Affiliated management and firm value: Evidence from China

Gaoping Zheng; Jian Xue; Xiao Chen

doi:10.1080/21697213.2014.921366

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

DeepDyve requires Javascript to function. Please enable Javascript on your browser to continue.

Affiliated management and firm value: Evidence from China

Zheng, Gaoping; Xue, Jian; Chen, Xiao 2014-07-03 00:00:00 China Journal of Accounting Studies, 2014 Vol. 2, No. 3, 172–199, http://dx.doi.org/10.1080/21697213.2014.921366 Afﬁliated management and ﬁrm value: Evidence from China a b b Gaoping Zheng *, Jian Xue and Xiao Chen Southwestern University of Finance & Economics, School of Accounting, Chengdu 611130, China; Tsinghua University, School of Economics and Management, Beijing 100084, China We provide evidence on how the presence of a top manager who is a controlling shareholder or the employee of a controlling shareholder (i.e. afﬁliated management) is associated with ﬁrm value. By analysing the data of Chinese listed ﬁrms in the A-share stock market from 2001 to 2009, we ﬁnd that afﬁliated management is neg- atively associated with ﬁrm value. Furthermore, although ﬁrms with afﬁliated man- agement have higher return on assets and return on sales, afﬁliated management is positively associated with the value and the frequency of loan-based related party transactions and with the likelihood of irregularity in information disclosure. Our ﬁndings imply that even if afﬁliated management may enhance shareholders’ ability to monitor managers and increase the operational proﬁtability, afﬁliated management can facilitate the expropriation of the interests of the minority shareholders by the controlling shareholders. Keywords: afﬁliated management; ﬁrm value; private beneﬁt of control 1. Introduction The co-ordination of various factors of production built up within a ﬁrm can reduce trans- action costs dramatically (Coase, 1937). However, as the size of a ﬁrm increases, the issue of the separation between ownership and control arises as well, which brings agency cost (Jensen & Meckling, 1976). Scholars and practitioners have proposed many mechanisms to alleviate the agency problem. For instance, Shleifer and Vishny (1986) argue that the controlling shareholders have incentives to monitor the management and reduce the agency problem between shareholders and managers. La Porta, Lopez-de-Silanes, and Shleifer (1999) show that in some cases the management of family ﬁrms in the United States is assigned to family members, while some controlling shareholders assign their employees to serve as the management in subsidiaries. Further empirical evidence from developed economies shows that family ﬁrms with afﬁliated management have higher ﬁrm value than other family ﬁrms, implying that afﬁli- ated management mitigates the agency problem in family ﬁrms (Anderson & Reeb, 2003; Villalonga & Amit, 2006). However, in a transition economy where the investor protec- tion is weak and the controlling shareholders are prevalent (Claessens, Djankov, & Lang, 2000; La Porta, Lopez-de-Silanes, Shleifer, & Vishny, 1998), controlling shareholders may control the ﬁrms for private beneﬁts (Chen, Chen, Schipper, Xu & Xue, 2012; Claessens, Djankov, Fan, & Lang, 2002; La Porta, Lopez-de-Silanes, Shleifer, & Vishny, 2000), causing afﬁliated management to have a different impact on ﬁrm value. *Corresponding author. Email: zhenggaoping2012@swufe.edu.cn Paper accepted by Kangtao Ye. © 2014 Accounting Society of China China Journal of Accounting Studies 173 Theoretically, afﬁliated management could serve not only as a channel through which the controlling shareholder can seize the private beneﬁt of control, but also as a channel through which the controlling shareholder can enhance the monitoring of the management of listed ﬁrms and hence reduce the agency problem. The former (latter) channel infers that afﬁliated management will be negatively (positively) associated with ﬁrm value. Thus, empirical analysis could clarify the association of afﬁliated manage- ment with ﬁrm value. To the best of our knowledge, although a few studies test the impact of the separation of cash ﬂow rights and control rights on ﬁrm value (Claessens et al., 2000, 2002), the association between afﬁliated management and ﬁrm value is still unknown. This paper contributes to the literature by empirically testing the association of afﬁliated management with ﬁrm value, using the unique data of afﬁliated manage- ment in China, the second largest economy in the world and the largest transition economy. Afﬁliated management is a common practice in Chinese listed ﬁrms, which pro- vides a rich dataset from which to test the association of afﬁliated management with ﬁrm value. Our data show that more than half of listed ﬁrms have afﬁliated manage- ment from 2001 to 2009 (Table 2, later). In an emerging market, afﬁliated management can facilitate the tunnelling behaviour of controlling shareholders, because the control- ling shareholders may abuse their control over the listed ﬁrm, given the weak investor protection (Claessens et al., 2000, 2002). Based on such concern, the China Securities Regulatory Commission (CSRC) established several regulations to constrain the emer- gence of afﬁliated management in listed ﬁrms. However, the State-Owned Assets Supervision and Administration Commission of the State Council (SASAC) holds the opposite view, urging a few giant ﬁrms owned by the central state, such as China National Petroleum Corporation, SINOPEC, and SHENHUA Group, to build up afﬁli- ated management. From the view of SASAC, afﬁliated management should bring better monitoring of the management of listed ﬁrms and reduce the agency problem between shareholders and management, which may increase the value of the ﬁrm and the wealth of the nation eventually. This paper also contributes to the long-lasting debate about the role of afﬁliated management in China. By analysing the data of Chinese listed ﬁrms in the A-share stock market from 2001 to 2009, we ﬁnd that the afﬁliated management is negatively associated with ﬁrm value. Furthermore, although ﬁrms with afﬁliated management have higher return on assets and return on sales, the afﬁliated management is positively associated with the value and the frequency of loan-based related party transactions and with the likelihood of irregularity in information disclosure in the listed ﬁrms. The ﬁndings imply that even if afﬁliated management may enhance the ability of shareholders to monitor managers, it can facilitate the expropriation of the interests of the minority shareholders by the controlling shareholder in a transition economy. Further evidence shows that the association between afﬁliated management and ﬁrm value does not depend on whether the largest shareholder in the ﬁrm owns more than 50% or not. Meanwhile, the negative association between afﬁliated management and ﬁrm value is weaker in ﬁrms with a high level of institutional investor holdings, which implies that institutional investors may restrict the harmful behaviour of afﬁliated man- agers. Our results show that the association between afﬁliated management and ﬁrm value is not signiﬁcant in the ﬁrms that are implementing their seasoned equity offer- ing, which is consistent with the idea that controlling shareholders do not have a high incentive to expropriate the interests of minority shareholders when the ﬁrm has a high demand for external ﬁnancing. We also ﬁnd that the negative association of afﬁliated 174 Zheng et al. management with ﬁrm value is mostly driven by non-state-owned listed ﬁrms. One pos- sible explanation is that compared with the controlling shareholder (i.e. the state) in the state-owned listed ﬁrms, the controlling shareholders (generally a person or a family) in the non-state-owned listed ﬁrms are more prone to expropriate the interests of minor- ity shareholders, with the help of afﬁliated managers, for their private beneﬁts, given the weak institutional environment. Meanwhile, the afﬁliated management in the state- owned listed ﬁrms is under more severe monitoring by the government than the afﬁli- ated management in the non-state-owned listed ﬁrms; thus, the afﬁliated managers’ behaviour that harms ﬁrm value is more likely to be restrained in the state-owned listed ﬁrms. Our main results hold after controlling for different types of afﬁliated manage- ment, and the different positions that the afﬁliated management hold in the controlling shareholder, as well as the split share structure reform. After investigating the associa- tion between afﬁliated management and the ROA (return on assets) and ROS (return on sales), we ﬁnd that afﬁliated management is positively associated with the ROA and ROS in the listed ﬁrms. The ﬁndings imply that the presence of afﬁliated management may improve the operational proﬁtability in the ﬁrm. However, we also ﬁnd that ﬁrms with afﬁliated management are more likely to manipulate information disclosure, including fraudulent accounting treatment (such as exaggerated proﬁt, false statement of assets), which may apparently increase ROA and ROS. Even if the increase of prof- itability is authentic, the controlling shareholder could turn it into a private beneﬁtof control, with the help of afﬁliated management, through related party transactions or other channels, which may result in a negative association between afﬁliated manage- ment and the ﬁrm value. The ﬁndings contribute to the literature in the following ways. First, this study empirically tests the role of afﬁliated management, from the corporate governance per- spective, using a large sample in a transition economy, which ﬁlls a gap left by previ- ous research. Theoretically, afﬁliated management could reduce ﬁrm value when it serves as a channel through which a controlling shareholder seizes the private beneﬁt of control, but it could also increase ﬁrm value when it serves as a channel through which a controlling shareholder enhances the monitoring of management. Previous studies focus on the role of afﬁliated management in a developed economy with good investor protection, paying little attention to the case in an emerging market. This paper shows that, given weak investor protection in an emerging market, the afﬁliated man- agement is negatively associated with the ﬁrm value, which implies that the cost related to afﬁliated management exceeds the beneﬁts in a transition economy. Secondly, our research on afﬁliated management in Chinese listed ﬁrms offers us a good opportunity to examine the impact of large shareholder control in an emerging market. Claessens et al. (2000) show that controlling shareholders use four channels to enhance their con- trol in East Asian countries, including a pyramid structure, cross-holdings, dual class shares and afﬁliated management. Claessens et al. (2002) test the impact of a pyramid structure, cross-holdings and dual class shares on ﬁrm value. They ﬁnd that these three channels induce separation of control rights and cash ﬂow rights in both private ﬁrms and state-owned ﬁrms, which enlarges the incentive for the controlling shareholder to pursue the private beneﬁt of control and reduce the ﬁrm value. They fail to test the impact of afﬁliated management on ﬁrm value due to data insufﬁciency. As a result, Claessens et al. (2002) mentioned that they may underestimate the effect of large share- holder control. In our setting, afﬁliated management is dominant in the Chinese stock market and the information disclosure of afﬁliated management in listed ﬁrms is man- dated, which means that our paper can provide another dimension to investigating the China Journal of Accounting Studies 175 impact of large shareholder control in an emerging market. Finally, this paper not only shows the negative association of afﬁliated management with ﬁrm value, but also discusses the channels through which afﬁliated management has this negative associa- tion with ﬁrm value. We ﬁnd that the presence of afﬁliated management is positively associated with the value and the frequency of related party transactions and with the likelihood of irregularity in information disclosure in the listed ﬁrms, which provides valuable evidence for further regulation. The rest of the paper is organised as follows. Section 2 reviews the institutional background; Section 3 develops research hypotheses; Section 4 describes the research design and data; Section 5 reports the empirical results; Section 6 presents the further analysis and a robustness test; and Section 7 concludes the paper. 2. Institutional background Afﬁliated management in China is different from afﬁliated management in other East Asian countries in the sense that the majority of afﬁliated management in China is not based on the family members of controlling shareholders (Bunkanwanicha, Fan, & Wiwattanakantang, 2013; Claessens et al., 2000). The disparity comes from the fact that the Chinese stock market is dominated by state-owned listed ﬁrms (62.7%), in which afﬁliated management cannot be family members, but usually are the employees of the controlling shareholder. Meanwhile, non-state-owned listed ﬁrms can still use family members as the afﬁliated management. State-owned listed ﬁrms include two categories (Zou & Chen, 2004). The ﬁrst category of state-owned listed ﬁrms is those ﬁrms directly controlled by government agencies, such as local government and SASAC. In these ﬁrms, political concerns other than economic factors serve as the major determinant of the personnel arrangements. Meanwhile, the largest shareholders for these ﬁrms are government-funded agencies, without any operational business or ﬁnancial needs, which means that they do not have high incentives to monitor the management or pursue private beneﬁt of control. As a result, the afﬁliated management in the listed ﬁrms directly controlled by government agencies may be arranged to achieve certain political goals instead of economic pur- poses. For example, local government (as they directly control the listed ﬁrms) may arrange the local governors to be the afﬁliated management in listed ﬁrms so that the listed ﬁrms may assist local government in propping local employment or achieving other political goals. Because the personnel arrangement in these ﬁrms may be mostly driven by political concerns other than economic factors, and political concerns are not the focus of our study, we drop this kind of ﬁrms in our sample. The second category of state-owned ﬁrms are the ﬁrms indirectly controlled by government agencies. Typically, the largest shareholders for these ﬁrms are parent state-owned ﬁrms with related business. This kind of controlling shareholder not only serves as the monitor of state-owned shareholdings and assets, but also has close opera- tional and ﬁnancial connections with the listed ﬁrms. On one hand, these controlling shareholders may have strong incentives to use listed ﬁrms to beneﬁt themselves. For example, the controlling shareholders may treat cash holdings of listed ﬁrms as their own, make listed ﬁrms guarantee their bank loans, and even sell assets to or purchase assets from listed ﬁrms at an unfair price. Such incentive partly comes from the fact that the controlling shareholder puts the most valuable business into the listed ﬁrm dur- ing the IPO process, and such assets infusion leads to the fact that the controlling shareholder has no core business to generate necessary cash ﬂows later on. In addition, 176 Zheng et al. a controlling shareholder may beneﬁt itself by making the strategy of listed ﬁrms con- sistent with its own strategy. On the other hand, since listed ﬁrms must achieve certain performance requirements before seasoned equity offering in the Chinese stock market, controlling shareholders may also sacriﬁce their own beneﬁts to ﬁnancially support the listed ﬁrms, in order to prop up the performance of listed ﬁrms in a short period. In this case, the controlling shareholders of this category of state-owned ﬁrms have a strong motivation to arrange the afﬁliated management in the listed ﬁrms to enhance their control over the listed ﬁrms for the beneﬁts of their own. Compared with the controlling shareholders of the above two categories of state- owned listed ﬁrms, the controlling shareholders of non-state-owned listed ﬁrms may have even stronger incentives to enhance their control. First, since the controlling shareholders for non-state-owned listed ﬁrms face more policy risks than those in the state-owned ﬁrms, including the weak protection of private property rights, the uncertainty of indus- trial policies and the discrimination in ﬁnancial policies enforced by government (Luo & Tang, 2009; Yu & Pan, 2008), strong control can facilitate these controlling shareholders to obtain security. Second, given weak investor protection and legal institution, these con- trolling shareholders have strong incentives to enhance their control over listed ﬁrms to get private beneﬁt of control (Claessens et al., 2000). In this case, afﬁliated management could not only help the controlling shareholders directly adjust the operation strategies of the listed ﬁrms according to the frequently changed policies, in order to ensure the con- trolling shareholders’ beneﬁts, but also reduce the transaction costs when the controlling shareholders try to grab beneﬁts by navigating around different government policies. More importantly, afﬁliated management can assure the strong control power of the con- trolling shareholders by directly controlling the management in the listed ﬁrms, since the effective enforcement of the decisions may be more important than making decisions given the weak institutional environment. Both in the state-owned listed ﬁrms and in the non-state-owned listed ﬁrms, afﬁli- ated management can enhance the large shareholder control and facilitate the tunnelling behaviour of controlling shareholders, given the weak investor protection (Claessens et al., 2000, 2002). Based on such concern, the China Securities Regulatory Commis- sion (CSRC) has established several regulations to constrain the emergence of afﬁliated management in listed ﬁrms. The CSRC issued a regulation ‘Announcement for Investi- gating the Restructuring in Firms that Plan to Go through IPO’ in 1998, stipulating that IPO ﬁrms should meet the personnel independence criteria, which require that the chairman in a listed ﬁrm must not serve as the legal representative of the controlling shareholder, that management such as president and vice president must not serve both in the listed ﬁrm and in the ﬁrm that is the controlling shareholder, and that the labour, personnel and wage management in a listed ﬁrm must be completely independent from the controlling shareholder. In 1999, the CSRC issued a special announcement, ‘Announcement for Forbidding Management in Listed Firms Serving in Controlling Shareholder’, requiring that a listed ﬁrm must have separate personnel from those of the controlling shareholder. Further regulation by the CSRC in 2001, ‘Regulations for Initial Public Offering’ requires that management such as the president and vice presi- dent in a listed ﬁrm must only serve in this listed ﬁrm. In 2002, CSRC issued the ‘Code of Corporate Governance for Listed Firms in China’ which enhanced the require- ment for personnel independence in relation to the managers of listed ﬁrms. The CSRC also took an enforcement action in 2007 to check whether or not listed ﬁrms follow the ‘Code of Corporate Governance for Listed Firms in China’. However, the enforcement of these regulations is quite weak, given the lack of monitoring and the lack of severe China Journal of Accounting Studies 177 punishment. Besides, the interventions from other government agencies make the enforcement of these regulations even weaker. For instance, the State-Owned Assets Supervision and Administration Commission of the State Council (SASAC) urges the giant ﬁrms owned by the central state to build up afﬁliated management in their listed ﬁrms, since SASAC holds the view that afﬁliated management should bring better monitoring of the management of listed ﬁrms and reduce the agency problem between shareholders and management, which may eventually increase the value of the ﬁrm and the wealth of the nation. 3. Hypothesis development The discussion about agency problems induced by the separation of ownership and control can be traced back to Adam Smith in the eighteenth century (Jensen & Meckling, 1976). The idea that the ownership in US ﬁrms is so dispersed that manage- ment in these ﬁrms may control the ﬁrm and harm ﬁrm value is well known, since the work of Berle and Means (1932), The Modern Corporation and Private Property, was published. As a result, scholars have investigated how to mitigate such a so-called type I agency problem, the agency problem that concerns the conﬂict of interests between management and shareholders. For the type I agency problem, Shleifer and Vishny (1986) argue that a controlling family can enhance their control over management by appointing family members to the board. Meanwhile, La Porta et al. (1999) show that using a family member as a manager is a common practice around the world. For example, they ﬁnd that 75% of family ﬁrms use family members as managers in the US. They do not study whether managers in family ﬁrms also work as employees of the parent ﬁrms owned by the controlling shareholders. Most empirical ﬁndings focus on the impact of afﬁliated management on ﬁrm value in family ﬁrms in developed economies. Anderson and Reeb (2003) ﬁnd that family ﬁrms exhibit superior performance compared with non-family ﬁrms, and even more superior performance when family members serve as CEOs. Villalonga and Amit (2006) also ﬁnd that family ﬁrms have higher value when the founder of a family ﬁrm serves as CEO or Chairman. However, the answer to the question of how afﬁliated management affects ﬁrm value may be different in an emerging market. In an emerging market, there is a greater prevalence of the type II agency problem, which concerns the conﬂicts of interest between the controlling shareholder and minority shareholders (La Porta et al., 2000). Claessens et al. (2000) show that controlling shareholders use four channels to enhance their control in East Asian countries, including a pyramid structure, cross-holdings, dual class shares and afﬁliated management. Claessens et al. (2002) test the impact of a pyramid structure, cross-holdings and dual class shares on ﬁrm value. They ﬁnd that these three channels induce the separation of control rights and cash ﬂow rights in both private ﬁrms and state-owned ﬁrms, which enlarges the incentives of controlling share- holders to pursue private beneﬁt of control and reduce the ﬁrm value signiﬁcantly. They fail to test the impact of afﬁliated management on ﬁrm value due to insufﬁciency of data. According to Claessens et al. (2002), afﬁliated management, playing a similar role as the other three channels in an emerging market, facilitates the tunnelling behaviour of controlling shareholders by enhancing their control over listed ﬁrms. Previous studies using Chinese data show that expropriations of the minority shareholder by controlling 178 Zheng et al. shareholder are quite common in China (Chen, Li, Su, &Yao, 2012; Fisman & Wang, 2010; Jiang, Lee, & Yue, 2010) and such tunnelling behaviour decreases ﬁrm performance (Chen, Chen, Schipper, Xu, & Xue, 2012; Jiang et al., 2010). If afﬁliated management does help controlling shareholders expropriate the interests of minority shareholders, we predict a negative relation between afﬁliated management and ﬁrm value. We test this prediction with the following hypothesis: H1(A): Firms with afﬁliated management have lower ﬁrm value than ﬁrms without afﬁliated management. However, even in a developed market with strong investor protection, management could harm the interests of shareholders (Jensen & Meckling, 1976). In an emerging market with weak investor protection, managers can easily achieve control over the ﬁrm, using the resources of the ﬁrm to beneﬁt themselves and reduce ﬁrm value (Fan, Wong, & Zhang, 2007). Afﬁliated management may increase ﬁrm value by alleviating this kind of agency problem between management and shareholders (Anderson & Reeb, 2003). Meanwhile, afﬁliated management can be regarded as a vertical interlock between controlling shareholder and listed ﬁrms (Arnoldi, Chen, & Na, 2013), an inter- lock that may help the controlling shareholder to prop the operation performance of listed ﬁrms more easily and increase the ﬁrm performance. From above perspective, afﬁliated management may be positively associated with ﬁrm value. This discussion leads to a competing hypothesis on the relation between afﬁliated management and ﬁrm value. H1(B): Firms with afﬁliated management have higher ﬁrm value than ﬁrms without afﬁli- ated management. 