Can mandatory dual audit reduce the cost of equity? Evidence from China

Abstract

In China, a mandatory dual audit system for domestic A-share firms cross-listed on the Hong Kong stock market (i.e. AH companies) was abolished in 2010. Since then, AH companies have been allowed to choose to have a dual audit or a single audit. We find that the mandatory dual audit regime before the deregulation is associated with a lower cost of equity than voluntary dual audit after the deregulation. Furthermore, the lower cost of equity under the mandatory dual audit regime is greater in companies exposed to stronger financial constraints and with higher agency costs, and is not attenuated by alternative voluntary audits. Our results are not affected by accounting standards convergence and audit quality, and are robust to various model specifications. Our results suggest that the role of mandatory dual audit in mitigating agency costs and information asymmetry is not replaceable by voluntary dual audit.

Keywords:

• mandatory dual audit
• deregulation
• compliance
• cost of equity
• China

Subject classification codes:

• G31
• G32
• M42

Acknowledgements

We are grateful to the editors Mark Clatworthy, Juan Manuel García Lara, and Edward Lee and two anonymous reviewers for their helpful comments on the paper. We also thank Liuchuang Li, Baolei Qi, Richard Macve, Liansheng Wu, Hanwen Chen, and participants at the 2018 Accounting and Business Research Special Issue Conference: Auditing in China for their useful suggestions on earlier drafts of the paper. This work was supported by the [National Natural Science Foundation of China] under Grant [number 71672141; 71502134; 71502139; 71602160].

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 A dual audit differs from a joint audit. Joint audit is defined as an audit in which audit areas are audited by two or more audit firms. In a joint audit process, the engaged audit firms share audit efforts, cross-check each other’s work, and finally issue one single audit report for the areas mutually audited by them (Francis et al. 2009, Ratzinger-Sakel et al. 2013, Holm and Thinggaard 2018). In a dual audit, however, the two engaged auditors audit respective audit areas and issue separate audit reports; they do not share the audit work, produce a mutual audit report, or share responsibility for the audit.

2 According to HKEx (2009), the listing rules on the Hong Kong stock market stipulate that an AH company can prepare its H-share financial statements using either IFRS or HKFRS. HKFRS became fully converged with IFRS from 1 January 2005. Since then, there have been no differences between HKFRS and IFRS; reconciliations between HKFRS and IFRS are no longer necessary. Meanwhile, as stated in Section 2, the Ministry of Finance (MOF) in China released the new accounting standards for enterprises on 15 February 2006, symbolising CAS’s convergence with IFRS. Given these developments in practice, the beginning year of our sample period (i.e. 2007) is not only the second year after the full convergence of HKFRS with IFRS, but also the year when the new CAS took effect. Therefore, during our sample period, there are no substantial changes in either the accounting standards used by an AH company to prepare its A-share financial statements or the accounting standards used to prepare its H-share financial statements.

3 On 10 December 2010, the China Securities Regulatory Commission (CSRC) and the Hong Kong Stock Exchange (HKEx) finally reached an agreement on a mutual recognition of accounting and auditing standards and audit firms from the other side. Since then, HKEx has accepted mainland accounting and auditing standards and allowed mainland audit firms for auditing mainland incorporated companies listed in Hong Kong (including AH companies).

4 During our sample collection, we delete the IPO observations and observations that are delisted, terminated, and under special treatment (see Section 5.1). This practice ensures that all sampled companies (including the single-audited sample for the analysis in the Appendix) are cross-listed on the A- and H- share markets and governed by both mainland and Hong Kong regulations when they experience a regime change from the mandatory dual audit to the voluntary dual audit.

5 Kausar et al.’s (2016) study on the real effects of U.K. private firms’ voluntary audit choice cautions that the economic effects of audit quality may confound those of the audit choice. As our study also involves voluntary audit choices (although with different research setting and orientation from those of Kausar et al. 2016), for rigorousness, we address the potential audit quality issue in our analyses.

6 Pure A-share companies and AB-share companies are listed only on China’s domestic stock market. Both kinds of companies issue A-shares to domestic investors, while the latter also issue B-shares to overseas investors. AH companies are cross-listed on both the domestic stock market and the Hong Kong stock market; they issue A-shares to domestic investors and H-shares to overseas investors.

7 Apart from protecting investors, imposing the DADRS also serves as a part of the strategy of growing the capacity of mainland China-based audit firms (Macve 2020).

8 Approved mainland China-based audit firms comprise BDO China Shu Lun Pan CPAs, Pan-China CPAs, BDO Da Hua CPAs, ShineWing CPAs, Ernst & Young Hua Ming CPAs, Crowe Horwath CPAs, Grant Thornton CPAs, PwC Zhong Tian CPAs, Deloitte & Touche Hua Yong CPAs, KPMG Hua Zhen CPAs, RSM China CPAs, and Wuyige CPAs.

9 We are enlightened by the disclosure literature to develop our main argument related to the dual audit regime change, but note that complying with a mandatory disclosure regime is not totally parallel with complying with a mandatory audit regime in playing the commitment device role. For example, in investors’ minds, mandatory disclosure implies that firms commit to disclose certain information (which is not necessarily verifiable), while mandatory audit suggests that firms commit to disclose trustful information verified by independent auditors (Ball et al. 2012).

