International Research Journal of Finance and Economics ISSN 1450-2887 Issue 42 (2010) © EuroJournals Publishing, Inc. 2010 http://www.eurojournals.com/finance.htm

To Pay or Not to Pay: Using Emerging Panel Data to Identify Factors Influencing Corporate Dividend Payout Decisions Duha Al-Kuwari Department of Economics and Finance, University of Qatar, P.O. Box: 2713, Doha, Qatar E-mail: [email protected] Tel: +974-4851868; Fax: +9744851564 Abstract In this study, the corporate dividend payout decisions of firms listed on the emerging Gulf Cooperation Council (GCC) states’ stock market were analysed. Panel data were obtained from 191 non-financial firms over the five year period from 1999 to 2003 and seven research hypotheses were tested. Random effects probit models were used to test the seven research hypotheses. The results revealed that government ownership, company profitability, and company size increases the probability of paying dividends, whereas the good investment opportunities decrease the probability of paying dividends. Overall, research findings indicated that companies listed on the GCC stock exchange paid dividends to reduce agency conflict, avoid exploiting minority shareholders, and enhance their company's reputation.

Keywords: Dividend payout decision, Emerging market, Non-financial firms, Panel data, Random effects probit model. JEL Classification Codes: JEL: G35 - Payout Policy

1. Introduction Since Miller and Modigliani (MM) presented their dividend irrelevance theory (DIT) in 1961, dividend policy has become one of the most puzzling matters in financial economics. MM’s irrelevance theory relies on the basis that all dividend decisions are equal. According to this theory, firm value and shareholder wealth are not related to the decision of whether or not a dividend should be paid. However, many researchers have argued that MM’s DIT was based on a perfect capital market; whereas in reality, the market is imperfect. The opponents to MM’s theory have developed competing theories, providing evidence that dividends are relevant, leaving financial researchers with a yet unsolved puzzle. To better explain this unsolved puzzle, several studies have illustrated that disputable dividends result from the bird-in-the-hand theory, arguing that investors value cash dividends (bird-in-the-hand) over capital gains (bird-on-the-bush), thus requiring firms to set a high dividend payout ratio to maximize firm share price (Gordon, 1956; Lintner, 1956; Fisher, 1961; and Brigham and Gordon, 1968). Conversely, the tax-effect hypothesis argues that investors prefer capital gains over the payment of higher dividends due to the tax effect, therefore, firms set a low payout policy to maximize their share price (e.g., Brennan, 1970; Elton and Gruber, 1970; Litzenberger and Ramaswamy, 1979; Litzenberger and Ramaswamy, 1982; Kalay, 1982; John and Williams, 1985; Poterba and Summers, 1984; Miller and Rock, 1985; Ambarish et al., 1987). Electronic copy available at: http://ssrn.com/abstract=1788944

International Research Journal of Finance and Economics - Issue 42 (2010)

20

Yet another theory, signalling theory, suggests that managers receive information about future profitability unavailable to outside shareholders. Implementing an announcement of dividends releases private information about a firm’s future performance, thus triggering an uneven asymmetrical response in shareholder’s investments (Aharony and Swary, 1980; Asquith and Mullins, 1986; Kalay and Loewenstein, 1985; Healy and Palepu, 1988). On the other hand, residual theory indicates that firms incurring large transaction costs will be required to reduce dividend payouts to avoid the costs of external financing (Mueller, 1967; Higgins, 1972; Crutchley and Hansen, 1989; Alli et al., 1993; Holder et al., 1998). A different explanation indicates the relevance of dividend policy, due to its tendency to reduce the effect of agency costs. Here, we find that high dividend payouts reduce internal cash flows, which forces managers to seek external financing, thereby making them liable to capital suppliers, and reducing overall agency costs (Rozeff, 1982; Easterbrook, 1984; Crutchley and Hansen., 1989; Dempsey and Laber, 1992; Alli et al., 1993; Moh’d et al., 1995; Glen et al., 1995; Holder et al., 1998; Saxena, 1999). Ultimately, the dividend policy issue and the actual motivation behind paying dividends remain unclear, due to varying hypotheses and controversial explanations. Dividend policy has been analysed for many decades, but no acceptable explanation for the observed dividend behaviour of companies has yet to be established. Black (1976: 5) defined this as the dividend puzzle. The majority of recent studies seem to be in line with Black. Brealey and Myers (2005) who defined dividend policy theory as one of the top ten most difficult and unsolved problems in financial economics research. Frankfurter et al. (2002) stated that forty years have been spent researching dividend policy, and thus far, it has not been resolved. Allen and Mikhaely (2003) suggest that it is important to develop more theories and empirical studies to achieve a general consensus among researchers. Researchers have primarily focused on developed markets; however, a consideration of developing countries may provide additional insight into the dividend policy debate. Dividend policy in the markets of developing countries often differs from the dividend policy already established in the developed markets. Firms in emerging markets that have originated more recently, have less information efficiency, more volatility, and are smaller in size (Kumar,1999). However, relatively little evidence exists for these emerging markets, particularly for the Gulf Cooperation Council (GCC) states. As a result, it is important and worthwhile to examine dividend policy in GCC states to provide new evidence about dividend policy in emerging markets. The GCC states are a group of six Arab oil-exporting countries that founded the GCC in May 1981: Bahrain, Kuwait, Qatar, Saudi Arabia, Oman, and the United Arab Emirates. The primary motive behind the creation of the GCC was to enhance the collective security of its member states in the face of the Iran-Iraq war, which had begun a year earlier. The stated objectives of the GCC constitution also included socio-economic cooperation and eventual economic integration. To achieve this goal, the GCC Economic Agreement of June 1981 (The Economic Agreement between GCC States, 1981) identified several distinct policy objectives including, among others: the creation of a free-trade area; a common market with the complete mobility of labour, financial capital, and commodities; the co-ordination of domestic economic policy; and the adoption of a common foreign trade policy. However, despite this strong initial commitment, after two decades, the pace of economic integration within the GCC slowed, achieving only a few of the stated economic goals as the GCC governments focused more on deficits and security. Due to their inability to achieve economic integration, combined with oil price fluctuation in countries entirely dependent on exporting crude oil, the governments were motivated to transfer part of their economic responsibility to the private sector and to initiate several steps to attract regional and foreign investment. Investment decisions for the company’s stock choices centred on the dividend policy these companies employed; however, the dividend decision was not considered adequate in any of the previous studies.