4. Research design 4.1. Data source and research sample We use the following steps to select our sample. First, all data on afﬁliated manage- ment in all the Chinese listed ﬁrms in the A share stock market from 2001 to 2009 are downloaded from the CSMAR database and we get 12,482 ﬁrm*year observations. Second, 173 ﬁrm*year observations from listed ﬁrms in the ﬁnancial sector are deleted, since they are not comparable to other ﬁrms in many perspectives, such as regulations and accounting standards. Third, we drop 2150 ﬁrm*year observations from the ﬁrst category of state-owned listed ﬁrms which are directly controlled by government agen- cies, such as SASAC and local government, because the personnel arrangement in these ﬁrms may be mostly driven by political concerns other than economic factors. Finally, we get 10,159 ﬁrm-year observations. All continuous variables are winsorized at the 1% level. Stock price, ﬁnancial data, ownership structure, board characteristics, related party transactions, irregularity and industry categories are all from the CSMAR database, while we use the CCER data- base to identify whether a ﬁrm is owned by the state or not. 4.2. Model identiﬁcation and variable deﬁnitions Consistent with Claessens et al. (2002), we use the following model to test the associa- tion of afﬁliated management with ﬁrm value. Standard errors in the regression analysis China Journal of Accounting Studies 179 are all corrected for clustering of the observations at the ﬁrm level. For each test sam- ple in the regression analysis, the number of observations may vary as the value of dependent variables (e.g., TQ, LoanRPTs) are missing, or the value of control variables (e.g., Lev, Bsize) are missing. TQ ¼ a þ a AM þ a Gr Sales þ a IND Q þ a SIZE þ a LEV þ a ROA þ a SOE þ a TOP1 0 1 2 3 4 5 6 7 8 X X þ a Herf 2 5 þ a BSIZE þ a OUTDIR þ a DUAL þ YEAR þ IND þ e 9 10 11 12 (1) Referring to Claessens et al. (2002), the dependent variable in model (1) is Tobin’sQ, which measures the value of the ﬁrm. TQ is deﬁned as the sum of book value of liabil- ity, market value of liquid shares and book value of illiquid shares, divided by book value of total assets. Consistent with the deﬁnition in Claessens et al. (2000), the management of a listed ﬁrm is deﬁned as afﬁliated management if Chairman or CEO of the ﬁrm is the controlling 1 2 shareholder, or the employee of the controlling shareholder. The independent variable AM equals one if the ﬁrm has afﬁliated management. Otherwise, AM equals 0. We focus only on Chairman and CEO because both Chairman and CEO are the key positions that directly affect the operational activities of Chinese listed ﬁrms in practice (Chen & Liu, 2003; Du & Zhou, 2010; Gong, 2001; Zhu, 2002; Zhao, Yang, & Bai, 2007). In our empirical analysis, we also calculate some variables as additional controls, such as growth, size, leverage and proﬁtability (Doidge, Karolyi, & Stulz, 2004; King & Segal, 2009). We use two measures, Gr_sales and IND_Q, to proxy for growth. Gr_sales is deﬁned as the average sales growth rate of the previous two years. We use the sales growth rate in the last year if there are missing observations. IND_Q is deﬁned as the median value of Tobin’s Q in the same industry. We deﬁne industries based on the industry classiﬁcation standard issued by CSRC in 2001. Note that we deﬁne manufacturing industry based on the secondary industry code, while we deﬁne other industries based on the ﬁrst industry code. SIZE is deﬁned as the natural loga- rithm of total assets. Generally speaking, ﬁrm value is positively associated with the growth rate and negatively associated with the ﬁrm size (King & Segal, 2009). We use Lev to proxy for ﬁrm leverage. Lev is deﬁned as the sum of short-term debts and long- term debts, divided by total assets. Proﬁtability is measured by ROA, return on assets， deﬁned as net income divided by total assets. Previous studies ﬁnd that the relation between debt and ﬁrm value is negative (Fama & French, 1998; Myers, 1977), and the relation between proﬁtability and ﬁrm value is positive (Doidge et al., 2004;King& Segal, 2009). We also use an industry dummy and a year dummy to control for the unobserved industry and time ﬁxed effects. Ownership structure and board characteristics are also controlled in our regression analysis (Rao, Zhang, & Peng, 2008; Shi, 2010). A dummy variable SOE equals 1 if the ﬁrm is controlled by the state, according to the CCER database. Otherwise, SOE equals 0. One widely accepted idea is that the ﬁrm value in the Chinese state-owned ﬁrms may be lower than the other ﬁrms, because the property rights of these ﬁrms are not well-deﬁned in practice and no one has a strong incentive to monitor the managers, leading to the low level of operation efﬁciency in these ﬁrms (Xia & Zhang, 2005). Variable TOP1 is deﬁned as the percentage of shares held by the largest shareholder. Variable Herf_2_5 measures the monitoring intensity by large shareholders other than the controlling shareholder and is deﬁned as the Herﬁndahl index from the second largest shareholder to the ﬁfth largest shareholder, multiplied by the sum of holdings 180 Zheng et al. from the second largest shareholder to the ﬁfth largest shareholder. BSIZE is deﬁned as the natural logarithm of the number of all board directors. The proxy for board independence is OUTDIR,deﬁned as the number of outside directors divided by the number of all directors. If one person serves as Chairman and CEO at the same time, dummy variable DUAL equals one, otherwise zero. A previous study shows that ﬁrm value is lower when board size is larger, when the board has fewer outside directors, and when one person serves as Chairman and CEO at the same time (Bai, Liu, Lu, Song, & Zhang, 2005). Table 1 shows the main variables and their deﬁnitions. 5. Empirical results 5.1. Descriptive statistics Table 2 reports the distribution of our sample. Panel A presents the sample distribution by year. It may be seen that the percentage of ﬁrms with afﬁliated management Table 1. Deﬁnitions of the main variables. Variable Deﬁnition AM A dummy variable equals 1, if the Chairman or CEO of the ﬁrm is the controlling shareholder, or an employee of the controlling shareholder. Otherwise, AM equals 0. TQ The sum of book value of liability, market value of liquid shares and book value of illiquid shares, divided by book value of total assets. LoanRPT The sum of annual loan-based related party transitions between listed ﬁrms and controlling shareholders, divided by total assets. LoanRPT_FREQ The frequency of loan-based related party transactions in a calendar year. IRREG If CSRC reports that the listed ﬁrm i in year t manipulated information disclosure, including fraudulent accounting treatment (such as exaggerated proﬁt, false statement of assets), misleading statements, and omission of substantial information, the dummy variable IRREG of the ﬁrm i in year t is equal to 1. Otherwise, IRREG is equal to 0. Gr_Sales The average sales growth rate of previous two years. We use the sales growth rate in last year if there are missing observations. IND_Q The median value of Tobin’s Q in the same industry. We deﬁne industries based on the industry classiﬁcation standard issued by CSRC in 2001. Note that we deﬁne manufactory industry based on secondary industry code, while we deﬁne other industries based on ﬁrst industry code. SIZE The natural logarithms of total assets. LEV The sum of short-term debts and long-term debts, divided by total assets. ROA The net income divided by total assets at the end of the accounting periods. AGE The number of years since the ﬁrm has been listed. SOE A dummy variable SOE equals 1, if the ﬁrm is controlled by the state and 0 otherwise. TOP1 The percentage of shares held by the largest shareholder. Herf_2_5 The Herﬁndahl index from second largest shareholder to ﬁfth largest shareholder, multiplied by the sum of holdings from second largest shareholder to ﬁfth largest shareholder following Chen et al. (2012). BSIZE The natural logarithm of the number of all board directors. OUTDIR The number of outside directors divided by the number of all directors. DUAL A dummy variable equals 1, if one person serves as Chairman and CEO at the same time and 0 otherwise. Note: The data source for all the variables is the CSMAR database, except for one variable SOE, which is downloaded from the CCER database. China Journal of Accounting Studies 181 Table 2. Sample distribution. Number of ﬁrms with Percentage of ﬁrms with Number of AM AM observations Panel A Distribution by year 2001 411 47.57% 864 2002 415 45.01% 922 2003 505 51.37% 983 2004 549 51.07% 1,075 2005 623 57.21% 1,089 2006 681 58.40% 1,166 2007 794 62.42% 1,272 2008 857 64.63% 1,326 2009 982 67.17% 1,462 Total 5,817 57.26% 10,159 Panel B Distribution by industry Industry Agriculture 117 54.17% 216 Mining 99 50.25% 197 Food & beverage 213 49.88% 427 Textiles & apparel 250 58.82% 425 Timber & furnishings 29 67.44% 43 Paper & printing 125 67.57% 185 Petrochemicals 649 56.29% 1,153 Electronics 267 64.18% 416 Metal/non-metal 520 60.54% 859 Machinery 894 58.17% 1,537 Pharmaceuticals 425 60.89% 698 Other manufacturing 43 56.58% 76 Utilities 177 47.20% 375 Construction 110 63.22% 174 Transportation 180 54.38% 331 Information technology 449 68.13% 659 Retailing 319 53.61% 595 Real estate 392 49.81% 787 Social Services 218 65.07% 335 Media 54 50.00% 108 Multi-industry 287 50.98% 563 Total 5,817 57.26% 10,159 Key: AM = Afﬁliated management. increases year by year in Panel A, with more than 50% of listed ﬁrms having afﬁliated management since 2003. Panel B shows the sample distribution by industry. The utili- ties industry has the lowest percentage (47.2%), and the information technology indus- try has the highest percentage (68.13%). Table 3 reports the descriptive statistics of the main variables. Panel A reports the full sample statistics. Panel B and Panel C report the statistics for ﬁrms with afﬁliated management and ﬁrms without afﬁliated management respectively. Panel D reports the difference between ﬁrms with afﬁliated management and those without afﬁliated management. We see that the mean TQ for the full sample is 1.6781, and ﬁrms with 182 Zheng et al. Table 3. Descriptive Statistics. Number of Mean Min. Median Max. Std. Observations Panel A Full Sample AM 0.5726 0.0000 1.0000 1.0000 0.4947 10,159 TQ 1.6781 0.9104 1.3648 6.9431 0.9519 9,975 LoanRPT 0.0419 0.0000 0.0000 0.4089 0.0852 10,156 LoanRPT_FREQ 1.1181 0.0000 0.0000 105.0000 3.7808 10,159 IRREG 0.0405 0.0000 0.0000 1.0000 0.1972 10,094 Gr_Sales 0.1806 –0.2169 0.1425 0.6920 0.2721 10,141 IND_Q 1.4837 0.9995 1.3088 2.8305 0.4082 10,159 SIZE 21.2015 18.7185 21.1274 24.0273 1.0305 10,156 LEV 0.2298 0.0000 0.2187 0.8149 0.1620 10,129 ROA 0.0319 –0.4623 0.0386 0.2603 0.0922 10,156 AGE 6.8474 0.0000 7.0000 19.0000 4.2336 10,159 SOE 0.6270 0.0000 1.0000 1.0000 0.4836 10,118 TOP1 39.2387 9.4500 37.3700 75.0000 16.1841 10,159 Herf_2_5 0.0058 0.0000 0.0010 0.0493 0.0098 10,149 BSIZE 2.2136 1.6094 2.1972 2.7081 0.2156 10,059 OUTDIR 0.3172 0.0000 0.3333 0.5000 0.1029 10,054 DUAL 0.1437 0.0000 0.0000 1.0000 0.3508 10,082 Panel B Firms with AM TQ 1.6613 0.9104 1.3629 6.9431 0.9191 5,685 LoanRPT 0.0463 0.0000 0.0000 0.4089 0.0877 5,815 LoanRPT_FREQ 1.3431 0.0000 0.0000 105.0000 4.2497 5,817 IRREG 0.0417 0.0000 0.0000 1.0000 0.2000 5,777 Gr_Sales 0.1944 –0.2169 0.1564 0.6920 0.2744 5,809 IND_Q 1.5070 0.9995 1.3164 2.8305 0.4252 5,817 SIZE 21.2438 18.7185 21.1612 24.0273 1.0175 5,815 LEV 0.2279 0.0000 0.2188 0.8149 0.1566 5,800 ROA 0.0386 –0.4623 0.0425 0.2603 0.0840 5,815 AGE 6.6582 0.0000 7.0000 19.0000 4.3186 5,817 SOE 0.5633 0.0000 1.0000 1.0000 0.4960 5,800 TOP1 39.6242 9.4500 38.3600 75.0000 15.8399 5,817 Herf_2_5 0.0054 0.0000 0.0009 0.0493 0.0095 5,808 BSIZE 2.2129 1.6094 2.1972 2.7081 0.2143 5,763 OUTDIR 0.3244 0.0000 0.3333 0.5000 0.0974 5,760 DUAL 0.1289 0.0000 0.0000 1.0000 0.3351 5,780 Panel C Firms without AM TQ 1.7004 0.9104 1.3654 6.9431 0.9934 4,290 LoanRPT 0.0361 0.0000 0.0000 0.4089 0.0813 4,341 LoanRPT_FREQ 0.8167 0.0000 0.0000 72.0000 3.0157 4,342 IRREG 0.0389 0.0000 0.0000 1.0000 0.1934 4,317 Gr_Sales 0.1620 –0.2169 0.1248 0.6920 0.2678 4,332 IND_Q 1.4525 0.9995 1.3033 2.8305 0.3822 4,342 SIZE 21.1449 18.7185 21.0794 24.0273 1.0450 4,341 LEV 0.2323 0.0000 0.2185 0.8149 0.1688 4,329 ROA 0.0230 –0.4623 0.0334 0.2603 0.1016 4,341 AGE 7.1009 0.0000 7.0000 19.0000 4.1039 4,342 (Continued) China Journal of Accounting Studies 183 Table 3. (Continued). Number of Mean Min. Median Max. Std. Observations SOE 0.7126 0.0000 1.0000 1.0000 0.4526 4,318 TOP1 38.7222 9.4500 36.1050 75.0000 16.6220 4,342 Herf_2_5 0.0064 0.0000 0.0012 0.0493 0.0102 4,341 BSIZE 2.2145 1.6094 2.1972 2.7081 0.2174 4,296 OUTDIR 0.3076 0.0000 0.3333 0.5000 0.1092 4,294 DUAL 0.1636 0.0000 0.0000 1.0000 0.3700 4,302 Panel D The difference between ﬁrms with AM and ﬁrms without AM Firms with AM-Firms Wilcoxon without AM T value Z value Mean Median difference difference TQ –0.0391 –0.0025 –2.01** –1.49 LoanRPT 0.0102 0.0000 5.96*** 8.72*** LoanRPT_FREQ 0.5264 0.0000 6.96*** 9.92*** IRREG 0.0028 0.0000 0.71 0.71 Gr_Sales 0.0324 0.0316 5.95*** 5.90*** IND_Q 0.0545 0.0127 6.66*** 4.59*** SIZE 0.0988 0.0818 4.79*** 4.40*** LEV –0.0044 0.0003 –1.36 –0.02 ROA 0.0156 0.0091 8.44*** 8.82*** AGE –0.4427 0.0000 –5.22*** –5.27*** SOE –0.1493 0.0000 –15.54*** –15.36*** TOP1 0.9020 2.2550 2.78*** 3.25*** Herf_2_5 –0.0009 –0.0003 –4.78*** –3.25*** BSIZE –0.0016 0.0000 –0.37 –0.18 OUTDIR 0.0168 0.0000 8.16*** 8.03*** DUAL –0.0347 0.0000 –4.93*** –4.92*** Notes: Firms with AM represents the ﬁrms that have afﬁliated management. Please refer to Table 1 for the deﬁnitions of variables. All continuous variables have been winsorized at the 1% level. ***, ** and * denote signiﬁcance at 1, 5 and 10% respectively. afﬁliated management clearly have lower TQ than ﬁrms without afﬁliated management. The mean TQ for the ﬁrms with afﬁliated management is 1.6613, while that for the ﬁrms without afﬁliated management is 1.7004. The difference in the means is –0.0391, which is signiﬁcant at the 5% level. The mean LoanRPTs and mean LoanRPTs-FREQ for the ﬁrms with afﬁliated management are 0.0463 and 1.3431 respectively, while those for the ﬁrms without afﬁliated management are 0.0361 and 0.8167 respectively, the differences of 0.0102 and 0.5264 are signiﬁcant at the 1% level. The foregoing uni- variate tests seem to suggest that ﬁrms with afﬁliated management have lower ﬁrm value and higher value and frequency of loan-based related party transactions than ﬁrms without afﬁliated management. Table 4 presents the correlation matrix of main variables. The Pearson correlation between AM and TQ is –0.0203 (signiﬁcant at the 5% level), which implies that afﬁliated management is negatively associated with ﬁrm value. Meanwhile, AM is positively corre- lated with LoanRPT and LoanRPT_FREQ at the 1% signiﬁcance level, which implies that afﬁliated management is positively associated with the value and the frequency of loan- based related party transactions between controlling shareholders and listed ﬁrms. 184 Zheng et al. Table 4. Correlation matrix of main variables. 　 AM TQ LoanRPT LoanRPT_FREQ IRREG Gr_Sales IND_Q SIZE LEV ROA AGE SOE TOP1 Herf_2_5 BSIZE OUTDIR DUAL AM −0.0150 0.0865 0.0984 0.0070 0.0586 0.0455 0.0437 0.0002 0.0875 −0.0523 −0.1527 0.0322 −0.0323 −0.0018 0.0801 −0.0490 TQ −0.0203 −0.1141 −0.1073 0.0039 −0.1141 0.6780 −0.3583 −0.1725 0.1821 0.1221 −0.1346 −0.2065 0.0803 −0.1274 −0.0117 0.0865 LoanRPT 0.0590 −0.0779 0.9695 −0.0009 0.0497 −0.0582 0.1196 0.2405 −0.0968 0.0153 0.0130 0.1195 −0.1571 0.0067 0.0852 −0.0415 LoanRPT_FREQ 0.0689 −0.0479 0.4698 −0.0043 0.0525 −0.0369 0.1363 0.2179 −0.0875 0.0231 0.0023 0.1038 −0.1438 0.0127 0.0930 −0.0425 IRREG 0.0070 −0.0013 0.0237 0.0064 −0.0647 −0.0185 −0.0901 0.0744 −0.1354 −0.0062 −0.0770 −0.0563 0.0498 −0.0221 −0.0259 −0.0001 Gr_Sales 0.0590 −0.0967 0.0258 0.0105 −0.0544 −0.0798 0.2307 0.0146 0.3109 −0.0751 0.0389 0.1109 −0.0157 0.0687 0.0064 −0.0233 IND_Q 0.0660 0.5417 −0.0412 0.0330 −0.0339 −0.0599 −0.0247 −0.1272 0.1488 0.0587 −0.0696 −0.0872 −0.0111 −0.0875 −0.0547 0.0633 SIZE 0.0475 −0.3381 0.0675 0.1229 −0.0937 0.2146 −0.0079 0.1487 0.1111 0.1741 0.2452 0.1976 −0.2159 0.2076 0.0586 −0.0910 LEV −0.0135 −0.0560 0.2468 0.1326 0.0800 −0.0202 −0.1115 0.0710 −0.3668 0.1260 0.0114 −0.0730 −0.0505 0.0286 0.0059 −0.0608 ROA 0.0835 −0.0017 −0.0835 −0.0365 −0.1621 0.2935 0.1243 0.1836 −0.3785 −0.2191 −0.0761 0.1494 0.0690 0.0316 0.0201 0.0521 AGE −0.0517 0.2050 0.0035 0.0327 −0.0052 −0.0782 0.1116 0.1500 0.1356 −0.1572 0.0751 −0.2182 −0.1661 −0.0530 0.1584 −0.0640 SOE −0.1527 −0.1552 0.0284 0.0111 −0.0770 0.0225 −0.1016 0.2503 −0.0114 −0.0088 0.0725 0.3080 −0.2557 0.1925 −0.1593 −0.1557 TOP1 0.0276 −0.2103 0.1078 0.0411 −0.0554 0.1052 −0.1116 0.2112 −0.0921 0.1417 −0.2197 0.3033 −0.5085 0.0178 −0.1051 −0.0553 Herf_2_5 −0.0474 −0.0337 −0.1046 −0.0684 0.0223 −0.0289 −0.0337 −0.0992 −0.0605 0.0386 −0.1918 −0.1153 −0.2741 0.0528 0.0524 0.0821 BSIZE −0.0037 −0.1316 −0.0171 0.0134 −0.0156 0.0669 −0.0950 0.2254 0.0051 0.0708 −0.0512 0.1898 0.0143 0.0481 −0.1884 −0.0883 OUTDIR 0.0811 0.0298 0.0797 0.0649 −0.0386 0.0224 0.0035 0.0741 0.0114 0.0144 0.2018 −0.1666 −0.1154 0.0138 −0.1351 0.0733 DUAL −0.0490 0.0752 −0.0358 −0.0240 −0.0001 −0.0209 0.0767 −0.0967 −0.0524 0.0245 −0.0615 −0.1557 −0.0578 0.0540 −0.0816 0.0532 　 Notes: Person (Spearman) coefﬁcients are below (above) the diagonal. Bolded coefﬁcients are signiﬁcant at p<10% (two-tailed test). Please refer to Table 1 for the deﬁnitions of variables. All continuous variables have been winsorized at the 1% level. China Journal of Accounting Studies 185 5.2. Afﬁliated management and ﬁrm value In Table 5, we investigate the association of afﬁliated management with ﬁrm value through regression analysis. All models in Table 5 are statistically signiﬁcant, with adjusted R squared higher than 40%, showing that our models explain 40% of the variation in ﬁrm value. After controlling for growth, size, leverage and proﬁtability, model 1 shows that afﬁliated management is negatively associated with ﬁrm value at the 1% signiﬁcance level, which means that ﬁrms with afﬁliated management have lower ﬁrm value. Consistent with Hypothesis 1, our result implies that, since controlling shareholders arrange afﬁliated management to get private beneﬁt of con- trol, afﬁliated management is negatively associated with ﬁrm value. Meanwhile, the coefﬁcients of our control variables are consistent with previous literature: growth, measured by Gr_Sales and IND_Q, is positively associated with ﬁrm value; ﬁrm size, measured by SIZE is negatively associated with ﬁrm value (King & Segal, 2009). After we control for ownership structure and board characteristics respec- tively in models 2 and 3, the coefﬁcient of AM is still signiﬁcantly negative at the 1% level, which supports our Hypothesis 1. The regression results in Table 5, descriptive statistics in Table 3 and the correlation matrix in Table 4 imply that although afﬁliated management may play a positive role in improving the ﬁrm’s operation efﬁciency, the cost that afﬁliated management brings about exceeds the beneﬁts, so that ﬁrms with afﬁliated management have lower ﬁrm value. Compared with the sample average of ﬁrm value (1.6781), the value of ﬁrms with afﬁliated management is 4.8% lower. 5.3. Channels through which afﬁliated management affect ﬁrm value If the controlling shareholders arrange afﬁliated management in the listed ﬁrms to enhance their control for private beneﬁts, related party transactions may be one channel through which the afﬁliated management can help the controlling share- holder to expropriate the listed ﬁrms. Fisman and Wang (2010) ﬁnd that the loan- based related party transactions (usually a loan guarantee from the listed ﬁrms) between listed ﬁrms and the controlling shareholders is the main form of RPTs that reduce the ﬁrm value. Based on their ﬁndings, we mainly focus on the relation between afﬁliated management and the loan-based related party transactions. The dependent variable LoanRPT in model 1 of Table 6 measures the value of loan- based related party transactions, deﬁned as the sum of annual loan-based related party transitions between listed ﬁrms and controlling shareholders, divided by total assets. For ﬁrms without these related party transactions, LoanRPT equals zero. The dependent variable LoanRPT_FREQ in model 2 of Table 6 is the frequency of loan-based RPTs in a calendar year. To be consistent with Arnoldi et al. (2013), we also control for AGE in the two models. AGE measures how long the ﬁrm has been listed. Considering that the loan-based RPTs are the main form of RPTs that harm ﬁrm value (Fisman & Wang, 2010), we expect that the afﬁliated management is positively associated with the value and the frequency of loan-based RPTs. The empirical results in Table 6 support our expectation. Compared with the sample average (0.0419), the value of loan-based RPTs with a controlling shareholder increases about 13% in the listed ﬁrms with afﬁliated management. Since the tunnelling behaviour of controlling shareholders will inevitably reduce ﬁrm performance in the long run, the controlling shareholders may disguise their illegal 186 Zheng et al. Table 5. Results for the OLS regression of afﬁliated management and ﬁrm value. (1) (2) (3) Model 1 Model 2 Model 3 AM −0.0847*** −0.0836*** −0.0811*** (4.03) (3.98) (3.91) Gr_Sales 0.0368 0.0462 0.0586* (1.04) (1.34) (1.70) IND_Q 0.9400*** 0.9280*** 0.9240*** (10.55) (10.45) (10.39) SIZE −0.3640*** −0.3450*** −0.3410*** (17.34) (17.81) (17.66) LEV 0.1860 0.0781 0.0809 (1.44) (0.66) (0.69) ROA 0.2310 0.2320 0.2260 (1.08) (1.11) (1.06) SOE −0.0075 0.0023 (0.29) (0.09) TOP1 −0.0055*** −0.0055*** (5.95) (5.93) Herf_2_5 −6.1680*** −6.2750*** (4.43) (4.56) BSIZE 0.0200 (0.38) OUTDIR 0.2430 (1.15) DUAL 0.0287 (0.85) IND Control Control Control YEAR Control Control Control adj. R 0.4370 0.4450 0.4440 F value 91.8300*** 84.9000*** 77.5000*** N 9,933 9,899 9,729 Notes: The dependent variable is TQ,deﬁned as the sum of book value of liability, market value of liquid shares and book value of illiquid shares, divided by book value of total assets. Please refer to Table 1 for the deﬁnitions of the other variables. Each column reports the result of a linear regression with robust standard error clustering and White (1980) heteroscedasticity adjustment at the listed ﬁrm level. Absolute value of T statistics is reported below each coefﬁcient. ***, ** and * denote signiﬁcance at 1, 5 and 10% respectively. behaviour through manipulating information disclosure while arranging for afﬁliated management to seize the private beneﬁt of control. In practice, afﬁliated management may also directly help the controlling shareholders expropriate the beneﬁts of other shareholders through manipulation of information disclosure. For example, afﬁliated management may exaggerate proﬁts to purposely increase the price of a seasoned equity offering. Ultimately, such manipulating behaviour in respect of information dis- closure may evolve into a great amount of irregularity in information disclosure. We use Logit regression to test the association of afﬁliated management with the likelihood of irregularity in information disclosure. If CSRC reports that the listed ﬁrm i in year t manipulated information disclosure, including fraudulent accounting treatment (such as exaggerated proﬁt, and false statement of assets), misleading statements, and omission of substantial information, the dummy variable IRREG of the ﬁrm i in year t equals 1. Otherwise, IRREG equals 0. The regression in Table 7 shows that ﬁrms with afﬁliated China Journal of Accounting Studies 187 Table 6. Results for OLS regression of afﬁliated management and loan-based RPTs. (1) (2) LoanRPT LoanRPT_FREQ AM 0.0056** 0.2910*** (2.18) (2.98) Gr_Sales 0.0034 −0.0080 (0.92) (0.06) IND_Q 0.0030 −0.2420 (0.43) (0.64) SIZE 0.0011 0.3330*** (0.70) (4.51) LEV 0.1400*** 3.1290*** (13.43) (8.14) ROA −0.0078 −0.7850* (0.53) (1.66) AGE −0.0012*** −0.0351** (3.01) (2.54) SOE 0.0046 0.1500 (1.34) (1.08) TOP1 0.0006*** 0.0108*** (5.81) (2.81) Herf_2_5 −0.4620*** −12.5400*** (3.43) (2.87) BSIZE −0.0065 0.0946 (0.99) (0.32) OUTDIR −0.0023 −0.8900 (0.14) (1.30) DUAL −0.0045 −0.1570 (1.43) (1.20) IND Control Control YEAR Control Control adj. R 0.1170 0.0630 F value 9.6560*** 5.2970*** N 9,883 9,883 Notes: The dependent variable in column (1) is LoanRPT,deﬁned as sum of annual loan-based related party transitions between listed ﬁrms and controlling shareholders, divided by total assets. The dependent variable in column (2) is the frequency of loan-based related party transactions in a calendar year. Please refer to Table 1 for the deﬁnitions of the other variables. Each column reports the result of a linear regression with robust standard error clustering and White (1980) heteroscedasticity adjustment at the listed ﬁrm level. Abso- lute value of T statistics is reported below each coefﬁcient. ***, ** and * denote signiﬁcance at 1, 5 and 10% respectively. management are more likely to manipulate information disclosure, which supports our view that the afﬁliated management is negatively associated with the ﬁrm value. 6. Further analysis and robustness tests 6.1. Impact of afﬁliated management on ﬁrm value in different subsamples The previous empirical results support our view that afﬁliated management is nega- tively associated with the ﬁrm value, but the degree to which the afﬁliated management correlates with ﬁrm value may differ under different circumstances. First of all, we use a dummy variable Control to measure whether or not the largest shareholder has abso- lute control over the ﬁrm. If the largest shareholder hold over 50% shares of the listed 188 Zheng et al. Table 7. Results for the Logit regression of afﬁliated management and irregularity. (1) (2) (3) Model 1 Model 2 Model 3 AM 0.3650*** 0.3140*** 0.3420*** (3.34) (2.82) (3.01) SIZE −0.3020*** −0.1950*** −0.1880*** (5.06) (3.13) (2.93) ROA −4.0030*** −4.0470*** −4.0880*** (8.98) (8.96) (8.80) LEV 0.9610*** 0.7520** 0.9110*** (3.02) (2.33) (2.77) SOE −0.6480*** −0.6410*** (5.47) (5.26) TOP1 −0.0071* −0.0066 (1.79) (1.62) Herf_2_5 2.8930 3.5750 (0.50) (0.61) BSIZE 0.1940 (0.75) OUTDIR 0.3390 (0.42) DUAL −0.0639 (0.40) IND Control Control Control YEAR Control Control Control pseudo R 0.1120 0.1250 0.1250 Chi-Square 380.7000*** 425.6000*** 416.1000*** N 10,022 9,972 9,794 Notes: The dependent variable of the Logit regression is IRREG. If CSRC reports that the listed ﬁrm i in year t manipulated information disclosure, including fraudulent accounting treatment (such as exaggerated proﬁt, false statement of assets), misleading statements, and omission of substantial information, the dummy variable IRREG of the ﬁrm i in year t equals 1. Otherwise, IRREG equals 0. Please refer to Table 1 for the deﬁnitions of the other variables. Each column reports the result of a Logit regression with robust standard error clustering at the listed ﬁrm level. Absolute value of Z statistics is reported below each coefﬁcient. ***, ** and * denote signiﬁcance at 1, 5 and 10% respectively. ﬁrms, Control equals 1; otherwise, Control equals 0. We ﬁnd that the coefﬁcient of AM*Control is positive but not signiﬁcant. The ﬁnding implies that the degree to which the afﬁliated management correlates with the ﬁrm value does not signiﬁcantly vary according to whether the largest shareholder has absolute control or not (Table 8). Second, we use a dummy variable Institutional Investor to measure the level of institutional investor holdings. If the level of institutional investor holdings is above the sample average, Institutional Investor equals 1; otherwise, Institutional Investor equals 0. We ﬁnd that the coefﬁcient of AM*Institutional Investor is positive at the 1% signiﬁ- cance level, implying that the degree to which afﬁliated management negatively corre- lates with ﬁrm value is smaller in the ﬁrms with a high level of institutional investor holdings (Table 9). Last of all, since CSRC requires the Chinese listed ﬁrms to reach certain perfor- mance standards if they want to start a seasoned equity offering (SEO) (Li, Yu, & Wang, 2005; Liu, Wei, & Zheng, 2008), the controlling shareholder may choose not to expropriate wealth from a listed ﬁrm immediately before the SEO. We divide our sam- ple into two subgroups according to whether the observation is before the SEO or after China Journal of Accounting Studies 189 Table 8. Results for the sub-group regression: Largest shareholders with and without absolute control. (1) (2) (3) Model 1 Model 2 Model 3 AM −0.0993*** −0.0899*** −0.0875*** (3.73) (3.47) (3.41) AM*Control 0.0517 0.0359 0.0353 (1.22) (0.86) (0.85) Control −0.0414 0.2260*** 0.2220*** (1.11) (4.57) (4.50) Gr_Sales 0.0365 0.0447 0.0573* (1.03) (1.30) (1.67) IND_Q 0.9410*** 0.9340*** 0.9300*** (10.56) (10.54) (10.48) SIZE −0.3630*** −0.3480*** −0.3440*** (16.98) (18.03) (17.91) LEV 0.1830 0.0853 0.0883 (1.40) (0.73) (0.76) ROA 0.2360 0.2440 0.2340 (1.10) (1.17) (1.11) SOE −0.0055 0.0031 (0.21) (0.12) TOP1 −0.0109*** −0.0108*** (7.96) (7.90) Herf_2_5 −5.3650*** −5.5010*** (3.93) (4.07) BSIZE 0.0330 (0.63) OUTDIR 0.2530 (1.22) DUAL 0.0300 (0.89) IND Control Control Control YEAR Control Control Control adj. R 0.4370 0.4500 0.4490 F value 86.9100*** 80.8200*** 74.0700*** N 9,933 9,899 9,729 Notes: In all columns, the dependent variable is TQ,deﬁned as the sum of book value of liability, market value of liquid shares and book value of illiquid shares, divided by book value of total assets. If the largest shareholder hold over 50% shares of the listed ﬁrms, Control equals 1, otherwise, 0. Please refer to Table 1 for the deﬁnitions of the other variables. Each column reports the result of a linear regression with robust standard error clustering and White (1980) heteroscedasticity adjustment at the listed ﬁrm level. Absolute value of T statistics is reported below each coefﬁcient. ***, ** and * denote signiﬁcance at 1, 5 and 10% respectively. the SEO. The results in Table 10 shows that the negative effect of afﬁliated manage- ment on ﬁrm value mainly happens in the years after the SEO, which implies that afﬁl- iated managers are more likely to be negatively associated with ﬁrm value when the controlling shareholders have a higher motivation to expropriate the value of