The Determinants of Corporate Voluntary Disclosure: Evidence from the Tunisian Capital Market

Transcription

1 The Determinants of Corporate Voluntary Disclosure: Evidence from the Tunisian Capital Market Mohamed Chakib Kolsi* The recent financial corporate scandals highlighted the role of accounting disclosure in misleading investors. While factors affecting voluntary disclosure policy in developed countries are well documented, little evidence is provided about those affecting such disclosure in emerging economies. The aim of this paper is to examine some determinants of voluntary disclosure policy adopted by a sample of Tunisian listed firms. By computing a disclosure index and running a multivariate regression analysis, our results show that firm leverage, audit quality, financial sector and profitability ratio are significant determinants of voluntary disclosure policy adopted by Tunisian firms. These results can be explained by the crucial role played by banks in financing the Tunisian economy. Contrary to our predictions, both ownership structure and firm size seem to have no effect on disclosure policy, since Tunisian firms are characterized by concentrated ownership structure with a family character and homogenous size on an average. Introduction Recent corporate financial scandals highlighted by the subprime crisis (i.e. Lehman Brothers Holding Inc. and Bear, Stearns & Co. Inc.) have raised much suspicion about the fairness of financial information disclosed by firms. Corporate disclosure is critical for the functioning of an efficient capital market (Healy and Palepu, 2001, p. 1). In that, higher quality disclosure policy attenuates information asymmetry and agency problems. Also, agency theory (Jensen and Meckling, 1976) stipulates that demand for discretionary disclosure arises from information asymmetry and agency conflicts between managers and outside investors. Such information is useful for investors as it can reduce fraud and manipulations. Ronen and Yaari (1993) argued that the information has a dual role monitors the actions of the management (the stewardship role) and provides information about the value of the firm which affects the market price (the informativeness role) in the context of efficient market hypothesis. Evidence concerning voluntary disclosure is controversial. While most of the previous researches assume that voluntary disclosure is beneficial for firms (e.g., Welker, 1995), it seems that such firms can incur additional costs and illegal concurrent practices when disclosing future perspectives to outsiders. As argued by Ozbilgin and Penno (2005), the * Assistant Professor, Accounting, Emirates College of Technology, Abu Dhabi, Unites Arab Emirates. chakibkolsi@yahoo.fr The 2012 Determinants IUP. All Rights of Corporate Reserved. Voluntary Disclosure: Evidence from the Tunisian Capital Market 1

2 Author pls. chk. this website. general notion that higher quality information is accompanied by more general disclosure is in general not true. While largely examined in the developed markets, research concerning voluntary disclosure in the emergent markets is not sufficiently explored ( see for eg. Haniffa and Cooke, 2002 for Malaysian companies; Barako et al., 2006 for Kenyan experience; and Dahawy, 2009 for Egyptian firms). Many factors affect the disclosure policy adopted by firms, such as economic, financial, political and cultural factors. Such factors differ across countries and accounting systems. So, we plan to check for the determinants of the Tunisian voluntary disclosure policy. The paper first presents a brief survey of the Tunisian economy and its eventual impact on disclosure policy and reviews the literature related to voluntary disclosure, especially in the emerging economies. Later, it develops the hypotheses and defines the variables and model specifications. The paper then discusses the results before finally concluding with some implications and limitations of the study. A Survey of the Tunisian Economy Turning to the Tunisian economy, a majority of the firms have a family character with a smaller and homogenous size. Since 1987, Tunisia is in the process of economic reformation and liberalization, after decades of heavy state direction in the national economy. Governmental control of economic affairs has gradually decreased over the past two decades with increasing privatization of trade and commerce, simplification of the tax structure, and a prudent approach to debt. Real growth which averaged 4.6% in , reached 5.6% in 1997 down from 6.9% in Inflation has been moderate. The diverse Tunisian economy hinges on tourism, mining, agriculture, manufacturing, and more recently, investments in new technologies. The main Tunisia Association Agreement with the European Union (EU) entered into force on 1 March,1998, the first accord of such kind between the EU and the Mediterranean countries to be activated. Under this agreement, Tunisia will gradually remove barriers to trade with the EU over the next decade. Further privatization, increase in foreign investment and improvements in government efficiency, are among the challenges for the future 1. Despite these main reforms, banks and other financial institutions are still playing a crucial role in the Tunisian economy and predominate the stock market (see Appendix 1 for the components of the TUNINDEX index) role as the principal sources of financing. Thus, the primary source of financing for firms is still debt rather than stock issuing via capital markets. The stock market in Tunis was opened in February 1969, to accompany the liberalization of the national economy. In 1986, the market capitalization barely represented 1% of the Gross Domestic Product (GDP). Nearly 40 years later, the Stock Exchange of Tunis (BVMT) has certainly become mature, but it is still fairly representative of the Tunisian companies because the contribution of banks and other financial institutions as the principal source of funds. In 2009, 54 companies were listed for a market capitalization of 8.7 bn Dinars ( 4.75 bn). This represents only 16% of the GDP. But after independence in 1956, the Tunisian authorities have adopted the accounting system of 1968 for approximately three decades ( ) which gives preference to the macro-economic tendency of the authorities. Then, in 1997, the country shifted to the new conceptual framework, 1 Available at 2 The IUP Journal of Accounting Research & Audit Practices, Vol. XI, No. 4, 2012

