how firm specific factors affect capital structure

13
Middle Eastern Finance and Economics ISSN: 1450-2889 Issue 13 (2011) © EuroJournals Publishing, Inc. 2011 http://www.eurojournals.com/MEFE.htm How Firm Specific Factors Affect Capital Structure: An Emerging Market Practice – Istanbul Stock Exchange (ISE) Basak Turan Icke Department of Business Administration, Istanbul University Beyazit Campus, Istanbul, Turkey E-mail: [email protected] Tel: +90-212-4400000 Hunkar Ivgen Odak Auditing Abstract In this study, we examine the firm-specific factors which are influential on capital structure decisions of 212 industrial firms listed in Istanbul Stock Exchange over period 2004 and 2009 with Panel Data Analysis. The results of this study show that firm size, liquidity, profitability and sales growth affect the leverage ratios of industrials firms significantly. Among these factors, firm size and profitability are the most significantly influential factors on capital structures of industrial firms, and these two factors are negatively correlated with leverage ratios. Growth factor is statistically significant and positively correlated with leverage ratios. Liquidity factor is also statically significant but negatively correlated with leverage ratios. The results of this study are consistent with most of the capital structure literature and especially support Pecking Order Theory. Keywords: Capital Structure, Static Trade-off Theory, Pecking Order Theory, Agency Cost Theory, Istanbul Stock Exchange, Panel Data Analysis 1. Introduction Determination of capital structure is one of the most important decisions that must be taken by firms. The importance of making a decision on determining the capital structure was firstly introduced by the article published by Modigliani and Miller (MM) in 1958. Following the pioneering works of MM in this field, many critical studies have been made about the subject. In fact, the MM theory is not able to provide a practical definition on how a company should finance its activities. According to MM theorem, it is not possible to talk about an optimum capital structure where there is no tax and investors have perfectly diversified portfolios. The assumptions made by MM had been widely criticized. Then, after such criticisms, they reviewed their capital structure theory with the new article they published in 1963. In this study, corporate tax factor is taken into consideration and dividends are excluded from the model. Then, in 1977, Miller published another article and included corporate tax and individual income tax in their models. An optimum capital structure in which highest level of debt is preferred is accepted as practical due to tax advantage of debt. However, such capital structure has two important disadvantages. Firstly,

Upload: mohammed-farid-fathalbab

Post on 10-Oct-2014

109 views

Category:

Documents


2 download

TRANSCRIPT

Page 1: How Firm Specific Factors Affect Capital Structure

Middle Eastern Finance and Economics ISSN: 1450-2889 Issue 13 (2011) © EuroJournals Publishing, Inc. 2011 http://www.eurojournals.com/MEFE.htm

How Firm Specific Factors Affect Capital Structure: An

Emerging Market Practice – Istanbul Stock Exchange (ISE)

Basak Turan Icke

Department of Business Administration, Istanbul University

Beyazit Campus, Istanbul, Turkey

E-mail: [email protected] Tel: +90-212-4400000

Hunkar Ivgen

Odak Auditing

Abstract

In this study, we examine the firm-specific factors which are influential on capital

structure decisions of 212 industrial firms listed in Istanbul Stock Exchange over period 2004 and 2009 with Panel Data Analysis. The results of this study show that firm size, liquidity, profitability and sales growth affect the leverage ratios of industrials firms significantly. Among these factors, firm size and profitability are the most significantly influential factors on capital structures of industrial firms, and these two factors are negatively correlated with leverage ratios. Growth factor is statistically significant and positively correlated with leverage ratios. Liquidity factor is also statically significant but negatively correlated with leverage ratios. The results of this study are consistent with most of the capital structure literature and especially support Pecking Order Theory. Keywords: Capital Structure, Static Trade-off Theory, Pecking Order Theory, Agency

Cost Theory, Istanbul Stock Exchange, Panel Data Analysis

1. Introduction Determination of capital structure is one of the most important decisions that must be taken by firms. The importance of making a decision on determining the capital structure was firstly introduced by the article published by Modigliani and Miller (MM) in 1958. Following the pioneering works of MM in this field, many critical studies have been made about the subject. In fact, the MM theory is not able to provide a practical definition on how a company should finance its activities. According to MM theorem, it is not possible to talk about an optimum capital structure where there is no tax and investors have perfectly diversified portfolios.

The assumptions made by MM had been widely criticized. Then, after such criticisms, they reviewed their capital structure theory with the new article they published in 1963. In this study, corporate tax factor is taken into consideration and dividends are excluded from the model. Then, in 1977, Miller published another article and included corporate tax and individual income tax in their models.

An optimum capital structure in which highest level of debt is preferred is accepted as practical due to tax advantage of debt. However, such capital structure has two important disadvantages. Firstly,

Page 2: How Firm Specific Factors Affect Capital Structure

Middle Eastern Finance and Economics - Issue 13 (2011) 91

high level of debt causes high bankruptcy costs. The second disadvantage is that taxation on equity revenues is better than the taxation on debt for investor in terms of individual income tax. After MM theorem, three fundamental theorems have been developed on capital structure. These are Static Trade-off Theory, Pecking Order Theory and Agency Cost Theory.

Considering the Static Trade-Off Theory in these three fundamental theories; joint analysis of taxes, the state of insolvency and agency costs underlies the model. Static Trade-Off Theory focuses on the advantages of tax shield and disadvantages of the state of insolvency. According to the Pecking Order Theory, while the firms are financing new investment opportunities, they firstly prefer internal financing resources and lastly issuing new shares. Lately, Jensen and Meckling developed Agency Cost Theory in 1977. In this theory, a firm has two types of agency costs. One is related to current equity holders and other one is the debts in capital structure.

