or not disclose transfer pricing methods in the notes to financial statements.2 This offers us an excellent setting to investigate managers’ decisions on transfer pricing disclosures. In particular, we aim to examine whether the voluntary disclosure of transfer pricing policies is associated with earnings management incentives and governance structure. We argue that independent directors and institutional shareholders can help to monitor the management’s decisions and lead to more voluntary disclosures of transfer pricing methods. Conversely, management is less likely to voluntarily disclose transfer pricing information if it manages earnings by manipulating transfer prices of related-party transactions and anticipates negative consequences (costs) of disclosing such information to stakeholders. The findings support our hypotheses and suggest that firms with a higher percentage of independent directors are more likely to voluntarily disclose the pricing details of their related-party transactions. Moreover, we find that firms with a higher percentage of government ownership are more likely to voluntarily disclose transfer pricing methods. This finding is consistent with the view that government-owned companies are more likely to make voluntary disclosures to reduce adverse selection problems. However, we find no significant association between institutional ownership and voluntarily disclosures of transfer pricing methods. Regarding the earnings management incentives, we find that firms with a higher level of earnings management resulting from related-party transactions (measured by abnormal related-party transactions) are less likely to voluntarily disclose transfer pricing methods. Our empirical results further suggest that managers who boost earnings to increase their bonuses or to meet earnings thresholds are less likely to engage in voluntary disclosures of transfer pricing methods. These results are robust in controlling for the type of goods being traded, tax effects, audit quality, and various firm characteristics (e.g., profitability, leverage, volume of related-party sales, capital needs, size, and growth) as well as the industry and year effects. In addition to investigating voluntary disclosure decisions, we conduct supplementary analyses to test the firms that are more likely to be disciplined by the Shanghai Stock Exchange or the China Securities Regulatory Commission (CSRC) for noncompliance with applicable mandatory disclosure requirements for related-party transactions (RPT sanctions). The results of our supplementary analyses show that firms with a higher percentage of independent directors, institutional investors, or government ownership are less likely to be sanctioned for noncompliance with mandatory disclosure requirements. However, we find no association between earnings management incentives and RPT sanctions. These findings imply that the decision to not comply with mandatory disclosure requirements is different from the decision to make additional voluntary disclosures; therefore, the same variables may not necessarily determine both decisions.3 This paper contributes to the existing literature by addressing an unexplored research issue: the voluntary disclosure of transfer pricing methods. The transfer pricing manipulation of companies can have far-reaching implications for the economies and societies in which they operate. In particular, outward income shifting by companies adversely affects government plans and policies with respect to tax revenue collection, foreign exchange, and market development (Lo et al., 2010a). Understanding managers’ motivations regarding the voluntary disclosure of transfer pricing methods should help policy makers to obtain more accurate information for improving economic, social, and financial policies. We believe that the findings of this paper may be generalizable to other Asian and developed countries. OECD (2009) suggests that the abusive manipulations of related-party transactions are one of the most significant corporate governance challenges faced by enterprises in China and other Asian countries, because many Asian enterprises are family-owned or state-owned. Given the similar institutional environments across the region, we believe that our findings are generalizable beyond the Chinese context.4 In addition, the empirical results of this study indicate that earnings management and governance structures influence management’s voluntary disclosure decisions and a firm’s risk of being sanctioned. The findings on RPT sanctions also provide the first form of evidence on regulatory disciplinary actions for noncompliance with mandatory disclosure regulations in China. Our data may help auditors to better evaluate risk when auditing related-party transactions and may provide insights for policy-makers seeking to prevent future financial scandals. Our findings are particularly important in countries where the monitoring and auditing processes for related-party transactions are extremely difficult (Levine et al., 1997 and AICPA, 2001) or where the presence of high-quality auditors may not effectively prevent or alleviate transfer-pricing manipulations (Lo et al., 2010a). The remainder of this paper is organized as follows. The next section discusses the institutional background. The third section formulates our research hypotheses, while the fourth section discusses the research methodology. The fifth section explains our results, and the sixth section presents our supplementary analyses on RPT sanctions. We provide concluding remarks in the final section. 