“Determinants of capital structure: evidence from Jordanian service companies”

This paper examines capital structure determinants for service companies in Jordan between 2014 and 2018. Secondary data from 45 companies were analyzed using the panel regression approach. The results show that the independent variables, suggested as capital structure determinants, have an effect on the debt ratio made by the service companies. Size and non-debt tax shield have a positive significant effect on the debt ratio, while profitability and business risk have a negative significant impact on the debt ratio. In general, the findings support the notion that the trade-off, bankruptcy cost, agency cost and pecking order theories are crucial in explaining the capital structure of Jordanian service companies except for non-debt tax shields and tangibility factors. Jordanian service companies do not use fixed assets as collateral or companies with higher collateral value tend to borrow less debt. Although the coefficient of insti- tutional investors is statistically insignificant, it is still negative and economically significant. This paper concludes that size, profitability, business risk, non-debt tax shields and institutional ownership factors are fundamental in terms of shaping the capital structure in Jordanian service companies. and pecking order developed by Fama and French (2002) and Frank and Goyal (2003), agency costs by Jensen (1986), and market timing by Baker and Wurgler (2002) and Hovakimian (2006). The purpose of this study is to examine the determinants of capital structure in Jordanian service companies listed on the Amman Stock Exchange (ASE) over the period from 2014 to 2018. In addition, to provide evidence to support or contradict some theories, such as trade-off, bankruptcy cost, agency and pecking-order theories, 45 companies from the sector were used. Why does this paper choose to examine the determinants of capital structure in the Jordanian companies in this


INTRODUCTION
The capital structure of a company is considered as one of the main issues of corporate finance, as it reflects most business decisions established by managers, which have an impact on the financial and economic values of the company.In particular, Modigliani and Miller (1958) provided the cornerstone of the capital structure theory.They conclude that market value of a company is not affected by financial leverage.Since then, many researchers have begun to examine the capital structure determinants (Harris & Raviv, 1991;MacKay & Phillips, 2005).In the context of these numerous works, there are many studies that deal with several theories and their relationship with capital structure, such as trade-off and pecking order developed by Fama and French (2002) and Frank and Goyal (2003), agency costs by Jensen (1986), and market timing by Baker and Wurgler (2002) and Hovakimian (2006).
The purpose of this study is to examine the determinants of capital structure in Jordanian service companies listed on the Amman Stock Exchange (ASE) over the period from 2014 to 2018.In addition, to provide evidence to support or contradict some theories, such as trade-off, bankruptcy cost, agency and pecking-order theories, 45 companies from the sector were used.Why does this paper choose to examine the determinants of capital structure in the Jordanian companies in this particular sector?Firstly, the service sector includes the companies responsible for the infrastructure in any country, therefore, these companies have a range of characteristics in this sector, that is, fixed capital is huge in these companies, which helps and facilitates access to credit (collateral value).Seasonality is another special characteristic, especially in hotel and education companies.Secondly, Jordan is an emerging market whose economy is considered as a service economy.Based on the ASE site, the number of companies in the Jordanian service sector is 46 out of 191, which means that this sector is of big size and represents 25% of ASE.Therefore, it deepens the knowledge of financing companies in the Jordanian services sector, which although it is an important part of ASE, does not receive due attention compared to other sectors, for example, financial and industrial sectors.Besides, the results of this study can be compared with those of previous studies, either in the same sector or other sectors in developed and developing countries.
Thirdly, according to the annual report of the Central Bank of Jordan (2017), the service sector contributes to the gross domestic product (GDP) for 32% in 2017 in the Jordanian economy.In addition, according to the same report (Central Bank of Jordan, 2017), the service sector is the most active and operating for the Jordanian workforce, especially in transport companies and educational services.These reasons motivate this study to shed light on examining the determinants of capital structure in companies of the Jordanian service sector and to recognize what factors must be taken into consideration when formulating the capital structure in this sector separately.
The current study uses some different variables, such as the percentage of stock owned by institutional investors from the total ownership of board members (PIO).In particular, this paper is the first to use the PIO to capture the ownership structure as a determinant of capital structure in Jordanian service sector companies.Furthermore, unlike previous studies in Jordan, such as Al-Najjar and Taylor (2008), or in developed countries, such as Bhaduri (2002), Titman and Wessels (1988), Welch (2004) and Gharaibeh and Khaled (2020), who calculate business risk based on the standard deviation of the firm's return on assets, and consistent with the Qiu and La's (2010) method, the current study addresses this issue by assuming that shareholders are essentially concerned about the part of the risk they cannot remove through diversification, which indicates the beta of the companies' stock.Therefore, the current study calculates business risk based on levered beta divided by one plus debt to equity ratio that has not yet been employed in empirical research as the determinants of capital structure in Jordan.This paper does not use the standard deviation of the return on assets as the business risk variable.Although this variable varies from one company to another, it is relatively stable during the period of analysis, and this can cause problem in the fixed effect model.This paper is organized as follows.Literature review is presented in the first section, while section 2 describes data and methodology.Section 3 discusses the results, and the last section concludes the paper.

