Capital Structure Differences between Public and Private Firms Evidence from China

Master Thesis

Capital Structure Differences between Public and Private Firms Evidence from China

by Tianling Bi

Dual Masters Award in Advanced International Business Management and Marketing (Newcastle University, Business School)

&

Dual Masters Award in Advanced International Business and Management (University of Groningen, Faculty of Economics and Business)

Student number: B1002267 (Newcastle)

S2259818 (Groningen) Supervisor: Dr. Joanna Berry (Newcastle)

Drs. A. Visscher (Groningen)

Table of Contents

Table  of  Contents  ...  2  

List  of  Tables  ...  4  

Acknowledgements  ...  5  

Abstract  ...  6  

Chapter  1  Introduction  ...  7  

Chapter  2  Literature  Review  ...  10  

2.1  Debt-­‐Equity  Choice  ...  10  

2.1.1  Information  Asymmetry  ...  10  

2.1.2  Ownership  Structure  ...  12  

2.2  Firm-­‐Level  Determinants  and  Leverages  in  Public  and  Private  Firms  ...  14  

2.2.1  Firm  Size  and  Leverages  in  Public  and  Private  Firms  ...  15  

2.2.2  Asset  Tangibility  and  Leverages  in  Public  and  Private  Firms  ...  16  

2.2.3  Profitability  and  Leverages  in  Public  and  Private  Firms  ...  17  

2.2.4  Growth  Opportunity  and  Leverages  in  Public  and  Private  Firms  ...  18  

2.2.5  Firm  Age  and  Leverages  in  Public  and  Private  Firms  ...  20  

Chapter  3  Methodology  ...  22  

3.1  Sample  Selection  ...  22  

3.2  Data  Collection  ...  23  

3.3  Variables  and  Measurements  ...  23  

3.3.1  Dependent  Variable  ...  24  

3.3.1.1  Total  Leverage  Ratio  ...  24  

3.3.2  Independent  Variables  ...  25  

3.3.2.1  Firm  Size  ...  25  

3.3.2.2  Asset  Tangibility  ...  25  

3.3.2.3  Profitability  ...  26  

3.3.2.4  Growth  Opportunity  ...  26  

3.3.2.5  Firm  Age  ...  27  

3.4  Analysis  ...  27  

3.4.1  Multiple  Linear  Regression  Model  ...  28  

3.4.2  Cross-­‐sectional  Regression  Model  ...  28  

Chapter  4  Results  ...  30  

4.1  Results  for  Debt-­‐Equity  Choice  of  public  and  private  firms  ...  30  

4.2  Results  for  Firm-­‐Level  Determinants  and  Leverage  ...  32  

4.2.1  Results  for  Firm  Size  and  Leverage  ...  34  

4.2.2  Results  for  Asset  Tangibility  and  Leverage  ...  35  

4.2.3  Results  for  Profitability  and  Leverage  ...  36  

4.2.4  Results  for  Growth  Opportunity  and  Leverage  ...  37  

4.2.5  Results  for  Firm  Age  and  Leverage  ...  38  

Chapter  5  Discussion  ...  39  

5.1  Discussion  of  results  for  Hypothesis  1  ...  39  

5.2  Discussion  of  results  for  Hypothesis  2.1  ...  40  

5.3  Discussion  of  results  for  Hypothesis  2.2  ...  41  

5.4  Discussion  of  results  for  Hypothesis  2.3  ...  42  

5.5  Discussion  of  results  for  Hypothesis  2.4  ...  42  

5.6  Discussion  of  results  for  Hypothesis  2.5  ...  43  

Chapter  6  Conclusions  ...  45  

Chapter  7  Limitations  ...  46  

Reference  ...  47  

List of Tables

Table 1: Hypotheses ... 21  

Table 2: Definitions of Variables ... 27  

Table 3: Descriptive Statistics: mean statistics for all selected variables ... 31  

Table 4: Results for Multiple Linear Regression ... 32  

Table 5: Results for Cross-sectional Regression ... 33  

Table 6: Results for Firm Size and Leverage ... 34  

Table 7: Results for Asset Tangibility and Leverage ... 35  

Table 8: Results for Profitability and Leverage ... 36  

Table 9: Results for Growth Opportunity and Leverage ... 37  

Table 10: Results for Firm Age and Leverage ... 38  

Acknowledgements

This thesis is conducted as the conclusion and the final assignment of my study at the Newcastle University and the University of Groningen. The time spent in both universities will be an unforgettable experience for me. I am sure this study experience will have very important contribution to both my future work and study.

I would like to thank several persons who have been supporting me and been giving advises during this period. First, I am grateful to my supervisors, Drs. A. Visscher and Dr. Joanna Berry for their constant supports. Their professionalism in the field help me to gain valuable knowledge during the research. Second, I would like to thank my family and friends for their continuous supports.

Groningen, December 2012

Abstract

What causes differences in capital structure between Chinese public and private corporations? To examine this question, I compare firm-level determinants of 153 Chinese public firms and 151 Chinese private firms in Orbis database in the year of 2011. I find significant differences in capital structure policies between these two groups. In particular, Chinese private firms have higher leverage ratios than their public counterparts. And the leverage of Chinese private corporations is less negatively related to firm size and tangibility, while more negatively related to profitability and firm age. However, the correlation between growth opportunity and leverage is not found in this study.

Keywords:

Capital structure, Firm-level determinants, Leverage, Public firms, Private firms

Chapter 1 Introduction

Starting with the theorem from Modigliani and Miller

(1958), scholars never stop exploring the field of firm’s capital structure decision, as it relates to a corporation’s capability to meet the demands of its stakeholders (Modigliani and Miller, 1963;

Simerly and Li, 2000).

