Prior evidence and hypotheses

Based on the theoretical framework primarily concerning information asymmetry, agency problem, signalling theory and political costs theory, many previous studies have attempted to formulate and test several hypotheses on the influencing factors of the company disclosure level. This section will review the findings of several prior researches as well as establishing hypotheses for the current study. 

(1). One of the most remarkable features of Chinese capital market is assumed to be its unique market structure, which comprises of three major segments: A shares which are only sold to domestic citizens in domestic currency; B shares which are only issued for foreign investors in foreign currency, but traded in domestic exchange markets; and H shares which are traded in SEHK in foreign currency. 

Given their different characteristics, such as listing market, listing requirements, accounting standards and reporting environment, their disclosure behaviours and disclosure policies are expected to vary systematically. Therefore, one of the basic intentions of the current study is to test whether companies, of which shares are belongs to the three different market segments, exhibit different disclosure patterns.

At a glance, foreign listing status is a major feature that distinguishes H shares-issuers from the other companies issuing only A or A+B shares. For this feature alone, compliance with Chinese GAAP and IFRS is mandatory for these Chinese firms that issue both A and H-shares. Accordingly, the IFRS-based annual report must be audited by an internationally recognized auditor; while the Chinese GAAP-based annual report may be audited by local accounting firms, and any difference in net incomes between these two sets of accounting information must be reconciled and presented in the financial statement. In that case, companies with A and H shares are subject to additional listing requirements as well as disclosure rules, consequently greater information disclosure can be expected from these companies than the other firms listed only in the domestic market.

Apart from regulatory requirement, H-share companies are also under greater market pressure to disclose more information. Assuming the primary objective for Chinese firms listing on international stock exchanges is to obtain capital at the lowest possible cost, they need to compete with the other SEHK-listed firms of which the westernized corporate governance systems are generally believed to be effective in terms of assuring a high quality financial report through proper internal control systems. Hard to deny that, comparing with other SEHK-listed firms, H-share PRC firms are commonly assumed to have significantly greater adverse selection and moral hazard problems due to their lack of prior trading history, the limited transparency of corporate governance and management control system, and foreign investors’ concern about the magnificent state ownership. 

Given these disadvantages raised from information asymmetry and the potential economic consequence of increased discount rate, H-share firms should have greater incentives to commit to more intensive information disclosure in order to reduce their agency cost. Previous empirical study by Ferguson, et. al(2002) found that companies with both A and H-share issuing disclose substantially more financial information than purely domestically listed PRC firms as well as other SEHK listed companies.

On the other hand, it is generally observed that companies only listed on the domestic exchanges (including companies with only A shares and companies with both A and B shares) tend to adopt a relatively more secret disclosure policy, which exhibit little voluntary disclosure if any information beyond the exchange requirements (Haw et al., 2000). Ferguson, et. al (2002) offer several explanations: Firstly, the concept of public information disclosure is relatively new to both the investors and corporate managers in PRC where the basic systematic accounting standard was first promulgated since 1992.

Given the less developed market-oriented accounting system and the weak disclosure culture, Tang (2000) points out that comparing to other exchange markets with mature accounting systems, accountability to outside investors is less concerned by most Chinese corporate managers; also, the majority individual investors are unfamiliar with the evaluation and use of financial statement disclosures. Secondly, Ferguson, et. al (2002) argue that because the current capital market is experiencing the transformation from the formerly state-controlled economy to the new market model, there remains some old concerns of investors emphasizing on the state plan.

That is, local investors still tend to focus on “inside” information such as anticipated actions by the controlling government entities rather than relying on public information like financial statement disclosure (DeFond et al., 1999). Hence, corporations’ incentives and investors’ desire for information disclosure appear to be less strong for companies only listed on domestic market than firms listed on foreign exchanges; consequently, greater extent of information disclosure is expected for companies issuing both A and H shares than firms listed only on domestic market. The resulting hypothesises are as following:

H1: Companies with both H shares and A shares tends to disclose more information than companies issuing only A shares;

H2: Companies with both H shares and A shares tends to disclose more information than firms issuing both A and B shares.

The most important difference between A share and B share is that A-shares can only be owned and traded by Chinese citizens in Chinese currency, while B-shares can only be owned and traded by foreign investors in either Hong Kong currency or US dollars. Accordingly, the accounting regulations applicable to firms issuing only A shares is Chinese GAAP; while, for companies issuing both A and B shares are required to apply with Chinese GAAP as well as IFRS. The IFRS-based annual report must be audited by an internationally recognized auditor, while the Chinese GAAP-based annual report may be audited by local accounting firms, and the discrepancy between the two sets of audited financial reports needs to be reconciled with the IFRS and displayed in the annual report for domestic investors.

Because of the different regulatory requirements, companies with both A-share and B-share issuing are expected to disclose more information than firms with only A shares. Since comparing with smaller CPA firms, which are assumed to be more sensitive to client demands due to the economic consequences associated with the loss of a client, larger and international well-known auditing firms have a greater incentive to maintain independence from clients' pressure for limited disclosure because of the economic consequences associated with potential damages to their reputation (Chow and Wong-Boren, 1986).

Therefore, larger CPA firms have a greater incentive to require adverse disclosures by the client, consequently increasing the level of information disclosure (Patteon and Zelenka, 1997). In contrast, accounting information audited by domestic auditing firms may be considered noisy because of sloppy information environment and inadequate regulation. (Fox, 1998; Rask, Chu, & Gottschang, 1998). Prior to 1996 no auditing standards existed with the exception of a few guidelines from the sponsoring governmental agencies; and until 1998 all domestic auditors were public employees, who tend to act as government agents and bore little responsibility for any improper behaviour due to the lack of litigation against them.

Thus, auditors usually were affiliated with their clients and lacked motivation to be independent from them, consequently information disclosure may be subject to management’s selective bias.

Despite of the recent institutional changes such as the reformation of the accounting-information system and the introduction of new auditing standards, which aims to impose stricter disciplinary rules, more intensive monitoring and sanctions, the effective implement of regulatory is still doubt by the market (Haw, 2008). Therefore, companies with only A shares are likely to make less information disclosure than companies with both A shares and B shares of which annual reports are influenced by internationally recognised auditing firms. The resulting hypothesises is:

H3: Companies with only A shares are likely to make less information disclosure than companies with both A shares and B shares.

(Ferguson, Lam and Lee, 2002)

2.4 Disclosure by PRC-listed Firms

PRC firms listed on the two domestic exchanges voluntarily disclose little, if any information beyond the exchange requirements (Haw et al., 2000). Explanations include (1) the lack of sophistication with respect to financial reporting on the part of both investors and corporate managers, and (2) investor reliance on “inside” rather than public information. Public financial statement disclosure is relatively new to the PRC. The first basic accounting standard was promulgated in 1992 and, of the 30 standards proposed in the intervening years, only eight have been adopted.

Thus, accountability to outside investors is new to most corporate managers, and most individual investors are unfamiliar with the evaluation and use of financial statement disclosures (Tang, 2000). Institutional investment in the PRC is in a fledgling state. Local investors are also likely to place greater weight on factors such as anticipated actions by the controlling government entities than on financial statement disclosures (DeFond et al., 1999). Thus, demand for, as well as supply of additional disclosures may be limited in the PRC domestic exchanges. The disclosure practices of PRC firms listed on international exchanges, in which they face sophisticated financial statement users with diminished access to inside information, have not been examined.

2.5 Hypotheses: Disclosure by H-Share Firms on the SEHK

Theory (Spence, 1973; Grossman, 1981) indicates that voluntary disclosure can be used to alleviate information asymmetry problems, including moral hazard and adverse selection. A rational strategy to avoid deep discounting of share prices is to disclose additional information to investors to signal firm value (Watts and Zimmerman, 1986). Compared to other SEHKlisted firms, H-Share firms are likely to present significantly greater adverse selection and moral hazard problems. In contrast to the westernized corporate governance systems in place in most SEHK-listed firms, many PRC SOEs still operate in a “vacuum” with respect to corporate governance and management control (World Bank, 1995).

For example, audit committees and shareholder litigation are nonexistent and independent; outside directors are not required (DeFond et al., 1999). Thus, in addition to H-Share firms’ lack of prior history, important investor concerns include management quality, the potential for asset stripping or misappropriation, de-capitalization through excessive wage increases, and the role of the government as a major shareholder (Chen and Firth, 1999). Therefore, ceteris paribus, H-Share firms face significantly greater incentives to voluntarily disclose additional information.

Proprietary costs, however, also affect disclosure (Verrecchia, 1983).

The benefits of voluntary disclosure must be weighed against the costs of providing information that may invite or assist competition or regulation. Compared to other SEHK-listed firms, H-Share firms also face significantly lower proprietary costs. Most operate in industries deemed by the PRC government to be of strategic importance and are hence shielded from international competition (Lin et al., 1998). Thus, additional disclosure by H-Share firms is also likely to be less costly. This potential for greater incentives and lower disclosure cost leads to our first hypothesis:

Hypothesis 1a: Voluntary disclosure by H-Share firms will be greater than that by other SEHK-listed firms.

Further, as the primary objective for PRC firms listing on international stock exchanges is to raise capital at the lowest possible cost, we expect that H-Share firms’ incentives will mainly affect disclosure of additional strategic and financial information. Such incentives will have little impact on the disclosure of additional non-financial, social accountability information.

Political costs are borne primarily in firms’ local operating environments and are driven by local norms. H-Share firms operate solely in the PRC and political costs within this environment are virtually non-existent. Thus, we expect that differences in disclosure will only be observed for financial and strategic rather than social accountability information:

(Sami and Zhou, 2004)

In the emerging A-share and B-share markets, however, the value relevance of accounting information has been questioned. Accounting information based on domestic standards may be considered noisy because of sloppy accounting, inadequate regulation, and crony capitalism (Fox, 1998; Rask, Chu, & Gottschang, 1998). Besides, accompanying the rapid development of securities markets are some inevitable problems such as lagging legislation issues and multiple regulatory authorities (Liu & Zhang, 1996). However, the institutional changes in emerging markets, including the reform of the accounting-information system, could increase market liquidity, reduce transaction cost, and improve pricing efficiency (Feldman & Kumar, 1995).

