The Determinants of company capital structure
Evidence from listed Chinese firms
Master thesis
Zhe Liu
Student no. 10630414Supervisor: Prof. Alex Clymo
Second reader: Prof. W.E. (Ward) Romp
Content
1. INTRODUCTION ... 4 1.1. OBJECTIVE OF THIS PAPER ... 5 2. LITERATURE REVIEW ... 7 3. CHINA’S FINANCIAL MARKET AND INDUSTRY ... 10 3.1. FINANCIAL MARKET ... 11 3.2 INDUSTRY FACTS ... 13 4. VARIABLE SELECTION ... 16 4.1. MEASUREMENTS OF CAPITAL STRUCTURE ... 17 4.2. DETERMINANTS OF CAPITAL STRUCTURE ... 18 4.2.1. Firm-specific ... 18 4.2.2. Monetary policy ... 21 4.2.3. Industry classification ... 24 5. DATA ... 26 5.1. DATA PROCESS ... 26 5.2. DATE DESCRIPTION ... 27 5.3. CORRELATION ANALYSIS ... 28 6. METHODOLOGY ... 30 6.1. MODEL ... 30 7. RESULTS AND ANALYSIS ... 33 7.1. FIRM-SPECIFIC DETERMINANT ... 35 7.2. MONETARY POLICY DETERMINANT ... 38 7.3. RESULTS FROM THE SECOND REGRESSION ... 42 8. CONCLUSION ... 47 9. LIMITATION AND FUTURE RESEARCH ... 49 10. REFERENCE ... 50 11. APPENDIX ... 55
Abstract
This research aims at investigating the determinants of capital structure of listed Chinese firms over 2010 to 2015 period. The categories of the determinants are firm-specific and monetary policy measurements. By using the panel data regression, the results show that firm size, borrowing capacity and profitability are significantly related to the short-term and long-term leverage. Specifically, size and borrowing capacity have the negative relations with the short-term debt ratio but the positive relations with the long-term leverage. Profitability negatively correlates with leverage concerning both short-term and long-term. Growth potential is only significantly associated with short-term leverage. All three monetary policy indicators in this paper are insignificant related to short-term leverage. Policy lending interest rate and inflation determine the long-term leverage positively, and deposit reserve ratio is negatively related to the long-term leverage. Also, this paper finds that the different sectoral implications of monetary policy on capital structure. Monetary policy shows its significance on mining, manufacturing, IT and wholesale & retail industries. By comparing with the government current industrial policy, this paper concludes that the effect of recent expansionary monetary policy imposed by the central bank of China (PBoC) on industrial policy is ambiguous.
1. Introduction
Research on the capital structure of the company has been increased considerably since Modigliani and Miller (1958) proposed their paper that the value of firms is independent of the financial market. Consequently, more and more researchers focus on what the company's financing choice is and what the determinants of capital structure are. In recent years, with the rapid growth of Chinese financial markets, a considerable amount of Chinese firms have gone public (Tong, 1999). It is worth to investigate on how these corporates are looking for external funding to fulfill their growth needs and what factors may influence their financing choices.
On the one hand, company’s specific factors are believed to exert the impact on their financing behavior. Most of the works concentrate on the determinants such as firm size, growth opportunity, tax shield, volatility and so on (Ozkan, 2001; Silva et al., 2015; Ferrarini et al., 2017). Small firms intend to take on more debt but are more likely to face financial constraints (Thomas & Vincenzo, 2006). Huang and Song (2002) also noticed that the uniqueness of listed Chinese firms regarding the enterprise characteristics impact on the financial decision. This paper documents four firm-specific factors to test whether or not they can be regarded as determinants of capital structure, which are the size, borrowing capacity, profitability and growth potential.
On the other hand, the effect may come from the monetary policy that imposed by Central Bank of China (PBoC) as well. China has an independent monetary policy framework, which allows the Central Bank to apply different monetary policies aiming at maintaining economic growth and financial stability (Goodfriend & Prasa, 2007). It is not hard to imagine that the use of monetary policy instruments can influence the company financing in the market level via different channels, such as interest rate and credit channels. Worldwide, three monetary policy tools are largely used: benchmark interest rate, reserves
requirement ratio and open market operations. These monetary policy instruments are the potential choices for most of the countries; however, different from other advanced economies, China uses conventional monetary policy more often (Sun, 2013). The Central Bank influences on the private deposits and lending interest rates directly by setting a policy interest rate and reserve requirement ratio, therefore, the central bank can either tighten up or inject liquidity into the market depends on different economic conditions. The reactions of commercial banks on various monetary policies are the keys for corporates financing due to that the individual firms borrow directly from the commercial banks instead of the central bank.
Also, industry classification is important to make a distinction between firms from different sectors that are potentially facing unequal financing opportunities. In recent years, the Chinese government has introduced several industrial policies to encourage or discourage the evolution of some areas. For example, the stricter operating requirements have been imposed on those high-energy consuming enterprises to deal with the pollution. At the same time, by awareness of the higher labor costs and increased market saturation, it is necessary to pay more attention to the new sources of economic growth. The traditional labour-intensive industry is no longer satisfying the rapid expansion needs; innovative and high-tech companies are thus in favored.
1.1. Objective of this paper
The major purpose of this paper is to detect how the firm-specific factors and monetary policy together affect the corporates’ financing behavior. The capital structures are assessed from the profiles of 7 Chinese industries: automobiles, construction, information technology (IT), manufacturing, mining, real estate and wholesale & retail.
Therefore, the basic research question of this paper is:
What are the capital structure determinants of listed Chinese company?
The following two sub-questions are helped to answer the research question. The first regression model focuses on what and how firm-specific factors and monetary policy affect the company financing decision, regardless of industry classification.
Due to the different monetary policies may generate sectoral effects and therefore, the second regression model aims at answering what the different implications of monetary policy on financing are, concerning the firms from various industries. Most importantly, whether or not the recent monetary policies can indeed assist the accomplishment of the government industrial policy, regarding stimulating or tightening up loans for some sectors.
The reason to investigate the implication of monetary policy and its effect across different industries is that there is a potential conflict of interests between the Chinese government and PBoC. For example, the government intends to stimulate the real estate market on the supply side, yet the PBoC aims at cool down the market by tightening up the loans to developers (Qian, 2010 and Lu, 2007).
To answer the above research question, the methodology used in this paper is panel data regression with pooled OLS, fixed effect, and random effect techniques. After applying F-test and Hausman test, fixed effect panel is regarded as the most appropriate technique to interpret the data. This paper observes that the firm-specific factors such as size, borrowing capacity, and profitability determine both companies’ short-term and long-term leverage. Growth potential is only significantly related to the short-term leverage. Moreover, monetary policies are the determinants for firms’ long-term debt ratio instead of
short-term one. Finally, monetary policies show their significance on the capital structure of firms that are from mining, manufacturing, IT and wholesale & retail industries.
This paper expresses as the following structure. Section 2 lists several previous empirical papers that are related to this research. Section 3 introduces background information of China’s financial market and industrial policy. Section 4 introduces the variable selection in detail, which is the baseline of conducting the regression models. Section 5 documents the dataset used in this paper. The methodology is represented in Section 6. Section 7 is the results of the models as well as the analysis. Finally, Section 8 shows the conclusion and Section 9 is the limitation and suggestion for future studies.