In recent years, researches on capital structure are mainly focusing on two main directions. One is attempting to explore capital structure decisions for some specific types of firms in some specific countries or regions. For example, Multinational Corporations (Doukas and Pantzalis, 2003; Lee, 1986; Abdulah, 1987; Chkir and Cosset, 2001; Eiteman and Stonehill, 1994; Shapiro, 1992), Domestic corporations (Madura, 1995; Lee and Kwok, 1988; Fatemi, 1988), joint ventures (Li et al., 2011), public corporations

(Bhabra et al., 2008; Chen, 2004; Su, 2010; Liu et al., 2011; Ni and Yu, 2008; Zou and Xiao, 2006) and so on. Among them, I find two gaps in this direction. First, very few studies focus on financing behaviors of private corporations

, which is probably due to difficulties in obtaining data. However, the private company plays a gradually more significant role in driving economic growth, creating productive jobs and raising incomes for people. Brav (2009) points out that in the United Kingdom, private corporations outnumber their public counterparts, representing 97.5% of all incorporated entities. Additionally, he indicates that private companies own over two-thirds of corporate assets in the UK. Likewise, according to the data from the World Bank, in 2011, domestic credit to private sector

(% of GDP)

1

Modigliani and Miller’s (1958) assumption is the cornerstone of capital structure theory, which indicates that in a perfect / frictionless market (e.g. no taxes, risk-free debt) financial leverage is irrelevant, in the other word, corporation’s capital structure does not affect its market value in that condition.

2

In this paper, public firms refer to those companies that are publicly listed in the stock market, so that they have unrestricted rights to offer debentures and issue shares publicly (Brav, 2009).

3

In this paper, private corporations means those firms that are not listed in the stock market.

4

According to the World Bank, domestic credit to private sector refers to financial resources provided to the

private sector, such as through loans, purchases of non-equity securities, and trade credits and other accounts

reaches to 129.4% in the worldwide. Thus, concerning the important position of private corporations in the world economy, it seems meaningful both academically and practically to explore the capital structure and financing policies of private firms.

Second, a large proportion of prior researches have shed light on firms’ capital structure decisions in developed nations (Myers, 1984; Myers and Majluf, 1984;

Akhtar and Oliver, 2009; Goyal et al., 2011; Brav, 2009), while small proportion of scholars focus on corporations’ financing behaviors in developing countries. For instance, Booth et al. (2001) investigate firms’ capital structure features in India, Korea, Jordan, Pakistan, Brazil, Turkey, Zimbabwe, Mexico, Thailand and Malaysia, but not including China. I believe that understanding leverage policies of Chinese corporations is essential, because China has become the world’s second largest economy in terms of nominal GDP

and is the fastest growing major economy in the world (data from Chinability, 2012). Lee and Rui (2010) indicate that there is still lack of knowledge about financing features of Chinese companies. Additionally, China has obvious institutional differences

comparing with Western countries, for example, taxation system, inflation, legal rights protection, contracts enforcement (Fan and So, 2000; Frank and Goyal, 2009; Goyal et al., 2011; Chen, 2004). Thus, it is important and interesting to examine whether corporates’ capital structures have distinct characters in Chinese context.

Another direction is about exploring the determinants of firms’ capital structure, which including internal / firm-related characteristics; such as firm age, firm size, asset tangibility, growth opportunity, profitability and corporate strategy (Titman and Wessels, 1988; Haris and Raviv, 1991; Li and Li, 1996; Singh et al. 2003; Lemmon et

receivable, that establish a claim for repayment.

5

According to the International Monetary Fund (IMF), China’s total nominal GDP reaches approximately US$7.298 trillion in 2012, which is only behind United States.

6

According to Chen (2004), the company law in Chinese firms is ambiguous about investors’ rights. Also,

there is still lacking of clearly defined about property rights, and lacking of effective capital markets for external

financing. Furthermore, compared with developed countries, the enforcement of law is poor in China.

al., 2008; Rajan and Zingales, 1995; Allen, 1991, 1993; Twite, 2001; Frank and Goyal, 2009) and external elements; like political risk, inflation, exchange rate risk, specific national institutions (Ramirez and Kwok, 2009; Doukas and Pantzalis, 2003;

Burgman, 1996; Chkir and Cosset, 2001; Chen et al., 1997; Frank and Goyal, 2009).

However, scholars have not further compared which factors are more important. I believe that the determinants of capital structure may play different important roles in explaining leverage features in different types of companies, which means that some factors may affect leverage policies more / less significantly in some kinds of firms, while others may not.

This paper aims to fulfill gaps that introduced above and enriches studies on capital structure. The main research question is formulated as: What causes differences in capital structure between Chinese public and private corporations? The purpose of this paper is to answer this question by analyzing the different debt-equity choices and different roles of capital structure determinants playing between private and public companies in China during the year of 2011, using the Orbis database

.

The rest of the paper proceeds as follows. Next chapter gives a comprehensive overview of previous studies on capital structure. It contains two sub-sections. Each sub-section develops one or more hypotheses, which are based on theoretical and empirical arguments presented in the literature review. There are 6 hypotheses in total.

In the third chapter, a detailed introduction of the database is provided, including sample selection and data collection. Furthermore, the definitions of dependent and independent variables as well as regression models are introduced in the following.

Chapter 4 details the results of the analysis. Chapter 5 discusses the result of each hypothesis. Chapter 6 summarizes and concludes the main findings. Chapter 7 considers limitations of this research.

7

Orbis database contains comprehensive information on companies around the globe. It covers over 100

million firms worldwide, including about 9 million companies located in Asia-Pacific, over 50 million firms in

Europe and more than 30 million corporations in Americas.

Chapter 2 Literature Review

In this chapter, I am going to theoretically explain some firm-related factors that cause capital structure differences between public and private corporations. In the meantime, a more extensive and systematic review of literature about this study will be presented.

This chapter includes two sub-sections, which are 2.1) Debt-Equity Choice, 2.2) Firm-Level Determinants and Leverages in Public and Private Firms. At the end of the first sub-section, a hypothesis will be developed, and 5 hypotheses will be formulated during the second sub-section. To begin with, I will analyze the different debt-equity policies of public and private companies in the following two aspects:

information asymmetry and ownership structure.

2.1 Debt-Equity Choice

2.1.1  Information  Asymmetry  

Information asymmetry, in corporate finance, refers to that firm insiders

have more or better information than firm outsiders

on the value of firm’s resources or investment opportunities (Klein et al., 2002). Information asymmetry plays an essential role in distinguish debt-equity choice between public and private companies, because it creates the probability that firm’s market value or the potential benefits of firm’s new project do not accord with its real value, hence offering a positive role for firm’s financing decision (Myers and Majluf, 1984; Klein et al., 2002). In short, firm’s leverage increases with the extent of the informational asymmetry (Harris and Raviv, 1991; Myers and Majluf, 1984).

Obviously, the level of information asymmetry is much higher for private firms than

8

Firm insiders refer to corporation’s managers, directors or even employees who have accurate and valuable information that only people in the firm have.

