The main purpose of this research is to investigate how the determinants of the capital structure (leverage) and the dividend payout policy impact the agency cost theory. Literature review part picked up the relevant material related to agency theory, leverage, and dividends payout policy.

The literature review section goes through the agency cost literature, and explores the financial policies; the capital structure (leverage), and the dividend payout policy and that these policies would influence the agency cost theory.

2.1 Agency theory Literature

The notion of the agency theory is widely used in economics, finance, marketing, legal, and social sciences; Jensen and Meckling (1976) initiated and developed it. Capital structure (leverage) for the firms is determined by agency costs, i.e., costs related to conflict of interests between various groups including managers, which have claims on the firm’s resources (Harris and Raviv, 1991).

Jensen and Meckling (1976) defined the agency relationship as “a contract under which one or more persons (the principal) engage another person (the agent), to perform some service on their behalf which involves delegating some decision making authority to the agent” pp.308. Assuming that both parties utility maximizes, the agents are not possible to act in the best interest of the principal.

Furthermore, Jensen and Meckling (1976) contended that the principal can limit divergences from his interest by establishing appropriate incentives for the agent, and by incurring monitoring costs (pecuniary and non pecuniary), which are designed to limit the aberrant activities of the agent. Jensen and Meckling (1976) argued that the agency costs are unavoidable, since the agency costs are borne entirely by the owner. Jensen and Meckling (1976) contended that the owner is motivated to see these costs minimized.

Authors who initiated and developed the agency theory have argued that if the owner manages a wholly owned firm, then he can make operating decisions that maximise his utility. The agency costs are generated if the owner - manager sells equity claims on the firms, which are identical to his. It also generated by the divergence between his interest and those of the outside shareholders, since he then bears only a fraction of the costs of any non-pecuniary benefits he takes out maximizing his own utility (Jensen and Meckling, 1976).

Jensen and Meckling (1976) suggested two types of conflicts in the firm; First of all, the conflict between shareholders and managers arises because managers hold less than a hundred percent of the residual claim. Therefore, they do not capture the entire gain from their profit enhancement activities, but they do bear the entire cost of these activities. For example, managers can invest less effort in managing firm resources and may be able to transfer firm resources to their own, personal benefit, i.e., by consuming “perquisites” such as a fringe benefits. The manager bears the entire cost of refraining from these activities but captures only a fraction of the gain.

As a result, managers over indulge in these interests relative to the level that would maximize the firm value. This inefficiency reduced the large fraction of the equity owned by the manager. Holding constant the manager’s absolute investment in the firm, increases in the fraction of the firm financed by debt increases the manager’s share of the equity and mitigates the loss from conflict between the managers and shareholders.

Furthermore, as pointed out by Jensen (1986), since debt commits the firm to pay out cash, it reduces the amount of free cash flow available to managers to engage in these types of interests. As a result, this reduction of the conflict between managers and shareholders will constitute the benefit of debt financing.

Second, they also suggested that the conflict between debt holders and shareholders arises because the debt contract, gives shareholders an incentive to invest sub optimally. Especially when the debt contract provides that, if an investment yields large returns, well above the face value of the debt, shareholders capture most of the gain. However, if the investment fails, debt holders bear the consequences. Therefore, shareholders may benefit from investing in very risky projects, even if they are under valued; such investments result in an adverse in the value of debt.

Lasfer (1995) argued that debt exacerbates the conflict between debt holders and shareholders. Shareholders will benefit from investments in risky projects at the expense of debt holders. If the investment yields higher return than the face value of debt, shareholders capture most of the gain, however, if the investment fails, debt holders lose, given that. Therefore, shareholders protected by the limited liability.

On the other hand, if the benefits captured by debt holders reduce the returns to shareholders, then an incentive to reject positive net present projects has created. Thus, the debt contract gives shareholders incentives to invest sub optimally. In addition, Myers (1977) argued that the firms with many growth opportunities should not be financed by debt, to reduce the negative net value projects. 

Furthermore, some of arguments have been debated that the magnitude of the agency costs varies among firms. It will depend on the tastes of managers, the ease with which they can exercise their own preferences as opposed to value maximization in decision making, and the costs of monitoring and bonding activities. Therefore, the agency costs depend upon the cost of measuring the manager’s performance and evaluating it (Jensen and Meckling, 1976). (Jensen, 1986) either points out that when firms make their financing decision, they evaluate the advantages that may arise from the resolution of the conflicts between managers, shareholders and from long run tax shields. 

In addition, Lasfer (1995) argues that debt finance creates a motivation for managers to work harder and make better investment decisions. On the other hand, debt works as a disciplining tool, because default allows creditors the option to force the firm into liquidation. Debt also generates information that can be used by investors to evaluate major operating decisions including liquidation (Harris and Raviv, 1990).

Jensen (1986) debated that when using debt without retention of the proceeds of the issue, bonds the managers to meet their promise to pay future cash flows to the debt holders. Thus, debt can be an effective substitute for dividends. By issuing debt in exchange for stock, managers are bonding their promise to pay out future cash flows in a way that cannot be accomplished by simple dividend increases.

Consequently, managers give recipients of the debt the right to take the firm to the bankruptcy court if they do not maintain their commitment to make the interest and principle payments. Thus, debt reduces the agency costs of free cash flow by reducing the cash flow available for spending at the discretion of managers. Jensen (1986) claimed that these control effects of debt are a potential determinant of capital structure.

In practice, it is possible to reduce the owner manager non pecuniary benefits; by using these instruments external auditing, formal control systems, budget restrictions, and the establishment of incentive compensation systems serve to identify the manager’s interests more closely with those of the outside shareholders (Jensen and Meckling, 1976).

Jensen (1986) suggested that leverage and dividend may act as a substitute mechanism to reduce the agency costs. Agency cost models predict that dividend payments can reduce the problems related to information asymmetry. Dividend payments might be consider also as a mechanism to reduce cash flow under management control, and help to mitigate the agency problems (Rozeff, 1982, and Easterbrook, 1984). Therefore, paying dividends may have a positive impact on the firm's value.

“Agency theory posits that the dividend mechanism provides an incentive for managers to reduce the costs related to the principal - agent relationship, one way to reduce agency costs is to increase dividends” Baker and Powell (1999). They also claim that firm use the dividends use as a tool to monitor the management performance. Moreover, Easterbrook (1984) and Jensen (1986) argue that agency costs exist in firms because managers may not always want to maximize shareholder’s wealth due to the separation of ownership and control.

Jensen (1986) addresses the free cash flow theory, in terms of this theory the conflict of interest between managers and stockholders is rooted in the presence of informational and self interest behavior. He defines the free cash flow as “cash flow in excess of that required to fund all projects that have positive net present value when discounted at the relevant cost of capital” (Jensen,1986). Within the context of the free cash flow hypothesis, firms prefer to increase their dividends and distribute the excess free cash flow in order to reduce agency costs. Consequently, markets react positively to this type of information. This theory is attractive because it is consistent with the evidence about investment and financing decisions (Jensen, 1986, Frankfurter and Wood, 2002).

2.2 Leverage Literature

This section reviews the determinants of capital structure by different relevant literatures. Titman and Wessels (1988) study is considered to be one of the leading studies in the developed markets. They tried to extend the empirical work in capital structure theory by examining a much broader set of capital structure theories, and to analyze measures of short term, long term, and convertible debt. The data covers the US industrial companies from 1974 to 1982, and they used a factor analytic approach for estimating the impact of unobservable attributes on the choice of corporate debt ratios.

As a result, the study confirms these factors, collateral values of assets, non-debt tax shields, growth, and uniqueness of the business, industry classification, firm size, and firm profitability. They also found that there is a negative relationship between debt levels and the uniqueness of the business. In addition, short term debt ratios have a negative relationship to firm size. However, they do not provide support for the effect on debt ratios arising from non debt tax shields, volatility, collateral value of assets, and growth.

In Jordan, Al-Khouri and Hmedat (1992) aimed to find the effect of the earnings variability on capital structure of Jordanian corporations from the period from 1980 to 1988. They included 65 firms. The study used a multivariate regression approach with financial leverage as the dependent variable measured in three ways; first, long term debt over total assets, secondly, short term debt over total assets, and finally, short term debt plus long term debt over total assets.

The standard deviation of the earnings variability and the size of the firm measured as independent variables. They concluded that the firm size is considered as a significant factor in determining the capital structure of the firm, and insignificant relationship between the earning variability and financial leverage of the firm. Furthermore, they suggest that the type of industry is not considered as a significant factor in determining the capital structure of the firm.

Rajan and Zingales (1995) provided international evidence about the determinants of capital structure. They examined the capital structure in other countries related to factors similar to those that influence United States firms. The database contains 2583 companies in the G7 countries. They used regression analysis with the firm’s leverage (total debt divided by total debt plus total equity) as the dependent variable.

Tangible assets, market to book ratio, firm size, and firm profitability used as independent variables. They found that in market bases firms with a lot of fixed assets are not highly levered, however, they supported that a positive relationship exists between tangible assets, and firms size, and capital structure (leverage). On the contrary, they confirmed that there is a negative relationship between leverage and the market to book ratio, and profitability.