In this study, we directly investigate the relative value relevance of accounting information in the two segments to provide further evidence on the value-relevance issue in the emerging market. Our basic intention is to test whether the two market segments differently value the major accounting information disclosed by the same company.

Under these regulations, listed companies prepare their financial statements based on the Chinese GAAP, as well as the IAS if they also issue B-shares. They should have their annual reports audited by authorized CPAs and submit copies to government agencies, such as state-owned-asset management agencies, tax authorities, securities regulatory agencies, and banks. They are also required to have copies available for investors.

In addition, listed companies are required to publish their annual reports in at least one of the authorized securities’ publications before April 30th the following year.10 For companies with both A-shares and B-shares, the audited annual reports for B-share investors are published in Hong Kong on the same day as those for A-share investors in China. The reconciliation information on the two sets of accounting statements is released to only A share investors, but not to B-share investors.

When there is a discrepancy between the two sets of audited financial reports, companies issuing both A shares and B-shares need to reconcile their accounting statements with the IAS for domestic investors. Because the IAS is considered to be of higher quality than local GAAP, and international auditors such as Big Five (Big Four) firms are thought to provide higher quality audits than their Chinese counterparts (Chui & Kwok, 1998; DeFond et al., 2000; Lam & Jing, 2000), the accounting information in the B-share market should be more relevant to the pricing process, compared with its counterpart in the A-share market.

(Sami and Zhou, 2008)

To shed light on the economic consequences of the implementation of new auditing standards, we investigate the Chinese emerging market where a set of auditing standards was introduced in a situation where, previously, no auditing standards existed with the exception of a few guidelines from the sponsoring governmental agencies.

In addition, in the information environment of an emerging market such as China, where the accounting disclosure was criticized for its low quality and quantity, the economic consequences of increased accounting disclosures due to the implementation of a set of auditing standards should be significant.

Moreover, auditors played the role of government agents and bore little responsibility for any improper behavior (Xiang, 1998). Because it was common practice to have a company audited by an auditing firm affiliated with the same level of government, auditors bent the rules under pressure from local government officials and company managers to pursue their own interests (Xiang, 1998; Graham, 1996). Additionally, there was no litigation against auditors (Graham,

1996; DeFond et al., 2000; Gul et al., 2003). Thus, auditors usually were affiliated with their clients and lacked motivation to be independent from them. Therefore, Zhou (2007) concludes that the implementation of new auditing standards helps reduce information asymmetry in an emerging market.

(Peng, Tondkar, Smith and Harless, 2008)

Chinese capital market development and market segmentation

A-shares can only be owned and traded by Chinese citizens, while B-shares can only be owned and traded by foreign investors.

The accounting regulations applicable to a Chinese listed firm depend on the type of security issued, A- or B-shares or both. Firms that issue A-shares are required to comply with

Chinese GAAP, while firms that issue B-shares are required to comply with IFRS. Firms that issue both A- and B-shares are required to issue two sets of annual reports, one based on Chinese GAAP and the other based on IFRS. The IFRS-based annual report must be audited by an internationally recognized auditor, but not necessarily a Big 4 firm, while the Chinese GAAP-based annual report may be audited by local accounting firms.

Reports must be released to the public simultaneously and any difference in net incomes between Chinese GAAP and IFRS must be reconciled and presented in the financial statement footnotes. Fig. 1 and Table 1 depict the Chinese capital market segmentation and the evolution of accounting regulations for Chinese listed A-share firms as of December 31, 2005.

Compliance with Chinese GAAP and IFRS is mandatory for Chinese firms that issue both A and B-shares. However, Tay and Parker (1990) remark that “even where compliance with standards is legally required, companies may not comply if it is perceived that the consequences of non-compliance are not serious” (p. 75). Street and Gray (2001) and Xiao (1999) find evidence that Chinese listed firms' compliance with accounting regulations is high.

(Sami and Zhou, 2008)

We mention the stricter disciplinary rules, monitoring, and sanctions imposed by the Chinese Institute of Certified Public Accountants (CICPA) and the Chinese Securities Regulatory Commission (CSRC) to effectively enforce the new auditing standards. The discussant points out that less effective corporate governance systems (characterized by dominant state and legal-entity ownership) and relatively low litigation risk in China's markets (compared to those in the West) could provide opportunities for managers of listed Chinese firms to act in the best interests of the government and its representative organizations rather than report high-quality accounting information or seek quality auditing of their financial reports (Ball, Kothari, & Robin, 2000).

Thus, the discussant is concerned whether auditing standards could be effectively implemented. While we agree that the Chinese markets are emerging markets, where accounting disclosure tends to be low in quality and quantity, as we mention in our introduction,

(Haw, 2008)

Corporate governance systems are less effective in China's markets than those in the West. In the government-controlled economy of China, managers of listed state-owned enterprises (SOEs) are frequently appointed by the government, who is the controlling shareholder. Recent studies show that such ownership structures adversely affect the information environment of these firms, which results in a high level of information asymmetry and a low level of informativeness of accounting earnings (Fan and Wong, 2002; Haw, Hu, Hwang, & Wu, 2004).

The managers of listed Chinese firms, where state and legal-entity (mostly SOEs) ownership dominate, are strongly motivated to act in the best interests of the government and its representative organizations, and have less incentive to report high-quality accounting information or seek quality auditing for their financial reports (Ball, Kathari, & Robin, 2000). Until 1998, all domestic auditors were public employees, and there was little incentive for high-quality audits, while litigation for audit failure was infrequent.2 In such an environment, it is doubtful whether auditing standards could be effectively implemented.

(2) The following set of hypotheses is concerned with the determinants of the extent of company disclosure. By reviewing the results of prior theoretical and empirical researches as well as considering the special feathers of the Chinese market, and data availability, the current study selects 10 relevant independent variables to be included in our model, which were further organized into three (not strictly mutually exclusive) categories, following the structure by Lang and Lundholm (1993),: Structure-related variables, Performance-related variables, Market-related variables and Other Monitoring factors.

Structure-related variables

The structural variables generally refer to firm characteristics that are widely known and likely to remain relatively stable over time. Size, leverage, state ownership, and board composition are included in this category.

Size

Apparently, among other possible influencing factors firm size has been the most commonly suggested variable in the disclosure literature, assumed to be positively associated with the level of company disclosure. Given the existence of information asymmetry in the capital markets and the agency problem raised from the separation between ownership and control, agency theory suggest that information disclosure can be used as a mean to reduce agency costs (Chow and Wong-Boren, 1987). According to Jensen and Meckling (1976) and Leftwich, Watts and Zimmerman (1981), larger firms with more reliance on external funds potentially are more subject to complicated conflicts among their wider range of stakeholders, consequently increasing agency costs.

Also, larger firms are assumed as more sensitive to political costs (Watts and Zimmerman, 1986). Besides, it is noted by Lang and Lundholm (1993) and McKinnon and Dalimunthe (1993) that in order to enhance firm value, large firms tend to suffer from greater pressures from analysts to disclose more information than smaller firms; as reluctance to disclosure may be interpretated by investors as unfavourable news. In that sense, larger companies have greater needs to engage in more intensive information disclosure in respect to their higher agency costs and greater disclosure demand. On the other hand, comparing to smaller companies, Singhvi and Desai (1971) argue that due to the generally better-established internal reporting systems of larger firms, the marginal cost for additional information disclosure is lower for larger companies than smaller ones. 

Furthermore, larger firms are assumed to have less exposure to competitive disadvantage than smaller companies when disclosing detailed company information (Firth, 1979). Therefore, compared to small firms large firms should have additional incentives for information disclosures. This argument has been confirmed as the influence of size on disclosure has been successfully tested by studies in various countries: the US (Singhvi and Desai, 1971 and Buzby, 1975), the UK (Firth, 1979), Canada (Kahl and Belkaoui, 1981), Mexico (Chow and Wong-Boren, 1987), Nigeria (Wallace, 1988), Sweden (Cooke, 1989), Austria (Wagenhofer, 1990), Japan (Cooke, 1991), Spain (Garcia and Monterrey, 1992, and Inchausti, 1997), New Zealand (Hossain et al., 1995), Czech (Patton and Zalenka, 1997), and Greece (Leventis and Weetman, 2004). To summarize, based on all the rationales discussed by previous studies and their supporting evidences, the first hypotheses can be formulated as: 

H1: firms with larger size disclose information to a greater extent than do those with smaller size.

(Patteon and Zelenka, 1997)

Several theoretical propositions from the voluntary disclosure literature support the expectation of greater financial report disclosure by larger firms: (1) lower incremental cost of producing information for larger firms (Lang and Lundholm, 1993); (2) transactions cost hypothesis (King, et a/,,1990), which suggests that incentives for private infonnation acquisition are greater for larger firms; (3) legal costs hypothesis (Skinner, 1994), which notes that damages in securities litigation are greater for larger firms; and (4) reluctance of small firms to inform competitors (Raffottmier, 1995). Although we will not be able to determine which of the above explanations is the actual cause, we expect a positive relationship between firm size and extent of disclosure.

Generally, firms with more employees are more complex and create the possibility of substantial infonnation asymmetry between the firm and market panicipants. Thus, firms with a greater number of employees might be expected to have more extensive disclosures in their annual reports.

(Malone, Fries and Jones, 1988)

- Singhvi and Desai (1971) provided several reasons why the extent of financial disclosure is different for firms of different sizes. Singhvi and Desai offered three justifications for their reasoning. First, the cost of accumulating certain infonnation is greater for small firms than for large firms. This difference is attributable to the more extensive internal reporting systems already in place in larger firms. Second, larger firms have a greater need for disclosure because their securities are typically distributed via a more diverse network of exchanges.