9

According to Jensen and Meckling (1976), firm outsiders refer to investors or minority shareholders who with

no direct role in the management of the corporation.

for public firms. Since public companies are required to disclose some financial information to their outside investors regularly, which usually contain a comprehensive detail of firm’s performance, capital flow and operation process.

However, private companies have no obligation to publicly announce much of detailed information about financing and operating. This directly results in higher equity cost for private firms than public ones. For example, when a private firm gets a profitable project but lack of funding, firm insiders have to persuade their stockholders by using a warranty of high returns or other attractive ways. Otherwise outside investors will probably not fund in a company that they have difficulty in obtaining necessary information, because it hides a lot of uncertainty. Thus, it is obviously that the cost of equity to debt is higher for private corporations than public corporations (Brav, 2009; Goyal et al., 2011).

For outside investors/suppliers

(Faulkender and Petersen, 2006), private companies are less transparent and offering lower disclosure. Thus, if they want to fund private corporations, investors have to collect additional information to understand the previous performance and current situation of these enterprises. During their investing, they have to spend a lot of time and expenditure

on monitoring managers’ activities, firms’ operating and financing behaviors. If monitoring is costly and imperfect, then for private firms that need to be more monitored and public firms that need to be less monitored, most investors

would choose funding public firms (Faulkender and

10

Faulkender and Petersen (2006) indicate that when considering a firm’s leverage, it is important to include not only the requirement of demand side (firms) but also the constraints from supply side (investors). This point of view is accepted by a lot of other scholars, for example, Brav (2009), Leary (2009), Lemmon and Roberts (2010), Goyal et al. (2011).

11

It is necessary for outside investors (minority shareholders) to monitor managers’ behaviors, because interests between corporations’ managers and outside investors will somewhat diverge. Thus the “expenditure”

refers to the price which outside investors pay for shares will include the monitoring costs and the effect of the divergence between managers’ benefits and theirs (Jensen and Meckling, 1976).

12

“Most investors” here refers to those lenders that prefer funding low-risk projects, so that they can enjoy the

stable income from interests of investing. However, some creditors that prefer high-risk high-return investments

are not included in “most investors”.

Petersen, 2006), since it costs less on monitoring, and thus the price of equity would be cheaper in public firms. To conclude, I assume that minority shareholders prefer investing external equity from public firms rather than from private firms, which majorly due to lower price.

Taken above considerations both from owner-managers and outside investors, I expect that equity cost are more expensive for private corporations than their public counterparts (Graham, 1999; Brav, 2009; Goyal et al., 2011), thus it is understandable that private companies prefer debt more than equity when they are raising capital.

2.1.2  Ownership  Structure  

Another fundamental factor that influences the choice of debt and equity to private and public firms is their ownership structure. Corporation's managers who value control will prefer raising capital by debt instead of equity, because issuing equity would dilute their holdings and increase the risk of losing control (Amihud et al., 1990; Stulz, 1988; Brav, 2009). Generally, public companies are usually held by a lot of minority shareholders without any control of the organization (Demsetz and Lehn, 1985), the behavior of managers from listed companies are restricted by their shareholders. Differently, private companies are mainly held by a few shareholders who own a large stake and significant control of the firm (Brav, 2009). As explained before, issuing external equity for finance means giving away control of firms. From this perspective, I believe that private corporates are more likely to use debt over equity compared with public firms when raising finance, because they already have higher control and of course do not want to lose it. In the following, I will further explain this opinion from managers’ and outside investors’ standpoints.

For owner-managers, financing investments by debt rather than by issuing external

equity increases the probability of maintaining control and reaping the advantages

associated with it (Amihud, et al., 1990; Harris and Raviv, 1988; Stulz, 1988). These

advantages include the authority over market accessing, operational decisions,

financial policies and so on. In addition, when conflicts between managers and outside investors (agency problems

) exist, managers in private firms do not have to worry about stockholders’ expectations and interference (Dorward, 2012). On the contrary, managers from public firms who have less control of the company usually have much pressure from their shareholders, they have to focus on recent earnings and increase profits in short time in order to raise the price of their stock and meet the demand of their shareholders (ib.). They also have to supervise and direct management and to make firm decisions on shareholders’ behalf. Considering the above advantages associated with controlling the firm and disadvantages generated from losing control, managers from private corporations are understandably unwilling to issue external equity, since it will make them losing control of the firm (Goyal, et al., 2011; Stulz, 1988; Amihud et al., 1990). However, managers from public firms may not that care about issuing debt or equity, because they did not have much control of the firm already.

For outside investors (minority shareholders), who are normally not involved in decision-making activities, have to rely on firm’s executives or board of directors to protect their interests. Thus, before they choose to invest in a company, they will consider deliberately about which firm is more transparent, which firm is more emphasis on the protection of minority shareholders’ interests. Obviously, public

13

In corporate finance, the agency problem often refers to two different kinds of conflicts. One is a conflict of interest between a firm’s managers and the firm’s stockholders. According to Jensen and Meckling (1976), agency problem generates because firm managers possess control over some resources, however, they may not use these resources to maximize stockholders’ wealth, but to satisfy their own preferences. Agency problem is associated with agency costs, which are the sum of the monitoring expenditure by stockholders (the principal), the bonding expenditures by the organization (the agent) and the residual loss. Another conflict is interests between shareholders and bondholders, which is often called ‘under-investment problem’ or ‘debt overhang problem’

(Myers, 1977). It means that the corporation usually tends to give up low-risk but positive net present value (NPV)

projects and choose to invest high-risk projects, which will increase shareholders’ value at the expense of debt

holders’ benefits. Because high-risk projects have larger profits than low-risk ones, the stockholders can profit

from the increased income, since debt holders only ask for a fixed stream of money (Investopedia, 2012). It is

important for understanding both kinds of agency problems when exploring capital structure differences between

public and private firms.

firms do better on information disclose than private firms. In addition, a large number of listed companies formulate policies and regulations to promote and protect minority shareholders’ interests, which increases confidence of outside investors to finance these firms. Nevertheless, private companies do not tend to provide minority stockholders the same disclosure and protections they would enjoy from listed companies (Brav, 2009). Unless private firms offer higher returns or extra bonus, outside investors are less likely to be attracted. Therefore, outside investors are more willing to purchase external equity from public firms over private firms, as it is cheaper and offers more protection for their benefits.

