1.1 Research Background

1.1.1 Practical Background

1.1.1.1 Emergence and Development of AH Companies

Alongside the deepening of economic globalization, international equity financing has experienced rapid development since the 1980s. Cross-listing, as an important way of international equity financing, has attracted much attention in academia (Stulz, 1999; Lang et al., 2003) ^[3,4] . With the constant development of China's economy, many mainland companies have been listed on overseas exchanges to raise capital. As a special administrative region of China, Hong Kong has a more internationalized and free capital market with sound regulatory systems than mainland China. Thus, Hong Kong has a good market guarantee for companies to finance. Nevertheless, the Hong Kong market and mainland market are similar to each other in terms of some rules and regulations, which, to an extent, facilitates mainland companies to list in Hong Kong. As a result, the Hong Kong market has become the prevailing choice for mainland companies to list abroad. In 1993, mainland Chinese company Tsingtao Brewery Co., Ltd.(Stock Code: 600600. SH; 00168. HK) was cross-listed on the A-share market in mainland China and the H-share market in Hong Kong as the first "AH company" . ⁽¹⁾Since then, a growing number of mainland companies have tried entering the Hong Kong market.

Mainland companies play an important role in the Hong Kong market. According to "Hong Kong Exchanges and Clearing Limited (HKEx) Fact Book 2018" , there were 2315 companies listed on the Hong Kong market at the end of 2018. Among them, 1146 were mainland companies, accounting for 49.50% of the total. The market capitalization (turnover value) of mainland companies was 20193 (15480) billion HKD, accounting for 67.51% (79.55%) of the market total (equity turnover). Among the mainland companies listed in Hong Kong, there were 243 H-share companies at the end of 2018, comprising 10.50% of total companies listed in Hong Kong. The market capitalization of H-share companies made up 19.98% of the equity total. Accordingly, H-share companies have occupied an important place in the Hong Kong market. The number of AH companies has been increasing constantly since the cross-listing of Tsingtao Brewery. At the end of 2018, 109 companies were cross-listed on the A-share and the H-share markets. Figure 1-1 illustrates the growth of AH companies from 1993 to 2018 on the basis of the data collected from the China Stock Market & Accounting Research (CSMAR) database. Listed in mainland China and Hong Kong and increasing in number, AH companies have incited the interest of many Chinese researchers on institutional environment, institution changes, and relevant accounting and auditing issues (Ke et al., 2015; Tian et al., 2017) ^[5,6] .

1.1.1.2 Persistent Existence of Dual Audit After Deregulation

Since the establishment of the H-share market in 1993, regulators have imposed the dual audit system on AH companies as a mandatory requirement. In the early 1990s, substantial differences existed between the Chinese Accounting Standards (CAS) and the International Financial Reporting Standards (IFRS) [or the Hong Kong Financial Reporting Standards (HKFRS)]. Thus, the information demands of Chinese and international investors largely differed from each other. In addition, mainland local audit firms were at the initial stage of their development and considered as having relatively low audit quality (MOF, 2009a) ^[7] . Under this circumstance, an AH company was required to adopt the dual audit system[specifically, hiring a mainland-based auditor to audit and issue an audit report for the A-share financial statements prepared based on the CAS and a Hong Kong-based auditor to audit and issue an audit report for the H-share financial statements prepared based on the IFRS (or HKFRS) each year]. The objective is to improve accounting information credibility and satisfy diversified information demands of Chinese and international investors.

In 2006, mainland China harmonized its accounting standards with the IFRS. The release of the new CAS by the Ministry of Finance (MOF) symbolized the CAS's convergence (equivalence) with the IFRS (HKFRS). Subsequent revisions on the CAS are on the road to achieving continuous and full convergence with the IFRS (MOF, 2010a) ^[8] . The convergence (and equivalence) of accounting standards between mainland China and Hong Kong, in turn, led regulators to consider that the dual audit system is no longer needed for most AH companies, although AH companies can continue to use the dual audit voluntarily. In addition, the strategic position of the H-share market in the internationalization of mainland local audit firms and the policy support of the Chinese government for mainland local audit firms to become bigger and stronger drove regulators to consider the futility of the existence of the mandatory dual audit system. Taken together, regulators deemed that the mandatory requirement should be abolished. Following a series of efforts by regulators (see Section 3.1 for details), the mandatory dual audit system was officially abolished in 2010.