3 Author pls. clarify are these dates? the Social Cognitive Theory (SCT) which is largely influenced not only by the US Generally Accepted Accounting Principles (GAAP) and the International Accounting Standards (IAS/ IFRS), but also depends largely on taxation. Such a system is sometimes called hybrid and gives preference to the market-based economic reform launched in the late 1980s. The Association of Chartered Accountants in Tunisia (OECT) was created in It is under the tutorship of the Ministry of Finance, and gathers professionals authorized to practice accounting and auditing job. The organization of the job of the accountant is governed by law of 18/8/88 and the Decree no of 25/5/89. As a result, factors affecting the disclosure policy for US firms and other accounting systems, including emerging economies,are not the same for the Tunisian companies. In my knowledge, this study is unique for using a wide range of voluntary disclosure items in the Tunisian context. Therefore, the following questions are to be answered: (1) What are the determinants of voluntary disclosure practices adopted by Tunisian listed companies? (2) And to what extent the disclosure policy in Tunisia differs from not only US firms but also companies in other emerging economies? The paper contributes to literature related to corporate disclosure policy in the Tunisian context, especially after the adoption of the Tunisian conceptual framework of Voluntary Disclosure of Financial Information: A Literature Review Voluntary disclosure of accounting information has great importance in the developed markets and economies, especially in the context of efficient market hypothesis. Research in this area takes concerns with incentives (Chow and Boren, 1987; Cook, 1992; Welker, 1995; and El-Gazzar et al., 2002); extent or measures (Botosan, 1997; and Sengupta, 1998); and consequences (Leuz and Verrechia, 2000; Miller, 2002; and Lang and Lundholm, 2003). For example, in the context of online disclosure, Da-Silva and Christensen (2005) argued that interest in voluntary disclosure through the Internet is growing throughout the world. For instance, the EDGAR database was introduced in the US in 1996, and the SEDAR database became the standard for Canadian firms since While some researches used a general disclosure index (Botosan, 1997), most of the studies have concentrated on the disclosure of specific information. For example, Morris and Shin (2002) found that the type of product market competition affects the level of voluntary disclosure over and above what the previous literature had documented as the firms external financing need, i.e., firms engaged in capacity competition disclose relatively more information than those firms engaged in price competition. Simpson (2008) estimated proxies for the proprietary costs and valuation benefits of advertising with data from the mandatory disclosure, and found that firms which experienced high proprietary costs (valuation benefits) from advertising during the mandatory disclosure period are less (more) likely to disclose their advertising expenditures in the discretionary disclosure period. Lim et al. (2007) examined the association between the board composition and different types of voluntary disclosure for 181 Australian companies. Based on 67 items and a global voluntary disclosure index, their results show a positive association between the board composition and voluntary disclosure in their annual reports. Also, independent board of directors provide more voluntary disclosure of forward looking information and strategic information. Entwistle et al. (2005) compared the pro forma earnings voluntary disclosure between the US and Canadian firms. The Determinants of Corporate Voluntary Disclosure: Evidence from the Tunisian Capital Market 3

4 Results show that US managers disclose pro forma earnings more frequently, place greater interest to non-gaap earnings measure when assessing firm performance, and manage earnings upward more than Canadian firms in calculating pro forma earnings. Chen et al. (2001) studied voluntary disclosure of balance sheet information in quarterly earnings announcements. Their results show that balance sheet disclosures are more likely among firms in high technology industries, reporting losses, with larger forecast errors, engaging in mergers or acquisitions and with more volatile stock returns. Some other researches evaluate voluntary disclosure in specific contexts. For example, in the context of earnings management, Kasznik (1999) investigated whether managers issuing annual earnings forecast, manage their reported earnings towards their forecasts. Results show that managers use positive discretionary accruals to manage earnings upwards when reported earnings are below management forecasts. Another example is provided by Lang and Lundholm (2003), around seasoned equity offering, i.e., voluntary disclosure grows dramatically six months before the offering, particularly for the categories of disclosure over which firms have the most discretion. Bertomeu et al. (2011) analyzed the impact of voluntary disclosure on the cost of capital. Results show that while forecasts of bad news experiences a significant increase in the cost of capital for the month after their disclosure, for good news, the cost of capital declines. But the reaction to bad news is more important in terms of absolute value. Al-Akra et al. (2010) investigated the impact of privatization on the extent of corporate voluntary disclosure in Jordan. It was found that accounting regulation reforms and foreign investments accompanying privatization had a significant impact on the levels of accounting disclosure in Jordan. Specifically, company size and industry sector are positively related to voluntary disclosure practices, whereas, liquidity and auditor type are negatively associated with voluntary disclosure. A review of literature helps us to investigate factors that may influence the voluntary disclosure practice of the Tunisian listed companies. Literature Review The literature on corporate policy decisions indicates that ownership structure is a significant determinant of voluntary disclosure decisions by firms. Results concerning this evidence are mixed. For instance, El-Gazzar (1998) argued that a higher concentration of institutional ownership in a particular firm is likely to motivate management to provide additional voluntary disclosures to maintain investor confidence. By contrast, the Agency theory developed by Jensen and Meckling (1976) suggests that the higher the ownership held by managers, the lower the divergence between their interests with the outsiders and the lower the need to offer voluntary information above that which is mandatory. In the same vein, Dal-Ri and Dos (2010) argued that companies with more dispersed ownership disclose more voluntary information in order to attenuate agency conflicts between insiders and outsiders. Makhija and Patton (2004) found a positive relationship between managerial ownership and voluntary disclosure. In fact, firms owned by a high number of investors should disclose more information in order to reduce information asymmetry. Welker (1995) predicted an inverse relation between the firm disclosure policy and the bid-ask spread as a proxy to information asymmetry during periods in which no disclosures are made. Bushee and Noe (2000) showed that institutional investors are attracted to firms with forthcoming disclosure. We predict that dispersed ownership 4 The IUP Journal of Accounting Research & Audit Practices, Vol. XI, No. 4, 2012