2. Institutional background 2.1. Reform of state-owned enterprises and related-party transactions in China Because of its unique institutional environment, Chinese listed firms are likely to have significant related-party transactions with their unlisted parent enterprises and with other related companies. Since the 1990s, state-owned enterprises (SOEs) in China have been experiencing reform aimed at improving their efficiency and profitability. As part of this reform, profitable units of SOEs are separated from the original enterprises and then listed on the stock exchange (Green, 2003 and Chow, 2007). These SOEs, which are wholly (100%) owned by agencies of the state or the local municipality, retain the unprofitable units of the business as well as engage in social activities (e.g., hospitals and schools). Even after separation, the SOEs usually retain significant ownership in the listed spin-off firms and exert substantial influence over them (Ding et al., 2007). One of the main influences is on the decision regarding related-party transactions (Lo et al., 2010b). Related-party transactions are a normal part of business, and many firms report a high volume of such transactions without committing accounting and financial fraud (Gordon et al., 2007). However, under certain conditions, related-party transactions can allow controlling shareholders or executives of a company to earn personal benefits and to expropriate wealth away from the minority shareholders of the listed firm5 (OECD, 2009). For example, Jian and Wong (2010) find evidence that Chinese-listed companies use related-party sales to achieve certain earnings targets. Lo et al. (2010b) find that firms with higher levels of government ownership are more likely to shift profits to their parent SOEs via transfer pricing. This transfer of profits to the parent SOEs allows them to subsidize their loss-making operations. 2.2. Disclosure requirements for related-party transactions in China Because of the significant volume of related-party transactions and their potential implications for profits and cash flows, adequate disclosure of related-party transactions and governance in approving such transactions are important for minority investors. In China, the disclosure requirements for related-party transactions are generally governed by the Accounting Standards for Business Enterprises (MOF, 1997) and the Shanghai Stock Exchange Listing Rules (Shanghai Stock Exchange, 2004). According to the Accounting Standards for Business Enterprises (MOF, 1997), listed companies must disclose information on any related enterprises in the notes to financial statements, including (i) the name, (ii) type of business entity, (iii) legal representative, (iv) place of registration, (iv) amount of registered capital and changes therein, (v) principal business, and (vi) proportion of shareholding and changes therein. The regulations require the disclosure of such information even though there may be no transactions between the listed companies and their related enterprises in a given reporting year. In cases where related-party transactions have occurred, the listed companies should disclose the nature of the relationships, types of transactions, and essential elements of the transactions (usually the amounts of related-party transactions) in the notes to financial statements. In addition to making regular disclosures of related-party transactions in the notes to financial statements, the Shanghai Stock Exchange Listing Rules require listed companies to follow specific procedures in approving related-party transactions and to provide timely disclosures of significant related-party transactions (Shanghai Stock Exchange, 2004). For example, if a director is involved in a related-party transaction or has controlling power over the transaction, s/he is prohibited from voting to approve or disapprove that transaction.6 If the related-party transaction is material (i.e., if the amount of a related-party transaction exceeds RMB 3 million and accounts for more than 0.5% of the absolute value of the listed company’s latest audited net assets), the listed company must announce to the public the details of the related-party transaction. The listed company must also submit relevant information to the stock exchange. Examples of relevant information are draft announcements, agreements or letters of intent relating to the transaction, board resolutions, competence reports, independent directors’ prior written approval for the transaction, independent directors’ opinions, and other related documents. It is important to