In this study, the factors affecting capital structure of 212 industrial firms being dealt at ISE (Istanbul Stock Exchange) between 2004–2009 and making business in industry sector are examined. Simple Panel Regression Model is adopted as a method. In evaluating the findings, it is observed that firm size, liquidity, profitability and growth are the factors that make impact on firms’ leverage ratios. Among these four factors, “firm size” and “profitability” make relatively higher impact and the correlation of these variables with leverage ratio results in negative direction. In analysis made on growth factor, statistically significant and positive relationship is revealed. With regards to liquidity variable, it can be said that its relationship with leverage is statistically significant and in negative direction.

2. Conceptual Foundations of Capital Structure In this section, theoretical approaches on capital structure and firm specific factors of capital structure are examined. These are discussed in the following sub-sections. 2.1. Theoretical Framework of the Term “Capital Structure”

Capital structure consists of the combination of various resources that establish equity and debt. The debt element in capital structure is a tool utilized in order to increase profitability. It is known that debt provides firms with some advantages. The most important one among these advantages is the tax shield brought by the interest due to debt. On the other hand, there are also some disadvantages that debt possesses. Debt is one of the reasons of financial insolvency for mismanaged firms (Ramagopal, 2008).

The history of contemporary criticisms on capital structure of firms dates back to Modigliani and Miller. The studies of MM published in 1958 have altered traditional view on corporate finance. According to traditional view, debt financing takes place with lower cost compared to equity financing. In contrast to traditional view, MM hypothesizes the existence of a perfect capital structure and utilizes a simple arbitrage mechanism introducing three hypotheses on firm value, equity cost and anticipated rate of return (Prasad, Green and Murinde, 2001).

MM discusses the capital structure from the point that there is no relation between capital structure and firm value. They emphasize market efficiency as the foundation of this idea. This perspective represents the first difference between traditional approach and MM approach. In this perspective, investment policies are adopted as data, and tax and agreement costs are ignored. In doing so, it is argued that the employed decision on financing will not affect market value of the firm. According to the studies of MM in 1958, the only element that affects firm value is investment decision (Brigham and Gapenski, 1996). In subsequent studies, it is seen that the said no relation hypothesis is ignored (Degryse, Goeij and Kappert, 2009).

The taxation variable which was ignored in earlier studies of MM was included in the analysis in their subsequent studies published in 1963 and known as review of the first article. In their first article, MM argue that there is no relation between capital structure and cost of capital, however, they regarded this argument as a mistake in their subsequent article (Gifford, 1998). Another important

Page 3: How Firm Specific Factors Affect Capital Structure

92 Middle Eastern Finance and Economics - Issue 13 (2011)

point is that bankruptcy and agency costs were not covered in both works of MM. Exclusion of these two variables from the model makes significant impact on the findings.

After such discussions, three widely-criticized different theories on capital structure were developed in finance literature; they criticize the ideas of MM. These are Static Trade-off Theory, Pecking Order Theory and Agency Cost Theory. In many works made in this field, the validity of said theories is tested.

Considering the Static Trade-Off Theory, the focus is on the advantages of tax shield and disadvantages of financial insolvency costs. The theory, referred to as “Tax Advantage Theory” in the literature, defines the best combination of benefit and costs of debt as the optimum capital structure of firm (Vernimmen, Quiry, Dallacchio, Le Fur and Salvi, 2005).

In Static Trade-Off Theory, the objective is to establish the trade-off between tax advantage of debt and its benefit in decreasing agency costs, and costs of bankruptcy resulting from leverage after debt and costs of financial insolvency (Buferna, Bangassa and Hodgkinson, 2005). In this model, agency costs, taxes and bankruptcy costs lead highly profitable companies to have high leverage.

Therefore, the positive relation is expected between financial leverage ratio and profitability. In this model, firms adopt an objective of financial leverage level and they try to reach that level (Fama and French, 2002). Ju et al., who argue that the main factors affecting optimum capital structure of firms are corporate tax and bankruptcy costs, test Trade-Off Theory, and they state in their studies that Trade-Off Model is successful in anticipating the capital structure of firms with characteristic debt level (Ju, Parrino, Poteshman and Weisbach, 2005).

It is accepted that firms may favour debt over equity or equity over debt until they maximize firm value. According to Myers (1984), Static Trade-Off Model is acknowledged in literature since this model looks like practical and specifies an internally appropriate debt rate. This model regards modest level of debt as realistic. However, it must be questioned whether this model explains financing policies of companies. If the model does not provide an answer fort that question, then a better theory is needed before informing directors on capital structure (Myers, 1984).