5 pls. clarify is considered as an incentive to enhance the level of voluntary disclosure by managers. As argued by Bova et al. (2010), firms whose non-manager employees have strong bargaining power provide less voluntary disclosure, whereas firms whose non-manager employees have employee stock ownership plans provide greater voluntary disclosure. Furthermore, the effect of employee ownership in generating better disclosure is particularly strong, the greater employees negotiation leverage (Bova et al. 2010, p. 1). As the Tunisian economy is growing and tends to shift to a market-based orientation, the role played by financial institutions must be reduced in favour of external investors and capital markets which create information asymmetry and diffused ownership. Therefore, Tunisian firms must disclose more voluntary information. The first hypothesis is stated as follows: H 1 : The higher the diffused ownership, the higher the information asymmetry and the higher the extent of voluntary disclosure. As argued by Lang and Lundholm (2003), a relationship between disclosure and firm size is expected if the disclosure cost is decreasing in firm size. An alternative reason is that disclosure increases with firm size because the incentives for private information acquisition are greater for larger firms (the transaction cost hypothesis) where the profit to trading on private information is higher. Further, consistent with the legal cost hypothesis, disclosure may increase with firm size because the dollar value of damages in securities litigation is a function of firm size. Using a sample of 52 listed Mexican firms and 24 disclosure items, Chow and Boren (1987) argued that larger firms disclose more voluntary data than smaller ones. This finding is based on the assertion that the proportion of outside capital tends to be higher for larger firms. So, the potential benefits from shareholder-manger contracting, including the extent of financial disclosure, increase with firm size (see Leftwich et al.,1981). Larger firms enjoy more resources and expertise than smaller ones. However, there is no reason for smaller firms to disclose additional information when they assume that the costs incurred will not be recovered. Several other studies, including Botosan (1997), Carcello et al. (2002) and El-Gazzar et al. (2002), documented a positive relationship between firm size and the amount of voluntary information that should be disclosed. Finally, Dal-Ri and Dos (2010), suggested that companies with higher visibility tend to disclose more information to improve corporate image. So, the second hypothesis is stated as follows: H 2 : Larger firms are likely to disclose more voluntary information than smaller ones. Jensen and Meckling (1976) suggested that agency costs are higher for firms with more debt in their capital structure. So, it will be useful for such firms to disclose more voluntary accounting data in order to reduce information asymmetry and monitoring costs (Ahmed and Courtis, 1999). Welker (1995) documented a negative association between financial analysts disclosures and the bid-ask spread as a component of the cost of equity capital. Botosan (1997) found a negative association between the disclosure measure and the cost of equity capital for firms with lower analyst following, whereas the findings are not similar for firms with higher analyst following. Based on the assumption that lenders and underwriters consider a firm s disclosure policy in their estimate of default risk, Sengupta (1998) argued that a timely and detailed disclosure reduces lenders and underwriters default risk for the disclosing firm, thereby reducing The Determinants of Corporate Voluntary Disclosure: Evidence from the Tunisian Capital Market 5