Last, management of a smaller corporation is likely to believe more strongly than the management of a larger corporation that the full disclosure of infonnation could endanger its competitive position.' Foster (1986, 111) suggested three possible proxies for firm size: total assets, net sales, and capitalized value of the firm. Among these, perhaps the one least subject to market fluctuations in the oil and gas industry is total assets. Sales and capitalized value of the firm are subject to relatively extreme fluctuations due to the volatility of oil and gas prices.

Total assets, although not completely unaffected by this volatility, is less affected because of the broad capital asset base that already exists in each firm.

(Meek, Roberts and Gray, 1995)

-As noted by Foster [1986, p. 44], "the variable most consistently reported as significant in studies examining differences across firms in their disclosure policy is firm size." Generally, large firms disclose more information than small ones. Unfortunately, it is unclear what size proxies. Larger firms may have lower information production costs, or they may have lower costs of competitive disadvantage associated with their disclosures. Larger firms are also likely to be more complex and have a wider ownership base than smaller firms. Agency theory suggests that large firms have higher agency costs [Jensen and Meckling 1976; Leftwich, Watts and Zimmerman 1981]. Finally, larger firms are more sensitive to political costs [Watts and Zimmerman 1986]. All of these reasons indicate that large firms should have additional incentives for voluntary disclosures, compared to small firms. Size is positively associated with voluntary disclosure levels in all of the country studies noted above.

(Raffournier and Geneva, 1995)

There is a general agreement that a positive relationship between the size of a company and its extent of disclosure is to be expected. Several reasons have been advanced in support of this influence (Singhvi and Desai, 1971; Firth, 1979). First, disclosing detailed information is relatively less costly for large firms because they are assumed to produce this information already for internal purpose. Secondly, because their annual report is the main source of information for their competitors, smaller firms may be reluctant to make a fuller disclosure of their activities which might place them at a competitive disadvantage.

It can also be assumed that large firms which, according to Watts and Zimmerman (1978), are more sensitive to political costs, will disclose more in order to allay public criticism or government intervention in their affairs. The influence of size is well documented. All empirical studies on the content of annual reports found a positive relationship between the size of a company and its extent of disclosure. Salamon and Dhaliwal (1980) noted a similar association for segmental information and Cowen et al. (1987) for social responsibility disclosure.

(Inchausti, 1997)

It is hypothesized that the larger the firm, the more need for external funds. Therefore there will be more potential conflicts among owners, creditors and managers, and information disclosures may be used to decrease agency costs and to reduce information asymmetries between the company and the providers of funds, and potential providers of funds. Larger firms are also subject to more political costs, and disclosure may be used to reduce such costs. On the other hand proprietary costs are smaller the larger the firm, so there are less incentives to withhold information. The independent variables initially considered as measures of size are total assets and sales. However in order to avoid the problems caused by heterocedasticity natural logarithms of these variables (LASSETS and

LSALES) were calculated.'" The influence of size on disclosure has been successfully tested by studtes in various countries: the US (Cerf, 1961; Singhvi and Desai, 1971; Buzby, 1975; Salamon and Dhaliwal, 1980), the UK (Firth, 1979), Canada (Kahl and Belkaoui, 1981), Mexico (Chow and Wong-Boren, 1987), Nigeria (Wallace, 1988), Sweden (Cooke, 1989), Austria (Wagenhofer, 1990), and Spain (Garcfa and Monterrey, 1993; Wallace et al. 1994).

(Ferguson, Lam and Lee, 2002)

Research indicates that voluntary financial statement disclosure is influenced by other factors. Larger firms face higher agency costs (Leftwich et al., 1981), higher political costs (Jensen and Meckling, 1976), greater information demand from financial analysts (Lang and Lundholm, 1993), and lower information production costs (Firth, 1979; Leftwich et al., 1981). Consistent with these arguments, a positive relationship between firm size and voluntary disclosure has been found in studies of US (Firth, 1979), Swedish (Cooke, 1989), New Zealand (Hossain et al., 1995) and Japanese firms (Cooke, 1991), as well as for firms listed on multiple exchanges (Meek et al., 1995).

(Hossain, perera and Rahman, 1995)

A number of disclosure studies (e.g. Cooke, 1991, 1989) find that firm size is an important factor in explaining variability in the extent of corporate voluntary disclosure. In the agency theory literature. Chow and Wong-Boren (1987, p. 539) argue that potential benefits of voluntary disclosure are likely to increase with agency costs. Moreover, Jensen and Meckling (1976) contend that agency costs increase with the proportion of outside capital. The proportion of outside capital tends to be higher for the larger firms (Leftwich, Watts and Zimmerman, 1981).

Thus, agency theory predicts a positive association between firm size and the extent of corporate voluntary disclosure. It is also argued, in the literature, that fiirm size is a comprehensive variable which can proxy for several corporate characteristics, such as competitive advantage and information production costs (see Buzby, 1975; Firth, 1979; Leftwich et al., 1981; Ball and Foster, 1982). For example. Firth (1979, p. 274) suggests that '... collecting and disseminating information is a costly exercise and perhaps it is the larger firms who can best afford such expenses. Furthermore, smaller firms may feel that fuller disclosure of their activities will put them at a competitive disadvantage with other, larger, companies in their industry'.

Another explanation put forward in the literature for the existence of a positive association between size of the firm and the extent of voluntary disclosure is the demand for information by financial analysts. For instance, Lang and Lundholm (1993) and McKinnon and Dalimunthe (1993) point out that 'large firms tend to have more analyst foUowings than small firms and therefore may be subjected to greater demand for information'. They argue that large firms have incentives to voluntarily disclose more information than smaller firms in order to enhance firm value because non disclosure may be perceived by investors as bad news.

(Leventis and Weetman, 2004)

Size is the variable most frequently found in disclosure studies, since it has repeatedly been identified as a significant explanatory factor. However, it proxies many theories (Ball & Foster, 1982). As a result, interpretation becomes complex. Agency theory (Jensen & Meckling, 1976) associates agency costs with the separation of management from ownership, which is likely to be greater in larger companies. Large companies are more visible than smaller ones and, in turn, more exposed to political attention (higher taxes, price controls, social responsibility). Hence, large companies are more likely to adopt strategies to reduce those political costs (Hagerman & Zmijewski, 1979; Watts & Zimmerman, 1978). Size has been used as a proxy for political costs by several studies (e.g. Holthausen & Leftwich, 1983).

It is expected that lower information costs (Verrecchia, 2001) for large companies and their investors will lead to a positive correlation between size and disclosure. Furthermore, larger companies that disclose information tend to have a lower competitive disadvantage than smaller companies (Dye, 1985; Meek et al., 1995). Larger companies may have a more intense impact on society, they are more capital market oriented and they have greater analyst following, which makes them more open to providing information (McKinnon & Dalimunthe, 1993). Our expectation, based on prior studies, is for a positive relation between size and extent of disclosure.

H1: firms with greater total assets disclose financial information to a greater extent than do those with fewer total assets.

Market capitalization, unlike both sales and asset size, represents an externally (as opposed to internally) determined measure of a firm's importance as seen by the investing public. This measure therefore seems more objective to us than the other two. This is in contrast to the decision by Malone et al. (1993, p. 253) to reject market capitalization because of its "volatility." However, the inclusion in our study of asset size, sales and market capitalization as proxies for corporate size was motivated by their use in the previous studies cited immediately above, the desire to compare our results with the results from previous studies, and the established practice of using these constructs as bases for distinguishing among firms in the non-accounting literature. As later discussion indicates, there is the potential for collinearity among sales, asset size, and market capitalization.

Leverage

Jensen and Meckling (1976) assert that similar to the arguments concerning firm size with respect to disclosure extent, agency theory suggests that firms with higher leverage are associated with higher agency costs, which can be improved through greater extent of disclosure. Myers (1977) and Schipper (1981) explain that in order to prevent the potential wealth transfer from debtholders to shareholders and managers, bond covenants are provided as a mean of contracting and monitoring the behaviour of management, which increase the complicity of the firms’ financial structure, consequently leading to greater need for information disclosure. 

Moreover, viewing from the perspective of risk, Patteon and Zelenka (1997) see leverage as a proxy for the financial risk of the firm, which can influence firms’ willingness to disclose as well as outside parties’ demand for monitoring information (Lang and Lundholm, 1993). And they assume that firms with higher leverage is financially more risky, then greater extent of disclosure may be required to assure investors’ confidence, thereby reducing the cost of financing. In that sense, a positive relation between leverage and the extent of disclosure is expected by the agency theory.

Nevertheless, by applying the concept of signalling theory, Leventis and Weetman (2004) argue that firms with lower level of leverage may be encouraged to show investor their advantage of being less financially risky by disclosing more information. Given the opposite expectations of agency and signalling theory, it is not surprising that contradicted empirical evidence have been found by previous studies. For instance, Malone, Fries and Jones (1988), and Ferpuson, Lam and Lee (2002) found a significant positive association between leverage and disclosure level; while, in the studies of Chow and Wong-Boren (1987), Garcia and Monterney (1992), Raffournier and Genera (1995), Hossain, Perera and Raham (1995), Inchausti (1997), and Leventis and Weetman (2004), leverage were found statistically insignificant in terms of explaining disclosure level. In contrast, Meek, Roberts and Gray (1995) report a significant, negative relationship between leverage and voluntary disclosure for US, UK, and continental European multinationals. Due to the mixed theoretical predications and empirical evidence, the second hypothesis of the present study is conducted to test the association between leverage and disclosure, but without any specification of the sign.

H2: company disclosure level is associated with leverage.