LITERATURE REVIEW
In an efficient capital market, companies should be indifferent selecting between debt and equity (Modigliani & Miller, 1958).Modigliani and Miller (1963) and Miller (1977Miller ( , 1988) ) reveal that high debt financing leads to higher corporate value when interest costs of debt are tax-free and equity costs are not.The tax benefit for debt financing constitutes 9.7 percent of the total market values of companies in the USA (Graham, 2000).In contrast, a relatively high level of leverage in-creases the risk of bankruptcy.Stiglitz (1974Stiglitz ( , 1988) ) points out that there is a positive relationship between the potential tax-saving benefits and potential bankruptcy costs arising from the use of leverage.Warner (1977) shows that big and profitable companies with essentially tangible assets tend to use more debt to minimize their tax commitments, as these companies have somewhat lower bankruptcy costs that each share incurs.
In the Jordanian stock market, Maghyereh (2005) investigates the determinants of the capital struc-ture of Jordanian manufacturing firms using a dynamic model.He shows that size, profitability, growth opportunity, earning volatility and tangibility have a significant impact on the capital structure of Jordanian firms.Using data for 86 Jordanian non-financial companies, Al-Najjar and Taylor (2008) explore the relationship between capital structure and ownership structure for the period 1994-2003.Pooled and panel regression analyses were used to examine this relationship.The results show that firm size, asset tangibility, business risk and growth opportunity are common factors in determining capital structure and ownership.Khrawish and Khraiwesh (2010) re-investigate the determinants of the capital structure of the industrial companies listed on the ASE over the period 2001-2005.They support the results of previous studies and show that size, tangibility, short-term debt and long-term debt are positively related to leverage, while profitability is negatively related to leverage.
Using 76 Jordanian non-financial companies for the period 2001-2006, Soumadi and Hayajneh (2012) show that there is a negative relationship between profitability and debt.Shubita and Alsawalhah (2012) confirm that profitability is negatively related to debt applied to 39 industrial companies listed on the ASE.In a more recent study, using a sample of 41 Jordanian industrial firms listed on the ASE for ten years (2007 to 2016), Al-Nsour and Jresat (2018) find that the profitability represented by return on assets (ROA) and return on equity (ROE) is negatively and significantly associated with the debt to equity ratio.Recently, in the context of industrial companies in Jordan, Al Abbadi (2019) has examined the capital structure determinants, using 15 companies listed on the Amman Stock Exchange for 2014-2016.The results show that profitability, interest rates, tangible assets, size and growth have significantly affect debt.
At the level of Jordanian service companies, Gharaibeh and Khaled (2020) examined the impact of capital structure and financial characteristics on the profitability of services companies for the period 2014-2018.The current study is similar to the latest one in terms of sector, that is, Jordanian service companies, but differs in terms of purpose, since the current study examines the determinants of capital structure at the level of Jordanian service companies.Gharaibeh and Khaled (2020) have measured the business risk using standard deviation of a company's return on assets, while the current study measures the business risk using company's equity beta because shareholders are essentially concerned about the part of the risk they cannot remove through diversification, which indicates the beta of stock.
The motivation of the current study is that although the service sector is an important part of the ASE, it has not received due attention compared to the industrial sector.Therefore, a lack of extensive research on the capital structure determinants, especially at the service sector level, is the main motivation behind this study.The other motivation is to examine various corporate financial characteristics provided as capital structure determinants in previous studies, such as size, tangibility, profitability, growth, business risk, non-debt tax shields and percentage of institutional investors.Some of these determinants were not used in studies in Jordan, such as business risk measured by levered beta divided by one plus debt to the equity ratio rather than volatility of ROA, as well as non-debt tax shields.Therefore, the purpose of this study is to fill this gap by examining the determinants of capital structure at the level of Jordanian service companies listed on the ASE.