From the capital structure literature, Ozkan (2001) also investigated that the determinants of the target capital structure of firms and the role of the adjustment process in the UK using a sample of 390 firms. The multiple regression approach (panel data) was used to measure the debts by total debt to total assets, on the one hand. He also used in his model, non debt tax shield, firm size, liquidity, firm profitability, and firm growth as an independent variables. He confirmed that the profit, liquidity, non debt tax shield, and growth opportunities have a negative relationship to capital structure (leverage). Finally, he supported that there is a positive effect arising from size of firms on leverage. The study provided evidence that the UK firms have long term target leverage ratios and that they adjust quickly to their target ratios.

The study by Booth et al. (2001) is considered as a one of the leading studies in the developing countries. It aimed to assess whether capital structure theory is applicable across developing countries with different institutional structures. The data include balance sheets and income statements for the largest companies in each selected country from the year 1980 to 1990. It included 10 developing countries: India, Pakistan, Thailand, Malaysia, Zimbabwe, Mexico, Brazil, Turkey, Jordan, and Korea.

The study used multivariate regression analysis with dependent variables; total debt ratio, long term book debt ratio, and long term market to debt ratio. The independent variables are; average tax rate, tangibility, business risk, firm size, firm profitability, and market to book ratio. Booth et al. found that the more profitable the firm the lower the debt ratio, regardless how the debt ratio is defined. In addition, the higher the tangible assets mix, the higher is the long term debt ratio but the smaller is the total debt ratio. Finally, it concluded that debt ratios in developing countries seem to be affected in the same way by the same set of variables that are significant in developed countries.

Voulgaris et al. (2004) investigated the determinants of capital structure for Greek manufacturing firms. The study used panel data of two random samples one for small and medium sized enterprises (SMEs) including 143 firms and another for large sized enterprises (LSEs) including 75 firms for the period from 1988 to 1996. It used a leverage model as a dependent variable (short run debt ratio, long run debt ratio, and total debt ratio).

On the other hand, It used firm size, asset structure, profitability, growth rate, stock level, and receivables as independent variables. The study suggested that there are similarities and differences in the determinants of capital structure among the two samples. The similarities include that the firm size and growth opportunities positively related to leverage. While, they confirm that the profitability has a negative relationship to leverage.

Moreover, they pointed out the differences that the inventory period, and account receivables collection period have been found as determinants of debt in SMEs but not in LSEs. Liquidity doest not affect LSEs leverage, but it affects the SMEs. Finally, they also suggested that there is a positive relationship between profit margins and short term debt ratio only for SMEs. Voulgaris et al. (2004) have debated this arguments as; ‘‘the attitude of banks toward small sized firms should be changed so they provide easier access to long-term debt financing’’. In addition, “enactment of rules that will allow transparency of operations in the Greek stock market and a healthier development of the newly established capital market for SMEs will assist Greek firms into achieving a stronger capital structure’’.

2.3 Dividends payout ratio literature

Dividend payout ratios vary between firms and the dividend payout policy will impact the agency cost theory. Rozeff (1982) investigated in his study that the dividends policy will be rationalize by appealing the transaction cost and agency cost associated with external finance. Moreover, Rozeff (1982) had found evidences supporting how the agency costs influence the dividends payout ratio. He found that the firms have distributed lower dividend payout ratios when they have a higher revenue growth, because this growth leads to higher investment expenditures. This evidence supports the view of the investment policy affect on the dividend policy; the reason for that influences is that would the external finance be costly.

Conversely, he found that the firms have distributed higher dividends payouts when insiders hold a lower portion of the equity and (or) a greater numbers of shareholders own the outside equity. Rozeff (1982) pointed out that this evidence supports that the dividend payments are part of the firm’s optimum monitoring and that bonding package reduces the agency costs. Moreover, if the agency cost declines when the dividend payout does and if the transaction cost of external finance increases when the dividend payout is increased as well, then minimization of these costs will lead to a unique optimum for a given firm.

In addition, Hansen, Kumar, and Shome [HKS], (1994) pointed out the relevance of the monitoring theory for explaining the dividends policy of regulated electric utilities. From an agency cost perspective, they emphasized their ideas that the dividends promote monitoring of what they call the shareholders - regulator conflict. Therefore, it is a monitoring role of dividends. On the contrary, Easterbrook’s (1984) has noted that the dividends monitoring of the shareholders - managers conflict. They also have observed that the utilities firms have a discipline of monitoring mechanism for controlling agency cost, depending on the relative cost effectiveness of those costs (Crutchly and Hensen, 1989).

The regulator process will impact the conflict between the shareholders and mangers, by mitigate the managers’ power to appropriate shareholders’ wealth and consume perquisites (Hansen et al. 1994). On the other hand, they argued this issue by the cost-plus concept, regulators may set into motion of managerial incentive structure that potentially conflicts with shareholders interests, this concept solve the shareholders-regulators concept since the sources of the conflict lies in differences in the perceptions of what constitutes fair cost plus. Therefore, the regulation can control some of the agency cost while exacerbating others. In their study, they conduct also that the managers and shareholders of unregulated firms have a several mechanisms whether, internal or external, for controlling agency cost.

In addition, they observed that the dividend policy to reduce the agency theory is not limited, depending on their findings they suggested that the cost of dividend payout policy might be below the costs paid by other types of firm. In fact the utilities company maintain high debt ratio that would maintain as well as equity agency costs.

Aivazian et al. (2003b) compare the dividend policy behaviour of eight emerging markets with dividend policies in the US firms in the period from 1980 to 1990. The sample included firms from; Korea, Malaysia, Zimbabwe, India, Thailand, Turkey, Pakistan, and Jordan. They found that it is difficult to predict dividend changes for such emerging markets. This is because the quality of firms with reputations for cutting dividends is somehow similar to those who increase their dividends, than for the US control sample. In addition, current dividends are less sensitive to past dividends than for the US sample of firms. They also found that the Lintner model[1] does not work well for the sample of emerging markets.

These results indicate that the institutional frameworks in these emerging markets make dividend policy a weak technique for signaling future earnings and reducing agency costs than for the US sample of firms.

Furthermore, Omran and Pointon (2004) investigated the role of dividend policy in determining share prices, the determinants of payout ratios, and the factors that affect the stability of dividends for a sample of 94 Egyptian firms. They found that retentions are more important than dividends in firms with actively traded shares, but that accounting book value is more important than dividends and earnings for non-actively traded firms.

However, when they combined both the actively traded and non-traded firms, they found that dividends are more important than earnings. In the determinants of payout ratios, they found that there is a negative relationship between the leverage ratio and market to book ratio, tangibility, and firm size on the one hand, to the payout ratios in actively traded firms. On the contrary, they also found that there is a positive relationship between the business risk, market to book and firm size (measured by total assets) to payout ratios in non-actively traded firms.

Furthermore, for the whole sample, leverage has a positive relationship with payout ratios, while firm size (measured by market capitalization) is negatively related to payout ratios. Finally, the stepwise logistic regression analysis shows that decreasing dividends is associated with lack of liquidity and overall profitability. In addition, increasing dividends is associated with higher overall profitability.

2.4 Summary

In this chapter the relevant literatures addressing the reviews of the agency cost theory related to the financial policies. It also gives a theoretical background on how the conflicts of interests arise between the agents (managers) and the principal (shareholders). The second and third sections present the determinants of leverage and dividend payout policy. The following chapter will go through the description of data, and data methodology was employed for this dissertation.

3. Methodology, Research Design and Data Description

The aim of the current study is to investigate firstly, the empirical evidence of the determinants of leverage and dividend policy under the agency theory concept for the period 2002-2007. The majority of the previous studies in the field of capital structure have made in the context of developed countries such as USA and UK. It is important to investigate the main determinants of leverage and dividend policy in developing countries where, capital markets, are less developed, less competitive and suffering from the lack of compatible regulations and sufficient supervision

This chapter will explain the research methodology of this study. This chapter also identifies the sample of the study. Moreover, it presents an illustration of the econometric techniques that have been employed. In addition, this chapter gives a brief explanation of the specification tests used in the study to identify which technique is the best for the data set.

This chapter structured as follows; Section (3.1) presents data description. Section (3.2) presents the sample of the study. Section (3.3) discusses the econometric techniques employed in the study. Finally, Section (3.4) provides a brief summary.

3.1 Data Description

The data used in the study are secondary data for companies listed at Amman Stock Exchange (ASE) for the period of 2002-2007. The data was extracted from the firm’s annual reports, and from Amman Stock Exchange’s publications (The Yearly Companies Guide, and Amman Stock Exchange Monthly Statistical Bulletins). Data is readily available in the form of CD and on the website of the Amman Stock Exchange.

The reason for the study period selection was to minimize the missing observations for the sample companies. Moreover, a different reporting system has been used since 2000. The application of the new reporting system was the result of the transparency act which was launched in 1999, and forced all companies listed in Amman Stock Exchange to disclose their financial information and publish their annual reports according to the International Financial Reporting Standards. In other words, this data series for the period from 2002-2007 was chosen in terms of consistency and comparability purposes.