(Patteon and Zelenka, 1997)

Risk may affect a firm's willingness to disclose as well as outside parties' demand for monitoring information (Chow and Wong-Boren, 1987; Lang and Lundholm, 1993). In this category we include two measures of risk: Percentage of Intangible Assets and the Leverage Ratio of the firm. The percentage of intangible assets is a proxy for the extent to which a firm's future performance depends on risky assets; the leverage ratio is a proxy for the financial risk of the firm. The higher the values on these independent variables, the higher the risk of the firm, and the greater the expected extent of disclosure.

(Malone, Fries and Jones, 1988)

The first question that became apparent was: Is the extent of financial disclosure a function of the targeted investor group? A high debt/equity ratio, for instance, may indicate a firm whose managers wish to provide financial disclosures directed at the needs of long-term creditors (and presumably shareholders as well). Conversely, a low debt/equity ratio may indicate a firm whose management's disclosures are targeted more toward shareholders than holders of the firm's debt securities. Myers (1977) described the conflict between bondholders and stockholders. He argued that holders of fixed claims anticipate that stockholders will attempt to expropriate those claims.

Schipper (1981) explained that because of this conflict, bond covenants are provided in debt instruments. Additional disclosure requirements may result from these conflicts. Jensen and Meckling (1976) asserted that more highly leveraged firms incur higher monitoring costs. If highly leveraged firms do, in fact, disclose financial information to a greater extent than those less leveraged, greater costs of disclosure would be the result. This would support the contention of Jensen and Meckling.

(Meek, Roberts and Gray, 1995)

Agency costs are higher for firms with proportionally more debt in their capital structures [Jensen and Meckling 1976; Smith and Warner 1979] since potential wealth transfers from debtholders to shareholders and managers increase with leverage. Thus, voluntary disclosures can be expected to increase with leverage. However, Chow and Wong-Boren [1987] find no association of this variable with voluntary disclosures for their sample of Mexican companies. One reason they suggest is that Mexican companies may use other mechanisms besides voluntary annual report disclosures for containing shareholder-debtholder-manager interest conflicts (p. 540)

(Raffournier and Geneva, 1995)

Financial disclosure in annual reports can also contribute to solve monitoring problems between stockholders and creditors. These problems are more likely to arise in firms making a large use of debt. A positive relation between leverage and the extent of disclosure can be expected, although studies which investigated this association did not support this hypothesis

(Chow and Wong-Boren, 1987; Garcia-Benau and Monterrey-Mayoral, 1992). This unexpected result may arise because leverage is a poor surrogate for external financing. Studies with more precise measures of external financing found results more consistent with the hypothesis. Choi (1973), in particular, noted that companies voluntarily increase the extent of disclosure prior to entering the European capital market. Similarly, Salamon and Dhaliwal (1980) showed that diversified firms obtaining long-term capital externally were more likely to disclose segmental financial data voluntarily.

(Inchausti, 1997)

Firms with a high rate of leverage disclose more information. The arguments that support this hypothesis are similar to those for the hypotheses relating to size and stock exchange listing. Information may be used to avoid agency costs and to reduce information asymmetries. It must be considered that leverage may also help to reduce agency costs in the relationship between owners and managers. Although the evidence has not validated this hypothesis when tested on other studies (Chow and Wong- Boren, 1987; Wagenhofer, 1990; Raffoumier, 1991; Garcia and Monterrey, 1992), it has nonetheless been considered in this study. The independent variable employed was the leverage ratio: total liabilities/equity (LEV).

(Ferguson, Lam and Lee, 2002)

Arguably, agency costs, and hence disclosure, will also be higher for firms with proportionately more debt because of the increased potential for wealth transfers from debtholders to shareholders and managers (Jensen and Meckling, 1976). Research on this relation has been mixed. Sengupta (1998) found that US firms with high analyst disclosure quality ratings enjoyed lower cost of issuing debt. Similarly, Hossain et al. (1995) found a marginally significant, and Bradbury (1992) a significant, positive relationship between leverage and disclosure for New Zealand firms.

Chow and Wong-Boren (1987), however, found no relationship between leverage and disclosure in their sample of Mexican firms, while Meek et al. (1995) report a significant, negative relationship between leverage and voluntary disclosure for US, UK, and continental European multinationals. Sengupta (1998) reports a marginally significant, negative relation between leverage and analyst disclosure quality ratings.

(Hossain, perera and Rahman, 1995)

Watts (1977), drawing on Jensen and Meckling (1976), suggests that the potential wealth transfers from fixed claimants to residual claimants increase as leverage increases. With debt holders price-protecting themselves, shareholders and managers have incentives to offer an increased level of monitoring such as voluntary disclosure of information in the published annual reports. Thus, agency theory posits that the extent of corporate voluntary disclosure will be an increasing function of leverage. While Craswell and Taylor (1992) found no statistically significant relation between leverage and voluntary disclosure of reserves by Australian oil and gas companies, Bradbury (1992) found a significantly positive association between leverage and the extent of voluntary segment disclosure in New Zealand diversified companies.

(Leventis and Weetman, 2004)

Agency costs (Jensen & Meckling, 1976) are expected to increase with gearing because the wealth transfer from debt-holders to managers and shareholders increases with gearing. Thus, highly geared companies may disclose more information to satisfy the needs of long-term creditors (Malone et al., 1993) and also to remove the suspicions of debt-holders regarding wealth transfer (Myers, 1977). In addition, a highly geared company has a greater obligation to satisfy the needs of long-term creditors for information (Wallace et al., 1994). However, applying the rationale of signaling theory (Akerlof, 1970) it could be argued that lower-geared companies may wish to draw attention to their financial structure by giving more voluntary disclosure. Hossain et al. (1995) found a significant positive association between gearing and disclosure. In contrast, Chow and Wong-Boren (1987), Hossain et al. (1994), Meek et al. (1995), Raffournier (1995), Inchausti (1997) and Patton and Zelenka (1997) found no significance. We have no specific expectation of the direction of association.

State ownership

Substantial state ownerships have been commonly observed among many listed Chinese companies, as prior to the economic reformation starting in 1980s all enterprises were owned and managed by the state and accounting practice were simply concerned with reporting compliance with state economic plans (Sami and Zhou, 2008). Resulting from decentralization of state-owned enterprises and separation of government from enterprises, under the current corporate system it seems that the state is entitled only to a dividend on its shares in the company’s assets (Ferguson, Lam and Lee, 2002). Based on the framework of agency cost, Ferguson et. al (2004) suggest that due to 

Investors’ concern of the potential for asset stripping or misappropriation, through excessive wage increases, and the role of the government as a major shareholder, companies with higher proportion of state ownership are subject to greater agency cost, thus greater need for information disclosure. Also, Lin et al (1998) argues that the extent of state ownership is positively related to the strategic importance of the companies deemed by the PRC government; hence companies with higher state ownership tend to face lower proprietary costs.

Therefore, firms with greater state ownership are assumed to have significantly greater incentives to disclose additional information. Nevertheless, Ferguson et. al (2004) observed that H-Share firms disclose significantly more information than Red-Chip firms despite the fact that both are majority owned by the state, indicating that state ownership alone does not drive disclosure policies. In contrast to Ferguson’s hypothesis of the positive relation between state ownership and disclosure level, Haw (2008) argues that in the case of listed state-owned enterprises, where state and legal-entity ownership dominate and managers are frequently appointed by the government, the managers are more likely to be strongly motivated to act in the best interests of the government and its representative organizations, and have less incentive to report high-quality accounting information or seek quality auditing for their public financial reports.

Considering the lack of proper incentives for managers together with the relatively weak corporate-governance structures in Chinese firms, Fan and Wong (2002) suggest an adverse affect of the state ownership on the information environment of these firms; hence, companies with greater state ownership may have less incentive to disclosure information for external investors. In view of the mixed theoretical predictions and ambiguous empirical result, the present study tends to only focus on the extent of the association between state ownership and disclosure level without specifying the direction of the association. Therefore, it is hypothesised as:

H3: companies disclosure is influenced by state ownership.

(In-Mu Haw, 2008)

Corporate governance systems are less effective in China's markets than those in the West. In the government-controlled economy of China, managers of listed state-owned enterprises (SOEs) are frequently appointed by the government, who is the controlling shareholder. Recent studies show that such ownership structures adversely affect the information environment of these firms, which results in a high level of information asymmetry and a low level of informativeness of accounting earnings (Fan and Wong, 2002; Haw, Hu, Hwang, & Wu, 2004). The managers of listed Chinese firms, where state and legal-entity (mostly SOEs) ownership dominate, are strongly motivated to act in the best interests of the government and its representative organizations, and have less incentive to report high-quality accounting information or seek quality auditing for their financial reports (Ball, Kathari, & Robin, 2000).

Another difficulty arises from the lack of proper incentives for Chinese managers and auditors as well as weak corporate-governance structures in China.

The discussant points out that less effective corporate governance systems (characterized by dominant state and legal-entity ownership) and relatively low litigation risk in China's markets (compared to those in the West) could provide opportunities for managers of listed Chinese firms to act in the best interests of the government and its representative organizations rather than report high-quality accounting information or seek quality auditing of their financial reports (Ball, Kothari, & Robin, 2000).

(Raffournier and Geneva, 1995)

Agency relations are likely to play a major role in the disclosure policy of companies because annual reports can be used to reduce monitoring costs. The separation of ownership and control generates agency costs resulting from conflicting interests between managers and stockholders. Jensen and Meckling (1976) showed that managers may be motivated to divert a part of the firm's wealth into perquisites. To the extent that this behaviour is anticipated by shareholders, managers bear the expected cost of these wealth transfers, so that they may have an incentive to engage in bonding activities to reassure shareholders that they are acting in a manner consistent with shareholders' interests.

