CFO’s Working as The Board Secretary Concurrently and Corporate Disclosure Quality: Based on Empirical Evidence of Listed Companies in Shenzhen Stock Exchange

This paper took the selected data listed companies in Shenzhen Stock Exchange in 2008-2015 as samples to study the relationship between the CFO’s working as the Board Secretary concurrently and corporate disclosure quality, and also to examine the impact of different government intervention levels and nature of property rights. The results indicate that the CFO’s doubling as the Board Secretary can distinctly improve the quality of corporate disclosure in listed companies; the CFO’s holding concurrently the post of the Board Secretary can improve noticeably the disclosure quality of listed companies in regions with a high degree of government intervention; the CFO’s also serving as the Board Secretary can improve the disclosure quality of non-state-owned listed companies. Moreover, this paper presents a reasonable explanation for the phenomenon that increasingly more CFOs are serving as the Board Secretaries simultaneously via empirical study. Lastly, conclusions of this study can provide empirical evidence for the appointment of the Board Secretary in listed companies.


I. Introduction
Corporate disclosure can reflect the financial and operational status, and other key information of companies. Playing an important role in investors' decision-making, it constitutes a critical part of the securities market that connects the listed companies and the securities company. As the bridge between the two, information disclosure can affect the resource allocation of the capital market. Therefore, reliable information disclosure can help investors make correct investment decisions, and reflect changes and relevant characteristics of the securities market on a real-time basis. Twenty years ago, Chinese listed companies started introducing the corporate disclosure system. Compared with that of listed companies in the developed areas of Europe and North America, Chinese counterparts still has a long way to go. At present, problems, such as low authenticity, accuracy, and completeness of corporate disclosure, are mainly responsible for internal information asymmetry and irrational resource allocation of listed companies on the capital market. The board secretary system originated from Britain, has been quite mature, and played an indispensable role in corporate governance of European and North American countries.
The system of the Secretary of the Board of Directors was introduced into China in the early 1990s. Yet the Secretary of the Board has not drawn much attention to the degree that some companies have not even established such a position. In this

CFO's Concurrently Serving as the Board Secretary and Corporate Disclosure Quality
From the perspective of risk avoidance, the Secretary will expose himself\herself to public condemnation and legal punishment when he\she fails to disclose information according to regulations, deliberately conceals negative news, or carelessly answers questions from shareholders or other relevant persons. Consequently, the risk-averse Secretary will disclose information more responsible manner, thereby improving the disclosure quality. In comparison with other Secretaries, the CFO who also serves as the Secretary generally has a stronger risk aversion motivation. First, as the company's financial staff, the CFO has a stronger sense of risk avoidance. Studies by (Devine, 1963;Staubus, 1985;Li & Lu, 2003) show that the company's financial reports demonstrate accounting robustness. Basu (1997) stresses that good news needs more evidence to confirm, compared with the confirmation of bad news in financial reports. Ball et al. (2000); Ball (2001) point out that one of the important features of financial reporting is the timely confirmation of bad news related to robustness or economic loss. Second, compared with other executives, the CFO faces more serve public condemnation and punishment. The CFO understands the company's financial and operating conditions better than other executives. Correspondingly, once negative news comes out of the company, the CFO is more liable to be suspected of violating regulations and thus subject to more serious public reproach and punishment.
From the perspective of information communication, the CFO who also serves as the Board Secretary can improve the quality and efficiency of information communication. First, the CFO has a fairly deep understanding of the company's business and financial situation. The core position of financial information cannot be shaken even though the capital market pays more and more attention to non-financial information. Graham et al. (2005) argue that since the core indicator that external investors are most concerned about is the company's accounting earnings, it is particularly important to achieve the target earnings of external investors. Yan & Qingquan (2004) also underscore that the company's accounting information is an essential indicator for institutional investors. Therefore, the Secretary needs to devote more attention to financial accounting information when disclosing company information.
Second, the CFO's concurrent role as the Board Secretary can simplify internal communication procedures. The communication between the Secretary and the CFO is crucially important. Any information errors in the communication between these two parts will directly impact corporate disclosure quality. However, when the CFO also holds the post of the Secretary, there is no need for the aforesaid communication and the potential information errors are avoided. The CFO and Secretary thus can disseminate company information more accurately and increase the professionalism, understandability, and completeness of the disclosure, therefore improving the efficiency of information communication and the disclosure quality. Based on risk avoidance and information communication effect, this paper believes that the CFO's doubling as the Secretary will help improve the disclosure quality. Hypothesis 1 is thus proposed: 3.2 The CFO Simultaneously Serving as the Board Secretary, Government Intervention, and Corporate Disclosure Quality. The research of Fan et al. (2017) shows that the degree of government intervention varies in different regions of China due to differences in resources, geographic locations, national policies, and traditional cultures. Most scholars have found that government intervention often gives rise to low operation efficiency and defects of governance structure in companies. Claessens et al. (2000) observe that in the case of the weak legal system, weakened capital market restraint mechanism and backward institutional environment, government intervention to some point negatively impacts the management efficiency, which leads to the lower level of information disclosure. The experiment of Bushman et al. (2004) demonstrates that when there are more direct government control and intervention of companies and banks and more hidden intervention from government officials, the country often displays poor financial transparency. On the other hand, Gao & Song (2007) note that the lower the degree of government intervention, the higher the company's transparency. Wang et al. (2007) also find that corporate disclosure quality is closely related to the governance environment: the lower the degree of government intervention in the area where the listed company is located, the higher the quality of corporate disclosure. It can be seen that more government intervention leads to lower disclosure quality whilst less government intervention attributes to higher disclosure quality. Meanwhile, the above analysis shows that the risk aversion and good information communication displayed by the CFO who doubles as the Board Secretary are conducive to improving the quality of corporate disclosure. Therefore, concerning controlling the risk related to corporate disclosure, investor protection is better in regions with a low level of government intervention and the company has a higher disclosure quality. In this case, the CFO's concurrent role as the Board Secretary has no noticeable influence on the improvement of the disclosure quality. On the other hand, for listed companies with severe government intervention, there is an urgent demand for the CFO's serving as the Secretary simultaneously. The CFO will proactively provide accurate and complete accounting information, thereby promoting corporate disclosure quality. Based on this, hypothesis 2 is proposed: H2. Compared with regions with a low level of government intervention, there is a more obvious positive correlation between the CFO's doubling as the Board Secretary and corporate disclosure quality in regions with a high level of government intervention.