note that there is no regulation requiring listed companies to disclose the transfer pricing policies and methods for related-party transactions in the notes to financial statements. During our sample period (from 2004 to 2005), the disclosures of pricing policies were conducted on a voluntary basis. Thus, the regulations encourage but do not require listed companies to disclose pricing policies and related details. This offers an excellent setting for investigating managers’ decisions on transfer pricing disclosures before such a disclosure policy becomes mandatory.7 3. Hypothesis development 3.1. Earnings management through related-party transactions Previous research studies provide empirical evidence that related-party transactions are widely used to manage earnings for financial reporting and tax purposes. For example, Jacob (1996) finds that firms shift their income via related-party transactions for tax savings. Aharony et al. (2010) find that related-party sales are used to manage earnings upward in the pre-IPO period. Jian and Wong (2010) conclude that firms use related-party sales, which serve as a substitute (rather than a complement) for earnings management via discretionary accruals to meet their earnings targets. Lo et al. (2010b) document that Chinese listed firms manage earnings through related-party transactions to reduce tax liabilities. Although it is difficult for the shareholders to prevent such opportunistic behavior, they can discipline such behavior by selling or refusing to buy the shares of companies that have related-party transactions (Naughton, 2006 and Kohlbeck and Mayhew, 2010). To minimize the negative impacts of earnings management on a firm’s values, the management may choose to withhold information that relates to its opportunistic behavior. Jo and Kim (2008) find that disclosure is negatively correlated with the degree of earnings manipulation. Furthermore, Zechman (2010) shows that managers of firms with incentives to defer cash payments and keep debt off the balance sheet via synthetic leases are less likely to voluntarily disclose the financial consequences of their synthetic leases. Considering these findings, we expect that if the management boosts earnings through related-party transactions, it is less likely to engage in voluntary disclosure regarding the pricing methods of such transactions, because disclosures would make shareholders more likely to uncover their manipulations. Therefore, we propose the following hypothesis: H1. Companies with a higher level of earnings management through related-party transactions are less likely to voluntarily disclose the pricing methods of related-party transactions. 3.2. Independent directors Managers, as agents of their shareholders, manage firms on their behalf. Problems may arise because of the separation of ownership and control (Jensen and Meckling, 1976). Agency theories suggest that the decisions of managers (i.e., agents) should be monitored to ensure that they act on behalf of the shareholders (i.e., the principals). To appropriately monitor management, directors are appointed by the shareholders. They work to ensure that management policies are in the best interests of the shareholders and that the financial statements reflect the true financial position of the company. This governance role played by independent directors is well documented in prior research (e.g., Byrd and Hickman, 1992, Beasley, 1996, Klein, 2002, Davidson et al., 2005 and Lo et al., 2010a). In addition to ensuring that the financial statements reflect the true financial position of the company, independent directors are responsible for ensuring greater transparency of company information in the shareholders’ interests (Ajinkya et al., 2005). Therefore, we expect that the presence of an increased number of independent directors is associated with high-quality disclosure or high voluntary disclosure. Prior studies present empirical evidence on the association between the proportion of independent directors and voluntary disclosures. For example, Karamanou and Vafeas (2005) report that a higher proportion of independent directors leads to a higher frequency of voluntary earnings forecasts. Xiao et al. (2004) and Kelton and Yang (2008) find that board independence is positively associated with voluntary Internet-based disclosures. Chen and Jaggi (2000) find a positive association between the proportion of independent directors and the comprehensiveness of financial disclosures in Hong Kong. Therefore, holding the level of earnings management through related-party transactions constant, we hypothesize that firms with a higher percentage of independent directors on their boards are likely to engage in more voluntary disclosures regarding the transfer pricing methods of their related-party transactions in the notes to financial statements. H2. Ceteris paribus, companies with a higher percentage of independent directors on their boards are more likely to voluntarily disclose the pricing methods of related-party transactions. 