Trade Credit Provision and National Culture

Trade Credit Provision and National Culture
Sadok El Ghoul*
Campus Saint-Jean, University of Alberta
Edmonton, AB T6C 4G9, Canada
[email protected]
Xiaolan Zheng
Nottingham University Business School (NUBS) China
Ningbo, China
[email protected]
Abstract
In this paper we investigate the effect of national culture on trade credit provision. We
generate testable hypotheses linking Hofstede’s four cultural dimensions
(collectivism/individualism, power distance, uncertainty avoidance, and
masculinity/femininity) to trade credit. Consistent with our predictions based on
several theories of trade credit, we find that after controlling for firm- and countrylevel factors as well as industry and year effects, trade credit provision is higher in
countries with higher collectivism, power distance, uncertainty avoidance, and
masculinity scores. These results are robust to using alternative measures of culture
and trade credit, addressing potential endogeneity concerns, and using alternative
estimation methods and sample compositions. In addition, we find that international
trade openness mitigates the influence of domestic culture on trade credit provision.
March 2015
JEL classification: G32; Z10
Key words: National culture, trade credit
* Corresponding author. This paper was originally prepared for the April 17-18, 2015 “Culture and Finance”
conference at Wake Forest University. We thank Jean-Marie Gagnon and Nabil Khoury for their insights on an earlier
version of this paper. Sadok El Ghoul acknowledges financial support from Canada’s Social Sciences and Humanities
Research Council.
Trade Credit Provision and National Culture
Abstract
In this paper we investigate the effect of national culture on trade credit provision. We
generate testable hypotheses linking Hofstede’s four cultural dimensions
(collectivism/individualism, power distance, uncertainty avoidance, and
masculinity/femininity) to trade credit. Consistent with our predictions based on
several theories of trade credit, we find that after controlling for firm- and countrylevel factors as well as industry and year effects, trade credit provision is higher in
countries with higher collectivism, power distance, uncertainty avoidance, and
masculinity scores. These results are robust to using alternative measures of culture
and trade credit, addressing potential endogeneity concerns, and using alternative
estimation methods and sample compositions. In addition, we find that international
trade openness mitigates the influence of domestic culture on trade credit provision.
March 2015
JEL classification: G32; Z10
Key words: National culture, trade credit
1. Introduction
Rather than require immediate payment for delivered goods and services, suppliers can
allow their customers to delay payment. This gives rise to trade credit. Trade credit is pervasive
around the world. For instance, in our sample of publicly listed firms from 51 countries, aggregate
trade credit reached nearly $US 5.6 trillion in 2012, with the average firm extending approximately
19% (17%) of its sales (assets) in the form of accounts receivable. However, compared to other
components of working capital, 1 prior literature tells us little about the determinants of crosscountry variation in the provision of trade credit.
In this paper we fill this gap in the literature by examining the impact of national culture
on cross-country differences in trade credit provision. Culture, “the collective programming of the
mind which distinguishes the members of one group or category of people from another” (Hofstede
(2001) page 9), influences how people interpret information and calibrates thoughts and behaviors
such that they are compatible with prevailing values (e.g., Licht et al., 2005; Lonner and
Adamopoulos, 1997; North, 1990). Cultural values thus serve as fundamental constraints on how
one evaluates problems and selects actions (Licht et al., 2005). Against this backdrop, we posit
that the decision to extend trade credit is shaped by suppliers’ national culture. To examine this
relation, we employ Hofstede’s four cultural dimensions (collectivism/individualism, power
distance, uncertainty avoidance, and masculinity/femininity) as proxies for culture. Using a sample
of 335,405 firm-year observations from 51 countries over the 1992 to 2012 period, and controlling
for firm- and country-level factors as well as industry and year effects, we find that suppliers
located in countries with higher collectivism, power distance, uncertainty avoidance, and
masculinity scores tend to offer more trade credit to their customers. These results continue to hold
when we control for potential endogeneity problems, when we employ alternative measures of
culture and trade credit, and when we consider alternative estimation methods and sample
compositions. Moreover, we find that a country’s openness to international trade mediates the
effect of national culture on trade credit provision.
1
For instance, a large number of studies examine the determinants of international differences in corporate cashing
holdings (e.g., Dittmar et al., 2003; Ferreira and Vilela, 2004; Kalcheva and Lins, 2007; Khurana et al., 2006; Kusnadi
and Wei, 2011; Lins et al., 2010; Pinkowitz et al., 2003; Pinkowitz et al., 2013) which in 2012 amounted to $US 5.5
trillion and account for roughly 17% of an average firm’s assets in our sample.