CFO Concurrently
Holding the Post of the Board Secretary, the Nature of Property Rights, and Corporate Disclosure Quality.
In China, the nature of property rights will affect corporate disclosure quality. Due to their property rights, state-owned enterprises have a natural political relationship with the government  and on this account will face more pressure from government regulation. They need to not only improve corporate performance but also deliver more social responsibilities (such as tax, employment, stability maintenance, and disaster relief). On top of that, state-owned enterprises have to set an example in building a good internal controlling system and actively disclosing information. Studies by Xingqiang and Riguang (2007) suggest that since state-owned listed companies assume more responsibilities of building their public images and improving credibility, they have less motivation for earnings management and the information disclosure is better regulated. Eng and Mak (2003) contend that when the state's shareholding ratio exceeds 25% of the listed company, the accounting information quality will be relatively reliable. Chen et al. (2013) find that the disclosure quality of accounting information is distinctly higher when the state is the actual controller of the company.
Since state-owned enterprises have their own powerful political connection networks, they have less motivation to maintain corporate images via manipulating accounting information compared with non-state-owned enterprises. This factor may have a certain positive effect on the quality of accounting information. In contrast, born with a shortage of resources and government guarantee, non-state-owned enterprises have to bear their operation risks and face high financing pressure. Under this circumstance, they are more inclined to manipulate accounting information. Therefore, in non-state-owned enterprises, the Board Secretary needs to bear a higher risk for corporate disclosure and has a stronger motivation for reputation protection. In this case, if the CFO also serves as the Secretary, he\she will improve the professionalism, understandability, accuracy, and thus the whole quality of the disclosure. To sum up, this paper believes that compared with state-owned enterprises, the CFO's doubling as the Secretary has a more noticeable improving effect on corporate disclosure quality in non-state-owned enterprises. Hence comes hypothesis 3.

Sample Selection
This paper selected listed companies in Shenzhen Stock Exchange from 2008 to 2015 as the initial sample, and screen them as follows: (1) excluding companies in the financial industry; (2) excluding companies with missing financial data; (3) excluding companies that have variable data with extreme values. After the aforementioned screening, 8971 sample companies were finally determined. The relevant data of the Board Secretary in this paper were obtained by manually retrieving and sorting through the annual reports of these listed companies. The rating data of corporate disclosure quality in the Shenzhen Stock Exchange were derived from the website of the Shenzhen Stock Exchange (www.sse.org.cn). The level of government intervention was measured by the "Government-Market Relationship Score" elaborated in the NERI Index of Marketization of China's Provinces by Fan et al. (2017). Other financial data were obtained from the China Stock Market & Accounting Research Database (CSMAR).