3.2 Sample of the study

The sample of the study consists of the Jordanian Manufacturing companies listed on the Amman Stock Exchange for the period of 2002-2007. The total number of the companies listed in ASE at the end of year 2007 was 215. Officially, these companies are divided into four main economic sectors; Banks sector, Insurance sector, Services sector and finally Industrial sector.

Moreover, this study is concerned only with Jordanian manufacturing companies that their stocks are traded in the organized market. It is important to note that the capital structure of financial firms has special characteristic when compared to the capital structure of non financial firms, they also have special tax treatment (Lester, 1995).

On the other hand, the financial firms have a higher leverage rate, which may tend to make the analysis results biased. Moreover, financial firms their leverage is affected by investor insurance schemes (Rajan and Zingals, 1995). For these reasons, the potential sample of the study consists of non financial (Manufacturing) companies that are still listed in Amman Stock Exchange. The total number of industrial companies listed in ASE at the end of year 2007 was 88 companies, which are 40.93% of the total number of the companies listed in that market.

The study conducts the following criteria in selecting the sample upon the Jordanian manufacturing companies by excluding all the firms that was incorporated after year 2002, and all the firms that have merged or acquired during this period, further, the firms have liquidated or delisted by the Amman Stock Exchange, and finally, the study have also excluded the firms that have information missing for that period.

The application for those criteria has resulted in 52 samples of manufacturing companies. The data for the variables that are included in the study models is tested using three different econometric techniques which will be discuss briefly in the next sections.

3.3 Econometrics techniques

Hairs et al. (1998) argued that the application of econometrics technique depends on the nature of data employed in the study, and to what extent it would be realised to the research objectives. In order to find a best and adequate data model, the current study employs pooled data technique and panel data analysis which is usually estimated by either fixed effect technique or random effects technique. The following sections provide a brief discussion on the econometrics techniques that the current study uses to estimate the empirical models.

3.3.1 Pooled Ordinary Least Square (OLS) technique

All the models used in the study have been tested by the pooled data analysis technique. The pooled data is the data that contains pooling of time series and cross-sectional observations (combination of time series and cross-section data) (Gujarati, 2003). The pooled data analysis has many advantages over the pure time series or pure cross sectional data. It generates more informative data, more variability, less collinearity among variables, more degrees of freedom, and more efficiency (Gujarati, 2003). The underlying assumption behind the pooled analysis is that, the intercept value and the coefficients of all the explanatory variables are the same for all the firms, as well as they are constant over time (no specific time or individual aspects). It also assumes that the error term captures the differences between the firms (across-sectional units) over the time.

However, (Gujarati, 2003) has pointed out that these assumptions are highly restrictive. He argues that although of it is simplicity and advantages, the pooled regression may distort the true picture of the relationship between the dependent and independent variables across the firms. Pooled model will be simply estimated by Ordinary Least Square (OLS). However, OLS will be appropriate if no individual (firm) or time specific effects exist. If they exist, the unobserved effects of unobserved individual and time specific factors on dependent variable can be accommodated by using one of the panel data techniques. 

According to (Gujarati, 2003) panel data is a special form of pooled data in which the same cross-sectional unit is surveyed over time. It helps researchers to substantially minimize the problems that arise when there is an omitted variables problems such as time and individual-specific variables and to provide robust parameter estimates than time series and (or) cross sectional data.

All the empirical models that have been tested by using pooled data analysis and tested again on the basis of panel data analysis techniques (Fixed Effects and Random Effects). 

3.3.2 The fixed effects model (FEM)

Fixed effects technique allows control for unobserved heterogeneity which describes individual specific effects not captured by observed variables. According to Gujarati (2003) the fixed effect model takes into account the specific effect of each firm “the individuality” by allowing the intercept vary across individuals (firms), but each individual’s intercept does not vary over time. However, it still assumes that the slope coefficients are constant across individuals or over time.

Two methods used to control for the unobserved fixed effects within the fixed effects model; the first differences and Least Square Dummy variables (LSDV) methods. For the purposes of the current study, (LSDV) was used where; two sets of dummy variables (industry, and year dummy variables). The additional dummy variables control for variables that are constant across firms but change over time. Therefore, the combine time and individual (firm) fixed effects model eliminates the omitted variables bias arising both from unobserved factors that are constant over time and unobserved factors that are constant across firms.

However, fixed effects model consumes the degrees of freedom, if estimated by the Least Square Dummy Variable (LSDV) method and, too many dummy variables are introduced (Gujarati, 2003). Furthermore, with too many variables used as regressors in the models, there is the possibility of multicollinearity. It is worth noting that OLS technique used in estimating fixed effects model.

3.3.3 The Random Effects Model (REM)

By contrast, fixed effects model, the unobserved effects in random effects model is captured by the error term (εit) consisting of an individual specific one (ui) and an overall component (vit) which is the combined time series and cross-section error. Moreover, it treats the intercept coefficient as a random variable with a mean value (α0) of all cross-sectional (firms) intercepts and the error component represents the random deviation of individual intercept from this mean value (Gujarati, 2003). Consequently, the individual differences in the intercept values of each firm are reflected in the error term (ui).

On the other hand, the Generalized Least Square (GLS) used in estimating random affects model. This is because the GLS technique takes into account the different correlation structure of the error term in the Random Effect Model (REM) (Gujarati, 2003).

3.3.4 Statistical specification tests

The study uses three specification tests to identify which empirical method is the best. These tests are used for testing the fixed effect model versus the pooled model (F-statistics), the random effect model versus pooled model (Lagrange Multiplier test) (LM), and the fixed effect model versus the random effect model (Hausman test). The following sub-sections offer brief discussion of these tests and their underlying assumptions.

3.3.4.1 Fixed effects model versus Random effects model

Greene (2003) has argued that the crucial distinction between the fixed and random effects model is whether the unobserved individual and time specific effects embodies elements that correlated with the regressors in the model. Hence, the answer to whether fixed effects or random effects should be used depends on the assumption about the likely correlation between the individual, or cross section specific, error component (εi) and the (Xit) regressors. If there is no substantial reason to assume that a significant correlation between the unobserved firm effects and the regressors, then the random effect model may be more powerful. If there is such a correlation, the random effects model would inconsistently estimated and the fixed effects model would be the model of choice (Gujarati, 2003).

Hausman (1978) provides a test for discriminating between fixed effect effects and random effects estimators. The Hausman’s essential result is that the covariance of an estimator with its difference from an efficient estimator is zero (Greene, 2003). It compares the differences in the estimated coefficients between the fixed effects model and the random effects model under the null hypothesis that these differences are not systematic[2].

The rejection of null hypothesis suggests that fixed effects estimations are more appropriate than random effects estimations. Hence, the difference between the estimators should be close to zero. While the acceptance of null hypothesis indicates random effects estimations are better than fixed effects estimations which suggest that the estimated coefficients are not significantly different (Johnston and Dinardo, 1997).

3.3.4.2 Pooled data model versus fixed and random effects models

The study uses two specification tests to investigate which one of the pooled and panel data analyses (Fixed and Random effects models) is most suitable. The Lagrange Multiplier (LM) test used to assess the random effect model versus the pooled Ordinary Least Square (OLS) model (Breusch, and Pagan, 1980). It has argued that pooled OLS model does not take in consideration the effect of unobserved individual and time specific factors on the dependent variable and the presence of individual and time specific effects may make the estimates obtained from Pooled model inconsistent. Gujarati (2003) has argued that these effects could be accommodated by using one of the panel data techniques (either fixed effect model or random effect model).

Regarding pooled and random effects model, (Breusch, and Pagan, 1980), Lagrange Model (LM) test will be used to test random effects model versus pooled model under the null hypothesis that the cross sectional variance components are zero (H0: б2v ­= 0). The significant Lagrange Multiplier (LM) test leads to the rejection null hypothesis. They suggest that the individual effect is not equal zero, and the estimate coefficients obtained from pooled model are not consistent. (Herwartz, 2006) argues that in the presence of individual effects, б2v > 0, Ordinary Least Square estimator will be unbiased but inefficient in general, while the Generalized Least Square (GLS) estimator will be blue.

3.4 The empirical models[3]

This study investigates the interactions between, leverage and dividend payout policy. This investigation may provide evidence to whether the agency theory is applicable in the Amman Stock Exchange (ASE), particularly in Jordanian manufacturing company, or not, depending on the relationship between dividend and leverage. It is important to note that there is a negative relationship between leverage and dividend support the agency theory of free cash flow. In addition the leverage and dividend have used as a substitute mechanism to reduce free cash flow under management control. While the positive relationship support the prediction of pecking order theory where the firm that pays dividend need to generate more external funds to finance their investment opportunities.