2. Literature review
Several previous papers focused on the similar topic, both from China and other countries. Ali (2011) used the data of Pakistan from 2003 to 2008 to investigate what variables are the determinants for capital structure of 368 Pakistani nonfinancial firms; he found that the significant empirical evidence to prove the relation between firm-specific factors and leverage. Among them, for example, profitability is negatively related to the leverage. Size, tangibility, growth opportunity and dividend are positively related to the leverage. Also, in his paper, the financing behavior of the companies in different industries shows the significant change. For instance, the size has the strongest effect for paper & board sector. Tangibility for engineering firms is negatively related to the total debt, but for other industries this relation is positive. He also added the inflation rate into the model and found that, in general, the positive relationship between inflation and total debt, however, the inflation is negatively related to the totaldebt in the engineering sector. Mahmud (2003) documented the capital structure of listed companies in Japan, Malaysia, and Pakistan as well; the period is from 1989 to 1998 with 718 firms included. The results demonstrated that the industry classification has the strongest impact in Japan and the size is positively related to leverage in Japan but the relation is negative in the other two nations. Also, the asset growth, fixed asset, and asset return are the essential determinants of leverage in all three countries. The prime interest rate is used as monetary policy indicator in his paper, and there is a positive relation between prime interest rate and leverage in Pakistan. Dennis et al. (2013) examined the monetary policy that is measured by interest rate, on 3,301 US public-traded firms; they found that the biggest effect contributed to the wholesale & retail sector, and the larger firms seem to mitigate the policy effect. Prasad and Ghosh (2005) studied the behavior of manufacturing companies in India with the monetary policy shock from 1992 to 2003 with 525 firms included. The monetary policy is measured as the yield on 364-day treasury bills. They demonstrated that the firms’ overall debt level decreases under the tight monetary policy but their short-term bank loan increases after the monetary tightening. Kashyap et al. (1994) examined the monetary policy transmission of US firms. Similarly, they suggested that under the tight monetary policy, companies preferred to be financed by issuing commercial paper rather than bank loans. Aliyev et al. (2014) investigated the impact of both monetary policy and firm-specific factors on the financing of Czech firms. They collected the information of 57,000 firms from 2003-2011, and the monetary policy measurement was the market interest rate. The result proved that the higher the interest rate, the lower the share of long-term and short-term bank loans. Also, the size and tangibility are positively related to leverage but the relationship between profitability and leverage is negative. Besides, Bougheas et al. (2006) argued that the riskier and smaller enterprises are harder to access to loans at
the time of tight monetary policy. Abuka et al. (2015) focused on the relationship between monetary policy rate and the loan that is granted by banks to 26,363 Uganda firms during 2010-2014, with the result that the increase of interest rate reduces the bank credit supply.
The empirical studies for listed Chinese firms are mostly focusing on the firm-specific factors. Huang and Song (2002) used more than 1,000 firms for the year of 2000 with size, profitability, tangibility, and tax as measurements. Size and growth opportunity are positively related to the leverage. Profitability and tangibility, instead, have the negative relation with leverage. Tse and Rodgers (2014) examined 831 listed Chinese firms and compared the debt-to-assets ratio of Chinese manufacturing sector to other industries, based on the firm-specific factors. The result showed that the debt ratio is not special for manufacturing industry even though it plays a leading role in China’s economy. However, in Tong’s paper (1999), he used the 1,577 firms’ data from 1993-1997, the result is that the leverage level for asset-intensive industries such as manufacturing is much higher than other industries. For the firm-specific determinants, he found that the profitability is negatively related to the leverage ratio, size and tangibility are positively correlated with the leverage. Also, the results show that the growth rate is not significant except for the year of 1995. Chen et al. (2014) examined 1,481 nonfinancial firms in China and illustrated that the companies from real estate sector tend to borrow more and manufacturing firms prefer more long-term debt financing. Similarly, the size shows a positive correlation with debt ratio and profitability is negatively related to leverage.
The papers on the implication of monetary policy in China are limited. Silva et al. (2015) examined the Chinese market from 2002 to 2011 with 10,181 observations. They observed that the inflation rate is negatively related to the leverage but hardly significant. Lastly, Wen (unknown) researched on the design
in companies of China by adding the independent variables of lending interest rate and reserve requirement ratio. Their sample consists 1,700 firms from 1998 to 2007, and the results show that the interest rate in positively related to the leverage, and the correlation towards the reserve requirement ratio is negative. Overall, various previous papers prove that the firm-specific factors such as size, profitability, growth opportunities, and tangibility are the determinants for capital structure but their effects on the capital structure are mixed. In most of the papers, size is positively related to the leverage and the negative relation between profitability and leverage. The effect of tangibility can be either positive or negative, which depends on the countries. Growth opportunity proved to be positively related to the leverage but its influence is less important than other determinants. The market interest rate is largely used as the monetary policy indicator and there is a negative relationship between interest rate and leverage such as in Czech and Uganda. However, this relation can also be positive in China and Pakistan, as shown in the paper of Wen (unknown) and Mahmud (2003), respectively. Concerning the implications of firm-specific factors in different industries, the results are ambiguous. Tong (1999) stated that the manufacturing firms tend to have higher leverage level yet Tse and Rodgers (2014) said the manufacturing is not special in China. Besides, Chen et al. (2014) confirmed that the real estate firms contain larger leverage than manufacturing firms in China.
3. China’s financial market and industry
This section introduces some basic information about Chinese financial environments such as stock market, monetary policy, and industrial policy. Since the economic reformation in 1978, China has gradually become a market-oriented economy with remarkable economic progress.
3.1. Financial market
Among different financing options, bank loans are still the essential method for Chinese firms, which is approximately 12 times bigger than the equity market (Sommers, 2002). There are 26 listed commercial banks in China with a 1,332,37 billion Yuan (RMB) net profit in the year of 2015, although the growth rate is positive, there is an obvious downward trend for the growth speed, especially for the “big five” state-owned banks1. However, the amount of social financing
continually increases in the recent year. As shown in Graph 1, the growing quantity of social financing regarding RMB loans enhances dramatically. In most of the countries, whether or not to issue the loan is determined by the market interest rate, which can also be seen as the benchmark interest rate. China has launched the Shanghai Interbank Offered Rate2 (SHIBOR) and prime lending
interest rate3 (PLR) in 2007 and 2013, respectively, as part of the interest rate
liberalization process; these rates have shown their success in guiding the decision of market. However, due to the incompleteness of the interest rate liberalization, the policy interest rate announced by the central bank of China (PBoC) is still regarded as one of the most significant monetary policies and to guide commercial banks in setting up their lending interest rate (Qi, 2013). PBoC policy interest rates contain several base rates such as the central bank lending rate, discount rate and the interest rates on required and excess reserves (Sun, 2013). The other two monetary tools are reserve requirement ratio and open market operation. As stated in the introduction section, normally the central bank from advanced countries seldom alter the reserve ratio requirement and policy lending interest rate, the situation in China is opposite. Graph 2 shows that 1 Source: Listed banks in China, 2015 review and outlook from EY. Industry and Commercial Bank of China (ICBC), Bank of China (BOC), China Construction Bank (CCB), Agricultural Bank of China (ABC) and Bank of 2 The SHIBOR is determined by 18 commercial banks that consist the price quotation group. SHIBOR contains overnight, 1-week, 2-week, 1-month, 3-month, 6-month, 9-month and 1-year maturities. The rate is published every working day at 11 am. http://www/shibor.org/ 3 1-year prime lending interest rate (PLR) is provided to the commercial banks’ best costumer. Now fixed at 4.3% that is 0.05% lower than the policy lending interest rate. http://www.icbc.com.cn
during the year of 2010 to 2011, the PBoC hiked policy interest rate by five times and has reduced the rate by eight times since 2012. Similarly, the deposit reserve ratio was increased ten times during 2010 to 2011 and the ratio has been cut by seven times since 2012. Therefore, the recent monetary policy has turned to be expansionary as the policy lending interest rate and reserve requirement ratio are gradually decreasing, since the year of 2012. Highly rely on the conventional monetary policy is the major difference from those advanced economies and due to the decision of issuing loans are implicitly influenced by policy interest rate, the monetary policy may have the considerable impact on the capital structure of Chinese companies.