Electronic copy available at: http://ssrn.com/abstract=1788944

21

International Research Journal of Finance and Economics - Issue 42 (2010)

Furthermore, the case of the GCC presents some interesting characteristics that enforce the relevance of this study in terms of policy recommendations. First, the GCC environment is unique in that tax is not paid on either dividends or capital gains. Second, the stock exchanges in these countries are more volatile and imply a certain degree of information asymmetry, together with an expectation that high agency costs will be incurred. Third, governments own a significant portion of the shares in the listed firms, especially in larger capital firms. However, government participation might also result in a complex setting of agency theory, as government involvement enhances the agency problem while, at the same time, serving management1. As a result of these characteristics and effects, a considerable amount of interest has emerged in the investigation of factors affecting the dividend payout decisions in emerging markets, using a panel dataset of companies listed on the GCC stock exchanges between 1999 and 2003. Exploratory data analysis will be used to answer the question of why firms pay dividends, and to determine how firms make the decision to pay dividends, since some firms pay dividends, while others do not: what characteristics of firms are responsible for the decision to pay, or not pay, dividends? To investigate this research question, seven research hypotheses are presented to discuss the factors affecting the dividend payment decision. The remainder of the paper is organised as follows. Section 2 discusses the factors affecting companies’ dividend decisions, the related literature on the subject, and the study hypotheses. Section 3 describes the data, while Section 4 presents the random-effects probit model used to investigate the research question and the relevant hypotheses. Section 5 reports the results and the associated discussions. The paper ends with a conclusion of the findings and recommendations for future research.

2. Factors Influencing Dividend Decisions Potential factors influencing corporate decisions to pay dividends will be addressed in this section. 2.1. Agency Theory and Government Ownership Modern organisational theory considers a firm to be comprised of shareholders and managers. In 1961, MM assumed that managers and shareholders had a similar objective: to maximize shareholder wealth. However, this has not been found to be true in the real world. Managers may have different objectives than shareholders, allowing them to operate in a way which is suboptimal for shareholders, leading to conflict between the two parties, and managers may obtain the entire benefit at the expense of the shareholders. The agency theory of dividends hypothesis states that dividend payments can be utilised to reduce agency problems (Rozeff, 1982; Easterbrook, 1984; Dempsey and Laber, 1992; Alli et al., 1993). Moh’d et al. (1995), Holder et al. (1998), Saxena (1999) explained that dividends can be a significant instrument in diminishing agency costs since paying high dividends would require a company to depend on capital markets as their key financing resource. Consequently, investment professionals such as bankers, people working for the securities exchange commission, and financial analysts may be able to monitor manager behaviour. Thus, dividend payments increase the scrutiny of management by outsiders and compel managers to disclose new information and reduce agency costs to acquire the necessary funds.

1

Gugler (2003, p.1301) was one of the first people to suggest that government intervention is an extension of the agency problem, where the principal-agent problem will be duplicated. That is, agency problems may appear between citizens and their representatives, the government, who may not play a role in the best interests of its citizens. The conflict might also appear between the government and other managers, where managers often seek to benefit themselves from the resources available within the firm, allowing them to increase their salaries and take advantage at the expense of other shareholders. This explanation was supported by Al-Malawi (2005), who examined dividend policy in the listed firms of the emerging Amman stock exchange.

International Research Journal of Finance and Economics - Issue 42 (2010)

22

Regardless, the dividend payment decision may depend on the ownership structure of the firm. Glen et al. (1995), Naser et al. (2004), and Al-Malkawi (2008), investigated emerging markets and found government ownership to be a key factor affecting dividend payment decisions. Glen et al. (1995) argued that, in developing markets, where legal protection for outside shareholders is weak, investors need to be protected. Government ownership, which may be seen as a safeguard for minority shareholders and serve as an influential investor, may force controlling shareholders to distribute dividends, thereby reducing agency conflict. Naser et al. (2004) that, in young stock exchanges, with no history of good treatment of minority shareholders and high uncertainty about the future cash flows of firms, governments have a strong desire to build a reputation of not exploiting minority shareholders by paying them high dividends. Al-Malkawi (2008) supported Gugler’s argument (2003) that firms with large government ownership may doublecate principal-agent conflict. That is, agency problems may emerge between citizens and their representative; the government, who may not play a role in the best interests of citizens. The conflict might also might appears between the government and other managers, where managers often seek to benefit themselves from the resources available to a firm, allowing them to increase their salaries and take advantage at the expense of other. Hence, high dividend payouts have been suggested to alleviate the doubled agency problem in these firms. By using the percentage of shares held by the government as an indicator of company ownership structure, the first hypothesis becomes: H1: Ceteris paribus, the percentage of government ownership positively affects the probability of paying dividends. 2.2. Free Cash Flow Jensen (1986) argued that in firms with high levels of free cash flow, the possibility of increasing agency conflict arises, because managers may use the free cash flow to benefit themselves at the expense of their shareholders. A number of other studies, including Smith and Watts (1992) and La Porta et al. (2000), supported Jensen’s argument and added that free cash flow allows managers to engage in wasteful projects, even when the protection for investors improves. Jensen (1986), Holder et al. (1998), and La Porta et al. (2000), among others, suggested that companies with a high free cash flow should increase dividend payments to lessen the agency costs of the free cash flow. This leads to the second hypothesis: H2: Ceteris paribus, a company’s free cash flow positively affects the probability of paying dividends. The free cash flow to total assets ratio will be used in this study as a proxy for the free cash flow variable. 2.3. Company Size Jensen and Meckling (1976), Lloyd et al. (1985), and Jensen et al. (1992) found that large companies were more likely to increase their dividend payouts to decrease agency costs and minimise the information asymmetry problem. In other words, large companies have a higher ownership dispersion than small companies. This ownership dispersion decreases the shareholders’ ability to monitor the internal and external activities of the company, resulting in a large information asymmetry. Paying high dividends can be a solution to this problem, because it leads to an increase in the need for external financing, thereby increasing creditor monitoring. Other studies have illustrated a positive association between dividends and company size from a transaction cost prospective, arguing that large companies characteristically have better access to capital markets and a better opportunity to raise financing at a lower cost. As a result, they are less dependent on internal funding and more likely to pay higher dividends to their shareholders (Eddy and Seifert, 1988; Redding, 1997; Holder et al., 1998; Fama and French, 2001).

