Political connection and bank lines of credit. Danglun Luo Qianwei Ying Lingnan College, Sun Yat-sen University, Guangzhou, China

Transcription

1 Political connection and bank lines of credit Danglun Luo Qianwei Ying Lingnan College, Sun Yat-sen University, Guangzhou, China Corresponding author. Address: Lingnan College, Sun Yat-sen University, 135 Xingang Xi Road, Guangzhou, China, Tel: (+8620) ; (+86) Author Details: Danglun Luo, Ph.D., Associate Professor, Lingnan College, Sun Yat-sen University, Guangzhou, China, Qianwei Ying (Corresponding Author), Ph.D., Assistant Professor, Lingnan College, Sun Yat-sen University, 135 Xingang Xi Road, Guangzhou, China, Tel: (+8620) ; (+86)

2 Political connection and bank lines of credit Abstract: As a low cost liquidity management tool, bank lines of credit are playing an important role in firms financing and operating activities. However, not all the firms are granted bank lines of credit. We analyzed the samples of companies listed in Shanghai and Shenzhen Stock Exchange in China from 2004 to 2009 and discovered that the political connections of a firm would help it obtain bank lines of credit, especially from state-own banks. Meanwhile, we discovered that for firms facing more financing constraints, non state-owned firms and firms in regions with intensive government intervention, political connections have a stronger impact on the acquisition of bank lines of credit. This paper not only deepens our understanding of bank lines of credit, but also deepens the understanding of the role political connections playing in firms financing activities. Keyword: Political connection, bank lines of credit, financial constraint 1. Introduction The use of bank credit lines plays an important role in corporate liquidity management. More than 70% of bank loans granted by major international commercial banks are drawn under certain credit lines. Sufi (2009) discovered that from 1996 to 2003, 85% of the firms he studied had been granted credit commitments, and the average amount of bank lines of credit accounted for 16% of their average total assets. Jimenez et al (2009) revealed that in Spain the amount of credit commitments took up 32% of the amount of loans recently granted by banks, and 42% of the amount firms borrowed from banks. In China, businesses in credit lines began to rise only in recent years, but they are growing at a tremendous speed. Taking the samples of the listed companies in the range of our study for example, the percentage of companies granted credit lines had risen from 16.43% in 2004 to 25.47% in 2009, with the amount granted from 97.6 billion yuan to billion. Considering the importance of bank credit commitments in corporate liquidity management, issues on how firms obtain credit lines is then worth further study. Jimenez et al.(2009) discovered that characteristics of firms, such as total assets, return on net assets and the company risks are major determinants of the acquisition of credit commitments. Sufi (2009) found out that factors like EBIT, cash flow and net assets are significantly correlated to credit commitments firms obtain. However, their research mainly focuses on the fundamentals of companies, without exploring into the effects of political connections. In the institutional context of China, this essay studies the effects of the political connections of firms on their acquisition of credit lines. More detailed, the following issues are discussed: whether political connections could help firms in obtaining credit lines; how state-own and private banks treat firms political connections; and what other factors would affect credit lines offering to firms with political connections. We analyzed the data during the period of 2004 and 2009 of the companies listed in the A Share market in China, and we discovered that with political connections firms are easier to obtain credit lines, especially credit lines from state-own banks. Moreover, we found out that for non state-owned firms, firms with strict financing constraint and firms in places with a high level of government intervention, political connections are more important in helping

3 firms obtain credit lines This paper makes the following contributions. Firstly, it extends the field of research on the effects of political connections on corporate financing activities. There has been some literature on this issue. For example, Khwaja and Mian (2005) found out that firms with political connections are able to get more loans and preferential interest rate. They pointed out that in places where the government has strong incentives in intervening banks credit lines offering decisions, establishing political connections help reduces financing costs of firms. Charumilind et al (2006) carried out additional analysis from the perspective of debt maturity structure. They pointed out that political connections help firms obtain more long-term loans because political connections are some sort of guaranteed collateral for banks, and obtaining more long-term loans reduces the costs of establishing political connections for firms. Faccio et al (2007) considered banks expectation that soft financing constrains exist in firms with political connections, the key factor that makes firms with political connections chooses higher debt ratio. His research indicated that firms with political connections are easier to get government subsidies or policy support in period of economic downturn. Claessens et al (2008) found out that firms that establish political connections by donation are able to obtain more loans after each election, which indicates that an important way to obtain credit support is to establish political connections. These essays, however, concern more about the indicators on balance sheet, including long-term or short-term debt and etc, without caring too much about the off balance sheet items. Since credit line is a very important off-sheet financing method that plays a more and more important role in corporate financing activities, effectively strengthening the financing ability of firms and shaping the financing behaviors of firms, reconsidering the effects of political connections and corporate financing activities will from the perspective of an off-sheet item--credit lines would provide some meaningful evidence. Secondly, the paper extends the field of research on bank credit lines. Previous research focused on the function of credit lines in operation and risk control from banks perspective (Kashyap et al.,2002;gatev and Strahan,2006). Recent research concerns more about the effects of the fundamental characteristics of firms on their acquisition of bank credit lines and the related liquidity management. (Saidenberg and Strahan,1999; Agarwal et al.,2004; Sufi,2009). Jimenez et al (2009) discovered that characteristics of firms, such as total assets, return on net assets and the company risks are major determinants of the acquisition of credit lines. Sufi (2009) found out that factors like EBIT, cash flow and net assets are significantly correlated to credit lines that firms obtain. The discoveries revealed by the above two paper basically come from the analysis of the qualifications of firms. They did not reveal the influence that firms political connections have on the acquisition of bank credit lines. In this paper we will make supplement to this topic, extending the research area on bank credit lines. Thirdly, we established a firm-level political-connection-index to study the impact of political connections on bank credit line offerings. Previous literature on political connections often used dummy variables to indicate whether firm executives had political connections or not. Faccio (2006) for example, regarded firms with executives or controlling shareholder serving as congressmen, prime minister and other government executives or have close relationship with