In sum, differences from information asymmetry and ownership structure are two major reasons that make external equity more costly for private firms than for public firms. Based on the above-mentioned explanation, the following hypothesis is formulated:

Hypothesis 1: Private firms prefer debt more than equity compared with their public counterparts when raising finance.

2.2 Firm-Level Determinants and Leverages in Public and Private Firms In this sub-section, I focus on whether the effects of firm-related factors on leverage differ between public and private companies. I select several factors

(i.e. firm size, asset tangibility, profitability, growth opportunity and firm age), which theoretical and empirical studies

have shown that they are important to firm’s capital structure

14

Actually, firm’s capital structure is related to a lot of firm-level factors. According to Harris and Raviv (1991), firm’s leverage ratio increases with fixed assets, non-debt tax shields, investment opportunities (in this paper, we name it growth opportunities), firm size; and decreases with volatility, advertising costs, the possibility of bankruptcy, profitability and uniqueness of the product. Goyal et al. (2011) also indicate that firm age is an important factor relating to capital structure. However, I only focus on five of above mentioned determinants (firm size, tangible asset, profitability, growth opportunities and firm age). I limit the study to these factors due to two reasons. First, these determinants are consistent with most previous researches (for example, Fernandes (2011), Rajan and Zingales (1995), Harris and Raviv (1991), Brav (2009), Bradley et al. (1984) etc.). Second, the data and time limitations not allow me to develop indicators for the other determinants.

15

The study of Rajan and Zingales (1995) shows that determinants of firm leverage identified in US (firm size,

choice. In the following, I will analyze the above determinants one by one to distinguish capital structure differences between private and public companies. At the end of each determinant explanation, one hypothesis would be developed.

2.2.1  Firm  Size  and  Leverages  in  Public  and  Private  Firms  

In previous studies, the relationship between firm size and firm’s leverage ratio is controversial. According to trade off theory

, large firms should be more highly leveraged because large firms tend to be more diversifies, to have more stable cash flows, more fixed assets for collateral, and thus have a smaller probability of bankruptcy (Warner, 1977; Ang, et al., 1982; Titman and Wessels, 1988; Rajan and Zingales, 1995; Wald, 1999; Booth et al., 2001). However, another important theory namely pecking order theory

holds the different view. It argues that small firms may be more leveraged due to equity is more expensive for small firms than large firms (Smith, 1977; Titman and Wessels, 1988; Kestor, 1986). As a consequence, small corporations prefer raising capital by debt rather than by equity, this results in higher leverage ratios for small sized companies than large ones. Although the relationship between firm size and capital structure is ambiguous in theoretical, a lot of empirical studies (Booth et al., 2001; Frank and Goyal, 2009) proved that they are

asset tangibility, profitability and growth opportunity) are similarly significant in other G-7 countries (Japan, Germany, France, Italy, the United Kingdom and Canada). In addition, Chen (2004) indicates that these factors that are relevant for explaining firm financing decision in developed countries are also relevant in China. Although many researches not treat firm age as a factor that influences capital structure, still a number of scholars (Goyal, et al., 2011; Mouamer, 2011; Serrasqueiro and Nunes, 2012) provide evidence to show that it is related to firm’s leverage choice.

16

Trade off theory is one of the most important theories to explain the capital structure choice. The idea of trade off theory, according to Mayers (1984), is that a company’s optimal leverage ratio can be determined by a tradeoff of the costs (i.e. bankruptcy, financial distress) and the benefits (i.e. the interest tax shields) associated with borrowing.

17

Pecking order theory, which was developed by Myers and Majluf (1984) and Myers (1984), is another very

important theory in describing firm’s leverage choice. This theory argues that information asymmetry between

firm insiders and outside investors makes external financing more expensive than internal funding. Thus, if

corporations want to raise capital, they may first rely on internal financing, then external debt, and external equity

may become the final choice.

positively related. This paper predicts a positive relationship between firm size and firm’s leverage ratio.

Firm size has different impacts on capital structure of public and private firms. Goyal et al. (2011) indicate that, compared with public corporations, firm size has a relatively smaller positive effect on leverage ratios of private firms. There are two main reasons for this prediction. First, the size of private firms is typically smaller than public ones, thus they may have weaker ability for borrowing, and they may less attractive for investors as well. Second, from the point of view of the large public companies, they generally have more diluted ownership, interests between managers and investors are more likely to go diverge. It means that public firms need more debt for monitoring in order to better control management behaviors, while private firms may not need to. Therefore, I expect that when firm size becomes larger, leverage ratios in public companies would increase more significantly than private firms.

Based on the above-mentioned explanation, the following hypothesis is formulated:

Hypothesis 2.1: Firm size has smaller positive effect on leverage of private firms than their public counterparts.

2.2.2  Asset  Tangibility  and  Leverages  in  Public  and  Private  Firms  

A positive relationship between tangible assets and leverage ratio is broadly accepted by lots of studies (for example: Faulkender and Petersen, 2006; Lemmon et al., 2008;

Giannetti, 2003; Rajan and Zingales, 1995; Titman and Wessels, 1988). These studies explain that tangible assets can be used as collateral, which diminish lender’s risk and firms’ probability of bankruptcy (Rajan and Zingales, 1995; Williamson, 1988, 1975;

Chen, 2003). Thus, the greater the proportion of tangible assets to total assets, the more willing should creditors be to fund loans, the higher leverage ratio should be reached (Rajan and Zingales, 1995).

Similar with firm size, asset tangibility has relatively smaller positive effect on the

leverage of private firms than public firms (Goyal et al., 2011). Because, as I assumed

in the first sub-section of literature review, the cost of equity is cheaper for public firms than private firms, thus public companies have higher equity ratio to debt than private companies. However, according to pecking order theory, firms prefer debt over equity when raising capital. Consequently, if a public firm increases its ratio of tangible assets, managers tend to use these additional assets to collateralize for debt, and then the ratio of equity would decrease immediately. Differently, private firms tend to access external equity market less often due to higher absolute costs, which means they have limited equity already. So when fixed assets rising, their increase of leverage ratios will not as significant as public firms do. Based on the above-mentioned explanation, the following hypothesis is formulated:

Hypothesis 2.2: Asset tangibility has smaller positive effect on leverage of private firms than their public counterparts.