23

International Research Journal of Finance and Economics - Issue 42 (2010)

According to the studies previously mentioned, the company size variable is expected to have a positive relationship with dividend payouts. The current study will make this same assumption. As a result, Hypothesis 3 is as follows: H3: Ceteris paribus, company size positively affects the probability of paying dividends. In this study, company size will be measured by the natural logarithm of market capitalisation. 2.4. Growth and Investment Opportunities Since external financing is costly, fast-growing companies require internally generated funds to finance their investment projects, thereby reducing dividends. In contrast, companies with slow growth and fewer investment opportunities have a greater ability to pay higher dividends. This negative relationship has been supported by a large number of studies (Rozeff, 1982; Lloyd et al., 1985; Jensen et al., 1992; Dempsey and Laber, 1992; Alli et al., 1993; Moh'd et al., 1995; Holder et al., 1998). This negative relationship between company growth opportunities and dividend payouts is consistent with the pecking order theory of Myers and Majluf (1984), where investment policy acts as an alternative to dividend payouts and reduces the agency problem by reducing free cash flow. To retain comparability, the current study also uses this proxy for growth opportunities by testing the following hypothesis: H4: Ceteris paribus, growth and investment opportunities negatively affect the probability of paying dividends. 2.5. Financial Leverage Financial leverage reveals how much a company utilises debt in their capital structure. Debt financing enables a company to leverage shareholders’ return on equity, but it involves risk, as debt must be repaid. Since risk is associated with high financial leverage, highly leveraged companies pay lower dividends to maintain their internal cash flow to fulfil their duty and protect their creditors. That is, highly leveraged firms pay lower dividends to reduce their transaction costs (Rozeff, 1982; Crutchley and Hansen, 1989; Jensen et al., 1992; Agrawal and Jayaraman, 1994; and Gugler and Yurtoglu, 2003). Jensen (1986), however, indicated that debt can serve as a substitute for dividends in reducing the agency costs of free cash flow. Moreover, when a company carries debt, it makes a fixed commitment to its creditors, reducing the discretionary funds available to their managers and subjecting managers to the scrutiny of debt-suppliers. This suggests that highly leveraged companies are expected to have low dividend payouts. To examine the extent to which debt can affect dividend payouts, this study employed the financial leverage ratio, defined as the ratio of liabilities to total shareholders’ equity. Using this definition, we formulated the following hypothesis: H5: Ceteris paribus, company leveraging negatively affects the probability of paying dividends. 2.6. Business Risk Business risk reflects the uncertainty of the direct relationship between current and expected future profits. Consequently, companies avoid commitments to pay high dividends, as the uncertainty about earnings increases (Jensen et al., 1992). Higgins (1972), McCabe (1979), Rozeff (1982), Chang and Rhee (1990), and Chen and Steiner (1999) observed that companies with volatile cash flows may choose to pay lower dividends because high cash-flow volatility is a sign that the company may be facing high business risks and need to avoid costly external financing. Moh'd et al. (1995) also reported an inverse relationship between the dividend ratio and business risk, since companies with volatile earnings pay lower dividends to avoid external fund costs.

24

International Research Journal of Finance and Economics - Issue 42 (2010)

The current study uses business beta to proxy business risk, following previous studies recommendations (Rozeff, 1982; Lloyd et al., 1985; Jensen et al., 1992; Alli et al., 1993; Moh’d et al., 1995; Holder et al., 1998; Saxena, 1999; and Manos, 2002). As a result, we formulated the following hypothesis: H6: Ceteris paribus, business risk negatively affects the probability of paying dividends. 2.7. Profitability The financial literature documents that a firm’s profitability is a significant and positive explanatory variable of dividend policy (Jensen et al., 1992; Han et al., 1999; Fama and French, 2001). However, Lintner (1956) indicates that firms follow stable dividend policies; in other words, firms focus their attention on distributing a moderate level of dividends based on long-run targets, rather than paying a higher dividend today followed by a lower one tomorrow. Hence, any change in the dividend amount is based on substantive changes in corporation operations. As a result, a corporation only increases dividends on the belief that there will be a permanent increase in earnings. Fama and Babiak (1968), Pruitt and Gitman (1991), and Baker and Powell (1999) support Lintner’s findings. The stability of dividend policy was commonly identified in many studies. McDonald et al. (1975) analysed dividend policy in French firms. Chateau (1979) and Shevlin (1982) investigated Canadian and Australian corporations, respectively. Kato and Loewnstein (1995) examined dividend policy in Japan. Lasfer (1996) explored the dividend policy of industrial and commercial British firms. Leithner and Zimmerman (1993) tested the dividend-profit relationship in the European market using Germany, France, the UK, and Switzerland, as case studies. La Porta et al. (2000) compared countries with strong legal protection for outside shareholders to those that have weak legal protection, and indicated that shareholders in countries with weak legal protection will take whatever cash dividends they can get, regardless of the dividend policy stability. In other words, when the outside shareholders’ legal protection is poor, dividend policy depends heavily on the firm profitability for the same year Since many emerging markets have poor legal protection, this study hypothesises that: H7: Ceteris paribus, company profitability positively affects the probability of paying dividends. The current study will measure company profitability by using return on equity (Aivazian et al., 2003; and ap Gwilym et al., 2004), To test our seven hypotheses and investigate factors that may influence the decision to pay dividends in the GCC listed companies, a functional relationship among dividends and their contributing factors can be expressed as follows, in which a positive sign indicates that the variable affects dividends positively, and vice versa: DIV = f {GOV(+), FCF (+), SIZE (+), GROW(-), LEV(-), BETA(-), PROF(+)}. (1) A description of the variables in Equation (1) and the selection of the relevant proxy variables are provided in Table 1. Table 1: Variables DIV GOV FCF SIZE GROW LEV BETA PROF

Summary of Variables and Research Hypotheses Description Paid dividends Percentage of shares owned by the government Free cash flows proxied by (net profit - changes in networking capital) / total assets Natural logarithm of market capitalisation Firm growth rate of sales Leverage ratio proxied by debt to equity ratio Business risk proxied by beta Firm profitability

Hypothesis (+) (+) (+) (-) (-) (-) (+)

25

International Research Journal of Finance and Economics - Issue 42 (2010)

3. Data Description The GCC stock exchange includes firms in Qatar, Saudi Arabia, Bahrain, Kuwait, and Oman. The United Arab Emirates (UAE) stock exchange, the Abu-Dhabi Stock Exchange, and the Dubai Financial Market were excluded from this study because of difficulties in acquiring sufficient data. In addition, the UAE stock exchange was recently established with a limited number of listed companies. Hence, the young age of the stock exchange and the small number of listed companies made it hard to perform a credible analysis. In total, there were 245 non-financial companies listed on the 5 stock exchanges in the GCC countries. The Muscat Stock Exchange (MSE) had the highest number of non-financial companies (75). Saudi Arabia had 62 companies, while Kuwait had 59, Qatar had 19, and Bahrain had 13. However, to examine the companies listed in the five countries for a specific number of years, it was essential to gather data for each country over the same time period. Missing data constrained this analysis to 191 non-financial companies (Table 2) and limited the study period to 5 years (1999 – 2003), as this was the longest period for which complete data for all GCC countries were obtainable. Table 2:

Non-financial companies on GCC states’ stock exchanges

Stock Exchange Kuwaiti Stock Exchange Saudi Stock Exchange Muscat Stock Exchange Doha Stock Exchange Bahrain Stock Exchange Total

Total Number of Firms 59 62 92 19 13 245

Total Number of Firms for which the Required Data were Available 37 57 75 10 12 191

% of Available Firms 63% 92% 82% 53% 92% 78%

Both the dividend payout ratio and the factors affecting the dividends for the 191 non-financial firms from 1999 to 2003 were collected. The primary data source was the 2004 Gulf Investment Guide (GIG). In addition, the directories of the national stock exchanges for each state were obtained to provide data that were not available in the GIG. However, it was difficult to obtain data on government ownership and business risk (BETA) for firms listed on the stock exchanges of Saudi Arabia, Kuwait and Qatar. Such data for Saudi Arabia were obtained from the Zughaibi and Kabbani (www.zkfc.com) financial consulting firm. Government ownership data for Kuwait were gathered from The National newspapers from 1999 to 2003. Government ownership data for Qatar were obtained from an unpublished report supplied by the Doha Stock Exchange. The dependent variable is whether a firm paid dividends in a specific year or not. If a firm paid dividends in a specific year, the dependent variable was 1, otherwise, it was 0. Of the 920 observations, 48.4% of the firms did not pay any dividends, while the remaining 51.6% of firms did. Table 3 presents the descriptive statistics of the variables included in the models. Table 3:

Descriptive Statistics of the Variables used in the Study for Non-Financial Companies Listed on GCC States’ Stock Exchanges for the period 1999-2003

Variables Dividend ratio (DIV) Government ownership (GOV) Free cash flow (FCF) Market capitalisation (MC) $000 Growth rate (GROW) Firm leverage (LEV) Business risk (BETA) Firm profitability (PROF)

Mean

Std. Dev.

41.8947 10.1226 0.0031 10.8353 0.3013 106.4513 0.3909 8.5110

54.9293 17.5390 0.2609 2.1534 1.4719 170.1568 0.4615 11.5267

Q1 0 0 -0.09 10010 -0.02 17 0.01 0

Quartiles Q2 8 1 0.02 53266 0.05 49 0.27 6

Q3 77 10 0.12 210600 0.20 128 0.72 4

International Research Journal of Finance and Economics - Issue 42 (2010)

26

4. Methodology The purpose of this study was to identify firm characteristics that lead to the decision to pay (or not pay) dividends. The review of the literature suggested that a suitable model for such a dependent variable was a probit or logit model (Long, 1997). Long (1997) suggested that the coefficients from a logit model are 1.6 to 1.8 times larger than those for the probit model. As a result, the probit model was considered in this study. The probit model is typically formulated in a latent (i.e., unobserved) variable framework. The general specification2 is as follows (Long, 1997): y i* = X i β ′ + ε i (2) * Where y i is a continuous, latent (unobserved) variable measuring the firm’s willingness to pay th X (or not pay) dividends for the i firm; i is a (k×1) vector of observed explanatory variables; β is a

ε

(k×1) vector of unknown parameters to be estimated; and i is the random error term that has a normal distribution with a mean of 0 and a variance of 1. * y The unobserved variable y i is linked to the observed binary variable i by the following measurement equation:  *  1 if y i > φ yi =  *  0 if y i ≤ φ (3) φ φ = 0 is the binary response variable. Where is the threshold and The probability that the outcome is 1 can be expressed as: Pr( y = 1 | X ) = Φ( X β ′) , (4) Φ Where is the cumulative distribution function for the standard normal distribution. The parameters included in β are typically estimated by the maximum likelihood method. Since the dataset used in this study is a panel dataset, a fixed-effects or random-effects probit model is appropriate to use (than the standard probit model), as these models are able to take into account the unobserved heterogeneity normally found in a panel dataset. Since the estimation of a fixed-effects probit model is complex (STATA, 2006), a random-effects probit model was used. The random-effects probit model can be expressed as: Pr( yit = 1 | X it ) = Φ( X it β ′ + ν i ) , (5) i = 1 ,......... ., N t = 1 ,......... ., T Where are the cross-sectional units; are the time-series units; and

ν i consists of the random-effects among the cross-section, which are assumed to be independently and σ2 identically distributed with a zero means and a constant variance. If ν represents the panel-level variance, then the total variance can be expressed as ρ (rho): σ ν2 ρ= 1 + σ ν2 , (6) If ρ is zero, the panel-level variance becomes unimportant, suggesting that the panel estimator was not different from the pooled estimator. A random-effects probit model can be estimated using the maximum likelihood method, in which the log likelihood is calculated either by using the adaptive or non-adaptive Gauss-Hermite quadrature (Liu and Pierce, 1994).

2

See Long (1997) for a more detailed description of the model.

27

International Research Journal of Finance and Economics - Issue 42 (2010)

5. Modelling Results and Discussions As discussed previously, the primary objective of this study was to identify firm characteristics that lead to a policy regarding whether to pay, or not pay, dividends. This was accomplished by using a random effects probit model to analyse a panel dataset from the non-financial firms listed on the GCC stock exchanges. Before estimating the model, the existence of multicollinearity and the bias of the standard errors of the coefficients were investigated. Both the pair-wise correlation matrix and the variance inflation function (VIF) were used. The results of the pair-wise correlation matrix among the explanatory variables are illustrated in Table 4, revealing that the correlation coefficients between any pair of explanatory variables were low (i.e., less than 0.3), suggesting that there was no multicollinearity problem among these explanatory variables. Table 4:

Correlation Coefficients among the Explanatory Variables

Variables Government ownership (GOV) Free cash flow (FCF) Firm size (MC) Growth rate (GROW) Firm leverage (LEV) Business risk (BETA) Firm profitability (PROF)