6 its cost of debt. He used two proxies for the cost of debt the yield to maturity on new debt issues, and the total interest cost of new debt issues. The results indicate that both measures of cost of debt are negatively associated with the disclosure measure. Ho and Wong (2001) and Chau and Gray (2002) stipulated that when the firm is closed to debt covenants, it discloses more optimistic information in order to signal future positive perspectives for both investors and debtors. So, the level of disclosure is positively related to the leverage ratio. El-Gazzar et al. (2006) argued that senior management disclosure decision, Report of Management's Responsibility (RMR), is significantly and positively related to firm size. By contrast, Chow and Boren (1987) find no relationship between leverage and the extent of voluntary disclosure for Mexican firms. Also, Bertomeu et al. (2011) predicted a negative association between firms cost of capital and the extent of information disclosed. However, more expensive voluntary disclosure does not cause firms cost of capital to decline. Dahawy (2009, p.199) argued that companies with higher levels of debt financing will be subject to more scrutiny than firms that mainly depend on equity financing. Consequently, it is predicted that higher leveraged firms tend to disclose more information in order to lower different costs of debt. H 3 : Higher leveraged firms are likely to disclose more voluntary information than lower leveraged firms. Audit quality plays a central role in the accounting process in general, and in disclosure policy in particular. In fact, the choice of a Big Four auditor by the firm is considered as a positive signal by the capital market, as the cost incurred by such firms is always high. Firms who engage a Big Four auditor have sufficient resources and are presumed to have more optimistic future perspectives than other firms. Results connecting the level of disclosure to the size of auditor have been found to be mixed (Dahawy 2009, p.199). Ahmed and Nicholls (1994) found a significant relationship between auditor quality and the level of disclosure. However, Naser (1998) does not find such a relation. As argued by Dal-Ri and Dos (2010), auditors incur in-costs when entering contracts with clients. In this sense, they tend to influence them to disclose as much information as possible in order to reduce the possible litigation costs due to the absence of information. El-Gazzar et al. (2006) investigated the relationship between the senior management disclosure decisions and two proxies of audit quality: the percentage of independent audit committee members and the number of audit committee meetings during the year. Results show that both variables are positive and significant. This confirms the influence of monitoring on the firm disclosure policy. Carcello et al. (2002) find that Big Four audited firms disclose more voluntary information than other firms. Dal-Ri and Dos (2010) concluded that Brazilian public companies audited by a Big Four auditor tend to disclose more information than companies audited by other auditing firms. Ahmed and Courtis (1999) investigated the relationship between the size of the audit firm and the extent of disclosure. They find a positive relationship between the audit firm and mandatory disclosure. However, there is no relationship between the size of the audit firm and voluntary disclosure. In the sample of this study, approximately 30% of the firms are audited by a Big Four auditor. Such firms must justify the choice of such costly auditing firms by disclosing more voluntary 6 The IUP Journal of Accounting Research & Audit Practices, Vol. XI, No. 4, 2012

7 information. Therefore, a positive relationship is expected between audit quality and the level of voluntary disclosure. In contrast, firms with poor future financial perspectives have no interest to be certified by a Big Four auditor. H 4 : Firms audited by Big Four auditors have enough resources and opportunities to disclose more voluntary information than other firms. Dahawy (2009) found a negative relationship between the financial sector and the level of disclosure quality due to the opacity of such sectors. In contrast, the listed companies on the BVMT are essentially banks and other financial institutions such as insurance firms and leasing companies. These companies always hold higher diffused ownership in other companies and require more disclosure of information in order to reduce information asymmetry. Carcello et al. (2002) documented that depository institutions (SIC 60) are more likely to make disclosure in the charter regarding private meetings with internal audit, given the highly regulated nature of the banking industry. This result is based on a comparison with three other sectors (chemicals and allied products (SIC 28), holding other investment offices (SIC 67) and business services (SIC 73)). El-Gazzar et al. (2006) argued that firms with greater percentage of equity owned by institutional investors disclose more voluntary information. They added that this result is consistent with the expectation that institutional investors exercise greater monitoring over firms, causing management to expand the disclosure for stockholders. Thus, the higher the presence of institutional investors, the greater the extent of voluntary disclosures adopted by firms. Dahawy (2009) argued that Egyptian banks have a higher level of disclosure than other companies since banks in Egypt are subject to additional disclosure requirements set by the Egyptian Central Bank. In the Brazilian context, Dal-Ri and Dos (2010) argued that firms pertaining to the electrical sector are more regulated than other firms and have strong incentives to disclose the social balance. Turning to the Tunisian context, banks and other financial institutions are governed by special regulated accounting rules (their activities and transactions are supervised by the BCT 2 ) and more severe internal control rules than other firms. Such firms represent approximately 21.5% of the TUNINDEX (see Appendix 1). Specifically, the study tests the following hypothesis: H 5 : Firms pertaining to higher regulated sectors disclose more voluntary information than other firms. Based on the signalling theory, Subramanyam (1996) argued that firms expecting future positive financial perspectives have stronger incentives to disclose more discretionary information than firms expecting poor financial perspectives. Specifically, managers disclose future positive discretionary accrual levels along with changes in the dividend policy. The disclosure of such information reduces information asymmetry between managers and outsiders along with agency costs. Ahmed and Courtis (1999) argued that higher profitability motivates managers to provide greater information because it increases the investors confidence, which in turn, increases management compensation. Dal-Ri and Dos (2010) argued that both accounting and capital market performances are positively associated with higher levels of voluntary disclosure. 2 BCT: Banque Centrale de Tunis. The Determinants of Corporate Voluntary Disclosure: Evidence from the Tunisian Capital Market 7

