Capital Structure of Real Estate Firms in Chinese Stock Market

This paper investigates the determinants of capital structure of real estate firms in China. An empirical study on determinants of capital structure of real estate in Chinese listed firms is conducted using a relatively regression of accounting data for 44 A-share financial listed companies over the quarter from 2008 to the third quarter of 2011. First, this paper identifies that the pecking order theory in China is different from western countries. Second, the results show that profitability, non-debt tax shields and liquidity are significant influence factors in financial sector, while others like size, growth, tangibility and non-circulating share should be judged by the size of the company. Moreover, firm size and non-circulating shares are almost positively related to the corporate leverage ratio. It is also found that Chinese institutional characteristic affects the capital choice decision and the largely state ownerships do affect capital structure choices.


INTRODUCTION
Real estate industry is the pivotal industry in national economy. The rapid gro wth of the real estate markets (both in quantity and property appreciation rates) in Chinese cities prov ides us an opportun ity to examine the influence of cap ital structure on the real estate industry (Brandt and Rawski, 2008). Over the past 10 years, Chinese real estate markets have joined the global universe of investable assets as exemp lified by vibrant skylines and the construction of startling commercial properties in Ch inese megacities. It is now difficult to ignore a market ranked as the third largest globally by invested stock (DTZ, 2012). Th is becomes evident after the 2008 financial crisis and the large amount of liquid ity injected by the government went into real estate.
Capital Structure not only affects the cost of capital and market value of the firm, but also the corporate governance and the integral operative performance. Capitalizat ion ratio delivers the key in sights to evaluate a company's capital position.
Therefore, analysts use different ratios to assess the financial strength of a company's capital structure. Firms raise investment capital through several sources such as initial public offerings (IPO), seasoned offerings of common stock and preferred stock which co mes under equity or alternatively they raised debt. The choice of funding sources is dependent on firms' capital structure, relative cost of funds, and possibly market value of firm's securitie s. Several theories have attempted to explain firms' choice of securities, but there was no consensus on this issue.
Lately, more and more China's real-estate enterprises raised funds from the capital market. They issue stocks and turned themselves into listed public co mpanies. But they remain controlled by some small group of investors mainly state owned enterprises or families. Therefore it is necessary and significant to conduct a study of the problem of optimizat ion of the capital structure of the listed real estate companies. Despite of western theories and models that influence firm's capital structure, what determine the capital structure of real estate firms in Ch ina had been a question to academic and practitioners alike. Therefore, this paper study the capital structure of real estate companies listed on the Chinese stock market to determine what affect capital structure decisions of real estate firms. In China, lack of capital is a common problem in the real estate industry; therefore most of its capital comes fro m bank loan, earnest money, advance receipts and outer investment. There is little capital fro m its own. This results in over dependence on loan and advance receipts and deep assets liabilities ratio.
This paper will main ly focus on the capital structure of real estate in Chinese listed companies to investigate the impacts of different measures or determinants on the choices of capital structure and attempt to point out the optimal cap ital structure for real estate firms in Chinese listed companies. Subsequently, the econometrics models such as static panel data model and dynamic panel data model will be adopted to analyze the data.

