financial development, bank discrimination and trade credit
DESCRIPTION
TRANSCRIPT
Financial Development, Bank Discrimination and
Trade Credit
Ying Gea and Jiaping Qiub
a School of International Trade and Economics, University of International Business and
Economics, Beijing, China
b Clarica Financial Research Centre, School of Business and Economics, Wilfrid Laurier University, 75 University Avenue West, Waterloo, ON, Canada, N2L 3C5
Abstract
Non-state owned firms in China grow tremendously with limited support from banks. This provides a unique setting to test how firms in a country with poorly developed financial institutions fund their prosperous growth opportunities. This paper compares the use of an important non-formal financial channel, trade credit, between state and non-state owned firms in China. We find that, compared to state owned firms, non-state owned firms demand more trade credit, and this higher demand is primarily for financing rather than transactional purpose. Moreover, non-state owned firms are more likely to choose trade credit over bank loans as the preferable financing source. The results suggest that, in a country with a poorly developed formal financial sector, firms can support their growth through non-formal financial channels that largely rely on implicit contractual relation such as culture, reputation and relationship. JEL Classification: O5, K0, G2 Key Words: Trade Credit, Financial Development, Chinese Economy
1. Introduction
The interaction between a country’s financial development and growth is a central
question in financial economics. The level of a country’s financial development typically
is measured by the services provided by formal financial intermediaries, e.g., the size of
bank to GDP, the size of equity to GDP and credit issued to private firms. King and
Levine (1993) show that the level of a country’s financial development is positively
associated with its economic growth, physical capital accumulation and the
improvement of economic efficiency. Rajan and Zingales (1998) find that firms in
industrial sectors with a greater need for external finance grow faster in countries with
well-developed financial markets. These studies support the notion that a
well-developed financial system can facilitate a country’s economic growth. Yet, they
also raise an important question: in countries with poorly developed financial system,
how do firms finance their growth opportunities?
One possible solution is that firms with high return projects in countries with
poorly developed financial system may finance through non-formal financing channels,
such as funds from friends and family, trade credit with business partners and joint
accounts with other companies. It is not clear, however, to what extent non-formal
financing channels can substitute for formal financial systems. On the one hand, the
development of a country’s formal and non-formal financing channels may be both
affected by its legal environment (e.g., La Porta, Lopez-de-Silanes, Shleifer, and Vishny,
1997, 1998). Thus, the levels of the development of formal and non-formal financing
2
channels are highly correlated; a country with a poor formal financial system is unlikely
to develop active and reliable non-formal financing channels. On the other hand,
culture, reputation and other implicit contractual relations may serve as substitutes for
the lack of legal and contract enforcement mechanisms (e.g., Greif, 1989, 1993, 1994,
Stulz and Williamson, 2002). Presumably, non-formal financing channels largely rely on
reputation and relationship. This view suggests that the possibility of well-functioning
non-standard financing channels in countries with poor financial and legal systems.
In this paper, we present a comprehensive study on the use of an important
non-formal financial channel - trade credits - for firms in China, the largest developing
country with a fast growth economy but a poorly developed formal financial system.
There are two primary reasons why a study on the use of trade credit in China is of
particular interest. First, the sharp differences in the availability of bank loans and
growth rate between state and non-state owned firms in China provide a natural setting
to test how (non-state) firms respond to the constraints of formal financial channels. On
the one hand, the non-state sector has very limited access to bank loans. According to
the World Bank Report 2000, in the late 1990s the non-state sector received less than one
percent of total bank loans. On the other hand, non-state owned firms dominate state
owned firms in growth. Between 1995 and 2000, the industrial output of the non-state
sector grew at 19% annually, while that of the state sector grew only at 4.6% annually
(China Statistical Yearbook, 2000). The stark contrast between state and non-state
sectors in their growth rate and the availability of bank loans provides a unique
opportunity to analyze how firms with prosperous growth opportunities exploit
3
alternative financial channels in the presence of formal financial channel constraints.
Second, trade credit is a loan that a supplier provides to its customers in conjunction
with product sales. The effectiveness of trade credit as an informal financing channel
largely relies on reputation and relationship, especially in a country like China where
there is relatively poor legal enforcement of contracts. Thus, testing the demand for
trade credit in China can help us understand if culture, reputation and other implicit
contractual relations are able to substitute for the lack of legal mechanisms.
To test the difference in the use of trade credit between state and non-state
sectors in China, we employ a unique survey data prepared by the Chinese Academy of
Social Sciences, which provides detailed financial information and survey
questionnaires on state and non-state owned firms. Using the information in the firms’
financial statements, we find that, compared to state owned firms, non-state owned
firms tend to use more trade credit. We then distinguish between transactional and
financing motives of trade credit by examining the trade credit outstanding, i.e., trade
credit expired but not repaid, and long-term trade credit, i.e., credit more than 30 days
after the delivery of goods. We find that non-state firms are more likely to have trade
credit outstanding and long term trade credit, which suggests that the higher demand
for trade credit by non-state owned firms is primarily for financing instead of
transactional purposes. Finally, using the information from survey questionnaires, we
find that managers in non-state firms, compared to those in state owned firms, are more
likely to choose trade credit over bank loans as a preferable financing source. Thus, the
4
results indicate that trade credit is used by non-state firms as an important alternative
financing channel to solve the problem of scarce bank loans.