Providing extended information may be seen as a bonding activity since it contributes to limit wealth transfers by making them more apparent. Deviations from wealth-maximizing behaviour are more likely in companies whose managers hold a small part of stock; these firms can be expected to disclose more information than those controlled by large shareholders. The annual report is the main, if not the exclusive, source of information for small shareholders who, because of their small part of the firm's wealth, cannot incur large expenditures in order to ascertain the wealth transfers of managers. Managers of firms whose ownership is diffuse have thus an incentive to disclose more information in order to help shareholders in monitoring their behaviour.

(Ferguson, Lam and Lee, 2002)

In recent years, many economies have sought to implement enterprise system reform as a means to improve productivity. These include not only the PRC and the former Soviet Bloc countries, but also developed countries such as France and New Zealand (World Bank, 1995). A predominant characteristic of SOEs in the PRC and other command-type economies was their lack of autonomy. The state determined production technology, product mix and input/output prices, while the SOEs delivered all revenues to the state. The influence of managers’ actions on profitability was marginal and as a result, discipline based on the level of profit alone was impossible (Lin et al., 1998).

Consequently, many SOEs were poorly run and incurring substantial losses. To improve efficiency, in 1979 the PRC initiated a series of reforms designed to foster transition to a market economy. In the first stage, SOEs were allowed to share in performance improvements under a profit retention program that gave 12 per cent of increased profits or reduced losses to the SOE. Next, managerial autonomy was increased via the replacement of the profit-retention system with a contract responsibility system whereby the SOEs agreed to deliver predetermined amounts of revenue to the state and retained the residual.

Finally, in the late 1980s the contract responsibility system was replaced by the current corporate system under which the state is entitled only to a dividend on its share in the SOEs’ assets. Increased management autonomy accompanied by the need to raise capital externally, however, led to the need for improved governance structures. The listing of shares on both domestic and international exchanges is viewed as an essential part of this reform. In addition to providing a conduit for external capital, listing is expected to improve corporate governance, information disclosure, and efficiency.

To date, the disclosure practices of former wholly SOEs, for whom accountability to external investors, public information disclosure and competition for capital are relatively new, have not been examined. The decision to disclose additional financial statement information is typically modeled within a cost–benefit framework (e.g., Choi and Levich, 1990; Meek et al., 1995). Costs include increased exposure to competitive and political costs via disclosure of proprietary information as well as the cost of preparing and disseminating additional information. Benefits include lower capital costs, improved marketability of company shares, and enhanced corporate image (Choi et al., 1999; Meek et al., 1995).

Thus, despite their status as relatively new entrants, disclosure by PRC SOEs appears sensitive to the information demands in competitive international capital markets. Further, state ownership alone does not appear to drive disclosure policies. H-Share firms disclose significantly more information than Red-Chip firms despite the fact that both are majority owned by the state. Although the presence of substantially higher costs and lower benefits of disclosure in the PRC could explain these differences, their magnitude suggests that disclosure by H-Share firms on the SEHK may also reflect the effects of state-encouraged disclosure policies.

Thus, in addition to H-Share firms’ lack of prior history, important investor concerns include management quality, the potential for asset stripping or misappropriation, de-capitalization through excessive wage increases, and the role of the government as a major shareholder (Chen and Firth, 1999). Therefore, ceteris paribus, H-Share firms face significantly greater incentives to voluntarily disclose additional information. Proprietary costs, however, also affect disclosure (Verrecchia, 1983). The benefits of voluntary disclosure must be weighed against the costs of providing information that may invite or assist competition or regulation.

Compared to other SEHK-listed firms, H-Share firms also face significantly lower proprietary costs. Most operate in industries deemed by the PRC government to be of strategic importance and are hence shielded from international competition (Lin et al., 1998). Thus, additional disclosure by H-Share firms is also likely to be less costly. This potential for greater incentives and lower disclosure cost leads to our first hypothesis:

Performance-Related Variables

Profitability

In the view of agency theory, Singhvi and Desai (1971) suggest a positive relation between profitability and company disclosure. They explain that companies with higher earnings are more willing to make greater disclosure to support management compensation and assure investors of the firm’s profitability. Political cost theory agrees that profit firms are under greater social and political pressures to disclose more information in order to justify the level of profits (Inchausti, 1997). Moreover, this positive relation between profitability and disclosure is also supported by signalling theory; as profitable and well-run firms are assumed to have more incentives to differentiate themselves from less profitable firms through greater disclosure in order to complete for funds (Foster, 1986). 

Alternatively, it can also be argued following a signalling scenario that intensive information disclosure can be used by less profitable firm for explaining poor performance (Leventis and Weetman, 2004). Similar to the contradicting theoretical indications, the prior empirical evidence were also ambiguous. Singhvi and Desai (1971), Garcia and Monterrey (1992), and Patton and Zelenka (1997) documented a statistically significant positive relation between profitability and disclosure using univariate tests; however, this influence disappears when profitability was included in multivariate model.

No statistically significant association was found in studies of McNally(1982), Malone(1993), Wallace (1994), Meek (1995), Inchausti (1997) and Leventis and Weetman (2004). Considering the current study of sample annual reports in year 2008 when majority companies were suffering from a serious global economy recession, it may be the case that companies with negative profitability would disclose more information concerning the adverse impact of the unfavourable overall economic statue in hoping of assuring the market about future growth. Overall, at the current stage there is no clear sign about the direction of association with profitability; thereby, this study tends to merely concern with the context of the association between profitability and disclosure rather than identifying the direction. Thus, the third hypothesis was stated as:

H3: company disclosure level is influenced by leverage.

(Patteon and Zelenka, 1997)

Although Raffoumier (1995, p.263) says 'The influence of profitability on voluntary disclostire is obvious' and hypothesizes a positive relationship between profitability and extent of disclosure, he also notes that this hypothesis has not been strongly supported in prior research. Also, Lang and Lundholm (1993, p. 250) note that both the theoretical and the empirical literatures are ambiguous conceming the effect of performance on disclosure. Thus, although we expect performance to affect the extent of disclosure in Czech annual repons, in this research we do not 'sign' our performance independent variable. Instead, we use two-tailed tests conceming the significance of the relationship between a firm's Retum on Equity performance and the extent of its annual report disclosure.

(Malone, Fries and Jones, 1988)

The fourth research question posed by the study was: Are earnings levels positively associated with the extent of financial disclosure? Singhvi and Desai (1971) showed a positive relationship between financial disclosure and both rate of return on equity and earnings margin (net income divided by net sales). Their premise for testing these variables was that as earnings rise, managers are willing to disclose more information to support management compensation contracts and assure investors of the firm's profitability.

At higher levels of earnings, the firm is also more capable of bearing higher disclosure costs. The inverse of these reasonings should hold true as earnings fall. The two variables used by Singhvi and Desai (1971), rate of return on net worth, designated RR, and earnings margin (net income divided by net sales), designated EM, were used as measures of profitability in the current study. The hypotheses tested, stated in their alternate form, were

(Meek, Roberts and Gray, 1995)

There is a cost in being perceived as a "lemon" [Akerlof 1970]. Thus, profitable, well-run firms have incentives to distinguish themselves from less profitable firms in order to raise capital on the best available terms. One way to do this is through voluntary information disclosure [Foster 1986, p. 32]. Thus, more profitable firms can be expected to disclose more voluntary accounting information. However, McNally et al. [1982] do not find their profitability measure to be significant in explaining voluntary disclosures by New Zealand companies. (The authors offer no explanation as to why.)

(Raffournier and Geneva, 1995)

The influence of profitability on voluntary disclosure is obvious. As argued by Singhvi and Desai (1971), when the rate of return is high, managers are motivated to disclose detailed information in order to support the continuance of their positions and remuneration. Inversely, when the rate of return is low, they may disclose less information in order to conceal the reasons for losses or declining profits. This hypothesis is not strongly supported by empirical evidence. Although Singhvi and Desai (1971) found, as expected, a positive relation when profitability alone is considered, this influence disappears when the rate of return is included in conjunction with other variables. Similar results have been obtained by Garcia-Benau and Monterrey-Mayoral (1992). Cowen et al. (1987) also noted that highly profitable companies do not tend to disclose more social responsibility information than less profitable firms.

(Inchausti, 1997)

The greater the profitability of a firm the greater the level of disclosure it will have. Agency theory suggests that managers of very profitable firms will use extemal information in order to obtain personal advantages. Therefore they will disclose detailed information in order to support the continuance of their positions and compensation arrangements. Signalling theory implies that owners will be interested in giving 'good news' to the market in order to avoid the undervaluation of their shares. Political process theory argues that firms with large profits will be interested in disclosing more information in order to justify the level of profits.

This hypothesis is consistent with the results of Cerf (1961) and Singhvi and Desai (1971). On the other hand Wagenhofer (1990) unsuccessfully tested the opposite hypothesis based on signalling theory namely that information is used as a mechanism for explaining 'bad news'. Therefore signalling theory may suggest two contradictory relationships between profitability and disclosure of information. Given the evidence and the underlying reasons for each hypothesis, it was considered more appropriate to suggest a positive relationship as in H3 above. Two ratios were tested as independent variables representing profitability: operating income/ total assets (PA), and net income/equity (PE).

(Leventis and Weetman, 2004)

Agency theory suggests that managers of very profitable companies will use external information in order to obtain personal advantages like continuance of their positions and compensation arrangements (Inchausti, 1997). Profitable companies may disclose more information in order to be distinguished from less profitable companies, following a signalling scenario (Akerlof, 1970). However, it may be the case that less profitable companies may disclose more information to explain the reasons for negative performance and reassure the market about future growth.

It may also be the case that companies, by disclosing “bad news,” reduce the risk of legal liability, severe devaluation of share capital and loss of reputation (Skinner, 1994). Profitability has been used as a variable in many studies. Singhvi and Desai (1971), Patton and Zelenka (1997), and Owusu-Ansah (1998) found a significant association. In contrast, McNally et al. (1982), Malone et al. (1993), Wallace et al. (1994), and Meek et al. (1995), found no association. In view of the mixed evidence from prior research, we have no specific expectation about the direction of association with profitability.