8 Similarly, a positive association between firm profitability and the level of voluntary disclosure is expected. H 6 : Firms with higher levels of profitability disclose more voluntary information than firms with lower levels of profitability. Methodology Sample The sample consists of 52 listed companies on the BVMT, pertaining to various industrial sectors 3. The period under study is from 2009 to This period was chosen in order to focus on the cross-sectional variation of voluntary disclosure determinants rather than their time-series variation. Botosan (1997) used one year study because disclosure policy appears to be relatively constant during that time and the study was limited to one industrial sector because different industries display numerous patterns of disclosure. For example, firms in the pharmaceutical industry tend to provide much more disclosure about their research and development (R&D) activities than do firms in other industries (Botosan, 1997). Computing a Disclosure Index 5 Empirical evidence on voluntary disclosure index offers several measures established by both professionals such as financial analysts and accounting committees. For example, Welker (1995), Lang and Lundholm (1996) and Sengupta (1998) used a set of corporate disclosure ranking produced by the Association for Investment Management and Research (AIMR) 6. El-Gazzar et al. (2006) used the RMR to represent the senior management s disclosure decision. However, as suggested by Botosan (1997), the AIMR sample tends to be limited to the largest most heavily followed firms in an industry. Since the disclosure levels are positively correlated with the firm size, AIMR firms are unlikely to display sufficient cross-sectional variation in disclosure levels. Based on Berton s (1994) work, Botosan (1997) computed a disclosure index with five fields disclosed by firms in their annual reports to shareholders 7. Then, the score was weighted based on the information accuracy within each item. Following Botosan (1997), Botosan and Harris (2000) used a two-step procedure to compute a simple disclosure index. 3 There are 14 industrial sectors in the BVMT and 55 listed companies (please see: companies/?view=listed). In our study, we use all firms for which accounting data are available. Only three firms have no sufficient data. So, we eliminate them from our study. For voluntary disclosure items, we had collected them directly from the CMF (Financial Market Council) official bulletin. 4 These two years correspond to the most recent period for which both accounting data and voluntary disclosure items are available. We use only two years because we are interested in the cross-sectional determinants of the voluntary disclosure practice. The choice of only two years is consistent with prior works of both Nalikka (2009) and Dal-Ri and Dos (2010), who used a three-year period. 5 There are many other measures for disclosure quality. For example, Chen et al. (2007) use conference calls as a proxy for voluntary disclosure. 6 Disclosure quality is provided by the annual volumes of the report of Financial Analysts Federation (FAF) branch of the Association for Investment Management and Research (AIMR). 7 Items included in Botosan s (1997) score reflect five categories of voluntary information: background information, summary of historical results, key non-financial statistics, projected information, and Management Discussion and Analysis (MD&A). 8 The IUP Journal of Accounting Research & Audit Practices, Vol. XI, No. 4, 2012

9 1. Counting the number of items considered as voluntary by Tunisian firms and then, the list of relevant information within each item. 2. Assigning the value of one if an item from the list is disclosed by the firm and zero otherwise. A disclosure score for a given firm is weighted by the total number of the relevant items. Finally, the disclosure index is computed as follows: DSCORE nj i 1 n j x ij where n j is the maximum number of items that should be disclosed by firm j. 45 voluntary items related to the eight disclosure areas were selected, where n j ranges from 0 to 45. The choice of these items is motivated by two reasons: (1) A wide review of the existing literature on voluntary disclosure items; and (2) from a careful examination of the firms activity reports. X 1 if the i th item is disclosed by firm j for the period under study and 0 otherwise, where ij 0 DSCORE 1, DSCOREacts as a continuous variable that takes the value of 1 as a maximum. Botosan (1997) conducted a set of validity checks for disclosure index and concluded that extending her index by the number of Wall Street Journal articles or the number of analysts might provide insights into the disclosure index alone. Because of the lack of these two sets of data, I do not conduct a validity check for my disclosure index. Table 1 displays eight relevant fields that should be disclosed by Tunisian firms, whereas Appendix 2 gives examples of disclosed items within each field. The definition of an item as voluntary is derived from two legal sources: (1) the conceptual framework of financial accounting (Decree n of December 30, 1996); and (2) the SCT (law n of December 30, 1996). We define a voluntary item as a disclosure that is not considered as mandatory both by the two previous legal sources. Barako et al. (2006) first defined voluntary disclosure as discretionary release of financial or non-financial information through annual reports over and above the mandatory requirements, either with regard to the Kenyan company law, professional accounting standards, or any other relevant requirements. Following Barako et al. (2006) who classified the disclosure items into five categories (general and strategic information, financial data, forward looking Table 1: The Number of Relevant Fields That Should Be Disclosed Fields Number of Items % Sector Data 2 4 Notes to the Balance Sheet Statement 3 6 Notes to the Income Statement 3 6 Share Data 1 2 Financial Market Data The Determinants of Corporate Voluntary Disclosure: Evidence from the Tunisian Capital Market 9