the ﬁrm. 6.2. Ownership and the impact of afﬁliated management on ﬁrm value We also investigate the different effect of afﬁliated managers on state-owned listed ﬁrms and non-state-owned listed ﬁrms. In model 1 of Table 11, we only considered the 190 Zheng et al. Table 9. Results for the sub-group regression: High and low institutional investor holdings. (1) (2) (3) Model 1 Model 2 Model 3 AM −0.1730*** −0.1670*** −0.1660*** (6.25) (5.95) (5.97) AM*Institutional Investor 0.1720*** 0.1670*** 0.1670*** (4.28) (4.16) (4.19) Institutional Investor 0.2170*** 0.2000*** 0.2050*** (6.67) (6.29) (6.42) Gr_Sales 0.0267 0.0436 0.0480 (0.67) (1.14) (1.24) IND_Q 0.9470*** 0.9370*** 0.9220*** (10.82) (10.68) (10.61) SIZE −0.3700*** −0.3500*** −0.3460*** (15.87) (16.88) (16.81) LEV 0.0049 −0.1020 −0.0681 (0.03) (0.83) (0.55) ROA 0.2320 0.2120 0.2830 (0.92) (0.89) (1.17) SOE −0.0091 0.0041 (0.35) (0.15) TOP1 −0.0043*** −0.0044*** (4.64) (4.72) Herf_2_5 −4.4030*** −4.4930*** (3.05) (3.15) BSIZE −0.0092 (0.16) OUTDIR 0.1820 (0.68) DUAL 0.0524 (1.41) IND Control Control Control YEAR Control Control Control adj. R 0.4750 0.4800 0.4820 F value 79.6100*** 73.8700*** 68.2200*** N 7,967 7,935 7,803 Notes: In all the columns, the dependent variable is TQ,deﬁned as the sum of book value of liability, market value of liquid shares and book value of illiquid shares, divided by book value of total assets. If the level of institutional investor holdings is above the sample average, Institutional Investor equals 1, otherwise, 0. Please refer to Table 1 for the deﬁnitions of the other variables. Each column reports the result of a linear regression with robust standard error clustering and White (1980) heteroscedasticity adjustment at the listed ﬁrms level. Absolute value of T statistics is reported below each coefﬁcient. ***, ** and * denote signiﬁcance at 1, 5 and 10% respectively. impact of the afﬁliated managers on ﬁrm value in state-owned listed ﬁrms, and we ﬁnd that even though afﬁliated managers negatively correlate with the value of the ﬁrm, the coefﬁcient is not signiﬁcant. Model 2 analyses the impact of afﬁliated managers on ﬁrm value in non-state-owned listed ﬁrms. We ﬁnd that the presence of afﬁliated managers is negatively correlated with ﬁrm value, and the coefﬁcient is signiﬁcant at the 1% level. Compared with the mean of TobinQ in non-state-owned listed ﬁrms (1.87), the negative effect of afﬁliated managers on ﬁrm value is 8.98%. In model 3, the coefﬁ- cient of AM is signiﬁcantly negative and the coefﬁcient of AM* SOE is signiﬁcantly positive. The absolute value of the two coefﬁcients is quite close, which implies that China Journal of Accounting Studies 191 Table 10. Results for the sub-group regression: Before and after seasoned equity offering. (1) (2) 　 (3) (4) SEO Non-SEO SEO(-2,0) Non-SEO(-2,0) AM 0.0153 −0.0846*** 　 −0.0415 −0.0866*** (0.22) (5.64) (1.19) (5.42) Gr_Sales −0.0869 0.0584* 0.0149 0.0436 (0.62) (1.78) (0.21) (1.26) IND_Q 0.9560*** 0.9170*** 0.9500*** 0.9080*** (3.39) (10.19) (4.71) (9.80) SIZE −0.2310*** −0.3470*** −0.1910*** −0.3620*** (4.48) (28.70) (8.80) (28.02) LEV −0.5490* 0.0995 −0.3600*** 0.1320 (1.79) (1.26) (2.74) (1.56) ROA 4.0290*** 0.1750 3.1500*** 0.1220 (2.99) (0.91) (5.04) (0.62) SOE 0.0555 0.0031 0.0276 0.0049 (0.71) (0.18) (0.72) (0.26) TOP1 −0.0104*** −0.0053*** −0.0069*** −0.0053*** (4.96) (9.58) (5.78) (8.94) Herf_2_5 0.1880 −6.4040*** −0.6540 −6.7090*** (0.03) (8.56) (0.26) (8.58) BSIZE 0.0711 0.0249 0.0679 0.0315 (0.45) (0.68) (0.80) (0.82) OUTDIR −0.3590 0.2860** 0.0861 0.2730* (0.46) (1.97) (0.25) (1.75) DUAL 0.0973 0.0261 0.0602 0.0224 　 (0.97) (1.04) 　 (1.25) (0.83) IND Control Control Control Control YEAR Control Control 　 Control Control adj. R 0.4080 0.4510 0.4650 0.4530 F value 7.3420*** 90.9000*** 16.7200*** 83.8000*** N 509 9,220 　 1,284 8,445 Notes: Column (1) reports the regression result of the sub-samples just in the SEO year. Column (2) reports the regression result of the sub-samples in the Non-SEO year. Column (3) reports the regression result of the sub-samples in the SEO year and two years before the SEO year. Column (4) reports the regression result of the sub-samples which are not in the SEO year or two years before the SEO year. In all the columns, the dependent variable is TQ,deﬁned as the sum of book value of liability, market value of liquid shares and book value of illiquid shares, divided by book value of total assets. Please refer to Table 1 for the deﬁnitions of the other variables. Each column reports the result of a linear regression with robust standard error cluster- ing and White (1980) heteroscedasticity adjustment at the listed ﬁrm level. Absolute value of T statistics is reported below each coefﬁcient. ***, ** and * denote signiﬁcance at 1%, 5% and 10% respectively. the negative effect of afﬁliated managers on ﬁrm value mainly happens in the non- state-owned listed ﬁrms. One possible explanation is that compared with the controlling shareholders (i.e. the state) in the state-owned listed ﬁrms, the controlling shareholders (usually a person or a family) in the non-state-owned listed ﬁrms are more prone to expropriate the minority shareholders with the help of afﬁliated managers for their own private beneﬁts given the weak institutional environment. Meanwhile, the afﬁliated management in the state-owned listed ﬁrms is under more severe monitoring by the government than the afﬁliated management in the non-state-owned listed ﬁrms; thus, the afﬁliated managers’ behaviours that harm ﬁrm value are restrained more strictly in the state-owned listed ﬁrms. 192 Zheng et al. Table 11. Results for the sub-group regression: State-owned listed ﬁrms and non-state-owned listed ﬁrms. (1) (2) (3) SOE Non-SOE Full sample AM −0.0239 −0.1680*** −0.1950*** (1.18) (4.21) (4.49) AM*SOE 0.1720*** (3.61) SOE −0.1020** (2.50) Gr_Sales 0.0147 0.1390** 0.0581* (0.38) (2.35) (1.68) IND_Q 0.8150*** 0.9600*** 0.9230*** (7.68) (6.78) (10.39) SIZE −0.2590*** −0.5010*** −0.3420*** (15.04) (13.70) (17.70) LEV −0.1440 0.3910** 0.0814 (1.31) (1.98) (0.70) ROA 0.3710 0.2850 0.2660 (1.34) (0.99) (1.26) TOP1 −0.0038*** −0.0109*** −0.0055*** (4.31) (5.91) (5.90) Herf_2_5 −2.2130 −14.2600*** −6.0230*** (1.61) (6.85) (4.41) BSIZE −0.0175 0.0405 0.0189 (0.33) (0.37) (0.36) OUTDIR −0.0017 0.5730 0.2510 (0.01) (1.39) (1.20) DUAL 0.0831* −0.0244 0.0306 (1.86) (0.54) (0.90) IND Control Control Control YEAR Control Control Control adj. R 0.4300 0.4800 0.4460 F value 50.6900*** 39.6100*** 76.0700*** N 6,171 3,558 9,729 Notes: Column (1) reports the regression result of the state-owned listed ﬁrms. Column (2) reports the regres- sion result of the non-state-owned listed ﬁrms. Column (3) reports the regression result of the full sample. In all the columns, the dependent variable is TQ,deﬁned as the sum of book value of liability, market value of liquid shares and book value of illiquid shares, divided by book value of total assets. SOE equals 1 if the ﬁrm is controlled by the state and 0 otherwise. Please refer to Table 1 for the deﬁnitions of the other variables. Each column reports the result of a linear regression with robust standard error clustering and White (1980) heteroscedasticity adjustment at the listed ﬁrm level. Absolute value of T statistics is reported below each coefﬁcient. ***, ** and * denote signiﬁcance at 1%, 5% and 10% respectively. 6.3. Different types of afﬁliated management and ﬁrm value First, we divide afﬁliated management into different types based on whether the Chairman (CEO) in the listed ﬁrm is afﬁliated management or not (Table 12). If the Chairman (CEO) in the listed ﬁrm is the controlling shareholder, or the employee of the controlling shareholder, the ﬁrm-year dummy variable AM_Chairman (AM_CEO) equals 1. Otherwise, AM_Chairman (AM_CEO) equals 0. Our empirical results show that no matter whether the Chairman is the afﬁliated management or the CEO is the afﬁliated management, afﬁliated management is negatively associated with ﬁrm value. Secondly, we divide afﬁliated management into two new types, based on whether the afﬁliated management in the listed ﬁrm also holds the chief position of the China Journal of Accounting Studies 193 Table 12. Results for the sub-group regression: Chairman and CEO. (1) (2) Chairman CEO AM_Chairman −0.0827*** (5.59) AM_CEO −0.0841*** (4.24) Gr_Sales 0.0639** 0.0709* (1.98) (1.77) IND_Q 0.9250*** 0.8510*** (10.59) (7.79) SIZE −0.3430*** −0.3440*** (28.76) (23.05) LEV 0.0891 0.1100 (1.13) (1.13) ROA 0.2160 0.0565 (1.10) (0.24) SOE 0.0009 −0.0091 (0.05) (0.41) TOP1 −0.0054*** −0.0053*** (9.95) (7.77) Herf_2_5 −6.1020*** −5.3180*** (8.15) (5.72) BSIZE 0.0286 −0.0110 (0.80) (0.25) OUTDIR 0.2340 0.3280* (1.59) (1.83) DUAL 0.0287 0.0228 (1.18) (0.89) IND Control Control YEAR Control Control adj. R 0.4440 0.4530 F value 93.4500*** 62.6400*** N 9,571 6,096 Notes: Column (1) reports the regression result of the sub-samples that the Chairman is the afﬁliated manage- ment. Column (2) reports the regression result of the sub-samples that the CEO is the afﬁliated management. In all the columns, the dependent variable is TQ,deﬁned as the sum of book value of liability, market value of liquid shares and book value of illiquid shares, divided by book value of total assets. If the Chairman (CEO) in the listed ﬁrm is the controlling shareholder, or the employee of the controlling shareholder, the ﬁrm-year dummy variable AM_Chairman (AM_CEO) equals 1. Otherwise, AM_Chairman (AM_CEO) equals 0. Please refer to Table 1 for the deﬁnitions of the other variables. Each column reports the result of a linear regression with robust standard error clustering and White (1980) heteroscedasticity adjustment at the listed ﬁrm level. Absolute value of T statistics is reported below each coefﬁcient. ***, ** and * denote signiﬁcance at 1%, 5% and 10% respectively. controlling shareholder or not (Table 13). If the afﬁliated management in the listed ﬁrm also holds the chief position of the controlling shareholder (such as Chairman, or CEO), the ﬁrm-year dummy variable AM_Chief equals 1, otherwise 0. If the afﬁliated management in the listed ﬁrms also holds the vice position of the controlling share- holder (such as Vice Chairman, Vice CEO, Vice Executive, or Directors), the ﬁrm-year dummy variable AM_Vice equals 1, otherwise, 0. The empirical study ﬁnds that no mat- ter whether the afﬁliated managers in the listed ﬁrms are in the chief position or the vice position of the controlling shareholders, the afﬁliated management harms the ﬁrm value signiﬁcantly. 194 Zheng et al. Table 13. Results for the sub-group regression: Chief and vice positions in controlling share- holders. (1) (2) Chief position Vice position AM_Chief −0.0580*** (3.57) AM_Vice −0.0934*** (5.36) Gr_Sales 0.0504 0.1020*** (1.33) (2.81) IND_Q 0.9280*** 0.8800*** (9.29) (8.62) SIZE −0.3530*** −0.3450*** (27.00) (24.81) LEV 0.1720** 0.0143 (1.97) (0.16) ROA 0.2800 0.0045 (1.31) (0.02) SOE −0.0050 −0.0152 (0.26) (0.75) TOP1 −0.0050*** −0.0060*** (8.11) (9.30) Herf_2_5 −5.8440*** −5.4340*** (6.92) (6.27) BSIZE 0.0201 −0.0125 (0.50) (0.29) OUTDIR 0.3980** 0.2770 (2.57) (1.64) DUAL 0.0779*** −0.0062 (2.69) (0.24) IND Control Control YEAR Control Control adj. R 0.4540 0.4460 F value 75.9700*** 72.4200*** N 7,628 7,295 Notes: Column (1) reports the regression result of the sub-samples that the afﬁliated management in the listed ﬁrm also holds the chief position of the controlling shareholder. Column (2) reports the regression result of the sub-samples that the afﬁliated management in the listed ﬁrms also holds the vice position of the control- ling shareholder. In all the columns, the dependent variable is TQ,deﬁned as the sum of book value of liabil- ity, market value of liquid shares and book value of illiquid shares, divided by book value of total assets. If the afﬁliated management in the listed ﬁrm also holds the chief position of the controlling shareholder (such as Chairman, or CEO), the ﬁrm-year dummy variable AM_Chief equals 1, otherwise, 0. If the afﬁliated man- agement in the listed ﬁrms also holds the vice position of the controlling shareholder (such as Vice Chairman, Vice CEO, Vice Executive, or Directors), the ﬁrm-year dummy variable AM_Vice equals 1, otherwise, 0. Please refer to Table 1 for the deﬁnitions of the other variables. Each column reports the result of a linear regression with robust standard error clustering and White (1980) heteroscedasticity adjustment at the listed ﬁrms level. Absolute value of T statistics is reported below each coefﬁcient. ***, ** and * denote signiﬁcance at 1%, 5% and 10% respectively. 6.4. Afﬁliated management and ﬁrm operation efﬁciency We also test the effect of afﬁliated management on Return on Assets (ROA), Return on Sales (ROS). We deﬁne Return on Assets as net income divided by total assets. Return on Sales is deﬁned as net income divided by sales. China Journal of Accounting Studies 195 Table 14. Results for the OLS regression of afﬁliated management and ROA. (1) (2) (3) Model 1 Model 2 Model 3 AM 0.0076*** 0.0080*** 0.0087*** (4.05) (4.21) (4.54) Gr_Sales 0.0751*** 0.0755*** 0.0753*** (23.62) (23.74) (23.64) IND_Q 0.0189*** 0.0193*** 0.0193*** (3.30) (3.38) (3.42) SIZE 0.0122*** 0.0124*** 0.0121*** (9.75) (9.89) (9.40) LEV −0.1920*** −0.1890*** −0.1880*** (22.03) (21.74) (21.88) TOP1 0.0003*** 0.0004*** 0.0004*** (4.84) (5.84) (5.83) SOE −0.0141*** −0.0134*** −0.0136*** (6.50) (6.17) (6.24) Herf_2_5 0.5090*** 0.4820*** (5.13) (4.84) BSIZE 0.0146*** (3.03) OUTDIR 0.0016 (0.09) DUAL 0.0052* (1.90) IND Control Control Control YEAR Control Control Control adj. R 0.2670 0.2700 0.2760 F value 37.5000*** 36.9000*** 33.9700*** N 10,071 10,061 9,883 Notes: The dependent variable is ROA,deﬁned as net income divided by total assets. Please refer to Table 1 for the deﬁnitions of the other variables. Each column reports the result of a linear regression with robust standard error clustering and White (1980) heteroscedasticity adjustment at the listed ﬁrm level. Absolute value of T statistics is reported below each coefﬁcient. ***, ** and * denote signiﬁcance at 1%, 5% and 10% respectively. The empirical results from Tables 14 and 15 show that the afﬁliated management is positively associated with the level of ROA and ROS. The empirical results support the view that the afﬁliated managers may play a positive role in increasing the proﬁtability in the ﬁrm. However, we also ﬁnd that ﬁrms with afﬁliated management are more likely to manipulate information disclosure, including fraudulent accounting treatment (such as exaggerated proﬁt, false statement of assets), which may cause the increase of ROA and ROS accordingly. Even if the increase of proﬁtability is authentic, the con- trolling shareholder may turn it into the private beneﬁt of control with the help of afﬁli- ated management through related party transactions or other channels, which may result in the negative impact of afﬁliated management on the ﬁrm value eventually. This paper also carries out other robustness tests (unreported for the space reason). Since the non-tradable shares exist in our capital market for a long time, a variety of methods could be chosen to calculate the ﬁrm value. Referring to Bai et al. (2005), we recalculate the value of the ﬁrm (TQ_R). TQ_R is deﬁned as ‘(book value of liabilities+tradable shares*stock price + non tradable shares*discounted stock price)/the book value of total assets’. The empirical results show that the association between the 196 Zheng et al. Table 15. Results for the OLS regression of afﬁliated management and ROS. (1) (2) (3) Model 1 Model 2 Model 3 AM 0.0273*** 0.0281*** 0.0288*** (3.09) (3.18) (3.19) Gr_Sales 0.2440*** 0.2450*** 0.2420*** (13.74) (13.78) (13.44) IND_Q 0.0496 0.0507* 0.0547* (1.62) (1.66) (1.85) SIZE 0.0594*** 0.0597*** 0.0570*** (9.59) (9.62) (9.17) LEV −0.6770*** −0.6700*** −0.6570*** (13.28) (13.12) (13.17) TOP1 0.0010*** 0.0012*** 0.0013*** (3.30) (3.74) (3.94) SOE −0.0152 −0.0137 −0.0172 (1.49) (1.34) (1.60) Herf_2_5 1.1830*** 1.0710** (2.64) (2.44) BSIZE 0.0652** (2.55) OUTDIR −0.0110 (0.13) DUAL 0.0138 (0.99) IND Control Control Control YEAR Control Control Control adj. R 0.1620 0.1630 0.1640 F value 14.3500*** 14.1200*** 12.9900*** N 10,055 10,045 9,871 Notes: The dependent variable is ROS,deﬁned as net income divided by sales. Please refer to Table 1 for the deﬁnitions of the other variables. Each column reports the result of a linear regression with robust standard error clustering and White (1980) heteroscedasticity adjustment at the listed ﬁrm level. Absolute value of T statistics is reported below each coefﬁcient. ***, ** and * denote signiﬁcance at 1%, 5% and 10% respectively. afﬁliated managers and the ﬁrm value remains the same. Meanwhile, after considering the split share structure reform and other factors, the main ﬁndings in this paper remain the same. 7. Conclusions By analysing the data of Chinese listed ﬁrms in A-share stock market from 2001 to 2009, we ﬁnd that the afﬁliated management is negatively associated with ﬁrm value. This empirical result implies that, in order to obtain more private beneﬁt of control, the controlling shareholder may enhance the control over the ﬁrm through afﬁliated man- agement. After further analysing the channels to explore how afﬁliated management is associated with ﬁrm value, we ﬁnd that the afﬁliated management is positively associ- ated with the value and the frequency of loan-based related party transactions and with the likelihood of irregularity in information disclosure in the listed ﬁrms, which imply that even if afﬁliated management may enhance the ability of shareholders to monitor managers, it also facilitates the expropriation of the interests of the minority sharehold- ers by the controlling shareholder in a transition economy. China Journal of Accounting Studies 197 The empirical results in this paper support the conjecture that the presence of a con- trolling shareholder enhances the control over the ﬁrm through afﬁliated management for the reason of obtaining the private beneﬁts of control. Meanwhile, afﬁliated manage- ment, as a prevailing phenomenon in Chinese listed ﬁrms, has attracted a great amount of attention from practitioners and from regulators. Because of the lack of large-sample empirical evidence, the issue of how to regulate afﬁliated management in Chinese listed ﬁrms has not reached a consensus, which may hurt the interests of minority sharehold- ers, and harm the long-term development of the Chinese capital market. The ﬁndings in this paper suggest that regulators should pay more attention to supervising the afﬁliated management, and strictly examine the economic behaviour and the ﬁnancial information disclosures of the ﬁrms that have afﬁliated management. Lastly, this paper empirically tests the association between afﬁliated managers and the value of the ﬁrm, addressing an inadequacy in the previous study (Claessens et al., 2002). There are several limitations in this paper. First, we analyse the effect of afﬁliated managers on ﬁrm value based only on agency theory, but there may be other theories that explain the relation between afﬁliated management and ﬁrm value. For example, the afﬁliated managers may be recognised as a direct vertical interlock between the control- ling groups and the listed ﬁrms and this kind of interlock may affect ﬁrm value (Arnoldi et al., 2013). Secondly, the ﬁndings in this paper suggest that even though afﬁliated management may enhance the ability of shareholders to monitor managers (decrease the type I agency problem) and increase the operational proﬁtability, it can facilitate the con- trolling shareholder to expropriate the minority shareholders (increase the type II agency problem) and reduce the ﬁrm value eventually in a transition economy. The cost that afﬁliated managers bring about may outweigh the beneﬁts that afﬁliated managers bring in. Although we offer systematic evidence to show how afﬁliated management is associ- ated with ﬁrm value, we do not have direct quantitative analysis and comparison between the costs and beneﬁts brought about by the afﬁliated management. Acknowledgements The authors acknowledge the useful comments and suggestions made by two anonymous reviewers and the editors. The authors take full responsibility for any errors. Funding Gaoping Zheng acknowledges ﬁnancial support from the Young Scholar Research Project of the Ministry of Education (Grant number 13YJC630236). Jian Xue acknowledges ﬁnancial support from National Natural Science Foundation of China (Grant number 71,272,025 and 71,322,201). Notes 1. If the controlling shareholder is one person, the term ‘afﬁliated management’ means that the Chairman or CEO of the listed ﬁrm is exactly the same person. If the controlling shareholder is a group of persons, our deﬁnition of afﬁliated management requires that the Chairman or CEO is one of the group members. 2. In many state-owned listed ﬁrms, and in some non-state-owned listed ﬁrms, the Chairman or CEO of the listed ﬁrms also works as an employee of the parent ﬁrm at the same time. 3. We get similar results when we use the Herﬁndahl index alone to measure the monitoring intensity of other large shareholders. Following Chen, Chen, Schipper, Xu, and Xue (2012), we multiply the Herﬁndahl index by the sum of holdings from the second largest shareholder to the ﬁfth largest shareholder, because the overall power of these shareholders also matters. 198 Zheng et al. References Anderson, R. C., & Reeb, D. M. (2003). Founding-family ownership and ﬁrm performance: Evidence from the S&P 500. The Journal of Finance, 58, 1301–1327. Arnoldi, J., Chen, X., & Na, C. (2013). Vertical interlocks of executives and performance of afﬁliated ﬁrms in state owned Chinese business groups. Working Paper. Bai, C., Liu, Q., Lu, Z., Song, M., & Zhang, J. (2005). An empirical study on Chinese listed ﬁrms’ corporate governance. Economic Research Journal, 2,81–91. (In Chinese) Berle, A. A., & Means, G. C. (1932). The Modern Corporation and Private Property. New York, NY: MacMillan. Bunkanwanicha, P., Fan, J., & Wiwattanakantang, Y. (2013). The value of marriage to family ﬁrms. Journal of Financial and Quantitative Analysis, 48,611–636. Chen, Q., Chen, X., Schipper, K., Xu, Y., & Xue, J. (2012). The sensitivity of corporate cash holdings to corporate governance. Review of Financial Studies, 25, 3610–3644. Chen, C., Li, Z., Su, X., & Yao, Y. (2012). Delegation and sensitivity of CEO turnover to ﬁrm performance within business groups: Evidence from China. Journal of Accounting and Public Policy, 31, 553–574. Chen, X., & Liu, S. (2003). An empirical study on the structural difference among the compensation of managers in Chinese public companies. Economic Research Journal, 8, 55–63. (In Chinese) Claessens, S., Djankov, S., Fan, J. P. H., & Lang, Larry H. P. (2002). Disentangling the incentive and entrenchment effects of large shareholdings. The Journal of Finance, 57, 2741–2771. Claessens, S., Djankov, S., & Lang, Larry H. P. (2000). The separation of ownership and control in East Asian corporations. Journal of Financial Economics, 58,81–112. Coase, R. H. (1937). The nature of the ﬁrm. Economica, 4, 386–405. Doidge, C. G., Andrew Karolyi, G., & Stulz, R. M. (2004). Why are foreign ﬁrms listed in the U.S. worth more? Journal of Financial Economics, 71, 205–238. Du, X., & Zhou, Z. (2010). Political connection and auditor choice. Auditing Research, 2,47–53. (In Chinese) Fama, E. F., & French, K. R. (1998). Taxes, ﬁnancing decisions, and ﬁrm value. The Journal of Finance, 53, 819–843. Fan, J. P. H., Wong, T. J., & Zhang, T. (2007). Politically connected CEOs, corporate governance, and Post-IPO performance of China’s newly partially privatized ﬁrms. Journal of Financial Economics, 84, 330–357. Fisman, R., & Wang, Y. (2010). Trading favors within Chinese business groups. American Economic Review, 100, 429–433. Gong, Y. (2001). Firm performance and top management turnover. Economic Research Journal, 10,75–82. (In Chinese) Jensen, M. C., & Meckling, W. H. (1976). Theory of the ﬁrm: Managerial behavior, agency costs and ownership structure. Journal of Financial Economics, 3, 305–360. Jiang, G., Lee, C., & Yue, H. (2010). Tunneling through intercorporate loans: The China experience. Journal of Financial Economics, 98(1), 1–20. King, M. R., & Segal, D. (2009). The long-term effects of cross-listing, investor recognition, and ownership structure on valuation. Review of Financial Studies, 22, 2393–2421. La Porta, R., Lopez-de-Silanes, F., & Shleifer, A. (1999). Corporate ownership around the world. The Journal of Finance, 54, 471–517. La Porta, R. L., Lopez-de-Silanes, F., Shleifer, A., & Vishny, R. (1998). Law and ﬁnance. Journal of Political Economy, 106,1113–1155. La Porta, R., Lopez-de-Silanes, F., Shleifer, A., & Vishny, R. (2000). Investor Protection and corporate governance. Journal of Financial Economics, 58,3–27. Li, Z., Yu, Q., & Wang, X. (2005). Tunneling, propping and M&A: Evidence from Chinese listed companies. Economic Research Journal, 1,95–105. (In Chinese) Liu, J., Wei, M., & Zheng, G. (2008). The interested-party investment controlled by controlling shareholders: Enhancing efﬁciency or transferring resources. Management World, 3, 133–141. (In Chinese) Luo, D., & Tang, Q. (2009). The performance of institutional environment: Evidence from China’s private listed companies. Economic Research Journal, 2, 106–118. (In Chinese) Myers, S. C. (1977). Determinants of corporate borrowing. Journal of Financial Economics, 5, 147–175. China Journal of Accounting Studies 199 Nguyen, B. D., & Nielsen, K. M. (2010). The value of independent directors: Evidence from sudden deaths. Journal of Financial Economics, 98, 550–567. Rao, Y., Zhang, Y., & Peng, D. (2008). Holding proportion, over guarantee and tunneling: Evidence from the guarantee of Chinese listed companies to their subsidiaries. Nankai Business Review, 1,31–38. (In Chinese) Shi, S. (2010). The transfer of control right, excessive control and the expropriation of large shareholders: Evidence from management turnover in the listed ﬁrms. Journal of Financial Research, 160–176. (In Chinese) Shleifer, A., & Vishny, R. (1986). Large shareholders and corporate control. Journal of Political Economy, 94, 461–488. Villalonga, B., & Amit, R. (2006). How do family ownership, control and management affect ﬁrm value? Journal of Financial Economics, 80, 385–417. White H. (1980). A heteroskedasticity-consistent covariance matrix estimator and a director test for heteroskedasticity. Econometrica, 48, 817–838. Xia, L., & Zhang, T. (2005). Government control, institutional environment and ﬁrm value: Evi- dence from Chinese securities market. Economic Research Journal, 5,40–51. (In Chinese) Yu, M., & Pan, H. (2008). The relationship between politics, institutional environments and private enterprises’ access to bank loans. Management World, 8,9–21. (In Chinese) Zhao, Z., Yang, Z., & Bai, C. (2007). Empirical study on the determinants of top management turnover in the Chinese Listed Firms. Journal of Financial Research, 8,76–89. (In Chinese) Zhu, H. (2002). Analysis of the phenomenon of the top management turnover in Chinese listed ﬁrms. Management World, 5, 126–131. (In Chinese) Zou, F., & Chen, X. (2004). An empirical study on the relationship between the ownership characteristics and the board leadership. Nankai Business Review, 7,49–56. (In Chinese) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png China Journal of Accounting Studies Taylor & Francis http://www.deepdyve.com/lp/taylor-francis/affiliated-management-and-firm-value-evidence-from-china-716mDfYvzp