Model Setting and Description of Variables
To examine the impact of the CFO's doubling as the Board Secretary on corporate disclosure quality, the following research model was constructed: CDQ denotes the explained variable, namely corporate disclosure quality of listed companies. Since the main research objects of this paper are Shenzhen-listed companies and drawing on the research method of Ying and Zhengfei (2006) the disclosure quality ratings from Shenzhen Stock Exchange were applied to measure the disclosure quality of each company. Number 4, 3, 2, and 1 were used to represent four quality levels of "excellent", "good", "pass", and "fail".
Drawing on the research of Wang and Fang (2018) and to avoid the interference of multicollinearity on the regression results, this paper did not use interactive variables to examine hypothesis 2 (H2) and hypothesis 3(H3). Instead, government intervention (GI) and the nature of property rights (SOE) were introduced as grouping variables to divide samples into four groups, and regression analysis was conducted separately on each group. It needs to be clarified that index of 2014 was  Table 1.

CDQ
Disclosure quality ratings from the Shenzhen Stock Exchange were used to measure the disclosure quality of each company. Number 4, 3, 2, and 1 were used to represent four quality levels of "excellent", "good", "pass" and "fail".

Explanatory Variable
The   Table 3 shows the disclosure quality ratings of sample companies from 2008 to 2015. Except for 2008, more than 60% of the companies ranked "good" in the disclosure quality rating, with the highest proportion reaching 70.58% in 2012. This indicates that the disclosure quality of listed companies is mostly in a good state. Companies that were rated "excellent" and "good" exceeded 70% of the total whilst those assessed as "fail" had a rather low proportion-the highest being 2.65% in 2008. This fact demonstrates that the disclosure quality of listed companies is relatively high generally. Also, the ratio of companies ranking "excellent" seemed to be basically on an upward trajectory. Though slightly down from 2011, this ratio is 2012 far surpassed that of 2008 to 2010 and continued to increase in 2013 and 2014 with a tiny decline in 2015. From 2008 to 2012, the proportion of companies rated as "good" increased year by year, but then dipped in both 2013 and 2014.  Table 4 represents the descriptive statistics of the observed data of the whole sample. The mean value of the disclosure quality (CDQ) stood at 2.986, with the median value being 3, indicating that the assessment of corporate disclosure quality mainly concentrated in the scored segment of 2-3. Among all these sample companies, 9.2% of them had the CFO double as the Board Secretary. On average, 55.8% of the listed companies were located in regions with a relatively high degree of government intervention, and 49.3% of the listed companies were state-owned enterprises. There were 29.8% of companies had the same person held the position of President and General Manager (DUAL). The average proportion of independent Board Members (DIRIND) in the Board was 36.5%, which met the requirement of regulatory agencies for the proportion to be no less than 1/3. Only 3.1% of the sample companies selected the "Big Four" accounting firms (BIG4). The minimum and maximum sizes of listed companies (SIZE) were 18.043 and 25.198 respectively, with the mean value close to the median. The return on assets (ROA) averaged 6.1%. The average ratio of financial leverage (LEV) was 41.6%. The average shareholding ratio of the largest shareholder (CONSHARE) stood at 34.5%, with more than half of the largest shareholder held more than 32.1% of the company's shares. The standard deviations of all variables were below 1.2, suggesting that these variables had relatively stable values.  Table 5 shows the Pearson correlation coefficients among variables. It can be seen that the CFO's doubling as the Board Secretary (CFO) is significantly positively correlated with corporate disclosure quality (CDQ) at a 1% level, with the correlation coefficient being 0.037. This indicates that the CFO's also serving as the Secretary is an essential factor affecting corporate disclosure quality and those listed companies with the CFO holding the post of the Secretary simultaneously will have higher disclosure quality. This result preliminarily verifies H1. The degree of government intervention (GI) and corporate disclosure quality (CDQ) has a significant negative association at a 1% level, which indicates that the more serious the government intervention, the lower disclosure quality. As the correlation coefficients among the main variables do not exceed 0.4, there is no serious multicollinearity.  Note: ***, ** and * denote a significant correlation at levels of 1%, 5% and 10% respectively.  Table 6 shows the regression results of the model (1). In column (1), the influences of industry and year were controlled. There is a significant correlation between the CFO's concurrently serving as the Secretary (CFO) and corporate disclosure quality (CDQ) at a 1% level, with the correlation coefficient being 0.060. When influences of all the variables were controlled in column (2), the CFO's doubling as the Secretary (CFO) shows a significant association with corporate disclosure quality (CDQ) at a 5% level with a correlation coefficient of 0.044. These facts indicate that the CFO's also holding the post of the Board Secretary is significantly positively correlated with corporate disclosure quality, namely, the CFO's also serving as the Secretary can attribute to the improvement of disclosure quality in listed companies. Therefore, hypothesis 1 is verified. Among control variables, there is a significant negative association between the General Manager's also serving as the President (DUAL) and corporate disclosure