Description
Capital Structure refers to the way a corporation finances its assets through some combination of equity, debt, or hybrid securities. A firm's capital structure is then the composition or 'structure' of its liabilities.
Factors Determining Capital Structure: A Case study of listed companies in Sri Lanka
Abstract The most debatable topic in capital structure revolves around the optimal capital structure. This article investigates the factors determining capital structure in Sri Lankan context during 2002 to 2006 (five years). In an attempt, the present study examined the determinants of capital structure such as tangibility, size, growth rate, profitability, liquidity and dividend payout with samples of 50 companies including 13 sectors representing 21% of the mean number of companies listed in Colombo Stock Exchange Ltd as of 2006. Descriptive and inferential statistics were used in the analysis of data using Statistical Package for Social Science (SPSS). Results revealed that the use of debt capital is relatively low in Sri Lanka and size, growth rate and profitability are statistically significant determinants of capital structure. According to regression analysis, above six factors have an impact on capital structure at the rate of 77 % which is significant at 0.05 levels. Keywords: Capital Structure, Determinant(s), Colombo Stock Exchange. 1. Introduction Capital structure remains as a controversial issue in modern corporate finance. A firm can raise fund either through debt or equity or mixture of both. There are two schools of thought in capital structure. One school pleads for optimal capital structure and other does against. Former school argues that judicious mixture of debt and equity capital can minimize the overall cost of capital and maximize the value of the firm. Hence, this school considers capital structure decision as relevant. Latter school of thought led by Modiglinai and Miller contends that financing decision does not affect the value of the firm. The modern theory of capital structure began with the introduction of Modiglinai and Miller (1958), Rajan and Zingales (1995), Harris and Raviv (1991). Four theoretical approaches can be distinguished namely the irrelevance theory of Modiglinai and Miller (1958) , the trade off theory (Bradley et al., 1984), agency cost theory (Jensen and Meckling, 1976) and pecking order theory (Myers and Majluf, 1984). The three conflicting theories of capital structure such as trade-off theory, agency cost theory and pecking order theories have been developed after the establishment of Modiglinai and Miller's theory. Even though his theory shows the irrelevance of capital structure to the value of the firm, this implies that, there does not exist an optimal capital structure. Because, a firm's value cannot be affected by its choice of financing. The logic of the Modigliani and Miller (1958) analysis is still accepted, despite the contradiction of their theoretical conclusions with empirical evidence. Various factors have been suggested as major players in determining how much of a company can borrow and in what circumstances. Academics don't seem to be in a complete agreement as to what determines the capital structure of a firm. Tangibility of assets, growth opportunities, size, uniqueness, business risk, and profitability are some of the major factors which determine the capital structure. However, the significance of these determinants may vary from country to country depending on their economy settings. In this study, determinants of capital structure in Sri Lankan context are examined based on a panel data set from 2002 to 2006 comprising 50 companies listed on Colombo Stock Exchange with reference to capital structure theories.
2.
Research Problem
The assets of a company can be financed by issuing ordinary shares or by retained earnings and borrowings. Capital structure refers to the different options used by a firm in financing its assets. It can combine bonds, lease financing bank loans or many other options with equity in an overall attempt to boost the market value of the firm. To maximize the overall value, firms differ with respect to capital structures which have given birth to different capital structure theories that attempt to explain the variation in capital structure firms over time or across regions. This study attempts to examine the factors which determine the capital structure of Sri Lankan listed companies. RQ: What are the factors that determine the capital structure of Sri Lankan listed companies? 3. Objective of Study The main objective of the study is to find out the factors which determine the capital structure of Sri Lankan listed companies and sub objectives are: identify the impact of those factors on capital structure of the companies. 4. Review of Literature