Affiliated management and firm value: Evidence from China

Zheng, Gaoping; Xue, Jian; Chen, Xiao

China Journal of Accounting Studies , Volume 2 (3): 28 – Jul 3, 2014

Download PDF

Share Full Text for Free

28 pages

Affiliated management and firm value: Evidence from China

Abstract

We provide evidence on how the presence of a top manager who is a controlling shareholder or the employee of a controlling shareholder (i.e. affiliated management) is associated with firm value. By analysing the data of Chinese listed firms in the A-share stock market from 2001 to 2009, we find that affiliated management is negatively associated with firm value. Furthermore, although firms with affiliated management have higher return on assets and return on sales, affiliated management is...

Loading next page...

Publisher: Taylor & Francis
ISSN: 2169-7221
eISSN: 2169-7213
DOI: 10.1080/21697213.2014.921366
Publisher site: See Article on Publisher Site

Abstract

China Journal of Accounting Studies, 2014 Vol. 2, No. 3, 172–199, http://dx.doi.org/10.1080/21697213.2014.921366 Afﬁliated management and ﬁrm value: Evidence from China a b b Gaoping Zheng *, Jian Xue and Xiao Chen Southwestern University of Finance & Economics, School of Accounting, Chengdu 611130, China; Tsinghua University, School of Economics and Management, Beijing 100084, China We provide evidence on how the presence of a top manager who is a controlling shareholder or the employee of a controlling shareholder (i.e. afﬁliated management) is associated with ﬁrm value. By analysing the data of Chinese listed ﬁrms in the A-share stock market from 2001 to 2009, we ﬁnd that afﬁliated management is neg- atively associated with ﬁrm value. Furthermore, although ﬁrms with afﬁliated man- agement have higher return on assets and return on sales, afﬁliated management is positively associated with the value and the frequency of loan-based related party transactions and with the likelihood of irregularity in information disclosure. Our ﬁndings imply that even if afﬁliated management may enhance shareholders’ ability to monitor managers and increase the operational proﬁtability, afﬁliated management can facilitate the expropriation of the interests of the minority shareholders by the controlling shareholders. Keywords: afﬁliated management; ﬁrm value; private beneﬁt of control 1. Introduction The co-ordination of various factors of production built up within a ﬁrm can reduce trans- action costs dramatically (Coase, 1937). However, as the size of a ﬁrm increases, the issue of the separation between ownership and control arises as well, which brings agency cost (Jensen & Meckling, 1976). Scholars and practitioners have proposed many mechanisms to alleviate the agency problem. For instance, Shleifer and Vishny (1986) argue that the controlling shareholders have incentives to monitor the management and reduce the agency problem between shareholders and managers. La Porta, Lopez-de-Silanes, and Shleifer (1999) show that in some cases the management of family ﬁrms in the United States is assigned to family members, while some controlling shareholders assign their employees to serve as the management in subsidiaries. Further empirical evidence from developed economies shows that family ﬁrms with afﬁliated management have higher ﬁrm value than other family ﬁrms, implying that afﬁli- ated management mitigates the agency problem in family ﬁrms (Anderson & Reeb, 2003; Villalonga & Amit, 2006). However, in a transition economy where the investor protec- tion is weak and the controlling shareholders are prevalent (Claessens, Djankov, & Lang, 2000; La Porta, Lopez-de-Silanes, Shleifer, & Vishny, 1998), controlling shareholders may control the ﬁrms for private beneﬁts (Chen, Chen, Schipper, Xu & Xue, 2012; Claessens, Djankov, Fan, & Lang, 2002; La Porta, Lopez-de-Silanes, Shleifer, & Vishny, 2000), causing afﬁliated management to have a different impact on ﬁrm value. *Corresponding author. Email: zhenggaoping2012@swufe.edu.cn Paper accepted by Kangtao Ye. © 2014 Accounting Society of China China Journal of Accounting Studies 173 Theoretically, afﬁliated management could serve not only as a channel through which the controlling shareholder can seize the private beneﬁt of control, but also as a channel through which the controlling shareholder can enhance the monitoring of the management of listed ﬁrms and hence reduce the agency problem. The former (latter) channel infers that afﬁliated management will be negatively (positively) associated with ﬁrm value. Thus, empirical analysis could clarify the association of afﬁliated manage- ment with ﬁrm value. To the best of our knowledge, although a few studies test the impact of the separation of cash ﬂow rights and control rights on ﬁrm value (Claessens et al., 2000, 2002), the association between afﬁliated management and ﬁrm value is still unknown. This paper contributes to the literature by empirically testing the association of afﬁliated management with ﬁrm value, using the unique data of afﬁliated manage- ment in China, the second largest economy in the world and the largest transition economy. Afﬁliated management is a common practice in Chinese listed ﬁrms, which pro- vides a rich dataset from which to test the association of afﬁliated management with ﬁrm value. Our data show that more than half of listed ﬁrms have afﬁliated manage- ment from 2001 to 2009 (Table 2, later). In an emerging market, afﬁliated management can facilitate the tunnelling behaviour of controlling shareholders, because the control- ling shareholders may abuse their control over the listed ﬁrm, given the weak investor protection (Claessens et al., 2000, 2002). Based on such concern, the China Securities Regulatory Commission (CSRC) established several regulations to constrain the emer- gence of afﬁliated management in listed ﬁrms. However, the State-Owned Assets Supervision and Administration Commission of the State Council (SASAC) holds the opposite view, urging a few giant ﬁrms owned by the central state, such as China National Petroleum Corporation, SINOPEC, and SHENHUA Group, to build up afﬁli- ated management. From the view of SASAC, afﬁliated management should bring better monitoring of the management of listed ﬁrms and reduce the agency problem between shareholders and management, which may increase the value of the ﬁrm and the wealth of the nation eventually. This paper also contributes to the long-lasting debate about the role of afﬁliated management in China. By analysing the data of Chinese listed ﬁrms in the A-share stock market from 2001 to 2009, we ﬁnd that the afﬁliated management is negatively associated with ﬁrm value. Furthermore, although ﬁrms with afﬁliated management have higher return on assets and return on sales, the afﬁliated management is positively associated with the value and the frequency of loan-based related party transactions and with the likelihood of irregularity in information disclosure in the listed ﬁrms. The ﬁndings imply that even if afﬁliated management may enhance the ability of shareholders to monitor managers, it can facilitate the expropriation of the interests of the minority shareholders by the controlling shareholder in a transition economy. Further evidence shows that the association between afﬁliated management and ﬁrm value does not depend on whether the largest shareholder in the ﬁrm owns more than 50% or not. Meanwhile, the negative association between afﬁliated management and ﬁrm value is weaker in ﬁrms with a high level of institutional investor holdings, which implies that institutional investors may restrict the harmful behaviour of afﬁliated man- agers. Our results show that the association between afﬁliated management and ﬁrm value is not signiﬁcant in the ﬁrms that are implementing their seasoned equity offer- ing, which is consistent with the idea that controlling shareholders do not have a high incentive to expropriate the interests of minority shareholders when the ﬁrm has a high demand for external ﬁnancing. We also ﬁnd that the negative association of afﬁliated 174 Zheng et al. management with ﬁrm value is mostly driven by non-state-owned listed ﬁrms. One pos- sible explanation is that compared with the controlling shareholder (i.e. the state) in the state-owned listed ﬁrms, the controlling shareholders (generally a person or a family) in the non-state-owned listed ﬁrms are more prone to expropriate the interests of minor- ity shareholders, with the help of afﬁliated managers, for their private beneﬁts, given the weak institutional environment. Meanwhile, the afﬁliated management in the state- owned listed ﬁrms is under more severe monitoring by the government than the afﬁli- ated management in the non-state-owned listed ﬁrms; thus, the afﬁliated managers’ behaviour that harms ﬁrm value is more likely to be restrained in the state-owned listed ﬁrms. Our main results hold after controlling for different types of afﬁliated manage- ment, and the different positions that the afﬁliated management hold in the controlling shareholder, as well as the split share structure reform. After investigating the associa- tion between afﬁliated management and the ROA (return on assets) and ROS (return on sales), we ﬁnd that afﬁliated management is positively associated with the ROA and ROS in the listed ﬁrms. The ﬁndings imply that the presence of afﬁliated management may improve the operational proﬁtability in the ﬁrm. However, we also ﬁnd that ﬁrms with afﬁliated management are more likely to manipulate information disclosure, including fraudulent accounting treatment (such as exaggerated proﬁt, false statement of assets), which may apparently increase ROA and ROS. Even if the increase of prof- itability is authentic, the controlling shareholder could turn it into a private beneﬁtof control, with the help of afﬁliated management, through related party transactions or other channels, which may result in a negative association between afﬁliated manage- ment and the ﬁrm value. The ﬁndings contribute to the literature in the following ways. First, this study empirically tests the role of afﬁliated management, from the corporate governance per- spective, using a large sample in a transition economy, which ﬁlls a gap left by previ- ous research. Theoretically, afﬁliated management could reduce ﬁrm value when it serves as a channel through which a controlling shareholder seizes the private beneﬁt of control, but it could also increase ﬁrm value when it serves as a channel through which a controlling shareholder enhances the monitoring of management. Previous studies focus on the role of afﬁliated management in a developed economy with good investor protection, paying little attention to the case in an emerging market. This paper shows that, given weak investor protection in an emerging market, the afﬁliated man- agement is negatively associated with the ﬁrm value, which implies that the cost related to afﬁliated management exceeds the beneﬁts in a transition economy. Secondly, our research on afﬁliated management in Chinese listed ﬁrms offers us a good opportunity to examine the impact of large shareholder control in an emerging market. Claessens et al. (2000) show that controlling shareholders use four channels to enhance their con- trol in East Asian countries, including a pyramid structure, cross-holdings, dual class shares and afﬁliated management. Claessens et al. (2002) test the impact of a pyramid structure, cross-holdings and dual class shares on ﬁrm value. They ﬁnd that these three channels induce separation of control rights and cash ﬂow rights in both private ﬁrms and state-owned ﬁrms, which enlarges the incentive for the controlling shareholder to pursue the private beneﬁt of control and reduce the ﬁrm value. They fail to test the impact of afﬁliated management on ﬁrm value due to data insufﬁciency. As a result, Claessens et al. (2002) mentioned that they may underestimate the effect of large share- holder control. In our setting, afﬁliated management is dominant in the Chinese stock market and the information disclosure of afﬁliated management in listed ﬁrms is man- dated, which means that our paper can provide another dimension to investigating the China Journal of Accounting Studies 175 impact of large shareholder control in an emerging market. Finally, this paper not only shows the negative association of afﬁliated management with ﬁrm value, but also discusses the channels through which afﬁliated management has this negative associa- tion with ﬁrm value. We ﬁnd that the presence of afﬁliated management is positively associated with the value and the frequency of related party transactions and with the likelihood of irregularity in information disclosure in the listed ﬁrms, which provides valuable evidence for further regulation. The rest of the paper is organised as follows. Section 2 reviews the institutional background; Section 3 develops research hypotheses; Section 4 describes the research design and data; Section 5 reports the empirical results; Section 6 presents the further analysis and a robustness test; and Section 7 concludes the paper. 2. Institutional background Afﬁliated management in China is different from afﬁliated management in other East Asian countries in the sense that the majority of afﬁliated management in China is not based on the family members of controlling shareholders (Bunkanwanicha, Fan, & Wiwattanakantang, 2013; Claessens et al., 2000). The disparity comes from the fact that the Chinese stock market is dominated by state-owned listed ﬁrms (62.7%), in which afﬁliated management cannot be family members, but usually are the employees of the controlling shareholder. Meanwhile, non-state-owned listed ﬁrms can still use family members as the afﬁliated management. State-owned listed ﬁrms include two categories (Zou & Chen, 2004). The ﬁrst category of state-owned listed ﬁrms is those ﬁrms directly controlled by government agencies, such as local government and SASAC. In these ﬁrms, political concerns other than economic factors serve as the major determinant of the personnel arrangements. Meanwhile, the largest shareholders for these ﬁrms are government-funded agencies, without any operational business or ﬁnancial needs, which means that they do not have high incentives to monitor the management or pursue private beneﬁt of control. As a result, the afﬁliated management in the listed ﬁrms directly controlled by government agencies may be arranged to achieve certain political goals instead of economic pur- poses. For example, local government (as they directly control the listed ﬁrms) may arrange the local governors to be the afﬁliated management in listed ﬁrms so that the listed ﬁrms may assist local government in propping local employment or achieving other political goals. Because the personnel arrangement in these ﬁrms may be mostly driven by political concerns other than economic factors, and political concerns are not the focus of our study, we drop this kind of ﬁrms in our sample. The second category of state-owned ﬁrms are the ﬁrms indirectly controlled by government agencies. Typically, the largest shareholders for these ﬁrms are parent state-owned ﬁrms with related business. This kind of controlling shareholder not only serves as the monitor of state-owned shareholdings and assets, but also has close opera- tional and ﬁnancial connections with the listed ﬁrms. On one hand, these controlling shareholders may have strong incentives to use listed ﬁrms to beneﬁt themselves. For example, the controlling shareholders may treat cash holdings of listed ﬁrms as their own, make listed ﬁrms guarantee their bank loans, and even sell assets to or purchase assets from listed ﬁrms at an unfair price. Such incentive partly comes from the fact that the controlling shareholder puts the most valuable business into the listed ﬁrm dur- ing the IPO process, and such assets infusion leads to the fact that the controlling shareholder has no core business to generate necessary cash ﬂows later on. In addition, 176 Zheng et al. a controlling shareholder may beneﬁt itself by making the strategy of listed ﬁrms con- sistent with its own strategy. On the other hand, since listed ﬁrms must achieve certain performance requirements before seasoned equity offering in the Chinese stock market, controlling shareholders may also sacriﬁce their own beneﬁts to ﬁnancially support the listed ﬁrms, in order to prop up the performance of listed ﬁrms in a short period. In this case, the controlling shareholders of this category of state-owned ﬁrms have a strong motivation to arrange the afﬁliated management in the listed ﬁrms to enhance their control over the listed ﬁrms for the beneﬁts of their own. Compared with the controlling shareholders of the above two categories of state- owned listed ﬁrms, the controlling shareholders of non-state-owned listed ﬁrms may have even stronger incentives to enhance their control. First, since the controlling shareholders for non-state-owned listed ﬁrms face more policy risks than those in the state-owned ﬁrms, including the weak protection of private property rights, the uncertainty of indus- trial policies and the discrimination in ﬁnancial policies enforced by government (Luo & Tang, 2009; Yu & Pan, 2008), strong control can facilitate these controlling shareholders to obtain security. Second, given weak investor protection and legal institution, these con- trolling shareholders have strong incentives to enhance their control over listed ﬁrms to get private beneﬁt of control (Claessens et al., 2000). In this case, afﬁliated management could not only help the controlling shareholders directly adjust the operation strategies of the listed ﬁrms according to the frequently changed policies, in order to ensure the con- trolling shareholders’ beneﬁts, but also reduce the transaction costs when the controlling shareholders try to grab beneﬁts by navigating around different government policies. More importantly, afﬁliated management can assure the strong control power of the con- trolling shareholders by directly controlling the management in the listed ﬁrms, since the effective enforcement of the decisions may be more important than making decisions given the weak institutional environment. Both in the state-owned listed ﬁrms and in the non-state-owned listed ﬁrms, afﬁli- ated management can enhance the large shareholder control and facilitate the tunnelling behaviour of controlling shareholders, given the weak investor protection (Claessens et al., 2000, 2002). Based on such concern, the China Securities Regulatory Commis- sion (CSRC) has established several regulations to constrain the emergence of afﬁliated management in listed ﬁrms. The CSRC issued a regulation ‘Announcement for Investi- gating the Restructuring in Firms that Plan to Go through IPO’ in 1998, stipulating that IPO ﬁrms should meet the personnel independence criteria, which require that the chairman in a listed ﬁrm must not serve as the legal representative of the controlling shareholder, that management such as president and vice president must not serve both in the listed ﬁrm and in the ﬁrm that is the controlling shareholder, and that the labour, personnel and wage management in a listed ﬁrm must be completely independent from the controlling shareholder. In 1999, the CSRC issued a special announcement, ‘Announcement for Forbidding Management in Listed Firms Serving in Controlling Shareholder’, requiring that a listed ﬁrm must have separate personnel from those of the controlling shareholder. Further regulation by the CSRC in 2001, ‘Regulations for Initial Public Offering’ requires that management such as the president and vice presi- dent in a listed ﬁrm must only serve in this listed ﬁrm. In 2002, CSRC issued the ‘Code of Corporate Governance for Listed Firms in China’ which enhanced the require- ment for personnel independence in relation to the managers of listed ﬁrms. The CSRC also took an enforcement action in 2007 to check whether or not listed ﬁrms follow the ‘Code of Corporate Governance for Listed Firms in China’. However, the enforcement of these regulations is quite weak, given the lack of monitoring and the lack of severe China Journal of Accounting Studies 177 punishment. Besides, the interventions from other government agencies make the enforcement of these regulations even weaker. For instance, the State-Owned Assets Supervision and Administration Commission of the State Council (SASAC) urges the giant ﬁrms owned by the central state to build up afﬁliated management in their listed ﬁrms, since SASAC holds the view that afﬁliated management should bring better monitoring of the management of listed ﬁrms and reduce the agency problem between shareholders and management, which may eventually increase the value of the ﬁrm and the wealth of the nation. 