Graph 1: The flow amount of three major ingredients of social financing from 2010-2015, which measures in 100 million Yuan (RMB). Source: National Bureau of Statistics of China (NBS) 0 20,000 40,000 60,000 80,000 100,000 120,000 2010 2011 2012 2013 2014 2015 RMB Loans Credit Loans Entrusted Loans
Graph 2: PBoC policy lending interest adjustment from 2010 to 2015 Source: PBoC database
3.2 Industry facts
Since the year of 1978, the industry in China has been rapidly restructured, from a traditional agricultural society to a more modern industrial economy. Moreover, the stock market has been reopened since the 1990s. Graph 3 shows that there are 2,827 companies listed on Shanghai Stock Exchange (SSE) and Shenzhen Stock Exchange (SZSE) in the year of 2015, respectively, and the number grows gradually. 0.00% 1.00% 2.00% 3.00% 4.00% 5.00% 6.00% 7.00% 2010.1 2010.6 2010.11 2011.4 2011.9 2012.2 2012.7 2012.12 2013.5 2013.10 2014.3 2014.8 2015.1 2015.6 2015.11Policy lending interest rate
Policy lending interest rateGraph 3: The number of listed companies from 2010-2015 and SSE&SZSE, respectively. Source: National Bureau of Statistics of China (NBS) Generally speaking, there are seven key sectors and their outputs that drive the modernization process: Steel, Cotton, Tobacco, Automobiles, Beer, Coal and High Technology. The industrial policy, which is designed by the government for different industries in terms of the competitive advantages, shows its importance as it works as a guideline to stimulate or slow down certain industries (Azarhoushang, 2013). As a matter of fact, China is the biggest consumer and producer of the most mining product, such as coal, gold and other rare minerals (Steen, 2016). However, two underlying problems may restrict the development of mining industry. Firstly, the air, water, and land pollutions appear to be more and more severe with those energy-intensive companies. Secondly, the overcapacity of the Chinese mining firms itself. The government has imposed several policies to regulate and standardize the entire industry. Back to the year of 2009, the high-emission companies were strictly controlled to avoid growing excessively and for those enterprises that lack of innovative methods to limit their emission to a certain level, were forced to closed until they 0 200 400 600 800 1000 1200 1400 1600 1800 2000 2010 2011 2012 2013 2014 2015 Shanghai Stock Exchange(SSE) Shenzhen Stock Exchange(SZSE)
met the requirements. Additionally, the market entries requirements were more demanding than before. The above industrial policies may result in the less accessibility to loans faced by mining companies.
The automobile industry takes the advantages of the policy, especially for hybrid and electric power vehicles. For example, the license-plate lottery policy has been launched since the year of 2011, yet it is not applicable to new-energy vehicles. Together with other finance-injection projects, the demand for loans is expected to expand.
The percentage of China’s manufacturing production to the world reached 19.8% in 2010, which made China became the world biggest manufacturing country. Meanwhile, the manufacturing faces several difficulties. For example, the growing costs of labor in recent years significantly erode the profitability of companies, especially the small and medium-sized enterprises. “12th five-year plan4” announced the goal of industry upgrading, which includes replacing labor
input by machinery, lowering the production costs and increasing productivity. Rise of financial support and demand for funding are assumed by manufacturing companies.
The wholesales & retail industry in China has grown for many years, yet the traditional way of managing cannot satisfy the consumption behavior nowadays. With the help of Internet, more and more companies peruse online selling. The “12th five-year plan” stated that two relevant policies to support industry growth. Firstly, encouraging the creation of online shops, secondly, promoting the big companies to go overseas by establishing retail terminals and distribution centers.
In addition, the IT industry, which apparently has a bright future as it was defined by one of the emerging industries in the “12th five-year plan” report, one 4 “Five-Year plan” for national economy and social development, which mainly focuses on the major construction projects, productivity distribution and important proportion of the national economy. “Five-Year plan” started from the year of 1953, the “12th Five-year plan” is from 2011 to 2015. http://www.china.com.cn/chinese/zhuanti/wngh/1163433.htm
of the major encouraging industrial policies is the tax incentive. Many emerging fields in China relied on the development of information technology such as electronic commerce and aerospace industry.
Regards the construction industry, the government has imposed notable support to the construction sector. Abundant capital flows into infrastructure facilities such as public transportations and buildings since the year of 2008 (ESC, 2015). Real estate, as one of the most important stanchion industries in China, has grown significantly in recent years, which can be reflected by the enormous housing pricing arises. The policies seem to be conflict; several actions show that the government would like to increase the shortage of housing supply by encouraging the bank loans to real estate market (Qiao, 2010). However, monetary policy targets at cooling the market by imposing the credit control (Lu, 2007). Moreover, due to the greater risk aversion, the growth of loan to the industry shrinks (Hsu, 2014).
4. Variable selection
The purpose of this paper is to examine what the determinants are of importance to impact the capital structure of listed Chinese companies. The previous empirical studies suggest a broad range of variables that can potentially be the influential factors, most of which goes to firm-specific factors and monetary policy perspectives.
This section introduces the variables used in this paper, firstly, the measurements of capital structures, which are short-term and long-term leverage. Moreover, four firm-specific determinants, size, borrowing capacity, profitability and growth potential, are interpreted. Third, the paper also includes three monetary policy measurements: policy lending interest rate, reserve requirement ratio, and inflation. In detail, policy lending interest rate and
required reserve ratio are exogenous variables that can be adjusted by PBoC directly. Inflation is typically a policy target and an endogenous variable, the reason to include inflation as a measurement of monetary policy is that normally the high inflation signifies loose monetary policy. Besides, the industry classification is recorded by dummy variables.
4.1. Measurements of capital structure
How to measure the capital structure of the company is the key to answering the research question. According to Titman and Wessels (1988), leverage level can be measured in many ways and are the appropriate proxy for capital structure. The most measurements contribute to the total, short-term and long-term debt to total debt with book or market value5. Additionally, Alivey et al. (2014) also added the trade credit to total asset ratio. Muthama et al. (2003) took the ratios of total liabilities to assets and total debt to equity into account. This paper applies the following two leverage ratios to investigate the accessibility of loans. STD The ratio of short-term debt to total assets, by book value. LTD The ratio of long-term debt to total assets, by book value. Firm’s default risk goes up with its debt ratio, yet the manager can remain the control of firm not like when using the equity financing. Therefore, there is a trade-off between the risk and leverage selection. Short-term debt refers to the outstanding loans that have to be repaid within a year. On the contrary, companies do not have to arrange the repayment within one year for long-term debt.