quality (CDQ). This suggests that the integration of these two positions will undermine the disclosure quality. The independence of the Board (DIRIND) shows a significant negative correlation with the disclosure quality (CDQ), meaning that independent Board Members have a malign influence on the disclosure quality. This result is consistent with the conclusion of Zhou & Zhuang (2016). The justification of such a phenomenon may be that independent Board Members are in a disadvantaged position regarding information access in comparison with internal Board Members. Selecting the "Big Four" accounting firms (BIG4) is significantly positively related to the disclosure quality (CDQ), indicating that high-quality external audits can improve the disclosure quality. The company size (SIZE) and return on assets (ROA) both have a significant positive association with the disclosure quality (CDQ), which demonstrates that the larger the company size or the higher the return on assets is, the higher the disclosure quality will be. This is consistent with previous research conclusions. There is a significant negative correlation between financial leverage (LEV) and disclosure quality (CDQ), indicating that the company with a higher asset-debt ratio is liable to have lower disclosure quality. This may be justified by the fact that the creditor will bring more pressure on the company as the asset-debt ratio increases; and in this scenario, managers will be prone to disclose only positive information while hiding negative information. The shareholding ratio of the largest shareholder (CONSHARE) is significantly positively related to the disclosure quality (CDQ), showing that the proportion of the largest shareholder's shares can positively affect the disclosure quality. Note: ***, ** and * denote a significant correlation at levels of 1%, 5% and 10% respectively. Table 7 presents the regression results of the model (1) grouped by the level of government intervention. In column (1), the CFO's doubling as the Secretary (CFO) shows a significant correlation with corporate disclosure quality (CDQ) at the level of 1%, with the correlation coefficient being 0.087. This implies that in regions with a high level of government intervention, the CFO's holding the post of the Secretary simultaneously is conducive to improving the quality of disclosure and alleviating the negative impact of government intervention on the disclosure quality. In column (2), the CFO's doubling as the Secretary (CFO) and corporate disclosure quality (CDQ) have a correlation coefficient of 0.002, showing no significant association. This suggests that in regions with a low level of government intervention, the CFO's also serving as the Secretary does not exert a noticeable influence on the disclosure quality. The reason may be that in regions with weak government intervention, there are relatively high degrees of marketization and legalization, a relatively sound protection mechanism for stakeholders and thus an already relatively high disclosure quality. In this circumstance, the CFO's doubling as the Secretary will not manifest an obvious positive influence on disclosure quality. To sum up, the CFO's holding the position of the Secretary concurrently can noticeably attribute to the disclosure quality only in regions with high levels of government intervention. Therefore, hypothesis 2 is corroborated. Note: ***, ** and * denote a significant correlation at levels of 1%, 5% and 10% respectively. Table 8 shows the regression results of the model (1) grouped by the nature of property rights. In column (1), there is not a significant association between the CFO's doubling as the Board Secretary (CFO) and corporate disclosure quality (CDQ) in state-owned enterprises. In column (2), however, the CFO's also serving as the Secretary (CFO) is significantly positively related to the disclosure quality (CDQ) at a 10% level in non-state-owned enterprises. These results indicate that the CFO's holding the post of the Secretary simultaneously exerts no obvious impact on the disclosure quality in state-owned enterprises but can noticeably improve the disclosure quality in non-state-owned enterprises. This may be due to the large differences in qualifications, appointment and removal procedures, responsibilities and authority of the CFO, and differences in political risks and severity of reputational damage in state-owned and non-state-owned enterprises. Based on the aforementioned analysis, hypothesis 3 is confirmed. Note: ***, ** and * denote a significant correlation at levels of 1%, 5% and 10% respectively.

Changing the Measurement of Dependent Variables
To enhance the reliability of the regression results, this paper changed the classification method of corporate disclosure quality ratings issued by the Shenzhen Stock Exchange and divided them into two categories. Companies rated as "excellent" and "good" took the value 1 and those ranked as "pass" and "fail" took the value 0. This explained variable was denoted by DISQ, and the remaining variables were the same as above. A logistic regression model was employed as follows: = 0 + 1 + 2 + 3 + 4 + 5 4 + 6 + 7 + 8 + 9 + ∑ + ∑ + After changing the measuring method of the dependent variables, empirical tests were conducted again on the research hypotheses. The regression results are shown in Table 9, Table 10, and Table 11. It can be seen that the empirical results are consistent with the above-mentioned results. Therefore, hypothesis 1, hypothesis 2, and hypothesis 3 are further verified by empirical evidence. Note: ***, ** and * denote a significant correlation at levels of 1%, 5% and 10% respectively. 3968 Note: ***, ** and * denote a significant correlation at levels of 1%, 5% and 10% respectively. Note: ***, ** and * denote a significant correlation at levels of 1%, 5% and 10% respectively.