3.3. Ownership structure Although shareholders are not involved in managing or operating their companies, they can still exercise significant influence over management decision making through block ownership. Healy et al. (1999) and Bushee and Noe (2000) report that firms with a higher level of institutional ownership have higher disclosure ratings. Similarly, Ajinkya et al. (2005) find that firms with a higher percentage of institutional ownership issue more voluntary earnings forecasts. Therefore, we hypothesize that firms with a higher level of institutional ownership are more likely to make voluntary disclosures of the transfer pricing methods of their related-party transactions. H3a. Ceteris paribus, companies with a higher level of institutional ownership are more likely to voluntarily disclose the pricing methods of related-party transactions. Given the peculiarities of the Chinese market, we investigate the impact of government ownership on voluntary disclosures. As mentioned previously, many SOEs in China have been reformed, such that the profitable units of the SOE are separated from the original enterprise and are listed on the stock market. After separation, however, many of these listed companies have remained under the control of the original SOEs (unlisted companies) (Wu, 2005 and Walter and Howie, 2006). This unique government ownership structure for Chinese listed firms exerts significant influence on management business decisions (say, the decision to switch auditors) (Wang et al., 2008). In the context of related-party transactions, management decisions to initiate related-party sales and to determine transfer prices are also affected by government ownership. For example, Jian and Wong (2010) find that state-owned firms are more likely to use related-party sales to manage earnings than non-state-owned firms are. Furthermore, Lo et al. (2010b) find that firms with a higher percentage of government ownership are more likely to tunnel profits out from the listed companies via transfer pricing manipulation. Since the transfer pricing manipulation by listed government-owned firms is an important concern for investors in China (CFA Institute, 2007 and OECD, 2009), we argue that there should be a great demand for communication with other (non-government) shareholders of government-controlled firms to reduce the high agency costs involved (Eng and Mak, 2003). Previous research provides empirical evidence supporting this view. Ferguson et al. (2002) find that the Chinese SOEs listed on the Hong Kong Stock Exchange voluntarily disclose more information than other listed firms do to reduce problems of adverse selection and moral hazard. Eng and Mak (2003) report similar results regarding the extent of voluntary disclosure by Singaporean government-controlled firms. Overall, these findings document that government-owned companies have higher incentives to make voluntary disclosure than non-government-owned companies to reduce adverse selection and alter external perceptions of certain activities. Therefore, we expect that, compared to non-government-owned firms, government-owned firms have higher incentives to voluntarily disclose their transfer pricing policies for minimizing investors’ negative perceptions of their related-party transactions. Therefore, we state the following: H3b. Ceteris paribus, companies with a higher level of government ownership, where the government is a controlling shareholder, are more likely to voluntarily disclose the pricing methods of related-party transactions. 4. Research methodology 4.1. Data collection Our initial sample includes all companies listed on the Shanghai Stock Exchange in 2004 and 2005, which consists of 1602 firm-year observations. We exclude 544 observations that did not involve the sale or purchase of raw materials, goods, or services to/from related companies in either 2004 or 2005. The details of the related-party transactions were manually collected. Our data, which include the amounts, types, pricing methods and pricing policies associated with the transactions, were extracted from the listed companies’ notes to financial statements. Appendix A illustrates actual examples of the disclosures of two listed firms regarding their related-party transactions. The first firm disclosed its pricing methods for all of its related-party transactions, while the second firm did not disclose its pricing methods for any of its related-party transactions. The financial information and ownership details of all companies in our sample, such as profitability, firm size, and ownership structure, were collected from the China Stock Market and Accounting Research (CSMAR) databases. Moreover, we manually collected the details of management remuneration, as disclosed in the companies’ annual reports. A total of 62 firm-year observations were excluded because of a lack of data. Our final sample consists of 996 firm-year observations (i.e., 498 firms), which meet all of our criteria. 