3. Hypothesis development The discussion about agency problems induced by the separation of ownership and control can be traced back to Adam Smith in the eighteenth century (Jensen & Meckling, 1976). The idea that the ownership in US ﬁrms is so dispersed that manage- ment in these ﬁrms may control the ﬁrm and harm ﬁrm value is well known, since the work of Berle and Means (1932), The Modern Corporation and Private Property, was published. As a result, scholars have investigated how to mitigate such a so-called type I agency problem, the agency problem that concerns the conﬂict of interests between management and shareholders. For the type I agency problem, Shleifer and Vishny (1986) argue that a controlling family can enhance their control over management by appointing family members to the board. Meanwhile, La Porta et al. (1999) show that using a family member as a manager is a common practice around the world. For example, they ﬁnd that 75% of family ﬁrms use family members as managers in the US. They do not study whether managers in family ﬁrms also work as employees of the parent ﬁrms owned by the controlling shareholders. Most empirical ﬁndings focus on the impact of afﬁliated management on ﬁrm value in family ﬁrms in developed economies. Anderson and Reeb (2003) ﬁnd that family ﬁrms exhibit superior performance compared with non-family ﬁrms, and even more superior performance when family members serve as CEOs. Villalonga and Amit (2006) also ﬁnd that family ﬁrms have higher value when the founder of a family ﬁrm serves as CEO or Chairman. However, the answer to the question of how afﬁliated management affects ﬁrm value may be different in an emerging market. In an emerging market, there is a greater prevalence of the type II agency problem, which concerns the conﬂicts of interest between the controlling shareholder and minority shareholders (La Porta et al., 2000). Claessens et al. (2000) show that controlling shareholders use four channels to enhance their control in East Asian countries, including a pyramid structure, cross-holdings, dual class shares and afﬁliated management. Claessens et al. (2002) test the impact of a pyramid structure, cross-holdings and dual class shares on ﬁrm value. They ﬁnd that these three channels induce the separation of control rights and cash ﬂow rights in both private ﬁrms and state-owned ﬁrms, which enlarges the incentives of controlling share- holders to pursue private beneﬁt of control and reduce the ﬁrm value signiﬁcantly. They fail to test the impact of afﬁliated management on ﬁrm value due to insufﬁciency of data. According to Claessens et al. (2002), afﬁliated management, playing a similar role as the other three channels in an emerging market, facilitates the tunnelling behaviour of controlling shareholders by enhancing their control over listed ﬁrms. Previous studies using Chinese data show that expropriations of the minority shareholder by controlling 178 Zheng et al. shareholder are quite common in China (Chen, Li, Su, &Yao, 2012; Fisman & Wang, 2010; Jiang, Lee, & Yue, 2010) and such tunnelling behaviour decreases ﬁrm performance (Chen, Chen, Schipper, Xu, & Xue, 2012; Jiang et al., 2010). If afﬁliated management does help controlling shareholders expropriate the interests of minority shareholders, we predict a negative relation between afﬁliated management and ﬁrm value. We test this prediction with the following hypothesis: H1(A): Firms with afﬁliated management have lower ﬁrm value than ﬁrms without afﬁliated management. However, even in a developed market with strong investor protection, management could harm the interests of shareholders (Jensen & Meckling, 1976). In an emerging market with weak investor protection, managers can easily achieve control over the ﬁrm, using the resources of the ﬁrm to beneﬁt themselves and reduce ﬁrm value (Fan, Wong, & Zhang, 2007). Afﬁliated management may increase ﬁrm value by alleviating this kind of agency problem between management and shareholders (Anderson & Reeb, 2003). Meanwhile, afﬁliated management can be regarded as a vertical interlock between controlling shareholder and listed ﬁrms (Arnoldi, Chen, & Na, 2013), an inter- lock that may help the controlling shareholder to prop the operation performance of listed ﬁrms more easily and increase the ﬁrm performance. From above perspective, afﬁliated management may be positively associated with ﬁrm value. This discussion leads to a competing hypothesis on the relation between afﬁliated management and ﬁrm value. H1(B): Firms with afﬁliated management have higher ﬁrm value than ﬁrms without afﬁli- ated management. 4. Research design 4.1. Data source and research sample We use the following steps to select our sample. First, all data on afﬁliated manage- ment in all the Chinese listed ﬁrms in the A share stock market from 2001 to 2009 are downloaded from the CSMAR database and we get 12,482 ﬁrm*year observations. Second, 173 ﬁrm*year observations from listed ﬁrms in the ﬁnancial sector are deleted, since they are not comparable to other ﬁrms in many perspectives, such as regulations and accounting standards. Third, we drop 2150 ﬁrm*year observations from the ﬁrst category of state-owned listed ﬁrms which are directly controlled by government agen- cies, such as SASAC and local government, because the personnel arrangement in these ﬁrms may be mostly driven by political concerns other than economic factors. Finally, we get 10,159 ﬁrm-year observations. All continuous variables are winsorized at the 1% level. Stock price, ﬁnancial data, ownership structure, board characteristics, related party transactions, irregularity and industry categories are all from the CSMAR database, while we use the CCER data- base to identify whether a ﬁrm is owned by the state or not. 4.2. Model identiﬁcation and variable deﬁnitions Consistent with Claessens et al. (2002), we use the following model to test the associa- tion of afﬁliated management with ﬁrm value. Standard errors in the regression analysis China Journal of Accounting Studies 179 are all corrected for clustering of the observations at the ﬁrm level. For each test sam- ple in the regression analysis, the number of observations may vary as the value of dependent variables (e.g., TQ, LoanRPTs) are missing, or the value of control variables (e.g., Lev, Bsize) are missing. TQ ¼ a þ a AM þ a Gr Sales þ a IND Q þ a SIZE þ a LEV þ a ROA þ a SOE þ a TOP1 0 1 2 3 4 5 6 7 8 X X þ a Herf 2 5 þ a BSIZE þ a OUTDIR þ a DUAL þ YEAR þ IND þ e 9 10 11 12 (1) Referring to Claessens et al. (2002), the dependent variable in model (1) is Tobin’sQ, which measures the value of the ﬁrm. TQ is deﬁned as the sum of book value of liabil- ity, market value of liquid shares and book value of illiquid shares, divided by book value of total assets. Consistent with the deﬁnition in Claessens et al. (2000), the management of a listed ﬁrm is deﬁned as afﬁliated management if Chairman or CEO of the ﬁrm is the controlling 1 2 shareholder, or the employee of the controlling shareholder. The independent variable AM equals one if the ﬁrm has afﬁliated management. Otherwise, AM equals 0. We focus only on Chairman and CEO because both Chairman and CEO are the key positions that directly affect the operational activities of Chinese listed ﬁrms in practice (Chen & Liu, 2003; Du & Zhou, 2010; Gong, 2001; Zhu, 2002; Zhao, Yang, & Bai, 2007). In our empirical analysis, we also calculate some variables as additional controls, such as growth, size, leverage and proﬁtability (Doidge, Karolyi, & Stulz, 2004; King & Segal, 2009). We use two measures, Gr_sales and IND_Q, to proxy for growth. Gr_sales is deﬁned as the average sales growth rate of the previous two years. We use the sales growth rate in the last year if there are missing observations. IND_Q is deﬁned as the median value of Tobin’s Q in the same industry. We deﬁne industries based on the industry classiﬁcation standard issued by CSRC in 2001. Note that we deﬁne manufacturing industry based on the secondary industry code, while we deﬁne other industries based on the ﬁrst industry code. SIZE is deﬁned as the natural loga- rithm of total assets. Generally speaking, ﬁrm value is positively associated with the growth rate and negatively associated with the ﬁrm size (King & Segal, 2009). We use Lev to proxy for ﬁrm leverage. Lev is deﬁned as the sum of short-term debts and long- term debts, divided by total assets. Proﬁtability is measured by ROA, return on assets， deﬁned as net income divided by total assets. Previous studies ﬁnd that the relation between debt and ﬁrm value is negative (Fama & French, 1998; Myers, 1977), and the relation between proﬁtability and ﬁrm value is positive (Doidge et al., 2004;King& Segal, 2009). We also use an industry dummy and a year dummy to control for the unobserved industry and time ﬁxed effects. Ownership structure and board characteristics are also controlled in our regression analysis (Rao, Zhang, & Peng, 2008; Shi, 2010). A dummy variable SOE equals 1 if the ﬁrm is controlled by the state, according to the CCER database. Otherwise, SOE equals 0. One widely accepted idea is that the ﬁrm value in the Chinese state-owned ﬁrms may be lower than the other ﬁrms, because the property rights of these ﬁrms are not well-deﬁned in practice and no one has a strong incentive to monitor the managers, leading to the low level of operation efﬁciency in these ﬁrms (Xia & Zhang, 2005). Variable TOP1 is deﬁned as the percentage of shares held by the largest shareholder. Variable Herf_2_5 measures the monitoring intensity by large shareholders other than the controlling shareholder and is deﬁned as the Herﬁndahl index from the second largest shareholder to the ﬁfth largest shareholder, multiplied by the sum of holdings 180 Zheng et al. from the second largest shareholder to the ﬁfth largest shareholder. BSIZE is deﬁned as the natural logarithm of the number of all board directors. The proxy for board independence is OUTDIR,deﬁned as the number of outside directors divided by the number of all directors. If one person serves as Chairman and CEO at the same time, dummy variable DUAL equals one, otherwise zero. A previous study shows that ﬁrm value is lower when board size is larger, when the board has fewer outside directors, and when one person serves as Chairman and CEO at the same time (Bai, Liu, Lu, Song, & Zhang, 2005). Table 1 shows the main variables and their deﬁnitions. 5. Empirical results 5.1. Descriptive statistics Table 2 reports the distribution of our sample. Panel A presents the sample distribution by year. It may be seen that the percentage of ﬁrms with afﬁliated management Table 1. Deﬁnitions of the main variables. Variable Deﬁnition AM A dummy variable equals 1, if the Chairman or CEO of the ﬁrm is the controlling shareholder, or an employee of the controlling shareholder. Otherwise, AM equals 0. TQ The sum of book value of liability, market value of liquid shares and book value of illiquid shares, divided by book value of total assets. LoanRPT The sum of annual loan-based related party transitions between listed ﬁrms and controlling shareholders, divided by total assets. LoanRPT_FREQ The frequency of loan-based related party transactions in a calendar year. IRREG If CSRC reports that the listed ﬁrm i in year t manipulated information disclosure, including fraudulent accounting treatment (such as exaggerated proﬁt, false statement of assets), misleading statements, and omission of substantial information, the dummy variable IRREG of the ﬁrm i in year t is equal to 1. Otherwise, IRREG is equal to 0. Gr_Sales The average sales growth rate of previous two years. We use the sales growth rate in last year if there are missing observations. IND_Q The median value of Tobin’s Q in the same industry. We deﬁne industries based on the industry classiﬁcation standard issued by CSRC in 2001. Note that we deﬁne manufactory industry based on secondary industry code, while we deﬁne other industries based on ﬁrst industry code. SIZE The natural logarithms of total assets. LEV The sum of short-term debts and long-term debts, divided by total assets. ROA The net income divided by total assets at the end of the accounting periods. AGE The number of years since the ﬁrm has been listed. SOE A dummy variable SOE equals 1, if the ﬁrm is controlled by the state and 0 otherwise. TOP1 The percentage of shares held by the largest shareholder. Herf_2_5 The Herﬁndahl index from second largest shareholder to ﬁfth largest shareholder, multiplied by the sum of holdings from second largest shareholder to ﬁfth largest shareholder following Chen et al. (2012). BSIZE The natural logarithm of the number of all board directors. OUTDIR The number of outside directors divided by the number of all directors. DUAL A dummy variable equals 1, if one person serves as Chairman and CEO at the same time and 0 otherwise. Note: The data source for all the variables is the CSMAR database, except for one variable SOE, which is downloaded from the CCER database. China Journal of Accounting Studies 181 Table 2. Sample distribution. Number of ﬁrms with Percentage of ﬁrms with Number of AM AM observations Panel A Distribution by year 2001 411 47.57% 864 2002 415 45.01% 922 2003 505 51.37% 983 2004 549 51.07% 1,075 2005 623 57.21% 1,089 2006 681 58.40% 1,166 2007 794 62.42% 1,272 2008 857 64.63% 1,326 2009 982 67.17% 1,462 Total 5,817 57.26% 10,159 Panel B Distribution by industry Industry Agriculture 117 54.17% 216 Mining 99 50.25% 197 Food & beverage 213 49.88% 427 Textiles & apparel 250 58.82% 425 Timber & furnishings 29 67.44% 43 Paper & printing 125 67.57% 185 Petrochemicals 649 56.29% 1,153 Electronics 267 64.18% 416 Metal/non-metal 520 60.54% 859 Machinery 894 58.17% 1,537 Pharmaceuticals 425 60.89% 698 Other manufacturing 43 56.58% 76 Utilities 177 47.20% 375 Construction 110 63.22% 174 Transportation 180 54.38% 331 Information technology 449 68.13% 659 Retailing 319 53.61% 595 Real estate 392 49.81% 787 Social Services 218 65.07% 335 Media 54 50.00% 108 Multi-industry 287 50.98% 563 Total 5,817 57.26% 10,159 Key: AM = Afﬁliated management. increases year by year in Panel A, with more than 50% of listed ﬁrms having afﬁliated management since 2003. Panel B shows the sample distribution by industry. The utili- ties industry has the lowest percentage (47.2%), and the information technology indus- try has the highest percentage (68.13%). Table 3 reports the descriptive statistics of the main variables. Panel A reports the full sample statistics. Panel B and Panel C report the statistics for ﬁrms with afﬁliated management and ﬁrms without afﬁliated management respectively. Panel D reports the difference between ﬁrms with afﬁliated management and those without afﬁliated management. We see that the mean TQ for the full sample is 1.6781, and ﬁrms with 182 Zheng et al. Table 3. Descriptive Statistics. Number of Mean Min. Median Max. Std. Observations Panel A Full Sample AM 0.5726 0.0000 1.0000 1.0000 0.4947 10,159 TQ 1.6781 0.9104 1.3648 6.9431 0.9519 9,975 LoanRPT 0.0419 0.0000 0.0000 0.4089 0.0852 10,156 LoanRPT_FREQ 1.1181 0.0000 0.0000 105.0000 3.7808 10,159 IRREG 0.0405 0.0000 0.0000 1.0000 0.1972 10,094 Gr_Sales 0.1806 –0.2169 0.1425 0.6920 0.2721 10,141 IND_Q 1.4837 0.9995 1.3088 2.8305 0.4082 10,159 SIZE 21.2015 18.7185 21.1274 24.0273 1.0305 10,156 LEV 0.2298 0.0000 0.2187 0.8149 0.1620 10,129 ROA 0.0319 –0.4623 0.0386 0.2603 0.0922 10,156 AGE 6.8474 0.0000 7.0000 19.0000 4.2336 10,159 SOE 0.6270 0.0000 1.0000 1.0000 0.4836 10,118 TOP1 39.2387 9.4500 37.3700 75.0000 16.1841 10,159 Herf_2_5 0.0058 0.0000 0.0010 0.0493 0.0098 10,149 BSIZE 2.2136 1.6094 2.1972 2.7081 0.2156 10,059 OUTDIR 0.3172 0.0000 0.3333 0.5000 0.1029 10,054 DUAL 0.1437 0.0000 0.0000 1.0000 0.3508 10,082 Panel B Firms with AM TQ 1.6613 0.9104 1.3629 6.9431 0.9191 5,685 LoanRPT 0.0463 0.0000 0.0000 0.4089 0.0877 5,815 LoanRPT_FREQ 1.3431 0.0000 0.0000 105.0000 4.2497 5,817 IRREG 0.0417 0.0000 0.0000 1.0000 0.2000 5,777 Gr_Sales 0.1944 –0.2169 0.1564 0.6920 0.2744 5,809 IND_Q 1.5070 0.9995 1.3164 2.8305 0.4252 5,817 SIZE 21.2438 18.7185 21.1612 24.0273 1.0175 5,815 LEV 0.2279 0.0000 0.2188 0.8149 0.1566 5,800 ROA 0.0386 –0.4623 0.0425 0.2603 0.0840 5,815 AGE 6.6582 0.0000 7.0000 19.0000 4.3186 5,817 SOE 0.5633 0.0000 1.0000 1.0000 0.4960 5,800 TOP1 39.6242 9.4500 38.3600 75.0000 15.8399 5,817 Herf_2_5 0.0054 0.0000 0.0009 0.0493 0.0095 5,808 BSIZE 2.2129 1.6094 2.1972 2.7081 0.2143 5,763 OUTDIR 0.3244 0.0000 0.3333 0.5000 0.0974 5,760 DUAL 0.1289 0.0000 0.0000 1.0000 0.3351 5,780 Panel C Firms without AM TQ 1.7004 0.9104 1.3654 6.9431 0.9934 4,290 LoanRPT 0.0361 0.0000 0.0000 0.4089 0.0813 4,341 LoanRPT_FREQ 0.8167 0.0000 0.0000 72.0000 3.0157 4,342 IRREG 0.0389 0.0000 0.0000 1.0000 0.1934 4,317 Gr_Sales 0.1620 –0.2169 0.1248 0.6920 0.2678 4,332 IND_Q 1.4525 0.9995 1.3033 2.8305 0.3822 4,342 SIZE 21.1449 18.7185 21.0794 24.0273 1.0450 4,341 LEV 0.2323 0.0000 0.2185 0.8149 0.1688 4,329 ROA 0.0230 –0.4623 0.0334 0.2603 0.1016 4,341 AGE 7.1009 0.0000 7.0000 19.0000 4.1039 4,342 (Continued) China Journal of Accounting Studies 183 Table 3. (Continued). Number of Mean Min. Median Max. Std. Observations SOE 0.7126 0.0000 1.0000 1.0000 0.4526 4,318 TOP1 38.7222 9.4500 36.1050 75.0000 16.6220 4,342 Herf_2_5 0.0064 0.0000 0.0012 0.0493 0.0102 4,341 BSIZE 2.2145 1.6094 2.1972 2.7081 0.2174 4,296 OUTDIR 0.3076 0.0000 0.3333 0.5000 0.1092 4,294 DUAL 0.1636 0.0000 0.0000 1.0000 0.3700 4,302 Panel D The difference between ﬁrms with AM and ﬁrms without AM Firms with AM-Firms Wilcoxon without AM T value Z value Mean Median difference difference TQ –0.0391 –0.0025 –2.01** –1.49 LoanRPT 0.0102 0.0000 5.96*** 8.72*** LoanRPT_FREQ 0.5264 0.0000 6.96*** 9.92*** IRREG 0.0028 0.0000 0.71 0.71 Gr_Sales 0.0324 0.0316 5.95*** 5.90*** IND_Q 0.0545 0.0127 6.66*** 4.59*** SIZE 0.0988 0.0818 4.79*** 4.40*** LEV –0.0044 0.0003 –1.36 –0.02 ROA 0.0156 0.0091 8.44*** 8.82*** AGE –0.4427 0.0000 –5.22*** –5.27*** SOE –0.1493 0.0000 –15.54*** –15.36*** TOP1 0.9020 2.2550 2.78*** 3.25*** Herf_2_5 –0.0009 –0.0003 –4.78*** –3.25*** BSIZE –0.0016 0.0000 –0.37 –0.18 OUTDIR 0.0168 0.0000 8.16*** 8.03*** DUAL –0.0347 0.0000 –4.93*** –4.92*** Notes: Firms with AM represents the ﬁrms that have afﬁliated management. Please refer to Table 1 for the deﬁnitions of variables. All continuous variables have been winsorized at the 1% level. ***, ** and * denote signiﬁcance at 1, 5 and 10% respectively. afﬁliated management clearly have lower TQ than ﬁrms without afﬁliated management. The mean TQ for the ﬁrms with afﬁliated management is 1.6613, while that for the ﬁrms without afﬁliated management is 1.7004. The difference in the means is –0.0391, which is signiﬁcant at the 5% level. The mean LoanRPTs and mean LoanRPTs-FREQ for the ﬁrms with afﬁliated management are 0.0463 and 1.3431 respectively, while those for the ﬁrms without afﬁliated management are 0.0361 and 0.8167 respectively, the differences of 0.0102 and 0.5264 are signiﬁcant at the 1% level. The foregoing uni- variate tests seem to suggest that ﬁrms with afﬁliated management have lower ﬁrm value and higher value and frequency of loan-based related party transactions than ﬁrms without afﬁliated management. Table 4 presents the correlation matrix of main variables. The Pearson correlation between AM and TQ is –0.0203 (signiﬁcant at the 5% level), which implies that afﬁliated management is negatively associated with ﬁrm value. Meanwhile, AM is positively corre- lated with LoanRPT and LoanRPT_FREQ at the 1% signiﬁcance level, which implies that afﬁliated management is positively associated with the value and the frequency of loan- based related party transactions between controlling shareholders and listed ﬁrms. 184 Zheng et al. Table 4. Correlation matrix of main variables. 　 