Our study contributes to extant literature in several important dimensions. It
presents a direct analysis of how non-state owned firms in China use non-formal
financial channels to finance their growth opportunities. How the non-state sector in
China sustains its high growth has been an interesting but puzzling question. A recent
study by Franklin, Qian and Qian (2005) compares the growth rate and formal financing
channels (e.g., bank loans and equity markets) of state owned firms and non-state
owned firms in China. They show that the growth rate of state owned firms is
dominated by non-state owned firms. Compared to state owned firms, non-state owned
firms have much poorer access to bank loans. Franklin, Qian and Qian (2005)
conjectures that very effective non-formal financing channels exist to support the
growth of the non-state owned sector in China. Using firm level data, our study directly
tests such a conjecture and shows that the non-state owned sector in China actively
exploits non-standard financing channels such as trade credits to finance its growth.
Our study, from a different angle, provides evidence on the use of trade credit as
a substitute for bank loans in developing countries. Fisman and Love (2003) examine
the use of trade credit in different countries and find that industries with higher
dependence on trade credit financing grow faster in countries with weaker financial
institutions. They argue that the results suggest that trade credit can substitute for bank
5
loans in countries with poor financial institutions.1 By focusing on firm level choices
and exploiting the unique information from the survey data, we are able to separate the
transactional from financing motives in the use of trade credit. Moreover, there is a clear
difference in the availability of bank loans between state and non-state firms, which
avoids the ambiguity in the measurement of financial constraint due to the banking
relationship. The separation of transactional and financing motives and the clean
measurement of the banking relationship allow us to draw a more direct conclusion
that the use of trade credit is for financing purposes to relax the constraint in obtaining
bank loans.
The remainder of the paper proceeds as follows: Section 2 describes the data
source and presents some summary statistics. Section 3 examines the difference in the
use of trade credit between state and non-state owned firms. Section 4 summarizes and
concludes the paper.
2. Data and Summary Statistics
Our data set comes from the enterprise surveys conducted by the Chinese Academy of
Social Sciences (CASS) in the year 2000. This survey provides annual data on 442 SOEs
from 1994 to 1999, which is the third part of a continuing survey of Chinese state-owned
enterprises since 1980. These surveys have been used in various studies of the Chinese
economy (e.g., Groves et al., 1994, 1995, Li, 1997). Different from previous surveys, the 1 In a study on the use of trade credit by the U.S. small businesses, Petersen and Rajan (1997) find that those small businesses with limited access to credit from financial institution (measured by longest relationship with lenders, previous history of loan rejection) held significantly higher levels of accounts payable. They suggest that the results imply that trade credit is used as an important alternative financing source by constrained firms.
6
year 2000 survey also provides annual information on 358 non-state enterprises. The
questionnaire was divided into two parts; the first, directed at the factory managers,
includes qualitative questions about the firm’s incentive and governance system. The
second part is directed at the enterprise accountants and includes quantitative questions
that ask for details of a firm's real and financial accounts.
Among 800 enterprises in our sample, around 55% of the firms are state-owned;
31% of the firms are non-state owned2; around 11% of the firms are publicly traded
companies; 3% of the firms are joint ventures. We exclude the publicly traded
companies and joint ventures, since these firms could finance through the stock market
or foreign market and not rely solely on domestic bank loans. The final sample includes
annual data for 677 firms from 1994 to 19993.
[Table 1 about here]
We use four measures to assess trade credit demand: accounts payable scaled by
total asset or by total sales; (accounts payable - accounts receivable) scaled by total
assets or total sales. The first two measure the total demand for trade credit while the
other two measure the net demand for trade credit. Rows 1 and 2 in Table 1 show the
summary statistics of total demand for trade credit. The sample mean of accounts
payable to total asset ratio is about 13%, with a standard deviation of 10%. The sample
mean of accounts payable to sales ratio is about 26%, with a standard deviation of 61%.
The high variation of this ratio is due to the high variation in sales across firms. Rows 5
2 Collective-owned firms and private firms 3 Four state owned firms changed to stock-limited companies and five state owned firms changed to joint ventures during the sample period. We exclude these nine firms in our sample.
7
and 6 of Table 1 show the summary statistics of net demand for trade credit. If scaled by
total asset, net trade credit demand has a sample mean of -1% if scaled by sales, net
trade credit has a sample mean of 3%.
Panel B in Table 1 report the summary statistics of various firm characteristics.
The average firm age is about 27 years, indicating that the sampled state owned firms
were well established ones with long histories. The capital-to-labor ratio is defined as
total fixed assets divided by the total number of employees. The sample mean is about
45, and the standard deviation is about 133. Since the survey selected firms operating in
more than 30 different industries, there are large variations in the capital-to-labor ratio.
Firm size is measured by the logarithm of the total number of employees. The average
size of firms is 7.05, with a relatively small variation. Two measurements are used to
assess a firm’s performance. One is accounting profit, which is defined as total profit
divided by total assets. The average profit rate is 8% and the cross firm variation is large.
The other measure is output per labor, which is defined as real output (in constant price
of 1990) divided by the total number of employees. It measures the physical production
capacity of firms. The sample mean is about 53 and standard variation is about 76.