4 some political parties as politically connected. Betrand (2004) treated those CEOs graduating from elite schools in France, serving or used to serve as government executives in French government as executives with political connections. Direct though, this depiction blurs the distinction among various hierarchical levels. As far as China is concerned, different hierarchical levels in government means differences in political resources and influences. If there are multiple executives with political background in a company, a simple dummy variable on whether the firm has political connection or not is obviously not a good measure of the firms relative political power comparing to others. Therefore, a more detailed political-connection-index is established in this paper to measure firms relative power of political connections. Here is the roadmap of this paper: Section 2 analyses the institutional background and propose the hypotheses, section 3 sets out the model, section 4 provides the empirical results, and section 5 concludes and gives remarks. 2. Institutional background and hypothesis development Bridging the gap between savings and investment, the services provided by financial institutions is the most fundamental catalyst in the development of a country s economy( Rajan and Zingales, 1998 ). In a typical transitional economy, the resource-allocation of the banking system is crucial to the development of the economy as a whole, given that its capital market is less developed. The effects of the role of banks as financial intermediations in promoting economic growth depend on the efficiency of bank credit capital allocation. Because of ownership discrimination and size discrimination, bank credit capital allocation is operated inefficiently in China (Brandt and Li, 2003). Of course, with the reform of the banking industry in recent years, China's bank credit capital allocation was improved. Firth et al. (2009) pointed out that banks will seriously consider the past financial performance and the agency cost of the firms when allocating credit resources. However, due to the underdevelopment of the product and credit markets and the lack of supportive institutional mechanism, the government often impose great control on the allocation of economic resources (McMillan, 1997), which makes it difficult for many companies to obtain economic resources through market system. First of all, when the banks are state-owned or state controlled, which is the case for the major banks in China, the appointment of the executives in these banks are decided by the government, who will thus also control or affect the key credit decisions of the banks. Secondly, the local government officials have strong incentives to intervene with the allocation of credit resources to stimulate local economic growth, because local economic growth is still the main indicator of assessment for the local government officials performance which is closely related to their promotion. Therefore, the local governments is still to a certain extent, affecting credit decisions in China. The most direct constraint in the development process of a firm is the financial constraint. With political connections, firms improve the relationship with governments and therefore improve the relationship with the state-owned banks, which facilitates firms financing activities. Leuz and Oberholzer (2005) discovered that political connections have an impact on firms financing decision, and that it is easier for firms with government background to obtain bank loans in

5 Indonesia. Charumilind et al. (2006) found out that empirical evidence in Thailand supports the hypothesis that political connections make it easier for firms to obtain long-term loans. Claessens et al. (2008) analyzed the data of Brazil and pointed that firms would establish political connections by donating in federal elections and firms that have established political connections would increase their loans after elections, indicating that political connections have an effect on obtaining bank loans. Fan et al. (2007) studied the debt financing activities of the listed companies related to 23 government executives after they felt a cropper. They found out that losing political connections result in lower ability in getting bank loans. The above analysis shows that the firms political connection can facilitate their access to financing. Since bank credit lines is also a valuable financial resources which is not easy to obtain for most of the firms in China, we expect that the same logic of the above analysis should also be applicable to the relationship between political connection and bank credit lines, i.e., firms with stronger political connection tend to obtain bank credit lines more easily 1. Especially, political connection should be more helpful for firms to get bank credit lines from state-owned banks which are controlled by the government and thus are more likely to be influenced by political connections on their decisions of granting loans and bank credit lines. Firms with smaller size and poorer credit rating are expected to be faced with severer financing constraints (Fazzari, 1988; Sufi, 2009). These firms are usually in greater need of bank credit lines but are less likely to obtain them. Political connection will be more helpful for these firms to get access to bank credit lines. We can conclude the above analysis as the following two hypotheses: H1: When seeking bank credit lines, stronger political connections would help a firm obtain them, especially from state owned banks. H2: Among the group of enterprises with relatively small size and poor credit position, there is a more significant positive correlation between political connections and bank credit lines. The four major state-own banks are leading the financial markets of China. They possess more than 70% of the credit funds, dominating the credit market. Although state-own banks are 1 There is a good real example to illustrate how government intervention can affect the offerings of bank lines of credit: In 2005, Sichuan Changhong Group Corp., one of China s largest electronic manufacturers, announced a huge loss of billion yuan in the previous year. Just after the announcement, the Sichuan branch of the Agricultural Bank of China signed a cooperation contract with Sichuan Changhong Group to provide it a huge number of credit lines amounting to 3 billion yuan. The vice secretary of Sichuan provincial committee, GAN Daoming attended the signing ceremony, and offered a intriguing speech. He noted that Changhong made a significant contribution to Sichuan province, and the Sichuan government will strongly support and help Changhong with its future development. The governor of Sichuan branch of Agricultural Bank of China even said that Changhong had no bad credit record before and its announcement of revenue loss should be considered as a responsible behavior for shareholders. Being informed of the government s supportive attitudes towards Changhong,more banks were engaged in the offering credit lines to Changhong Group Corp. Finally, the four major state owned banks including Industrial and Commercial Bank of China, Agricultural Bank of China, Bank of China, and China Construction Bank provided a total number of 8.5 billion credit lines in In addition, other commercial banks, including China Merchants Bank, CITIC Industrial Bank, Industrial Bank, and China Everbright Bank are also actively involved in the cooperation with Changhong, granting it 4 billion lines of credit in total in The details can be seen at