Continuing the dual audit system or not has become a voluntary choice for an AH company since the deregulation of the mandatory dual audit system in 2010 (HKEx, 2009) ^[9] . The company only needs to retain the A-share audits if it considers using the dual audit system unnecessary. Regulators expected most AH companies to follow this policy change and dismiss their Hong Kong-based auditors. However, this expectation was not the case in reality. For example, the news showed that AH companies that canceled the dual audit system were only 10%to 20% of total H-share companies in the initial years after the deregulation of the mandatory dual audit system. This percentage was quite below the original expectation and made some practitioners doubt the achievements of this policy change. ⁽²⁾Figure 1-2 illustrates the number and proportion of companies that continue to keep the dual audit system in AH companies from 2010 to 2018 based on the data collected from the CSMAR database. Only 1 firm responded to the abolition of the mandatory dual audit system in the year of the official announcement (i. e., 2010). Despite showing a generally declining trend, the proportion of companies keeping the dual audit system accounted for 62.39% (68/109) of AH companies even 8 years after the deregulation. This figure, as well as the practical news, suggests that the majority of AH companies persistently use the dual audit system in the absence of the mandatory requirement. The deregulation of the mandatory dual audit system does not mean the disappearance of the dual audit system. The implementation of this policy change is not as successful as regulators originally expected.

The contributions of the dual audit system cannot be denied. These include ensuring auditor independence, improving corporate governance, providing accurate and complete financial information for investors, and accumulating international business experience for mainland local audit firms (Sun, 2014) ^[1] . Even after the abolition of the mandatory requirement, the dual audit system continues to comprise a large proportion of AH companies. Therefore, the dual audit system has inherent rationality and value for academic research compared with the traditional single audit. In addition, sufficient variation in audit systems adopted by A-share companies (including AH companies) provides me a good opportunity to study issues about the dual audit system, specifically, the economic consequences of the deregulation of the mandatory dual audit system in the Chinese market. Issues such as whether AH companies change their audit systems and whether their audit quality is affected have become investors' major concerns since the deregulation (Sun, 2014) ^[1] . An in-depth investigation on the economic consequences of the deregulation of the mandatory dual audit system can enhance understanding of the influences of this audit system on financial reporting quality, audit quality and investors' perceptions and decisions. Such an investigation can provide regulators and policy-makers useful references for future audit regime arrangements in China.

1.1.2 Theoretical Background

1.1.2.1 Research on the Economic Consequences of the Dual Audit System and Its Change Persists

Academics pay continuous attention to the dual audit system and its change. Some researchers discuss the potential consequences of this policy change through theoretical and qualitative analyses (Zheng et al., 2011; Weng et al., 2012) ^[10,11] . They suggest that abolishing the mandatory dual audit system and encouraging mainland local audit firms to undertake the audits of H-share companies bring not only opportunities but also challenges for these audit firms. Other researchers explore the audit-related consequences of this audit system and its change from the empirical aspect. Among these researchers, some suggest that the dual audit system helps ensure higher audit quality based on comparisons among A-share listed companies (including domestic-listed-only A-share companies and AH companies) (Ke et al., 2015; Tian et al., 2017; Lin et al., 2014; Wang, 2014) ^[5,6,12,13] . Other researchers focus on the effect of this policy change on market share and audit fees (Sun, 2013) ^[14] . Wang and Xin (2011) ^[15] show the differences in motivation between Big 4 and non-Big 4 auditors when performing the dual audit. The abolition of the mandatory dual audit system for AH companies in 2010 does not mean that the dual audit system has disappeared since then. Zhong and Zhang (2019) ^[16] suggest that a deep investigation on this audit system contributes to enhancing understanding of its influences and providing useful references for audit system arrangements in the future.