10 disclosure, corporate social disclosure, board and senior management information), I identify eight voluntary disclosure fields summarized in Table 1. Table 2 displays both the endogenous and the independent variables derived from our disclosure hypotheses. The exogenous variables include: firm size, reputation of the auditor, firm leverage, ownership structure, industrial sector and firm profitability. While my objective is to check the impact of several variables on voluntary disclosure policy adopted by Tunisian firms in their annual reports, I run a multivariate regression analysis taking the disclosure index as the dependent variable, while firm size, leverage, audit quality, ownership structure and industrial sector act as the independent variables. Model Specification Following Chow and Boren (1987), Botosan (1997), El-Gazzar et al (2002), I regress the ratio of a firm i disclosure index for a year t on a number of exogenous determinants discussed above. DSCORE t Table 1 (Cont.) Fields Number of Items % Forecast Data 5 10 Data About Firm Product and Concurrence General Non-Financial Information Total Table 2: Variables Definition Variable Label Description Voluntary Disclosure Index DSCORE Number of items that are voluntarily disclosed by a given firm divided by the total number of relevant items that should be disclosed (Botosan 1997). Firm Size SIZE Natural logarithm of total assets.. Leverage LEV Long-term debt divided by capital equity. Audit Quality AUD A dummy variable which equals 1 if the firm is audited by a Big Four auditor, and 0 otherwise. Ownership Structure 8 OWNER A dummy variable which equals 1 if none of the investors maintains more than 50% of the voting rights, and at least 10 shareholders, and 0 otherwise. Industrial Sector SECTOR A dummy variable which equals 1 if the firm pertains to the financial sector, and 0 otherwise. Profitability ROA Net income divided by total assets to proxy for firm performance. 1 β0 β1 SIZE1 β2lev1 β3aud1 β4owner1 β5sector1 β6roa ε 8 There are other measures for diffused ownership. We proxy for this measure because Tunisian companies do not disclose enough information about ownership structure. 10 The IUP Journal of Accounting Research & Audit Practices, Vol. XI, No. 4, 2012

11 where DSCORE is the index of disclosure quality at year t. Such index is computed as follows: t Disclosure Index Number of items that are voluntarily disclosed by a given firm The total number of relevant items that should be disclosed All other variables are defined above. Statistical problems such as residual autocorrelations and heteroscedasticity among independent variables were resolved and do not bias our results. Discussion of the Results Descriptive Statistics Table 3 displays descriptive statistics for the explanatory variables. As shown, the mean of leverage ratio is about 82%, suggesting that our sample firms are highly leveraged. The principal owner of the firm equity capital held approximately 61% of the firm s total equity, suggesting a higher concentrated ownership among our sample firms. In fact, banks and other institutional investors (leasing, mutual funds, etc.) represent the principal financing source for Tunisian Table 3: Descriptive Statistics Variable Statistic Mean STD Error STDV Variance MAX MIN Size Leverage Owner Auditor ROA DSCORE companies. Approximately, 35% of our sample firms are audited by a Big Four auditor. Finally, the mean of the disclosed items by our sample firms is about over the 45 items. Table 4 displays the Pearson correlation coefficients among the explanatory variables. The greater correlation coefficient is of about (sector and ownership structure), but not Table 4: Correlation Matrix DSCORE Sector Size Leverage Auditor Owner Profitability DSCORE Sector Size Leverage Auditor Owner Profitability The Determinants of Corporate Voluntary Disclosure: Evidence from the Tunisian Capital Market 11

12 significant at conventional levels. Thus, the multi-colinearity problem among the exogenous variables is attenuated. Results of the Multivariate Regression Contrary to my expectations, diffused ownership does not affect the level of voluntary disclosure (the coefficient is 0.124, while t=1.058 and p=0.297) (Table 5). This result does not conform Table 5: Estimated Coefficients Hypothesis Explanatory Coefficient Student test Sig Result variables ( ) (t) (p) H 1 Ownership Structure Rejected H 2 Firm Size Rejected H 3 Leverage Confirmed H 4 Audit Quality Confirmed H 5 Sector Confirmed H 6 Profitability Confirmed with the previous researches (Haniffa and Cooke, 2000; Chau and Gray, 2002; and Makhija and Patton, 2004). In fact, Tunisian listed companies have concentrated ownership and a familiar character (See also Chen et al., 2007 for family firms). Although Tunisian economy is shifting to a market-based perspective, outside investors still do not play a central role in providing funds to firms, as do banks and other institutions (see the introduction for details). Thus, firms capital equity is concentrated on a few institutional investors who make voluntary disclosure useless, while the information asymmetry problem is thereby resolved. My first hypothesis suggesting that firms with higher diffused ownership disclose more voluntary information is therefore rejected. As ownership structure and firm size do not affect the disclosure level (the coefficient is 0.027, t=0.255 and p=0.8). This result, contrary to many other researches in the US capital markets, can also be explained as follows: Almost 80% of Tunisian listed companies are Small and Medium Enterprises (SMEs) with a familiar character. Although listed, Tunisian companies also have homogenous size, which in turn does not affect the endogenous variable (see Chen et al., 2007). Small firms have not enough financial and human resources to adopt a well-developed disclosure policy. Again, my second hypothesis related to the impact of firm size on the voluntary disclosure level is also rejected. Recent literature, including Sengupta (1998), Chau and Gary (2002) and Bertomeu et al. (2011), documented a negative relationship between voluntary disclosure and the cost of debt. By contrast, higher leveraged firms should disclose more voluntary information than other 12 The IUP Journal of Accounting Research & Audit Practices, Vol. XI, No. 4, 2012