3. The Demand for Trade Credit: State vs. Non-state Firms 3.1 Some Tabulated Facts
We start with a simple mean difference test to look at the difference in the demand for
trade credit between state and non-state firms. The average demand for trade credit
between state and non-state firms is reported in Table 2. Row 1 reports the difference
for the whole sample period. Columns 2 and 3 in row 1 show that, for the whole sample
8
period, the average ratio of accounts payable/total assets is 15.5% for non-state firms,
compared to 11.2% for state firms. Non-state firms on average own 3.4% less accounts
payable/total assets than state firms; this difference is statistically significant at the 5%
level. The difference is even larger if we look at the difference in accounts payable/sales
ratio. Column 5 to 7 in row 1 shows that the average ratios of accounts payable/sales
are 49.1% and 13.1% for non-state and state owned firms respectively; the difference is
36% and is significant at the 5% level. The results indicate that non-state firms tend to
have larger accounts payable than state firms.
[Table 2 about here]
Columns 8 to 13 present the difference in the net trade credit demand, accounts
payable minus accounts receivable normalized by total assets or sales. The results show
that the average ratios of net trade credit are positive for non-state firms but negative
for state firms, suggesting that on average non-state firms are net credit demanders
while state firms are trade credit suppliers. To check the consistency of this pattern
across different sample years, rows 2 to 7 further report the mean difference test for
each sample year. The results show that the difference in demand for trade credit
between state and non-state firms is quite consistent and significant across different
sample years. Thus, a simple mean difference test clearly indicates that non-state firms
demand more trade credit than state firms.
3.2 Regression Analysis
The mean difference tests show a significant difference in the use of trade credit
between state and non-state owned firms. Of course, state and non-state owned firms
9
could differ in various aspects, e.g., firm size, age and profitability, which might affect
the demand for trade credit. In this section, we test the effect of ownership (state vs.
non-state) on the demand for trade credit but control for various firm characteristics
that could potentially affect a firm’s demand for trade credit. The regression
specification is following:
tititititi
tititititi
VLocationbIndustrybProfitbLaborCapitalbAgebSizebOwnershipbaTC
,,7,61,5
1,4,31,2,1, /ελ ++++++
++++=
−
−−
(1)
where the dependent variable, , is firm i’s demand for trade credit at time t. Five
measurements are employed to assess the demand for trade credit: accounts
payable/total assets, accounts payable/sales, (accounts payable – accounts
receivable)/total assets, (accounts payable – accounts receivable)/sales, a net trade
credit demander indicator which is equal to 1 if a firm is net trade credit demander, i.e.,
accounts payable is greater than accounts receivable, and 0 otherwise. The independent
variable of central interest, , is a dummy variable indicating a firm’s
ownership, which is equal to 1 if the firm is state-owned and 0 otherwise. Other
controlling variables are conventional and include: is the lagged firm’s size
measured by the logarithm of the number of employees;
tiTC ,
tiOwnership ,
1, −tiSize
4 is the firm’s age
measured by the length of time a firm has existed. In the absence of effective legal
enforcement, the use of trade credit relies on implicit contractual relations such as trust
and reputation. Larger and older firms usually have better established records and may
tiAge ,
4 The variable is lagged one year to mitigate the endogenous issue.
10
be easier to finance through trade credit. Yet larger and older firms might have better
banking relationships and have less need for trade credit. Thus, the effect of firm size
and age on the demand for trade credit may be ambiguous. is the
capital labor ratio, which is measured by a firm’s total fixed assets divided by the
number of employees. is the profit rate measured by total profit divided by
total assets. Profit rate could be a signal for the future growth opportunities of a firm.
Profitable firms may find it easier to get trade credit since lenders are more comfortable
with their repayment ability. However, profitable firms may have more internal cash
flow to sustain their growth and have less need for trade credit. is the
industry a firm belongs to. Ng, Smith, and Smith (1999) and Fisman and Love (2003)
document a large variation of trade credit across industries but little variation within a
industry, which suggests that industry is one of the primary determinants in the
demand for trade credit. is a firm’s location, indicating the province in
which a firm is located. There are significant unbalances in financial development
among Chinese provinces, which may affect the use of trade credit for firms in different
provinces. is a firm’s unobservable individual effect;
1,/ −tiLaborCapital
1i,tProfit −
tiIndustry ,
tiLocation ,
iV tλ is year effect; itε is the
error term. We estimate the model using random effect.5
[Table 3 about here]
The determinants of the demand for trade credit are shown in Table 3. Columns 2
and 3 report the results on the determinants of trade credit demand. Columns 4 and 5
5 Note that fixed effect estimation is not applicable since the ownership (state or non-state) will be absorbed in the fixed effect.