6 adopting economic principles in making credit decisions as the reform of the banking industry goes deeper, the government, as the ultimate controller of those banks, is still dominating the allocation of credit resources. As a result, credit resources are overmuch allocated to state-own enterprises whereas non state-owned firms endure difficulties in getting support from banks (Allen et al.,2005). Additionally, a lot literature show that obvious credit discrimination over non state-owned firms exits in China (Brandt and Li,2003;Gordon and Li,2003;Boyreau-Debray and Wei, 2005). The discrimination over non state-owned firms is a result of the restriction of the banks semi-public-financial system under excessive administrative intervention. For example, with regard to the same amount of bad loans, bank staff would not be or would not be seriously punished if the borrower is a state-own enterprise but would be held responsible or even prosecuted if the firm is a private one. Afraid of punishment, many bank staff would rather lend less or nothing to non state-owned firms in order to strictly control the total amount credited to non state-owned firms. In the meantime, the cumbersome procedure of getting bank loans, the scathing restriction on collateral, and the low mortgage rate let many non state-owned firms flinch. La Porta et al. (2004) pointed out that China s performance in protecting political freedom and Intellectual property is one of the worst in the world 2, which means that enterprises would rather integrate interpersonal networking into their strategies instead of acquiring resources through the market system or establishing strategic alliance (Choi,Lee and Kim, 1999). Xin and Pearce (1996) discovered in their research that non-state owned enterprises pay more attention to connections and allocate more resources to establishing connections than state-own firms in order to get protection that they cannot get from the legal system. Both Chen et al (2005) and Bai et al (2006) found out that since legal protection is low in China, the political connections would still protect enterprises from being encroached by the authority. When politics dominates resource-allocation, enterprises would adopt some informal mechanisms to support their development under various political constraints (Allen et al., 2005). Political connection brought by getting involved in politics is one of those mechanisms. Du and Girma (2007) studied the behavior that non-state owned enterprises obtain political asylum and thus support from local or even higher level authority by adopting red-hat strategy. Political connections of non state-owned firms can greatly improve the surviving prospects of them, facilitating their development. This effect is more marked in capital-intensive firms that demand more financial resources. Non-stated owned firms with political connections are easier to get financed (Li et al., 2008), including bank lines of credit. Comparatively, the effect of political connections on lines of credit should be less significant for state-owned enterprises. Accordingly, we put forward the hypothesis in this paper: H3: Political connections are more helpful for non state-owned firms in obtaining credit lines than for state-own firms. 2 According to the ranking of economic freedom index jointly constructed by The Heritage Foundation and the Wall Street Journal, based on 50 economic variables (including some indicators relating to corruption in the judiciary, the rule of law, and enforcement of contracts) in 2012, the United States ranked No.10 among 179 countries, while China only ranked No Please refer to the / research / features / index / search.cfm for the details.

7 Boubakri (2006) discovered that with more concentrated political power and shorter the term of office, it is more likely that the authority would appoint some government officials as executives in privatized firms. In the mean time, in countries where the judicial system is less independent, firms are more likely to get involved in politics. Studies have shown that in countries or regions where financial industry is less developed, government officials are more corrupted, the legal system is poorer, or the protection of property rights is weaker, non state-owned firms are more likely to establish political connections (Chen et al.,2005;faccio, 2006;Li et al.,2006). The results mean that in regions where institutional environment is good, non state-owned firms can rely more on market principles to obtain bank credit; however, in regions with worse institutional environment, since it is difficult for them to obtain resources following market principles, firms seek for political connections in order to get bank credit or credit lines. In economies with more government intervention, market plays a less important role in allocating credit resources and the allocation is more likely to be controlled by government officials. In such economies, political connections role in helping non state-owned firms obtain bank credit lines will be more important than that in an economy where the level of marketization is high and the level of government intervention is low. Therefore we put forward the fourth hypothesis. H4: In regions with more government intervention, political connections role in helping non state-owned firms obtain bank credit is more important. 3. Empirical study design 1) Data sources and description a) Collecting and processing of the data on political connections The data on firms political connections is manually sorted out from the record of the backgrounds of listed companies executives in Wind database. Wind database records the background information of the senior management group of all companies listed in the Shanghai Stock Exchange and the Shenzhen Stock Exchange, including information on the positions the executives once held in the government or on other important positions. Based on the information, we sorted out the political connections of the senior executives of different companies, and calculated the strength of their political connections. Details are as follow. The political background of the executives can be classified into the following categories. In the first category are people who once served as NPC representatives, CPPCC members or once held other positions that have relatively strong influence on politics. The second category consists of people who used to serve in the central government. In the third one are people who once served in provincial governments. Those who used to work in other local government institutions are classified into the fourth category. There is a strict division of the administrative level in China s administration system, ranging from ordinary clerks and clerks to the state-level officials. With regard to the executives in the second to fourth categories we score the strength of the political