The second theory is Pecking Order Theory, also known as Asymmetric Information Theory. The origin of this theory dates back to the works of Myers and Majluf. Pecking Order Theory is based on minimum effort principle. According to this theory, financial decisions of firm directors depend on a hierarchical structure (Bhaduri, 2002). Hierarchical decisions of directors can be summarized as follows (Myers, 1984):

i. Firms prefer internal financing. ii. Although the dividends are fixed, firms try to adapt their target dividend payment rates to

investment opportunities. iii. Internally generated cash flows can be either higher or fewer than capital costs due to sticky

dividend policies and unexpected fluctuations in the profitability of an investment opportunity. If the generated cash flows are less than cost of capital, the first action of firm will be decreasing cash balance or marketable securities portfolio.

iv. If external financing is needed, the firms issue the most reliable security at first. In other words, the firms start financing with debt, and then mixed securities such as convertible bonds are preferred. In this context, there is no specific debt-equity target for the firm since there exist two types of equity financing; internal financial and external financing. According to Pecking Order, internal financial is at the top of such hierarchy while external financing is undermost. Observed debt rate of every firm reflects its cumulative external financial need. In Pecking Order Theory, there is an idea that “a firm should issue its securities with lowest

information costs at first before issuing securities with higher information cost”. This statement recommends the consumption of short-term debt tools before issuing long-term debt tools (Frank and Goyal, 2003).

Pecking Order Theory is a theory that aims to explain why firms with high profitability utilize less debt and why firms with low profitability prefer financing by utilizing high rates of debt. According to the theory, a firm with high profitability will not need external fund. However, a firm

Page 4: How Firm Specific Factors Affect Capital Structure

Middle Eastern Finance and Economics - Issue 13 (2011) 93

with low profitability will prefer external financing over share issue since it does not perform sufficient fund-raising and debt is less costly compared to share issue (Drobetz and Fix, 2003).

The Trade-Off Theory and the Pecking Order Theory have a number of predictions regarding the debt-equity choice. The predictions related to the two theories are summarized in Table 1. Table 1: Capital Structure Theory And Expected Sign On Leverage For Explanatory Variables

Trade-Off Theory Pecking Order Theory

Firm Size + + Collateral + + Profitability + - Growth Opportunities - +

Source: Degryse, H., et al., 2009. “The Impact of Firm and Industry Characteristic on Small Firms’ Capital Structure: Evidence from Dutch Panel Data”, Discussion Paper, European Banking Center Discussion Paper No. 2009-03, Tilburg University, s. 4.

According to Ross (1977), a director avoids utilizing debt that will distress the firm. It debt is utilized, and then it will signal the high quality of firm. In his study, Ross advocates the argument that firm value and debt/total equity rate are positively correlated (Ross, 1977).

Agency Cost Theory indicates that optimum capital structure can be defined by minimizing the costs arising from conflict between the parties since agency costs play an important role due to potential conflicts between partners and company creditors (Buferna et al., 2005).

According to Jensen and Meckling (1976), agency costs are the sum of monitoring costs, bonding costs, and residual loss. In their studies evaluating the advantages of external financing, Jensen and Meckling define agency relationship as a contract assigned by decision authority to another person in order to have one or more persons perform specific services in the name of such authority. By raising fund from the represented (investor), the agent (director) will utilize such funds for the benefit of investor. According to the theory, both investor and director are rational in financial decisions.

Even if a new firm is financed completely by equity and risk-free debt, it will experience agency costs. It is the case in Agency Cost Theory that the composition of a firm’s assets affects the decision to be made on that firm’s capital structure. In Agency Cost Theory, the partners of a company with high financial leverage ratio have an investment trend under optimum level (Bhaduri, 2002).

Considering the abovementioned theories, decisions on capital structure consisting of debt and securities are specific to both firm and industrial determinants. In Static Trade-Off Theory, it is seen that tax rates, type of asset, business risk, profitability and bankruptcy costs are among the elements that affect capital structure. In Agency Cost Theory, type of assets and growth opportunities for the firm are important determinants.

Pecking Order Theory is associated with transaction costs, asymmetric information and the capacity to undertake new investments with internal funds. In case the firm needs external funds, then it will prefer debt with less asymmetric information over equity issue (Booth, Aivalian, Demirguc-Kunt and Maksimovic, 2001). 2.2. Firm Specific Factors of Capital Structure

In empirical and theoretical studies made by this time on capital structure, it has been introduced that there are many internal and external determinants affecting leverage levels of firms. It can be said that such determinants are important elements constituting firms’ decisions on debt-equity choice.

In studies made in 80s, it was presented that the determinants affecting capital structure of firms are generally classified as active structure, growth opportunities, size, non-debt tax shield, variability in earnings and profitability (Titman and Wessel, 1988). Today, the efficiency level of different variables has been increased. Periodically, there appears the need to re-evaluate the impact of some determinants

Page 5: How Firm Specific Factors Affect Capital Structure

94 Middle Eastern Finance and Economics - Issue 13 (2011)

on capital structure. In fact, the main objective has been to examine the relation between capital structure and firm value.

When the studies in literature are considered, it can be seen that the relation between firm value and leverage is fundamental to analysis in capital structure analyses. The reasons for that are the development of theories on leverage, accurate determination of shift date in leverage and the use of leverage in estimating bankruptcy and bond valuation (Bowman, 1995).

Many studies (Bradley et al. (1984) [BJK], Chaplinsky and Niehaus (1990) [CN], Friend and Hasbrouck (1988), and Friend and Lang (1988) [FH/L], Gonedes et al. (1988) [GLC], Long and Malitz (1985) [LM], Kester (1986) [Kest], Kim and Sorensen (1986) [KS], Marsh (1982) [Mar.], Titman and Wessels (1988) [TW]) show that leverage levels of firms increase through fixed assets, non-debt tax shield, growth opportunities and firm size; however, decrease through volatility, advertising costs, R&D costs, bankruptcy probability, profitability and uniqueness. The results of these studies are shown in Table 2 (Harris and Raviv, 1991). Table 2: Determinants of Leverage

Characteristic BJK CN FH/L GLC LM Kest. KS Mar. TW

Volatility - - -* + -* Bankruptcy

Probability

-

Fixed Assets + + + + +* Non-Debt Tax

Shields

+ + - -*

Advertising Costs - - R & D Costs - - Profitability - -* +* - - Growth

Opportunities

-* + - -*

Size -* +* -* -* + -* Free Cash Flows - Uniqueness -

* Indicates that the result was either not statistically significantly different from zero at conventional significance levels or that the result was weak in non-statistical sense. Source: Harris M., Raviv A.1991. The Theory of Capital Structure, The Journal of Finance, 46 (1- March): s. 336.