4.2. Determinants of capital structure
Refer to the previous papers and particular Chinese financial environment, the following proxies are selected as independent variables, together with their hypothesizes of the influence on capital structure.
4.2.1. Firm-specific
Size
Most of the previous papers use the firm size as a proxy to investigate the determinant of capital structure. The measurement of size in this paper is the nature logarithm of company’s total assets.
It is of great pain to the economy if companies go bankrupt, due to that the large and listed firms are critical in China, therefore, the potential support by Chinese government more likely goes to the big companies in China, which leads to the low cost of default and the relatively high proportion of debt financing (Tong, 1999). It is assumed that the relation between size and leverage is positively correlated. Indeed, the positive relation confirmed by many researchers, Wen (unknown) illustrated that the size shows the positive relationship with both overall and long-term debt ratios. Silva et al. (2015) stated that the positive relation between firm size and long-term debt ratio.
Instead, Myers and Majluf (1984) disagreed with this theory and proved the negative link. Larger firms prefer to choose internal funding rather than external debt to financing. Kester (1986) illustrated in his paper that the larger firms hold the greater capability to issue equity.
Though the relation between size and leverage can be ambiguous, considering the governmental influence in China, the following hypothesis is made:
long-term debt ratios. Borrowing capacity The borrowing capacity of the company is believed to have the similar effect to size and the indicator for borrowing capacity in this paper is the ratio of tangible fixed assets to total assets. Banks are more likely to issue the loans to those corporates with relatively high tangibility due to that the tangible fixed assets can be easily collateralized for the loan. Wen (unknown) explained that the borrowing capacity is of particular importance among Chinese market, which contains higher uncertainty and abundant emerging corporates, especially for the short-term debt due to that the repayment has to be repaid within a year. Therefore, the hypothesis here is: H2: There exists a positive relationship between company’s borrowing capacity and leverage ratios, regarding both short-term and long-term. Profitability Profitability is another crucial variable that may determine the financing choice. This paper uses the ratio of earning before interest and tax (EBIT) to total asset as the indicator of profitability (Yang, 2016). The relation between profitability and leverage is ambiguous. Intuitively, firms with higher profit could have a relatively higher proportion of debt due to the lower default risk. However, it is also reasonable that the firm uses retained earning to financing since the transaction cost disadvantages the exploit of external debt financing. Besides, Singh (1995) observed that in developing countries, firms prefer equity financing to debt if external financing is necessary. Thus the relation between profitability and leverage could be negative as well, which is consistent with most of the empirical findings that are introduced in
Section 2. Besides, Rajan and Zingales (1995) also proved that the negative relationship emerges between leverage and profitability. Following the logic of Singh (1995) and the findings of previous papers, the hypothesis of the profitability is, therefore: H3: Company’s profitability has a negative correlation with its leverage, for both short-term and long-term debt ratios. Growth potential The relation between growth potential and leverage is rather obvious. The firm with higher growth potential results in lower default risk so that the more growth followed by more leverage (Tong, 1999). The rapid growth leads to higher leverage level is also suggested by Gupta (1969) and Ginn & Young (1995).
However, the indicator of growth rate shows controversy. Tong (1999), and Titman and Wessels (1998) used the growth of total assets as their indicator. Instead, Ferrarni et al. (2017) documented the ratio of revenue growth to total assets growth as the measurement. Also, Tse & Rodgers (2014) employed the percentage change of EBIT. This paper uses the percentage change of total asset for the reason that the growth of assets shows greater stability than the modification of EBIT or revenue, which can be temporary. The hypothesis is, therefore: H4: The growth potential of the total asset is assumed to be positively related to the leverage, concerning both short-term and long-term debt ratios.
4.2.2. Monetary policy
Not only the firm characteristics may determine the capital structure but also the macroeconomic factors can do so, as suggested by Abuka et al. (2015). The purpose for including monetary determinants in this paper is to test on whether or not the monetary policy set by PBoC can effectively impact the real economy, via the aggregation of credit. Also, do these monetary policies influence the industries differently? As introduced in Section 3, the monetary policy rate frequently alters during 2010 to 2015 with both expansionary and contractionary policies and thus, it is assumed to have the significant influence on the real economy, which in particularly to the company’s capital structure. It is necessary to understand the channels between capital structure and monetary, as well as the measurements of the monetary policy rate. Following part in this section introduces three indicators used as monetary policy rates in this paper. PBoC policy lending interest rate The decision for the company on whether to borrow is highly dependent on the interest rate. The challenge is to find an interest rate that can appropriately represent the monetary rate regard to Chinese financial environment. This paper uses the PBoC policy lending interest rate as the indicator.
Normally, the market interest rates are chosen such as London interbank offered rate (LIBOR) and Federal Reserve interest rate; these market indicators mostly influence the credit aggregate in advanced countries. However, as also mentioned before, the PBoC depends more on influencing directly on personal lending and saving amount by imposing the policy interest rate. Even though the prime interest rate and Shanghai interbank interest rate are launched, the policy interest rate is still regarded as the most critical rate in Chinese credit market. Based on this policy rate, commercial banks are permitted to alter their lending
and deposit interest rates within a certain range6 (Sun, 2013). The policy
interest rate is such important to impact the individual firms’ financing choices due to that the commercial banks are extensively regulated.
The previous paper pointed out to use 3-month PRIBOR7 (Aliyev et al., 2014),
Treasury bill rate (Muthama et al., 2013) or long-term and short-term market interest rates as the indicators for monetary policy (Mokhova & Zinecker, 2013). Considering the less rely on market interest rates of PBoC, the indicator in this paper is thus the PBoC policy lending interest rate that is 4.35% currently.
The effect of lending interest rate to company’s capital structure is ambiguous. Bank lending channel shows that the higher lending interest rate induces the cost of borrowing, and subsequently, the company lowers their leverage ratio. However, empirical studies sometimes show the different stories. Mokhova and Zinecker (2013) in their paper stated that the total leverage is firmly and positively related to both long-term and short-term interest rates for Germany and France. This relation also works for Hungary with the long-term interest rate. Muthama et al. (2013) documented a positive relationship between long-term debt ratio and interest rate. According to Wen (unknown), the interest rate set by PBoC shows a positive relation to overall leverage ratios. Therefore, the following hypothesis has been made after considering the prior results:
H5: The effect of policy lending interest rate is ambiguous to the capital
structure. 6 Currently the floating range for lending interest rate is [0.9, ∞] 7 Prague Interbank Offered Rate
Deposit reserve ratio
Adjusting the required reserve ratio is another monetary policy instruments that are frequently used by PBoC to influence the real economy, which is a discretionary and more direct tool. Deposit reserves ratio can be considered as the minimum requirement that the commercial bank must keep of private deposits. Intuitively, increases the reserves ratio by PBoC limits the amount of commercial bank’s loan availability, which drives up the lending costs (Sun, 2013). In reality, to stimulate the economy, central bank normally applies an expansionary monetary policy that lowers the deposit reserve ratio. The lower lending costs appear to increase company’s leverage level and vice versa. The hypothesis is, therefore: H6: The higher deposit reserve ratio set by PBoC, the lower leverage ratios of the company. Inflation
Apart from the above two rates that PBoC can adjust directly, inflation is an endogenous variable and typically a target, one of the goals of monetary policy is to prevent the inflation rate goes extremely or too low. The raises of inflation rate can be regarded as the loose monetary policy.