8 connections according to the administrative levels they used to serve. Details are shown as follow. Vice-ministerial level and higher: 7 points; departmental level: 6 points; deputy departmental level: 5; division level: 4; deputy division level: 3; section level: 2 points; deputy section level and below, 1 point. As for those who used to serve as Representatives of the People s Congress and members of the CPPCC, detailed scores are as follow. National level: 6 points; provincial level 4 points; municipal and below: 2 points. With the scores of the strength of the political background of each executive, we establish the political-connection-index of each company. We first sum up the scores of the political background of the executives of each company and thus get the executive-political-connection-index of each firm. Secondly, taking into account the fact that the CEOs (presidents or general managers) and the chairmen of the boards play a major role in company decision-making, we specifically sum up the scores of CEOs and the board chairmen to get the political-connection-index of the top decision-makers. Meanwhile, we standardized the index, making the highest score the value of 1, and the rest are adjusted according to the proportion. The sample interval is from 2004 to b) Data of bank credit lines We collected the detailed data of bank credit lines from the bank credit database of RESSET, which keeps records of the details of each credit commitment, including the amount and the specific sources disclosed in the bulletins of all listed companies. We processed the data from the database manually, and classified the bank credit lines into two categories, bank credit lines from state-own banks and those from non-state-owned banks, according to the nature of the ownership of the credit granting banks. The sample interval is from 2004 to c) Data of credit rating Data on corporate credit rating in this paper comes from the China Corporate Risk Index database (CCRI). The risk index system provides two major indicators of risk rating. One is established according to the firm s profitability, security, corporate governance and other key financial statements; the other one is the threshold credit rating set up according to the threshold of the revenue size, and the stability of controlling right and business operation. These two indicators are highly correlated, and the higher their scores are, the higher is the firm s credit risk. d) Data of other control variables The remaining control variables include firm size, firm age, growth, asset-liability ratio, return on assets, and cash flow from operations, net tangible assets ratio and the ultimate controller s nature of ownership. The moderate variables include the nature of ownership, and the local government s degree of intervention. Data on the local government s degree of intervention is adopted from the

9 Report of Marketization Process in China by Fan et al. (2010). Data of the ultimate controller s nature of ownership is adopted from CCERDATA, and the remaining data is from REESET. Detailed definitions of the variables used in this paper are listed in Table 1. 2) Sample selection and modeling The samples in this paper are all the listed companies in the A-share Market in Shanghai Stock Exchange and Shenzhen Stock Exchange during 2004 and In accordance with convention, we removed samples of financial companies, and finally got 8809 samples in total. We classified the companies by two-digit industry code according to the CSRC Industry Classification Standard, into 33 industry categories. In order to overcome the impact of outliers, we have winsorized the main variables at 1% level. The basic empirical model we adopted in this paper is shown in model (1). J Creditline = a + a PC + a Pr ivate + β Control + Fixed effects + e (1) it 0 1 it 2 it j jit 1 it j = 1 In this model, Creditline represents the bank credit lines that firms obtain from banks, which is measured by three index, Cline (Whether firms obtain bank credit or not), Cliner (the ratio of bank credit to total asset) and Scline (the proportion of bank credit lines obtained from state-own banks to total bank credit lines). PC represents the political background of firms. We use the overall political-connection-index described in the preceding text for our main empirical test, and use the political-connection-index of firms top executives (chairmen of the board or CEOs) for robustness test. Private is a dummy variable characterizing the nature of the ownership of the firm, assigned value 1 if the ultimate controlling party is non state-owned, or 0 otherwise. Control represents other control variables, including net tangible assets ratio, return on assets, revenue growth, operating cash flow, firm size, debt ratio, firm age and corporate credit rating according to earlier studies (e.g. Sufi, 2009; Jimenez and Lopez, 2009). In order to overcome the endogeneity and simultaneity problem to some extent, we adopted values of last period for all control variables. Additionally, we controlled the fixed effects of industry, year and province that the firm locates. To further study the impact that political connections have on the acquisition of bank lines of credit in different ownership structure and regional market environment, we adopted a further model shown as follows: Creditline = a + a PC + a Pr ivate + a Highgov + a PC Highgov it 0 1 it 2 it 3 it 4 it it J + a5pc Pr ivate + β Control + Fixed effects + e (2) it it j jit it j = 1 In model (2), Highgov is a dummy variable that characterize the level of government intervention of the region where the firm locates. If the level of government intervention excesses the median,

10 we assigned it value 1, or else 0. The coefficient in front of the cross term, represents whether political connections have a stronger effect in regions with higher level of government intervention. The coefficient before the cross term indicates whether there is a stronger effect of political connections in helping non state-owned firms obtain bank credit. 4. Empirical results and analysis 1) Descriptive statistics Table 2 gives yearly statistics regarding bank lines of credit and firms political connections from 2004 to Table 2:Yearly average for bank lines of credit and firms political connections Average Total bank lines of credit of all listed companies (Billion Yuan) Total bank lines of credit of all listed companies /Total bank loans at the end of the year of listed companies Proportion of companies granted bank lines of credit Average ratio of bank credit lines to bank loans for each individual firm granted credit lines Proportion of bank credit line offered by state-own banks Proportion of companies having senior executives with political connections Average senior executive political-connection score Proportion of companies having executives with political connections 16.4% 13.3% 15.1% 20.1% 23.7% 25.5% 19.4% % 51.4% 46.0% 51.5% 48.4% 48.4% 48.8% 57.4% 57.6% 56.1% 55.8% 56.1% 52.3% 55.7% % 23.1% 23.2% 23.9% 25.2% 23.5% 23.7% Average executive political-connection score It can be seen from the ratio of total bank credit lines to total bank loans for all listed companies that the bank credit lines financing is becoming increasingly important year by year, reaching 578 billion Yuan by the end of 2009, which is around 1/5 of the total bank loan of all listed companies. The average ratio of bank credit lines to bank loans for each individual company granted lines also increased year by year, surpassing 0.5 by the end of Although the proportion of companies granted bank lines of credit has gradually increased after the year 2006, it is still at a low level afterall, only accounting for about 1/4 of all listed companies. This shows that only a small part of the firms are able to obtain bank credit lines although it is an important external source of liquidity. This shows that for most businesses, bank lines of credit is a kind of scarce financial resource, and that the study on what kind of firms acquire this scarce resource is a meaningful issue with practical emphasis. Looking into the sources of bank lines of credit, state-own banks accounted for

