The economic effects of mandated ICFR disclosure in China: an insider perspective

ABSTRACT

Using a survey of corporate insiders from 1,148 Chinese listed firms, this paper examines the direct effects of compliance with mandated Internal Control over Financial Reporting (ICFR) disclosure. In contrast with their U.S. counterparts, vast majorities of respondents recognize compliance benefits and perceive these benefits to outweigh the costs. However, the degree to which the firm is susceptible to market forces is a major determinant of the effects of compliance. The paper shows the evidence which supports the idea that the effectiveness of mandated ICFR disclosure depends crucially on the broader market and institutional environment in which firms operate.

KEYWORDS:

• ICFR disclosure
• regulation
• cost-benefit
• marketization
• China

Disclosure statement

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

Notes

1. Several studies report strong relations between disclosure quality and market outcomes, such as cost of capital (Francis, Nanda, and Wang 2006).

2. Cross-country evidence indicates that disclosure and securities regulations are associated with lower cost of capital (e.g. Hail and Leuz 2006) and greater financial development and economic growth (e.g. La Porta et al. 1998).

3. For example, Canada, Japan, and France introduced rules closely resembling Section 404; Australia, Germany, and the United Kingdom introduced ‘comply or explain’ rules; while Mexico and Brazil maintained a regime of voluntary compliance.

4. Using the average exchange rate of 614.28 RMB per 100 U.S. dollar in 2014, this figure corresponds to a mean total annual cost of compliance with EICS of $1.047 million (raw value, untabulated). Alexander et al. (2013) report a mean total annual cost of compliance with Section 404 of $1.207 million. Internal control information technology expenses account for over 70% of EICS compliance cost (raw value).

5. In 2012, the CSRC tasked the China Association for Public Companies, Shanghai Stock Exchange, Shenzhen Stock Exchange, Shenzhen DIB Enterprise Risk Management Technology Co. Ltd (hereafter DIB), and the Internal Control Research Center for Enterprises and Nonprofit Organizations of Sun Yat-sen University to design and administer the survey.

6. To be consistent with the regression analysis sample, we restrict our descriptive statistics sample consisting of A-share listed firms excluding ST or *ST firms (21 firms) and financial industry firms (37 firms) in Table 1 and Table 2.

7. For example, if the respondent holds the position of the Chairman and CEO (Internal Auditor and IC Director), we retain the Chairman’s (Internal Auditor’s) responses for the overlapping questions.

8. There is a concern that the respondents may consider that any negative response or comment would be sanctioned by the government afterwards, so vast majority of the respondents ascribe benefits to ICFR disclosure. However, the purpose of the survey was to know the listed firms’ attitudes towards EICS and their difficulties when implementing EICS, and the CSRC would further adjust EICS implementation schedule based on these survey data. Hence, the survey is a good lobbying channel to express their real thoughts, and the respondents may not overstate the benefit of EICS since the mandatory regulatory requirements would burden their costs.

9. To estimate the total cost of compliance, we require non-missing values for all three categories. Given that EICS-related audit fees were not required to be disclosed, we proceed as follows when the data is missing in the DIB database: (1) if the firm disclosed EICS-related audit fees in at least one year, we use the mean of disclosed fees in place of missing values; (2) if the firm never disclosed EICS-related audit fees, missing values are set to the predicted values from the regression of observed costs on total assets, their square, and the years of compliance experience.

10. The empirical p values of Firmage, MVE, Multisegment, and ICDcurrent in column (4) versus column (6) in Table 3 are 0.01, 0.02, 0.01 and 0.22 respectively. While the difference on ICDcurrent is insignificant, we only find some partial evidence that ICDcurrent is more pronounced in non-SOEs.

11. The empirical p values of Firmage, MVE, R&D, Crosslist, OwnBalance, and Excessret in column (2) versus column (3) in Table 4 are 0.01, 0.01, 0.10, 0.24, 0.01, and 0.07 respectively. While the differences on Crosslist is insignificant, we only find some partial evidence that Crosslist is more pronounced in SOEs.

12. The empirical p values of Multisegment, Dual, Restatement, and Excessret in column (4) versus column (6) in Table 5 are 0.01, 0.01, 0.69, and 0.01 respectively. While the differences on Restatement is insignificant, we only find some partial evidence that Restatement is more pronounced in SOEs.

13. The empirical p values of Benefits in column (6) versus column (7) in Tables 6 and 7 are 0.19 and 0.50 respectively. While the difference on Benefits is statistical insignificant, it is economically significant, and we only find some partial evidence that Benefits is more pronounced in non-SOEs.

Additional information

Funding

This work was supported by the National Social Science Foundation of China [21CGL013].