Kayhan and Titman (2003) state that cash flows, investment costs and share price history affect capital structure choice. When the findings are examined, after it is specified that the relation between debt rate and firm value is weak, it is stated that especially the stock yield is one of the important elements affecting leverage ratio.

Titman and Wessels (1988) suggest that security value of assets, non-debt tax shield, growth, uniqueness, sector, size, volatility and profitability are the variables affecting capital structure. Considering the findings, a negative relationship is identified between debt rates and uniqueness of activity field. Also, transaction costs are also specified as an important variable that possibly affect capital structure choice. Short-term debt is negatively correlated with firm size. In the study, it is also observed that there is no impact originating from debt rates, non-debt tax shield, volatility, security value of assets or growth opportunities.

Weston (1963) examines the relation between growth in earnings and equity cost with regression analysis and concludes that this relation is statistically significant and negative. In study by Fama and French (1998) the relation between firm value and leverage is analyzed. Dependent variable in the study is identified as the difference between total market value and total asset book value while explanatory variable is identified as tax, dividend, profitability, investment and R&D costs.

According to Harris and Raviv (1991), leverage increases depending on fixed assets, non-debt tax shield, investment opportunities and size while it decreases depending on volatility, advertising costs, bankruptcy probability, profitability and uniqueness.

Page 6: How Firm Specific Factors Affect Capital Structure

Middle Eastern Finance and Economics - Issue 13 (2011) 95

Rajan and Zingales (1995) establish their study on tangibility of assets, (fixed assets/total assets), market value/book value (as a sign of investment opportunities, firm size and profitability. The authors state two reasons for limiting their study only to these elements. The first one is that these elements have been associated with leverage in previous studies. The second one is that the collected data is limited as it does not allow examining other determinants. Tangibility of assets shows the effect of security value of assets on leverage level of firm (Rajan and Zingales, 1995). When the studies made on the influence of this determinant on leverage of firms are considered, there are studies positive and negative results as well as the ones with presenting its ineffectuality.

In terms of firm size, there are studies identifying that these element is positive, negative and insignificant with leverage ratios of firms. However, such studies also indicate that there is significance about the positivity of relationship between leverage and firm size. For example, the studies by Rajan and Zingales and Jordan et al. support this result. Also, the study which examines 25 developing countries and capital structures of these countries reveals significant results especially on the influence of size. Accordingly, as firms gain size, risk of failure will decrease and leverage ratios will increase (Bas, Muradoglu and Phylaktis, 2009). Huang and Song summarize the determinants of capital structure, definitions and predicted signs of these determinants and results of major empirical studies in Table 3. Table 3: Summaries of Determinants of Capital Structure, Theoretical Predicted Signs and the Results of

Previous Empirical Studies

Proxy Definitions Theoretical Predicted

Signs

Major Empirical Studies’

Results

Profitability (ROA) EBITDA / Total Assets +/- - Size Natural Logarithm of Sales +/- + Tangibility Fixed Assets / Total Assets + + Tax Effective Tax Rate + + Non-Debt Tax Shields Depreciation / Total Assets - - Growth Opportunity Tobin’s Q - - Volatility Standard Deviation of

EBITDA +/- -

Managerial Equity Ownership

Total Percentage of Directors and Top Managers

+/- +/-

Ownership Structure (Institute)

Institutional Shareholding ? ?

Source: Huang, G., Song, F. M. 2006. The Determinants of Capital Structure: Evidence from China, China Economic Review, 17, s. 22.

Deari and Deari (2009) use independent variables of profitability, tangibility, size, growth and non-debt tax shield in their analysis on publicly listed firms in Macedonia between 2005–2007. The findings of study are in coherence firstly with the assumptions of Pecking Order Theory and secondly with Static Trade-Off Theory. The results do not confirm the assumptions of Agency Costs Theory.

Huang and Song (2006) state in their study that the determinants affecting leverage ratios of firms in other countries are the same with Chinese firms. In the study, it is identified that leverage ratio ([Short Term Debts+Long Term Debts]/Total Assets) decreases when elements like profitability, non-debt tax advantage and managerial equity ownership are taken into consideration; however, it increases when it is evaluated in terms of firm size. Tangibility and tax rate have positive impact on long-term debt rate and total debt rate. It is also found that firms with higher growth opportunities have less leverage ratios.

In studies made by Miguel and Pindado (2001) on analyzing Spanish firms, it is identified that Spanish firms have more on-debt tax shield compared to American firms. Also, it is stated that there is a negative correlation between insolvency costs and debt.

Page 7: How Firm Specific Factors Affect Capital Structure

96 Middle Eastern Finance and Economics - Issue 13 (2011)

Booth et al. (2001) revealed coherent findings with Pecking Order Theory in their study examining capital structure of firms in ten developing countries. Additionally, they identified that there is a significant problem of asymmetric information. According to these findings, external financing is high cost and firms tend to avoid external financing.