Theoretical and empirical capital structure studies have generated many results that attempt to explain the determinants of capital structure. There exist a number of determinants of capital structure derived from various theories. Haris and Raviv (1991) demonstrate in their review article that the motives and circumstances could determine capital structure choices which have been seen nearly uncountable. According to Titman and Wessels (1988), the required explanatory variable may frequently be imperfect proxies for the defined corporate attributes. Most of the capital structure studies to date are based on data from developed countries. For example, Rajan and Zingales (1995) used data from the G.7 countries while Bevan and Danbolt (2000 and 2002) utilized data from UK, Antoious et al. (2002) analyzed data from UK, Germany and France and Hall et al. (2004) used data from Europe. There are only a few studies that provide evidence from developing countries. For example, Booth et al. (2001) analyzed data from developing countries (Brazil, Mexico, India, South Korea Jordan, Malaysia, Pakistan and Thailand. Turkey and Zimbabwe, Pandey (2001) used data from Malaysia. Chen (2004) utilized data from China, Omet and Nobanee (2001) used data from Jordan and Al-sakaran (2001) analyzed data from Saudi Arabia. Some have used cross country comparisons based data from particular region. For example, Decsomsak et al. (2004) analyzed data from the Asian pacific region. There is no universal theory of capital structure and no reason to expect one (Myers, 2001). Different theories give different predictions on how tangibility is related to leverage. On the relationship between tangibility and capital structure, the trade-off and agency theories state that tangibility is positively related to leverage. If a company's tangible assets are high, then these assets can be used as collateral, diminishing the lender's risk of suffering such agency costs of debt. Moreover, in firms with more intangible assets, the costs of controlling capital expenditures are higher, as monitoring is more difficult. Hence, a high fraction of tangible assets is expected to be associated with high leverage. Several empirical studies confirm this suggestion, as Friend and Lang (1988), Rajan and Zingales (1995), Wald (1999), Frank and Goyal (2004) found out. On the other hand, Bevan and Danbolt (2000), Booth et al. (2001), and Huang and Song (2002) experienced a negative relationship between tangibility and leverage in some cases. In line with the tradeoff theory, size is expected to be positively related to leverage. However, as Rajan and Zingales (1995) stated, if the costs of financial distress are low, the positive relationship should not be strong. Agency theory also suggests a positive relationship between size and leverage. According to Pecking order theory, larger firms have less asymmetric information problems which imply that they will have a higher preference for equity than smaller ones. A negative relationship is thus predicted for size and leverage. Some authors find positive relationship between size and leverage, for example Friend and Lang(1988), Rajan and Zingales (1995) for all G7 countries except for Germany, Wald (1999), Wiwattanakantang (1999), Bevan and Danbolt (2000), Pandey (2001), Al-Sakaran (2001),
237
Huang and Song (2002), Frank and Goyal (2004). On the other hand, in some studies a negative relationship is reported, for example Titman and Wessels (1988) and Chaplinsky and Niehaus(1993). Whereas the trade-off, signaling and agency theories expect a positive relationship between profitability and leverage, the pecking order theory predicts a negative one. Most empirical studies observe a negative relationship between leverage and profitability. For example Kester (1986), Friend and Lang(1988), Titman and Wessels (1988), Rajan and Zingales (1995) for G7 countries except for Germany, Wald(1999), Wiwattanakantang (1999), Bevan and Danbolt (2000), Booth et al. (2001), Huang and Song (2002).On the other hand, Frank and Goyal (2004) experienced a positive relationship between profitability and leverage in some models. Velnampy and Nimalathasan (2010) examined about firm size on profitability between Bank of Ceylon and Commercial Bank of Ceylon in Sri Lanka during ten years period from 1997 to 2006 and found that there is a positive relationship between Firm size and Profitability in Commercial Bank of Ceylon Ltd, but there is no relationship between firm size and profitability in Bank of Ceylon. Various studies identified the determinants of profitability (Islam and Mili, 2012, Velnampy, 2005 & 2005, 2013, Velnampy and Pratheepkanth, 2012, and Niresh and Velnampy, 2012) The agency cost theory and pecking order theory explain the contradictory relation between the growth rate and capital structure. Growth rate is negatively related with long-term debt level (Jensen and Meckling, 1976). Some empirical studies also show the negative relationship, such as Kim and Sorensen (1986), and Titman and Wessels (1988), Rajan and Zingales (1995), Wiwattanakantang (1999), Bevan and Danbolt (2000), Huang and Song (2002) and Frank and Goyal (2004). On the other hand, Booth et al. (2001) demonstrate, in some models, a positive relationship between growth opportunities and leverage. On the other hand, Pecking order theory, contrary to the agency cost theory, shows the positive relation between the growth rate and debt level of enterprises based on that a higher growth rate implies a higher demand for funds, and, ceteris paribus, a greater reliance on external financing through the preferred source of debt. Pecking order theory contends that management prefers internal to external financing and debt to equity if it issues securities. Velnampy and Nimalathasan (2008) pointed out the association between organizational growth and profitability in Commercial Bank of Ceylon limited in Sri Lanka and concluded that organizational growth has a greater impact on all profitability ratios such as net profit, operating profit, ROE, ROI and RAA. The bankruptcy costs theory pleads for adverse relation between the dividend payout ratio and debt level in capital structure. The low bankruptcy cost implies the high level of debt in the capital structure. But, the pecking order theory shows the positive relation between debt level and dividend payout ratio. Further, management prefers the internal financing to external one. Instead of distributing the high dividend, and meeting the financial need from debt capital, management retains the earnings. A negative relationship between liquidity and leverage is expected in market oriented economies (Ghassan omet & Fadi Mashharawe, 2003). Indeed, this result is supported by the empirical findings of Ozkan (2001) and Antonious (2002). 