To identify the important financial factors that affect the firm's capital structure, dividend policy pooled and panel regression models applied for the Jordanian manufacturing companies data obtained from the Amman Stock Exchange (ASE). The following regression models; 

3.4.1 The Leverage model;

LEVit= β0 + β1 DPOr + β2 PROFit-1 + β3TANGit-1 + β4MBrit-1 + β5 SZit_1+ β6 LIS + -uit

3.4.2 The dividend policy model;

DPOrit= β0 + β1 LEV + β2 PROFit-1 + β3TANGit-1 + β4MBrit-1 + β5 SZit-1+ β6 LIS + -uit

The variables are described below;

Leverage

Total debt to total assets (LEV)

Dividends

Dividends to EBIT (DPOr)

Profitability

EBIT to total assets (PROF)

Tangibility

Fixed asset to total assets (TANG)

Growth

Market to book ratio (MBr)

Ownership Structure

Shareholders who owned more than 5% in the company (LIS); and

Size

log of total assets (SZ)

3.5 Summary

This chapter focused on the research methodology employed to achieve the research objectives. The first section of the study has discussed the data of the study where a documentary secondary data is used. Secondly, the sample of the study and the criteria was used to select the sample. Thirdly, it explains the econometric techniques that would be used in the study (pooled, fixed effects and random effects regression models). Finally, the study has discussed the specification tests to identify which of method will be the most suitable. The next chapter estimates and test the results upon the structural equations for the leverage and dividends payout policy.

4. Empirical Evidence for the Leverage

The leverage of a firm is actually a mix of different types of debt, whether, long or short term to finance its assets, and also to supplement their investment. Motivation to use debt arise from the fact that it raises cash, provides a tax shield advantage which adds value to the firm and finally it mitigates the agency of the conflicts between managers (agents) and shareholders (principals). (Modigliani and Miller, 1958) under the irreverence theory of capital structure argued that the value of the firm is independent of its capital structure. M.M (1958) built their argument on a restrictive assumption which can not be held in the real world. They assumed that the capital market is perfect, with no taxes, transaction and bankruptcy costs.

Moreover, information is symmetrical; management and investors can borrow at the same rate, in addition, no bankruptcy exists. The application of their theory is that a firm’s investment decision is totally determined by the profitability of the new investment opportunities, not by the way it chooses to finance this opportunities. For this reason, M.M also argued that the firm value is invariant regardless of the way of financing, whether it uses debt or equity.

However, (M.M, 1963) argued that debt provides tax shields benefits because interest payments are tax deductible from taxable income. Hence, the firm value is maximized by using debt. Once again, they ignored the bankruptcy and agency costs of debt. In fact, debt has costs such as the bankruptcy and agency costs which are at least outweighing its benefits. This is either the gripe of the trade off theory of capital structure.

Therefore, debts are not freely used because the excusive use of debts will increase the probability of bankruptcy. According to (Myers, 1977) firms are forced to forgo some of rich investment opportunities if their borrowing reserve capacity is being closed to the maximum. Jensen and Meckling (1976) addressed the effect of agency costs of debt between shareholders and debt holders on the firm’s financing behaviour due to the assets substitution problem.

However, the agency theory of (Jensen and Meckling, 1976) and (Jensen, 1986) added another motivation or benefit for using debt due to the role of debt in mitigating the conflicts between the managers (agents) and the shareholders (principal) that might arise because of the separation between ownership and control.

This chapter illustrates the following sections as (4.1) addresses the model and predictions, section (4.2) presents the regression results, and finally section (4.3) discusses an estimation and testing results.

4.1 The model and predictions

4.1.1 The model

This study will investigate the interaction between the leverage and agency cost using the following model:

LEV it= β0 + β1 DPOr + β2 PROFit-1 + β3TANGit-1 + β4GRTHit-1 + β5 SZit-1+ β6 LIS +-uit

Where: leverage (LEV) is measured by total debt to total asset (is the firm’s total debt divided by its total assets where the total debt consists of short term and long term debt) (Adedeji 1998, and 2002), Allen (1993), Baskin (1989), finally (Bevan and Danbolt 2002, and 2004) suggest the same indicator. (DPOr) is the dividend payout ratio, (PROF) is profitability; (TANG) is fixed assets ratio; (MBr) is the market to book ratio; (SZ) is firm size and measured as the log of total assets; and (LIS) is the percentage of shares measured as who owned over than 5%.

4.1.2 The predictions

In this section, the predictions relating to each variable in the model will be developed according to the existing theories and the literature review.

4.1.2.1 Dividends (DPOr)

Firms with a reputation for paying dividends face less asymmetric information when they enter to the equity market. Dividend payment represents a signal of improved financial health and hence of more debt issuing capacity (Bhaduri, 2001; John and Williams, 1985; Miller and Rock, 1985). This argument is supported by the signaling theory of capital structure (leverage). The descriptive statistics show that Jordanian firms paid very low dividends where the mean value is 0.025. This may suggest that any dividend payment might be considered as a positive signal in the market. Furthermore, the dividend payout policy is one of the measures available to managers for controlling agency behaviour.

However, the agency theory of capital structure (leverage) suggests a negative relationship between leverage and dividend. Furthermore, Jensen and Meckling (1976) and Jensen (1986) confirmed that the leverage and dividend are a substitute mechanism for mitigating the agency cost of free cash flow.

4.1.2.2 Firm profitability (PROF)

Tong and Green (2005) and Baskin (1989) argued that the high profitable firms are less likely subject to the bankruptcy risk, because these firms will be able to meet its debts repayment obligation, so the probability of bankruptcy for high profitable firm is expected to be low, which in turn reduces the cost of capital and enable them to raise more debts.

More profits means more funds under management control, which may induce mangers to consume more perquisites, or to invest in less profitable project to pursue their own benefits. Therefore, high profitable firms must employ higher leverage to reduce the amount of cash available for possible over investments or perquisites consumption since debt will shift the discipline action to the capital market (Jensen, 1986).

Under this argument, a positive relationship between profitability and leverage is expected. Moreover, Fama and French (1998), analysing that the high leverage degree generates agency problems between the shareholders and creditors. Therefore, that predicts negative relationship between the leverage and profitability.

Some literatures confirm these arguments, such as, Titman and Wessels (1988), Jensen et al. (1992), Allen (1993) and Adedeji (1998). This study uses the ratio of the earnings before interest and, tax (EBIT) to total assets as a measure of firm profitability.

4.1.2.3 Asset structure (TANG)

From the agency theory perspective, the potential agency problems for companies arise between shareholders and the managers with the separation of ownership and control that occurs with the development of the joint stock form of organization (Jenson, and Meckling, 1976). Long and Malitz (1992) argued that the tangible assets using as an arrangement placed as a fix charge. Therefore that would reduce adverse selection and moral hazard costs.

Grossman and Hart (1982) argued that the agency cost of consuming more than the optimal level of perquisites is higher for firms with lower levels of assets that can be used as collateral. Hence, firms with less collateralized assets are more vulnerable to such risk since monitoring the capital outlays is more difficult for such firms (higher monitoring costs) (Bhaduri, 2001). Therefore, one would expect a negative association between debt and collateralized assets.

Firm with collateralized assets can restrict such behaviour, for this reason, (Bhaduri, 2001) expected that there is a positive association between collateralized assets and debt. Hutchinson and Hunter (1995) argued that the asset structure could have a negative impact on capital structure.

The study uses the ratio fixed assets to total assets as a measure of asset structure (tangibility). Rajan and Zingales (1995), Bevan and Danbolt (2002) have used this ratio to measure the asset structure.

4.1.2.4 Growth (MBr)

Agency problems are likely to be more severe for growing firms, because they are more flexible in their selection of future investments. Thus, the expected growth rate could be negatively related to long term leverage (Titman, and Wessels, 1988). Myers (1977) suggested that firms with higher growth rates tend to use less long term debt and more short term debt in their capital structure in order to reduce agency costs. 

On the contrary, Green (1984), Jensen and Meckling (1976), and Smith and Warner (1979) suggested that the convertible debt would mitigate the agency cost. Thus, this implies a positive relationship between the convertible debt and growth opportunities.

4.1.2.5 Ownership structure (LIS)

Institutional shareholders are considered to be experts in collecting and explaining information relating to firms in which they invest. The agency theory of capital structure suggests that the optimal capital structure and ownership may be used for reducing the agency costs.

Jensen and Meckling (1976) and Jensen (1986) suggested that there is a negative relationship between leverage and institutional shareholders (ownership structure).

4.1.2.6 Firm size (SZ)

Literatures in leverage find a positive relationship between leverage and firm size. In the context of agency cost theory, it has argued that larger firms are likely to have lower agency costs associated with the asset substitution and under investment problems (Chung, 1993), because small firms may not have larger institutional follow up. Therefore, moral hazards are expected to be more severe for these firms than it would be for the large ones. Consequently, managers of small firms are more likely work to avoid the market discipline by depending on internally generated funds to finance new investment opportunities.

Jensen and Meckling (1976) have hypothesized that the total agency costs are increasing function with size. Therefore, the lager the firm size, the higher is the agency costs. This might be attributed to the fact that monitoring task is likely to be more difficult and expensive in a larger organization.

Furthermore, Chittenden et al. (1996) claimed that the moral hazard and adverse selection problems may be greater for small firms because of the closely held nature of their equity, and they may not be required to disclose much information. Thus, they incur significant costs in providing such information to outsiders for the first time. Therefore, monitoring device could be costly for these firms.