11 half of the sources and non-state-own ones accounted for the other half. In the sample of all listed companies, senior executives (including chairmen, general managers, deputy general managers, the CFOs, board members, chief accountants, and etc.) with political background account for around 56% of the population, far higher than the proportion of firms granted bank credit lines. Although in this article we expect that political connections would help firms obtain bank credit, merely having political connections cannot guarantee superiority in obtaining bank credit because of the existence of fierce competition between these politically connected firms looking for bank credit lines. Therefore we further measured the strengths of the political background of each firm. According to the statistical results in Table 2, if the value of the political-connection-index of the firm with the strongest political background is normalized to 1, the average value of the political-connection-index of listed is about 0.08, indicating that there is a great variance in the political background of executives in listed companies. Meanwhile, we also adopted the common practice of literature in the field of the political connections, using the method whether corporate top executives (CEOs or chairmen of the board only) have political background or not to portray the political connections of the firms. As shown in table 2, in the sample of listed companies, companies having top executives with political connections accounted for around 24%. The average value of the political-connection-index is about 0.1, which is 10% the level of firms with strongest political background. This shows that the strength coefficient of corporate leaders political connections various tremendously. Table 3:Descriptive statistics for main variables Variable Sample size Minimum Maximum Average Standard Error Medium Cliner Cline PC Chairceopc Private Gov Highgov Brating Thrating Age SIZE Tang LEV ROA Growth Cashflow The overall descriptive statistics of the main variables used in this article are shown in Table 3. In order to overcome the influence of outliers in the financial data, the lowest and highest 1% values of the variables are winsorized. 2) Regression analysis a) Political connections and the acquirement of bank lines of credit

12 We first use the overall corporate political connection index based on the political background of all corporate executives to characterize the influence of political background of firms on the acquisition of bank lines of credit. The results are shown in table 4. Table 4: Regression results of the effect of political connection on bank credit lines (1) (2) (3) (4) (5) (6) Cline Cliner Scline Cline Cliner Scline PC *** 0.278*** *** 0.287*** (0.90) (5.57) (3.22) (1.62) (5.32) (3.39) Private * (0.68) (0.95) (0.39) (0.15) (1.84) (-0.23) Brating *** ** (-2.99) (-2.15) (-1.19) ROA 1.215*** 0.207* (4.73) (1.95) (-1.45) Growth 0.063** (2.04) (-1.19) (-1.02) Tang (1.41) (0.97) (-0.94) Cashflow *** (-4.33) (-1.23) (1.11) Size * 0.039** (1.01) (-1.70) (2.38) LEV *** *** (-2.95) (-0.39) (-2.85) Age *** ** (-3.14) (-2.10) (-0.33) Constant *** ** *** (-0.60) (2.59) (-1.11) (-2.46) (-0.73) (4.12) Industry YES YES YES YES YES YES Year YES YES YES YES YES YES Province YES YES YES YES YES YES N Pseudo R2/R Note:*** ** * represents significance level of 1%, 5% and 10% respectively. The t-statistics, computed using the robust standard error clustered at firm level, are in parentheses. In model (1)-(3) in table 4, we included a group of control variables like the nature of ownership, return on assets, revenue growth, the ratio of tangible assets, internal cash flow, firm size, debt ratio, firm age, and etc. We also controlled dummy variables like industry, province and year. The dependent variable in Model (1) is whether the firm has access to bank lines of credit or not. Regression results from Probit model indicate that the overall relation between corporate political connection and bank lines of credit is positive but not significant. Firms with better profitability (with higher return on assets), higher growth rate in operating income, larger proportion of tangible assets and lower debt ratio are more likely to be granted bank lines of credit. This is