11
report results on the determinants of net trade credit demand. Inspection of the results
in Table 3 reveals that the variables, , are significantly negative and consistent
across different measurements of trade credit. For example, in column 4 on the
determinants of the net trade credit/total assets ratio, the coefficient on is
-0.045, suggesting that state owned firms on average have 4.5% less net trade credit than
non-state owned firms, other things being equal. Column 6 further investigates the
likelihood of a firm being a net trade credit demander, using a random effect Logit
model. Again, the coefficient on the ownership dummy is significantly negative,
implying that non-state owned firms are more likely to be a net trade credit demander
than state owned firms. In contrast, there are no consistent effects of other firm’s
characteristics on the demand for trade credit. The effects of firm size, age and
profitability are significant for some measurements of trade credit but insignificant for
others. The results suggest that the ownership is a primary determinant of the use of
trade credit for Chinese firms. The results from regression analysis are consistent with
the simple mean difference test and show that non-state owned firms use more trade
credit than state owned firms, even after controlling for the various firm characteristics.
tiState ,
tiOwnership ,
4. The Motives of Using Trade Credit
One limitation of using accounts payable as the measurement of trade credit is that it
can’t distinguish between transactional motives and financing motives of trade credit
demand. Firms may use trade credit for various reasons, such as price discrimination by
12
supplier, sales motives for quality guarantee or financial support by suppliers. 6
Following Nilsen (2002), we separate the use of trade credit into transactional motives,
the use of trade credit related to transaction service, and the financing motives, the
trade credit provided by suppliers which allows the financially constrained customers
to delay the payments. Separation between these two motives is important, since our
story is that non-state owned firms use trade credit as a means of external finance to
circumvent the difficulty in financing through formal channels.
In this section, we will take a closer look at firms’ motives with the use of trade
credit. Moreover, if non-state owned firms are constraint in getting bank loans and use
trade credit as a substitute for bank loans, the difference between state and non-state
owned firms in their preference between bank loans and trade credit as external
financing sources could provide direct evidence on the effect of bank discrimination on
a firm’s finance choice. Thus, we will also examine the difference between state and
non-state owned firms in their preferences for bank loans or trade credit.
4.1 Trade Credit Outstanding
We first investigate the difference in the structure of trade credit outstanding between
state and non-state firms. Trade credit outstanding refers to trade credit that has
expired but is not repaid. Firms are usually reluctant to have trade credit outstanding
because they may face significant late payment penalties, including the explicit cost of
pecuniary penalties as well as implicit costs of damaging long-term relationships with
6 For a detailed discussion, see Peterson and Rajan (1997), Nilsen (2002) and Fisman and Love (2003).
13
customers (Petersen and Rajan, 1997). Unless firms lack funds, they won’t delay trade
credit repayment because of the significant penalty. Thus trade credit outstanding tends
to be used for financing purposes in the presence of constraint in bank loans.
The data provides information on the sources of outstanding debt which allows
us to investigate the trade credit outstanding. In the survey questionnaire, managers are
asked, does your firm currently have outstanding debt from any source? If the answer
is yes, managers are asked to indicate the sources of outstanding debt. 288 firms in our
sample, including 163 state owned firms and 65 non-state owned firms, have debt
outstanding at the end of 1999. The sources of debt outstanding include bank loans,
trade credit, credits from rural credit cooperatives, non-bank financial institutions and
others. Bank loans and trade credit are two main sources of debt outstanding,
comprising 89% of the total debt outstanding. The average shares of state bank loans
and trade credit in the debt outstanding are 76% and 13% respectively. Though bank
loans are the dominant source in the debt outstanding, there is a clear difference in the
structure of debt outstanding between state and non-state firms. State bank loans
comprise 86% of debt outstanding for state firms, compared with only 51% for the
non-state firms. In contrast, trade credit only comprises 6% of debt outstanding for state
firms, compared to 31% for non-state firms.
[Table 4 about here]
Table 4 examines the shares of bank loans and trade credit in total debt
outstanding. Columns 2 and 3 report the results on the shares of bank loans and trade
credit outstanding, using the OLS estimation conditional on a firm having debt
14
outstanding. In the equation of bank loan share, the coefficient on is
positive and significant. In contrast, in the equation of trade credit share, the coefficient
on is negative and significant. The results indicate that state owned firms
have a larger proportion of bank loans in debt outstanding than non-state owned firms,
while the non-state owned firms have a larger proportion of trade credit in debt
outstanding than state owned firms.
tiOwnership ,
tiOwnership ,
The results of conditional OLS regressions show that non-state owned firms have
a higher average share of trade credit but lower bank loans in debt outstanding than
state owned firms do. However, the results cannot tell if trade credit or bank loans
outstanding is the major source of debt outstanding for non-state owned firms. For
example, if a state owned firm has a 70% bank loans and 20% trade credit in their debt
outstanding and a non-state owned firm has a 60% bank loans and 30% trade credit, the
non-state firm has a larger share of trade credit and lower share of bank loans in debt
outstanding. But bank loans are still the major source of debt outstanding for non-state
firms. To further investigate the relative importance of bank loans and trade credit for
state owned versus non-state owned firms, we define a dummy variable which is equal
to 1 if the share of trade credit is larger than the share of bank loans in a firm’s debt
outstanding, and 0 otherwise. This variable measures the relative importance of bank
loans versus trade credit. We then test to see if a firm’s ownership affects the relative
importance of trade credit to bank loans in its debt outstanding.