5. Methodology 5 .1. Sampling Techniques For the purpose of this study, population has been defined in terms of the number of companies listed on Colombo Stock Exchange (CSE) for the period from 2002 to 2006. As on this period, the total number of such companies falling in twenty different sectors such as Banks finance and insurance, Beverage food and tobacco, Chemicals and pharmaceuticals, Construction and engineering, Diversified holdings, Foot ware and textiles, Health care, Hotels and travels, Information technology, Investment trusts, Land and property, Manufacturing, Motors, Oil palms, Plantations, Power and energy, Services, Stores and supplies, Telecommunications and Trading were around 235. The sample of firms is selected on the basis of following criteria. -financial firm. The companies in Banks finance and insurance, and Investment trust sectors are excluded, since the use and determinants of leverage of these companies are likely to be different to the ones in companies in non-financial sectors.
238
Availability of the required firm-specific data for a consecutive period of five years starting from the financial year 2002 and ending with the financial year 2006.Firms with any missing observations for any variable in the model during the above range also dropped. Firms having negative book equity are excluded since that can lead to negative leverage, which will induce a downward bias to the mean estimates of leverage. r any of capital structure are also excluded. These criteria result in a sample of fifty companies. This represents 21% of the mean number of companies listed in the CSE during the sample period. The table 1 summarizes how individual sectors are represented in the sample. Table 1: Sectoral distribution of sample No of companies in the population 18 09 04 10 06 33 21 32 07 05 18 06 11 235 No of companies in the sample 8 5 1 3 2 6 7 10 2 1 3 1 1 50
No 1 2 3 4 5 6 7 8 9 10 11 12 13
Sector Beverage food and tobacco Chemicals and pharmaceuticals Construction and engineering Diversified holdings Health care Hotels and travels Land and property Manufacturing Motors Oil palms Plantations Services Trading Total
5 .2. Data Collection The secondary data were collected for the study during the period of five years (2002 -2006). And the data used for the empirical analysis was derived from the data base maintained by Colombo Stock Exchange (CSE). This data base contains balance sheet, profit and loss account and investor guide. The data were averaged over five years to smooth the leverage and explanatory variables. While Titman and Wesseles (1988) adopted three year averages, Rajan and Zingales (1995) used five year averages. Following Rajan and Zingales, this study used five year averages and some
239
necessary data were hunted from online (official website of CSE).Further, annual reports of the companies, books, journals, magazines and research reports were also used for data collection. 5 .3. Mode of Analysis In this study, various statistical methods have been employed to analyze data collected from 50 companies listed on CSE. A well know statistical package called "SPSS" (Statistical Package for Social Science) version 13 has been used to analyze the data researcher collected. The upper level of statistical significance for hypotheses testing was set at 5%. All statistical test results were computed at the 2-tailed level of significance. Statistical analysis involves both descriptive and inferential statistics. According to research objective and research questions, this study has set the variables used in the study and their measurement is largely adopted from existing literatures. The following table 2 shows the variables and their measures. Table 2: Determinants of capital structure No Variables Measurement
1
Leverage
Long term debt / Total capital
2
Tangibility (Assets structure)
Book value of fixed assets/ Total assets (FA/TA)
3
Size
natural logarithm of sales Return on equity (ROE) (Profit before interest & taxes/ Total Assets X 100) Market to book ratio.
4
Profitability
5
Growth
6
Liquidity
Current assets/ current liabilities (CA/CL)
7
Dividend payout
Dividend per share (DPS)/Earnings per share (EPS)
5 .4. Development of Hypotheses This study has tested the following hypothesis on relations between the defined variable and capital structure of the listed companies. H1: There is a positive relationship between tangibility and leverage of the firm. H2: There is a positive relationship between size and leverage of the firm. H3: There is a negative relationship between profitability and leverage of the firm. H4: There is a positive relationship between growth rate and leverage of the firm. H5: There is a positive relationship between dividend payout and leverage of the firm. H6: There is a positive relationship between liquidity and leverage of the firm.