Liquidity

Similar to leverage, liquidity can also be considered as a proxy for risk, indicating the company’s ability to meet its short-term financial obligations, assuming the firm’s ongoing operations without liquidating its fixed assets. It is a critical factor investigated by investors, lenders and regulatory institutions in terms of evaluating the firm. Therefore, it is predicted by signalling theory that firms with strong liquid position have more incentives to disclose more information in order to differentiate themselves from companies with less liquidity (Leventis and Weetman, 2004). However, comparing to the other frequently tested variables (such as size and leverage), as the relationship between liquidity and disclosure level is less obvious (Leventis and Weetman, 2004), a limited number of previous researches attempted the variable of liquidity in their regression models, such as Wallace et.al (1994), Wallance and Naser (1995), Owusu-Ansah (1998), and Leventis and Weetman (2004).

Among these studies, only Wallance et. al. (1994) managed to document a statistically significant association between liquidity and company disclosure level. To examine the relation between the liquidity and the level of company information disclosure with respect to the sample Chinese companies, it is hypothesized that:

H5: company’s disclosure level is associated with firm’s liquidity.

(Leventis and Weetman, 2004)

Liquidity may be viewed as a measure of risk. The ability of a company to meet its short-term financial obligations without having to liquidate its long-term assets or cease operations is an important factor in the evaluation of the company by interested parties such as investors, lenders and regulatory bodies (Wallace & Naser, 1995). Companies that have strong liquidity may disclose more information to distinguish themselves from less financially strong companies. The rationale for this variable lies in signaling theory (Belkaoui & Kahl, 1978).

However, the relationship between liquidity and voluntary disclosure may not be straightforward. Wallace et al. (1994) observed significant correlation. No association was reported by Wallace and Naser (1995), and Owusu-Ansah (1998). In view of the mixed evidence from prior research, we have no specific expectation about the direction of association with liquidity.

Market-related Variables

Market return

According to Leuz and Verrecchia (2000), company’s commitment to increase levels of disclosure can lower the information asymmetry component of the firm’s cost of capital. Because greater information disclosure should enhance liquidity and thereby lower the transaction costs, consequently reducing the cost of equity. Also, increasing disclosure can be used as a mean to improve information risk which is a function of asset returns. In that sense, Bhattacharyya, Raychauduri and Sadhalaxmi (2005) contend that appropriate and timely information should assist investors to predict the future cashflows and the uncertainty related to these cashflows, thereby evaluating companies more effectively.

In other word, reduction in information asymmetry through greater information disclosure can help to reduce the cost of equity capital and thus increase the valuation of such companies. For instance, mentioned by Bhattacharyya, Raychauduri and Sadhalaxmi (2005), previous empirical study by Collins (1975) assesses the impact of the introduction of segment reporting on stock market for US listed companies, higher abnormal returns were observed for the portfolio of companies disclosing greater segmental information, compared to the firms with less segmental information disclosure.

From the point view of signalling theory, Leventis and Weetman (2004) also suggest that high return companies may be encouraged to disclose more information in order to attract greater investors’ attention, thereby distinguishing themselves from low return companies. However, Leventis and Weetman (2004) found no significant relation between market returns of companies listed on the Athens stock exchange and their disclosure level.

On the other hand, considering the current study of company reports of year 2008, where most companies lost substantial value in stock market, signalling theory may also argue that companies with significant negative market returns may disclose more information to explain their poor market performance in hope of assuring investors about their future returns. In that sense, the current study prefers to merely test the content of the association between market returns and disclosure level without specifying the direction of association. The resulting hypothesis is:

H6: Company disclosure extent is associated with market return.

Most of the empirical research linking greater disclosure to lower cost of capital has been conducted on the United States stock markets. Stigler (1964) studies the effect of Securities Exchange Act1, 1934, on the volatility of new stock issues. He documents that the volatility of the returns of new issues has reduced subsequent to the Act. Jarrell (1981) confirms the results obtained by Stigler (1964) with improved statistical techniques. Friend and Herman (1964) view that lower volatility during the post-SEC period has attracted more risk-averse investors and has increased the level of investment in the United States capital markets.

On the contrary, Benston (1973) does not find any impact of the Act on abnormal returns and variability, subsequent to the applicability of the Act. Friend and Westerfield (1975) defy the results obtained by Benston (1973) by demonstrating that the result is due to wrong classification of firms. Majority of the empirical results related to this Act confirm that increased disclosure subsequent to the SEC Act is beneficial to the investors.

The Securities Exchange Commission has made segment reporting mandatory for listed companies in United States starting from the year 1970. Collins (1975) assesses the impact of segment reporting on stock market and finds that the portfolio of companies disclosing greater segmental information has earned higher abnormal returns than the group with less segmental information.

We conjecture that good governance coupled with timely and appropriate information dissemination will reduce the information asymmetry between the managers and the investors. The increased information flow and congenial governance system will enable investors to correctly predict future cashflows and uncertainty related to these cashflows, and thus help in appropriate valuation of the company. Based on the literature review we argue that reduction in information asymmetry will reduce the cost of equity capital and thus increase the valuation of such companies.

(Leventis and Weetman, 2004)

A positive association between share returns and disclosure has been suggested in the literature (Horngren, 1957). It has been also argued that higher information asymmetry between market participants leads to higher transaction costs and lower liquidity which raises the required rate of return and lowers current stock prices (Bartov & Bodnar, 1996). In the context of signaling theory the view could be that high return companies may wish to distinguish themselves from low return companies. Our expectation is that there will be a positive association between voluntary disclosure and share return.

Industry type

Intuitively, it is assumed that company’s disclosure level may vary across different industry types. Because each industry with its specific characteristics (such as nature of product, industry structure, risks, legal requirement, cultural and historical factors) is likely to be subject to different proprietary cost, consequently leading to diversity of disclosure level across industries (Verrecchia, 1983; Watts and Zimmerman, 1986). For instance, regulated industries with higher potential political costs tend to engage in greater information disclosure; on the other hand, companies operating in highly competitive industries (like high tech industries) are likely to be more concerned with the possible leakage of proprietary information, thereby, may be more reluctant to disclose information regarding research and development of new products (Meek, Roberts and Gray, 1995).

Furthermore, Inchausti (1997) argue that within the same industry, companies tend to follow the similar disclosure policy adopted by the industry leaders. The underlying rational of this “following leader” strategy is driven from signalling theory, which assumes that if a firm adopt a different corporate reporting strategy from the one used by the dominant companies, it could be viewed by the market as a signal of 'bad news'. Some evidence of empirical researches also supports the inclusion of industry type as a variable to explain the extent of disclosure.

For example, Cooke (1989; 1992) document the significantly greater disclosure by Swedish non-trading firms and Japanese manufacturing comparing with the Swedish trading firms and Japanese non-manufacturing respectively. Similar results were concluded by Meek et al (1995); while Raffournier (1995), Patton and Zelenka (1997), Inchausti (1997) and Leventis and Weetman found no significance for the industry factor. Based on the mixed empirical evidence, this study prefers to mainly focus on the association between industry types and disclosure level without specify the direction. Thus, the hypothesis is established as:

H6: Industry type exerts an influence on voluntary disclosure.

Empirical research also supports the inclusion of type of industry as a variable in our study. For example, Cooke (1989b, p. 188) found that voluntary disclosure by trading firms in Sweden is significantly lower than voluntary disclosure by non-trading firms, while Cooke (1991, p. 186;

1992, p. 236) found that manufacturing firms disclose more than non-manufacturing firms in a study of both listed and unlisted Japanese firms. But there is no definite theory to expect one industry to outperform any other in disclosure. However, we suggest that HK firms which operate in the property category may be expected to have less information to disclose than HK firms which operate in the trading or industrial category, as trading and industrial activities are more complex and generate more reportable events than property activities.

(Patteon and Zelenka, 1997)

Finally, some researchers (see Marston and Shrives, 1996 for a review), have suggested that a firm's industry type is associated with the extent of disclosure. For example, Cooke (1992) found that in Japan manufacturing firms disclose more than non-manufacturing firms. Raffotimier (1995) also hypothesized that manufacturing firms have incentives to disclose more information.

(Meek, Roberts and Gray, 1995)

Proprietary (i.e., competitive disadvantage and political) costs vary across industries [Verrecchia 1983]. For example, because of the nature of their products and their research and development, chemical companies are likely to be more sensitive about disclosures to competitors and the public than companies in certain other industries. The relevance of selected items of disclosure can also vary across industries. Research and development is more relevant for companies in high tech industries, for example. Therefore, industry membership may exert an influence on voluntary disclosure. Cooke [1989, 1991] finds mild evidence of an industry effect in his studies of Swedish and Japanese companies.

(Raffournier and Geneva, 1995)

In the case of Japan, Cooke (1992) found that manufacturing companies disclose more information than non-manufacturing firms, regardless of their listing status. Apart from historical and other country-specific reasons, Cooke suggests that this fact could arise from the international exposure of the manufacturing sector. A similar phenomenon may occur in

(Inchausti, 1997)

Firms in the same industry will disclose similar information to third parties. This hypothesis tries to explain the disclosure policy adopted by firms as a way of following other companies in the same industry. According to Wallace et al. (1994) industry, as well as listing status and auditor type, are factors that influence the corporate reporting culture of firms. It is suggested that firms from a particular industry may adopt disclosure practices additional to those mandatory for firms in all industries.

So it is reasonable to think that the policy of financial information disclosure will differ across industries. Although this argument is not based on a clear causal relationship, it may be justified using political cost theory or signalling theory. As Watts and Zimmerman (1986: 239) suggest industry membership probably affects the political vulnerability of a firm, but in their opinion firm size proxies for industry because firms within an industry have similar sizes.