The effect of inflation varies among countries; Mokhova and Zinecker (2013) empirically proved that there is a strong positive relation between total leverage and the inflation rate for France and Greece, but for Germany, Poland, and Hungry this relation is opposite. Elkhaldi and Daadaa (2015) stated that the inflation rate is positively related to the long-term debt ratio for Tunisian firms. The hypothesis for inflation rate is:
Noting that the company’s balance sheet is reported by the end of each accounting year, but the above three monetary policy rates are reported and vary during the year. Therefore, the monetary policies have lagged effect on company’s capital structure. That is, the firms make their financing decision based on the previous year’ monetary policy. Also, to avoid the potential endogeneity problem, the monetary policy indicators in this paper are included with one-year lagged.
4.2.3. Industry classification
It is believed that the implication of monetary policy differs across industries on company’s capital structure; the questions are what the magnitude of the effect is and whether or not all industries are sensitive to the monetary policy adjustment (Jansen et al., 2013).Answering the second sub-question of this research, this paper creates seven industry dummy variables to represent Automobile, Construction, Information Technology (IT), Manufacturing, Mining, Real estate and Wholesale & retail, respectively. Xu and Wang (1997) suggested that the manufacturing and real estate companies have a significant influence among listed Chinese firms. Besides, Tong (1999) stated that firms that are more asset-intensive are assumed to have higher leverage, such as manufacturing and real estate companies. Therefore, the hypothesis is:
H8: Monetary policy has the stronger effect on the capital structure for the firms
in manufacturing and real estate sectors in China.
Table 1: summarizes both dependent and independent variables that are used in the model of this paper
Type Measurement Predicted sign of capital structure determinants Dependent variable Short-term leverage (STD) The ratio of book value short-term debt to total assets Long-term leverage (LTD) The ratio of book value long-term debt to total assets Independent variable Firm-specific Size (Size) Natural logarithm of total assets + Borrowing capacity (Borrcap) The ratio of tangible fixed assets to total assets + Profitability (Pro) The ratio of EBIT to total assets - Growth potential (Gro) The percentage change of total assets: (𝑇𝑜𝑡𝑎𝑙 𝑎𝑠𝑠𝑒𝑡𝑠!− 𝑇𝑜𝑡𝑎𝑙 𝑎𝑠𝑠𝑒𝑡𝑠!!!)/ 𝑇𝑜𝑡𝑎𝑙 𝑎𝑠𝑠𝑒𝑡𝑠!!! + Monetary policy rate PBoC policy lending interest rate (L1.InterestR)
One-year lagged PBoC benchmark lending interest rate
?
Deposit reserve ratio (L1.ReserveR)
Inflation rate (L1.Inflation) One-year lagged inflation rate CPI based ? “+” means that the leverage ratio goes up with the factor “-“ means that the leverage ratio decreases with the factor “?” means that the effect of the factor is ambiguous Industry classification DUM 1 Dummy for Automobile DUM 2 Dummy for Construction DUM 3 Dummy for Information Technology (IT) DUM 4 Dummy for Manufacturing DUM 5 Dummy for Mining DUM 6 Dummy for Real estate DUM 7 Dummy for Wholesale & Retail
5. Data
5.1. Data process
The data of firm-specific information used in this paper are from Bureau van Dijk (BvD) database, which contains the profiles of listed Chinese companies’ financial statements. The dataset is extracted from the balance sheet and profit & loss statement of each company. In detail, the short-term debt belongs to the category “Loan” and Long-term debt refers to the “non-current interest-bearing debt” under firms’ liability. Total assets and tangible fixed assets are from the“asset” side of the balance sheet. Besides, EBIT goes to operating profit and loss under firms’ profit & loss statement. Regards to the monetary policy rate indicators, policy lending interest rate and deposit reserve ratio are from the PBoC’s database, and the inflation rate is from the World Bank database. The period in this paper is from 2010 to 2015, the data of 2016 is not included since during the data collection, many companies annual reports have not been disclosed or updated yet.
After narrowing down the period to 2010-2015 and neglecting all the missing data, the dataset in this paper contains 1,037 Chinese listed firms, which are from 7 different industries: 43 automobile, 44 construction, 771 manufacturing, 38 mining, 17 IT, 59 real estate and 65 wholesale & retail companies. The currency is converted into US dollars, and the measurement is in thousands. Therefore, the dataset is finalized with 6,222 observations sorted by six years and 1,037 companies.
Moreover, when processing the data, due to the monetary rates do not vary on an annual basis, to be consistent with other firm-specific data on the time-series dimension, the policy lending interest rates, deposit reserve ratio are yearly averaged.
5.2. Date description
Table 2 provides the descriptive statistics for the overall dataset, from 2010 to 2015. Results show that the selected firms prefer to use short-term debt financing; the mean for STD is 0.1845 but for LTD is 0.1078. However, as shown in Appendix A, the construction and real estate companies contain higher long-term leverage ratios that may due to their longer production cycle than firms from other industries, the means of LTD for them are 0.1729 and 0.1966, respectively. Also, the wholesale & retail companies depend on the short-term
debt financing most, which the ratio is 0.1983. Automobile companies have the lowest mean of long-term debt ratio. It is not surprising that mining firms have the highest mean of borrowing capacity and size. That is they contain the largest tangible fixed assets and total assets on their balance sheets. Mining companies in China are mainly supported by the government and need the advanced and vast amount of mining facilities.
Table 2: Overall Summary Statistics
Observations Mean Min Max Std. Dev.
STD 6,222 0.1845 0.0000 21.7570 0.3654 LTD 6,222 0.1078 0.0000 1.2323 0.1068 Size 6,222 5.9536 2.7342 8.5945 0.6001 Borrcap 6,222 0.3022 0.0000 0.9425 0.2011 Pro 6,222 0.0396 -2.2821 0.5130 0.0730 Gro 6,222 0.2734 -0.9941 47.4502 1.2695 InterestR 6,222 5.80% 4.98% 6.29% 0.0044 ReserveR 6,222 19.46% 17.20% 20.50% 0.0116 Inflation 6,222 2.88% 1.4% 5.4% 0.0127
5.3. Correlation analysis
Table 3 shows the results of Pearson correlation analysis, which is a quick look at the data; the formal panel data regressions to investigate the relations between variables are introduced in the Methodology and Result Sections. The size is significantly correlated with all other variables yet with less than 0.1 correlation coefficients. Interestingly, there is a negative relationship between the borrowing capacity and size, which indicates that the firms with larger size tend to have a lower proportion of the tangible fixed assets. Moreover,borrowing capacity is negatively related to the profitability as well as the growth potential show that the firms contain less tangible fixed assets if they are facing higher profitability and growth potential.