240
5 .5. Conceptual Model Based on data availability and some other constraints, only six variables such as tangibility, size, profitability, growth, dividend payout and liquidity are tested in this study. This can be conceptualized as follows;
Size (H2)
Tangibility
(H1 )
There is a positive relationship Profitability (H3)
Determinantbeof Capital s tween Structure
firm. H2: There is a Dividend payout (H5) iqui ive Lpositdity (H6) tangibility leverage of and the
Capital Structure
Growth rate (H4)
relationship Figure 1: Research Model between size and
6. Results and Discussions 6.1. Descriptive Statistics Descriptive statistics provide information on the key vleverage theoftud the d are presented below. Means, medians ariables in s y an and standard deviations for the each regression variables are presented in table 3.
firm. H3: There is a negative relationship between profitability leverage firm. H4: There is a positive 241 of and the
relationship
Table 3: Mean Values of Regression Variables (2002-2006) Variable Leverage 15 .62 Tangibility 0.47 Sales Growth Profitability Liquidity Dividend payout 8 .948 29.5 14 .99 15 0.40 0.28 0.81 83.9 14 .56 26 .54 0.59 19 .53 Mean Standard deviation
The table 3 indicates that the means for the leverage, tangibility, size, growth, profitability, liquidity and dividend payout ranged from a low of 0.40 to a high of 29.5. It therefore appears that the mean leverage (the proportion of long term debt in the total capital) in Sri Lanka is estimated as 15.62 % . Thus, the use of debt capital by Sri Lanka firm is quite low compared to G 7 countries (32.5%). The mean leverage, which is estimated using data for the financial year 2005, is 28.989%, drastically higher than the five year mean, which was observed earlier (15.62%). Table 4 below, reports the yearly mean and median figures for the leverage mentioned above. Table 4: Yearly Mean values of Leverage (2002-2006) Year 2002 2003 2004 2005 2006 Average Mean 15.496 17.218 14.757 28.989 1 .623 15.617 Median 9.02 10.8 9 .8 11.545 0 .775 8 .388
According to the above table, yearly mean leverage amount to 15%, and the corresponding median amounts to 8%. While we can observe a tendency of increasing mean leverage values in 2002, 2003 and 2005, a sudden decline observed in 2006 ( 1.623%).
242
The table 3 shows that, mean tangibility over the five year is 47% and the market value of assets is 29.5 times the book value of assets on mean. Further, the means of sales of the companies in the sample is Rs 16219170118 (means of natural logarithm of sales is 8.948). During the five year period, firms earned a mean ROE of 14.99%. Moreover, it indicates that the companies in the sample have current assets, which is 15 times greater than their current liability. The average dividend available to shareholders out of earnings is 0.4 times. 6 .2. Correlation Analysis
The Pearson's Moment Correlation Coefficient was computed for purposes of determining the following relationship: The relationship between leverage and tangibility The relationship between leverage and size The relationship between leverage and growth The relationship between leverage and profitability The relationship between leverage and liquidity The relationship between leverage and dividend payout. In order to delineate the relationship between the various determinants of capital structure, leverage was correlated and is presented in table 5. Table 5: Correlation matrix of regression variables Variable Leverage Tangibility Size Growth Profitability Liquidity Dividend Payout
Tangibility .019 Size .449(**) Growth -.096 Profitability .569(**) Liquidity .387(**) Dividend .127 payout
-.073 -.014 -.304(*) -.084 .347(*)
.282(*) .379(**) .188 -.020
.292(*) -.120 -.104
.391(**) -.112 -.034 1
** Correlation is significant at the 0.01 level (2-tailed) ** Correlation is significant at the 0.05 level (2-tailed) The results indicate that there is a direct and positive (weak) relationship between tangibility and leverage (r=0.019). This supports the hypothesis that there is a positive relationship between tangibility and leverage. However, the estimated coefficient on tangibility is statistically insignificant. (P > 0.01). A significant, strong and positive correlation is shown between size and leverage (r =0.449, P 0.01), rejecting the hypothesis that there is a positive relationship between growth and leverage. There was also a significant strong and positive relationship between profitability and leverage (r = 0.569, P
Capital Structure refers to the way a corporation finances its assets through some combination of equity, debt, or hybrid securities. A firm's capital structure is then the composition or 'structure' of its liabilities.