AM TQ LoanRPT LoanRPT_FREQ IRREG Gr_Sales IND_Q SIZE LEV ROA AGE SOE TOP1 Herf_2_5 BSIZE OUTDIR DUAL AM −0.0150 0.0865 0.0984 0.0070 0.0586 0.0455 0.0437 0.0002 0.0875 −0.0523 −0.1527 0.0322 −0.0323 −0.0018 0.0801 −0.0490 TQ −0.0203 −0.1141 −0.1073 0.0039 −0.1141 0.6780 −0.3583 −0.1725 0.1821 0.1221 −0.1346 −0.2065 0.0803 −0.1274 −0.0117 0.0865 LoanRPT 0.0590 −0.0779 0.9695 −0.0009 0.0497 −0.0582 0.1196 0.2405 −0.0968 0.0153 0.0130 0.1195 −0.1571 0.0067 0.0852 −0.0415 LoanRPT_FREQ 0.0689 −0.0479 0.4698 −0.0043 0.0525 −0.0369 0.1363 0.2179 −0.0875 0.0231 0.0023 0.1038 −0.1438 0.0127 0.0930 −0.0425 IRREG 0.0070 −0.0013 0.0237 0.0064 −0.0647 −0.0185 −0.0901 0.0744 −0.1354 −0.0062 −0.0770 −0.0563 0.0498 −0.0221 −0.0259 −0.0001 Gr_Sales 0.0590 −0.0967 0.0258 0.0105 −0.0544 −0.0798 0.2307 0.0146 0.3109 −0.0751 0.0389 0.1109 −0.0157 0.0687 0.0064 −0.0233 IND_Q 0.0660 0.5417 −0.0412 0.0330 −0.0339 −0.0599 −0.0247 −0.1272 0.1488 0.0587 −0.0696 −0.0872 −0.0111 −0.0875 −0.0547 0.0633 SIZE 0.0475 −0.3381 0.0675 0.1229 −0.0937 0.2146 −0.0079 0.1487 0.1111 0.1741 0.2452 0.1976 −0.2159 0.2076 0.0586 −0.0910 LEV −0.0135 −0.0560 0.2468 0.1326 0.0800 −0.0202 −0.1115 0.0710 −0.3668 0.1260 0.0114 −0.0730 −0.0505 0.0286 0.0059 −0.0608 ROA 0.0835 −0.0017 −0.0835 −0.0365 −0.1621 0.2935 0.1243 0.1836 −0.3785 −0.2191 −0.0761 0.1494 0.0690 0.0316 0.0201 0.0521 AGE −0.0517 0.2050 0.0035 0.0327 −0.0052 −0.0782 0.1116 0.1500 0.1356 −0.1572 0.0751 −0.2182 −0.1661 −0.0530 0.1584 −0.0640 SOE −0.1527 −0.1552 0.0284 0.0111 −0.0770 0.0225 −0.1016 0.2503 −0.0114 −0.0088 0.0725 0.3080 −0.2557 0.1925 −0.1593 −0.1557 TOP1 0.0276 −0.2103 0.1078 0.0411 −0.0554 0.1052 −0.1116 0.2112 −0.0921 0.1417 −0.2197 0.3033 −0.5085 0.0178 −0.1051 −0.0553 Herf_2_5 −0.0474 −0.0337 −0.1046 −0.0684 0.0223 −0.0289 −0.0337 −0.0992 −0.0605 0.0386 −0.1918 −0.1153 −0.2741 0.0528 0.0524 0.0821 BSIZE −0.0037 −0.1316 −0.0171 0.0134 −0.0156 0.0669 −0.0950 0.2254 0.0051 0.0708 −0.0512 0.1898 0.0143 0.0481 −0.1884 −0.0883 OUTDIR 0.0811 0.0298 0.0797 0.0649 −0.0386 0.0224 0.0035 0.0741 0.0114 0.0144 0.2018 −0.1666 −0.1154 0.0138 −0.1351 0.0733 DUAL −0.0490 0.0752 −0.0358 −0.0240 −0.0001 −0.0209 0.0767 −0.0967 −0.0524 0.0245 −0.0615 −0.1557 −0.0578 0.0540 −0.0816 0.0532 　 Notes: Person (Spearman) coefﬁcients are below (above) the diagonal. Bolded coefﬁcients are signiﬁcant at p<10% (two-tailed test). Please refer to Table 1 for the deﬁnitions of variables. All continuous variables have been winsorized at the 1% level. China Journal of Accounting Studies 185 5.2. Afﬁliated management and ﬁrm value In Table 5, we investigate the association of afﬁliated management with ﬁrm value through regression analysis. All models in Table 5 are statistically signiﬁcant, with adjusted R squared higher than 40%, showing that our models explain 40% of the variation in ﬁrm value. After controlling for growth, size, leverage and proﬁtability, model 1 shows that afﬁliated management is negatively associated with ﬁrm value at the 1% signiﬁcance level, which means that ﬁrms with afﬁliated management have lower ﬁrm value. Consistent with Hypothesis 1, our result implies that, since controlling shareholders arrange afﬁliated management to get private beneﬁt of con- trol, afﬁliated management is negatively associated with ﬁrm value. Meanwhile, the coefﬁcients of our control variables are consistent with previous literature: growth, measured by Gr_Sales and IND_Q, is positively associated with ﬁrm value; ﬁrm size, measured by SIZE is negatively associated with ﬁrm value (King & Segal, 2009). After we control for ownership structure and board characteristics respec- tively in models 2 and 3, the coefﬁcient of AM is still signiﬁcantly negative at the 1% level, which supports our Hypothesis 1. The regression results in Table 5, descriptive statistics in Table 3 and the correlation matrix in Table 4 imply that although afﬁliated management may play a positive role in improving the ﬁrm’s operation efﬁciency, the cost that afﬁliated management brings about exceeds the beneﬁts, so that ﬁrms with afﬁliated management have lower ﬁrm value. Compared with the sample average of ﬁrm value (1.6781), the value of ﬁrms with afﬁliated management is 4.8% lower. 5.3. Channels through which afﬁliated management affect ﬁrm value If the controlling shareholders arrange afﬁliated management in the listed ﬁrms to enhance their control for private beneﬁts, related party transactions may be one channel through which the afﬁliated management can help the controlling share- holder to expropriate the listed ﬁrms. Fisman and Wang (2010) ﬁnd that the loan- based related party transactions (usually a loan guarantee from the listed ﬁrms) between listed ﬁrms and the controlling shareholders is the main form of RPTs that reduce the ﬁrm value. Based on their ﬁndings, we mainly focus on the relation between afﬁliated management and the loan-based related party transactions. The dependent variable LoanRPT in model 1 of Table 6 measures the value of loan- based related party transactions, deﬁned as the sum of annual loan-based related party transitions between listed ﬁrms and controlling shareholders, divided by total assets. For ﬁrms without these related party transactions, LoanRPT equals zero. The dependent variable LoanRPT_FREQ in model 2 of Table 6 is the frequency of loan-based RPTs in a calendar year. To be consistent with Arnoldi et al. (2013), we also control for AGE in the two models. AGE measures how long the ﬁrm has been listed. Considering that the loan-based RPTs are the main form of RPTs that harm ﬁrm value (Fisman & Wang, 2010), we expect that the afﬁliated management is positively associated with the value and the frequency of loan-based RPTs. The empirical results in Table 6 support our expectation. Compared with the sample average (0.0419), the value of loan-based RPTs with a controlling shareholder increases about 13% in the listed ﬁrms with afﬁliated management. Since the tunnelling behaviour of controlling shareholders will inevitably reduce ﬁrm performance in the long run, the controlling shareholders may disguise their illegal 186 Zheng et al. Table 5. Results for the OLS regression of afﬁliated management and ﬁrm value. (1) (2) (3) Model 1 Model 2 Model 3 AM −0.0847*** −0.0836*** −0.0811*** (4.03) (3.98) (3.91) Gr_Sales 0.0368 0.0462 0.0586* (1.04) (1.34) (1.70) IND_Q 0.9400*** 0.9280*** 0.9240*** (10.55) (10.45) (10.39) SIZE −0.3640*** −0.3450*** −0.3410*** (17.34) (17.81) (17.66) LEV 0.1860 0.0781 0.0809 (1.44) (0.66) (0.69) ROA 0.2310 0.2320 0.2260 (1.08) (1.11) (1.06) SOE −0.0075 0.0023 (0.29) (0.09) TOP1 −0.0055*** −0.0055*** (5.95) (5.93) Herf_2_5 −6.1680*** −6.2750*** (4.43) (4.56) BSIZE 0.0200 (0.38) OUTDIR 0.2430 (1.15) DUAL 0.0287 (0.85) IND Control Control Control YEAR Control Control Control adj. R 0.4370 0.4450 0.4440 F value 91.8300*** 84.9000*** 77.5000*** N 9,933 9,899 9,729 Notes: The dependent variable is TQ,deﬁned as the sum of book value of liability, market value of liquid shares and book value of illiquid shares, divided by book value of total assets. Please refer to Table 1 for the deﬁnitions of the other variables. Each column reports the result of a linear regression with robust standard error clustering and White (1980) heteroscedasticity adjustment at the listed ﬁrm level. Absolute value of T statistics is reported below each coefﬁcient. ***, ** and * denote signiﬁcance at 1, 5 and 10% respectively. behaviour through manipulating information disclosure while arranging for afﬁliated management to seize the private beneﬁt of control. In practice, afﬁliated management may also directly help the controlling shareholders expropriate the beneﬁts of other shareholders through manipulation of information disclosure. For example, afﬁliated management may exaggerate proﬁts to purposely increase the price of a seasoned equity offering. Ultimately, such manipulating behaviour in respect of information dis- closure may evolve into a great amount of irregularity in information disclosure. We use Logit regression to test the association of afﬁliated management with the likelihood of irregularity in information disclosure. If CSRC reports that the listed ﬁrm i in year t manipulated information disclosure, including fraudulent accounting treatment (such as exaggerated proﬁt, and false statement of assets), misleading statements, and omission of substantial information, the dummy variable IRREG of the ﬁrm i in year t equals 1. Otherwise, IRREG equals 0. The regression in Table 7 shows that ﬁrms with afﬁliated China Journal of Accounting Studies 187 Table 6. Results for OLS regression of afﬁliated management and loan-based RPTs. (1) (2) LoanRPT LoanRPT_FREQ AM 0.0056** 0.2910*** (2.18) (2.98) Gr_Sales 0.0034 −0.0080 (0.92) (0.06) IND_Q 0.0030 −0.2420 (0.43) (0.64) SIZE 0.0011 0.3330*** (0.70) (4.51) LEV 0.1400*** 3.1290*** (13.43) (8.14) ROA −0.0078 −0.7850* (0.53) (1.66) AGE −0.0012*** −0.0351** (3.01) (2.54) SOE 0.0046 0.1500 (1.34) (1.08) TOP1 0.0006*** 0.0108*** (5.81) (2.81) Herf_2_5 −0.4620*** −12.5400*** (3.43) (2.87) BSIZE −0.0065 0.0946 (0.99) (0.32) OUTDIR −0.0023 −0.8900 (0.14) (1.30) DUAL −0.0045 −0.1570 (1.43) (1.20) IND Control Control YEAR Control Control adj. R 0.1170 0.0630 F value 9.6560*** 5.2970*** N 9,883 9,883 Notes: The dependent variable in column (1) is LoanRPT,deﬁned as sum of annual loan-based related party transitions between listed ﬁrms and controlling shareholders, divided by total assets. The dependent variable in column (2) is the frequency of loan-based related party transactions in a calendar year. Please refer to Table 1 for the deﬁnitions of the other variables. Each column reports the result of a linear regression with robust standard error clustering and White (1980) heteroscedasticity adjustment at the listed ﬁrm level. Abso- lute value of T statistics is reported below each coefﬁcient. ***, ** and * denote signiﬁcance at 1, 5 and 10% respectively. management are more likely to manipulate information disclosure, which supports our view that the afﬁliated management is negatively associated with the ﬁrm value. 6. Further analysis and robustness tests 6.1. Impact of afﬁliated management on ﬁrm value in different subsamples The previous empirical results support our view that afﬁliated management is nega- tively associated with the ﬁrm value, but the degree to which the afﬁliated management correlates with ﬁrm value may differ under different circumstances. First of all, we use a dummy variable Control to measure whether or not the largest shareholder has abso- lute control over the ﬁrm. If the largest shareholder hold over 50% shares of the listed 188 Zheng et al. Table 7. Results for the Logit regression of afﬁliated management and irregularity. (1) (2) (3) Model 1 Model 2 Model 3 AM 0.3650*** 0.3140*** 0.3420*** (3.34) (2.82) (3.01) SIZE −0.3020*** −0.1950*** −0.1880*** (5.06) (3.13) (2.93) ROA −4.0030*** −4.0470*** −4.0880*** (8.98) (8.96) (8.80) LEV 0.9610*** 0.7520** 0.9110*** (3.02) (2.33) (2.77) SOE −0.6480*** −0.6410*** (5.47) (5.26) TOP1 −0.0071* −0.0066 (1.79) (1.62) Herf_2_5 2.8930 3.5750 (0.50) (0.61) BSIZE 0.1940 (0.75) OUTDIR 0.3390 (0.42) DUAL −0.0639 (0.40) IND Control Control Control YEAR Control Control Control pseudo R 0.1120 0.1250 0.1250 Chi-Square 380.7000*** 425.6000*** 416.1000*** N 10,022 9,972 9,794 Notes: The dependent variable of the Logit regression is IRREG. If CSRC reports that the listed ﬁrm i in year t manipulated information disclosure, including fraudulent accounting treatment (such as exaggerated proﬁt, false statement of assets), misleading statements, and omission of substantial information, the dummy variable IRREG of the ﬁrm i in year t equals 1. Otherwise, IRREG equals 0. Please refer to Table 1 for the deﬁnitions of the other variables. Each column reports the result of a Logit regression with robust standard error clustering at the listed ﬁrm level. Absolute value of Z statistics is reported below each coefﬁcient. ***, ** and * denote signiﬁcance at 1, 5 and 10% respectively. ﬁrms, Control equals 1; otherwise, Control equals 0. We ﬁnd that the coefﬁcient of AM*Control is positive but not signiﬁcant. The ﬁnding implies that the degree to which the afﬁliated management correlates with the ﬁrm value does not signiﬁcantly vary according to whether the largest shareholder has absolute control or not (Table 8). Second, we use a dummy variable Institutional Investor to measure the level of institutional investor holdings. If the level of institutional investor holdings is above the sample average, Institutional Investor equals 1; otherwise, Institutional Investor equals 0. We ﬁnd that the coefﬁcient of AM*Institutional Investor is positive at the 1% signiﬁ- cance level, implying that the degree to which afﬁliated management negatively corre- lates with ﬁrm value is smaller in the ﬁrms with a high level of institutional investor holdings (Table 9). Last of all, since CSRC requires the Chinese listed ﬁrms to reach certain perfor- mance standards if they want to start a seasoned equity offering (SEO) (Li, Yu, & Wang, 2005; Liu, Wei, & Zheng, 2008), the controlling shareholder may choose not to expropriate wealth from a listed ﬁrm immediately before the SEO. We divide our sam- ple into two subgroups according to whether the observation is before the SEO or after China Journal of Accounting Studies 189 Table 8. Results for the sub-group regression: Largest shareholders with and without absolute control. (1) (2) (3) Model 1 Model 2 Model 3 AM −0.0993*** −0.0899*** −0.0875*** (3.73) (3.47) (3.41) AM*Control 0.0517 0.0359 0.0353 (1.22) (0.86) (0.85) Control −0.0414 0.2260*** 0.2220*** (1.11) (4.57) (4.50) Gr_Sales 0.0365 0.0447 0.0573* (1.03) (1.30) (1.67) IND_Q 0.9410*** 0.9340*** 0.9300*** (10.56) (10.54) (10.48) SIZE −0.3630*** −0.3480*** −0.3440*** (16.98) (18.03) (17.91) LEV 0.1830 0.0853 0.0883 (1.40) (0.73) (0.76) ROA 0.2360 0.2440 0.2340 (1.10) (1.17) (1.11) SOE −0.0055 0.0031 (0.21) (0.12) TOP1 −0.0109*** −0.0108*** (7.96) (7.90) Herf_2_5 −5.3650*** −5.5010*** (3.93) (4.07) BSIZE 0.0330 (0.63) OUTDIR 0.2530 (1.22) DUAL 0.0300 (0.89) IND Control Control Control YEAR Control Control Control adj. R 0.4370 0.4500 0.4490 F value 86.9100*** 80.8200*** 74.0700*** N 9,933 9,899 9,729 Notes: In all columns, the dependent variable is TQ,deﬁned as the sum of book value of liability, market value of liquid shares and book value of illiquid shares, divided by book value of total assets. If the largest shareholder hold over 50% shares of the listed ﬁrms, Control equals 1, otherwise, 0. Please refer to Table 1 for the deﬁnitions of the other variables. Each column reports the result of a linear regression with robust standard error clustering and White (1980) heteroscedasticity adjustment at the listed ﬁrm level. Absolute value of T statistics is reported below each coefﬁcient. ***, ** and * denote signiﬁcance at 1, 5 and 10% respectively. the SEO. The results in Table 10 shows that the negative effect of afﬁliated manage- ment on ﬁrm value mainly happens in the years after the SEO, which implies that afﬁl- iated managers are more likely to be negatively associated with ﬁrm value when the controlling shareholders have a higher motivation to expropriate the value of the ﬁrm. 6.2. Ownership and the impact of afﬁliated management on ﬁrm value We also investigate the different effect of afﬁliated managers on state-owned listed ﬁrms and non-state-owned listed ﬁrms. In model 1 of Table 11, we only considered the 190 Zheng et al. Table 9. Results for the sub-group regression: High and low institutional investor holdings. (1) (2) (3) Model 1 Model 2 Model 3 AM −0.1730*** −0.1670*** −0.1660*** (6.25) (5.95) (5.97) AM*Institutional Investor 0.1720*** 0.1670*** 0.1670*** (4.28) (4.16) (4.19) Institutional Investor 0.2170*** 0.2000*** 0.2050*** (6.67) (6.29) (6.42) Gr_Sales 0.0267 0.0436 0.0480 (0.67) (1.14) (1.24) IND_Q 0.9470*** 0.9370*** 0.9220*** (10.82) (10.68) (10.61) SIZE −0.3700*** −0.3500*** −0.3460*** (15.87) (16.88) (16.81) LEV 0.0049 −0.1020 −0.0681 (0.03) (0.83) (0.55) ROA 0.2320 0.2120 0.2830 (0.92) (0.89) (1.17) SOE −0.0091 0.0041 (0.35) (0.15) TOP1 −0.0043*** −0.0044*** (4.64) (4.72) Herf_2_5 −4.4030*** −4.4930*** (3.05) (3.15) BSIZE −0.0092 (0.16) OUTDIR 0.1820 (0.68) DUAL 0.0524 (1.41) IND Control Control Control YEAR Control Control Control adj. R 0.4750 0.4800 0.4820 F value 79.6100*** 73.8700*** 68.2200*** N 7,967 7,935 7,803 Notes: In all the columns, the dependent variable is TQ,deﬁned as the sum of book value of liability, market value of liquid shares and book value of illiquid shares, divided by book value of total assets. If the level of institutional investor holdings is above the sample average, Institutional Investor equals 1, otherwise, 0. Please refer to Table 1 for the deﬁnitions of the other variables. Each column reports the result of a linear regression with robust standard error clustering and White (1980) heteroscedasticity adjustment at the listed ﬁrms level. Absolute value of T statistics is reported below each coefﬁcient. ***, ** and * denote signiﬁcance at 1, 5 and 10% respectively. impact of the afﬁliated managers on ﬁrm value in state-owned listed ﬁrms, and we ﬁnd that even though afﬁliated managers negatively correlate with the value of the ﬁrm, the coefﬁcient is not signiﬁcant. Model 2 analyses the impact of afﬁliated managers on ﬁrm value in non-state-owned listed ﬁrms. We ﬁnd that the presence of afﬁliated managers is negatively correlated with ﬁrm value, and the coefﬁcient is signiﬁcant at the 1% level. Compared with the mean of TobinQ in non-state-owned listed ﬁrms (1.87), the negative effect of afﬁliated managers on ﬁrm value is 8.98%. In model 3, the coefﬁ- cient of AM is signiﬁcantly negative and the coefﬁcient of AM* SOE is signiﬁcantly positive. The absolute value of the two coefﬁcients is quite close, which implies that China Journal of Accounting Studies 191 Table 10. Results for the sub-group regression: Before and after seasoned equity offering. (1) (2) 　 (3) (4) SEO Non-SEO SEO(-2,0) Non-SEO(-2,0) AM 0.0153 −0.0846*** 　 −0.0415 −0.0866*** (0.22) (5.64) (1.19) (5.42) Gr_Sales −0.0869 0.0584* 0.0149 0.0436 (0.62) (1.78) (0.21) (1.26) IND_Q 0.9560*** 0.9170*** 0.9500*** 0.9080*** (3.39) (10.19) (4.71) (9.80) SIZE −0.2310*** −0.3470*** −0.1910*** −0.3620*** (4.48) (28.70) (8.80) (28.02) LEV −0.5490* 0.0995 −0.3600*** 0.1320 (1.79) (1.26) (2.74) (1.56) ROA 4.0290*** 0.1750 3.1500*** 0.1220 (2.99) (0.91) (5.04) (0.62) SOE 0.0555 0.0031 0.0276 0.0049 (0.71) (0.18) (0.72) (0.26) TOP1 −0.0104*** −0.0053*** −0.0069*** −0.0053*** (4.96) (9.58) (5.78) (8.94) Herf_2_5 0.1880 −6.4040*** −0.6540 −6.7090*** (0.03) (8.56) (0.26) (8.58) BSIZE 0.0711 0.0249 0.0679 0.0315 (0.45) (0.68) (0.80) (0.82) OUTDIR −0.3590 0.2860** 0.0861 0.2730* (0.46) (1.97) (0.25) (1.75) DUAL 0.0973 0.0261 0.0602 0.0224 　 (0.97) (1.04) 　 (1.25) (0.83) IND Control Control Control Control YEAR Control Control 　 Control Control adj. R 0.4080 0.4510 0.4650 0.4530 F value 7.3420*** 90.9000*** 16.7200*** 83.8000*** N 509 9,220 　 1,284 8,445 Notes: Column (1) reports the regression result of the sub-samples just in the SEO year. Column (2) reports the regression result of the sub-samples in the Non-SEO year. Column (3) reports the regression result of the sub-samples in the SEO year and two years before the SEO year. Column (4) reports the regression result of the sub-samples which are not in the SEO year or two years before the SEO year. In all the columns, the dependent variable is TQ,deﬁned as the sum of book value of liability, market value of liquid shares and book value of illiquid shares, divided by book value of total assets. Please refer to Table 1 for the deﬁnitions of the other variables. Each column reports the result of a linear regression with robust standard error cluster- ing and White (1980) heteroscedasticity adjustment at the listed ﬁrm level. Absolute value of T statistics is reported below each coefﬁcient. ***, ** and * denote signiﬁcance at 1%, 5% and 10% respectively. the negative effect of afﬁliated managers on ﬁrm value mainly happens in the non- state-owned listed ﬁrms. One possible explanation is that compared with the controlling shareholders (i.e. the state) in the state-owned listed ﬁrms, the controlling shareholders (usually a person or a family) in the non-state-owned listed ﬁrms are more prone to expropriate the minority shareholders with the help of afﬁliated managers for their own private beneﬁts given the weak institutional environment. Meanwhile, the afﬁliated management in the state-owned listed ﬁrms is under more severe monitoring by the government than the afﬁliated management in the non-state-owned listed ﬁrms; thus, the afﬁliated managers’ behaviours that harm ﬁrm value are restrained more strictly in the state-owned listed ﬁrms. 192 Zheng et al. Table 11. Results for the sub-group regression: State-owned listed ﬁrms and non-state-owned listed ﬁrms. (1) (2) (3) SOE Non-SOE Full sample AM −0.0239 −0.1680*** −0.1950*** (1.18) (4.21) (4.49) AM*SOE 0.1720*** (3.61) SOE −0.1020** (2.50) Gr_Sales 0.0147 0.1390** 0.0581* (0.38) (2.35) (1.68) IND_Q 0.8150*** 0.9600*** 0.9230*** (7.68) (6.78) (10.39) SIZE −0.2590*** −0.5010*** −0.3420*** (15.04) (13.70) (17.70) LEV −0.1440 0.3910** 0.0814 (1.31) (1.98) (0.70) ROA 0.3710 0.2850 0.2660 (1.34) (0.99) (1.26) TOP1 −0.0038*** −0.0109*** −0.0055*** (4.31) (5.91) (5.90) Herf_2_5 −2.2130 −14.2600*** −6.0230*** (1.61) (6.85) (4.41) BSIZE −0.0175 0.0405 0.0189 (0.33) (0.37) (0.36) OUTDIR −0.0017 0.5730 0.2510 (0.01) (1.39) (1.20) DUAL 0.0831* −0.0244 0.0306 (1.86) (0.54) (0.90) IND Control Control Control YEAR Control Control Control adj. R 0.4300 0.4800 0.4460 F value 50.6900*** 39.6100*** 76.0700*** N 6,171 3,558 9,729 Notes: Column (1) reports the regression result of the state-owned listed ﬁrms. Column (2) reports the regres- sion result of the non-state-owned listed ﬁrms. Column (3) reports the regression result of the full sample. In all the columns, the dependent variable is TQ,deﬁned as the sum of book value of liability, market value of liquid shares and book value of illiquid shares, divided by book value of total assets. SOE equals 1 if the ﬁrm is controlled by the state and 0 otherwise. Please refer to Table 1 for the deﬁnitions of the other variables. Each column reports the result of a linear regression with robust standard error clustering and White (1980) heteroscedasticity adjustment at the listed ﬁrm level. Absolute value of T statistics is reported below each coefﬁcient. ***, ** and * denote signiﬁcance at 1%, 5% and 10% respectively. 