All the correlation coefficients are relatively small and can be neglected except for the relation between one-year lagged interest rate and one-year lagged deposit reserve ratio. In reality, the PBoC sometimes adjusts the policy interest rate and reserve ratio simultaneously so that the high correlation between them is understandable. Concerning the multicollinearity issue, the variance inflation factor (VIF) is estimated to test for the multicollinearity among these independent variables. The result in Appendix B shows that the highest VIF is 5.06 that is smaller than 10; therefore, the multicollinearity is not a concern in the analysis. The same method used in the paper of Aliyev et al. (2014) as well.
Table 3: Pearson Correlation Analysis for 2010 to 2015
Size Borrcap Pro Gro L1.InterestR L1.ReserveR L1.Inflation
Size 1 Borrcap -0.0313* (0.0135) 1 Pro 0.0941* (0.0000) -0.0368* (0.0037) 1 Gro 0.0489* (0.0001) -0.0520* (0.0000) 0.0732* (0.0000) 1 L1.InterestR 0.0359* (0.0097) 0.0238 (0.0869) -0.0502* (0.0003) -0.0199 (0.1516) 1 L1.ReserveR 0.0869* (0.0000) 0.0165 (0.2337) -0.1199* (0.0000) -0.0412* (0.0030) 0.8563* (0.0000) 1 L1.Inflation -0.0802* (0.0000) 0.0054 (0.6964) 0.0772* (0.0000) 0.0108 (0.4365) 0.2732* (0.0000) 0.0158 (0.2553) 1 * Significant at 5% level, the first number is the Pearson correlation coefficient, P-value in parentheses
6. Methodology
The dataset contains 1,037 Chinese public companies that from 7 industries as during 2010 to 2015, consequently, the panel regression methods are applied to determine the relations between leverage ratios and independent variables. This paper borrows and revises the method of Aliyev (2014) by using policy lending interest rate, deposit reserve ratio, and inflation. The first panel regression is run by pooled OLS, fixed effect and random effect panel techniques. The F-test for fixed effect is applied to select between pooled OLS and fixed effect, and Hausman test is used to choose between fixed and random effect. After comparing the results, the most appropriate technique can be decided. The second panel regression is estimated by only fixed effect technique.
6.1. Model
To answer the research question, what are the determinants of listed Chinese companies’ capital structure; the first regression model is conducted in terms of two different leverage ratios: 𝑌!,! = 𝛼! + 𝛼!𝑆𝑖𝑧𝑒!,!+ 𝛼!𝐵𝑜𝑟𝑟𝑐𝑎𝑝!,!+ 𝛼!𝑃𝑟𝑜!,!+ 𝛼!𝐺𝑟𝑜!,!+ 𝛽!𝐿1. 𝐼𝑛𝑡𝑒𝑟𝑒𝑠𝑡𝑅! + 𝛽!𝐿1. 𝑅𝑒𝑠𝑒𝑟𝑣𝑒𝑅!+ 𝛽!𝐿1. 𝐼𝑛𝑓𝑙𝑎𝑡𝑖𝑜𝑛! + 𝜀!,! 1 &(2)
Where 𝑌!,! represents 𝑆𝑇𝐷!,! and 𝐿𝑇𝐷!,!, respectively. 𝑆𝑖𝑧𝑒!,!, 𝐵𝑜𝑟𝑟𝑐𝑎𝑝!,!, 𝑃𝑟𝑜!,! and 𝐺𝑟𝑜!,! are firm-specific determinants. 𝑖 denotes the different companies that 𝑖 = 1,2,3 … 1037. 𝑡 stands for the different years, which in this paper equals to 𝑡 = 2010,2011,2012,2013,2014,2015. Moreover, 𝐿1. 𝐼𝑛𝑡𝑒𝑟𝑒𝑠𝑡𝑅!, 𝐿1. 𝑅𝑒𝑠𝑒𝑟𝑣𝑒𝑅! and 𝐿1. 𝐼𝑛𝑓𝑙𝑎𝑡𝑖𝑜𝑛! are one-year lagged monetary policy factors that are independent from the cross-section, but are time variant. 𝜀!,! is the
error term.
This first regression contains two models that have different dependent variables. The model one regresses on STD and model two uses LTD as the leverage ratio. These models try to test that, in general, what are the determinants of capital structure, so that the industry classification is not classified in this case. The results can answer the first sub-question of the research. There are three ways to deal with this panel dataset: pooled OLS, fixed effect, and random effect. Specifically, different from the random effect, fixed effect also estimates the firm-specific error terms that are believed to be correlated with independent variables. Moreover, The panel data can be pooled if the heterogeneity problem is not significant (Tse & Rodgers, 2014). Otherwise, fixed effect or random effect is more appropriate. Whether to use fixed or random effect in this model has to be decided as well. Theoretically, fixed effect model is in favor if there is unobserved heterogeneity that correlated with independent variables. That is, the high leverage ratios may come from the company individual effect that is unobserved. If this is the case, ignoring the firm-specific effect may cause omitted variable bias8. The random effect method
can be used without the doubt if assumed that there are no unobserved factors that related with independent variables.
Intuitively, as mentioned before, it is assumed that there exist some unobserved factors such as the manager’s intelligence or age that are correlated with the firm-specific factors so that the fixed effect model is preferred. Furthermore, the Hausman test is used to decide which method is better explained the dataset statistically. As well as the F-statistic observed under the fixed effect is used to determine between pooled OLS and fixed effect techniques. The results of the first regression together with Hausman test and F-test are displayed in Table 4.