Factors Determining Capital Structure: A Case study of listed companies in Sri Lanka
Abstract The most debatable topic in capital structure revolves around the optimal capital structure. This article investigates the factors determining capital structure in Sri Lankan context during 2002 to 2006 (five years). In an attempt, the present study examined the determinants of capital structure such as tangibility, size, growth rate, profitability, liquidity and dividend payout with samples of 50 companies including 13 sectors representing 21% of the mean number of companies listed in Colombo Stock Exchange Ltd as of 2006. Descriptive and inferential statistics were used in the analysis of data using Statistical Package for Social Science (SPSS). Results revealed that the use of debt capital is relatively low in Sri Lanka and size, growth rate and profitability are statistically significant determinants of capital structure. According to regression analysis, above six factors have an impact on capital structure at the rate of 77 % which is significant at 0.05 levels. Keywords: Capital Structure, Determinant(s), Colombo Stock Exchange. 1. Introduction Capital structure remains as a controversial issue in modern corporate finance. A firm can raise fund either through debt or equity or mixture of both. There are two schools of thought in capital structure. One school pleads for optimal capital structure and other does against. Former school argues that judicious mixture of debt and equity capital can minimize the overall cost of capital and maximize the value of the firm. Hence, this school considers capital structure decision as relevant. Latter school of thought led by Modiglinai and Miller contends that financing decision does not affect the value of the firm. The modern theory of capital structure began with the introduction of Modiglinai and Miller (1958), Rajan and Zingales (1995), Harris and Raviv (1991). Four theoretical approaches can be distinguished namely the irrelevance theory of Modiglinai and Miller (1958) , the trade off theory (Bradley et al., 1984), agency cost theory (Jensen and Meckling, 1976) and pecking order theory (Myers and Majluf, 1984). The three conflicting theories of capital structure such as trade-off theory, agency cost theory and pecking order theories have been developed after the establishment of Modiglinai and Miller's theory. Even though his theory shows the irrelevance of capital structure to the value of the firm, this implies that, there does not exist an optimal capital structure. Because, a firm's value cannot be affected by its choice of financing. The logic of the Modigliani and Miller (1958) analysis is still accepted, despite the contradiction of their theoretical conclusions with empirical evidence. Various factors have been suggested as major players in determining how much of a company can borrow and in what circumstances. Academics don't seem to be in a complete agreement as to what determines the capital structure of a firm. Tangibility of assets, growth opportunities, size, uniqueness, business risk, and profitability are some of the major factors which determine the capital structure. However, the significance of these determinants may vary from country to country depending on their economy settings. In this study, determinants of capital structure in Sri Lankan context are examined based on a panel data set from 2002 to 2006 comprising 50 companies listed on Colombo Stock Exchange with reference to capital structure theories.
2.
Research Problem
The assets of a company can be financed by issuing ordinary shares or by retained earnings and borrowings. Capital structure refers to the different options used by a firm in financing its assets. It can combine bonds, lease financing bank loans or many other options with equity in an overall attempt to boost the market value of the firm. To maximize the overall value, firms differ with respect to capital structures which have given birth to different capital structure theories that attempt to explain the variation in capital structure firms over time or across regions. This study attempts to examine the factors which determine the capital structure of Sri Lankan listed companies. RQ: What are the factors that determine the capital structure of Sri Lankan listed companies? 3. Objective of Study The main objective of the study is to find out the factors which determine the capital structure of Sri Lankan listed companies and sub objectives are: identify the impact of those factors on capital structure of the companies. 4. Review of Literature
Theoretical and empirical capital structure studies have generated many results that attempt to explain the determinants of capital structure. There exist a number of determinants of capital structure derived from various theories. Haris and Raviv (1991) demonstrate in their review article that the motives and circumstances could determine capital structure choices which have been seen nearly uncountable. According to Titman and Wessels (1988), the required explanatory variable may frequently be imperfect proxies for the defined corporate attributes. Most of the capital structure studies to date are based on data from developed countries. For example, Rajan and Zingales (1995) used data from the G.7 countries while Bevan and Danbolt (2000 and 2002) utilized data from UK, Antoious et al. (2002) analyzed data from