In a study made in Nepal, an analysis was made by using seven explanatory variables; business risk, size, growth, earning ratio, dividend distribution, debt capacity and activity leverage ratio. Based on the findings, growth, size and earning ratio have 72% rate of explaining financial leverage, and the remaining variables have 5% explanatory capacity on leverage. It is identified in the study as a conclusion that firm size, growth rate and profitability variables play important role in determining leverage in Nepal (Baral, 2004).

3. Model, Analysis of the Study and Evaluation of Findings In this section, the scope of study and evaluated data set, analysis method utilized in the study and analyses made to select the used panel data method are explained. 3.1. Determination of Scope, Data Set and Method of the Study

In this study, the determinants affecting capital structure of firms in ISE (İstanbul Stock Exchange) are examined with panel data analysis. 212 industrial firms operating consistently between 2004 and 2009 are included within the scope of study. Firms which are industry sector but excluded from ISE for bankruptcy, merger and other reasons are not included in this study. Financial data on firms are obtained from the website of ISE. EViews 7.0 and Stata 11 applications are used in statistical analysis of data. In the study, the determinants which are established in literature as having influence on capital structures are taken as basis and firm specific variables are identified as in Table 4. Table 4: Variables Used in the Study and Definitions

Variable Proxy (Representation) Definition

Leverage Ratio TD/TA (Short Term Debts+Long Term Debts)/Total Assets

Agency Cost Expenses Rate Operating Costs/Sales Asset Utilization Rate Sales/Total Assets Taxes Tax Rate Tax Reserves/Profit Before Tax Non-Debt Tax Shield Depreciation Rate Depreciation Cost/Total Asset Growth Growth in Sales (Sales1-Sales0)/Sales0

Firm Size Total Assets Total Assets Tangibility of Assets Net Fixed Asset Rate Net Fixed Asset/ Total Assets Profitability Return on Assets Net Profit/Total Assets Liquidity Current Ratio Current Assets/Short Term Debts

In literature, three different basic representative rates are utilized for financial leverage. For

example, Gaver and Gaver (1993) utilize Long term liabilities/common equity; Remmers et al. (1974), and Rahman (1990) utilize Long-term Liabilities/Total Assets, and Bradley, et al. (1984) and Wald (1999) utilize Total Debt/Total Assets [(Short Term Debts+Long Term Debts)/Total Assets] as representation for capital structure. Since short-term debt rates is high in Turkey, the definition of leverage used by Bradley (1984) and Wald (1999) is utilized.

Panel data regression analysis is utilized in explaining the relation between leverage ratio and variables. Simultaneous examination of both time and units is called as panel data analysis. In other words, panel data analysis is the method to estimate financial relations by using cross-sectional data in time (Greene, 1997). Simple panel regression model is described as follows:

Yi,t = αi + β’Xi,t + εit

Page 8: How Firm Specific Factors Affect Capital Structure

Middle Eastern Finance and Economics - Issue 13 (2011) 97

Panel data has multi directional surveys. The situation of each i=1…N cross sectional survey is examined in t=1…T term. In this context, the study includes 1272 surveys in total as 212 sections and 6 terms. 3.2. Panel Data Analysis and Evaluation of Findings

In order to test determinants affecting capital structure, a panel data regression analysis is performed between leverage ratio and explanatory variables of agency cost [representative variables operating expenses/sales (OES) and asset turnover (AT)], tax (TX), non-debt tax shield (DVK), growth (G), firm size (TA), change in profits (CIP), tangibility of assets (TOA), profitability (ROA) and liquidity (CO). Before regression analysis, stationarity test and methodology study in panel data regression analysis were made. Then, the existence of autocorrelation and constant variance in model was examined. The findings of study are summarized below. 3.2.1. Stationarity

Granger and Newbold (1974) revealed that fake regression problems can be experienced in case non-stationary time series are examined. In panel data analysis, the series must be stable in order to prevent fake correlations.

Since cross-sectional size of panel is bigger than time section in panel data (N>T), Levin-Lin-Chu (LLC) Test is appropriate for stationarity test. This method is especially recommended for panel data analyses containing industrial and firm level data (Levin, Lin, Chu, 2002). The hypothesis on the stationarity of LLC is as follows:

Ho: General unit root is available. H1: General unit root is not available Null hypothesis of LLC test indicates that the series is not stationary. Rejection of null

hypothesis states that the series is stationary. As is seen in Table 5, Ho hypothesis is rejected at 5% of significant level for all series related to dependent and independent variables used in the study. In other words, the series are stationary. Table 5: Results Of Levin, Lin & Chu Unit Root Test

Variables Statistic Probability

ROA -16.5385 0.0000 TDTA -15.5088 0.0000 TA -5.39466 0.0000 AT -79.9758 0.0000 TX -12.2951 0.0000 CIP -2127.00 0.0000 OES -7.31711 0.0000 TOA -9.29876 0.0000 CR -13.7126 0.0000 G -717.211 0.0000

3.2.2. Selection of Panel Data Method

There are three regression techniques used with panel data: Pooled Ordinary Least Square, Fixed-Effects Models and Random-Effects Model. Among these three models, Breusch-Pagan test (BP) (Greene, 2003) is applied in order to see whether classical model (pooled) can be performed, and Hausman test is applied in choice among random and fixed effects models. Apart from these, Redundant Fixed-Effects Test which is used in non-fixed panels and indicating the availability to use fixed-effects method, can be utilized.