Bhaduri (2002) and Booth et al (2001) have used the natural logarithm of the total assets as a proxy to measure the firm size.

In addition, this section establishes the model for the leverage, and also shows the predictions of the relationship between the leverage, and the dividends payout ratio, profitability, asset structure, large institutional shareholders, and the firm size on the other hand; based on the past literatures. The rest of this section shows regression results, and then, the paper explores the estimation and testing the results.

4.2 Regression results

In this section, the regression results of the leverage model will report the regression analyses including the analysis of the pooled and the panel models, as mentioned in section (3.3). The data includes Jordanian manufacturing firms for the period from 2002 to 2007. The sample includes 52 firms, which provides the required financial information without any significant gaps. (Table.2) shows the correlation matrix between the variables. The following main comments are:

There is a positive correlation between the dependent variable (LEV) and the following variables; dividend payout (DPOr), tangibility (TANG), and growth (MBr). This means that there is a direct relationship between the leverage of the firm and each one of these variables.

On the contrary, there is a negative correlation between the leverage ratio and the profitability (PROF), firm size (SZ), and the institutional shareholders (LIS). This indicates that the higher the profitability the lower the leverage ratio and vice versa.

(Table.2) Correlation matrix between variables

4.2.1 Regression analysis of the pooled and the panel models

After analysing the bivariate correlations the next step is to investigate the regression models. (Table.3) shows the regression analysis of the pooled and the panel models. The first column reports the pooled model (OLS), where the assumption of no specific group effects has made. The second and the third columns show the panel models, the fixed and random effects models, respectively.

The fixed effects model embodies the panel effect by allowing the intercept to vary across groups. However, the random effects model has a unique composite error. The R-square (R2) value indicates the explanation power of the model. The highest R-square (R2) of 67.5% is for the random effects model, which means that 67.5% of the variation in the dependent variable is explained by the model. The pooled model has R-square (R2) of 68.9%, and the fixed effects model has R-square (R2) of 65.4%.

The Lagrange Multiplier (LM) test is a test statistic with the null hypothesis that the pooled model is appropriate rather than the random effects model. The Lagrange Multiplier statistic is 223.56 and statistically significant. This means that the panel models are more appropriate for the data than the pooled model. To test between the fixed effects model and the random effects model the Hausman test has used. The Hausman test Chi-square (Chi2) is 6.26 and not statistically significance. This indicates that the random effects model is the most appropriate model for the data. However, the acceptance of random effects model does not necessary mean that the fixed effect is not consistent.

tes: All independent variables have taken by one year lag. Leverage (TD/TA) is the total debt over total assets. Dividend payout ratio (DPOr): Dividends to EBIT, Profitability (PROF) measure by EBIT over total assets. Tangibility (TANG) measured as fixed assets over total asset. Size (SZ) the log of total assets. Growth (MBr) is market to book ratio.

The following three conclusions can be drawn base on the (Table.3):

There is a significant negative relationship between profitability and leverage. While there is a significant positive relationship has found between dividend payments, and asset structure, firm size, and

There is an insignificant a negative relationship between institutional ownership (LIS) and leverage (LEV). Furthermore, there is insignificant relationship between the growth (MBr) and leverage, but positive.

The R-square (R2) statistic shows that the random effects model has the highest explanation power with 69%.

The study uses the variance inflating factor (VIF) to test the multicollinearity of the model, the result shows the variance inflating factor for the overall model is 1.23. This finding suggests that multicollinearity problem does exist (Gujarati, 2003)[4]. The pooled models and the fixed effects models are corrected for heteroskedasticity using Breusch-Pagan, and White methods, respectively. Thus, the models in this chapter do not suffer from multicollinearity and heteroskedasticity. 

4.3 Estimation and testing results[5]

Section (4.3) shows the regression analysis. The following is the analysis for the interactions between the leverage and the other variables impact on the agency cost for the Jordanian manufacturing companies. The results have taken from the best models according to Lagrange Multiplier (LM) and Hausman test.

4.3.1 Dividends (DPOr)

The results indicate that there is a positive significant relationship between dividend pay out ratio and the leverage of the firm. However, the study finds strong evidence of the positive relationship between dividend payments and the leverage of the firm. This relationship does not support the prediction that there is a negative relationship between the dividend payments and leverage.

The result is consistent with the signalling theory not the agency theory, the dividend payment is considered as a signal of good performance which provides such Jordanian manufacturing firms with the ability to borrow more. Moreover, Bhaduri (2001), John and Williams (1985), and Miller and Rock (1985), confirmed the same results that indicate the positive relationship among leverage and dividends payout policy.

4.3.2 Profitability (PROF)

The results indicate that there is a strong significant negative relationship between profitability and leverage. This means that the Jordanian manufacturing firms prefer internal financing (equity) rather than debt financing.

This result is in term with the pecking order theory of capital structure (leverage). Capital structure literatures find the same result such as Jensen et al. (1992), Rajan and Zingales (1995), Booth et al. (2001), Ozkan (2001) and Huang and Song (2005).

4.3.3 Asset structure (TANG)

The results show that there is a positive significant relationship between the tangible assets and leverage. This result indicates that the Jordanian manufacturing firms using fixed asset by pledging the assets as collateral. In addition, this results support the agency theory of leverage. 

Jensen and Meckling (1976), Titman and Wessels (1988), Rajan and Zingales (1995), Bhaduri (2001), and Booth et al. (2001), confirmed the same results that indicate the positive relationship between the leverage and asset structure (Tangibility).

4.3.4 Growth (MBr)

The results show that there is a positive insignificant relationship between the potential growth rate and leverage, and this contradicts the expected negative sign which is predicted by the agency theory. This result does not accept the prediction that the negative relationship between the growth opportunities and leverage. 

This means that the Jordanian manufacturing firms with high growth opportunities tend to face different financing alternatives, and they prefer debt financing as a way to finance their investment opportunities. In addition, one can argue that such firms have a low probability of bankruptcy. Hence, they have more access to debt financing than low growth firms. This result is consistent with Bhaduri (2002), Booth et al. (2001), Cassar and Holmes (2003), and Voulgaris et al. (2004).

4.3.5 Ownership structure (LIS)

The results show an insignificant negative relationship between the leverage and the institutional shareholders. However, the positive sign is consistent with the agency theory of leverage. This means that institutional owners became significant monitors of the firm’s managers, to reduce the agency costs by forcing them to create more debt to reduce the amount of cash under management control.

This result is consistent with Leland and Pyle (1977), Berger et al. (1997), and Chen and Steiner (1999) who showed that managerial ownership and leverage are positively related. Here one can argue that there is a dual effect for the institutional owners, the role of agents to monitor the firm to minimize the agency costs, and the role of insider or managerial ownership.

4.3.6 Firm size (SZ)

The results show that there is a significant positive relationship between firm size and leverage. This means that large Jordanian manufacturing firms tend to be more diversified and hence less likely to be susceptible to financial distress.

Moreover, this result supports the bankruptcy theory of leverage. Other studies in the financial literature have the same result such as; Rajan and Zingales (1995), Booth et al. (2001), Bhaduri (2002), Cassar and Holmes (2003), Voulgaris et al. (2004), and Huang and Song (2005).

5. Empirical Evidence for the Dividend Payout ratio

Many studies have explained the corporate dividend policy, also researchers have argued that the dividends policy may have real economic consequences due to signalling (Ross, 1977), and (Bhattacharya, 1979) and clientele effects (Pettit, 1977), and (Lewellen et al., 1978). Moreover, Easterbrook (1984), and Rozeff (1982) argued that the dividends payout policy also impacts the agency costs. Miller and Modigliani (1961) viewed dividend payout policy as irrelevant. Moreover, this paper defines the dividends payout policy as the dividends divided by EBIT. Dividends payout policy would mitigate the agency costs through increased monitoring by capital market.

This section explores the determinants of the dividend policy in Jordanian manufacturing firms, and whether these firms follow the same determinants of dividend policy as suggested by the theories of dividend policy that have been investigated whether, in developed and developing economic countries.

5.1 The model and predictions

5.1.1 The model

This study will investigate the interaction between the dividends and the agency cost theory using the following model:

DPOrit= β0 + β1 LEV + β2 PROFit-1 + β3TANGit-1 + β4GRTHit-1 + β5 SZit-1+ β6 LIS +-uit

Where: Dividends (DPOr) is measured by dividends to EBIT, (LEV) is the total debt to total assets; (PROF) is the EBIT to total assets; (MBr) is the market to book ratio; (TANG) is the fixed assets to total assets; (LIS) is the percentage of shareholders who owned more than 5%; and (SZ) is the firm size measured as the log of total assets.