8 The original model of fixed effect should be 𝑌!,!= 𝛼!+ 𝛽!+ 𝛼!𝑆𝑖𝑧𝑒!,!+ 𝛼!𝐵𝑜𝑟𝑟𝑐𝑎𝑝!,!+ 𝛼!𝑃𝑟𝑜!,!+
𝛼!𝐺𝑟𝑜!,!+ 𝛽!𝐿1. 𝐼𝑛𝑡𝑒𝑟𝑒𝑠𝑡𝑅!+ 𝛽!𝐿1. 𝑅𝑒𝑠𝑒𝑟𝑣𝑒𝑅!+ 𝛽!𝐿1. 𝐼𝑛𝑓𝑙𝑎𝑡𝑖𝑜𝑛!+ 𝜀!,!, where 𝛽! stands for the
The second regression aims at answering what the sectorial effects of monetary policy on company's capital structure are. 𝑌!,!,! = 𝛼!+ 𝛼!𝑆𝑖𝑧𝑒!,! + 𝛼!𝐵𝑜𝑟𝑟𝑐𝑎𝑝!,! + 𝛼!𝑃𝑟𝑜!,!+ 𝛼!𝐺𝑟𝑜!,! + 𝛾!𝐿1. 𝐼𝑛𝑡𝑒𝑟𝑒𝑠𝑡𝑅! ! !!! ∙ 𝐷𝑈𝑀!,!+ 𝜀!,! 3 &(7) 𝑌!,!,! = 𝛼!+ 𝛼!𝑆𝑖𝑧𝑒!,!+ 𝛼!𝐵𝑜𝑟𝑟𝑐𝑎𝑝!,!+ 𝛼!𝑃𝑟𝑜!,! + 𝛼!𝐺𝑟𝑜!,! + 𝛿!𝐿1. 𝑅𝑒𝑠𝑒𝑟𝑣𝑒𝑅! ! !!! ∙ 𝐷𝑈𝑀!,! + 𝜀!,! 4 &(8) 𝑌!,!,! = 𝛼!+ 𝛼!𝑆𝑖𝑧𝑒!,!+ 𝛼!𝐵𝑜𝑟𝑟𝑐𝑎𝑝!,!+ 𝛼!𝑃𝑟𝑜!,!+ 𝛼!𝐺𝑟𝑜!,! + 𝜃!𝐿1. 𝐼𝑛𝑓𝑙𝑎𝑡𝑖𝑜𝑛! ! !!! ∙ 𝐷𝑈𝑀!,! + 𝜀!,! 5 &(9) 𝑌!,!,! = 𝛼!+ 𝛼!𝑆𝑖𝑧𝑒!,! + 𝛼!𝐵𝑜𝑟𝑟𝑐𝑎𝑝!,! + 𝛼!𝑃𝑟𝑜!,!+ 𝛼!𝐺𝑟𝑜!,! + 𝛾!𝐿1. 𝐼𝑛𝑡𝑒𝑟𝑒𝑠𝑡𝑅! ! !!! ∙ 𝐷𝑈𝑀!,!+ 𝛿!𝐿1. 𝑅𝑒𝑠𝑒𝑟𝑣𝑒𝑅! ! !!! ∙ 𝐷𝑈𝑀!,! + 𝜃!𝐿1. 𝐼𝑛𝑓𝑙𝑎𝑡𝑖𝑜𝑛! ! !!! ∙ 𝐷𝑈𝑀!,! + 𝜀!,! 6 &(10)
Similar as before, 𝑌!,!,! represents for 𝑆𝑇𝐷!,!,! and 𝐿𝑇𝐷!,!,!, respectively. The regression adds subscript 𝑗 to distinguish industries, where 𝑗 = 1,2,3,4,5,6,7 for automobile, construction, IT, manufacturing, mining, real estate and wholesale & retail, respectively. Firm-specific factors are the same as before yet they are considered as control variables in this case. 𝐷𝑈𝑀!,! is the dummy variable used for different industries that refer to 𝐷𝑈𝑀!,!=1 if firm 𝑖 is in the industry 𝑗, otherwise 𝐷𝑈𝑀!,!=0. Most importantly, the interaction terms of one-year lagged monetary policy rate and dummy variable (e.g.
𝛾!𝐿1. 𝐼𝑛𝑡𝑒𝑟𝑒𝑠𝑡𝑅! !
!!! ∙ 𝐷𝑈𝑀!,!) are used to explain the different implications of monetary policy for industries on leverage. In total, there are 7 coefficient
estimators (𝛾!, 𝛿!, 𝜃!) for each monetary policy rate. 𝜀!,! is the error term.
Firstly, the monetary policy measurement is included solely regarding policy lending interest rate (model 3&7), deposit reserve ratio (model 4&8) and inflation rate (model 5&9). Secondly, the three monetary policy rates are estimated in the model simultaneously (model 6&10). Therefore, there are 8 different models with short-term and long-term leverage, respectively. Besides, these models are estimated by only using fixed effect technique. To avoid collinearity, the terms with monetary policy factors alone, are dropped. The results are shown in Table 5.
7. Results and Analysis
Table 4 First regression results of leverage ratios on firm-specific factors and one-year lagged monetary policy measurements from 2010 to 2015, regardless of industry classification. Fixed effect technique is chosen due to F-test and Hausman test reject the Pooled OLS and Random effect. STD=short-term debt/Total asset, LTD=long-term debt/Total asset. Observations 5,185 Number of firms 1,037 STD LTD9 Model 1 Model 2Variable Pooled OLS Fixed Effect Random Effect Pooled OLS Fixed Effect Random Effect Size -0.0519*** -0.176*** -0.0799*** 0.0571*** 0.106*** 0.0656***
(0.00798) (0.0288) (0.0132) (0.00238) (0.00850) (0.00395)
Borrcap 0.0188 -0.133*** -0.0406 0.0814*** 0.137*** 0.104*** (0.0231) (0.0510) (0.0335) (0.00688) (0.0151) (0.00999) Pro -2.707*** -2.143*** -2.285*** -0.268*** -0.164*** -0.178*** (0.0635) (0.0590) (0.0551) (0.0189) (0.0174) (0.0163) Gro 0.000727 0.00735** 0.00396 0.00266** -4.66e-05 0.00142* (0.00382) (0.00290) (0.00270) (0.00114) (0.000857) (0.000799) L1.InterestR 7.253* 2.511 4.864* 3.709*** 4.094*** 3.426*** (3.973) (2.589) (2.531) (1.184) (0.765) (0.748) L1.ReserveR -2.828*** -0.928 -1.922*** -0.797*** -1.027*** -0.726*** (0.833) (0.585) (0.540) (0.248) (0.173) (0.160) L1.Inflation 0.332 -0.182 0.144 0.0589 0.183** 0.0642 (0.455) (0.299) (0.290) (0.136) (0.0882) (0.0857) Constant 0.699*** 1.395*** 0.840*** -0.316*** -0.611*** -0.374*** (0.123) (0.178) (0.105) (0.0367) (0.0526) (0.0311) 𝑅! 0.280 0.267 0.140 0.077 F-value 8.68 8.90 Prob>F 0.0000 0.0000 Standard errors in parentheses *** Significant at 1% level, ** Significant at 5% level, * Significant at 10% level. Hausman Test 𝐻!: 𝑑𝑖𝑓𝑓𝑒𝑟𝑒𝑛𝑐𝑒 𝑖𝑛 𝑐𝑜𝑒𝑓𝑓𝑖𝑐𝑖𝑒𝑛𝑡𝑠 𝑛𝑜𝑡 𝑠𝑦𝑠𝑡𝑒𝑚𝑎𝑡𝑖𝑐 𝑥!-value 61.11 61.60 𝑃𝑟𝑜𝑏 > 𝑥! 0.0000 0.0000
7.1. Firm-specific determinant
Table 4 reports the first regression results toward the alternatives leverage ratios by Pooled OLS, fixed effect, and random effect, respectively. Model 1 contains STD and Model 2 with LTD. Before discussion about the determinants of listed Chinese firms’ capital structure, it is necessary to decide which technique is the most appropriate to describe the panel dataset. The p-values of F-test under both models are zero, indicating that the null hypothesis, individual fixed effects are jointly zero, is rejected. Therefore, comparing to pooled OLS, fixed effect is the better panel technique. Moreover, Hausman test is applied to compare fixed and random effect techniques. The result of the test shows that the null hypothesis is rejected due to the zero p-value, which means that simply ignore the unobserved individual fixed effect may cause omitted variable bias. Therefore, the fixed effect technique is the most appropriate one, and it is finally selected to explain the results.