Page 9: How Firm Specific Factors Affect Capital Structure

98 Middle Eastern Finance and Economics - Issue 13 (2011)

Table 6: Panel Data Model Choice

1.BREUSH/PAGAN LM TEST Chi_Square P- Value

Ho; Var (u) = 0 1803.4 0.0

2.HAUSMAN TEST Chi_ Square P- Value

Ho: Random Effects is Available. 527.04 0.0

3.REDUNDANT FIXED-EFFECTS TEST Chi_ Square P-Value

Ho: No Fixed Effects in Section 2156.4 0.0

Ho: No Fixed Effects in Period 110.8 0.0

Ho: No Fixed Effects in Section and Period 2177.9 0.0

PANEL DATA METHOD CHOICE Invariant in Section, Invariant in Period

The results of tests made on panel data method choice are given in Table 6. According to BP

test results, Ho hypothesis is rejected at 5% of significance level and thus, there is no need to pool the data. Then, Hausman test is applied in choice between random effects-fixed effects. According to test results, Ho hypothesis is rejected at 5% of significance level. In other words, Hausman test showed the need to use fixed effects model. The results of Redundant Fixed Effects Test which indicate the rejection of Ho hypothesis at 5% of significance level support the results of Hausman test. As is seen in Table 6, the results of Redundant Fixed Effects Test on two-way (cross-section and period) show the need to employ Fixed Effect Method in cross-section and period. 3.2.3. Autocorrelation, Heteroscedasticity and Cross-Sectional Dependence, Contemporaneous

Correlation

Sequential dependence in panel data is associated with units in time or section. Wooldridge test is applied in order to reveal sequential dependence. Hypothesis and results of the test are as follows:

H0: No first order autocorrelation. F ( 1,211) = 35.22 Probability > F = 0.00 The results of autocorrelation show that there is autocorrelation at 5% of significance level.

Wald test, which can be used in fixed effects regression analysis methods, is applied in order to identify the availability of heteroscedasticity. Hypothesis and results of the test are as follows:

H0: No heteroscedasticity (Constant variance is available). Chi_Square (212) = 1.2e+06 Probability>Chi_Square = 0.00 The results of Wald Test show the need to reject Ho hypothesis at 5% of significance level.

Therefore, it is identified that heteroscedasticity is available in the model. Pesaran Test is applied in testing cross sectional dependence. The results of Pesaran Test are as follows;

Pesaran Value = 29.72 Probability = 0.00 The results of Pesaran test show that there is cross-sectional dependence. Therefore, Panel

Corrected Standard Errors (PCSE) method is applied in correcting standard errors since both cross-sectional dependence and heteroscedasticity problem exist (Bjørnstad and Nymoen, 2008). 3.2.4. Results of Panel Data Regression Analysis

The results of two-way fixed effects panel data regression analysis are given in Table 7. As is seen in the table, the result of F statistic shows that model can be employed in aggregate; and the result of R2 value as 0.85 shows that independent variables in the model can explain 85% of variation in dependent variable.

Page 10: How Firm Specific Factors Affect Capital Structure

Middle Eastern Finance and Economics - Issue 13 (2011) 99

Table 7: Results of Panel Data Regression Analysis

Dependent Variable: TDTA Date: 2004 - 2009 Total Survey: 1216 Cross-section SUR (PCSE) standard errors & covariance Variable Coefficient Std. Error t-Statistic Prob.

TA -0.258411 0.075495 -3.422864 0.0006 AT -6.16E-05 5.88E-05 -1.047623 0.2951 TX -6.37E-05 0.001367 -0.046570 0.9629 CIP 0.000635 0.000724 0.876552 0.3809 OESS 0.021208 0.015539 1.364804 0.1726 TOA -0.017122 0.104156 -0.164392 0.8695 CO -0.002711 0.000827 -3.278317 0.0011 G 0.007100 0.001197 5.932440 0.0000 ROA -0.342275 0.114681 -2.984580 0.0029 C 2.678042 0.589893 4.539878 0.0000 R-squared 0.850286 Adjusted R-squared 0.816630 F-statistic 25.26439 Prob(F-statistic) 0.000000

According to findings, Firm Size (TA) Liquidity (CO), Growth (G) and Profitability (ROA)

have effect on leverage ratio (TDTA). From these four determinants, “firm size” and “profitability” make the most effect on leverage ratio and there is a negative relation between these determinants and leverage ratio. The results of this study are in parallel with the results of significant part of studies previously made in this field.

4. Conclusion In this study, the factors affecting capital structure of 212 industrial firms in ISE (Istanbul Stock Exchange) are examined between 2004 and 2009. According to the findings of this study, four firm specific factors; firm size, liquidity, growth in sales and profitability make statistically significant effect on leverage ratios of firms (at 5% significance level). The result, 0.85, explanatory power of independent variables of utilized model on leverage ratio shows highly important effect of statistically significant four factors (firm size, liquidity, growth in sales and profitability) on explaining variations in leverages of the firms.

Considering these four determinants, “firm size” and “profitability” make the most effect on leverage ratio and there is a significant and negative relation between firm size and profitability, and leverage ratio. Firstly, the negative relation found between leverage and profitability shows us that Pecking Order Theory, which is one of the capital structure theories, is valid for the evaluated firms. As mentioned above, this theory states that in case firms are in need of resource, they firstly try to meet such need through their internal resources. Accordingly, since profitable firms will need relatively less external financing, it is stated that there will be adverse relation between profitability and leverage ratios. Considering the studies in this field, there have been found a negative relation between profitability and leverage ratio in many studies as in Kester (1986), Friend and Lang (1988), Titman and Wessels (1988), Allen and Mizuno (1989), Chowdhury and Miles (1989), Rajan and Zingales (1995) Jordan et al. (1998).