1
The results are consistent with our predictions, which are based on several theories of trade
credit. In high collectivism societies, suppliers extend more trade credit because they can obtain
information about customer creditworthiness at lower cost and can rely on collective punishment
in the event customer defaults. In high power distance countries, which are characterized by large
inequalities between rich (cash) and poor (credit) customers, suppliers have incentives to use trade
credit to price discriminate. In high uncertainty avoidance countries, where customers are more
concerned about the quality of their purchases, suppliers are likely to use more trade credit to
provide an implicit warranty on their products. And in high masculinity countries, where borrowers
are likely to display more opportunistic behaviors, sellers are likely to use more trade credit as it
is a more effective constraint on borrower opportunism than bank credit.
Our paper contributes to several streams of research. First, our paper is related to the
literature on the determinants of firms’ trade credit policies. A number of single-country papers
examine this question, specifically, for the U.S. (e.g., Giannetti et al., 2011; Petersen and Rajan,
1997) (e.g., Long et al., 1993; Molina and Preve, 2009; Ng et al., 1999), the U.K. (Atanasova,
2007; Cunat, 2007), China (Cull et al., 2009; Ge and Qiu, 2007), and Vietnam (McMillan and
Woodruff, 1999). In comparison, cross-country studies are relatively scarce. For a sample of
publicly traded firms from 40 countries, Demirgüç-Kunt and Maksimovic (2001) find that trade
credit provision is higher in countries with more developed banking systems while it is lower in
countries with more protective legal environments. Beck et al. (2008) study large versus small
firms in 48 countries and find that small firms have reduced access to external (e.g., bank) finance,
particularly when property rights are not well protected, but that trade credit does not help fill this
gap. For a sample of five formerly communist countries, Johnson et al. (2002) find that legal
enforcement of contracts leads suppliers to extend more trade credit. Finally, in a review of the
trade credit contract terms offered by suppliers in the U.S. and Europe,2 Klapper et al. (2012) find
2
Several studies in this literature examine trade credit terms. For instance, Ng et al. (1999) observe that trade credit
terms take one of two forms: net terms, whereby payment is due within a specified period after delivery, or two-part
terms. Under “net 30” terms, for example, the customer can delay payment up to 30 days after delivery. The most
common two-part terms is “2/10 net 30”, which means that the customer receives a 2% markdown if they pay during
the discount period (i.e., within 10 days of delivery) but otherwise pays the full price during the remainder of the net
period (i.e., between days 11 and 30 after delivery). If the customer forgoes the early payment discount, the annual
100
360⁄(#  −# )
100
360⁄(30−10)
implicit interest rate would be (
)
−1=(
)
− 1 = 43.9%.
100−2
100−%
This example illustrates that trade credit can be an expensive source of finance. Unfortunately, data on trade credit
terms are limited. Accordingly, in this paper we focus on the determinants of the volume of trade credit extended to
customers by suppliers as reported in the Compustat database.
2
that large customers receive better trade credit terms (i.e., longer maturity) while riskier customers
face higher implicit rates. However, Klapper et al. (2012) do not study the impact of the
institutional environment on trade credit contracts. We contribute to this literature by showing that
national culture, an informal country-level institution, plays an important role in shaping crosscountry variation in trade credit provision. Specifically, we show that different dimensions of
national culture have a significant impact over and above that of previously documented countryand firm-level determinants of trade credit provision.
Our paper is also related to the literature on the effect of culture on firms’ financial decision
making. Studies in this literature examine the impact of culture on debt ratios (Chui et al., 2002),
debt maturity (Zheng et al., 2012), dividend policies (Shao et al., 2010), earnings management
(Han et al., 2010), mergers and acquisitions (Ahern et al., forthcoming), investment (Shao et al.,
2013), profit reinvestment (El Ghoul et al., 2014), and executive compensation (Tosi and
Greckhamer, 2004). Our paper contributes to this literature by showing that national culture affects
yet another firm decision, namely, trade credit.
Our paper is also related to the economic growth literature. Fisman and Love (2003) find
that industries relying on trade credit financing grow faster in countries with an underdeveloped
financial system. However, the authors do not investigate what affects suppliers’ decision to extend
trade credit in such environments. Our paper complements their analysis by showing that informal
institutions (i.e., national culture) can act as substitutes for formal institutions (e.g., financial
markets) in channeling external funds to firms through trade credit. More generally, we contribute
to research, starting with (Weber, 1905 [2001]), that identifies channels through which culture
affects economic growth, for example, countries’ saving rates and fiscal redistribution policies
(Guiso et al., 2006), financial system (Aggarwal and Goodell, 2009; Kwok and Tadesse, 2006),
legal institutions (Licht et al., 2005; Stulz and Williamson, 2003), and corruption in bank lending
(Zheng et al., 2013).