2.2.3  Profitability  and  Leverages  in  Public  and  Private  Firms  

The relationship between profitability and financial policy of a company is

theoretically controversial (Rajan and Zingales, 1995). According to trade off theory,

higher profitability or more free cash flow should be more leveraged in order to offset

corporate taxes by using the tax shield of debt (Fernandes, 2011; Ross, 1977). On the

supply side, outside investors are more like to lend money to firms with much free

cash flows (Rajan and Zingales, 1995). However, pecking order theory holds the

opposite view, which suggests that higher profitability would result in lower leverage

ratio (Frank and Goyal, 2003, 2008, 2009). The reason is due to information

asymmetry, firms prefer raising finance first from internal capital / retained earnings /

profitability, and then from external source (Fama and French, 2002; Myers and

Majluf, 1984; Donaldson, 1961). Moreover, Fischer et al. (1989) support this negative

relationship by analyzing the effect of adjusting leverage. When a corporate gets

earnings, debt gets paid off and debt ratio decreases automatically (Frank and Goyal,

2003). This paper predicts a negative relationship between firms’ profitability and

leverage ratios. Because most empirical evidences support this prediction even if the

tradeoff assumption is at work (Chkir and Cosset, 2001; Friend and Lang, 1988;

Baskin, 1989; Kestor, 1986; Fernandes, 2011; Frank and Goyal, 2003).

The factor profitability has different impacts on capital structure choice of public and private firms. Based on studies of Brav (2009) and Goyal et al. (2011), the leverages of private companies are more sensitive to firms’ profitability / operating performance.

There are two reasons for this prediction. First, as explained previously, private firms cost much more on external capital than public firms, thus when they get earnings they may pay off debt immediately in order to decrease their costs of capital raising.

This directly results in the decline of debt ratios. In terms of public firms, due to relative cheap cost of external finance, public corporates probably would not stockpile cash in response to the enhancement in performance. On the contrary, they are more likely to use retained earnings to increase their investments (Brav, 2009). Second, when a firm has much free cash flows, the manager has tendency to appropriate perquisites for his own consumption (Jensen and Meckling, 1976). Outlays associated with agency problem are more costly in public firms than private firms. Using retained profits for investing is one of effective ways to diminish agency costs.

Considering the above two reasons, I assume that leverages of public firms are less negatively related to their profitability (Brav, 2009; Goyal et al., 2011). Therefore, the following hypothesis is formulated:

Hypothesis 2.3: Profitability has larger negative effect on leverage of private firms than their public counterparts.

2.2.4  Growth  Opportunity  and  Leverages  in  Public  and  Private  Firms  

A negative relationship between growth opportunity and leverage ratio is accepted by

majority of studies (Rajan and Zingales, 1995; Frank and Goyal, 2009; Titman and

Wessels, 1988; Chen et al., 1997; Jensen and Meckling, 1976). Two possible reasons

are found to explain this inverse relation. First, firms with high growth opportunity

are understandably having more flexibility in their choice of further investments

(Titman and Wessels, 1988). Thus they tend to give up profitable investment chances

when they are highly leveraged (Jensen and Meckling, 1976). Second, future growth opportunities can be regarded as a form of intangible assets. Compared with fixed resources, growth opportunities tend to borrow less because they cannot be collateralized (Chen, 2004) and do not generate current taxable benefits (Titman and Wessels, 1988). Thus, as trade off theory predicts, growth opportunities increase costs of financial distress and raise the probability of bankruptcy, finally it will reduce firm’s leverage ratios. Most of empirical studies support this negative relationship (for example, Kim and Sorensen, 1986; Wald, 1999; Smith and Watts, 1992; Booth et al., 2001).

Similar with other factors, growth opportunity plays an important role in explaining capital structure differences between private firms and public firms. According to Brav (2009) and Goyal et al. (2011), compared with public companies, leverages of private corporates are expected to be less sensitive for further growth opportunities.

One possible reason can be used to answer this prediction. As we know that equity-controlled firms (generally are public firms) tend to invest suboptimally in order to expropriate benefits from the corporate’s bondholders (Titman and Wessels, 1988), which increases conflicts between shareholders and bondholders. As agency problem arising, bondholders may reduce their invests in this firm while shareholders may purchase more in order to obtain more interests, which obviously lowers public firm’s leverage ratio and higher its equity ratio at the same time (Berens and Cuny, 1995). However, for private firms, the more costly is financing by external equity, the less often are they accessing the external equity market, the less serious conflicts generates between shareholders and bondholders, thus the smaller negative effect of growth opportunity is to leverage ratio. Based on the above-mentioned explanation, the following hypothesis is formulated:

Hypothesis 2.4: Growth opportunity has smaller negative effect on leverage of private

firms than their public counterparts.

2.2.5  Firm  Age  and  Leverages  in  Public  and  Private  Firms  

Firm age as a relevant factor of firm’s capital structure does not receive enough attention by many researchers. As I see, firm age can be treated as a proxy for information asymmetry. As firms become older, they are probably becoming more transparent and providing more disclosure to minority shareholders. Older firms are well known in the market and their access to investment improves with age (Goyal et al., 2011). Similarly, Serrasqueiro and Nunes (2012) argue that age is effective in reducing information asymmetry, decreasing agency costs, accumulating wealth and increasing firm’s reputation (Diamond, 1989). Consequently, as firm age increasing, firm’s financing by equity will arise and financing by debt will decrease.

Comparing with private and public firms, firm age has a significant negative effect on the leverage of private corporates’ (Goyal et al., 2011). The older private firms are, the smaller debt ratios will be. There are two potential reasons for this fact. First, greater age of firms can contribute to a reduction of problems associated with information asymmetry, thus decreasing the costs of issuing equity (Serrasqueiro and Nunes, 2012). This reduction in information asymmetry is particularly significant for private corporations (Goyal et al., 2011). Second, as private corporations get older, they increase their ability to retain earnings. According to pecking order theory, firms prefer internal funding more than external financing, thus older private companies would rely less on debt. Based on the above-mentioned explanation, the following hypothesis is formulated:

Hypothesis 2.5: Firm age has larger negative effect on leverage of private firms than their public counterparts

Based on the literature presented above, factors that cause capital structure differences

between public and private firms are expected. Table 1 summarizes all the hypotheses

in this paper.

Table 1: Hypotheses

Hypothesis 1: Private firms prefer debt more than equity compared with their public counterparts when raising finance.

Hypothesis 2.1: Firm size has smaller positive effect on leverage of private firms than their public counterparts.