13 firms in order to give more insurance for investors and creditors. In this case, a higher level of voluntary disclosure is positively priced by investors and creditors. My results show that the coefficient related to leverage ratio is positive and significant (0.277, t=2.907 and p=0.006). This confirms the results by Raffournier (1995), but not those by Chow and Boren (1987). In the Tunisian economy, financial institutions play a central role, which makes debt as the first financing means for a majority of the firms, whereas stock market financing is not sufficiently integrated in firms financing policy. Thus, firms looking for additional funds from banks and other credit institutions have to disclose more voluntary information about their future perspectives. As suggested by the fourth hypothesis, firms audited by a high quality auditor disclose more voluntary information than other firms (the coefficient is 0.47, t=4.988 and p=0.000). The choice of a Big Four auditor is positively priced by investors, and it is viewed as a motive to disclose more detailed information. Although audit quality can be defined under many sides, many studies in earnings management literature (DeFond and Subramanyam, 1998) use the dichotomy (Big Four/ non-big Four) to proxy for auditor quality. The appointment of such auditor is also seen as a guarantor of information quality especially after recent financial scandals. As a result, it will be harmful for a higher quality auditor to certify a firm with poor financial perspectives. The industrial sector has also a positive and significant effect on the disclosure level (coefficient is 0.443, t=3.668 and p=0.001). In fact, financial companies disclose more information than other firms. Such result can be explained by the fact that the most listed companies in our sample pertain to the financial sector 9. Also, the disclosure culture is more developed for financial institutions than other firms as they pertain to a higher regulated sector. Thus, my fifth hypothesis is also confirmed. Finally, and as expected, the coefficient related to the firm profitability (ROA) is positive (0.290) and significant at conventional levels (t=1.966 and p=0.054), indicating that higher levels of disclosure are associated with higher firm profitability. This confirms the findings of Dal-Ri and Dos (2010) for Brazilian companies. In fact, firms disclosing more voluntary information get more investors confidence and positive share price variation. A widespread disclosure policy contributes to the reduction of information asymmetry and agency costs. Conclusion The literature related to voluntary disclosure argued that factors affecting such financial policy may differ from countries and financial markets. While much of the existing literature focuses on the US, little evidence is provided about the emerging economies. The aim of this paper is to check for factors affecting firm disclosure policy adopted by a sample of listed Tunisian companies. Computing a disclosure index (Botosan, 1997) and using a multivariate regression analysis, I find that factors affecting the disclosure level are financial sector, leverage ratio, audit quality and profitability. By contrast, both ownership structure and firm size have no effect on disclosure level. This result can be explained as follows: 1. Tunisian firms are of homogenous size, which in turn does not affect the endogenous variable. 2. Although listed, many firms have a family character with a concentrated ownership structure. 9 Approximately 45% of the sample consists of companies pertaining to the financial sector. The Determinants of Corporate Voluntary Disclosure: Evidence from the Tunisian Capital Market 13

14 3. Although Tunisian authorities have shifted to capital market-based economy, banks and other financial institutions are still playing a crucial role in financing firms. Implications of the Study: The study has many implications. 1. It can help investors in their decision making and leads to fair allocation of resources. 2. It helps those who prepare financial information to enhance financial reporting quality. 3. Accounting standard setters should take into account factors affecting firms voluntary disclosure policy when issuing new standards leading to more transparency of financial information. Limitations of the Study: The present study includes some limitations. 1. All voluntary disclosure items such as the characteristics of the firm s board of directors, human resources and environmental data disclosed by Tunisian firms 10 are not taken into account. 2. Simple and general disclosure index was used instead of other more specific and weighted index for each industry, year or firm characteristics. 3. The lack of voluntary disclosure is a widespread feature among the population of the firms. This limitation will be of less significance with the mandatory shift to the International Accounting Standards IAS/IFRS in 2014, which in turn enhances more disclosure transparency. Other voluntary disclosure determinants remain unexplored for Tunisian firms such as culture and tax incentives. Future research should focus on specific disclosure items (Entwistle et al., 2005). Then, it would be useful to check the stock market reaction in response to voluntary information disclosed by the Tunisian firms. Bibliography 1. Ahmed K and Courtis J (1999), Association between Corporate Characteristics and Disclosure Levels in Annual Reports: A Meta Analysis, British Accounting Review, Vol. 31, No. 1, pp Ahmed K and Nicholls D (1994), The Impact of Non-Financial Company Characteristics on Mandatory Disclosure Compliance in Developing Countries: The Case of Bangladesh, The International Journal of Accounting, Vol. 29, pp Al-Akra M, Eddie I A and Ali M J (2010), The Association between Privatization and Voluntary Disclosure: Evidence from Jordan, Accounting and Business Research, Vol. 40, No. 1, pp Paragraph 83 of the conceptual framework of financial accounting for Tunisian firms (SCT, 1997) suggests separate statements for: (1) Human resources data; (2) Environmental performances; and (3) Technology innovations. (Decree n for the December 30 th 1996 related to the conceptual framework of financial accounting). Despite these recommendations, very few firms disclose such kind of information. 14 The IUP Journal of Accounting Research & Audit Practices, Vol. XI, No. 4, 2012