The remainder of the paper is organized as follows. Section 2 reviews existing literature
on national culture and trade credit and develops our main hypotheses. Section 3 discusses the data
and variables, and provides descriptive statistics. Section 4 presents our main empirical analysis.
Section 5 reports results of robustness tests. Section 6 concludes.
2. Literature Review and Hypothesis Development
3
In this section, we first discuss Hofstede’s cultural framework. We then build on several
theories of trade credit to develop predictions linking Hofstede’s (2001) four cultural dimensions
(individualism/collectivism, power distance, uncertainty avoidance, and masculinity/femininity)
to trade credit provision.
2.1.
Hofstede’s cultural framework
Hofstede’s cultural framework views culture as the collective mental programming that
prescribes ways of thinking, feeling, and behaving (Hofstede, 2001). This framework comprises
four dimensions that capture fundamental aspects of a culture and thus can be used to capture
differences across cultures (Hofstede, 1983, 1991, 2001; Hofstede and Bond 1988). The
individualism/collectivism dimension characterizes the relationship between self and group. In
high individualism countries, people value their independence, maintain loose ties with others, and
are expected to look after themselves and their immediate families only. In high collectivism
countries, in contrast, people have interdependent self-construals and place the interests of their
in-groups (e.g., extended families) above their own interests. The power distance dimension
captures “the extent to which the less powerful members of institutions and organizations within
a country expect and accept that power is distributed unequally” (Hofstede, 2001: 98). An unequal
distribution of power has more legitimacy in high power distance countries. The uncertainty
avoidance dimension focuses on the extent of anxiety a society feels in the face of uncertainty and
how people respond to unstructured situations. Countries with strong uncertainty-avoidance
sentiment value institutions and beliefs that minimize variation in unknown outcomes. Finally, the
masculinity/femininity dimension captures the extent to which male assertiveness (e.g.,
achievement, recognition, and material success) is a dominant value in a society as opposed to
female nurturance (e.g., modesty, caring for others).
2.2.
Hypotheses3
2.2.1. Trade credit and networks
One might posit that a well-functioning legal system is a prerequisite for the provision of
trade credit—after all, a well-functioning legal system allows suppliers to collect unpaid debts,
3
We note that there exist additional theories of trade credit, but they do not apply to our setting. For completeness,
these theories include transaction cost(Emery, 1987; Ferris, 1981), financing motive (Schwartz, 1974), tax (Brick and
Fung, 1984; Desai et al., 2012), and bargaining power and relationship (Cunat, 2007; Dass et al., 2014; Wilner,
2000)theories of trade credit.
4
reducing buyers’ incentives to cheat. However, trade credit can flourish where contract
enforceability is limited or nonexistent. Greif (1989, 1993)’s studies of “Maghribi traders”,
eleventh-century Jewish merchants operating in the Muslim Mediterranean, provide one elegant
example. The Maghribis had two options: sail along with their goods, or recruit overseas agents to
supply this service. Absent contractual problems, the second option should reduce traders’ cost,
time, and risk of travelling and allow them to better diversify activities across trading centers.
However, for the second option to work, the Maghribis needed assurance that the agents would
not misappropriate their goods. In the absence of a legal system that provided contract enforcement,
the Maghribis created an informal network of traders that shared information about agents (i.e., a
monitoring mechanism) and engaged in collective punishment of agents that cheated (i.e., an
enforcement mechanism).4,5 For instance, cheating agents may not be hired again and coalition
members could cheat them in return (e.g., withhold debt repayment) without fear of collective
punishment. Greif (1994) attributes the Maghribis’ particular choice of institution (i.e., the
coalition) to their collectivist beliefs.6 Therefore, culture—in particular, collectivism—acted as a
substitute for a legal system in fostering pre-Modern trade and trade credit in the Mediterranean.