Therefore it has been considered convenient to include the industry variable in order to separate this factor from the size variable. Moreover, in the author's opinion, if a firm does not adopt the same corporate reporting strategy as others from the same industry, it could be interpreted by the market as a signal of 'bad news'. This argument follows directly from signalling theory as discussed above. When this hypothesis was tested in other studies it did not always produce satisfactory results. Amemic and Maiocco (1981) and Wagenhofer (1990) found an industry effect, but Wallace et al. (1994) did not.

(Ferguson, Lam and Lee, 2002)

Competitive and political costs are also likely to vary across industries (Verrecchia, 1983; Watts and Zimmerman, 1986). For example, higher potential political costs in regulated industries are likely to result in higher voluntary disclosure, whereas firms in highly competitive industries may curtail disclosure to avoid leakage of proprietary information. Empirically, Cooke (1989, 1991) and Meek et al. (1995) report evidence of an industry effect on the level of disclosure. Finally, firms listed in multiple capital markets face additional demands for information.

As the number of shareholders increase and ownership becomes more dispersed, monitoring costs, and hence demand for additional information, increase (Fama and Jensen, 1983; Schipper, 1981). Consistent with this, prior research has found that voluntary disclosure in firms’ original place of listing increases for multiple-listing firms (Cooke, 1989, 1991; Hossain et al., 1995; Meek et al., 1995). To control for these factors, we include size, leverage, industry and multiple-listing status in our regression model.

(Leventis and Weetman, 2004)

Industry has been identified by theoretical and empirical research as a factor with explanatory potential for voluntary disclosure. Industries have different characteristics which may relate to competition, product differentiation, industry structure (e.g. oligopolies), growth, demand instability, quasi-legal constraints, risks and also to the specific culture which may be related to historical factors (Hambrick & Abrahamson, 1995). These may provide scope for differential disclosure policy as suggested by Dye and Sridhar (1995). Furthermore, the existence of a dominant company in an industry with high levels of voluntary disclosure may have bandwagon effects for all companies within this industry.

This “follow the leader effect” has been detected in many studies of accounting disclosure (e.g. Belkaoui & Kahl, 1978; Cooke, 1991). Moreover, different industries have different proprietary costs that give incentives for companies belonging to the same industry to disclose either more or less (Verrecchia, 1983). Industry may also affect the political vulnerability of a company (Watts & Zimmerman, 1986, p. 239). Results for industry effects tend to differ. While Belkaoui and Kahl (1978), Cooke (1989), Meek et al. (1995), found significance, Raffournier (1995), Patton and Zelenka (1997) and Inchausti (1997) found no significance for the industry factor. In view of the mixed evidence from prior research, we have no specific expectation about the direction of association with industry type.

Listing status

In many prior studies, company’s listing status is assumed as a function of information disclosure. Meek (1995) states that firms with foreign listing face additional capital market pressures for the information disclosure, compared to companies of which shares are only listed domestically. According to Malone (1988), firms listed overseas are usually subject to more reporting requirement levied by the foreign government in excess of the domestic listing requirements.

In that sense, if these firms had already accumulated additional data for complying with these special reporting requirements, they are more likely to disclose that information rather than withholding it, as the marginal disclosure cost is relatively lower. Additionally, Healy et al, (1995) argue that firms with foreign listing may well have economic ambitions of obtaining funds internationally; thereby, they would be motivated to disclose more information to overcome the informational asymmetry that could lead to lower prices for their securities.

Moreover, Patteon and Zelenka (1997) suggest that another rational for foreign listed firms to disclose more information can be the concerns of their reputation and image with bankers, investors, and regulatory institutions in the international capital market. This argument is supported by Ferguson et. al (2002)’s study of former wholly state-owned People’s Republic of China (PRC) enterprises, listed on the Stock Exchange of Hong Kong (SEHK).

According to Ferguson et. al (2002), these firms have been selected for “showcasing” in international capital markets, so that they tend to adopt a “state-encouraged” disclosure policies, which exhibit substantially higher disclosure than purely domestic listed firms despite of the fact that the disclosure costs may excess the benefits of disclosure. In the author’s opinion, the reason advocated by Ferguson is of great relevance to the current study. Together with the evidence that listing status have significant effects on disclosure extent from previous empirical researches (Cooke, 1989, 1991; Gray et al.1993; Wallace et al. 1994; Hossain et al.1995 ; Meek et al.1995; Inchausti, 1997; Leventis and Weetman, 2004), the present study will test the hypothesis concerning listing status as:

H7: Foreign listing status is associated with company’s disclosure level.

(Patteon and Zelenka, 1997)

While the Prague Stock Exchange required listed firms to submit their full financial statements (including footnotes) to the Exchange, it imposed no specific disclosure requirements conceming firms' annual reports. Thus, listing on the exchange does not automatically mean that such firms will have a greater extent of disclosure in their annual repons than unlisted firms. However, firms that are listed may well have economic ambitions that will require additional capital.

This should lead firms to disclose more information to overcome the informational asymmetry that could lead to lower prices for their securities (Healy et al, 1995). In addition, firms that list may be expected to be more concemed with their reputation and image with bankers, investors, the govemment, etc. One means of disclosure that might be used to deal with these concems is a firm's annual report. Thus, we expect Czech firms that are listed on the Prague Stock Exchange to disclose more in their annual financial repons than firms that are not listed.^

(Malone, Fries and Jones, 1988)

Because of different registration requirements by the various exchanges, and thus different sets of information having been gathered, several authors (e.g., Leftwich, Watts, and Zimmerman [1981]; Buzby [1975]; Singhvi and Desai [1971]) have suggested that annual report disclosures may vary between listed and unlisted firms. Past studies are not in agreement on whether or not listing status is associated with the quality or extent of financial disclosure (e.g., Cerf [1961]; Singhvi and Desai [1971]; Buzby [1975]). An additional reason that firms with foreign operations might provide more extensive financial disclosures is that firms dealing with more than one political entity may have to fulfill separate different reporting requirements by each entity. If firms accumulate additional data in order to comply with these special reporting requirements, the firms are more likely to disclose that information than firms not subject to special reporting requirements.

The subject of marginal information requirements made by foreign entities has the same characteristics as a firm's industry diversification, or diminishing marginal information requirements. Information requirements may be levied by the foreign government of the country in which the firm was operating and/or by the foreign exchange on which the firm's securities were traded.

(Meek, Roberts and Gray, 1995)

Firms whose shares are listed internationally face additional capital market pressures for the disclosure of information, compared to companies whose shares are only listed domestically. Cooke [1989, 1991] finds an international listing effect for Swedish and Japanese companies, respectively. As noted earlier, Gray et al. [1993] find the same for their sample of U.S., U.K. and Continental European MNCs

(Hossain, perera and Rahman, 1995)

However, more recent studies suggest that there is a complementary (positive) association between foreign listing status and the extent of voluntary disclosure. For instance. Meek and Gray (1989) found that continental European multinational corporations listed on the London Stock Exchange voluntarily disclosed information in excess of that which was required under London Stock Exchange rules. Likewise, Hossain et al. (1994) found that Malaysian multinationals listed on London Stock Exchange voluntarily disclosed more information in their annual reports than companies listed only on Kuala Lumpur Stock Exchange. Overall, the earlier research suggests a positive association between foreign listing status and the extent of corporate voluntary disclosure. Accordingly, in the New Zealand context it is hypothesized that:

(Leventis and Weetman, 2004)

Listing status has been proposed in many studies as a factor explaining variability in accounting disclosure. Fama and Jensen (1983) argue that stock exchange listing is a mechanism for mitigating incentive conflicts between contracting parties. Cooke (1991) suggests that companies with different listing status may disclose different levels of voluntary information because they may have different objectives in raising capital. Companies having a listing status, or a more prestigious listing status, may be more willing to disclose enhanced voluntary information in annual reports. The significance of listing status has been supported empirically by Cooke (1989, 1991), Wallace et al. (1994), Hossain et al. (1994), Hossain et al. (1995), Meek et al. (1995) and Inchausti (1997). Our expectation is that there will be a positive association between voluntary disclosure and listing status.

Auditor

Previous researches have suggested that auditing firms also play a role in determining the disclosure policy of their clients. Singhvi and Desai (1971) argue that companies audited by large and well-known auditing firms may be incited to make greater disclosure. According to Jansen and Mechking (1976), auditing can be used as a means to reduce agency costs; thus, companies experiencing greater agency costs tend to have a greater demand for higher-level audit quality.

Comparing with smaller CPA firms, which are assumed to be more sensitive to client demands due to the economic consequences associated with the loss of a client, larger and international well-known auditing firms have a greater incentive to maintain independence from clients' pressure for limited disclosure because of the economic consequences associated with potential damages to their reputation (Chow and Wong-Boren, 1986). Therefore, larger CPA firms have a greater incentive to require adverse disclosures by the client, consequently increasing the level of information disclosure (Patteon and Zelenka, 1997).

Moreover, in the view of signalling theory, Morris (1987) further suggests that the company’s choice of a well-known auditor can serve as a signal to the market about the quality of a firm's disclosure, as larger international CPA firms are commonly believed to bring enhanced credibility to financial reports, compare to local auditing firms. Despite of the general agreement about the positive relation between auditing firm’s size/reputation and the level of company disclosure, the results of prior studies were not consistent.

Singhvi and Desai (1971), Craswell and Taylor (1992), Raffournier and Geneva (1995), Patton and Zalenka (1997) and Inchausti (1997) confirmed the theoretical hypothesis; while, Firth (1979), Malone, Fries and Jones (1988), Garcia and Monterrey (1992), Wallace et al. (1994), and Hossain, perera and Raham (1995) did not find any significant relation. In short, in the current study the relation between the choice of auditing firm and the information disclosure is hypothesized as that:

H7: the extent of disclosure is positively associated with a company having a Big4 intemational audit firm instead of a domestic auditor.