UK, Germany and France and Hall et al. (2004) used data from Europe. There are only a few studies that provide evidence from developing countries. For example, Booth et al. (2001) analyzed data from developing countries (Brazil, Mexico, India, South Korea Jordan, Malaysia, Pakistan and Thailand. Turkey and Zimbabwe, Pandey (2001) used data from Malaysia. Chen (2004) utilized data from China, Omet and Nobanee (2001) used data from Jordan and Al-sakaran (2001) analyzed data from Saudi Arabia. Some have used cross country comparisons based data from particular region. For example, Decsomsak et al. (2004) analyzed data from the Asian pacific region. There is no universal theory of capital structure and no reason to expect one (Myers, 2001). Different theories give different predictions on how tangibility is related to leverage. On the relationship between tangibility and capital structure, the trade-off and agency theories state that tangibility is positively related to leverage. If a company's tangible assets are high, then these assets can be used as collateral, diminishing the lender's risk of suffering such agency costs of debt. Moreover, in firms with more intangible assets, the costs of controlling capital expenditures are higher, as monitoring is more difficult. Hence, a high fraction of tangible assets is expected to be associated with high leverage. Several empirical studies confirm this suggestion, as Friend and Lang (1988), Rajan and Zingales (1995), Wald (1999), Frank and Goyal (2004) found out. On the other hand, Bevan and Danbolt (2000), Booth et al. (2001), and Huang and Song (2002) experienced a negative relationship between tangibility and leverage in some cases. In line with the tradeoff theory, size is expected to be positively related to leverage. However, as Rajan and Zingales (1995) stated, if the costs of financial distress are low, the positive relationship should not be strong. Agency theory also suggests a positive relationship between size and leverage. According to Pecking order theory, larger firms have less asymmetric information problems which imply that they will have a higher preference for equity than smaller ones. A negative relationship is thus predicted for size and leverage. Some authors find positive relationship between size and leverage, for example Friend and Lang(1988), Rajan and Zingales (1995) for all G7 countries except for Germany, Wald (1999), Wiwattanakantang (1999), Bevan and Danbolt (2000), Pandey (2001), Al-Sakaran (2001),
237
Huang and Song (2002), Frank and Goyal (2004). On the other hand, in some studies a negative relationship is reported, for example Titman and Wessels (1988) and Chaplinsky and Niehaus(1993). Whereas the trade-off, signaling and agency theories expect a positive relationship between profitability and leverage, the pecking order theory predicts a negative one. Most empirical studies observe a negative relationship between leverage and profitability. For example Kester (1986), Friend and Lang(1988), Titman and Wessels (1988), Rajan and Zingales (1995) for G7 countries except for Germany, Wald(1999), Wiwattanakantang (1999), Bevan and Danbolt (2000), Booth et al. (2001), Huang and Song (2002).On the other hand, Frank and Goyal (2004) experienced a positive relationship between profitability and leverage in some models. Velnampy and Nimalathasan (2010) examined about firm size on profitability between Bank of Ceylon and Commercial Bank of Ceylon in Sri Lanka during ten years period from 1997 to 2006 and found that there is a positive relationship between Firm size and Profitability in Commercial Bank of Ceylon Ltd, but there is no relationship between firm size and profitability in Bank of Ceylon. Various studies identified the determinants of profitability (Islam and Mili, 2012, Velnampy, 2005 & 2005, 2013, Velnampy and Pratheepkanth, 2012, and Niresh and Velnampy, 2012) The agency cost theory and pecking order theory explain the contradictory relation between the growth rate and capital structure. Growth rate is negatively related with long-term debt level (Jensen and Meckling, 1976). Some empirical studies also show the negative relationship, such as Kim and Sorensen (1986), and Titman and Wessels (1988), Rajan and Zingales (1995), Wiwattanakantang (1999), Bevan and Danbolt (2000), Huang and Song (2002) and Frank and Goyal (2004). On the other hand, Booth et al. (2001) demonstrate, in some models, a positive relationship between growth opportunities and leverage. On the other hand, Pecking order theory, contrary to the agency cost theory, shows the positive relation between the growth rate and debt level of enterprises based on that a higher growth rate implies a higher demand for funds, and, ceteris paribus, a greater reliance on external financing through the preferred source of debt. Pecking order theory contends that management prefers internal to external financing and debt to equity if it issues securities. Velnampy and Nimalathasan (2008) pointed out the association between organizational growth and profitability in Commercial Bank of Ceylon limited in Sri Lanka and concluded that organizational growth has a greater impact on all profitability ratios such as net profit, operating profit, ROE, ROI and RAA. The bankruptcy costs theory pleads for adverse relation between the dividend payout ratio and debt level in capital structure. The low bankruptcy cost implies the high level of debt in the capital structure. But, the pecking order theory shows the positive relation between debt level and dividend payout ratio. Further, management prefers the internal financing to external one. Instead of distributing the high dividend, and meeting the financial need from debt capital, management retains the earnings. A negative relationship between liquidity and leverage is expected in market oriented economies (Ghassan omet & Fadi Mashharawe, 2003). Indeed, this result is supported by the empirical findings of Ozkan (2001) and Antonious (2002). 