GOV 1.0000 0.1044 0.2522 -0.0130 -0.1572 0.0828 0.1278

FCF

MC

GROW

LEV

BETA

PROF

1.0000 0.0227 -0.0650 -0.0900 0.1073 0.2470

1.0000 -0.0245 -0.0177 -0.1569 0.0237

1.0000 0.0877 -0.0576 -0.0345

1.0000 -0.1713 -0.0792

1.0000 0.1448

1.0000

The results using the VIF are presented in Table 5. The mean VIF was 1.08, which was very low relative to the threshold value of 10. The VIF for each variable was also very low, indicating that the explanatory variables included in the model were not substantially correlated with each other. Table 5:

The Variance Inflation Factor

Variables Government ownership (GOV) Firm size (MC) Firm profitability (PROF) Free cash flow (FCF) Business risk (BETA) Firm leverage (LEV) Growth rate (GROW) Mean VIF

VIF 1.11 1.09 1.09 1.08 1.08 1.07 1.01 1.08

Tolerance 0.8971 0.9147 0.9147 0.9240 0.9259 0.9386 0.9870

Two random-effects probit models were then estimated. In the first model, no country dummies were used to control for the impact of different stock exchanges, while country dummies were employed in the second model (Table 6). The parameters of the models were estimated using a maximum likelihood (ML) method in which the likelihood function was calculated using the adaptive Gauss-Hermite quadrature method. The computational time was found to be roughly proportional to the number of points used for the quadrature. To obtain a reliable estimate of the parameters, a model should be estimated for different values of the quadrature points. According to Liu and Pierce (1994), if the coefficients estimated for different quadrature points do not change by more than a relative difference of 0.01%, the choice of quadrature points does not significantly affect the outcome and the results may be confidently interpreted. However, if the results change appreciably (such as being greater than a relative difference of 1%), then questions should be raised as to whether the model can be reliably fitted using the chosen quadrature method and the number of integration points. It is interesting to note that two aspects of the random-effects models have the potential to make the quadrature approximation inaccurate: (1) large group sizes and (2) large correlations within groups.

28

International Research Journal of Finance and Economics - Issue 42 (2010)

In this study, both models were estimated using the adaptive Gauss-Hermite quadrature method, where the quadrature points included 16, 20, and 24. A relative difference among the coefficients was found to be 0.01% or less, suggesting that the results were stable. Table 6 illustrates the estimation results for the quadrature point of 20. Table 6:

Estimation Results from a Random Effects Probit Model

Explanatory Variables Government ownership (GOV) Free cash flow (FCF) In(Firm size (MC) in U.S.$) Growth rate (GROW) Firm leverage (LEV) Business risk (BETA) Firm profitability (PROF) Country-specific Dummies Qatar (Reference) Bahrain Saudi Arabia Kuwait Oman Constant Descriptive Statistics Observations Panel-level variance (rho) Log-likelihood ratio test Log-likelihood at convergence Pseudo R-squared Akaike information criterion (AIC)

Model A: without dummies Coeff t-stat 0.0463 4.23 0.1923 0.55 0.3176 3.83 -0.1718 -2.64 0.0013 -1.73 0.0491 -0.21 0.0835 7.51

Model B: with dummies Coeff t-stat 0.0344 3.28 0.2812 0.28 0.3484 3.37 -0.2548 -2.99 -0.0013 -1.58 -0.0184 -0.08 0.0767 6.94

-4.2592

11.6339 -2.2471 -1.6301 -1.9985 -2.6924

-4.71

920 0.784 p-value=0.00 -338.5493 0.2 693.0986

0 -2.4 -1.53 -1.92 -2.07

920 0.742 p-value=0.00 -320.7349 0.23 668.0514

As can be seen from Table 6, the panel-level variance component (rho) was found to be important in both models, suggesting that the panel estimator (i.e., a random effects probit model) was better than the pooled estimator (i.e., a probit model). This result was obtained by a log-likelihood ratio test, which rejected the pooled estimator in favour of the panel estimator. All goodness-of-fit statistics presented in Table 6 reveal that the model with country-specific dummies fits the data better than the model without the country-specific dummies. Therefore, the rest of the discussion will be based on the random-effects model using country-specific dummies. The dummy variable representing Qatar was taken as a reference case, while the effects of the other countries (i.e., Bahrain, Saudi Arabia, Kuwait, and Oman) were estimated. The results imply that there were no significant differences in dividend policies among the firms listed on the Qatar and Bahrain stock exchanges. Firms listed on the Saudi Arabia, Kuwait, and Bahrain stock exchanges were found to be negatively associated with the willingness to pay a dividend. Four factors were found to be statistically significant in the models, suggesting that these four firm characteristics were associated with the firm’s willingness to pay (or not pay) dividends to their shareholders. These four factors were: government ownership, log of market capitalisation, firm growth and investment opportunities, and firm profitability. The interpretation of each of these variables is subsequently provided.

5.1. Agency Theory and Government Ownership Government ownership (%) was a significant factor affecting the dividend payout decision of the companies listed on the GCC stock exchanges, which supports H1. The slope coefficient of this variable was 0.034 with a t-statistic of 3.28. This result indicates that, in GCC countries, where the legal protection of outside shareholders is poor, government ownership may play a significant role in

29

International Research Journal of Finance and Economics - Issue 42 (2010)

the decision-making process regarding dividend payments. This is informed by the belief that investors need protection in countries with weak legal protection and because the government may be a powerful investor acting on behalf of minority shareholders. As a result, companies will be forced to pay high dividends. Consequently, agency conflict will be reduced and firm reputation will be maintained. This finding is in line with the results of Glen et al. (1995) and Naser et al. (2004). The result is also in line with Al-Malkawi (2008) and Gugler’s (2003) argument that indicates that firms pay dividends when governments own a portion of a firms shares to shrink the double agency conflict. There might be a further explanation for the positive association between dividend decision and government ownership, related, in particular, to the GCC states. That is, the governments of GCC countries are looking to diversify their economic resources because of the ongoing deficits in state budgets and the negative impact on the GCC countries’ economies in terms of fluctuations in the price of, or decreased demand for, crude oil. One way to diversify their economic resources and reduce the dependency on oil revenue and the government sector would be to develop and encourage investment in the private sector. Therefore, governments may force firms to pay large dividends to enhance the reputation of the private sector by suggesting that the exploitation of minority shareholders is avoided. This respectable reputation may then attract small investors to invest in such companies. The estimated coefficients presented in Table 6 were used to determine the probability that a company paid a dividend for a given value of government ownership. Figure 1 illustrates how the predicted probability of being paid a dividend changes with the variation in the percentage of shares owned by the government. If all other explanatory variables included in the model were held constant at their means, it is more likely that a firm pays dividends if the number of shares of the firm held by the government increases. It is noticeable that the probability of paying a dividend becomes 1 if the number of shares owned by the government exceeds 70% or more, ceteris paribus. Figure 1: The predicted probability that a firm pays a dividend (i.e., y = 1) for different values of government ownership 1.00 0.95 0.90

Pr(y=1)

0.85 0.80 0.75 0.70 0.65 0.60 0.55 0.50 0

10

20

30

40

50

60

70

80

90

100

Shares owned by the Government (% )

5.2. Firm Size Company size is the second factor that was found to be a statistically significant determinant of paying dividends. This result supports H3, which predicts that company size and the dividend ratio should have a positive association. The log of market capitalisation was used to represent firm size. The slope coefficient of the log of market capitalisation was found to be 0.348 with a t-statistic of 3.37. This positive association might reflect the concern that large-sized companies have widespread ownership. Hence, as a firm’s size increases, the shareholders’ ability to monitor company performance closely decreases. In other words, as the firms size increases, the agency conflict will also increase.

