PREFACE

China abolished the mandatory dual audit system for mainland Chinese companies cross-listed in Hong Kong (AH companies) in 2010. At the end of 2018, over 62% of AH companies continued to use the dual audit system. Arguably, the dual audit system contributes to ensuring auditor independence and improving corporate governance. It also helps provide investors accurate and complete financial information and facilitates mainland local audit firms to accumulate international business experience. The impacts of the deregulation of the mandatory dual audit system greatly concern investors (Sun, 2014) ^[1] . The dual audit system and its change have drawn persistent attention from researchers. A comprehensive investigation on the consequences of the deregulation of the mandatory dual audit system can enhance understanding of the influences of the dual audit system on financial reporting quality, audit quality and investor protection. Such an investigation can also provide important implications for future audit regime arrangements. However, existing studies are limited to investigating the audit-related consequences of this deregulation, with little attention paid to its economic consequences on the capital market. Moreover, studies focusing on the audit-related consequences have not deeply explored, through the empirical process, about a mechanism whose loss is likely to be the main driver of the decreased audit quality following companies' cancelation of the dual audit system. Given these limitations of previous literature, this study investigates the consequences of the deregulation of the mandatory dual audit system on the capital market and audit market.

The main conclusions and innovative work of this study are embodied in three aspects.

First, this study theoretically elaborates and empirically verifies that mandatory dual audit and voluntary dual audit can reduce the cost of equity to different extents, providing empirical evidence for the economic consequences of the deregulation of the mandatory dual audit system on the capital market. Previous studies focus on the audit-related consequences of the deregulation. These studies only compare companies that have the dual audit with those that do not or companies in the regulated period with those in the period after the deregulation. To the contrary, this study classifies dual audits into mandatory dual audit (before the deregulation) and voluntary dual audit (after the deregulation) and analyzes how both separately affect the cost of equity. The theory of commitment effect of complying with mandatory requirements and the theory of signaling effect of making voluntary choices are used as bases. This study derives the cost-of-equity measure from the GLS model (Gebhardt et al., 2001) ^[2] and effectively controls for the effects of the audited financial information in the empirical process. This study finds that companies have a lower cost of equity when they mandatorily have the dual audit than when they voluntarily use the dual audit; companies voluntarily using the dual audit have a lower cost of equity than companies using only the single audit in the same period. This study also finds that the effects of mandatory dual audit and voluntary dual audit in reducing the cost of equity are greater in companies with higher demand for equity financing and in companies with higher agency costs; the effect of mandatory (voluntary) dual audit in reducing the cost of equity is irreplaceable (substitutable) by alternative voluntary audit mechanisms. The theoretical elaborations and relevant results suggest that the deregulation of the mandatory dual audit system affects investors' perceptions and decisions. Such effects yield significant economic consequences on the capital market. These results not only enrich the research on the dual audit system and that on the cost of equity, but also enhance understanding of the differential albeit important roles that mandatory and voluntary dual audits separately play in the cost of equity. Furthermore, these results serve as new references for regulators and policy-makers to learn about the influences of the mandatory dual audit system on investor protection. To a certain extent, these results imply why a majority of AH companies persist using the dual audit system even years after the abolition of the mandatory requirement. The results warn regulators to cautiously reassess the costs and consequences of the deregulation of the mandatory dual audit system.

Second, this study verifies that the loss of the positive spillover effect from Hong Kong-based auditors is the main driver of the declining audit quality of companies canceling the dual audit. Although previous studies discuss the audit-related consequences of the deregulation of the mandatory dual audit system, they lack thorough analysis and verification of a mechanism that possibly drives the audit quality change under this policy change. This study measures audit quality by abnormal accruals, quantifies the spillover effect, and uses the analyses of changes. This study finds that companies keeping the dual audit system have higher audit quality when they are subject to a more positive spillover effect from Hong Kong-based auditors. This study also finds that after companies cancel the dual audit system, there is a positive association between the loss of the positive spillover effect from Hong Kong-based auditors and the decrease in audit quality. These findings provide initial evidence confirming that the positive spillover effect from Hong Kong-based auditors helps ensure the higher audit quality of companies keeping the dual audit, whereas the loss of this effect is primarily responsible for the decline in the audit quality of companies canceling the dual audit. Compared with previous studies, the analyses of this study effectively rule out the potential impacts of cross-listing, accounting standards convergence, auditors' shared reputational concern, and unobserved confounding factors. These analyses help to reach rigorous and robust conclusions. Extending the research on the dual audit system and that on audit quality, the results of this study contribute to further understanding of the indispensable role played by the positive spillover effect from Hong Kong-based auditors in the dual audit regime. Furthermore, these results confirm practitioners' concern on the likely damage to audit quality when companies cancel the dual audit and accept the single audits performed by mainland-based auditors. The results caution regulators and mainland local audit firms that becoming truly big and strong is a long and arduous task for local audit firms.

Third, this study develops a measurement method to directly and effectively capture the spillover effect from Hong Kong-based auditors in the dual audit process. Owing to its unobservability, this effect is limited to a description by previous studies, which restricts researchers to deeply exploring the audit-quality issues in the dual audit setting from this perspective. According to relevant rules and previous studies, this study assumes that the stricter H-share audits of a company are likely to have a positive spillover effect on the company's A-share audits by helping prevent earnings management in its A-share financial statements. Accordingly, a closer gap (shorter distance) between a company's H-share and A-share audit quality represents a larger constraint effect of the H-share audits performed by its Hong Kong-based auditor on the A-share audits performed by its mainland-based auditor, that is, a more effective positive spillover effect from the Hong Kong-based auditor. Following this rationale and from the perspective of earnings management, this study creates a proxy for the spillover effect based on the calculation of the distance between the H-share's and A-share's absolute abnormal accruals of a company keeping the dual audit. Compared with previous studies, this study intuitively presents the spillover effect from Hong Kong-to mainland-based auditors through the dual audit process by constructing a novel measure. This study takes an initial step toward relevant empirical work. It promotes the development of research on the dual audit system and that on audit quality.