(Patteon and Zelenka, 1997)

Firms are likely to choose an extemal auditor as pan of their overall financial disclosure strategy. Large, intemational CPA firms are commonly believed to bring enhanced credibility to financial reports. Thus, we expect that extent of disclosure will be positively associated with a company having a Big 6 intemational audit firm instead of a domestic auditor.

(Malone, Fries and Jones, 1988)

Of additional interest was the possible difference in a firm's financial disclosures as required by its external auditor. A variety of studies, although mixed in conclusions, examined differences in certified public accounting (CPA) firm size as the source of disparity among disclosures. (See for example Wright [1974] and DeAngelo [1980].) The principal argument forwarded by these studies was that smaller CPA firms are more sensitive to client demands because of the economic consequences associated with the loss of a client. Larger firms have a greater incentive to require adverse disclosures by the client. A fifth question was thus posed: Is the extent of financial disclosure greater for firms that engage larger audit firms than for those that engage smaller ones?

(Raffournier and Geneva, 1995)

Several authors have suggested that auditors play a role in defining the disclosure policy of their clients. Specifically, it has been argued that large and well-known auditing firms may incite firms to disclose more information (Singhvi and Desai, 1971; Firth, 1979). The empirical evidence is conflicting, since the former study found that clients of national 'Big Eight'^exhibited a higher level of disclosure, while the latter study did not reveal differences in the extent of disclosure in relation to the size of auditing firms.

(Inchausti, 1997)

Firms audited by one of the Big Six audit firms will disclose more information.

Auditing firms may use the information disclosed by their clients as a way of signalling about their own quality. According to DeAngelo's argument (1981) the larger audit firms have incentives to supply a higher level of audit quality, and they risk losing some of their reputation if they are associated with clients whose reporting practices are considered as offering 'bad quality'.

Therefore as Craswell and Taylor (1992) suggest a firm's choice of auditor is likely to be associated with the decision to disclose more or less information. Big Six audit firms are larger than others, and it is suggested that clients of these firms will disclose more information. This hypothesis might also be based on the argument that companies audited by Big Six firms have substantial agency costs, and try to reduce them by contracting with these auditing firms.

Auditing may be considered as a means of reducing agency costs (Jensen and Meckling, 1976; Chow, 1982; Watts and Zimmerman, 1983), and as Francis and Wilson (1988: 680) argue: 'It follows that when agency costs are greater there is increased demand for higher-level audit quality." As noted above there is a positive relationship between agency costs and the disclosure decision. This qualitative aspect was considered using a dichotomous variable, with the value 1 if an auditing firm belongs to the Big Six, and 0 if it does not. The results of previous studies are not consistent. Singhvi and Desai (1971) and Craswell and Taylor (1992) confirmed the hypothesis, but Firth (1979), Raffoumier (1991) and Wallace et al. (1994) did not find any relation.

(Hossain, perera and Rahman, 1995)

Schipper (1981) and Watts and Zimmerman (1986) put forward the view that choice of external auditors is a mechanism which helps to alleviate conflicts of interest between principals and agents. Moreover, DeAngelo (1981) and Chow and Wong-Boren (1986) consider that Big-6 audit firms have incentives to maintain independence from clients' pressure for limited disclosure because of the economic consequences associated with potential damages to their 'brand name' (reputation). Therefore, they encourage their clients to disclose a greater amount of information in published annual reports, indicating that the level of voluntary disclosure is likely to be higher for companies audited by Big-6 audit firms.

The signalling literature, which may to a certain extent be complementary to agency theory literature (Morris, 1987), further suggests that the choice of an external auditor can serve as a signal of firm value. For instance, Bar-Yosef and Livnat (1984) show that entrepreneurs are likely to choose a Big-6 audit firm, since such an action signals to investors the expectation of high cash flow. Likewise, Datar, Feltham and Hughes (1991, p. 4) analytically demonstrate that the selection of an auditor is a signal to the market about the quality of a firm's disclosure.

Craswell and Taylor (1992) found a significant positive association between type of auditor (i.e., Big-6 or non-Big-6) and voluntary disclosure of oil and gas reserves by Australian companies, while McNally, Eng, and Hassledine (1982) found no significant association between type of auditor and the extent of discretionary disclosure by New Zealand manufacturing companies.

Board composition

Proportion of independent non-executive directors within the board is another variable that has been tested by several previous researches with respect to the extent of disclosure. It is pointed out by Leftwich, Watts, and Zimmerman (1981) that the similar to the publication of annual reports, presence of independent directors on the board can also serve as a form of monitoring, aiming to improve the agency problem initiated from the separation of ownership and management.

In that sense, either a substitutive or a complementary relationship might be expected between the different forms of monitoring. According to Malone et.al. (1988), assuming the relationship is substitutive, the extent of disclosure is dependent on the monitoring requirements placed on the agents of the firm, thus the disclosure level is expected to decline as bigger proportion of independent non-executive directors are added to the board.

That is, by enhancing the presence of independent directors, which can be seen as the agents (management of the firm)’s efforts to satisfy the owners’ threshold of monitoring requirements, the firm has less need to provide additional information disclosures at the same time. Nevertheless, Malone’s hypothesis of the substitutive relationship was not supported by his empirical test concerning a sample of 125 oil and gas companies, as no significant association was found.

On the other hand, if the relationship is complementary, the presence of independent directors should be rather seen as a "more intensive monitoring package" selected by the managers of the firm; hence, a greater extent of financial disclosure is expected (Malone, Fries and Jones, 1988). This opposite assumption was adopted by Cheng & Courtenay (2006), they explain that the effectiveness of a board in monitoring management is primarily determined by its independence which is positively linked with the proportion of independent directors.

Therefore, firms with a higher proportion of independent directors on the board are expected to disclosure more information. Their hypothesis based on the complementary relation between proportion of independent directors and disclosure level was confirmed by their empirical result that boards with a majority of independent directors have significantly higher levels of voluntary disclosure than firms with balanced boards.

In view of the mixed empirical evidence, the resulting hypothesis is:

HI: The extent of disclosure is associated with the proportion of independent non-executive directors on board.

(Malone, Fries and Jones, 1988)

Finally, Leftwich, Watts, and Zimmerman (1981) pointed out that one form of monitoring is the presence of outside directors on the board. These directors are often present as trustees for the nonmanager owners of a firm and serve the function of providing their expertise to the chief executive officer (CEO) of a firm. If the extent of disclosure is dependent on the monitoring requirements placed on the agents of the firm, the extent of financial disclosure may decrease as outside directors are added to the board.

In a study published after the current study was committed to the proportion of outside directors as a variable. Chow and Wong-Boren (1987) affirmed the need to examine alternative mechanisms for the containment of shareholder-bondholder conflicts. They suggested the proportion of outside directors elected by the shareholders of a firm as one such mechanism. The last question addressed by this study was, therefore: Is the extent of financial disclosure less for firms with a greater proportion of outside directors than for those firms with a smaller proportion of outside directors? Because the publication of 10-K and annual reports and the presence of outside directors are both forms of monitoring, one might expect a relationship between the two.

But, as Leftwich, Watts, and Zimmerman (1981) pointed out, it is unclear whether the relationship is complementary or substitutive. If it is complementary, one would expect a greater extent of financial disclosure due to a "more intensive monitoring package" selected by the managers of the firm. On the other hand, if the relationship is substitutive, one might expect that in the presence of outside directors, the firm has less need to provide additional financial disclosures. Because the costs of providing disclosure are nontrivial costs, the study proposed that the relationship between disclosure and the presence of outside directors on the board is a substitutive one.

This argument takes the position that owners of a firm have a threshold of monitoring requirements that the agents try to satisfy. Note that in the case of the agency relationship, where monitoring costs are positive, the level of monitoring sought represents a solution that is pareto optimal. The presence of these costs simply implies that the benefits gained outweigh the costs. Thus, it seemed reasonable to tentatively hypothesize that the relationship between extent of disclosure and the presence of outside directors was a substitutive one. To test this issue, the hypothesis used, in its alternate form, was

HIO: Extent of financial disclosure is less for ESIC 1311 firms that have a greater proportion of outside directors on the board than for those with a smaller proportion of outside directors.

(Cheng & Courtenay, 2006)

This study, which examines the association between board monitoring and the level of voluntary disclosure, finds new evidence that firms with a higher proportion of independent directors on the board are associated with higher levels of voluntary disclosure. Although board size and CEO duality are not associated with voluntary disclosure, boards with a majority of independent directors have significantly higher levels of voluntary disclosure than firms with balanced boards.

Using a direct measure of voluntary disclosure, we find that boards with a larger proportion of independent, nonexecutive directors (our proxy for board-monitoring effectiveness) are significantly and positively associated with higher levels of voluntary disclosure. This is a notable contribution to the research on board monitoring, as it demonstrates for the first time the link between board independence and a direct measure of voluntary disclosure that can be generalized to the overall market. In addition, the results also indicate that firms with boards with a majority (N50%) of independent directors have higher levels of voluntary disclosure than firms with boards that do not have a majority of independent directors.

We also demonstrate that the level of independence suggested by the Singaporean regulator (33%) does not provide a monitoring capacity that is supportive of higher levels of voluntary disclosure. As discussed in John and Senbet (1998), the effectiveness of a board in monitoring management is determined by its composition, independence, and size. Composition and independence are closely related, since board independence increases as the proportion of independent, outside directors increases.

Fama (1980) views outside directors as referees whose task is to ensure that the board, as the ultimate internal monitor of managerial decision-making, protects the interests of the security holders. Fama and Jensen (1983) suggest that boards composed of a higher proportion of independent, outside directors (directors not involved in the direct operations of the firm) have greater control (ratification and monitoring) over managerial decisions. Independent directors have incentives to exercise their decision control in order to maintain reputational capital