6.3. Different types of afﬁliated management and ﬁrm value First, we divide afﬁliated management into different types based on whether the Chairman (CEO) in the listed ﬁrm is afﬁliated management or not (Table 12). If the Chairman (CEO) in the listed ﬁrm is the controlling shareholder, or the employee of the controlling shareholder, the ﬁrm-year dummy variable AM_Chairman (AM_CEO) equals 1. Otherwise, AM_Chairman (AM_CEO) equals 0. Our empirical results show that no matter whether the Chairman is the afﬁliated management or the CEO is the afﬁliated management, afﬁliated management is negatively associated with ﬁrm value. Secondly, we divide afﬁliated management into two new types, based on whether the afﬁliated management in the listed ﬁrm also holds the chief position of the China Journal of Accounting Studies 193 Table 12. Results for the sub-group regression: Chairman and CEO. (1) (2) Chairman CEO AM_Chairman −0.0827*** (5.59) AM_CEO −0.0841*** (4.24) Gr_Sales 0.0639** 0.0709* (1.98) (1.77) IND_Q 0.9250*** 0.8510*** (10.59) (7.79) SIZE −0.3430*** −0.3440*** (28.76) (23.05) LEV 0.0891 0.1100 (1.13) (1.13) ROA 0.2160 0.0565 (1.10) (0.24) SOE 0.0009 −0.0091 (0.05) (0.41) TOP1 −0.0054*** −0.0053*** (9.95) (7.77) Herf_2_5 −6.1020*** −5.3180*** (8.15) (5.72) BSIZE 0.0286 −0.0110 (0.80) (0.25) OUTDIR 0.2340 0.3280* (1.59) (1.83) DUAL 0.0287 0.0228 (1.18) (0.89) IND Control Control YEAR Control Control adj. R 0.4440 0.4530 F value 93.4500*** 62.6400*** N 9,571 6,096 Notes: Column (1) reports the regression result of the sub-samples that the Chairman is the afﬁliated manage- ment. Column (2) reports the regression result of the sub-samples that the CEO is the afﬁliated management. In all the columns, the dependent variable is TQ,deﬁned as the sum of book value of liability, market value of liquid shares and book value of illiquid shares, divided by book value of total assets. If the Chairman (CEO) in the listed ﬁrm is the controlling shareholder, or the employee of the controlling shareholder, the ﬁrm-year dummy variable AM_Chairman (AM_CEO) equals 1. Otherwise, AM_Chairman (AM_CEO) equals 0. Please refer to Table 1 for the deﬁnitions of the other variables. Each column reports the result of a linear regression with robust standard error clustering and White (1980) heteroscedasticity adjustment at the listed ﬁrm level. Absolute value of T statistics is reported below each coefﬁcient. ***, ** and * denote signiﬁcance at 1%, 5% and 10% respectively. controlling shareholder or not (Table 13). If the afﬁliated management in the listed ﬁrm also holds the chief position of the controlling shareholder (such as Chairman, or CEO), the ﬁrm-year dummy variable AM_Chief equals 1, otherwise 0. If the afﬁliated management in the listed ﬁrms also holds the vice position of the controlling share- holder (such as Vice Chairman, Vice CEO, Vice Executive, or Directors), the ﬁrm-year dummy variable AM_Vice equals 1, otherwise, 0. The empirical study ﬁnds that no mat- ter whether the afﬁliated managers in the listed ﬁrms are in the chief position or the vice position of the controlling shareholders, the afﬁliated management harms the ﬁrm value signiﬁcantly. 194 Zheng et al. Table 13. Results for the sub-group regression: Chief and vice positions in controlling share- holders. (1) (2) Chief position Vice position AM_Chief −0.0580*** (3.57) AM_Vice −0.0934*** (5.36) Gr_Sales 0.0504 0.1020*** (1.33) (2.81) IND_Q 0.9280*** 0.8800*** (9.29) (8.62) SIZE −0.3530*** −0.3450*** (27.00) (24.81) LEV 0.1720** 0.0143 (1.97) (0.16) ROA 0.2800 0.0045 (1.31) (0.02) SOE −0.0050 −0.0152 (0.26) (0.75) TOP1 −0.0050*** −0.0060*** (8.11) (9.30) Herf_2_5 −5.8440*** −5.4340*** (6.92) (6.27) BSIZE 0.0201 −0.0125 (0.50) (0.29) OUTDIR 0.3980** 0.2770 (2.57) (1.64) DUAL 0.0779*** −0.0062 (2.69) (0.24) IND Control Control YEAR Control Control adj. R 0.4540 0.4460 F value 75.9700*** 72.4200*** N 7,628 7,295 Notes: Column (1) reports the regression result of the sub-samples that the afﬁliated management in the listed ﬁrm also holds the chief position of the controlling shareholder. Column (2) reports the regression result of the sub-samples that the afﬁliated management in the listed ﬁrms also holds the vice position of the control- ling shareholder. In all the columns, the dependent variable is TQ,deﬁned as the sum of book value of liabil- ity, market value of liquid shares and book value of illiquid shares, divided by book value of total assets. If the afﬁliated management in the listed ﬁrm also holds the chief position of the controlling shareholder (such as Chairman, or CEO), the ﬁrm-year dummy variable AM_Chief equals 1, otherwise, 0. If the afﬁliated man- agement in the listed ﬁrms also holds the vice position of the controlling shareholder (such as Vice Chairman, Vice CEO, Vice Executive, or Directors), the ﬁrm-year dummy variable AM_Vice equals 1, otherwise, 0. Please refer to Table 1 for the deﬁnitions of the other variables. Each column reports the result of a linear regression with robust standard error clustering and White (1980) heteroscedasticity adjustment at the listed ﬁrms level. Absolute value of T statistics is reported below each coefﬁcient. ***, ** and * denote signiﬁcance at 1%, 5% and 10% respectively. 6.4. Afﬁliated management and ﬁrm operation efﬁciency We also test the effect of afﬁliated management on Return on Assets (ROA), Return on Sales (ROS). We deﬁne Return on Assets as net income divided by total assets. Return on Sales is deﬁned as net income divided by sales. China Journal of Accounting Studies 195 Table 14. Results for the OLS regression of afﬁliated management and ROA. (1) (2) (3) Model 1 Model 2 Model 3 AM 0.0076*** 0.0080*** 0.0087*** (4.05) (4.21) (4.54) Gr_Sales 0.0751*** 0.0755*** 0.0753*** (23.62) (23.74) (23.64) IND_Q 0.0189*** 0.0193*** 0.0193*** (3.30) (3.38) (3.42) SIZE 0.0122*** 0.0124*** 0.0121*** (9.75) (9.89) (9.40) LEV −0.1920*** −0.1890*** −0.1880*** (22.03) (21.74) (21.88) TOP1 0.0003*** 0.0004*** 0.0004*** (4.84) (5.84) (5.83) SOE −0.0141*** −0.0134*** −0.0136*** (6.50) (6.17) (6.24) Herf_2_5 0.5090*** 0.4820*** (5.13) (4.84) BSIZE 0.0146*** (3.03) OUTDIR 0.0016 (0.09) DUAL 0.0052* (1.90) IND Control Control Control YEAR Control Control Control adj. R 0.2670 0.2700 0.2760 F value 37.5000*** 36.9000*** 33.9700*** N 10,071 10,061 9,883 Notes: The dependent variable is ROA,deﬁned as net income divided by total assets. Please refer to Table 1 for the deﬁnitions of the other variables. Each column reports the result of a linear regression with robust standard error clustering and White (1980) heteroscedasticity adjustment at the listed ﬁrm level. Absolute value of T statistics is reported below each coefﬁcient. ***, ** and * denote signiﬁcance at 1%, 5% and 10% respectively. The empirical results from Tables 14 and 15 show that the afﬁliated management is positively associated with the level of ROA and ROS. The empirical results support the view that the afﬁliated managers may play a positive role in increasing the proﬁtability in the ﬁrm. However, we also ﬁnd that ﬁrms with afﬁliated management are more likely to manipulate information disclosure, including fraudulent accounting treatment (such as exaggerated proﬁt, false statement of assets), which may cause the increase of ROA and ROS accordingly. Even if the increase of proﬁtability is authentic, the con- trolling shareholder may turn it into the private beneﬁt of control with the help of afﬁli- ated management through related party transactions or other channels, which may result in the negative impact of afﬁliated management on the ﬁrm value eventually. This paper also carries out other robustness tests (unreported for the space reason). Since the non-tradable shares exist in our capital market for a long time, a variety of methods could be chosen to calculate the ﬁrm value. Referring to Bai et al. (2005), we recalculate the value of the ﬁrm (TQ_R). TQ_R is deﬁned as ‘(book value of liabilities+tradable shares*stock price + non tradable shares*discounted stock price)/the book value of total assets’. The empirical results show that the association between the 196 Zheng et al. Table 15. Results for the OLS regression of afﬁliated management and ROS. (1) (2) (3) Model 1 Model 2 Model 3 AM 0.0273*** 0.0281*** 0.0288*** (3.09) (3.18) (3.19) Gr_Sales 0.2440*** 0.2450*** 0.2420*** (13.74) (13.78) (13.44) IND_Q 0.0496 0.0507* 0.0547* (1.62) (1.66) (1.85) SIZE 0.0594*** 0.0597*** 0.0570*** (9.59) (9.62) (9.17) LEV −0.6770*** −0.6700*** −0.6570*** (13.28) (13.12) (13.17) TOP1 0.0010*** 0.0012*** 0.0013*** (3.30) (3.74) (3.94) SOE −0.0152 −0.0137 −0.0172 (1.49) (1.34) (1.60) Herf_2_5 1.1830*** 1.0710** (2.64) (2.44) BSIZE 0.0652** (2.55) OUTDIR −0.0110 (0.13) DUAL 0.0138 (0.99) IND Control Control Control YEAR Control Control Control adj. R 0.1620 0.1630 0.1640 F value 14.3500*** 14.1200*** 12.9900*** N 10,055 10,045 9,871 Notes: The dependent variable is ROS,deﬁned as net income divided by sales. Please refer to Table 1 for the deﬁnitions of the other variables. Each column reports the result of a linear regression with robust standard error clustering and White (1980) heteroscedasticity adjustment at the listed ﬁrm level. Absolute value of T statistics is reported below each coefﬁcient. ***, ** and * denote signiﬁcance at 1%, 5% and 10% respectively. afﬁliated managers and the ﬁrm value remains the same. Meanwhile, after considering the split share structure reform and other factors, the main ﬁndings in this paper remain the same. 7. Conclusions By analysing the data of Chinese listed ﬁrms in A-share stock market from 2001 to 2009, we ﬁnd that the afﬁliated management is negatively associated with ﬁrm value. This empirical result implies that, in order to obtain more private beneﬁt of control, the controlling shareholder may enhance the control over the ﬁrm through afﬁliated man- agement. After further analysing the channels to explore how afﬁliated management is associated with ﬁrm value, we ﬁnd that the afﬁliated management is positively associ- ated with the value and the frequency of loan-based related party transactions and with the likelihood of irregularity in information disclosure in the listed ﬁrms, which imply that even if afﬁliated management may enhance the ability of shareholders to monitor managers, it also facilitates the expropriation of the interests of the minority sharehold- ers by the controlling shareholder in a transition economy. China Journal of Accounting Studies 197 The empirical results in this paper support the conjecture that the presence of a con- trolling shareholder enhances the control over the ﬁrm through afﬁliated management for the reason of obtaining the private beneﬁts of control. Meanwhile, afﬁliated manage- ment, as a prevailing phenomenon in Chinese listed ﬁrms, has attracted a great amount of attention from practitioners and from regulators. Because of the lack of large-sample empirical evidence, the issue of how to regulate afﬁliated management in Chinese listed ﬁrms has not reached a consensus, which may hurt the interests of minority sharehold- ers, and harm the long-term development of the Chinese capital market. The ﬁndings in this paper suggest that regulators should pay more attention to supervising the afﬁliated management, and strictly examine the economic behaviour and the ﬁnancial information disclosures of the ﬁrms that have afﬁliated management. Lastly, this paper empirically tests the association between afﬁliated managers and the value of the ﬁrm, addressing an inadequacy in the previous study (Claessens et al., 2002). There are several limitations in this paper. First, we analyse the effect of afﬁliated managers on ﬁrm value based only on agency theory, but there may be other theories that explain the relation between afﬁliated management and ﬁrm value. For example, the afﬁliated managers may be recognised as a direct vertical interlock between the control- ling groups and the listed ﬁrms and this kind of interlock may affect ﬁrm value (Arnoldi et al., 2013). Secondly, the ﬁndings in this paper suggest that even though afﬁliated management may enhance the ability of shareholders to monitor managers (decrease the type I agency problem) and increase the operational proﬁtability, it can facilitate the con- trolling shareholder to expropriate the minority shareholders (increase the type II agency problem) and reduce the ﬁrm value eventually in a transition economy. The cost that afﬁliated managers bring about may outweigh the beneﬁts that afﬁliated managers bring in. Although we offer systematic evidence to show how afﬁliated management is associ- ated with ﬁrm value, we do not have direct quantitative analysis and comparison between the costs and beneﬁts brought about by the afﬁliated management. Acknowledgements The authors acknowledge the useful comments and suggestions made by two anonymous reviewers and the editors. The authors take full responsibility for any errors. Funding Gaoping Zheng acknowledges ﬁnancial support from the Young Scholar Research Project of the Ministry of Education (Grant number 13YJC630236). Jian Xue acknowledges ﬁnancial support from National Natural Science Foundation of China (Grant number 71,272,025 and 71,322,201). Notes 1. If the controlling shareholder is one person, the term ‘afﬁliated management’ means that the Chairman or CEO of the listed ﬁrm is exactly the same person. If the controlling shareholder is a group of persons, our deﬁnition of afﬁliated management requires that the Chairman or CEO is one of the group members. 2. In many state-owned listed ﬁrms, and in some non-state-owned listed ﬁrms, the Chairman or CEO of the listed ﬁrms also works as an employee of the parent ﬁrm at the same time. 3. We get similar results when we use the Herﬁndahl index alone to measure the monitoring intensity of other large shareholders. Following Chen, Chen, Schipper, Xu, and Xue (2012), we multiply the Herﬁndahl index by the sum of holdings from the second largest shareholder to the ﬁfth largest shareholder, because the overall power of these shareholders also matters. 198 Zheng et al. References Anderson, R. C., & Reeb, D. M. (2003). Founding-family ownership and ﬁrm performance: Evidence from the S&P 500. The Journal of Finance, 58, 1301–1327. Arnoldi, J., Chen, X., & Na, C. (2013). Vertical interlocks of executives and performance of afﬁliated ﬁrms in state owned Chinese business groups. Working Paper. Bai, C., Liu, Q., Lu, Z., Song, M., & Zhang, J. (2005). An empirical study on Chinese listed ﬁrms’ corporate governance. Economic Research Journal, 2,81–91. (In Chinese) Berle, A. A., & Means, G. C. (1932). The Modern Corporation and Private Property. New York, NY: MacMillan. Bunkanwanicha, P., Fan, J., & Wiwattanakantang, Y. (2013). The value of marriage to family ﬁrms. Journal of Financial and Quantitative Analysis, 48,611–636. Chen, Q., Chen, X., Schipper, K., Xu, Y., & Xue, J. (2012). The sensitivity of corporate cash holdings to corporate governance. Review of Financial Studies, 25, 3610–3644. Chen, C., Li, Z., Su, X., & Yao, Y. (2012). Delegation and sensitivity of CEO turnover to ﬁrm performance within business groups: Evidence from China. Journal of Accounting and Public Policy, 31, 553–574. Chen, X., & Liu, S. (2003). An empirical study on the structural difference among the compensation of managers in Chinese public companies. Economic Research Journal, 8, 55–63. (In Chinese) Claessens, S., Djankov, S., Fan, J. P. H., & Lang, Larry H. P. (2002). Disentangling the incentive and entrenchment effects of large shareholdings. The Journal of Finance, 57, 2741–2771. Claessens, S., Djankov, S., & Lang, Larry H. P. (2000). The separation of ownership and control in East Asian corporations. Journal of Financial Economics, 58,81–112. Coase, R. H. (1937). The nature of the ﬁrm. Economica, 4, 386–405. Doidge, C. G., Andrew Karolyi, G., & Stulz, R. M. (2004). Why are foreign ﬁrms listed in the U.S. worth more? Journal of Financial Economics, 71, 205–238. Du, X., & Zhou, Z. (2010). Political connection and auditor choice. Auditing Research, 2,47–53. (In Chinese) Fama, E. F., & French, K. R. (1998). Taxes, ﬁnancing decisions, and ﬁrm value. The Journal of Finance, 53, 819–843. Fan, J. P. H., Wong, T. J., & Zhang, T. (2007). Politically connected CEOs, corporate governance, and Post-IPO performance of China’s newly partially privatized ﬁrms. Journal of Financial Economics, 84, 330–357. Fisman, R., & Wang, Y. (2010). Trading favors within Chinese business groups. American Economic Review, 100, 429–433. Gong, Y. (2001). Firm performance and top management turnover. Economic Research Journal, 10,75–82. (In Chinese) Jensen, M. C., & Meckling, W. H. (1976). Theory of the ﬁrm: Managerial behavior, agency costs and ownership structure. Journal of Financial Economics, 3, 305–360. Jiang, G., Lee, C., & Yue, H. (2010). Tunneling through intercorporate loans: The China experience. Journal of Financial Economics, 98(1), 1–20. King, M. R., & Segal, D. (2009). The long-term effects of cross-listing, investor recognition, and ownership structure on valuation. Review of Financial Studies, 22, 2393–2421. La Porta, R., Lopez-de-Silanes, F., & Shleifer, A. (1999). Corporate ownership around the world. The Journal of Finance, 54, 471–517. La Porta, R. L., Lopez-de-Silanes, F., Shleifer, A., & Vishny, R. (1998). Law and ﬁnance. Journal of Political Economy, 106,1113–1155. La Porta, R., Lopez-de-Silanes, F., Shleifer, A., & Vishny, R. (2000). Investor Protection and corporate governance. Journal of Financial Economics, 58,3–27. Li, Z., Yu, Q., & Wang, X. (2005). Tunneling, propping and M&A: Evidence from Chinese listed companies. Economic Research Journal, 1,95–105. (In Chinese) Liu, J., Wei, M., & Zheng, G. (2008). The interested-party investment controlled by controlling shareholders: Enhancing efﬁciency or transferring resources. Management World, 3, 133–141. (In Chinese) Luo, D., & Tang, Q. (2009). The performance of institutional environment: Evidence from China’s private listed companies. Economic Research Journal, 2, 106–118. (In Chinese) Myers, S. C. (1977). Determinants of corporate borrowing. Journal of Financial Economics, 5, 147–175. China Journal of Accounting Studies 199 Nguyen, B. D., & Nielsen, K. M. (2010). The value of independent directors: Evidence from sudden deaths. Journal of Financial Economics, 98, 550–567. Rao, Y., Zhang, Y., & Peng, D. (2008). Holding proportion, over guarantee and tunneling: Evidence from the guarantee of Chinese listed companies to their subsidiaries. Nankai Business Review, 1,31–38. (In Chinese) Shi, S. (2010). The transfer of control right, excessive control and the expropriation of large shareholders: Evidence from management turnover in the listed ﬁrms. Journal of Financial Research, 160–176. (In Chinese) Shleifer, A., & Vishny, R. (1986). Large shareholders and corporate control. Journal of Political Economy, 94, 461–488. Villalonga, B., & Amit, R. (2006). How do family ownership, control and management affect ﬁrm value? Journal of Financial Economics, 80, 385–417. White H. (1980). A heteroskedasticity-consistent covariance matrix estimator and a director test for heteroskedasticity. Econometrica, 48, 817–838. Xia, L., & Zhang, T. (2005). Government control, institutional environment and ﬁrm value: Evi- dence from Chinese securities market. Economic Research Journal, 5,40–51. (In Chinese) Yu, M., & Pan, H. (2008). The relationship between politics, institutional environments and private enterprises’ access to bank loans. Management World, 8,9–21. (In Chinese) Zhao, Z., Yang, Z., & Bai, C. (2007). Empirical study on the determinants of top management turnover in the Chinese Listed Firms. Journal of Financial Research, 8,76–89. (In Chinese) Zhu, H. (2002). Analysis of the phenomenon of the top management turnover in Chinese listed ﬁrms. Management World, 5, 126–131. (In Chinese) Zou, F., & Chen, X. (2004). An empirical study on the relationship between the ownership characteristics and the board leadership. Nankai Business Review, 7,49–56. (In Chinese)