5. Methodology 5 .1. Sampling Techniques For the purpose of this study, population has been defined in terms of the number of companies listed on Colombo Stock Exchange (CSE) for the period from 2002 to 2006. As on this period, the total number of such companies falling in twenty different sectors such as Banks finance and insurance, Beverage food and tobacco, Chemicals and pharmaceuticals, Construction and engineering, Diversified holdings, Foot ware and textiles, Health care, Hotels and travels, Information technology, Investment trusts, Land and property, Manufacturing, Motors, Oil palms, Plantations, Power and energy, Services, Stores and supplies, Telecommunications and Trading were around 235. The sample of firms is selected on the basis of following criteria. -financial firm. The companies in Banks finance and insurance, and Investment trust sectors are excluded, since the use and determinants of leverage of these companies are likely to be different to the ones in companies in non-financial sectors.
238
Availability of the required firm-specific data for a consecutive period of five years starting from the financial year 2002 and ending with the financial year 2006.Firms with any missing observations for any variable in the model during the above range also dropped. Firms having negative book equity are excluded since that can lead to negative leverage, which will induce a downward bias to the mean estimates of leverage. r any of capital structure are also excluded. These criteria result in a sample of fifty companies. This represents 21% of the mean number of companies listed in the CSE during the sample period. The table 1 summarizes how individual sectors are represented in the sample. Table 1: Sectoral distribution of sample No of companies in the population 18 09 04 10 06 33 21 32 07 05 18 06 11 235 No of companies in the sample 8 5 1 3 2 6 7 10 2 1 3 1 1 50
No 1 2 3 4 5 6 7 8 9 10 11 12 13
Sector Beverage food and tobacco Chemicals and pharmaceuticals Construction and engineering Diversified holdings Health care Hotels and travels Land and property Manufacturing Motors Oil palms Plantations Services Trading Total
5 .2. Data Collection The secondary data were collected for the study during the period of five years (2002 -2006). And the data used for the empirical analysis was derived from the data base maintained by Colombo Stock Exchange (CSE). This data base contains balance sheet, profit and loss account and investor guide. The data were averaged over five years to smooth the leverage and explanatory variables. While Titman and Wesseles (1988) adopted three year averages, Rajan and Zingales (1995) used five year averages. Following Rajan and Zingales, this study used five year averages and some
239
necessary data were hunted from online (official website of CSE).Further, annual reports of the companies, books, journals, magazines and research reports were also used for data collection. 5 .3. Mode of Analysis In this study, various statistical methods have been employed to analyze data collected from 50 companies listed on CSE. A well know statistical package called "SPSS" (Statistical Package for Social Science) version 13 has been used to analyze the data researcher collected. The upper level of statistical significance for hypotheses testing was set at 5%. All statistical test results were computed at the 2-tailed level of significance. Statistical analysis involves both descriptive and inferential statistics. According to research objective and research questions, this study has set the variables used in the study and their measurement is largely adopted from existing literatures. The following table 2 shows the variables and their measures. Table 2: Determinants of capital structure No Variables Measurement
1
Leverage
Long term debt / Total capital
2
Tangibility (Assets structure)
Book value of fixed assets/ Total assets (FA/TA)
3
Size
natural logarithm of sales Return on equity (ROE) (Profit before interest & taxes/ Total Assets X 100) Market to book ratio.
4
Profitability
5
Growth
6
Liquidity
Current assets/ current liabilities (CA/CL)
7
Dividend payout
Dividend per share (DPS)/Earnings per share (EPS)
5 .4. Development of Hypotheses This study has tested the following hypothesis on relations between the defined variable and capital structure of the listed companies. H1: There is a positive relationship between tangibility and leverage of the firm. H2: There is a positive relationship between size and leverage of the firm. H3: There is a negative relationship between profitability and leverage of the firm. H4: There is a positive relationship between growth rate and leverage of the firm. H5: There is a positive relationship between dividend payout and leverage of the firm. H6: There is a positive relationship between liquidity and leverage of the firm.