15 Author pls. chk. Underlined refs are not cited in the text. 4. Barako D G, Hancock P and Izan H Y (2006), Factors Affecting Voluntary Corporate Disclosure by Kenyan Companies, Corporate Governance an International Review, Vol. 14, No. 2, pp Bertomeu J, Beyer A and Dye R A (2011), Capital Structure, Cost of Capital and Voluntary Disclosure, The Accounting Review, Vol. 86, No. 3, pp Berton L (1994), Companies Pressure Accounting Panel to Modify Demands for More Data, Wall Street Journal, August 16, A Botosan C A (1997), Disclosure Level and the Cost of Equity Capital, The Accounting Review, Vol. 71, No. 3, pp Botosan C A and Harris M S (2000), Motivations for Change in Disclosure Frequency and its Consequences: An Examination of Voluntary Quarterly Segment Disclosure, The Accounting Review, Vol. 38, No. 2, pp Bova F, Dou Y and Hope O K (2010), Employee Ownership and Firm Disclosure, Working Paper, University of Toronto, available at: Bushee B J and Noe C P (2000), Corporate Disclosure Practices, Institutional Investors and Stock Return Volatility, Journal of Accounting Research, Vol. 38, Supplement, pp Carcello J V, Hermanson D R and Neal T L (2002), Disclosures in audit Committee Charters and Reports, Accounting Horizons, Vol. 16, No. 4, pp Chau G K and Gray S J (2002), Ownership Structure and Corporate Voluntary Disclosure in Hong Kong and Singapore, The International Journal of Accounting, Vol. 37, No. 2, pp Chen F, Hope O K, Li Q and Wang X (2010), Financial Reporting Quality and Investment Efficiency of Private Firms in Emerging Markets, Working Paper, University of Toronto, available at: Chen S, Defond M L and Park W (2001), Voluntary Disclosure of Balance Sheet Information in Quarterly Earnings Announcement, Working Paper, University of Washington, available at: Chen S, Chen X and Cheng Q (2007), Do Family Firms Provide More or Less Voluntary Disclosure, Journal of Accounting Research, Vol. 46, No. 3, pp Chow C and Boren W A (1987), Voluntary Disclosure by Mexican Corporations, The Accounting Review, Vol. 62, No. 3, pp Clarkson P, Kao J and Richardson G D (1999), "Evidence that Management Discussion and Analysis (MD&A) is a Part of a Firm's Overall Disclosure Package", Contemporary Accounting Research, Spring, pp Cook T E (1992), The Effect of Size, Stock Market Listing and Industry Type on Disclosure in the Annual Reports of Japanese Listed Corporations, Accounting and Business Research, Vol. 22 No. 87, pp The Determinants of Corporate Voluntary Disclosure: Evidence from the Tunisian Capital Market 15

16 19. Dahawy K (2009), Company Characteristics and Disclosure Level: The Egyptian Story, International Research Journal of Finance and Economics, Vol. 34, December, pp Dal-Ri M F and Dos S A (2010), Determinants of Corporate Voluntary Disclosure in Brazil, Working Paper, Sao Paulo University, available at: Da-Silva M and Christensen T E (2005), Determinants of Voluntary Disclosure of Financial Information on the Internet by Brazilian Firms, Working Paper, Brigham Young University, available at: DeFond M L and Subramanyam K R (1998), Auditor Changes and Discretionary Accruals, Journal of Accounting and Economics, Vol. 25, No. 1, pp El-Gazzar S M (1998), Pre-disclosure Information and Institutional Ownership: A Cross Sectional Examination of Market Revaluations During Earnings Announcement Periods, The Accounting Review, Vol. 73, No. 1, pp El-Gazzar S M, Fornado J M and Jacob R A (2006), An Examination of the Determinants and Contents of Corporate Voluntary Disclosure of Management s Responsibilities for Financial Reporting, Faculty Working Papers No. 56, available at pace.edu/lubinfaculty_workingpapers/ Entwistle G, Feltham G and Mbagwu C (2005) The Voluntary Disclosure of Pro Forma Earnings: A US-Canada Comparison, Journal of International Accounting Research, Vol. 4, No. 3, pp Hail L (2002), The Impact of Voluntary Corporate Disclosures on the ex ante Cost of Capital for Swiss Firms, The European Accounting Review, Vol. 11, No. 4, pp Haniffa R M and Cooke T E (2002), Culture Corporate Governance and Disclosure in Malaysian Corporations, Abacus, Vol. 38, No. 3, pp Healy P M and Palepu K G (2001), Information Asymmetry, Corporate Disclosure and the Capital Markets: A Review of the Empirical Disclosure Literature, Journal of Accounting and Economics, Vol. 31, Nos. 1-3, pp Ho S M and Wong K S (2001), A Study of the Relationship between Corporate Governance Structures and the Extent of Voluntary Disclosure, Journal of International Accounting, Auditing and Taxation, Vol. 10, No. 2, pp Jensen M and Meckling W (1976), Theory of the Firm: Managerial Behaviour, Agencycosts and Ownership Structure, Journal of Financial Economics, Vol. 3, No. 4, pp Kasznik R (1999), On the Association between Voluntary Disclosure and Earnings Management, Journal of Accounting Research, Vol. 37, No. 1, pp Lang M H and Lundholm R J (1996), Corporate Disclosure Policy and Analyst Behaviour, The Accounting Review, Vol. 71, No. 4, pp Lang M H and Lundholm R J (2003), Voluntary Disclosure and Equity Offering: Reducing Information Asymmetry or Hyping the Stock, Working Paper, University of North Carolina, available at: Author pls. chk. Underlined refs are not cited in the text. 16 The IUP Journal of Accounting Research & Audit Practices, Vol. XI, No. 4, 2012