15
Another issue is that a firm’s ownership may also affect the likelihood of a firm
having debt outstanding, and thus a cross section regression conditional on firms with
positive debt outstanding may suffer from a selection bias. To deal with this issue, we
estimate the relative importance of trade credit versus bank loans using the Probit
model with sample selection. In the Probit equation, the dependent variable is the
dummy variable indicating the relative importance of trade credit as defined above. In
the sample selection equation, the dependent variable is a dummy variable which is
equal to 1 if a firm has debt outstanding and 0 otherwise. The model is estimated using
the method by Van de Ven and Van Pragg (1981).
The results of the Probit model with sample selection are reported in Columns 4
and 5 in Table 4. In the selection equation, the coefficient on is positive
and significant suggesting that state owned firms are more likely to have debt
outstanding. This result is reasonable given that state firms have good banking
relationships that allow them to extend debt payment easily. Conversely, in the Probit
equation, the coefficient on is negative and significant indicating that
trade credit is more likely to be the dominant source of debt outstanding for non-state
firms. In fact, 3% of state owned firms in the sample, compared to 31% of non-state
owned firms, reported a larger share of trade credit than bank loans in their debt
outstanding.
tiOwnership ,
tiOwnership ,
16
4.2 Long Term Trade Credit
Another approach to distinguish between the transactional motive and external
financing motive is to investigate the maturity structure of trade credit. Compared to
bank loans, trade credit is usually repaid within a short term. As Ng et al. (1999)
describe, the most common repayment plan is “2/10, net 30”: a supplier offers a 2%
discount on the purchase price if the customer can repay within 10 days; otherwise full
repayment is required within 30 days. The short term trade credit generally provides
transactional service for firms. If the repayment term of trade credit is unusually long,
then it’s likely that trade credit is used by suppliers as a financial support to customers.
The survey questionnaire provides unique information to differentiate the long
and short term trade credit. In the survey, managers are asked, what proportion of your
company’s payments to the supplier is made at the following times (in percent)? There
are six schedules: (1) when the order is placed; (2) on delivery; (3) 1-7 days after
delivery; (4) 8-30 days after delivery; (5) more than 30 days after delivery; and (6) other
schedule. We thus define the payments made more than 30 days after delivery as long
term trade credit.
[Table 5 about here]
We then perform two tests on the determinants of a firm’s use of long term trade
credit: one is the determinants of the access to long term trade credit using the Logit
model. The dependent variable is a dummy variable indicating if a firm has long term
trade credit; the other is the determinants of the share of long term credit in the trade
credit. The dependent variable is the share of long term trade credit in total trade credit.
17
The independent variables are the same as those in Table 4. The results are reported in
Table 5. Column 2 shows the determinants of the availability of long term trade credit.
The coefficient on is negative and significant at the 5% level, indicating
that the state owned firms are less likely to demand long term trade credit. The tobit
estimation of the determinants of the share of long term credit is reported in Column 3.
Again, the coefficient on is negative and significant at the 1% level,
suggesting that state owned firms tend to have a lower share of long term trade credit.
Thus, the results on the maturity structure of trade credit show that the non-state
owned firms are more likely to use long term credit and have a greater share of long
term trade credit. The results are consistent with analysis on the trade credit extension
and indicate that the greater demand for trade credit by non-state owned firms is
primarily for financing purpose.
tiOwnership ,
tiOwnership ,
4.3. Firm Preferences: Bank Loans versus Trade Credit
That non-state owned firms are discriminated against in bank lending has been well
documented and studied in the literature. There are many reasons for banks preferring
to lend to state-owned firms. For example, the Chinese government tends to pressure
banks to provide financial support to state owned firms through cheap bank credit (e.g.,
Lardy, 1998, Cull and Xu, 2003). Bank managers are less responsible for loans defaulted
by state owned firms since both enterprises and banks are owned by the state. The
income of bank employees does not depend highly on the profitability of the bank, and
this weakens the incentive for bank employees to identify high return projects. Local
18
government remains an influence on bank decision and encourages the banks to extend
credit to state owned firms that lack money to pay their employees.
Clearly, such discrimination could affect state and non-state owned firms’ choices
on financing resources. Presumably, state firms will prefer bank loans over trade credit,
given that bank loans are generally a cheaper financing source. In contrast, non-state
owned firms may prefer trade credit over bank loans given the difficulty in borrowing
from banks. Thus, a test of a firm’s choice between bank loans and trade credit can
provide direct evidence of the consequences of bank discrimination on a firm’s
financing behavior.
[Table 6 about here]
In the survey, managers are asked to rank favorable external financing sources.
Based on the manager’s answer, we first define a dummy which is equal to 1 if a firm
prefers trade credit over bank loans, and 0 otherwise. We then test the effect of a firm’s
ownership on the likelihood of preferring trade credit over bank loans, using the Logit
model. The result is reported in Column 2 of Table 6. One can see that the coefficient on
is significantly negative, implying that non-state firms are more likely to
choose trade credit over bank loans as their preferred financial sources. Trade credit is
generally viewed as a less attractive financing source than are bank loans. The distorted
preference of non-state owned firms in China suggests that the unavailability of bank
loans is due to bank discrimination discourages non-state owned firms to use bank
loans and the cost of trade credit may be effectively reduced through effective informal
contractual mechanisms, such as reputation, culture and relationship.
tiOwnership ,
19
5． Summary and Conclusion
One important feature in the Chinese high growth economy is that non-state owned
firms dominate state firms in growth and profitability. Non-state owned firms, however,
are discriminated against by banks in receiving loans. How non-state owned firms in
China finance their growth has been an interesting question. In this paper, we focus on
the use of an informal financing channel, trade credit, for firms in China. We find that,
compared to state owned firms, non-state owned firms tend to use more trade credit.