Modern Theories of Capital Structure
Modern finance theories on capital structure are dominated by these three famous western theories; MM, trading off theory (TOM) and pecking order theory (POT).
Modigliani and Miller (1958) points out one of the most classical modern theories which ind icates that, in a perfect capital market, the value of enterprise is irrelevant with the firm's capital structure. This theory assumes that there are no bankruptcy costs, individuals can borrow and lend at the risk-free rate and there are only two types of finance which is risk-free debt and risky equity. In the hypothes is of MM theory, all firms are in the same level of risk, no growth, symmetry information and no agency costs.
Trade off theory was put fo rward by Scott (1977) demonstrating that as the increase of the ratio of bonds, the risk of the company will also increase, thus raising the bankruptcies, then decreasing the value of the company. Therefore, the optimu m capital structure of company should be the equilibriu m point caused by financial crisis cost and bankruptcy cost between section tax benefit and the rise of ratio of debt capital. In 1984, Myers and Majluf carry out pecking order theory considering the impact of information sources between insiders and outside investors on the corporation investment and financing behaviors. Due to adverse selection, they illustrate the retained earn ings as the internal funds dominated the first place in the corporate financing p reference, followed by debt financing and equity financing. When outside funds are necessary, firms prefer debt to equity because of lo wer in formation costs associated with debt issues. As an alternative capital structure theory, pecking order theory takes account of the asymmetric information and the existence of transaction costs.
Fro m the asymmetric informat ion perspective, it could be imagined that manager, as the insider, could access more relevant and reliable information about firms than the less informed outside investors. With the information advantage, managers could have more opportunity to issue bonds thus avoiding adverse signaling informat ion of issuing equity. This theory is usually regarded as a competitor to the Trade-off theory.
Baker and Wurgler (2002) also suggest a new theory of cap ital structure: the "market timing theory of capital structure".
This theory argues that firms time their equity issues in the sense that they issue new stock when the stock price is perceived to be overvalued, and buy back own shares when there is undervaluation. Consequently, fluctuations in stock prices affect firms' capital structures. They find that leverage changes are strongly and positively related to their market timing measure, so they conclude that the capital structure of a firm is the cumulat ive outcome of past attempts to time the equity market.
Fama and French (2002) have attempted to compare TOM and POT, indicat ing that although the rate of the adjustment is quite slow, regressions of firms' debt ratios show reliable ev idence that leverage is mean -revert ing. Hackbarth et al.
(2007) argues that the TOM is not only able to measure a certain debt level, but also could in fer the optimal debt structure in terms of bank and market debt to borrow or issue.

Capital Structure Research in China
Wang (2003)  shows that institutional differences and financial constraints in the capital markets have also significantly con stituted to the influencing factors for companies' capital structure.

RESEARCH METHODOLOGY
The research method of this empirical study involves the econometrics model with a series of Descriptive Statistic Analysis and quantitative methods from the Ch inese listed companies' financial and accounting database. Due to the data limitat ion of state-owned share percentage and fixed assets ratio, the analysis is conducted by using a panel data linear regression model where the dependent variable is leverage rat io during the time period fro m the first quarter of 2008 to the third quarter of 2011. In this paper, static and dynamic panel data models will be used, and the results from these two models will be run in the STATA program in the following empirical analysis.

Static Panel Data Model
The init ial basic regression model for analyzing panel data is the ordinary least squares regression (OLS). The formulation of this model in mathematics is: In this model, it y represents the capital structure of real estate firm i in year t;  is the constant term; ' it x is a vector of seven independent variables which is the explanatory indicators;  is coefficient vector o f parameters wh ich is independent variables that is constant over quarters; it  is the unobserved error term or d isturbance term. In Table 1, The practical and theoretical reasons why we use the OLS estimators is due to the do minant method used in practice.
OLS has become the common language for regression analysis throughout economics, finan ce and social sciences more generally. The using OLS p resents results indicating that you are 'speaking the same lan guage' as other economists and statisticians. The OLS formu las are built into virtually all spreadsheet and statistical software packages, making OLS easy to use. The desirable theoretical properties are also the advantages of OLS estimators. The OLS estimator is unbiased and consistent which is also efficient among a certain class of unbiased estimators; however, additional special conditions need to be discussed and more result analysis should be deferred (Stock and Watson, 2007).

Dynamic Panel Data Model
Dynamic PDM contains a lagged dependent variable as a regression. It considers a dynamic model with variable intercepts. Take the following equations as an example: In the above equation, it y is the dependent variable of unit i at time t;  is the dynamic coefficient with -1 <  <1; the firm and the debt ratio of the firm can't response to the current change in company situation. Therefore, the independent variables in the earlier periods could have impacts on the debt ratio.

Ahn and Schmidt (1995) introduce a GMM (Generalized Method of Moments) framework to utilize all available mo ment
conditions understandard assumptions. This study will apply optimal GMM estimat ion by using longer lags of the dependent variable as addit ional instruments. In the empirical study of Li (2010), pointing out because of linear correlated with the error term due to the structure of fixed effect model, the capital structure of the firm in the last year can be determined as endogenous. The endogenourity of the other variab les could be tested by comput ing the correlat ion and some other specific tests in the model.