13 because with better profitability, higher growth, larger proportion of tangible assets and lower debt burden, firms possess stronger ability in repaying debt, and are thus safer for banks to grant them bank lines of credit. The results, to some extent, indicate that when banks in China are granting bank lines of credit, they basically keep to the market principles, acting to maximize their profits and minimize their risks. Additionally, there is a negative correlation between internal cash flow and the bank lines of credit, representing the substitution effect between the internal liquidity, i.e. the internal cash flow, and the external liquidity, i.e. the bank lines of credit in the practice of corporate liquidity management (Sufi, 2009). With more internal cash flow, firms rely less on bank credit lines. It should be noted, however, that we cannot conclude that political connections are of no use in helping firms obtain bank lines of credit for the possibility that some firms with strong political connections did not obtain bank lines of credit just because they do not need any. It is possible that they adopted some other cheap and effective method of financing. However, it is difficult to strictly sort out which firms are really in need of bank lines of credit while others are not. We thus run regressions on the sub-sample of firms which already obtained bank lines of credit in Model (2) and Model (3) to partially overcome this problem 3. In Model (2), the dependent variable is the proportion of bank lines of credit to total assets, and the regression result indicates that among the firms with bank lines of credit, those with stronger political connections obtained more. In Model (3) the ratio of bank lines of credit from state-own banks is used as dependent variable in the regression. The result reveals that for firms with stronger political background, the proportion of credit lines from state-own banks is higher. The results in Model (2) and Model (3) indicate that political connections play a role in helping firms obtain bank lines of credit, especially from state-own banks. In Model (4)-(6) basic corporate credit rating is used as control variable. Basic corporate credit rating is derived from the profitability of firms, security, corporate business activity, corporate governance and other financial statistics. The measurement of corporate size, age of the firm, asset-liability ration, profitability, growth and other indicators are included in this variable. Therefore, in the regression, no other control variables are included besides of political background, nature of property, credit rating, as well as dummy variables of industry, year and province. The results model (4)-(6) are basically the same as those of (1)-(3), indicating that for firms in need of bank lines of credit, their political background is indeed beneficial in helping them obtain bank lines of credit, especially from state-own banks. This provides empirical support to hypothesis 1. b) The impact of political connections on obtaining bank lines of credit under financial constraint A further regression analysis based on the corporate credit risk rating is conducted in table 5. The credit risk rating data adopted in this article is scored into 10 levels. With the increasing risk, the rating scores increase from 1 to 10. Score 10 indicates that the firm has once defaulted. According to the grading criteria in the data base, companies rated as 7 points or above are regarded as companies with high credit risk. In this article we classified them into the group with low credit rating. Generally speaking, in a free financial market, companies with lower credit rating are facing more difficulties in obtaining bank loans, bond financing and other borrowings. These firms, 3 Having bank lines of credit can be considered as an indicator that the firm needs the bank lines of credit.

14 therefore, are facing severer financing constraint, and thus are in greater need of bank lines of credit. Based on previous analysis, we expect that political connections are more helpful for firms with lower credit ratings. Firms with higher credit rating, on the other hand, are facing less financing constraints and are easier to obtain bank loans, bank lines of credit and other borrowings, so that they do not need to rely too much on political connections. Therefore, we expect that the effect of political connections on the acquisition of bank lines of credit would be less significant in the sample firms with higher credit ratings. The grouped regression results based on credit ratings shown in table 5 support the above hypothesis. From the coefficient in front of PC and its significance level, we can see that in the group where credit rating level is relatively low, firms with stronger political connections are more likely to obtain bank lines of credit, especially from state-own banks. While in the group with higher credit rating level, political background has no significant effects on obtaining bank lines of credit or the proportion of bank lines of credit from state-own banks accounts for. Table 5:Grouped regression results based on credit rating The group with lower credit rating The group with higher credit rating Cline Cliner Scline Cline Cliner Scline PC *** 0.449*** (1.60) (4.04) (2.91) (0.73) (-0.90) (0.08) Private (0.75) (0.84) (-0.05) (1.12) (1.04) (-0.22) ROA 1.145*** (3.28) (1.01) (-1.34) (-0.77) (-0.41) (0.75) Growth 0.099** (2.25) (0.40) (-1.13) (-0.31) (-0.64) (1.15) Tang 1.362* 0.310* * (1.88) (1.71) (-1.67) (-0.60) (-0.32) (0.91) Cashflow * ** *** ** (-1.92) (0.33) (2.24) (-3.33) (-1.40) (-2.12) Size 0.153*** *** ** (4.09) (-0.43) (1.00) (-2.87) (-0.99) (1.99) LEV *** * *** (-4.16) (-0.41) (-0.39) (1.96) (0.92) (-3.66) Age * ** ** (-1.41) (-1.80) (-2.46) (-1.08) (-0.44) (2.44) Constant *** * ** (-3.45) (-0.36) (1.12) (1.68) (1.42) (-2.36) Industry YES YES YES YES YES YES Year YES YES YES YES YES YES Province YES YES YES YES YES YES N Pseudo R2/R Note:*** ** * represents significance level of 1%, 5% and 10% respectively. The t-statistics, computed using the robust standard error clustered at firm level, are in parentheses.

15 Either in earlier literature (e.g. Fazzari et al, 1988) or recent literature (e.g. Sufi, 2009), firm size is regarded as an important criteria in distinguishing whether a firm is facing financing constraint or not. Therefore in this article we conduct further regression analysis according to the categorization based on firm size. The smallest one thirds of the firms are classified into the group with high financing constraint, while the biggest one thirds of the firms are classified into the group with low financing constraint. The regression results are shown in table 6. Table 6: Regression results based on classification of the firm size The group of firms with smaller scale The group of firms with larger scale Cline Cliner Scline Cline Cliner Scline PC 1.874*** 0.413*** 0.385*** (4.94) (6.39) (3.11) (-0.87) (0.73) (0.74) Private 0.273*** (2.84) (1.49) (0.32) (-0.17) (-1.14) (-1.33) ROA 1.561*** ** (3.79) (1.07) (-1.08) (0.77) (1.21) (-2.15) Growth (1.28) (-0.59) (-1.25) (1.56) (0.46) (-0.73) Tang 1.563** * (1.98) (0.72) (-1.91) (-0.05) (-0.03) (-0.21) Cashflow *** (-2.65) (0.15) (1.55) (-1.53) (0.30) (0.53) Size 0.599*** *** (6.39) (0.55) (1.04) (-3.28) (0.10) (0.87) LEV ** ** (-2.36) (-1.20) (-1.09) (1.04) (1.32) (-2.44) Age *** (-2.75) (-1.10) (-0.91) (0.66) (-0.75) (-0.01) Constant *** (-6.26) (-0.34) (-0.38) (1.33) (-0.36) (0.42) Industry YES YES YES YES YES YES Year YES YES YES YES YES YES Province YES YES YES YES YES YES N Pseudo R2/R Note:*** ** * represents significance level of 1%, 5% and 10% respectively. The t-statistics, computed using the robust standard error clustered at firm level, are in parentheses. From the above table we can see that in the group with smaller firm scale, firms with stronger political background are indeed more likely to obtain bank lines of credit, especially from state-own banks; in the group with larger firm size, however, there is no significant correlation between political connections and the acquisition of bank credit lines. The result further proves that when firms are facing financing constraint and are in need of bank lines of credit, their political background would indeed help them obtain bank lines of credit.