4.2. Regression model We test our hypotheses using the following logistic regressions: equation(1) DISCLOSE=α0+α1ABN_RPT+α2IND_DIR+α3INSTITUTE+α4GOV+α5TYPE+α6TAX+α7TOP10+α8ROE+α9RPT+α10D/E+α11GROWTH+α12NEW_CAP+α13SIZE+α14LEI+α15CMI+α16MII+α17Year2005+α18COMPETITIVE+α19IndustryDummies+ϵ.DISCLOSE=α0+α1ABN_RPT+α2IND_DIR+α3INSTITUTE+α4GOV+α5TYPE+α6TAX+α7TOP10+α8ROE+α9RPT+α10D/E+α11GROWTH+α12NEW_CAP+α13SIZE+α14LEI+α15CMI+α16MII+α17Year2005+α18COMPETITIVE+α19IndustryDummies+ϵ. Turn MathJax on equation(2) DISCLOSE=α0+α1ABN_RPT+α2HRPT_BONUS+α3HRPT_ROE+α4IND_DIR+α5INSTITUTE+α6GOV+α7TYPE+α8TAX+α9TOP10+α10ROE+α11RPT+α12D/E+α13GROWTH+α14NEW_CAP+α15SIZE+α16LEI+α17CMI+α18MII+α19Year2005+α20COMPETITIVE+α21IndustryDummies+ϵ.DISCLOSE=α0+α1ABN_RPT+α2HRPT_BONUS+α3HRPT_ROE+α4IND_DIR+α5INSTITUTE+α6GOV+α7TYPE+α8TAX+α9TOP10+α10ROE+α11RPT+α12D/E+α13GROWTH+α14NEW_CAP+α15SIZE+α16LEI+α17CMI+α18MII+α19Year2005+α20COMPETITIVE+α21IndustryDummies+ϵ. Turn MathJax on 4.3. Dependent and explanatory variables The dependent variable of both models, DISCLOSE, is a dummy variable8 that is equal to 1 if the companies voluntarily disclose their transfer pricing methods (regardless of the extent of disclosure).9 The first explanatory variable, ABN_RPT, represents the abnormal related-party transactions that measure the level of earnings management resulting from related-party transactions. Regarding abnormal related-party transactions, we follow the model developed by Jian and Wong (2010) and run an OLS regression year by year to remove any normal components of related-party transactions associated with industry classifications (as classified by a firm’s two-digit industry code assigned by the Shanghai Stock Exchange), size (as measured by the natural logarithm of a firm’s total assets), leverage (as measured by a firm’s debt-asset ratio), and growth (as measured by a firm’s market-to-book equity ratio). The residual term is our measure of abnormal related-party transactions (ABN_RPT). We expect ABN_RPT to be negatively correlated with DISCLOSE, as hypothesized in H1. Prior studies suggest that management boosts earnings (i) to increase its own compensation (Lo et al., 2010b) and (ii) to meet earnings targets for issuing new shares or avoiding losses through related-party transactions (Jian and Wong, 2010). To further investigate the relation between the voluntary disclosure of transfer pricing methods and these two earnings management incentives, we incorporate two proxies, HRPT_BONUS and HRPT_ROE, in Model (2). HRPT_BONUS captures the earnings manipulation incentives from management bonus packages and is equal to 1 if a firm has high abnormal related-party transactions (i.e., ABN_RPT above the median) and its management remuneration contains performance-linked bonuses, and 0 otherwise. Keeping in mind Marquardt and Wiedman’s (2007) argument that the use of bonus contracts for CEOs tends to reduce the quality of disclosures, we expect a negative coefficient on HRPT_BONUS. HRPT_ROE captures earnings thresholds incentives and is a dummy variable that is equal to 1 if a firm has high abnormal related-party transactions (i.e., ABN_RPT above the median) and a strong incentive to meet earnings targets, and 0 otherwise. In China, the regulations on share issuance and delisting lead to strong incentives for listed companies to manage earnings. Chinese listing rules require from companies a minimum return on equity (ROE) prior to the offering of rights. Specifically, firms should have an average ROE of 6% or more during the 3 years prior to their rights issues to existing shareholders. To qualify for a public offering, firms should have a minimum ROE of 10% for the year immediately before the placement and an average ROE of 10% or more over the previous 3 years (CSRC, 2002 and Shanghai Stock Exchange, 2004).10 In addition, Chinese-listed companies cannot report losses for three consecutive years; if such is the case, their shares will be suspended and delisted (CSRC, 2001). Following Jian and Wong (2010), we classify a firm as having stronger incentives to achieve earnings targets if the firm’s ROE is just above the profitability threshold for qualifying for a share issue, if its ROE is just above the delisting threshold, or if it is seeking a shares issue in the following year. In particular, HRPT_ROE is equal to 1 if the firm’s ABN_RPT is above the median and its ROE is between 6% and 8%, 10% and 12%, or 0% and 2%, or if it plans to issue shares in the following year. We expect that there will be less voluntary disclosure for transfer pricing methods if these earnings management incentives are strong (i.e., a negative coefficient on HRPT_ROE). IND_DIR measures the percentage of directors on the board that are deemed independent. INSTITUTE indicates the percentage of shares owned by institutional investors (including both domestic investors and foreign institutional investors that invest in China via the Qualified Foreign Institutional Investor (QFII) program). GOV measures the percentage of shares owned by the government as the controlling shareholder (we traced all pyramid shareholding structures to identify the ultimate controlling investor). As hypothesized in H2, H3a, and H3b, we expect IND_DIR, INSTITUTE, and GOV to be positively correlated with DISCLOSE in both models. 