Hypothesis 2.2: Asset tangibility has smaller positive effect on leverage of private firms than their public counterparts.

Hypothesis 2.3: Profitability has larger negative effect on leverage of private firms than their public counterparts.

Hypothesis 2.4: Growth opportunity has smaller negative effect on leverage of private firms than their public counterparts.

Hypothesis 2.5: Firm age has larger negative effect on leverage of private firms than their

public counterparts.

Chapter 3 Methodology

After formulating hypotheses that guide this paper, this chapter is attempting to present how samples will be selected, how data will be collected, measurements and analysis I am going to use for testing the hypotheses.

3.1 Sample Selection

In order to select samples appropriately, it is necessary to distinguish the definition of two types of firms - public companies and private companies. According to Brav (2009), the capability of raising external capital is the most significant distinction between public and private corporations. More specifically, public firms have unlimited rights to offer debentures and issue shares publicly, while private firms are restricted to enjoy these funds. It means that only public corporations are eligible to be listed on a stock market, thus public firms can be defined as those corporations that are listed, while private firms can be defined as those corporations that are not listed (Brav, 2009). Based on the definition, I searched samples by their listing statuses. I chose publicly listed companies (not include formerly publicly listed companies) for public firms, and unlisted companies for private firms. Furthermore, in order to reduce alternative influences, I use three control criteria in selecting the sample. First, controlling for firm type. Regulated firms (SIC cods 4000-5000) and financial firms such as banks and insurance companies (SIC codes 6000-6999) are excluded, because these firms are fundamentally different in the nature, and there may exist a systematic relationship between regulations and firms’ leverage ratios (Chkir and Cosset, 2001;

Lee and Kwok, 1988; Brav, 2009; Goyal et al., 2011; Burgman, 1996). Second, controlling for firm size

. Very large companies (VL), large companies (L) and Medium sized companies (M) are included. Since Small companies (S) are considered to be small and do not have a complete record of the variables needed in this study,

18

According to ORBIS database, a firm’s size category is determined by annual turnover, total assets or total

number of employees for the last available year.

thus I have to exclude them due to difficulty in obtaining data. Finally, controlling for location. As the objective of this paper is to find out distinct characters of Chinese public and private firms’ capital structure, hence I only focus on firms in China.

3.2 Data Collection

The source of firms’ financial data for both public and private corporations collects from Orbis electronic databases. It includes keys financials, balance sheet, profit and loss account for both listed and unlisted companies. 304 firms’ financial information is collected in this study, since Thomas (2004) indicates a sample scope that fewer than 200 is expected as not sufficient scope for analysis of survey. Among all sample companies, there are 153 Chinese public firms and 151 Chinese private firms. I only collect sample firms’ financial data in the year of 2011, it is due to three reasons. First, some firms are established in recent years, they lack of data a few years ago, so I have to collect data in 2011 due to data availability reason

. Second, firms’ financial information of 2011 is the latest data I can obtain so far, and it can greatly represent the current situation of the development of sample enterprises. Third, Rajan and Zingales (1995) use four-year average data (on the period of 1987 to 1991) for their variables. However, Bevan and Danbolt (2000) replicate the analysis of Rajan and Zingales (1995) but use only one-year (in the year of 1991) financial data, the results were proved have no significant change. Therefore, I believe that using one-year data is not a serious limitation.

3.3 Variables and Measurements

In this sub-section, I will introduce one dependent variable (total leverage ratio) and five independent variables, including firm size, asset tangibility, profitability, growth opportunity and firm age. During the presenting of these variables, I will introduce how to measure them in detail, because a lot of variables are usually calculated

19

For example, a sample firm is born in the late of 2010. Therefore, for this firm, only financial data of 2011 is

available in the database. It is impossible to collect data before that year.

differently across previous finance researches. At the end of this sub-section, definitions of all variables though this study will be summarized in Table 2.

3.3.1  Dependent  Variable  

I employ one dependent variable in this study. It is total leverage ratio. In the following, I will present the reason I choose it and how to measure it.

3.3.1.1 Total Leverage Ratio

Total leverage ratio is the dependent variable in this research, since it will change as other firm-level factors change. It measures as the total liabilities (long-term liabilities plus short-term liabilities) to total assets ratio

. The leverage measure is based on book values rather than market values. Since the objective of this paper is comparing financial preferences between Chinese public and private firms, while market values for Chinese private companies are unobservable (Brav, 2009). So I have to employ book value to measure private corporations. In addition, although market values of Chinese public firms are available from database, book values are also applied to measure Chinese public companies in order to facilitate comparisons across two groups (Brav, 2009). According to Goyal et al. (2011), Fama and French (2002), Leary and Roberts (2005), only using book values for leverage measure is not a serious limitation, because a lot of previous researches applying both book values and market values, but they have shown that these two measurements behave similarly in the analysis (Goyal et al. 2011; Rajan and Zingales, 1995; Marsh, 1982;

20

According to Brav (2009), Bevan and Danbolt (2002), Rajan and Zingales (1995), Goyal et al. (2011), empirical studies suggest a lot of ratios to measure firm’s leverage ratio. For example, total liabilities to total assets (measures as long-term liabilities plus short-term liabilities and then divide firm’s total assets); total debt to total assets (total assets here is calculated as adjusting assets subtract the book value of equity and add back the market value); total debt to capital ratio (the capital here is calculated as total debt plus book equity which including preference shares); adjusted debt to adjusted capital (measures as adjusted debt to adjusted debt plus book value of adjusted equity). In this study, I employ total liabilities to total assets as firm’s leverage ratio due to two reasons.

First, this measurement is proved to be suitable and accurate to represent firms’ leverage ratio (Goyal et al., 2011;

Rajan and Zingales, 1995; Bevan and Danbolt, 2002). Second, being limited to time for collecting data, I have to

apply this method to measure leverage ratio, since it is easy both in data obtaining and calculating.

Myers, 1977; Taggart, 1977; Deangelo and Roll, 2011).

3.3.2  Independent  Variables  

In this paper, independent variables include firm size, asset tangibility, profitability, growth opportunity and firm age. In the following, I will introduce them one by one.

3.3.2.1 Firm Size

Previous researches employ diverse measurements for firm size. For example, Rajan and Zingales (1995) calculate firm size as the natural logarithm of the net sales.