Statistically significant (5% significance level) negative relation is identified between firm size and leverage. This can be a result of the fact that relatively large firms enter equity markets more easily and with less cost. The findings show parallelism with the result of studies made by Archer and Faerber (1966), Gupta (1969), Scott and Martin (1975), Titman and Wesels (1988), and Anderson and Makhija (1999) In analysis made for growth factor, it can be said that there is statistically significant

Page 11: How Firm Specific Factors Affect Capital Structure

100 Middle Eastern Finance and Economics - Issue 13 (2011)

and positive relation. In can be stated that coherent findings with the studies of Gupta (1969), Higgins (1977) and Ellsworth (1983) are identified considering the results related to the growth. In terms liquidity variable, it is identified that the relation between leverage and liquidity is statistically significant and negative. These findings are in parallel with the studies of Titman and Wessels (1988) and Baskin (1989), Deesomsak, Paudyal and Pescetto (2004). At this point, it is important to say that liquid firms prefer internal sources more compared to debt. It is also presented that the analyses did not reveal significant results for other variables covered in this study (agency costs, tax, tangibility of assets, non-debt tax shield, change in profits).

References [1] Anderson, C. W., and A. K. Makhija, 1999. “Deregulation, Disintermediation, and Agency

Costs of Debt: Evidence from Japan”, Journal of Financial Economics, 51, pp. 309–399. [2] Baral, K.J., 2004. “Determinants of Capital Structure: A Case Study of Listed Companies of

Nepal”, The Journal of Nepalese Business Studies, 1 (1-December), pp. 1-13. [3] Bas, T., Muradoglu, G., and Phylaktis, K., 2009, “Determinants of Capital Structure in

Developing Countries”, Working Paper, http://efmaefm.org/0EFMSYMPOSIUM/China-2010/papers/determinants%20of%20capital%20structure%20in%20developing%20countries.pdf

[4] Bhaduri, S.N., 2002. “Determinants of Capital Structure Choice: A Study of the Indian Corporate Sector”, Applied Financial Economics, 12 (9), pp. 655-665.

[5] Bjørnstad, R. and Nymoen R., 2008. “The New Keynesian Phillips Curve Tested on OECD Panel Data”, Economics Discussion Papers, No 2008-4.

[6] Booth L., Aivazian V., Demirgüç-Kunt A., Maksimovic V., 2001. “Capital Structure in Developing Countries”, The Journal of Finance, 56 (1-February), pp. 87-130.

[7] Bowman, R.G., 1995. “Information Content of Financial Leverage; An Empirical Study: A Comment”, Journal of Business Finance & Accounting, 22 (3), pp. 455-460.

[8] Brigham E.F., Gapenski L.C., 1996. Intermediate Financial Management, 5th Edition, Illinois: The Dryden Press, Hinsdale.

[9] Buferna F., Bangassa K., Hodgkinson L., 2005. “Determinants of Capital Structure Evidence from Libya”, Research Paper Series, No: 2005/8, The University of Liverpool.

[10] Deari, F., Deari, M., 2009. “The Determinants of Capital Structure: Evidence from Macedonian Listed and Unlisted Companies”, Analele Stiintifice ale Universitatii Alexandu Ioan Cuza din Iasi, 56 (November), pp. 91-102.

[11] Degryse H., De Goeij P., Kappert P., 2009. “The Impact of Firm and Industry Characteristics on Small Firms’ Capital Structure: Evidence from Dutch Panel Data”, European Banking

Center Discussion Paper, No: 2009-03, CentER, Tilburg University. [12] Demirhan, D., 2009. “Sermaye Yapısını Etkileyen Firmaya Özgü Faktörlerin Analizi: İMKB

Hizmet Firmaları Üzerine Bir uygulama”, Ege Academic Review, 9(2), pp. 677-697. [13] Driscoll, J. and Kraay, A. 1998. “Consistent Covariance Matrix Estimation with Spatially

Dependent Panel Data”, Review of Economics and Statistics, 80(4), pp. 549-560. [14] Drobetz W. and Fix R., 2003. “What are the Determinants of Capital Structure? Some Evidence

for Switzerland”, WWZ Department of Finance, Working Paper, No. 4/03. [15] Fama, E.F., and French K.R., 1998. “Taxes, Financing Decision, and Firm Value”, The Journal

of Finance, 53 (3- June), pp. 819-843. [16] Fama, E.F., and French K.R., 2002. “Testing Trade-off and Pecking Order Predictions About

Dividends and Debt”, The Review of Financial Studies, 15 (1 – Spring), pp. 1-33. [17] Frank M.Z., and Goyal V.K., 2003. “Testing the Pecking Order Theory of Capital Structure”,

Journal of Financial Economics, 67, pp. 217-48. [18] Gifford, D., 1998. “After the Revolution”, CFO Magazine, 1 (July), pp. 1-5.