10 Rock (2002) suggests that the SEC disclosure regulation in the U.S. is implemented along with the powerful enforcement of the U.S. regulators and securities laws, which makes violating this regulation costly. Hence, firms’ mandatory disclosure under the SEC disclosure regulation becomes an ex-ante credible commitment to investors and thereby helps address agency problems effectively. Leuz and Verrecchia (2000) that study German firms propose similar argument. The regulation unifies the format and quantity of disclosure and constantly makes adjustment on firms’ reporting duties, which are also contributory to solving agency issues (Rock 2002).

11 Apart from the oversight by the CSRC and the MOF in mainland China, AH companies are under the supervision of the Hong Kong regulatory agencies, including the HKEx, the HKICPA, the Hong Kong Securities and Futures Commission, and the Independent Commission against Corruption. Independent from the government, they are powerful in monitoring AH companies, investigating their audit engagements, and sanctioning irregularities (Ke et al. 2015), forming a good complement to the mainland enforcement. The combination of the mainland China and Hong Kong enforcement infrastructures ensures the effective implementation of the mandatory dual audit requirement.

12 Owing to data limitations, we focus on the financial data reported in the A-share (rather than the H-share) market for all the sample companies when conducting the empirical tests.

13 According to the Securities Laws in China, a firm that reports losses in 2 consecutive years or more should be labelled with a special treatment sign (ST/*ST). These firms are subject to a daily price fluctuation limit of 5% and will be terminated from listing if their financial or operating conditions become even worse (e.g. reporting losses in 4 consecutive years and reporting negative net assets in 3 consecutive years).

14 As we aim to examine whether any change in the cost of equity from mandatory dual audit to voluntary dual audit will be attributed to a loss of the role that complying with the mandatory dual audit requirement plays in reducing agency costs and information asymmetry rather than a change in audit quality, we limit our analyses by comparing mandatory dual audit with voluntary dual audit; both of which observations have received the dual audit. As a consequence, the economic effects of the audited financial information can be effectively measured and controlled in the empirical process for both observations mandatorily having the dual audit and observations voluntarily choosing the dual audit.

15 In particular, the 67 observations becoming single-audited post-2010 comprise 1 observation in 2010, 11 observations in 2011, 12 observations in 2012, 13 observations in 2013, 14 observations in 2014, and 16 observations in 2015. These observations are deleted in the tests of the Hypothesis. However, in addition to the main tests, we compare these observations with those voluntarily choosing the dual audit to further address the alternative explanation regarding signalling. The results are shown in the Appendix and discussed in Section 6.4.1.

16 Among these methods, the most commonly used models are the GLS model (Gebhardt et al. 2001), the PEG model (Easton 2004), the CT model (Claus and Thomas 2001), and the OJ model (Ohlson and Juettner-Nauroth 2005). We conduct a series of validation tests among the cost-of-equity estimates derived from these models, combining the methods of Botosan and Plumlee (2005) with those of Mao et al. (2012). We also make references to prior literature (e.g. Lu and Ye 2004, Chen et al. 2011) in choosing the GLS model that performs better in the Chinese capital market than the measures derived from the other three models.

17 We exclude the loss firm-years when calculating the industry median ROE over the prior 10 years for each industry. As suggested by Lu and Ye (2004) and Zeng and Lu (2006), in the long run, the loss firms would quit an industry eventually. Therefore, they suggest that ROE of the profitable firm-years can better reflect the long-term equilibrium return of an industry. We follow this view, and exclude the loss firm-years in the calculation of the industry median ROE.

18 We also produce descriptive statistics of variables used in other tests, and the discussion of these variables will be provided in later respective sections.

19 Owing to limited time-series (i.e. 2010–2015 (six-year)) data in this analysis, we do not include firm fixed effects. Instead, we include industry fixed effects in the model to control for other factors that may affect the cost of equity. We caveat that with industry fixed effects included, we cannot rule out the potential endogeneity brought by unobserved firm characteristics.

20 We acknowledge that given the small sample size, the insignificant results could also be due to a lack of statistical power issue. We thank the Editor’s comment for this potential research limitation.

21 Kausar et al. (2016) document that the role of a mechanism (specifically, audit choice in their study) in reducing financing frictions is of a higher value for firms with stronger financial constraints.

22 SA Index is calculated as: – 0.737 × firm size + 0.043 × firm size²– 0.04 × firm age. This index is commonly used to measure financial constraints in Chinese studies (e.g. Ju et al. 2013, Dou et al. 2014, Lu and Chen 2017).

23 Similar to the calculation of the ITCV, the impact of each control variable is the product of the partial correlation between COE_t,i and the control variable and the partial correlation between Mandatory_t,i and the control variable.

24 Only 1 AH company (i.e. Tsingtao Brewery Co., Ltd.) opted out of the dual audit in 2010 (http://finance.sina.com.cn/stock/hkstock/hkstocknews/20110418/08099706108.shtml).

Additional information

Funding

This work was supported by the [National Natural Science Foundation of China] under Grant [number 71672141; 71502134; 71502139; 71602160].