Non-state owned firms also have more trade credit outstanding and long term trade
credit, suggesting that the greater demand for trade credit by non-state owned firms is
primarily driven by financing motives instead of transactional motives. Moreover,
non-state firms are more likely to choose trade credit as a preferable financing source
rather than bank loans. The results suggest that non-state owned firms are able to
exploit informal financing channels to fund their growth in response to the constraint of
formal financing channels.
Our results also have implications for the role of culture and reputation in firms’
external financing in countries lacking legal and contractual enforcement mechanisms.
The effectiveness of trade credit largely relies on reputation and relationship. The
significant use of trade credit by non state firms in China, a country with a relatively
poorly developed legal system, implies that implicit contract enforcement mechanisms,
such as culture and relationship, could to a certain extent substitute for the legal system.
20
However, one should interpret our results as informal financing channels could
replace a well functioning formal financial system. In fact, trade credit for finance is a
highly unattractive substitution for bank loans. Petersen and Rajan (1997) estimate the
cost of trade credit in U.S. small firms and find it is more expensive than 99.8 percent of
the loans. Moreover, trade credit is tied to the purchase of goods which is less flexible
than bank loans. Thus, even though trade credit could play a significant role in
providing external finance to support the growth of firms, an effective formal financial
system may be necessary to sustain a country’s long run growth.
21
Reference Cull, R. and L. Xu, 2003, Who Gets Credit? The Behaviour of Bureaucrats and State
Banks in Allocating Credit to Chinese State-Owned Enterprises, Journal of Development Economics 71, 533-559.
Franklin, A., J. Qian, and M. Qian, 2005, Law, Finance and Economic Growth in China,
Journal of Financial Economics, forthcoming. Fisman, R. and Love, I., 2003, Trade Credit, Financial Development and Industry
Growth, Journal of Finance LVIII, 353-373 Greif, A., 1989, Reputation and Coalitions in Medieval Trade: Evidence on the Maghribi
Traders, Journal of Economic History 49, 857-882. Greif, A., 1993, Contract Enforceability and Economic Institutions in Early Trade: The
Maghribi Traders Coalition, American Economic Review 83, 525-548. Greif, Avner 1994, Cultural Beliefs and the Organization of Society: A Historical and
Theoretical Reflection on Collectivist and Individualist Societies, Journal of Political Economy 102, 912-950.
Groves, T., Hong, Y., McMillan, J. and Naughton, B., 1994, Autonomy and Incentives in
Chinese State Enterprises, Quarterly Journal of Economics, 184-209.
Groves, T., Hong, Y., McMillan, J. and Naughton, B, 1995, China's Evolving Managerial Labor Market, Journal of Political Economy 103, 873-892.
King, R. G., and R. Levine, 1993, Finance and growth: Schumpeter Might Be Right,
Quarterly Journal of Economics 108, 717-737. La Porta, R., Florencio Lopez-de-Silanes, Andrei Shleifer, and Robert Vishny, 1997.
Legal Determinants of External Finance, Journal of Finance 52, 1131-1150. La Porta, R., F. Lopez-de-Silanes, A. Shleifer, and R. Vishny, 1998, Law and Finance,
Journal of Political Economy. 106, 113-1155. Lardy, N. R., 1998, China’s Unfinished Economic Revolution, Brookings Institution Press,
Washington.
22
Li, W., 1997, The Impact of Economic Reform on the Performance of Chinese State Enterprises, 1980-1989, Journal of Political Economy 105, 1080-1106.
Ng, C. K., J. K. Smith, and R. Smith, 1999, Evidence on the Determinants of Credit
Terms Used in Interfirm Trade, Journal of Finance 54, 1109-1129. Nilsen, J., 2002, Trade Credit and the Bank Lending Channel of Monetary Policy
Transmission, Journal of Money,Credit, and Banking, forthcoming. . Petersen, M., and R. Rajan, 1997, Trade credit: Theories and Evidence, Review of Financial
Studies 10, 661-691. Rajan, R., and L. Zingales, 1998, Financial Dependence and Growth, American Economic
Review 88, 559-586. State Statistical Bureau, 2000, China Statistical Yearbook, Beijing: China Statistical Press. Stulz, R., and R. Williamson, 2002, Culture, Openness, and Finance, Journal of Financial
Economics, forthcoming.
Van de Ven, W.P.M.M. and B.M.S. Van Pragg, 1981, The Demand for Deductibles in Private Health Insurance: A Probit Model with Sample Selection, Journal of Econometric, 17: 229-252.
World Bank, 2000, China’s Emerging Private Enterprises: Prospects for the New
Century, International Finance Corporation.
23
Table 1
Summary Statistics of Sample SOEs
This table presents summary statistics of sample SOEs. The sample is from the enterprise surveys conducted by the Chinese Academy of Social Sciences (CASS) in the year 2000.