Size
Many previous studies document that the size of a company has a positive relationship with leverage (Albayrak, 2019;Ang, 1992 Scott (1976) points out that a big company tends to have high asset values, offers the best debt guarantee.Big companies attract institutional investors because they believe that big companies have a low risk of bankruptcy.O'Brien and Bhushan (1990) and Tang and Jang (2007) show that big companies are less subject to bankruptcy cost risk and have the required resources.Recently, at the level of the Gulf Cooperation Council (GCC) and UK real estate companies, Yousef (2019) has confirmed that company size has a significant efhttp://dx.doi.org/10.21511/imfi.17(2).2020.28fect on debt, and this result supports the tradeoff theory.The study uses the natural logarithm of total assets as an indicator of the company size (SIZE).
H1: Company size has a positive effect on debt.

Tangibility
The trade-off, agency cost and pecking-order theories suggest a positive relationship between tangibility and debt.Based on the agency cost theory, the stockholders of a leverage company tend to have a motivation to invest sub-optimally (Titman & Wessels, 1988).According to the trade-off and pecking-order theroies, the type of assets owned by companies determines their choices in capital strucute, since tangible firm's assets can be used as a collateral (Myers, 1984;Scott, 1976).However, Dakua (2019) contradicts previous results and finds that debt is negatively related to tangibility in the Indian steel industry.This means that the majority of leverage is not raised against fixed assets (Dakua, 2019).Doan (2019) confirms this result and finds that tangibility is negatively correlated with capital structure.Yousef (2019) has found that tangibility has a significant positive effect on debt in the GCC, while it has a significant negative effect in the UK.To determine the effect of tangiblitliy, the fixed assets divided by total assets ratio is used as a proxy for company tangibility in this study (TANG).
H2: Tangibility assets have a positive effect on debt.

Profitability
The trade-off theory suggests a positive relationship between profitability, since achieving more profits increases the company's ability to borrow and pay financial liabilities.Thus, bankruptcy costs will be minimized.In addition, debt plays a corrective role in the relationship between owners and managers, so it helps reduce agency costs among those stakeholders (Doan, 2019).Conversely, according to the pecking-order theory, past profitability is negatively related to debt because company will prefer using internal funds more than other sources, but can resort to debt if internal sources are not enough.1988).Recently, at the level of the GCC and UK real estate companies, Yousef (2019) has confirmed the previous result and finds that profitability has a significant negative effect on debt, and this result is in line with the pecking order theory.The current study uses earnings before interest and tax divided by the total assets ratio as an indicator of company profitability (PROFIT).
H3: Profitability has a negative effect on debt.

Growth
The trade-off theory assumes that there is a negative relationship between growth opportunities and debt.In particular, agency costs increase for growing companies, and this will lead to increased bankruptcy costs and increased obstacles to have external credit, thus reducing indebtedness.Serrasqueiro and Nunes (2014) find a negative relationship between growth opportunities and the level of company indebtedness in the hotel industry.They justify their result based on the fact that these companies are afraid of the risk of default associated with financing investment opportunities using debt.Yousef (2019) has found that growth has a significant positive effect on debt in the GCC, while it has a significant negative effect in the UK.Marketto-book ratio is used as a proxy for the growth opportunities of a company (GROW).
H4: Growth opportunities have a negative effect on debt.