In more recent studies, McMillan and Woodruff (1999) and Johnson et al. (2002) examine trade
credit in Vietnam and the formerly planned economies of Eastern Europe and the Soviet Union
(Poland, Romania, Russia, Slovakia, Ukraine), respectively. These samples comprise collectivist
4
A somewhat similar example can be found in Landa (1981) study of middlemen in the rubber trade from the HokkienChinese ethnic group operating in West Malaysia and Singapore in the late 1960s. These middlemen faced contract
uncertainty due to poor enforcement of contract law. As a result, they established an ethnic-based pecking order of
trading partners to reduce information asymmetry and the costs of contract enforcement. This pecking order was as
follows: 1. near kinsmen from family, 2. distant kinsmen from extended family, 3. clansmen, 4. fellow villagers from
China, 5. Fellow Hokkiens, 6. non-Hokkiens, and 7. non-Chinese. The first five layers comprised “insiders” sharing
the same ethnicity and abiding by the Confucian code of ethics, while the last two were “outsiders”. Interestingly, the
middlemen used credit transactions with insiders and cash transactions with outsiders. Landa establishes theoretically
that ethnically homogeneous middleman groups arise as an efficient response to poor contract enforcement. In a
follow-up paper, Landa (2008) argues that the emergence of other ethnic-based merchant groups around the world
such as Indian traders in South Africa and Lebanese traders in West Africa is consistent with her theory of ethnically
homogeneous middleman groups.
5
Greif et al. (1994) argue that (formal) merchant guilds in medieval Europe played a similar institutional role as the
Maghribis’ (informal) coalition. Rulers of trade centers often confiscated merchandise of foreign merchants. Absent
proper enforcement of property rights, foreign merchants had little incentive to trade with overseas trade centers. This
situation resulted in less trade, which was costly to both rulers and foreign merchants. Rulers favored the emergence
of merchant guilds—formal associations of foreign merchants—that were empowered to coordinate collective
sanctions (e.g., embargos) on delinquent rulers. The creation of merchant guilds resulted in more trade to the benefit
of rulers and foreign merchants alike.
6
In contrast, Genoese traders, who held individualist beliefs, developed costlier formal (e.g., legal) institutions to
manage their agency relationships (Greif, 1994).
5
countries with relatively underdeveloped legal systems. The authors find that firms in these
countries that belong to business and social networks extend more trade credit to their customers.
This evidence is consistent with firms sharing information about customers (“information effect”)
and sanctioning customers that default (“sanctions effect”). McMillan and Woodruff (1999) argue,
however, that the information effect should be valuable early in a supplier-customer relationship
while the sanctions effect should be more persistent. They find that network membership does
indeed have a persistent effect on trade credit provision, consistent with the dominance of the
sanctions effect over the information effect.
Based on this discussion, we predict that suppliers are likely to extend more trade credit in
collectivist countries, where firms can obtain more information about customer creditworthiness
and can trigger collective punishment of delinquent customers. This leads to our first hypothesis:
Hypothesis 1 (H1): Ceteris paribus, suppliers extend more trade credit in high collectivism
countries.
2.2.2. Trade credit and price discrimination
Brennan et al. (1988) argue that provided the reservation prices of cash customers are
higher than those of credit customers, or put differently, the price elasticity of demand of cash
customers is higher than that of credit customers, sellers will profit from offering lower prices to
credit customers. However, in several countries sellers face practical or legal constraints on using
price discrimination (Mian and Smith, 1992).7 To circumvent such constraints, sellers may extend
trade credit at subsidized (i.e., low) rates, lowering the effective price credit customers pay. In
related research, Petersen and Rajan (1997) argue that sellers with high profit margins have an
incentive to sell additional units on credit. The rationale is that sellers with high margins will make
a net profit on the additional units as long as the revenues from the additional sales outweigh the
costs of providing subsidized trade credit to poor customers. Consistent with this argument, they
find a positive relationship between a firm’s gross profit margin and its accounts receivable.
Motivated by these studies and Hofstede’s (1984: page 98)observation that “inequality in
power and inequality in wealth go hand in hand”, we next predict that sellers are likely to extend
7
Antitrust laws generally prohibit firms from price discriminating. See, for instance, the Clayton Act, which was the
basis for the Robinson-Patman Act in the U.S., the Competition Act in Canada, and Article 82(c) of the European
Community Treaty.
6
more trade credit in high power distance countries, which exhibit large inequalities between rich
(cash) and poor (credit) customers, to price discriminate. This leads to our second hypothesis:
Hypothesis 2 (H2): Ceteris paribus, suppliers extend more trade credit in high power distance
countries.
2.2.3. Trade credit and implicit warranties
Extant theoretical models suggest that sellers can use trade credit to certify product quality
(Emery and Nayar, 1998; Lee and Stowe, 1993; Long et al., 1993; Smith, 1987). To the extent that
customers cannot observe product quality at the time of purchase, trade credit can provide them
time to verify quality before making payment—if product quality falls short of expectations, they
can withhold payment and return the product to the seller.