5.1.2 The predictions

In this section, the predictions of the dividend policy have been developed according to the literature and the theories of the dividend payout policy

5.1.2.1 Leverage (LEV)

Literatures have argued that leverage considered as being a one of the important factors that may affect the dividend policy of the firm. Further, studies in this field have found different results about how financial leverage might affect the dividend payout decision. Leverage negatively related to dividends, this means that the firms with low debt ratio tend to pay more dividends. Aivazian et al., (2003) argued that there is relationship between dividend payments and leverage, his arguments states: “Firms with relatively less debt and more tangible assets have greater financial slack and more able to pay and maintain their dividends”. This result support the agency costs theory of dividend policy. Jensen et al. (1992) also confirmed this argument.

On the contrary, literatures argued that there is a positive relationship between leverage and the dividend policy, this result supports the signaling theory. “Firms with high payout ratio tend to debt financed, while firms with low payout ratios tend to equity financed” (Chang and Rhee, 1990). This study either hypothesizes that there is a positive relationship between the leverage position of the firm and the dividend payments, because if the firm has a reputation for paying dividends, this will give a signal of a healthy financial position, and banks tend to lend them at a preferable interest rate.

Moreover, the positive relationship between leverage and dividend payments infers a negative relationship between leverage and retention rate which implies the preference of retentions to debt financing. Hence, this result in terms of the pecking order hypothesis. This study uses the ratio of total debt to total equity to measure the leverage for the Jordanian manufacturing company. 

5.1.2.2 Firm Profitability (PROF)

Profitability is a significant factor that affects on the dividend payout policy at the firm. Signaling theory suggests that there is a positive relationship between the profitability of the firm and dividend payments. Empirical studies indicate a positive relationship between the dividend payout policy and profitability. This explains that the highly profitable firms tend to pay high dividends. Thus, they would be lead to high payout ratio. Baker et al. (1985), and Pruitt and Gitman (1991) support this positive relationship between the profitability and dividends payout policy. This study uses the rate of return to measure the profitability of the Jordanian manufacturing firms.

5.1.2.3 Asset Structure (TANG)

Literatures have argued that the measure of tangibility is one of the determinants of dividend payout policy. Studies also attained different results regarding the tangible assets. They suggested there is a positive relationship between asset structure (tangibility) and dividend payout policy, the higher tangible assets, tends to use assets as collateral. Thus, they reduce the agency costs of debt (Rajan and Zingals, 1995).

On the contrary, Aivazian et al. (2003) found a negative relationship between asset tangibility and the dividend policy, that explanation of the results states: “when the assets are more tangible, fewer short-term assets are available for banks to lend against. This imposes financial constraints on firms operating in more primitive financial systems, where the main source of debt is short-term financing”.

Furthermore, (Ho, 2003) supported the negative relationship between the asset structure and dividend payout policy using agency the theory of dividend policy. This study uses a ratio of fixed assets to total assets as a proxy of tangibility of the Jordanian manufacturing firms.

5.1.2.4 Growth (MBr)

Higher growth rate implies that the firm attracts a high opportunities for investment. Further, it would be enhanced for the company to finance their investments, and more likely to retain earnings than pay it as dividends (Chang and Rhee, 2003). Moreover, Myers and Majluf (1984) found a negative relationship between the growth and dividend payout policy.

However, they suggest that firms with high growth opportunities will have low dividend payout ratios and thus this result has supported by the pecking order theory. This negative relationship between growth rates and dividends was supported by the agency theory of dividend policy Holder et al. (1998), Gul (1999), Chang and Rhee (2001), Ho (2003), and Aivazian et al., (2003) confirmed the negative relationship between the growth and dividend payout ratio. This paper uses market to book ratio as an indicator of the growth opportunities of a firm.

5.1.2.5 Ownership Structure (LIS)

Large institutional shareholders considered as a one of the important factors that may affect the dividend policy decision, either the institutional shareholder may act as a monitoring device, therefore, will reduce the use of capital markets as an external monitoring system (Zeckhauser and Pound, 1990). Jensen (1986), Zeckhauser and Pound, 1990, and Short et al. (2002) found that there is a positive relationship between the institutional shareholders and dividend payout policy.

Moreover, In terms of signalling theory there is a negative relationship between institutional ownership and dividend pay out ratio. This is because dividends and institutional ownership are alternative signalling devices. The existence of institutional ownership mitigates the need for dividends to signal good performance (Short et al., 2002). The study uses the percentage of the institutional shareholders who owned more than 5% at the firm.

5.1.2.6 Firm Size (SZ)

A firm size plays an important role for the shareholders as a proxy if the company is able to declare or pay dividends. This indicates that, the larger firms can afford to pay higher dividends than the smaller firms. Holder et al. (1998), Gul (1999), Koch and Shenoy (1999), and Chang and Rhee (2001) support that; there is a positive relationship between the firm size and the dividend payout policy by using the transaction cost explanation of dividend policy. This study measures the firm size as a logarithm of total assets for the Jordanian manufacturing firms.

This section investigates the relationship between the dividends payout policy and the other variables based on the model for the dividends. In addition, it predicts how these relationships impact the agency cost for the Jordanian manufacturing firms. The rest of this section includes the regression results and finally show the estimation and testing results section.

5.2 Regression results

This section presents the regression results of the estimated model; the regression analysis includes the analysis of pooled and panel models. The data is composed of Jordanian manufacturing firms for the period from 2002 to 2007. The sample of the study consists of 52 firms which provide the required financial information without any significant gaps. (Table.4) shows the correlation matrix between the variables, the following correlations are below:

The following are the main comments on the correlation matrix between the variables:

There is a positive correlation between the dependent variable (DPOr) and the following variables; profitability (PROF), and the firm size (SZ). This means that there is direct relationship between dividend payout policy and each of these variables.

On the contrary, there is a negative correlation between the dividend payout ratio, the growth (MBr), leverage (LEV), assets structure (TANG), and the institutional shareholders (LIS) on the other hand.

5.2.1 Regression analysis for the pooled and the panel models

After analysing the bivariate correlations, the next step is to investigate the regression models of the dividend policy. (Table.5) shows the regression analysis of the pooled model and the panel models.

(Table.5) OLS, Fixed and Random effect estimation results

Notes: All independent variables have taken by one year lag. Leverage (TD/TA) is the total debt over total assets. Dividend payout ratio (DPOr): Dividends to EBIT, Profitability (PROF) measure by EBIT over total assets. Tangibility (TANG) measured as fixed assets over total asset. Size (SZ) the log of total assets. Growth (MBr) is market to book ratio.

(Table.5) shows the following conclusions:

There is a significant positive relationship between the leverage (LEV), profitability (PROF), and the firm size (SZ) on one hand, and the dividend payout policy (DPOr) on the other hand. While there is an insignificant positive relationship between the dividends payout and the growth opportunities (MBr) at the Jordanian manufacturing firms.

On the contrary there is an insignificant negative relationship between tangibility (TANG), large institutional shareholders (LIS), and the firm size (SZ) on one hand, and the dividend payout policy on the other hand.

The R-square (R2) statistic shows that the random effects model has the highest explanation power with 18.38%, while 11.40% for fixed effect model.

The variance inflating factor (VIF)[6] has used in the model to test for multicollinearity; the result shows that all the variables have coefficients of variance inflating factor (VIF) less than 10.

The pooled models and the fixed effects models were corrected for heteroskedasticity using Breusch-Pagan. Thus, the models in this chapter do not suffer from multicollinearity and heteroskedasticity. Moreover, the Lagrange Multiplier (LM) test shows that the panel model is more favourable than the pooled model where the Lagrange multiplier is found 23.40 and significant. Therefore, this indicates that the random effects model is the favourable one, while Hausman test shows that the random effects model is favourable than fixed effects model, when panel data analysis is used. The value of Hausman test is statistically insignificant. Hence, we could not reject the null hypothesis that the random effect is the best specification for the dataset. 

5.3 Estimation and testing results

Section (5.3) presents the regression analyses using pooled and panel models. The following are the main determinants of firm’s dividend policy.

5.3.1 Leverage (LEV)

The results indicate that there is a positive significant relationship between the leverage of the firm and the dividend payout ratio. This finding does not support prediction that the leverage and dividend payout are negatively related. This result generally supports the pecking order theory not the agency theory. This implies that the shareholders are not strong enough to force managers who have free cash flow to pay dividends. Another explanation might arise due to the fact that the Jordanian manufacturing firms have no more free cash flow. Therefore, paying dividend may require these firms to generate more debt to finance their investment opportunities.

5.3.2 Profitability (PROF)

The results indicate that there is a positive relationship between profitability and dividend payout ratio. This means that the profitable Jordanian manufacturing firms are more likely to pay dividends for their shareholders.

This result supports the signaling theory of the dividend policy. Therefore, the more profitable firm has a chance to pay dividends. Moreover, the pecking order hypothesis suggests that firms finance investment opportunities in a certain order; first, with the retained earnings, second, with debt financing finally, from external financing sources (Myers 1984; Myers, and Majluf, 1984). Thus, profitable firms will find it more significant to pay dividends and to have more retained earnings (the pecking order theory). Jensen et al. (1992) and Aivazian et al. (2003) confirm the same results.