Obs. Mean S.D. 25% 50% 75%
A: Trade Credit Demand
Accounts payable / total assets 4062 0.13 0.10 0.05 0.10 0.17
Accounts payable / sales 4062 0.26 0.61 0.06 0.13 0.25 (Accounts payable - accounts receivable) / total assets 4062 -0.01 0.12 -0.08 -0.02 0.05
(Accounts payable - accounts receivable) / sales 4062 0.03 0.55 -0.08 -0.02 0.07
B: Firm Characteristics
Age 4062 26.96 18.07 10.00 28.00 39.00
Capital/labor 4062 45.00 132.5 17.00 26.52 42.56
Size 4062 7.05 1.00 6.51 7.03 7.62
Profits 4062 0.08 0.28 -0.03 0.02 0.12
Output per employer 4062 53.28 76.29 17.39 32.49 60.96
24
Table 2
Mean Difference Test on the Demand for Trade Credit: State versus Non-state Owned Firms This table reports the mean difference test on the demand for trade credit between state and non-state owned firms. Row 2 reports the results for the whole sample period. Rows 2 to 8 report the results for each sample year. Four measurements are used to assess the trade credit demand: accounts payable scaled by total asset or by total sales, (accounts payable – accounts receivable) scaled by total assets or total sales. For each measurement, first and second columns report the sample mean for non-state and state firms respectively; the third column reports the difference. t values are reported in parenthesis.
Accounts Payable / Total Assets
Accounts Payable / Sales Accounts Payable – Accounts Receivable) / Total Assets
(Accounts Payable – Accounts Receivable) / Sales
Year
Non-state State Diff. Non-state State Diff. Non-state State Diff. Non-state State Diff.All 0.155 0.112 0.043**
(11.87) 0.491 0.131 0.360**
(14.36) 0.022 -0.026 0.048**
(11.82) 0.103 -0.011 0.114**
(4.83) 1994
0.157 0.103 0.054**(6.14)
0.379 0.151 0.228**(6.64)
0.020 -0.025 0.045**(4.45)
0.067 -0.023 0.090**(2.37)
1995 0.154 0.109 0.045**(5.05)
0.385 0.149 0.236**(6.44)
0.024 -0.011 0.035**(3.58)
0.067 0.004 0.063*(1.94)
1996 0.153 0.113 0.040**(4.74)
0.440 0.130 0.310**(5.32)
0.025 -0.016 0.041**(4.42)
0.076 -0.002 0.078(1.28)
1997 0.151 0.114 0.037**(4.46)
0.474 0.119 0.355**(5.67)
0.025 -0.026 0.051**(5.25)
0.091 -0.010 0.101*(1.58)
1998 0.155 0.116 0.039**(4.34)
0.573 0.122 0.451**(6.82)
0.020 -0.033 0.053**(5.32)
0.100 -0.016 0.116*(1.76)
1999 0.156 0.117 0.039**(4.36)
0.696 0.113 0.583**(6.43)
0.019 -0.043 0.062**(5.85)
0.217 -0.023 0.240**(3.18)
Table 3
The Determinants of the Demand for Trade Credit
This table reports the determinants of the demand for trade credit. The regression specification is = + + + +b b ba
where the dependent variable, TCi,t, is the firm i’s demand for trade credit at time t,. Five measurements are employed to assess the demand for trade credit: accounts payable/total assets, accounts payable/sales, (accounts payable – accounts receivable)/total assets, (accounts payable – accounts receivable)/sale, a net trade credit demander indicator which is equal to 1 if the firm is net trade credit demander, accounts payable is greater than accounts receivable, and 0 otherwise. Ownershipi,t, is a dummy variable indicating a firm’s ownership, which is equal to 1 if the firm is state-owned, and 0 otherwise. Sizei,t-1 is lagged firm size, which is measured by logarithm of number of employees; Agei,t is firm age, indicating how long the firms have been in existence; capital/Labori,t-1 is the lagged capital labor ratio, which is measured by a firm’s total fixed assets divided by number of employees; profiti,t-1 is lagged profit rate measured by accounting profit divided by total assets. Industry is a firm’s industry; Locationi,t is a firm’s location indicating the province in which a firm is located. V is a firm’s individual effect.i tλ is year effect; itε is the error term. The model is estimated using random effect. t-values reported in parenthesis.