Business risk
According to the bankruptcy theory, business risk is negatively related to debt.Bhaduri (2002), Titman and Wessels (1988) and Welch (2004) and others show that companies with high volatility incomes tend to have less leverage because debt includes a commitment of periodic payment, and increased debt ratio will expose a company to financial distress cost.On the other hand, Qiu and La's (2010) use the beta of stock as a measure for a company's business risk.They are based on the fact that stockholders have a great ability to diversify in the finanhttp://dx.doi.org/10.21511/imfi.17(2).2020.28cial markets, so their primary focus should be on systematic and non-diversifiable risks of their stock holding.To maintain the effect that financial leverage has on the stock beta, Qiu and La (2010) employ Hamadeh's equation to get the unlevered beta of stock as a measure of a company's risk.Following Qiu and La's (2010) approach, the current study calculates business risk based on levered beta divided by one plus debt to equity ratio (BR).
H5: Business risk has a negative effect on debt.

Non-debt tax shields
The trade-off theory assumes that non-debt tax shields are negatively related to debt.DeAngelo and Masulis (1980) debate that the presence of non-debt tax shields, such as fixed asset depreciation and amortization, lower a company's capacity of debt tax benefit.They claim that the companies with greater non-debt tax shields tend to have relatively lower debt in their capital structure.This result is confirmed by several previous studies such as Barton, Hill, and Sundaram (1989) and Prowse (1990).Bowen, Daley, and Huber (1982) show that the capital structure at the industry level is significantly affected by non-debt tax shields.Titman and Wessels (1988) find no signficant relationship between non-debt tax shields and the debt ratio, while Grier and Zychowicz (1994) and Chang, Lee, and Lee (2009) show a positive relationship.MacKie-Mason (1990) significantly contributes to this issue and suggests that there are two parts related to non-debt tax shields used in this study: the first part of NDTS (non-debt tax shield ) is investment tax credits that are positively related to the debt level, while the second part is tax loss carry-forwards that are negatively related to debt level.A proxy of non-debt tax shields used in this study is depreciation divided by total assets (NDTS).
H6: Non-debt tax shields have a negative effect on debt.

Institutional ownership
The agency theory suggests that agency costs can be minimized by an optimal ownership structure and capital structure (Jensen, 1986 Tong and Ning (2004) find that managerial ownership is negatively related to leverage when company faces a future of financial difficulties.This paper uses PIO as an indicator of company profitability.
H7: Institutional investors have a negative effect on debt.

Sample selection and data
This study examines the capital structure determinants of all Jordanian service companies.The yearly data are collected from the Amman Stock Exchange (ASE).The study is based on the financial 2014-2018 data, which include 225 year observations from a final sample of 45 Jordanian service companies.One company was excluded due to the lack of related information on the shares.
The study period begins from 2014 to avoid the global 2009 financial crisis and its effects, and extends until 2013.Therefore, the study period was selected to have more stable results related to the capital structure determinants of Jordanian service companies.Two companies were removed from the sample due to insufficient data availability.Panel data regression is employed to investigate the effect of the six company characteristics on the debt ratio using a balanced panel data.This paper chooses the fixed effect model based on the Hausman test result.

Research methodology
The study examines the capital structure determinants of Jordanian service companies, using a panel regression analysis:

RESULTS
The characteristics of Jordanian service companies are listed in Table 1.On average, the Jordanian service companies use only the 34 percent debt ratio in their capital structure.This is considered a low debt ratio, and one explanation is that these companies prefer reducing the probability of bankruptcy by minimizing debt financing.Given the profitability of Jordanian service companies, they achieve an average of 5 percent from operating activities.The percentage of institutional investors is relatively high, 68 percent on average.This means that the institutional investors play an important role in governing and managing these companies.
To determine the degree of variable integration, the LLC unit root tests with intercept and with intercept and trend are used to check whether the variables include a unit root or not.The results of the LLC test are detailed in Note: The first row reports the name of each variable addressed in this study, while the second row details the average value of each variable for the Jordanian service companies.The number of observations for each variable is in the last row.
both the level and the first difference of each variable in Panel A and B, respectively.Panel A of Table 2 shows that the series include a unit root, especially SIZE, TANG and NDTS.Therefore, the variables are non-stationary in their levels.
On the other hand, Panel B of Table 2 reveals that the variables are stationary in the first difference.This indicates that the series should be investigated at the level of the first difference.This section presents and discusses the empirical analysis of the capital structure determinants.Table 3 shows the correlation matrix for all explanatory variables used in this study.Table 2 shows that all correlation coefficients are not high.As a result, there are no multicollinearity problems among explanatory variables.Given the Lagrange multiplier test, it is statistically significant.Therefore, the panel model is recommended rather than the pooled model.Next, if the Hausman test provides a significant result, then the fixed effect result is recommended rather than the random effect result.The Hausman test is statistically significant for the capital structure model.This means that the fixed effects model is preferred over the random effects model for capital structure.This indicates that firm-specific characteristics that determine differences between the leverage of Jordanian companies are not randomly determined.Table 4 shows that the value of the R-squared adjusted in the model is 0.93, which means that explanatory variables explain 93% of the variation in the debt ratio, while 7% of the variation in the debt ratio cannot be explained by the explanatory variables used in this study.
Table 4 shows that the leverage in Jordanian service sector companies is positively and significantly affected by size.This means that big Jordanian service companies are more diversified and less possible to be subject to financial distress, and larger firms tend to borrow more.This finding is in line with the trade-off and bankruptcy theories of capital structure, since big companies can meet creditor commitments and have lower bankruptcy costs, and this result is consistent with most previous results, such as Rajan and Zingales (1995), Bhaduri The asset tangibility is not a significant determinant of leverage for Jordanian service companies.
However, the asset tangibility is negatively affected by debt ratio.In terms of sign, this finding is consistent with Dakua (2019), who finds a negative significant relationship between tangibility and debt in the Indian steel industry, as well as with Doan (2019), who shows that tangibility is negatively correlated to capital structure in Vietnam between 2008 and 2018.This indicates that companies do not use fixed assets as a collateral or companies with higher collateral value tend to borrow less debt.The result contradicts the expected positive sign predicted by the trade-off, pecking-order and agency theories of capital structure.Yousef (2019) show that profitability and leverage are negatively correlated.The effect of the growth variable, as referred by the market to book rate, is not significant for the leverage model.This implies that Jordanian service companies with high-growth opportunities do not rely heavily on debt to finance their investment opportunities.The lack of statistical significance of the growth effect can also be somewhat explained by its significant correlation with most other independent variables (see Table 3).However, the negative sign of the growth coefficient is in line with Al-Najjar and Taylor (2008), who show that the potential growth rate is positively and significantly related to debt ratio in Jordanian non-financial firms, suggesting that the agency cost theory and pecking-order theory are more relevant than the trade-off theory in the context of service companies in Jordan Referring to Table 4, in accordance with the results of the existing literature, this study shows that business risk has a negative significant impact on debt ratio of Jordanian service companies.This finding is in line with the bankruptcy cost theory of capital structure, as risky companies with a high debt ratio are more likely to suffer from high financial distress costs and underinvestment problems.Therefore, companies with high business risk tend to be less leveraged.Furthermore, a negative relationship between business risk and debt ratio is in line with the trade-off theory, where a risky company will find debt less attractive (Wald, 1999).According to the trade-off theory, companies are unable to repay their debt due to higher bankruptcy and financial distress costs.This finding is in line with previous results such as Albayrak (2019).
Table 4 shows a significant positive effect of nondebt tax shields on debt ratio.This contrasts the expected negative sign predicted by the trade-off theory.This means that Jordanian service companies with lager non-debt tax shields tend to include more debt in their capital structure.The finding is consistent with Chang, A. C. Lee, and C. F. Lee (2009).PIO is statistically insignificant, but the coefficient of PIO is still negative and economically large -1.1914.This indicates that increased percentage of stock owned by institutional investors from the total ownership of board members in Jordanian service companies with 1 percent leads to largely reduced debt ratio with 119 percent.The evidence suggests that institutional investors do not prefer using debt to finance their investments.This result confirms the finding of Al-Najjar and Taylor (2008), who show that PIO is negatively and significantly related to debt ratio in Jordanian firms.