Journal

China Journal of Accounting Studies – Taylor & Francis

Published: Jul 3, 2014

Keywords: affiliated management; firm value; private benefit of control

References

Founding-family ownership and firm performance: Evidence from the S&P 500
An empirical study on Chinese listed firms’ corporate governance
The value of marriage to family firms
The sensitivity of corporate cash holdings to corporate governance
Delegation and sensitivity of CEO turnover to firm performance within business groups: Evidence from China
An empirical study on the structural difference among the compensation of managers in Chinese public companies
Disentangling the incentive and entrenchment effects of large shareholdings
The separation of ownership and control in East Asian corporations
The nature of the firm
Why are foreign firms listed in the U.S. worth more?
Political connection and auditor choice
Taxes, financing decisions, and firm value
Politically connected CEOs, corporate governance, and Post-IPO performance of China’s newly partially privatized firms
Trading favors within Chinese business groups
Firm performance and top management turnover
Theory of the firm: Managerial behavior, agency costs and ownership structure
Tunneling through intercorporate loans: The China experience
The long-term effects of cross-listing, investor recognition, and ownership structure on valuation
Corporate ownership around the world
Law and finance
Investor Protection and corporate governance
Tunneling, propping and M&A: Evidence from Chinese listed companies
The interested-party investment controlled by controlling shareholders: Enhancing efficiency or transferring resources
The performance of institutional environment: Evidence from China’s private listed companies
Determinants of corporate borrowing
The value of independent directors: Evidence from sudden deaths
Holding proportion, over guarantee and tunneling: Evidence from the guarantee of Chinese listed companies to their subsidiaries
The transfer of control right, excessive control and the expropriation of large shareholders: Evidence from management turnover in the listed firms
Large shareholders and corporate control
How do family ownership, control and management affect firm value?
A heteroskedasticity-consistent covariance matrix estimator and a director test for heteroskedasticity
Government control, institutional environment and firm value: Evidence from Chinese securities market
The relationship between politics, institutional environments and private enterprises’ access to bank loans
Empirical study on the determinants of top management turnover in the Chinese Listed Firms
Analysis of the phenomenon of the top management turnover in Chinese listed firms
An empirical study on the relationship between the ownership characteristics and the board leadership

Sign up for your 14-Day Free Trial Now!

Read and print from thousands of top scholarly journals.

System error. Please try again!

or

By signing up, you agree to DeepDyve’s Terms of Service and Privacy Policy.

Already have an account? Log in

Cite

Copy and paste the desired citation format or use the link below to download a file formatted for EndNote

APA
Zheng, G., Xue, J., & Chen, X. (2014). Affiliated management and firm value: Evidence from China. China Journal of Accounting Studies, 2(3), 172-199.
MLA
Zheng, Gaoping, Jian Xue, and Xiao Chen. "Affiliated management and firm value: Evidence from China." China Journal of Accounting Studies 2.3 (2014): 172-199.

Reference Managers
Export to EndNote
Export in RIS Format

Our policy towards the use of cookies

All DeepDyve websites use cookies to improve your online experience. They were placed on your computer when you launched this website. You can change your cookie settings through your browser.

cookie policy

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Affiliated management and firm value: Evidence from China

Affiliated management and firm value: Evidence from China

Affiliated management and firm value: Evidence from China

Abstract

Journal

Recommended Articles

References

Our policy towards the use of cookies