Page 12: How Firm Specific Factors Affect Capital Structure

Middle Eastern Finance and Economics - Issue 13 (2011) 101

[19] Granger, C.W.J., and Newbold P., 1974. “Spurious Regressions in Econometrics”, Journal of

Econometrics, 2, pp. 111-120. [20] Greene, William H., 1997. Econometric Analysis, Prentice Hall, Third Edition, New Jersey. [21] Greene, William H. 2003. Econometric Analysis, International Edition, Pearson Education Inc.,

Fifth Edition. [22] Gupta, M.C., 1969. “The Effect of Size, Growth, and Industry on the Financial Structure of

Manufacturing Companies”. Journal of Finance, 24, pp. 517-29. [23] Harris M., and Raviv A., 1991. “The Theory of Capital Structure”, The Journal of Finance, 46

(1-March), pp. 297-355. [24] Huang, G., and Song, F.M., 2006. “The Determinants of Capital Structure: Evidence from

China”, China Economic Review, 17, pp. 14-36. [25] Jensen M. C. and Meckling W.H., 1976. “Theory of the Firm: Managerial Behavior, Agency

Cost and Ownership Structure”, Journal of Financial Economics, 3 (4), pp. 305-60. [26] Ju, N., Parrino, R., Poteshman A.M. and Weisbach, M.S., 2005. “Horses and Rabbits? Trade-

Off Theory and Optimal Capital Structure”, The Journal of Financial and Quantitative

Analysis, 40 (2 – June), pp. 259-281. [27] Kayhan A. and Titman S., 2003. “Firms’ Histories and Their Capital Structure”, URL:

http://papers.ssrn.com/sol3/papers.cfm?abstract_id=552144 ]; 06.11.2010. [28] Korkmaz, T., Başaran, Ü., Gökbulut, R.İ., 2009. “İMKB’de İşlem Gören Otomotiv ve

Otomotiv Yan Sanayi İşletmelerinin Sermaye Yapısı Kararlarını Etkileyen Faktörler: Panel Veri Analizi”, İktisat, İşletme ve Finans, 24 (277), pp. 29-60.

[29] Kraus, A. and Litzenberger, R.H., 1973. “A State-Preference Model of Optimal Financial Leverage”, Journal of Finance, pp. 911-922.

[30] Levin, A., Lin C. and Chu, C.J., 2002. “Unit Root Tests In Panel Data: Asymtotic and Finite-Sample Proporties”, Journal of Econometrics, 108, pp. 1-24.

[31] Miguel, A. and Pindado, J., 2001. “Determinants of Capital Structure: New Evidence from Spanish Panel Data”, Journal of Corporate Finance, 7, pp. 77-99.

[32] Miller, M., 1977. “Debt and Taxes”, The Journal of Finance, 32 (2), Papers and Proceedings of the Thirty-Fifth Annual Meeting of the American Finance Association, Atlantic City, New Jersey, September 16-18, pp. 261-275.

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

[34] Modigliani, F. and Miller, M., 1958. “The Cost of Capital, Corporation Finance and the Theory of Investment”, American Economic Review, pp. 261-297.

[35] Myers, S., 1984. “The Capital Structure Puzzle”, The Journal of Finance, 39 (3- July), pp. 575-592.

[36] 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 (2), pp. 187-221.

[37] Pettit, J., 2007. Strategic Corporate Finance: Applications in Valuation and Capital Structure, John Wiley & Sons, Inc., New Jersey.

[38] Prasad S., Green C.J., and Murinde V., 2001. “Company Financing, Capital Structure, and Ownership: A Survey, and Implications for Developing Economies”, Société Universitaire Européenne de Recherches Financiérea, Vienna, SUERF Studies: 12.

[39] Qureshi, M.A. and Azid, T., 2006. “Did They do it Differently? Capital Structure Choices of Public and Private Sectors in Pakistan”, The Pakistan Development Review, 45 (4- Winter), pp.701-709.

[40] Rajan, R.G., and Zingales, L., 1995. “What do We Know about Capital Structure? Some Evidence from International Data”, The Journal of Finance, 50 (5-December), pp. 1421-1460.

[41] Ramagopal, C., 2008. Financial Management, New Age International, Daryaganj, Delhi, IND.

Page 13: How Firm Specific Factors Affect Capital Structure

102 Middle Eastern Finance and Economics - Issue 13 (2011)

[42] Ross, S.A., 1977. “The Determination of Financial Structure: The Incentive Signalling Approach”, The Bell Journal of Economics, 8 (1), pp. 23-40.

[43] Sayılgan, G., Karabacak, H., and Küçükkocalıoğlu, G., 2006. “The Firm-Specific Determinants of Corporate Capital Structure: Evidence from Turkish Panel Data”, International Journal of

Investment Management and Financial Innovations, 3. [44] Scott, D. F., Jr., and J. D. Martini, 1975. “Industry Influence on Financial Structure”, Financial

Management, pp. 67-73. [45] Terim, B. and Kayalı, C.A., 2009. Sermaye Yapısını Belirleyici Etmenler: Türkiye’de İmalat

Sanayi Örneği, Celal Bayar Üniversitesi Sosyal Bilimler Dergisi, 7 (1), pp. 125-154. [46] Titman, S. and Wessels R., 1988. “The Determinants of Capital Structure Choice”, The Journal

of Finance, 43, pp. 1-19. [47] Vernimmen P., Quiry P., Dallocchio M., Le Fur Y., and Salvi A., 2005. Corporate Finance:

Theory and Practice, John Wiley & Sons, Inc., England. [48] Weston, F.J., 1963. “A Test of Cost of Capital Proposition”, The Southern Economic Journal,

30 (2- October), pp. 105-112.