5.3.3 Asset Structure (TANG)

The results show that there is strong evidence of a negative insignificant relationship between the tangible assets and dividend payout ratio. This negative relationship has supported, by the agency theory of dividend policy (Ho, 2003). More tangible assets in the firm, the lower size of the short-term assets which can used as collateral for short term debt.

Therefore, firms will depend more on their retained earnings, which means that the lower chance to pay dividends. This result supports the prediction for the negative relationship between the asset structure (tangibility) and dividend payout policy.

5.3.4 Growth (MBr)

The regression results to some extent support the argument that the growth plays an important role in the dividend policy decision. The results show there is a positive relationship between the potential growth rate and dividend payout ratio. This result is consistent with the expected negative predicted by the agency theory.

This means that Jordanian manufacturing firms with high growth opportunities restrict their ability to generate debt, because growth opportunities have no collateral value, and therefore that tends to increase agency costs of debt. This suggestion will turn those firms to internal funds for financing instead of paying dividends. 

5.3.5 Ownership structure (LIS)

The results show an insignificant negative relationship between the percentage of institutional ownership and dividend. This result is consistent with the signaling theory of dividend policy. This result confirm of the prediction that there is a negative relationship between the institutional shareholders and dividend payout policy.

This means that the dividends and the ownership structure are alternative for signaling devices. The existence of large institutional shareholders mitigates the need for dividends to signal good performance. Jensen et al. (1992) confirmed the same result when they use the insider ownership as a factor of ownership structure. 

5.3.6 Firm size (SZ)

The results indicate that there is a positive relationship between firm size and dividend payout ratio. This means that large Jordanian manufacturing firms tend to followed by large institutional shareholders. This result is consistent with the signaling theory of dividend policy.

6. Summary and findings 

The purpose of this research is to investigates the role of the agency theory, related with the determinants of capital structure (leverage) and the dividends payout policy for the Jordanian manufacturing companies, Further, this research predicts the relationships between the financial policies as a dependent variables, with the firm profitability, growth, assets structure (tangibility), firm size, and large institutional shareholders used as independent variables.

In addition, the research illustrates empirically the relationship among the variables based on the sample of the study.

Over the years the capital structure theories have been developed in order to determine the factors that influence on the corporate capital structure decisions. Some of these theories have been dominated by search the optimal capital structure. (Table.6) will summarize the corporate financial policies theories.

(Table.6) Summary of the main corporate finance theories

Agency cost theory

The agency cost theory developed by Jensen and Meckling (1976), this theory investigates the relation ship between the managers (agents) and the principals (shareholders). This theory has been debated under the uncertainty conditions and inadequate information asymmetry. Agency costs include the monitoring, bonding and structuring a form of contracts between agents with conflicting interests, plus the residual costs incurred because the cost of enforcement of contracts exceed the benefits (Jensen and Meckling, 1976).

Furthermore, the agency cost caused by the separation of ownership and control. However, the separation of ownership and control leads to potential conflict of interests between the manager and shareholders. Therefore, the higher separation of ownership and control, will raise the cost of agency between the managers and owners, that would need to increase the monitoring by shareholders.

According to the issuance of corporate capital structure, Myers (1984) and Jensen and Meckling (1976) confirm that the external financing has agency costs as well. Therefore, it would raise the conflicts between the shareholders and lenders. Further, the benefits for issuing debt from the reduction in agency cost of external equity financing.

Signaling theory

Under the signalling theory, there is a different behavioural model of dividend policy have been used to make inferences about information asymmetry and agency confliscts between the managers and shareholders (Dewenter and Warther, 1998), and also Kao and Wu (1994) argued that the dividends play a role in signalling information about future earnings. Therefore, the corporate mangers use the dividends as a signal about the firms’ future earnings.

Moreover, the leverage and dividend use as mechanism of signals. Firms depend on the dividend signaling policy may be associated with the high ratio of leverage compared to non signalling firms. Ross (1977) developed a signaling model for capital structure (leverage), he predicts that the higher leverage will be associated with higher cash flow. As a result, the capital structure and dividend indicate that interact to provide a significant information (signal) about future earnings.

Trade off theory

This theory asserts that the firms set a target debt to value rate and gradually move towards it. Furthermore, the bankruptcy, financial distress (Myers, 1977) and agency costs (Jensen and Meckling, 1976) constitute the trade off theory.

However, there is a positive relationship between those costs and the level of debt. Therefore, if the level of debt increases then the bankruptcy, financial distress, and agency cost will increase, and hence decreases the firm value.

Under this theory, to reach an optimal capital structure should establish equilibrium between the tax shields as an advantage for the firm, and bankruptcy, financial distress, and agency costs of debt. In order to establish this equilibrium firms should seek debt levels at which the costs of possible financial distress offset the tax advantages of additional debt.

Pecking order theory

The pecking order theory is an alternative of trade off theory, originally developed by Myers and Mjluf (1984). They assumed that there is asymmetric information between the investors. In general, investors have less information than insiders; common stock would be undervalued by the capital market.

Furthermore, firms do not have target capital structures. This states that firms prefer internal financing, instead of external sources and debt to equity if they issue securities. Myers (1984) provides that the firms adapt its target dividend payout ratios to the investment opportunities, though the dividends here is sticky, then the payouts adjusted to shifts in the positive projects.

Moreover, he also argued that sicky dividends plus instability in profits, and investment opportunities mean that the cash flow in the firm inadequate. In this case the firm draws on its cash and marketable securities.

Further, the external financing for the company contains the debt, then hybrid, then the equity as a last source. Therefore, Myers (1984) and Myers and Majluf (1984) concluded that the pecking order theory has no optimal debt ratio. Because of asymmetric information and signalling problem would be associated with the external financing. As a result, the pecking order theory emphasises the asymmetric information while the trade off theory underlines the taxes.

Bibliography

Adedeji, A. (1998). Does the Pecking Order Hypothesis Explain the Dividend Payout Ratios of Firms in the UK. Journal of Business Finance and Accounting, 25, 127-157.

Allen, D. (1993). The Pecking Order Hypothesis: Australian Evidence. Applied Financial Economics, 3, 101-120.

Al. Najjar, B. and Taylor, P. 2008, The relationship between capital structure and ownership structure: new evidence from Jordanian panel data, Managerial Finance, 32, pp. 919-933.

Aivazian, Varouj, Laurence Booth, and Sean Cleary, 2003b. Dividend Policy and the Organisation of Capital Markets. Journal of Multinational Financial Management,13, pp.101-121.

Amidu, M. and Abor, J., (2006), Determinants of Dividend Payout Ratios in Ghana, The Journal of Risk Finance, 7, pp. 136-145.

Baker, H.K., Farrelly, G.E. and Edelman, R.B. (1985), “A survey of management views on dividend policy”, Financial Management, Vol. 14 No. 3, pp. 1007-1034.

Baskin, J. (1989). An Empirical Investigation of the Pecking Order Theory. Financial Management, 18, 26-35.

Berger, P., Ofek, E. and Yermack, D. (1997), ‘‘Managerial entrenchment and capital structure decisions’’, Journal of Finance, Vol. 52, pp. 1411-38.

Breusch T.S. and Pagan, A.R. 1980, The Largange Multiplier Test and its Applications to Model Specification in Econometrics. Rev. Econ. Stud. 47 pp. 239-253.

Bevan, A. & Danbolt, J. (2002). Capital Structure and Its Determinants in the UK – A Decomposition Analysis. Applied Financial Economics, 12(3), 159-170.

Bevan, A. & Danbolt, J. (2004). Testing for Inconsistencies in the Estimation of UK Capital Structure Determinants. Applied Financial Economics, 14(1), 55-66.

Bhattacharya, S., (1979), Imperfect information, dividend policy, and “the bird in the hand” fallacy, Bell Journal of Economics and Management Science, 10(1), 259 - 270.

Booth, L., Aivazian, V., Demirguc-Kunt, A. & Maksimovic, V. (2001). Capital Structures in Developing Countries. Journal of Finance, 56(1), 87-130.

C. Crutchley and R. S. Hansen 1989, "A Test ofthe Agency Theory of Managerial Ownership, Corporate Leverage, and Corporate Dividends" Financial Management, pp. 36-46.

Chang, R.P. and Rhee, S.G., 1990. The Impact of Personal Taxes on Corporate Dividend Policy and Capital Structure Decisions. Financial Management, 19(2), pp. 21-31.

Chen, C. and Steiner, T. (1999), ‘‘Managerial ownership and agency conflicts: a nonlinear simultaneous equation analysis of managerial ownership, risk taking, debt policy, and dividend policy’’, Financial Review, Vol. 34, pp. 119-36.

Chung, K. H., 1993. Asset characteristics and corporate debt policy: an empirical test. Journal of Business Finance and Accounting, 20, 83-98.

Dewenter, K. L., and V. A. Warther, 1998, Dividends, Asymmetric Information, and Agency Conflicts: Evidence from a Comparison of the Dividend Policies of Japanese and U.S. Firms. Journal of Finance, 53 ,879-904.

Easterbrook, F., 1984. Two Agency-cost Explanations of Dividends. American Economic Review, 74, pp. 650-659.