1.1.2.2 Research on the Cost of Equity Is Constantly Enriched

As a major reflection of investors' expectation and evaluation of risks on their investments (Modigliani and Miller, 1959; Solomon, 1963) ^[17,18] , researchers have paid considerable attention to the cost of equity for decades. Consequently, theoretical and empirical research on the cost of equity is rich and constantly growing. Some scholars make progress on how to measure the cost of equity through theoretical and model analyses (Gebhardt et al., 2001; Sharpe, 1964; Feltham et al., 1995; Gordon et al., 1997; Claus et al., 2001; Easton, 2004; Ohlson et al., 2005) ^[2,19-24] . Other scholars explore the determinants of the cost of equity based on theoretical and empirical studies. The literature reaches a consensus that a strong institutional environment, for example, sound legal system and strong investor protection, lowers the cost of equity (Levitt, 1998; Hail et al., 2005; Himmelberg et al., 2004; Jiang et al., 2008; Xiao et al., 2008; Hong and Chen, 2016) ^[25-30] . Moreover, the cost of equity can be affected by governmental intervention, and the impact of governmental intervention varies with factors, like ownership property (Xu and Lv, 2007) ^[31] , political connection (Xiao et al., 2010; Zhao et al., 2012) ^[32,33] , earnings management level (Wang, 2013) ^[34] , and financial ecological environment (Zou and Yang, 2014) ^[35] . Firm size is another important determinant of the cost of equity, although no uniform conclusion on the direction of the impact has been reached. Some researchers hold that firm size is negatively associated with the cost of equity (Botosan, 1997; Botosan and Plumlee, 2002; Cao et al., 2012) ^[36-38] , whereas other researchers obtain opposite findings (Tian et al., 2003; Ye et al., 2004) ^[39,40] . Previous studies suggest that leverage (or debt ratio) also affects the cost of equity, although their conclusions are not totally consistent (Modigliani et al., 1959; Tian et al., 2003; Fama et al., 1992; Richardson et al., 2001; Chen et al., 1999; Tang et al., 2005; Jin, 2008) ^{[17,39,41-45]} . Researchers also find that book-to-market ratio has a significant impact on the cost of equity. Some scholars show that this impact is positive (Fama et al., 1992; Berk et al., 1999; Cheng et al., 2006; Lei et al., 2014) ^[41,46-48] , whereas other scholars find it negative (Ye et al., 2004) ^[40] . Corporate governance is suggested to significantly affect the cost of equity, and scholars draw similar conclusions that strong corporate governance helps reduce the cost of equity (Cheng et al., 2006; Ashbaugh et al., 2004; Huang, 2005; Mazzotta et al., 2014; Jiang, 2009; Jiang et al., 2009; Yan, 2011; Lai et al., 2016; Ke, 2017) ^[47,49-56] . Disclosure quality is yet another determinant of the cost of equity suggested by the previous literature. Most studies agree with the negative relationship between disclosure quality and the cost of equity (Richardson et al., 2001; Cheng et al, 2006; Diamond et al., 1991; Mangena et al., 2016; Wang et al., 2004; Zeng et al., 2006; Shen et al., 2010; Li et al., 2016; Zhou et al., 2016) ^{[42,47,57-63]} . Other studies demonstrate that no significant or even a positive relationship exists (Botosan et al., 2002; Kim, 1993; Yang et al., 2012; Li et al., 2013; Che et al., 2016; Zhang et al., 2017) ^[37,64-68] . Accounting conservatism can also affect the cost of equity, with most findings supporting a negative association between them (Easley and O' Hara, 2004; Suijs, 2008; Lafond et al., 2008; Lara et al., 2011; Guay et al., 2018; Wang et al., 2013; Lu et al., 2016; Li et al., 2016; Wang et al., 2016; Ma et al., 2017) ^[69-78] . Audit quality is suggested to be another important determinant of the cost of equity by previous studies, and high audit quality contributes to reducing the cost of equity (Choi et al., 2014; Houqe et al., 2017; Yu et al., 2008; Wang et al., 2009; Chen et al., 2011; Li et al., 2017; Zhu et al., 2017) ^[79-85] . The cost of equity is also affected by liquidity. Most researchers find that the cost of equity is lower when stock liquidity is higher (Amihud et al., 1986; Jacoby et al., 2000; Saad et al., 2017; Wang et al., 2012; Yang, 2016) ^[86-90] , whereas a few Chinese researchers have contrary findings (Shen et al., 2005) ^[91] . In addition, market risk (beta) is suggested as a major factor that determines the cost of equity (Sharpe, 1964; Fama et al., 1973; Ross, 1976) ^[19,92,93] . However, this factor shows different impacts on the cost of equity in the Chinese market in various studies (Shen et al., 2005; Wang, 2013; Wang et al., 2014) ^[91,94,95] .

Overall, the cost of equity is closely related to many participants in the capital market as a major concern in companies' financing activities. Previous studies propose a variety of measurements on the cost of equity and a wealth of research on its determinants. I assume that after the deregulation of the mandatory dual audit system, investors may perceive information risk to increase and thereby require higher risk compensation (rate of return) than before. However, existing studies have not yet paid enough attention to the economic consequences of this policy change or this potential important determinant of the cost of equity.