4.4. Control variables The first control variable, TYPE, is the proportion (expressed as a percentage) of total related party transactions that are trades of tangible goods. The price of intangible goods (e.g., services) is more difficult to determine than that of tangible goods, as intangible assets lack physical substance and are usually very difficult to value (Gordon et al., 2007). Second, we include TAX, which measures the actual tax rate of the listed company, in our regression model. Gramlich et al. (2004) suggest that companies can shift profits from high-tax jurisdictions to low-tax jurisdictions to reduce the groups’ overall tax liabilities. These tax-induced profit-shifting practices may influence management decisions regarding disclosure of transfer pricing methods. For example, firms entitled to lower tax rates have greater incentives to shift profits through transfer pricing manipulations, which may make them less likely to voluntarily disclose the pricing mechanisms of related-party transactions. We thus include TAX to control for the potential tax effects on voluntary transfer pricing disclosures. Third, we include TOP10, which is equal to 1 if the company is audited by a top 10 auditor11 and 0 otherwise, to control for possible influence from external auditors. We also include several financial performance indicators suggested by prior studies to control for effects on voluntary disclosures (Ho and Wong, 2001, Eng and Mak, 2003, Ajinkya et al., 2005 and Karamanou and Vafeas, 2005). These indicators are ROE, which measures a firm’s return on equity; RPT, which measures the natural logarithm of a firm’s related-party sales12; D/E, which measures a firm’s debt-equity ratio; GROWTH, which measures a firm’s market-to-book equity ratio; and NEW_CAP, which is equal to 1 if the company raises additional capital in the current year or does so in the following two years, and 0 otherwise. We also include SIZE, which measures the natural logarithm of the company’s total assets, as one of the control variables. Larger firms may exhibit more transfer pricing disclosures if their unit fixed costs of disclosure are lower or if their agency costs are higher. On the other hand, larger firms may also engage in fewer disclosures if their political costs are higher. Therefore, we cannot predict the sign of this variable. China is characterized as having great disparities in economic, legal, and social development across its provinces. These regional differences may help explain the differences in the management behaviors of Chinese firms. We include a legal environment index (LEI), which measures the level of development of the legal environment in each province (where a higher LEI indicates a more developed legal environment); a credit market index (CMI), which measures the development of the local credit market (again, higher is better); and a market intermediary index (MII), which measures the development of market intermediaries in each province (a higher level indicates greater development). These indices aim to control for the possible effects of the legal and business environment in each province on firms’ management disclosure decisions. These market development indexes are constructed by Fan and Wang (2004), and previous studies in China have used these indices to control for the impact of regional development on various corporate decisions (e.g., Wang et al., 2008, Chan et al., 2006, Firth et al., 2011 and Lo et al., 2010b). Finally, we include the year indicator, the Herfindahl index, and a set of industry dummies as control variables. Year2005 is equal to 1 if the data are for 2005, and 0 otherwise. The Herfindahl index (COMPETITIVE) is a measure of industry concentration equal to the sum of the squared market shares of firms in the industry. The industry dummies are dummy variables classified by a firm’s two-digit industry code as assigned by the Shanghai Stock Exchange.13 The competitive environment varies across industries, and companies may be reluctant to disclose certain types of internal information to their competitors. Companies in different industries may also have different disclosure practices. Thus, we include the aforementioned variables to control for the possible effects of industry type on firm disclosure behavior. 5. Results 5.1. Descriptive statistics and univariate tests Table 1 lists descriptive statistics on all 996 firm-year observations. These results show that 67.8% of our sample firms voluntarily disclose details on the transfer pricing methods adopted in their related-party transactions. The mean abnormal related-party transactions value (ABN_RPT) is 0.007. Table 1 also indicates that 20.7% of the sample firms have high abnormal related-party transactions and use a management remuneration package that contains performance-linked bonus elements (HRPT_BONUS) and 18.3 percent of the sample firms have high abnormal related-party transactions and strong incentives to boost profits upward to issue new shares or to avoid losses (HRPT_ROE). The mean percentage of independent directors on the board (IND_DIR) is 34.5%, the mean percentage of shares owned by institutional shareholders (INSTITUTE) is 9.9%, and the mean percentage of shares owned by the Chinese government (GOV) is 34.3%.14 The mean values for debt-equity ratio, market-to-book equity ratio, return on equity, volume of related-party sales, and firms’ total assets are 1.230, 2.254, 0.050, RMB 602 million and RMB 3.526 billion, respectively. Table 1. Descriptive statistics. 