Size
Hypothesis 1 is rejected by the results from Table 4. For both models, size (natural logarithm of the total asset) is significant at the 1% level, which illustrates that the size of the firm does affect the leverage so that it is a determinant of the capital structure. However, the directions are distinct. On the one hand, there is a negative relation between firm size and short-term leverage; one unit increases of size leads to 0.176 units of short-term debt ratio decreases. One the other hand, corporate with bigger size seems to have higher long-term leverage level. By enhancing of one unit of size followed by 0.106 units increase of long-term debt ratio. The latter result supports the idea in the most previous paper that the larger companies tend to borrow more due to their relatively low costs of financial distress. Contrast to the results from Wen (unknown), who stated that larger firms have easier access to both long-term and short-term debt;
this paper proves that the long-term debt is more favored instead of short-term debt by larger companies, and smaller firms use more short-term debt financing. The observation is consistent with the results of Marsh (1982) that the size is negatively related to short-term debt but positively correlated with long-term debt. The reason that larger firms are still using more long-term debt to finance instead of internal funding, which is different from the idea of Rajan and Zingales (1995), may be that many larger companies are wholly or partly controlled by the government in China and thus are more likely to be sustained by the banking system than companies from other states. Borrowing capacity
The impact of borrowing capacity (the ratio of the tangible fixed asset to total asset) on firm’s capital structure is similar to that of the size. Borrowing capacity is another essential determinant that is negatively related to the short-term leverage and positively associated with the long-term debt ratio; both relations are significant at 1% level. The bigger amount of tangible fixed asset, the higher level of long-term leverage. It is reasonable that creditors would like to lend more once they are securely guaranteed by sufficient collateral. By increasing one unit of the tangible fixed asset, the long-term debt raises by 0.137 units. The surprise goes to the negative relation between borrowing capacity and short-term leverage. One may ask why the firms with less tangible fixed assets can still reach higher short-term debt ratio? Berger and Udell (1994) believed that in the environment of the Chinese market, the relationship between creditors and firms is somehow more important than the physical collateral. The tangible fixed asset (collateral) can be substituted by the good relationship between both parties when the firm is asking for debt, especially for the short-term debt. It is then acceptable that the relation between short-term debt and borrowing capacity is negative, 0.133 units of short-term debt reduce when
the tangible fixed asset goes up by one unit.
Therefore, the second hypothesis is rejected, as there is no positive relation between short-term leverage and firm borrowing capacity.
Profitability
Consistent with the hypothesis 3, the profitability (ratio of EBIT to total assets) of the firm is negatively correlated with leverage levels concerning both short-term and long-term debt at 1% significant level. Profitability is regarded as another determinant of capital structure. The results consistent with the ideas from Booth et al. (2001) and Chen (2004), that when firms obtain higher profit, they prefer to be financed by retained earning rather than external debt. Noting that the short-term debt shows greater sensitivity towards profitability than long-term leverage, which a unit raises of the profitability, the short-term leverage decreases by 2.143 units. The reason is that the short-term debt is typically required for financing the projects with the smaller amount of funds needed, and these funds are more likely to be replaced by the current retained earnings. Moreover, on average, the Chinese company uses more short-term debt financing can be another reason for the significant sensitivity.
Growth potential
The results from Table 4 only show a significant positive relation between growth potential (percentage change of total assets) and short-term leverage at 5% level, for the long-term debt, this relation is negative but insignificant. The insignificance of the coefficient is also proved in other papers (See Chen et.al 2014 and Tong 1999). According to Wen (unknown), the insignificance may come from the regulation of initial public offerings (IPO) in China. That is, the firm’s profitability is the one that can mostly decide whether a firm can go public or not, instead of its growth potential or market share, making that the growth
potential is less important than other determinants for the Chinese company. Even though there exists a significant relation between short-term debt and growth, the coefficient is relatively small compared to other factors. By inducing one unit of the growth rate, the short-term debt level only increases by 0.007 units. Thus, hypothesis 4 is rejected due to that the results show neither robust positive nor significant relation between firm’s growth potential and long-term leverage.
7.2. Monetary policy determinant
Rest of this section depicts the implication of monetary policies on the capital structure of listed Chinese firms, regardless of industry classification; the effects show heterogeneity among three monetary policies. Regards to the Model 1, all monetary policy measurements are not significantly related to the short-term leverage. The reason for the above phenomenon falls into the uniqueness of Chinese financial market. As explained before, firms with less tangible fixed assets may still borrow more from the bank due to they possess the good relationship with the bank, especially for the short-term loan. Similarly, the adjustment of monetary policy by PBoC may not affect the use of short-term debt financing by firms. In other words, the financing decision of whether or not to borrow short-term loan is independent of the general monetary policy in China. The results from Table 4 show significant relations between long-term leverage ratio and three monetary policy variables, which are interpreted in detail in this section. Policy Lending Interest Rate
Interest rate channel tells that the hiking of interest rate leads to the costs for external debt financing grow, which accordingly drives down the amount of debt
proportion employed by the firm. However, the results in Table 4 picture that there is a positive relation between long-term debt and lagged policy lending interest rate, which is significant at 1% level. Even though the lending interest rate goes up, firms are still willing to be financed by long-term debt issued by the bank. Therefore, hypothesis 5 is rejected. Increasing of the policy lending interest rate by one unit leads to a 4.09 units improvement of long-term debt ratio. The result supports the paper of Wen (unknown), who used the data from 1998 to 2007 and stated that the cost of borrowing is not a major concern of Chinese companies.
There are five potential reasons for the positive relation in China. Most importantly, the reverse causality may still exist even though the monetary policy indicators are one-year lagged. The PBoC raises lending interest rate when the current or the predicted leverage is high in order to cool down the economy and lower the leverage, which causes the positive relationship. Secondly, borrowing from the bank is still the priority due to the costs of issuing long-term corporate bonds are relatively high and continue increasing, which makes the financing by bank loans still attractive, especially for those large and credible corporates (Zhou, 2017). Thirdly, the interest rate floating mechanism in China makes the commercial banks decide their actual lending interest rate. Public firms in China are easier to receive the lower lending interest rate from the commercial bank that is close to the policy rate, than medium and small enterprises, which have to bear the higher floating interest rate. Thus, the raise of interest rate is not a serious obstacle for public companies in case they need financing (Yu, 2017). Fourthly, the market reaction of the interest rate adjustment may lag longer than expected. The model in this paper includes the one-year lagged monetary policy variables; yet, it is possible that the response by financial market takes more years, particularly for borrowing the long-term
debt10. Moreover, the transmission channel is immature in China that may also
lead to the late response to the interest rate adjustment. Lastly, Mokhova and Zinecker (2014) stated that the company increases their long-term debt with interest rate is due to the tax benefit.
Deposit Reserve Ratio
The implication of lagged deposit reserve ratio on leverage level is as assumed by the hypothesis 6. There is a significant negative relation between firm’s long-term leverage and reserve ratio. With the stricter reserve ratio requirement faced by the commercial bank, the economy is tightened up by monetary austerity; consequently, companies receive less amount of debt financing. A unit higher of one-year lagged deposit reserve ratio causes 1.027 units of long-term leverage decreases. Inflation Rate Hypothesis 7 is rejected due to that a unit rises in one-year lagged inflation rate indicates approximately 0.183 units of long-term leverage increases in capital structure, which is significant at 5% level. Thus, the inflation rate is another determinant of the capital structure. Obviously, the listed Chinese firms take advantage of inflation to borrow more long-term debt. Higher inflation is a benefit to the borrower due to the relatively less repayment in the future. Moreover, the benefit grows larger with the longer maturity of the loan, which is reflected by the significant positive relation between lagged inflation and long-term debt. On the one hand, companies increase their leverage level if the inflation is under control (Lima et. al., 2011). On the contrary, by facing the potential loss, the commercial banks may reluctant to issue more loans under a high inflation environment. These compensated effects lead to the low elasticity
10 By taking the three-year lagged variable into account, the coefficient of lending interest rate is negative