1.1.2.3 Research on Audit Quality Is Popular and Enduring

As an important mechanism to mitigate information asymmetry, alleviate agency conflicts and protect investors, the quality of independent auditing lies at the core of accounting and auditing research (DeFond and Zhang, 2014) ^[96] . Audit quality has been a hot topic in academia for nearly 40 decades, with a large volume of relevant theoretical and empirical research. Audit quality jointly depends on an auditor's competence and independence (DeFond and Zhang, 2014; DeAngelo, 1981a; Watts and Zimmerman, 1983; Palmrose, 1986; Teoh and Wong, 1993; Lee and Gu, 1998; Francis, 2011; Zhang and Zhang, 1997; Feng, 2002) ^[96-104] . Similar to the cost of equity, researchers propose various methods to measure audit quality (DeFond and Zhang, 2014) ^[96] . Prior studies suggest that audit firm size is an important determinant of audit quality. Most researchers suggest a positive relationship between audit firm size and audit quality (DeAngelo, 1981a; Dopuch and Simunic, 1982; Firth and Smith, 1992; DeFond et al., 1999; Wu and Li, 2006; Lin and Wang, 2013; Wu et al., 2015; Lu et al., 2017) ^[97,105-111] . Others have different opinions (Wu et al., 2005; Guo, 2011) ^[112,113] . Researchers generally find that auditor industry expertise also has a significant positive impact on audit quality (Owhoso et al., 2002; Krishnan, 2003; Liu et al, 2010; Xie and Sun, 2010; Fan et al., 2013) ^[114-118] . By contrast, others draw opposite conclusions in the Chinese setting (Cai and Xian, 2007) ^[119] . Additionally, audit firm internal governance is suggested to significantly affect audit quality (Muzatko et al., 2004; Firth et al., 2012; Zhu, 2013; Chen, 2014; Liu and Wang, 2014; Liu et al., 2015; Huang et al., 2017) ^[120-126] . From the audit client's side, agency conflicts of clients have a significant impact on audit quality, that is, severer agency conflicts drive their demand for higher audit quality (Chow, 1982; Francis and Wilson, 1988; Lennox, 2005a; Blouin et al., 2007; Francis et al., 2009; Guedhami et al., 2009; Wang et al., 2006; Gao and Zhang, 2011; Qi and Han, 2016) ^[127-135] . Moreover, corporate governance of clients significantly influences audit quality. Many findings support that clients with stronger corporate governance demand higher audit quality (Carcello and Neal, 2000; Beasley and Petroni, 2001; Beasley and Salterio, 2001; Stewart and Kent, 2006; Engel et al., 2010; Cassell et al., 2012; Xiao, 2006; Yu and Liu, 2007; Shen et al., 2009; Pan and Wen, 2017) ^[136-145] . Nonetheless, a few studies have different findings (Ye et al., 2015) ^[146] . The relationship between auditors and clients is also suggested to play an important role in determining audit quality. For example, literature suggests that auditor tenure affects audit quality. Some studies note that long auditor tenure helps increase audit quality (Loebbecke et al., 1989; Petty and Cuganesan, 1996; Johnson et al., 2002; Geiger and Raghunandan, 2002; Myers et al., 2003; Litt et al., 2014; Xia et al., 2005; Chen and Xia, 2006; Liu and Tang, 2009) ^[147-155] , whereas other studies argue that long auditor tenure adversely affects audit quality (DeAngelo, 1981b; Shockley, 1981; Arrunada and Cá ndido, 1997; Arel et al., 2005; Carey and Simnett, 2006; Ye et al., 2011; Liu, 2006; Jiang and Li, 2011; Li et al., 2014) ^[156-164] . Mixed evidence is also found (Yu and Li, 2003; Dong and Zhang, 2010; Shen et al., 2010) ^[165-167] . Another determinant of audit quality relevant to the relationship between auditors and clients is client importance. Some researchers argue that client importance has an adverse impact on audit quality (DeAngelo, 1981b; Choi et al., 2010; Kanagaretnam et al., 2010; Hossain et al., 2016; Zhang, 2017; Lu et al., 2012; Cao et al., 2012; Jia and Li, 2015; Xie et al., 2017; Yang and Yang, 2019) ^{[156,168-176]} , whereas others hold the contrary viewpoint or find mixed evidence (Stice, 1991; Bonner et al., 1998; Zhang, 1999; Reynolds and Francis, 2000; Gaver and Paterson, 2007; Fung et al., 2016; Fang et al., 2004; Yu et al., 2008; Chen et al., 2010; Liu, 2012; Li and Liu, 2013) ^[177-187] . In addition, private relationships between auditors and clients are suggested to affect audit quality. In particular, many studies suggest that private relationships, such as employment relationships (Dowdell and Krishnan, 2004; Menon and Williams, 2004; Lennox, 2005b; Wu, 2006; Liu, 2011; Liu and Zhang, 2014; Chen et al., 2015; Wu et al., 2015) ^[188-195] and social ties (Guan et al., 2016; He et al., 2017) ^[196,197] , adversely affect audit quality. A few studies draw different conclusions (Geiger et al., 2005; Geiger and North, 2006; Baber et al., 2014; Deng et al., 2017) ^[198-201] . Scholars also find that external environmental factors affect audit quality, such as institutional environment (Lang et al., 2003; Ke et al., 2015; Tian et al., 2017; Khurana and Raman, 2004; Huijgen and Lubberink, 2005; Choi et al., 2008; Francis and Wang, 2008; Francis et al., 2011; Wang and Tan, 2010; Xin and Wang, 2010) ^{[4-6,202-208]} , litigation risk (Feng and Liang, 2010; Zhai and Liao, 2011) ^[209,210] , regulatory risk (Song and Xiao, 2012; Liu, 2016; Huang and Li, 2016) ^[211-213] , and media coverage (Zhou and Yao, 2015; Zhang et al., 2016; Wu and Li, 2017) ^[214-216] .