30

International Research Journal of Finance and Economics - Issue 42 (2010)

Consequently, large companies have been found to choose to pay dividends with the aim of lowering agency costs (Jensen and Meckling, 1976; Lloyd et al., 1985; and Jensen et al., 1992). Another explanation for this positive relationship is that large-sized companies acquire lower issuance costs from external funds and are beneficially placed in the capital market to raise external financing at low costs (Eddy and Seifert, 1988; Redding, 1997; Holder et al., 1998; and Fama and French, 2001). Figure 2: The predicted probability that a firm pays a dividend (i.e., y = 1) for different values of firm size 1 0.9

Pr(y=1)

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 4

6

8

10

12

14

16

18

20

ln(MC)

Figure 2 illustrates how the predicted probability of being paid a dividend varies with a change in firm size. The probability ranges from 0.1 to 1, when the firm size increases from 176 thousand US$ to 43,500,000 thousand US$. In other words, if all other explanatory variables included in the model are held constant at their means, it is more likely that a firm would pay a dividend if the firm is large.

5.3. Growth and Investment Opportunities Growth and investment opportunities emerged as a statistically significant variable in explaining a firm’s decision to pay a dividend. The coefficient was negative, suggesting that a firm might decide not to pay a dividend when there were growth opportunities, supporting H4. The payment of dividends removes resources from the firm and helps to mitigate the agency costs of free cash flows. Therefore, firms with high growth opportunities are more likely to not pay dividends, since they have lower free cash flows (Rozeff, 1982; Lloyd et al., 1985; Jensen et al., 1992; Dempsey and Laber, 1992; Alli et al., 1993; Moh'd et al., 1995; and Holder et al., 1998). However, the negative relationship between a company’s growth opportunities and its dividend payouts is consistent Myers and Majluf (1984), where investment policy acts as an alternative to dividend payouts and reduces the agency problem by reducing the free cash flow.

31

International Research Journal of Finance and Economics - Issue 42 (2010)

Pr(y=1)

Figure 3: The predicted probability that a firm pays a dividend (i.e., y = 1) for different growth rates 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 -6

-4

-2

0

2

4

6

8

10

12

14

16

18

20

Firm growth rate of sales

Figure 3 illustrates how the predicted probability of being paid a dividend changes with a change in the growth rate of sales. The probability changes from 1 to 0, while the growth rate increases from -2.7 to 18 percent. In other words, it is more likely that the firm will not pay a dividend if its growth opportunities are high.

5.4. Firm Profitability Company profitability is the last influential factor, and was found to positively affect the decision to pay dividends, which provides support for H7. The positive association suggests that the profitability of companies listed on emerging stock exchanges is a critical determinant of the dividend decision. This finding implies that, since legal protection is poor in developing countries, shareholders might take whatever they can get from company profits, regardless of dividend stability (La Porta et al., 2000). Figure 4: The predicted probability that a firm pays a dividend (i.e., y = 1) for different values of firm profitability 1.00 0.90 0.80

Pr(y=1)

0.70 0.60 0.50 0.40 0.30 0.20 0.10 0.00 -100

-75

-50

-25

0

25

50

Firm profitability (%)

75

100

32

International Research Journal of Finance and Economics - Issue 42 (2010)

Figure 4 illustrates how the predicted probability of being paid a dividend changes with the corresponding change in firm profitability. The probability sharply changes from 0 to 1, while firm profitability increases from -30 to 30 percent. As a result, it is more likely that a firm does not pay a dividend if the growth rate of sales is high. The insignificant factors in the model were free cash flow and business risk. The insignificance of free cash flow might be related to the significant influence of government ownership for these companies, which may force those with high free cash flow to distribute dividends. Moreover, the appearance of the common transaction cost variables, such as business risk as an insignificant variable, suggests that transaction costs do not have a direct influence on the dividend payout policy. The companies listed on the GCC states’ stock exchanges may consider the agency problem and company reputation more than the transaction costs when making decisions to pay dividends. Since the dependent variable - the firm’s willingness to pay a dividend - is a latent variable, it is more interesting to interpret the partial change in the probability that a firm has paid a dividend to their shareholders. This is known as the marginal effect. Table 7 illustrates the marginal effects of the probability of being paid a dividend for changes in the explanatory variables. Table 7:

Marginal effects

Explanatory Variables Government ownership (GOV) Free cash flow (FCF) Firm size (MC) in 000 U.S. $ Growth rate (GROW) Firm leverage (LEV) Business risk (BETA) Firm profitability (PROF)

Marginal effect 0.012 0.102 0.126 0.092 0.000 0.007 0.028

t-test 3.28 0.82 3.37 2.99 1.58 0.08 6.94

Measured at 0.000 0.003 10.835 0.301 106.451 0.391 8.511

For instance, the marginal effect for the probability of being paying a dividend with respect to government ownership (%) is positive and the value is 0.012: ∂ Pr( y = 1 | x) = 0.012 ∂(GOV ) The findings of this study, in terms of the signs of the coefficients, the predicted probabilities, and the marginal effects, would facilitate the identification of factors that are influencing the decision of whether or not to pay dividends to shareholders of firms listed on the stock exchanges of the GCC states.

6. Conclusions The current study has explored factors that affect the dividend policy decisions of companies listed on stock exchanges in emerging markets, namely those of the GCC. The models were developed using probit specifications based on a panel dataset consisting of observations from 191 companies over a period of five years (1999-2003). The results illustrated that four factors were influential in the dividend decisions of the non-financial companies listed on the GCC states’ stock exchanges. These factors include government ownership, company size, and company profitability, which positively affect the decision to pay dividends. The fourth factor, growth rate, negatively affects the company dividend decision. Furthermore, the free cash flow, firm leverage, and business risk factors were found to be statistically insignificant. The changes in the probability of paying a dividend due to the changes in the individual variables were estimated and the marginal effects were calculated. As expected, there were differences in the effects of the specific variables among countries.