Eisenhardt, K. M., 1989. Agency Theory: An Assessment and Review. Academy of Management Review, 14(1), pp. 57-74.

EL-Khouri, R. and Hmedat, W. (1992). The Effect of Earning Variability on Capital Structure: The Case of Jordanian Firms. Abhath Al-Yarmouk, 8(3), 49-64.

Fama, E. F., and Jensen, M. C. (1983). Separation of ownership and control. Journal of Law and Economics, 26, 288–397.

Fama, E. F., and French, K. R., 1998, Taxes, Financing Decisions, and Firm Value. Journal of Finance, 53 (3), pp.

Frankfurter, G. M. and Wood, Jr., 2002. Dividend Policy Theories and Their Empirical Tests. International Review of Financial Analysis, 11, pp. 111-138.

Garret I., and Priestley R., 2000, Dividend Behaviour and Dividend Signaling, Journal of Financial and Quantitative Analysis, 35, pp. 173-189.

Gujarati, Damodar N., 2003. Basic Econometrics, 4th ed., McGraw-Hill.

Grossman, S. & Hart, O. (1982). Corporate Financial Structure and Managerial Incentives. The Economics of information and Uncertainty, Edited by J.McCall. Chicago: University of Chicago press, 107-137.

H. Herwartz, 2006, Testing for random effects in panel data under cross sectional error correlation – A bootstrap approach to the Breusch Pagan test, Computation Statistics & Data Analysis. 50, pp. 3567- 3591.

Hansen, R.S., R. Kumar and D.K. Shome, (1994), Dividend policy and corporate monitoring: Evidence from the regulated electric utility industry, Financial Management, 23(1), 16-22.

Harris, Milton & Raviv, Artur, (1990). Capital Structure and the Informational Role of Debt. Journal of Finance, 45, 321-349.

Harris, M. & Raviv, A. (1991). The Theory of Capital Structure. Journal of Finance, 32, 297-356.

Ho, H., 2003. Dividend Policies in Australia and Japan. International Advances in Economic Research, 9(2), pp. 91-100.

Hutchinson, R.W., and Hunter, R. L., 1995. Determinant of capital structure in the retailing sector in the UK. The International Review of Retail, Distribution and Consumer Research, 5, 1, 63-78.

Huang, S. G. and Song, F. M., 2005. The Determinants of Capital Structure: Evidence from China. China Economic Review, pp. 1-23. 

J.F. Hair, R.E. Anderson, R.L. Tathan and W.C. Black, 1998, Multivariate Data Analysis 3rd ed. MacMillan Publishing Co.

Jensen, M. & Meckling, W. (1976). The theory of the Firm: Managerial Behaviour, Agency Costs, and Ownership Structure. Journal of Financial Economics, 3(4), 305-360.

Jensen, M. (1986). Agency Cost Free Cash Flow, Corporate Finance, and Takeovers. American Economic Review, 76(2), 323-329.

Jensen,.M, Geralg, P, & Thomas, S. 1992. Simultaneous Determination of Insider Ownership, Debt, and Dividend policies. Journal of financial and Quantitative Analysis, 27, 247-263.

John, K., and J. Williams, 1985 “Dividends, Dilution, and Taxes: A Signaling Equilibrium,” Journal of Finance pp. 1053-1070.

Kao, C, and C, Wu, Tests of Dividend Signaling using the Marsh-Merton Model: A Generalized Friction Approach, Journal of Business, 67, 45-68

Karadeniz, E., Kandir, S., Balcilar, M., and Onal, Y., 2009, Detreminants of capital structure: evidence from Turkish lodging companies, International Journal of Contemporary Hospitability Management, 21 pp. 594-609

Koch, P., and Shenoy, C.,1999, The Information Content of Dividends and Capital Structure Policies, Financial Management, 28, pp. 16-35.

Lasfer, M.A. (1995). Agency Costs, Taxes and Debt: The UK Evidence’ European Financial Management, 1(3), 265-285.

Leland, H. E. and Pyle, D. H., 1977. Informational Asymmetries, Financial Structure, and Financial Intermediation. Journal of Finance, 32(2) pp. 371-387.

Lester V.M. (1995), Victorian Insolvency, Bankruptcy, Imprisonment for Debt, and Company

Winding-Up in Nineteenth-Century England, Oxford, Clarendon Press, 354p.

Lewellen, W., K. Stanley, R. Lease, and G. Schlabaum, 1978, Some direct evidence on the dividends clientele phenomena, Jouranl of finance, 33 pp. 1385-99

Long, M. and Malitz, I., 1992. The investment-financing nexus : some empirical evidence. In J. Stern and D. Chew (eds) The Revolution in Corporate Finance, Oxford: Blackwell, 156-162.

Miller, M., and K. Rock, 1985, Dividend Policy Under Asymmetric Information, Journal of Finance, pp. 1031-1052.

Manos, R., 2001 Capital Structure and Dividend Policy: Evidence From Emerging financial Markets, Unpublished PhD thesis, University of Birmingham.

Modigliani, F. and Miller, M.H., 1958. The Cost of Capital, Corporation Finance and the Theory of Investment. American Economic Review, 48(3), pp. 261-297.

Modigliani, F. and Miller, M.H., 1963. Corporate Income Taxes and the Cost of Capital: A Correction. American Economic Review, 53(3), pp. 433-443.

Mougoue, M., and Rao, R. 2003, The Information Signaling Hypothesis of Dividends: Evidence From Cointegration and Causality Tests. Journal of Business Finance, and Accounting, 30, pp. 441-477.

Myers, S.C., 1977. Determinants of Corporate Borrowing. Journal of Financial Economics, 5,pp.147-175.

Myers, S.C., 1984. The Capital Structure Puzzle. Journal of Finance, 39(3), pp. 575-592.

Myers, S.C. and Majluf, N.S., 1984. Corporate Financing and Investment Decisions When Firms Have Information That investors Do Not Have. Journal of Financial Economics, 13, pp.187-221.

Omran, M and Pointon, J., 2001.Does Dividend Policy, Trading Characteristics and Share Prices: Empirical Evidence From Egyptian Firms. International Journal of Theoretical and Applied Finance, 7(2),pp 121-133.

Ozkan, A., 2001. Determinants of Capital Structure and Adjustment to Long Run Target: Evidence From UK Company Panel Data. Journal of Business Finance and Accounting, 28(1/2), pp.175-198.

Pettit, R., 1977, Taxes, transactions costs and clientele effects of dividends, Journal of Financial Economics 5, 419-436.

Pruitt, S.W. and Gitman, L.W. (1991), “The interactions between the investment, financing, and dividend decisions of major US firms”, Financial Review, Vol. 26, pp. 409-430.

Rajan, R. & Zingales, L. (1995). What Do We Know about Capital Structure? Some Evidence from International Data. Journal of Finance, 50, 1421-1460.

Ross, S. (1977). The Determination of Financial Structure: the Incentive-Signalling Approach. Bell Journal of Economics, 8(1), 23-40.

Rozeff, Michael S., 1982. Growth, Beta and Agency Costs As Determinants of Dividend Payout Ratios. Journal of Financial Research, 5, pp. 249-259.

Schooley, Diane K., Barney, Jr., L., 1994, Using Dividend Policy and Managerial Ownership to Reduce Agency Costs, Journal of Financial research, 17 (3) pp. 363-373.

Short, H., Zhang, H. and Keasey, K., 2002. The Link Between Dividend Policy and Institutional Ownership. Journal of Corporate Finance, 8(2), pp. 105-122.

Tong, G. and Green, C.J. (2005), “Pecking order or trade-off hypothesis? Evidence on the capital structure of Chinese companies”, Applied Economics, Vol. 37 No. 19, pp. 2179-89.

Titman, S. and Wessels, R., 1988. The Determinants of Capital Structure Choice. Journal of Finance, 43(1), pp. 1-19.

Voulgaris, F., Asteriou, D. and Agiomirgianakis, G., 2004. Size and Determinants of Capital Structure in the Greek Manufacturing Sector. International Review of Applied Economics, 18(2), pp. 247-262.

Viviani, L.J., 2008, Capital structure determinants: and empirical study of French companies in wine industry.

W.H. Greene. Econometric Analysis 2003 5th ed. Pearson Education.

[1] A model stating that dividend policy has two parameters: (1) the target payout ratio and (2) the speed at which current dividends adjust to the target. (Investopedia.com)

[2] The random effects estimator is efficient and consistent under the null hypothesis, and inconsistent under the alternative hypothesis, and the fixed effects model is consistent under both the null and the alternative hypothesis.

[3] The study used STATA Statistics/Data Analysis (9.2).

[4] Gujarati (2003) states that “As a rule of thumb, if the VIF of the variable exceeds 10, which will happen if exceeds 0.90, that variable is said be highly collinear”

[5] When the dividend payout ratio (DPOr) is used, the results in the previous tables were the same (in terms of the significant values and the signs), except for the DPOr which is insignificant in all cases. In addition, the random effects models are more favourable than the fixed effects model.

[6] The variance inflating factor (VIF) shows how the estimator’s variance is inflated if multicollinearity does exist (Gujarati, 2003).