Overall, as an enduring hot topic in accounting and auditing research, audit quality has been measured through many proxies from different aspects, such as earnings quality, audit opinion, earnings restatements and audit fees. A variety of studies on its determinants exist, and research on audit quality is ongoing. Specific to the research setting of this study, previous studies assume a mechanism through which the dual audit system ensures high audit quality from the perspective of the institutional environment faced by auditors (Ke et al., 2015) ^[5] . However, the mechanism remains an abstract description without any measurement or tests on its validation.

1.1.2.4 Limitations of Existing Literature

Taken together, academics pay continuous attention to the consequences of the dual audit system and its change and have obtained enlightening research findings. However, research on this topic may have limitations such as the ones listed below.

(1) The existing research on the mandatory dual audit system for AH companies and its abolition pays attention only to the audit-related consequences, such as audit quality, audit market share and audit fees. Although the mandatory dual audit system was officially abolished in 2010, a remarkable proportion of AH companies choose to use this system after this deregulation, as shown in Figure 1-2. The abolition of the mandatory dual audit system does not mean that the dual audit system has disappeared since 2010. Further investigation on this audit system can explain its influences on financial reporting quality, audit quality and investor protection, and provide useful references for audit system arrangements in the future. Under this circumstance, learning about whether and to what extent the deregulation of the mandatory dual audit system has consequences on the capital market is of academics and practitioners' interest. However, evidence on the capital market-related consequences of this policy change, such as the cost of equity, is limited.

(2) The existing research on the dual audit system for AH companies reaches a consensus that the AH companies using the dual audit have higher audit quality compared with other A-share listed companies that do not use the dual audit (including non-cross-listed A-share companies and some AH companies). Based on this consensus, it can be implied that AH companies potentially experience an audit quality deterioration after they cancel the dual audit system. However, people cannot rule out the possibility that the consistent finding in the existing research is driven by systematic differences between cross-listings and non-cross-listings.

(3) The existing research on the dual audit system for AH companies provides a detailed explanation of the positive spillover effect from Hong Kong-to mainland-based auditors in the dual audit process (Ke et al., 2015) ^[5] . From the perspective of the institutional environment faced by the engaged auditors, this positive spillover is suggested to serve as an important mechanism by which the dual audit system ensures high audit quality. Nonetheless, this positive spillover effect stays an abstract description. The narratives of this effect have not yet been translated into any valid empirical measure. Whether and to what extent the existence (loss) of this positive spillover affects the audit quality of companies keeping (canceling) the dual audit system remain open questions.

(4) The existing research on the cost of equity puts forward measurements and main determinants of the cost of equity, providing solid foundations for the design of the empirical research on the cost of equity in this study. Nonetheless, while independent auditing is suggested to have a significant impact on the cost of equity by previous literature, it is underexplored whether a change in a particular audit system, for example, the deregulation of the mandatory dual audit system, can induce any variation in the cost of equity. Based on my research questions in Section 1.2.1, I assume that investors may perceive information risk to increase and thereby require higher risk compensation (rate of return) after the deregulation of the mandatory dual audit system. However, previous literature has paid limited attention to this factor that potentially affects the cost of equity in the Chinese capital market.

(5) The existing research on audit quality proposes various measurements and suggests main determinants of audit quality, serving as useful references for the design of the empirical research on audit quality in this study. Nonetheless, specific to the audit quality-related issue investigated in this study, prior studies have largely overlooked how to empirically measure and test the mechanism [as stated in (3)] by which the dual audit system ensures high audit quality from the perspective of the institutional environment.