16 c) The impact of political connections on bank lines of credit with different ownership and local governance environment The next direct question is about whether political connections have a stronger effect in helping non state-owned firms obtain bank lines of credit than in helping state-owned ones. Also, in places where the level of government intervention is relatively high and thus the government is relatively more powerful in influencing resource-allocation, will political connections be more helpful? In order to answer the above two questions, we select only firms with relatively poor credit rating or small firm size as regression samples. This is to ensure that the sample companies are facing financing constraint and are thus in urgent need of bank lines of credit. In this way we eliminate the interference of sample firms that could easily obtain bank lines of credit but do not need any, so that the impact of political connections is better reflected. The related results are shown in table 7. Table 7:Ownership of the firms, level of local government intervention and the financing effect of political background whether granted bank credit lines or not the ratio of credit lines to total asset (sample firms with lines of credit) (1) (2) (3) (4) (5) (6) PC (0.02) (-1.17) (-1.57) (-0.20) (-0.26) (-1.34) Private (-0.17) (1.42) (0.30) (-1.06) (0.75) (-0.78) ROA 1.387*** 1.384*** 1.393*** 0.194* 0.186* 0.199* (4.35) (4.33) (4.34) (1.74) (1.69) (1.80) Growth 0.080** 0.073** 0.076** (2.13) (1.97) (2.02) (0.40) (0.15) (0.35) Tang 1.305** 1.385** 1.330** (2.14) (2.24) (2.15) (1.21) (1.57) (1.24) Cashflow *** *** *** (-3.51) (-3.34) (-3.35) (-0.37) (-0.07) (-0.26) Size 0.144*** 0.154*** 0.157*** (4.47) (4.86) (4.95) (-0.40) (-0.34) (-0.26) LEV *** *** *** * (-3.99) (-4.12) (-4.08) (-1.62) (-1.61) (-1.68) Age *** *** *** ** ** ** (-3.01) (-2.77) (-2.82) (-2.54) (-2.21) (-2.39) Private PC 1.887*** 1.077* 0.509*** 0.381*** (3.66) (1.86) (5.00) (2.96) Highgov PC 2.386*** 2.043*** 0.460*** 0.273* (4.83) (3.75) (3.95) (1.96) Constant *** *** *** (-4.60) (-5.08) (-5.00) (1.20) Industry YES YES YES YES YES YES

17 Year YES YES YES YES YES YES Province YES YES YES YES YES YES N Pseudo R2/R Note:*** ** * represents significance level of 1%, 5% and 10% respectively. The t-statistics, computed using the robust standard error clustered at firm level, are in parentheses. As shown in table 7, the coefficients of the cross-terms, Private PC and Highgov PC, are both significantly positive, indicating that political connections have a more significant effect on helping obtain credit lines for non state-owned firms and firms in regions with high level of government intervention. We can draw the same conclusion from either bank lines of credit or the ratio of bank lines of credit granted to the firm s total assets. 3) Sensitivity test In order to test the robustness of the empirical results, we substituted the overall political connection index with the political background of the board chairmen and CEO, and ran the same regressions. No change was made to our major conclusions 4 : For firms with low credit rating, small scale and are thus facing financing constraint, their political connection would help them obtain bank lines of credit, especially from state-own banks. Also, to non state-owned firms or firms in regions with high level of government intervention, political background is more helpful. For firms with high credit rating, large scale and are not facing financing constraint, political background has no significant positive influence on bank lines of credit. 5. Conclusion and discussions Expanding in recent years, bank lines of credit businesses is playing an increasingly essential role in corporate business operation and liquidity management. Till now, only a small part of companies are able to obtain bank credit resources among the group of listed firms. To most companies, it is a sort of scarce financial resources. We analyzed the samples of companies listed in Shanghai and Shenzhen Stock Exchange from 2004 to 2009 and discovered that political connections of a firm would help it obtain bank lines of credit, especially from state-own banks. Meanwhile, by classifying firms into different groups in accordance with the financing constraints they are facing, we discovered that only when firms are facing financing constraints (e.g. firms with low credit rating, small scale) political connections have a significant impact on the acquisition of bank lines of credit. To put it in another way, although bank lines of credit is a kind of resource that is difficult to obtain, firms with high credit rating and are not facing financing constraint do not need to rely on political connections to get bank credit. To firms with low credit rating, facing strict financing constraints, however, political connections are useful. Furthermore, this article reveals that in places with high level of government intervention, the positive effects that political connections have on obtaining bank 4 The detailed regression results are available upon request.