Accounts
Payable / Total Assets
Accounts Payable /
Sales
(Accounts Payable – Accounts
Receivable) / Total Assets
(Accounts Payable – Accounts
Receivable) / Sales
Net Trade Credit
Demander Dummy
Ownership -0.020* (1.78)
-0.685*** (-9.37)
-0.049*** (-3.67)
-0.225*** (-3.25)
-4.524*** (-6.71)
Size -0.014*** (-3.75)
-0.14 (-0.62)
-0.003 (-0.78)
-0.011 (-0.48)
0.127 (0.88)
Age -0.07 (-1.29)
0.205*** (5.87)
0.003 (0.53)
0.090*** (2.65)
0.846*** (3.27)
Capital/labor -0.0001*** (-4.34)
-0.0002 (-0.14)
-0.00001 (-0.30)
-0.0003 (-0.19)
-0.005*** (2.63)
Profit -0.004 (-0.86)
-0.065** (-2.02)
-0.023*** (-3.91)
-0.041 (-1.28)
-0.838* (-1.94)
Constant Yes Yes Yes Yes Yes
Location dummies
Yes Yes Yes Yes Yes
Year dummies Yes Yes Yes Yes Yes
R square 0.078 0.074 0.072 0.007 -
Observations (# of firms)
3385 (677)
3385 (677)
3385 (677)
3385 (677)
3385 (677)
titit,it , 1− i
t , 1− iLabor t,itit , it , iTC O 1 b wnership Size Age Capital / VLocationb 6 Pr 5 ofit b ,
431−2 ,
+ + + + λ + ε
Table 4
The Determinants of Shares of Banks Loans and Trade Credit in Debt Outstanding
This table reports the regression analysis on the shares of bank loans and trade credit in debt outstanding. Columns 2 and 3 report the results on the shares of bank loans and trade credit from the OLS estimation, conditional on firms have debt outstanding, respectively. Column 4 and 5 report the results of the Probit model with sample selection. In the Probit equation, the dependent variable is a dummy variable which is equal to 1 if the share of trade credit is larger than the share of state bank loans in a firm’s debt outstanding, and 0 otherwise. In the selection equation, the dependent variable is a dummy variable indicating if a firm has debt outstanding. Ownership, is a dummy variable indicating a firm’s ownership which is equal to 1 if the firm is state-owned, and 0 otherwise. Size is firm size, which is measured by logarithm of number of employees; Age is firm age indicating how long the firms have been existing; Capital/Labor is the lagged capital labor ratio which is measured by a firm’s total fixed assets divided by number of employees; Profit is lagged profit rate measured by accounting profit divided by total assets. Industry is a firm’s industry; Location is a firm’s location indicating the province in which a firm is located. t-values are reported in parenthesis.
Conditional OLS Probit Model with Selection Bank Loan
Share Trade Credit
Share Trade Credit
vs. Bank Loans Selection Equation
Ownership 27.161*** (2.86)
-18.907*** (-2.97)
-1.438** (-2.53)
0.780*** (3.31)
Size 5.268** (2.39)
-1.941 (-1.09)
-0.123 (-0.84)
0.181*** (2.91)
Age -0.286 (-0.05)
-2.470 (-0.61)
0.116 (0.33)
-0.274* (-1.75)
Capital/labor -0.007 (-1.35)
0.007** (2.30)
-0.009 (-1.41)
0.0002 (0.55)
Profit 1.068 (0.14)
5.415 (1.27)
0.516 (1.12)
-1.169*** (-3.75)
Constant 20.667 (1.05)
45.702*** (2.93)
-1.466 (-1.10)
-1.434* (-1.73)
Industry dummies Yes Yes Yes Yes
Location dummies Yes Yes Yes Yes
R square 0.301 0.284 - -
Observations 228 228 228 677
27
Table 5
The Determinants of Long Term Trade Credit This table reports the regression analysis on the determinants of long term trade credit. Column 2 report the results on the likelihood of having long term trade credit using the Logit estimation. Column 3 report the results on share of long term trade credit in total trade credit using Tobit model. Ownership is a dummy variable indicating a firm’s ownership, which is equal to 1 if the firm is state-owned and 0 otherwise. Size is firm size, which is measured by logarithm of number of employees; Age is firm age indicating how long the firms have been in existence; Capital/Labor is the lagged capital labor ratio which is measured by a firm’s total fixed assets divided by number of employees; Profit is lagged profit rate measured by accounting profit divided by total assets. Industry is a firm’s industry; Location is a firm’s location indicating the province in which a firm is located. t-values are reported in parenthesis.
Long Term Trade Credit Dummy
Long Term Trade Credit Share
Ownership -1.272*** (2.67)
-0.166*** (-2.63)
Size -0.096 (-0.83)
0.00004 (0.001)
Age 0.947*** (3.06)
0.122*** (2.91)
Capital/labor -0.001 (-0.80)
-0.0001 (-0.74)
Profit -0.002 (-0.01)
-0.014 (-0.32)
Constant -18.287*** (16.33)
-0.401** (-2.55)
Industry dummies Yes Yes Location dummies Yes Yes R-square 0.053 0.081 Uncensored observations - 133 Observations 677 677
28
Table 6
The Determinants of Preferred Financing Sources This table reports the Logit regression analysis on the firm choices of bank loans versus trade credit as the preferable financing source. The dependent variable is a dummy, which is equal to 1 if a firm chooses trade credit over bank loans as the favorable external financing sources, and 0 otherwise. Size is firm size, which is measured by logarithm of number of employees; Age is firm age indicating how long the firms have been in existence; Capital/labor is the lagged capital labor ratio which is measured by a firm’s total fixed asset divided by number of employees; Profit is lagged profit rate measured by accounting profit divided by total assets. Industry is a firm’s industry; Location is a firm’s location, indicating the province in which a firm is located. t-values are reported in parenthesis.
Trade Credit is Preferred over Bank Loans Ownership -0.092**
(-2.18) Size 0.018*
(1.77) Age 0.046
(1.63) Capital/labor 0.00001**
(2.68) Profit 0.039
(1.31) Constant 0.781***
(8.91) Industry dummies Yes
Province dummies Yes
R-square 0.030
Observations 677
29