DISCUSSION
The  Yousef (2019).As for the growth factor, this paper finds that growth of a company has no statistically significant effect on a company's choice of a debt ratio.This means that companies may prefer to use debt to increase equity when internal funds are exhausted.
Although the findings support the theories related to capital structure, two variables, namely, tangibility and non-debt tax shields, contradict these theories.For example, tangibility result contradicts the agency theory of capital structure in the sense that Jordanian service companies with more fixed assets cannot employ fixed assets as a collateral.In particular, if a company offers high tangible assets, this will lead to reduced agency costs of the debt, since tangible assets are easy to collateralize; therefore, decreased agency cost will provide more productivity in the company value (Rajan & Zingales, 1995).This indicates that companies do not use fixed assets as a collateral or companies with higher collateral value tend to borrow less debt.This finding is in line with Dakua (2019)

CONCLUSION
This study examines the capital structure determinants of companies in the Jordanian service sector for 2014-2018, using data from a sample of 45 companies.The service sector is considered unique due to its huge tangibility and seasonality.Therefore, the results relatively differ from those documented in the empirical studies in other sectors.For example, tangibility has no significant impact on debt decisions.
This paper shows that size and non-debt tax shield have a positive significant effect on the debt in the Jordanian service companies, while profitability and business risk have a negative significant impact on the debt ratio.In spite of the coefficient of institutional investors is statistically insignificant, it is still negative and economically significant.
This paper contributes to the capital structure literature in various ways.First, it expands knowledge of the financing of the Jordanian service sector, which, although an important component of the ASE, has not received due attention compared to other sectors.Second, the results of the analysis confirm that most researchers use similar variables when studying the capital structure determinants.After reviewing the literature, this paper attempts to build a new model combining the strongest variables, which plays an important role in the capital structure determinants.In this study, for example, unlevered beta is used, rather than volatility of ROA, to measure business risk; this indicates that stockholders mainly worry about systematic risk of stocks they cannot evade by diversification.Furthermore, this study is the first to use the PIO variable, which is the percentage of stock owned by institutional investors from the total ownership of board members to capture the ownership structure as a determinant of capital structure in Jordanian service sector companies.

Table 2 , for Table 1 .
Characteristics of Jordanian service companies

Table 2 .
LLC test results

Table 3 .
Correlation coefficients between independent variables

Table 4
details the results of the capital structure equation.Seven variables are used to reflect the capital structure determinants.To determine the preferable set of results statistically, the Lagrange multiplier and Hausman test result is presented.

Table 4 .
The effect of service company characteristics on the capital structure Harris and Raviv (1991)8))his finding contradicts most of the previous studies, such asJensen and Meckling (1976),Myers (1984),Titman and Wessels (1988),Harris and Raviv (1991), Rajan and Zingales (1995), Sayılgan, Karabacak, and Küçükkocaoğlu (2006), Tang and Jang (2007), and Al-Najjar and Taylor (2008).This paper finds a significant positive effect of non-debt tax shields on the debt ratio.This contrasts with the expected negative sign predicted by the trade-off theory.This implies that Jordanian service companies with lager non-debt tax shields tend to include more debt in their capital structure.The result is consistent with Chang, A. C. Lee, and C. F. Lee (2009).Although PIO is statistically insignificant, the coefficient of PIO is still negative and economically large.This finding confirms the results of the Al-Najjar and Taylor's (2008) study that shows there is a negative relationship between PIO and the debt ratio in Jordanian non-financial firms.In general, it is concluded that size, profitability, business risk, non-debt tax shields and PIO are important factors in terms of shaping capital structure in Jordanian service companies.