Further Controlling Other Variables
To further incorporate other variables, this paper referred to the research method of Zhou et al. (2011) to add two control variables-age (AGE) and education level (EDU). Five education levels in CFOs concurrently hold the posts of the Board Secretary: high school, junior college, bachelor, master, and doctor, which were represented by the numbers 1, 2, 3, 4, and 5 respectively. After adding the two control variables of age (AGE) and education level (EDU) to the model (1), the hypotheses were tested again. The regression results are shown in Table 12, Table 13, and Table 14 and are consistent with the previous results. In this sense, hypothesis 1, hypothesis 2, and hypothesis 3 is further corroborated by empirical evidence.  Note: ***, ** and * denote a significant correlation at levels of 1%, 5% and 10% respectively. Note: ***, ** and * denote a significant correlation at levels of 1%, 5% and 10% respectively.

Reclassification of Government Intervention
The mean value of the "Government-Market Relationship Score" was used to divide sample companies into two groups: those with a high level of government intervention and those with a low level of government intervention. Then these two groups were tested in the model. The regression results are shown in Table 15 and are consistent with the previous results. Therefore, hypothesis 2 is further supported by empirical evidence. Note: ***, ** and * denote a significant correlation at levels of 1%, 5% and 10% respectively.

Solving the Problem of Endogeneity
Propensity score matching was used on sample companies since there may be relatively big differences in their initial conditions (such as industry, size, and nature of property rights) and thus the impact of endogeneity should be avoided. This paper adopted the CFO's doubling as the Board Secretary as the grouping variable. "1" = treatment group; "0" = control group. The stepwise regression was employed on the independent variables to identify the statistically significant ones. Finally, the company size (SIZE), asset-debt ratio (LEV), industry dummy variable (IND), and year dummy variable (YEAR) were determined as covariates for propensity score matching. The logit model was used for fitting. Due to the limited sample size of the control group, it was only suitable for 1: 1 matching. The maximum distance allowed between the treatment and control groups is 0.05. Table 16 represents the results of the propensity score matching. The balance test results in Table 17 show that from U to M, the p-value of the industry dummy variable (IND) remained insignificant and p-values of all the remaining covariates changed from significant to insignificant, passing the balance test. These variables can be considered as balanced after matching between the treatment group and the control group. At the same time, the density profiles before and after the matching in Fig. 1 and Fig. 2 show that the matching effect is satisfying and the common support assumption is satisfied.  Table 18, Table 19, and Table 20 represent the results of regression analysis on matched samples. It can be seen that the empirical results are consistent with the previous ones. Therefore, after solving the problem of endogeneity, hypothesis 1, hypothesis 2, and hypothesis 3 are still supported by empirical evidence. 1642 Note: ***, ** and * denote a significant correlation at levels of 1%, 5% and 10% respectively. Note: ***, ** and * denote a significant correlation at levels of 1%, 5% and 10% respectively. Note: ***, ** and * denote a significant correlation at levels of 1%, 5% and 10% respectively.

Conclusion
This paper took the selected data of listed companies in Shenzhen Stock Exchange in 2008-2015 as samples to study the relationship between the CFO's working as the Board Secretary concurrently and corporate disclosure quality, and also to examine the impact of different government intervention levels and nature of property rights. The results indicate that the CFO's doubling as the Board Secretary can distinctly improve the quality of corporate disclosure in listed companies; the CFO's holding concurrently the post of the Board Secretary can improve noticeably the disclosure quality of listed companies in regions with a high degree of government intervention; the CFO's also serving as the Board Secretary can improve the disclosure quality of nonstate-owned listed companies. The research in this paper has the following implications: first, the CFO of listed companies concurrently serves as the Board Secretary, which is conducive to improving the quality of corporate disclosure and improving the corporate governance efficiency. Second, state-owned enterprises must first solve the problems in the governance structure, otherwise, the role of CFO's doubling as the Board Secretary in corporate disclosure quality will not play. Third, the government should appropriately reduce the degree of interference in corporate disclosure, improve the corporate disclosure system, and strengthen the supervision of corporate disclosure for giving full play to the role of CFO's doubling as the Board Secretary in promoting corporate disclosure quality.