18 lines of credit is more significant. Studies of this article not only deepen the understanding of bank lines of credit but also the understanding of the impact of politics on corporate finance. Of course, whether there are rent seeking activities when firms are going after credit lines awaits further exploration. Rent seeking may result in lower investment efficiency after the firms are granted bank lines of credit. References: [1] Agarwal, S., S. Chomsisengphet, J. Driscoll, Loan commitments and private firms, FEDS Working Paper No , 2004, April. [2] Allen, F., Qian, M., Qian, J., Law, finance, and economic growth in China, Journal of Financial Economics, 2005, 77, [3] Berger, A.N., and G.F. Udell, Relationship lending and lines of small firm finance, Journal of Business, 1995, 68, [4] Berkovitch, E., S. Greenbaum, The loan commitment as an optimal financing contract, Journal of Financial and Quantitative Analysis, 1991,26, [5] Boot, A.W., A.V. Thakor, and G. F. Udell, Competition, risk neutrality and loan commitments, Journal of Bankig and Finance, 1987, 11, [6] Boubakri, N., Cosset, J.C., Saffar, W., Political Journal of Corporate Finance 14, [7] Boyreau-Debray, G. and S. J. Wei, Pitfalls of a State Dominated Financial System: The Case of China, NBER Working Paper, No , [8] Brandt, L. and H. Li, Bank Discrimination in Transition Economies: Ideology, Information or Incentives?, Journal of Comparative Economics, 2003, 31, [9] Charumilind, C., R. Kali, Y. Wiwattanakantang, Connected lending: Thailand before the financial crisis, The Journal of Business, 2006, 79, [10] Chen, C., Z. Li and X. Su, Rent Seeking Incentives, Political Connections and Organizational Structure: Empirical Evidence from Listed Family Firms in China, Working Paper, City University of Hong Kong, [11] Choi,C.J.,Lee,S.H.,, J.B., A Note on Counter trade:contractual Uncertainty and Transaction Governance in Transition Economies, Journal of International Business Studies, 1999, 30, [12] Claessens, S., E. Feijen, and L. Laeven, Political Connections and Preferential Access to Finance: The Role of Campaign Contributions, Journal of Financial Economics, 2008, 88, [13] Cull, Robert, Xu, Lixin Colin, Institutions, ownership, and finance: The determinants of profit reinvestment among Chinese firms, Journal of Financial Economics, 2005, 77 (1), [14] Du, J., and S. Girma., Red Capitalists: Political Connections and the Growth and Survival of Start-up Companies in China. University of Nottingham. GEP Research Paper 2007/40, [15] Faccio, M., Politically connected firms, The American Economic Review, 2006, 96, [16] Fan, Joseph P.H., Wong T.J., T. Zhang, Politically-connected CEOs, Corporate Governance and Post-IPO Performance of China's Newly Partially Privatized Firms, Journal of Financial Economics,2007, 84,

19 [17] Fan, G., Wang, X., Zhu, H, Chinese Marketization Index 2009 report of Regional Maketization process of China, Beijing: Economic Science Publishing House, [18] Fazzari,S.,Hubbard,R.and Petersen,B., Financing Constraints and Corporate Investment, Brookings Papers on Economic Activity,1988,19, [19] Firth Michael, Chen Lin, Ping Liu, Sonia M.L. Wong, Inside the black box: Bank credit allocation in China s private sector, Journal of Banking and Finance, 2009, 33, [20] Gatev, E., P. Strahan, Banks' advantage in hedging liquidity risk: Theory and evidence from the commercial paper market, Journal of Finance, 2006, 61, [21] Gordon, R. and W. Li, Government as a Discriminating Monopolist in the Financial Market: The Case of China, Journal of Public Economics, 2003, 87, [22] Jimenez, G., J. Lopez, J. Saurina, Empirical analysis of corporate credit lines, Review of Financial Studies, 2009, 22,5069. [23] Kashyap, A., R. Rajan, J. Stein, Banks as liquidity providers: An explanation for the coexistence of lending and deposit-taking, Journal of Finance, 2002, 57, [24] Khwaja, A., Mian, A., Do lenders favor politically connected firms? Rent provision in an emerging financial market, Quarterly Journal of Economics,2005, 120, [25] LaPorta,R.,F. Lopez-de-Silane,C. Pop-Eleches,and A. Shleifer. Judicial Checks and Balances, Journal of Political Economy, 2004, 112(2), [26] Li, H., L. Meng and J. Zhang, Why Do Entrepreneurs Enter Politics?Evidence from China, Economic Inquiry, 2006, 44, [27] Li, H., Meng, L., Wang, Q., and Zhou, L., Political Connections, Financing and Firm Performance: Evidence from Chinese Private Firms, Journal of Development Economics, 2008,87, [28] McMillan, J., Markets in Transition, Advances in Economics and Econometrics, Volume II, edited by David M. Kreps and Kenneth F. Wallis, Cambridge: Cambridge University Press, 1997, [29] Rajan,R.G. and L. Zingales,Financial Dependence and Growth,American Economic Review, 1998, 88(3), [30] Saidenberg, M., P. Strahan, Are banks still important for financing large businesses?, Current Issues in Economics and Finance, 1999, 5(12). [31] Sufi, A., Bank lines of credit in corporate finance: An empirical analysis, Review of Financial Studies, 2009,22, [32] Xin,K. R. and Pearce,J. L.,Guanxi: Connections as substitutes for formal institutional support,academy of Management Journal,1996,39(6),