240
5 .5. Conceptual Model Based on data availability and some other constraints, only six variables such as tangibility, size, profitability, growth, dividend payout and liquidity are tested in this study. This can be conceptualized as follows;
Size (H2)
Tangibility
(H1 )
There is a positive relationship Profitability (H3)
Determinantbeof Capital s tween Structure
firm. H2: There is a Dividend payout (H5) iqui ive Lpositdity (H6) tangibility leverage of and the
Capital Structure
Growth rate (H4)
relationship Figure 1: Research Model between size and
6. Results and Discussions 6.1. Descriptive Statistics Descriptive statistics provide information on the key vleverage theoftud the d are presented below. Means, medians ariables in s y an and standard deviations for the each regression variables are presented in table 3.
firm. H3: There is a negative relationship between profitability leverage firm. H4: There is a positive 241 of and the
relationship
Table 3: Mean Values of Regression Variables (2002-2006) Variable Leverage 15 .62 Tangibility 0.47 Sales Growth Profitability Liquidity Dividend payout 8 .948 29.5 14 .99 15 0.40 0.28 0.81 83.9 14 .56 26 .54 0.59 19 .53 Mean Standard deviation
The table 3 indicates that the means for the leverage, tangibility, size, growth, profitability, liquidity and dividend payout ranged from a low of 0.40 to a high of 29.5. It therefore appears that the mean leverage (the proportion of long term debt in the total capital) in Sri Lanka is estimated as 15.62 % . Thus, the use of debt capital by Sri Lanka firm is quite low compared to G 7 countries (32.5%). The mean leverage, which is estimated using data for the financial year 2005, is 28.989%, drastically higher than the five year mean, which was observed earlier (15.62%). Table 4 below, reports the yearly mean and median figures for the leverage mentioned above. Table 4: Yearly Mean values of Leverage (2002-2006) Year 2002 2003 2004 2005 2006 Average Mean 15.496 17.218 14.757 28.989 1 .623 15.617 Median 9.02 10.8 9 .8 11.545 0 .775 8 .388
According to the above table, yearly mean leverage amount to 15%, and the corresponding median amounts to 8%. While we can observe a tendency of increasing mean leverage values in 2002, 2003 and 2005, a sudden decline observed in 2006 ( 1.623%).
242
The table 3 shows that, mean tangibility over the five year is 47% and the market value of assets is 29.5 times the book value of assets on mean. Further, the means of sales of the companies in the sample is Rs 16219170118 (means of natural logarithm of sales is 8.948). During the five year period, firms earned a mean ROE of 14.99%. Moreover, it indicates that the companies in the sample have current assets, which is 15 times greater than their current liability. The average dividend available to shareholders out of earnings is 0.4 times. 6 .2. Correlation Analysis
The Pearson's Moment Correlation Coefficient was computed for purposes of determining the following relationship: The relationship between leverage and tangibility The relationship between leverage and size The relationship between leverage and growth The relationship between leverage and profitability The relationship between leverage and liquidity The relationship between leverage and dividend payout. In order to delineate the relationship between the various determinants of capital structure, leverage was correlated and is presented in table 5. Table 5: Correlation matrix of regression variables Variable Leverage Tangibility Size Growth Profitability Liquidity Dividend Payout
Tangibility .019 Size .449(**) Growth -.096 Profitability .569(**) Liquidity .387(**) Dividend .127 payout
-.073 -.014 -.304(*) -.084 .347(*)
.282(*) .379(**) .188 -.020
.292(*) -.120 -.104
.391(**) -.112 -.034 1
** Correlation is significant at the 0.01 level (2-tailed) ** Correlation is significant at the 0.05 level (2-tailed) The results indicate that there is a direct and positive (weak) relationship between tangibility and leverage (r=0.019). This supports the hypothesis that there is a positive relationship between tangibility and leverage. However, the estimated coefficient on tangibility is statistically insignificant. (P > 0.01). A significant, strong and positive correlation is shown between size and leverage (r =0.449, P 0.01), rejecting the hypothesis that there is a positive relationship between growth and leverage. There was also a significant strong and positive relationship between profitability and leverage (r = 0.569, P