33

International Research Journal of Finance and Economics - Issue 42 (2010)

This study provides new evidence that firms’ dividend decisions are based on the type of government ownership, firm size, firm profitability, growth rate and investment opportunities. More specifically, government ownership, firm size, and firm profitability were found to affect both the decision to pay, or not pay, a dividend, as well as the level of the dividend paid. Firm growth and investment opportunity were found to affect the decision to make a dividend. .The significance of these four factors to gather support for our hypotheses provides new evidence in relation to emerging markets: that firms make decisions to pay dividends to reduce agency conflict, avoid exploiting minority shareholders, and enhance their company's reputation. One way to extend this study is to investigate the dividend payout ratios of individual states and compare these results with those of the GCC countries. Another is to extend this analysis by disaggregating the firms into sectors, such as the service and industry sector. Such an analysis will assist in identifying the sectors where firms hand out the highest and lowest dividends.

References [1] [2] [3]

[4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17]

Agrawal, A. and N. Jayaraman, 1994. “The Dividend Policies of All-Equity Firms: A Direct Test of the Free Cash Flow Theory”, Managerial and Decision Economics 15, pp. 139-148. Aivazian, V., L. Booth, and S. Cleary, 2003. “Do Emerging Market Firms Follow Different Dividend Policies from U.S. firms?”, Journal of Financial Research 26, pp. 371-387. Allen, F. and R. Mikhaely, 2003. “Payout Policy.” In G.M. Constantinides, M. Harris and R.M. Stulz (Eds.), “Handbook of the Economics of Finance, Edition 1”,Elsevier, Amsterdam, pp. 337-429. Alli, K., A. Khan, and G. Ramirez, 1993. “Determinants of Corporate Dividend Policy: A Factorial Analysis”, The Financial Review 28, pp. 523-547. Al-Malkawi, H., 2008. “Factors Influencing Corporate Dividend Decision: Evidence from Jordanian Panel Data”, International Journal of Business 13, pp. 177-195. Ambarish, R, K. John and J. Williams, 1987. “Efficient Signalling with Dividends and Investments”, Journal of Finance 42, 321-343. ap Gwilym, O., J. Seaton, and S. Thomas, 2004. “Dividends Aren’t Disappearing: Evidence from the UK”, Working Paper, University of Southampton. Asquith, P. and W. Kraske. 1985. “Changes in Dividend Policy and Stock Trading Volume”, Working Paper, Harvard Business School. Asquith, P. and D. Mullins, 1986. “Signalling with Dividends, Stock Repurchases and Equity Issues”, Financial Management 15, 27-44. Baker, H. and Powell, E., 1999. “How Corporate Managers View Dividend Policy”, Quarterly Journal of Business and Economics 38, pp. 18-35. Black, F., 1976. “The Dividend Puzzle”, Journal of Portfolio Management 2, pp. 5-8. Brealey, R., and Myers, S., 2003. “Principles of Corporate Finance”, McGraw Hill, 7th edition. Brennan, M., 1970. “Taxes, Market Value and Corporate Financial Policy”, National Tax Journal 23, 417-427. Brigham, F. and J. Gordon, 1968. “Leverage, Dividend Policy, and the Cost of Capital”, Journal of Finance 23, 85-103. Chang, R. and S. Rhee, 1990. “The Impact of Personal Taxes on Corporate Dividend Policy and Capital Structure Decisions”, Financial Management 19, pp. 21-31. Chateau, J., 1979. “Dividend Policy Revisited: Within-and Out-of-Sample Tests”, Journal of Business, Finance and Accounting 6, pp. 355-372. Chen, C. and T. Steiner, 1999. “Managerial Ownership and Conflicts: A Nonlinear Simultaneous Equation Analysis of Managerial Ownership, Risk-Taking, Debt Policy, and Dividend Policy”, Financial Review 34, pp. 119-136.

International Research Journal of Finance and Economics - Issue 42 (2010) [18] [19]

[20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42]

34

Crutchley, C. And R. Hansen, 1989. “A Test of the Agency Theory of Managerial Ownership, Corporate Leverage and Corporate Dividends”, Financial Management 18, pp. 36-46. Dempsey, S. and G. Laber, 1992. “Effects of Agency and Transaction Costs on Dividend Payout Ratios: Further Evidence of the Agency-Transaction Cost Hypothesis”, Journal of Financial Research 15, pp. 317-321. Easterbrook, F., 1984. “Two Agency-Costs Explanations of Dividends”, American Economic Review 74(4), pp. 650-659. Eddy, A., and B. Seifert, 1988. “Firm Size and Dividend Announcements”, Journal of Financial Research 11, pp. 295-302. Elton, J. and Gruber, M., 1970. “Marginal Stockholder Tax Rates and the Clientele Effect”, Review of Economics and Statistics 52, pp. 68-74. Fama, E. and Babiak, H., 1968. “Dividend Policy: An Empirical Analysis”, Journal of the American Statistical Association 63, pp. 1132 -1161. Fama, E. and K. French, 2001. “Disappearing Dividends: Changing Firm Characteristics or Lower Propensity to Pay?”, Journal of Financial Economics 60, pp. 3-43. Fisher, G., 1961. “Some Factors Influencing Share Prices”, Economic Journal LXXI, 121-141. Frankfurter, G. and Wood, B., 2002. “Dividend Policy Theories and Their Empirical Tests”, International Review of Financial Analysis 11, pp. 111-138. Glen, J., Y. Karmokolias, R. Miller, and S. Shah, 1995. “Dividend Policy and Behavior in Emerging Markets”, Discussion Paper No. 26, International Financial Corporation. Gordon, M. and Shapiro, M., 1956. “Capital Equipment Analysis: The Required Rate of Profit”, Management Science 3, pp. 102-110. Greene, W., 2003. “Econometric Analysis”, Pearson Education, New Jersey. Gugler, K., 2003. “Corporate Governance, Dividend Payout Policy, and the Interrelation Between Dividends, R&D, and Capital Investment”, Journal of Banking and Finance 27, pp. 1297-1321. Gugler, K. and B. Yurtoglu, 2003. “Corporate Governance and Dividend Pay-Out Policy in Germany”, European Economic Review 47, pp. 731-758. Gul, F., 1999. “Government Share Ownership, Investment Opportunity Set and Corporate Policy Choices in China”, Pacific-Basin Finance Journal 7, pp. 157-172. Han, K., S. Lee, and D. Suk, 1999. “Institutional Shareholders and Dividends”, Journal of Financial and Strategic Decisions 12, pp. 53-62. Healy, P. and Palepu, K., 1988. “Earning Information Conveyed by Dividend Initiation and Omissions”, Journal of Financial Economics 21, pp. 149-176. Higgins, R., 1972. “The Corporate Dividend-Saving Decision”, Journal of Financial and Quantitative Analysis 6, pp. 1527-1541. Holder, M. E., F.W. Langrehr, and J.L. Hexter, 1998. “Dividend Policy Determinants: An Investigation of the Influences of Stakeholder Theory”, Financial Management 27, pp. 73-82. Jensen, G., D. Solberg, and T. Zorn, 1992. “Simultaneous Determination of Insider Ownership, Debt, and Dividend Policies”, Journal of Financial and Quantitative Analysis 27, pp. 274-263. Jensen, M., 1986. “Agency Costs of Free Cash Flow, Corporate Finance, and Takeovers”, American Economic Review 76(2), pp. 323-329. Jensen, M. and W. Meckling, 1976. “Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure”, Journal of Financial Economics 3(4), pp. 305-360. Kalay, A., 1982a. “The Ex-Dividend Day Behaviour of Stock Prices”, Journal of Finance 37, pp. 1059-1070. Kalay, A. and U. Loewenstein, 1985. “Predictable Events and Excess Returns: The Case of Dividend Announcements”, Journal of Financial Economics 14, pp. 423-449. Kato, K. and Loewenstein, U., 1995. “The Ex-Dividend-Day Behaviour of Stock Prices: The Case of Japan”, Review of Financial Studies 8, pp. 816-847.