Titman and Wessels (1988) use the natural logarithm of sales for measuring small sized firms and use quit rates for measuring large sized firms. Fernandes (2011), Brav (2009), Goyal et al. (2011), Chkir and Cosset (2001) measure firm size as the natural logarithm of the total assets. It seems there is no uniform opinion on the measurement of firm size. In this paper, I follow Fernandes’ (2011) measurement, because firm’s total assets are easier than net sales to obtain in Orbis database.

3.3.2.2 Asset Tangibility

A number of studies define tangible assets in different ways. For instance, Rajan and Zingales (1995), Fernandes (2011) and Giannetti (2003) define tangible assets as the ratio of property, equipment and plant to total assets. Brav (2009) employ tangible assets plus investments to total assets to measure tangible assets. In addition, Jong et al. (2008) define tangibility as net fixed asset over book value of firm’s total assets.

Furthermore, Goyal et al. (2011) point out that the ratio of fixed assets and inventory to total assets is more suitable for measuring tangible assets. In this study, I define asset tangibility as fixed assets to total assets, where fixed asset includes tangible fixed assets (e.g. plant, equipment, property etc.) and intangible fixed assets (e.g.

patents, copyrights, trademarks, franchises etc.). Since inventory is a kind of liquid

asset for firms, thus I exclude inventory for calculating tangible assets.

3.3.2.3 Profitability

There are two main measurements

of profitability that are widely accepted in capital structure researches. One measures profitability as the ratio of earnings before interest and taxes to total assets (EBIT), it is supported by Brav (2009)

, Chkir and Cosset (2001). Another way is applying the ratio of earnings before interest, taxes, depreciation and amortization to total book value of assets (EBITDA) (Goyal et al., 2011) as a proxy for profitability. Due to data availability, I employ the first measurement – the ratio of earnings before interest and taxes to total assets – to define profitability.

3.3.2.4 Growth Opportunity

Several indicators of growth opportunity are suggested in previous studies. For example, a) the increasing or decreasing ratio of total assets (measures as the percentage change in the book value of total assets); b) the change of research and development to sales; c) the increasing or decreasing ratio of earnings (Titman and Wessels, 1988); d) the changing ratio of turnover (Brav, 2009); e) the ratio of market value of total assets to the book value of total assets (Rajan and Zingales, 1995; Myers, 1977) and so on. In this paper, I measure growth opportunity as the natural logarithm of the sales in the year of 2011 minus the natural logarithm of sales in the year of 2010. I use this measurement for two reasons. First, the data of sales is easy to obtain in Orbis database. Second, Goyal et al. (2011) suggest that the increasing or decreasing of sales can be served as a suitable indicator for the growth opportunity attribute.

21

There also exist other measurements for profitability. For example, Titman and Wessels (1988) use the ratios of operating income over sales and operating income over total assets as indicators of profitability. Fernandes (2011) apply return on assets (ROA) as the measurement of profitability. Rajan and Zingales (1995) measure profitability as cash flow operations normalized by the book value of total assets. However, the ratio of EBIT and EBITDA are two definitions that are widely used in previous studies.

22

Note that Brav (2009) only include those observations which earning before interest and taxed (EBIT) is

positive. However, this paper includes all the observations with positive and negative EBIT.

3.3.2.5 Firm Age

There is no controversial on the measurement of firm age. I define firm age (in years) as the age of firm relative to its year of incorporation (Goyal et al., 2011).

In order to clearly present the measurements of all variables, Table 2 provides definitions of dependent and independent variables used throughout the paper.

Table 2: Definitions of Variables Dependent Variable Definitions

Total leverage ratio Total liabilities (long-term liabilities plus short-term liabilities) to total assets

Independent Variables Definitions

Firm size Natural logarithm of the total assets in USD

Asset Tangibility Fixed assets (tangible fixed assets plus intangible fixed assets) to total assets

Profitability Earnings before interest and taxes to total assets (EBIT)

Growth opportunity Natural logarithm of sales in 2011 minus natural logarithm of sales in 2010

Firm age 2012 minus the year of incorporation (in years)

3.4 Analysis

To measure whether the effects of firm related factors on leverage differ between

public and private companies, two different kinds of regression models will be

applied. One is multiple linear regression model, it is used to test the effects of

selected determinants on capital structure in public firms and private firms. Another

model is cross-sectional regression, which analyses whether the firm related factors

differ across different firm groups. In the following, I will formulate the two

regression models and then explain them in detail.

3.4.1  Multiple  Linear  Regression  Model  

A linear regression model is appropriate

to explore the relationship between five selected determinants and capital structure in public and private firms. The multiple linear regression model I estimate is:

Leverage

= β

+ β

size

+ β

tangibility

+ β

profitability

+ β

growth

+ β

age

+ε

where i stands for a certain firm among samples, j stands for firm type (public firms, private firms). Leverage represents dependent variables including total leverage ratio, long-term debt ratio and short-term debt ratio. 𝛽

is a constant and 𝜀 is the error term. All the variables are defined in Table 2. This equation is used to compare the estimated coefficients of firm characteristics between public and private firms.

3.4.2  Cross-­‐sectional  Regression  Model  

A cross-sectional regression model is used to test the difference between public and private firms on the coefficients of each firm related factor. This method is widely accepted by a lot of studies (for example, Doukas and Pantzalis, 2003; Goyal et al., 2011; Hovakimian et al., 2001; Fama and French, 2002). The cross-sectional regression model I estimate is:

leverage

= α

+α

public × size

+α

public × tan gibility

+α

public × profitability

+α

public × growth

+α

public × age

+α

private × size

+α

private × tan gibility

+α

private × profitability

+α

private × growth

+α

private × age

+ε

where leverage

represents the leverage of firm i in firm type j (firm type refers to public firm or private firm). In this paper, the leverage measure is total liabilities to total book value of assets. The dummy variables public and private are indicator variables, which are used to define the public status of the sample company. Indicator variable public that takes the value of one, if the firm is a public firm, and zero if it is a private corporation. On the contrary, the variable private takes a value of one, if the

23

A number of researches employ linear regression to test the relationship between determinants and leverage.

For example, Fernandes (2011), Rajan and Zingales (1995), Chkir and Cosset (2001).

firm is a private firm, and zero otherwise. Size, tangibility, profitability, growth and

age are independent variables that have been indicated to be essential determinants of

capital structure in section 2. 𝛼

is a constant and 𝜀 is the error term. This equation

is applied to test whether the estimated coefficients of firm determinants on leverage

differ between public and private firms. In the next section, the results of both

multiple linear regression and cross-sectional regression will be presented.