Data Collection
This empirical research will adopt Descriptive Statistic Analysis to fully explore the feature of capital structure for Therefore, th is study is going to follow the literatures and check whether the init ial theoretical results are still co mpatib le with the current capital structure of the real estate firms in Chinese markets.

RESULTS AND DISCUSSION
As for the big real estate firms, there is a negative relationship between profitability, growth, tangibility, liquid ity and non-debt tax shields to the leverage. There is a positive relat ionship between size and non -circulat ing shares to the leverage. It proves that the growth and tangibility are the non-significant variables in this regression.
According to the above negative relat ionship, it demonstrates that high leverage company has low profitability, gro wth and high level of bankruptcy. Due to that, they have to incur more debt, because if higher leverage co mpanies want to issue shares, not many people will be interested in buying the shares. Thereby, the only option for the high leverage company is to issue debt, because if the companies are profitable, they don't need to raise finance.
In this case, debt is the last choice, because the company has retained earnings. This is in line with the findings in this paper where the first choice of cap ital in China is equity, and debt is the last choice. Therefore, only co mpany without much p rofitability and growth will have to borro w money fro m the debt market. In this case, most of the debts come fro m the banks. Big real estate companies mainly in Ch ina are controlled by the government, they are not like the real shareholder. The companies actually are from their parent which is the government.
As for the s mall real estate firms, there is a negative relationship between size, profitability, growth, tangibility, liquid ity and non-debt tax shields to the leverage. Only non-circulating shares have the positive relationship with the leverage.
While, in this regression, growth, tangibility and non-circulating shares are the non-significant variables.
Based on the positive relationship of non-circulating shares to the small leverage real estate firms, it could be judged that most of the s mall real estate companies are not state owned enterprises. They are s mall enterprises who don't have major banks support them due to the non-significant of the non-circulating shares in the regression.
The positive relationship between size and non-circulating share represent they are state own enterprise. The large real estate companies are mainly own by the government. Thus, if they are profitable, the government will buy their share that is why the equity is the first choice, because they can sell the shares to the gov ernment very easily and quickly. Un like the small companies, they have to sell the shares to the public.
As to the whole samp le size of the real estate firms in China, there is a negative relationship between profitability, gro wth , tangibility, liqu idity, non-debt tax shields to the leverage, while there is a positive relat ionship between size and non-circulating shares to the leverage. In this regression, size, growth and tangibility are non -significant variables.
Large co mpanies are more dominant because there is more large co mpany in the market place. China real estate companies are dominated by the large company. Our findings show that China real estate industry is controlled or dominated by large companies. When analyzing whether and how strongly ope rating risk affects leverage, it shows that firms with a h igh level of operating risk have lo wer leverage, due to the purpose to maintain a moderate total risk profile, avoiding to adding financial risk.

CONCLUS ION
The results show that the real estate firms in Ch ina choose equity financing first, then debt financing and the last choice is the internal financing. Th is is opposite of the pecking order theory which suggest that firm choose internal financing, debt and lastly equity. In China many real es tate firms are govern ment owned and therefore they have advantages in raising capital through issuing of stocks. Since minority shareholders do not demand high dividend, it is a relatively cheap way of raising capital.
Second, there is a negative relationship between profitability, gro wth, tangibility, liquid ity and non -debt tax shields to the leverage rat io for the big real estate firms. There is a positive relat ionship between size and non-circulating shares to the leverage. This means that large firms have lo w nu mber of available shares for trade which suggest that they are owned by large shareholders in China case most likely State own enterprise (SOE).
Co mpared to the s mall real estate firms, there is a negative relationship between size, profitability, growth, tangibility, liquid ity and non-debt tax shields to the leverage ratio. Only non-circulating shares have the positive relat ionship with the leverage. Hence, when comb ing the big and small real estate firms together, there is a negative relations hip between profitability, growth, tangibility, liquid ity, non-debt tax shields to the leverage, while there is a positive relationship between size and non-circulating shares to the leverage.
Third, th is emp irical study proves that the growth and tangib ility are the non-significant variab les in this regression in the big real estate. While, in the s mall real estate firms, growth, tangibility and non -circulating shares are the non-significant variables. A comb ination of the big and small firms indicates t hat the size, growth and tangibility are non-significant variables.