35 [43] [44] [45] [46] [47]

[48] [49] [50] [51] [52] [53] [54]

[55] [56] [57] [58] [59] [60]

[61] [62] [63] [64]

International Research Journal of Finance and Economics - Issue 42 (2010) Kumar, P and G. Tsetsekos, 1999. “The Differentiation of 'Emerging' Equity Markets”, Applied Financial Economics 9, pp. 443-453. La Porta, R., F. Lopez-De-Silanes, A. Shleifer, and R. Vishny, 2000. “Agency Problems and Dividend Policy Around the World”, Journal of Finance 55, pp. 1-33. Lasfer, M., 1996. “Taxes and Dividends: the UK Evidence”, Journal of Banking and Finance 20, pp. 455-472. Lintner, J., 1956. “Distribution of Incomes of Corporations Among Dividends, Retained Earnings, and Taxes”, American Economic Review 46, 97-113. Litzenberger, R. and K. Ramaswamy, 1979. “The Effect of Personal Taxes and Dividends on Capital Asset Prices: Theory and Empirical Evidence”, Journal of Financial Economics 7, pp. 163-195. Litzenberger, R. and K. Ramaswamy, 1982. “The Effects of Dividends on Common Stock Prices: Tax Effects or Information Effects?”, Journal of Finance 37, pp. 429-443. Liu, Q. and D. Pierce, 1994. “A Note on Gauss-Hermite Quadrature”, Biometrika 81, pp. 624-629. Lloyd, W.P., J.S. Jahera, and D.E. Page, 1985. “Agency Costs and Dividend Payout Ratios”, Quarterly Journal of Business and Economics 24, pp. 19-29. Long, S.,1997).Regression Models for Categorical and Limited Dependent Variables: Advanced Quantitative Techniques in the Social Sciences,, Sage Publications Manos, R., 2002. “Dividend Policy and Agency Theory: Evidence on Indian Firms”, Working Paper, Institute for Development Policy and Management, University of Manchester. McCabe, G., 1979. “The Empirical Relationship Between Investment and Financing: A New Look”, Journal of Financial and Quantitative Analysis 14, pp. 119-135. McDonald, J., Jacquillat, B. and Nussenbaum, M., 1975. “Dividend, Investment and Financing Decisions: Empirical Evidence on French Firms”, Journal of Financial and Quantitative Analysis 10, pp. 741- 755. Miller, M. and F. Modigliani, 1961. “Dividend Policy, Growth, and the Valuation of Shares”, Journal of Business 34, pp. 411-433. Miller, M. and K. Rock, 1985. “Dividend Policy under Systematic Information”, Journal of Finance 40, pp. 1031-1051. Moh'd, M., L. Perry, and J. Rimbey, 1995. “An Investigation of the Dynamic Relationship Between Agency Theory and Dividend Policy”, The Financial Review 30, pp. 367-385. Mueller, D., 1967. “The Firm Decision Process: An Econometric Investigation”, Quarterly Journal of Economics 81, pp. 58-87. Myers, S. and N. Majluf, 1984. “Corporate Financing and Investment Decisions When Firms Have Information That Investors Do Not”, Journal of Finance 39, pp. 187-221. Naser, K., R. Nuseibeh, and D. Al-Kuwari, 2004. “Dividend Policy of Companies Listed on Emerging Stock Exchanges: Evidence from Banking Sector of the Gulf Co-operation Council”, Middle East Business and Economic Review 16, pp. 1-12. Poterba, J. and L. Summers, 1984. “New Evidence that Taxes Affect the Valuation of Dividends”, Journal of Finance 39, pp. 1397-1415. Pruitt, S. and Gitman, L., 1991, “The Interactions between the Investment, Financing, and Dividend Decisions of Major US Firms”, Financial Review 26, pp. 409-430. Redding, L., 1997. “Firm Size and Dividend Payouts”, Journal of Financial Intermediation 6, pp. 224-248. Rozeff, M., 1982. “Growth, Beta and Agency Costs as Determinants of Dividend Payout Ratios”, The Journal of Financial Research 5, pp. 249-259.

International Research Journal of Finance and Economics - Issue 42 (2010) [65] [66] [67] [68]

36

Saxena, A., 1999. “Determinants of Dividend Policy: Regulated Versus Unregulated Firms”, Financial Management Association Conference. Shevlin, T., 1982. “Australian Corporate Dividend Policy: Empirical Analysis”, Accounting and Finance, pp. 1-22. Smith, C. And R. Watts, 1992. “The Investment Opportunity Set and Corporate Financing, Dividend and Compensation Policies”, Journal of Financial Economics 32(3), pp. 263-292. STAT, 2006. Stata Statistical Software: Release 9.0, Stata Corporation, College Station, Texas