1% 25% 50% 75% 99% Mean Std. deviation DISCLOSE 0.000 0.000 1.000 1.000 1.000 0.678 0.466 ABN_RPT −0.069 −0.032 −0.019 −0.004 0.520 0.007 0.128 HRPT_BONUS 0.000 0.000 0.000 0.000 1.000 0.207 0.405 HRPT_ROE 0.000 0.000 0.000 0.000 1.000 0.183 0.387 IND_DIR (%) 21.4 33.3 33.3 36.4 45.5 34.5 4.2 INSTITUTE (%) 0.0 0.1 1.5 13.0 67.8 9.9 16.2 GOV(%) 0.0 0.0 38.6 56.8 75.6 34.3 25.6 TYPE (%) 0.0 43.5 95.5 100.0 100.0 71.3 39.2 TAX (%) 0.0 15.0 15.0 33.0 33.0 22.3 10.0 TOP10 0.000 0.000 0.000 1.000 1.000 0.280 0.449 ROE −0.835 0.020 0.061 0.108 0.301 0.050 0.167 RPT 0.000 0.015 0.076 0.252 8.913 0.602 2.651 D/E 0.041 0.575 0.965 1.554 7.559 1.230 1.650 GROWTH 0.646 1.467 1.992 2.816 7.399 2.254 2.081 NEW_CAP 0.000 0.000 0.000 0.000 1.000 0.034 0.182 ASSET 0.305 1.058 1.791 3.720 30.713 3.526 6.529 LEI 2.670 4.620 5.720 9.560 11.040 6.466 2.475 CMI 1.550 3.880 5.310 7.290 8.600 5.382 2.049 MII 0.440 2.360 2.890 6.920 28.110 6.271 7.405 Year2005 0.000 0.000 0.500 1.000 1.000 0.500 0.500 COMPETITIVE 0.000 0.026 0.094 0.111 0.882 0.100 0.091 Definitions of variables: DISCLOSE = 1 if the companies have voluntarily disclosed the pricing methods of related-party transactions; 0 otherwise. ABN_RPT = abnormal related-party transactions. HRPT_BONUS = 1 if ABN_RPT is above the median and the management remuneration includes bonus elements; 0 otherwise. HRPT_ROE = 1 if ABN_RPT is above the median and the company has strong incentives to meet earnings targets; 0 otherwise. IND_DIR = percentage of independent directors on the board. INSTITUTE = percentage of shares owned by institutional shareholders. GOV = percentage of shares owned by the government as the controlling shareholder. TYPE = percentage of related-party transactions involving the trade of tangible goods between related companies. TAX = corporate income tax rate. TOP10 = 1 if the company is audited by a top 10 auditor; 0 otherwise. ROE = return on equity. RPT = volume of related-party transactions (in RMB billion). D/E = debt-equity ratio. GROWTH = market-to-book equity ratio. NEW_CAP = 1 if the company issues new shares in the current year or in the next two years; 0 otherwise. ASSET = total assets of the listed company (in RMB billion). LEI = legal environment index; a higher LEI indicates a better legal environment in that province. CMI = credit market index; a higher CMI indicates a more developed local credit market in that province. MII = market intermediary index; a higher MII indicates a more developed system of market intermediaries in that province. Year2005 = 1 if the data is for 2005; 0 otherwise. COMPETITIVE = Herfindahl index of competitiveness. Full-size table Table options * View in workspace * Download as CSV Table 2 compares the characteristics of firms that voluntarily disclose the pricing methods of related-party transactions with those of firms that do not. The findings reveal that high-ABN_RPT firms (i) with management remuneration packages that contain performance-linked bonuses (HRPT_BONUS) and (ii) that have strong incentives to achieve earnings thresholds (HRPT_ROE) are less likely to voluntarily disclose the pricing methods of related-party transactions. Moreover, Table 2 shows that firms with voluntary disclosure of transfer pricing methods have a higher percentage of independent directors (IND_DIR), which is consistent with our prediction in hypothesis H2. Table 2 also shows that firms with voluntary disclosure of transfer pricing methods have a larger volume of related-party transactions (RPT), have a higher percentage of transactions involving the trading of tangible goods (TYPE), are less likely to issue new shares in subsequent years (NEW_CAP), and are more likely to be located in provinces with less legal protection (LEI) and less developed systems of market intermediaries (MII).Table 3 shows the correlations of the variables of Models (1) and (2). As shown in the table, ABN_RPT, HRPT_BONUS, and HRPT_ROE are negatively correlated with DISCLOSE, which implies that firms with earnings management through related-party transactions are less likely to voluntarily disclose their transfer pricing methods, no matter what type of incentive may be offered (e.g., performance-linked bonus contracts or earnings targets incentives). The findings also reveal that MII, NEW_CAP, and GROWTH are negatively correlated with DISCLOSE, which implies that a less developed system of market intermediaries, a lower need for additional capital, and a smaller market-to-book equity ratio are associated with a higher likelihood of voluntary disclosures of pricing methods of related-party transactions in China.