Chapter 4 Results

Results that are analyzed by SPSS statistic software will be introduced in this chapter.

I apply mean statistics of leverage ratios to test the Hypothesis 1 (results are reported in Table 3), and I use two kinds of regressions (multiple linear regression and cross-sectional regression) to test Hypothesis 2.1 to 2.5 (results are reported in Table 4 and Table 5). Each sub-section will explain the tests and results for each hypothesis.

4.1 Results for Debt-Equity Choice of public and private firms

Table 3 reports mean statistics for all selected variables of sample public and private firms in China. Through the table, we can see distinct differences between capital structure of public and private firms. First, the average size of pubic firms is approximately 9 times larger than private firms (9.705 vs. 1.084). Second, public firms have a little more ratio of fixed assets compared with private firms (35.4% vs.

34.5%), but the average ratio of profitability for public companies is lower than their private counterparts (7.5% vs. 8.5%). A strikingly large difference can be seen in the comparison of sales growth, where public firms growing much slower than private companies (9.5% vs. 16.4%). Also, the average age of public firms is a little older than private firms (14.53 vs. 13.8). These two types of firms also differ in terms of their liability structure. The short-term debt ratio is 43.6% for public companies, while 51.7% for private companies. In addition, the long-term debt ratio is about 11.7%

for public firms, while the same ratio for private companies is almost 40% higher at 16.4%. Furthermore, equity ratio comprises about 40% for the total financing in public firms, but this ratio is a little lower for private firms (36.6%). In terms of total leverage ratio, as predicted, this figure is lower for public firms than private firms (60%

vs. 63.4%).

To conclude, since private firms have higher total leverage ratio and lower equity

ratio than public firms, hypothesis 1: private firms prefer debt more than equity

compared with their public counterparts when raising financing is accepted.

Table 3: Descriptive Statistics: mean statistics for all selected variables

Variables Chinese Public Firms Chinese Private Firms

Total assets ($ dollars) 10,264,613,868 1,911,336,969

Liabilities Chinese Public Firms Chinese Private Firms

Long-term debt ($ dollars) 2,004,164,513 256,010,325

Long-term debt ratio 0.164 0.117

Short-term debt ($ dollars) 4,044,040,296 1,026,324,852

Short-term debt ratio 0.436 0.517

Total liabilities ($ dollars) 6,048,204,809 1,282,335,177

Total leverage ratio 0.600 0.634

Equity Chinese Public Firms Chinese Private Firms

Total equity ($ dollars) 4,216,409,053 630,009,443

Equity ratio 0.400 0.366

Firm-level determinants Chinese Public Firms Chinese Private Firms

Firm size 9.705 1.084

Tangible assets ($ dollars) 5,141,209,164 163,128,838

Tangibility ratio 0.354 0.345

Profitability ($ dollars) 695,060,947 153,801,786

Profitability ratio 0.075 0.085

Sales growth 0.095 0.164

Firm age 14.53 13.8

Observations 153 151

Notes: Table 3 reports the mean leverage ratios, leverage factors for selected variables of

public and private firms in 2011. Data sample contains 153 publicly listed companies and 151

unlisted companies in China, which collects from Orbis database. All variables are defined in

Table 2, Section 3.

4.2 Results for Firm-Level Determinants and Leverage

Table 4 shows results from multiple linear regressions, Table 5 reports results from cross-sectional regression. Both of them are applied to test Hypothesis 2.1 to 2.5. In the following, I will present results for hypothesis 2.1 to 2.5 one by one.

Table 4: Results for Multiple Linear Regression

Independent

variables Firm types

Dependent variables Total leverage ratio coefficient t-statistics

Firm size Public -0.238 -2.889***

Private -0.140 -0.843**

Tangibility Public -0.131 -1.542*

Private -0.616 -9.446***

Profitability Public -0.448 -4.943***

Private -0.520 -5.819***

Growth Public 0.110 1.238

Private 0.041 0.594

Firm age Public -0.161 -2.159**

Private -0.251 -0.930***

Constant Public -0.688

Private 7.562***

R

Public 0.226

Private 0.400

R

-adjusted Public 0.199

Private 0.379

Observations Public 153

Private 151

Notes: Table 4 reports multiple linear regression results of leverage on firm-level determinants for 304 Chinese firms, using annual average data of 2011. More specifically, data sample contains 153 publicly listed companies and 151 unlisted companies in China, which collects from Orbis database. The multiple linear regression model has been introduced in Section 3.4.1. The dependent variable is total leverage ratio and the independent variables are firm size, asset tangibility, profitability, growth opportunity (sales growth) and firm age.

Book values are applied to calculate the variables. All variables are defined in Table 2, Section 3.3.2.5. Standardized coefficients beta and t-value are presented in this table.

***Indicates significance at the 1% level.

**Indicates significance at the 5% level.

*Indicates significance at the 10% level.

Table 5: Results for Cross-sectional Regression

Variables All Firms

coefficient t-statistics

Public×Firm size -0.565 -1.682***

Private×Firm size -0.080 -0.174***

Public×Tangibility -0.045 -0.492**

Private×Tangibility -0.354 -4.193***

Public×Profitability -0.390 -5.367***

Private×Profitability -0.426 -5.383***

Public×Growth 0.106 1.316

Private×Grwoth 0.059 0.943

Public×Firm age -0.017 -0.082**

Private×Firm age -0.110 -1.161***

Constant 8.821***

R

0.223

R

-adjusted 0.197

Observations 304

Notes: Table 5 reports cross-sectional regression results of leverage. All 304 firms are analyzed in this regression model. Data collects from Orbis database. The cross-sectional regression model has been introduced in Section 3.4.2. It is used to test whether the effects of firm related determinants on leverage differ between public and private corporations. The dependent variable is total leverage ratio, which measures as total liabilities to total book value of assets. Indicator variable public that takes the value of one, if the firm is a public firm, and zero if it is a private corporation. On the contrary, the variable private takes a value of one, if the firm is a private firm, and zero otherwise. All variables are defined in Table 2, Section 3.3.2.5. Standardized coefficients beta and t-value are presented in this table.

***Indicates significance at the 1% level.

**Indicates significance at the 5% level.

*Indicates significance at the 10% level.