Sellers may alternatively certify product quality through money-back guarantees and
product warranties. But these mechanisms are imperfect substitutes for trade credit. First, they
require that customers prove that product quality is substandard, which may be difficult to do for
products where the question of quality may be subjective. Under trade credit, however, by
withholding payment, customers place the burden of proof on the seller. Moreover, money-back
guarantees and product warranties require that the seller survives while customers inspect their
purchase; if the seller goes bankrupt during this time, then money-back guarantees and product
warranties often become worthless. Under trade credit, in contrast, customers do not make
payment upfront so are protected against this risk. In short, in the presence of information
asymmetry between sellers and customers with respect to product quality, trade credit provides
customers stronger protection than money-back guarantees or product warranties. Consistent with
this idea, Long et al. (1993) find that firms with longer production cycles (a proxy for higher
product quality), smaller firms with little reputation in product markets, and technology firms
whose product quality is difficult to determine extend more trade credit.
Based on this discussion, we predict that in high uncertainty avoidance countries, where
customers tend to be more anxious about the quality of products they purchase, sellers offer trade
credit to provide an implicit warranty, thereby inducing sales. This leads to our third hypothesis:
Hypothesis 3 (H3): Ceteris paribus, suppliers extend more trade credit in high uncertainty
avoidance countries.
7
2.2.4. Trade credit and bank lending
Prior literature indicates that relative to traditional credit extended by banks, trade credit
extended by sellers has a comparative advantage in terms of the lender’s ability to acquire
information, repossess goods, and control moral hazard. With respect to information acquisition,
extant literature suggests that relative to banks, trade creditors generate superior information about
borrowers. For instance, Smith (1987) argues that suppliers can induce trade credit borrowers (i.e.,
customers) to reveal their creditworthiness by providing deep discounts for early payments while
charging higher effective interest rates on late payments—customers that forgo the early payment
discounts signal that they are high default-risk customers that may require increased monitoring
or stricter terms, such as cash-on-delivery only on future sales. Suppliers may also obtain superior
(credit) information about customers in the normal course of business if, for instance, sale
representatives visit the premises of their customers on a repeated basis (Mian and Smith, 1992).
In addition, to the extent that suppliers work with a portfolio of similar customers, they can more
easily differentiate between an industry decline and a deterioration in the prospects of certain
customers (Ng et al., 1999). Reflecting sellers’ comparative advantage in information acquisition,
Jain (2001) argues that banks with asymmetric information about borrowers prefer to lend through
better-informed suppliers, that is, banks prefer to provide credit to suppliers, which in turn extend
trade credit to customers.8
Turning to the ability to repossess goods, Frank and Maksimovic (2005) argue that in the
event a customer defaults, a trade creditor is likely to suffer less of a loss than a bank, as the trade
creditor is better able to repossess and resell the product sold to the defaulting customer on
favorable terms to another customer. While banks generally demand seniority over other creditors,
Longhofer and Santos (2003) explain that banks allow trade creditors to collateralize the products
they sell because trade creditors value these goods more highly than banks do. Finally, with respect
to moral hazard, Burkart and Ellingsen (2004) argue that relative to banks, trade creditors are in a
8
In a related paper, Biais and Gollier (1997) show that trade credit and bank credit are complements. The authors
develop a game theoretic framework with three players: a bank, a seller and a customer. If the seller decides to extend
trade credit to the customer then trade credit provision reveals the seller’s private information to the bank, which in
turn decides to provide credit.
8
better position to control moral hazard as suppliers lend illiquid assets while banks lend cash, and
opportunistic borrowers find it more difficult to divert illiquid assets than cash.9
Building on the argument that the importance of achievement in high masculinity countries
leads individuals to show off (Hofstede, 2001), Zheng et al. (2012) conjecture that borrowers in
high masculinity countries engage in high-risk overinvestment at the expense of their lenders. They
find that, anticipating this incentive problem, lenders in such countries extend shorter-maturity
debt to mitigate borrower opportunism. Based on this argument and the discussion above, we
predict that in high masculinity countries, where customers are more likely to display opportunistic
behaviors, sellers are likely to use more trade credit as it is a more effective constraint on borrower
opportunism than traditional credit by banks. This leads to our fourth hypothesis:
Hypothesis 4 (H4): Ceteris paribus, suppliers extend more trade credit in high masculinity
countries.
2.2.5. Trade credit, national culture and international trade openness
An economy’s openness to international trade not only brings domestic managers greater
exposure to the values and norms of foreign countries but also exerts stronger competitive
pressures on them to understand foreign cultures and practices in the pursuit of firms’ international
success (e.g., Heuer et al., 1999). Thus, firms may be willing to relinquish their cultures to reap
the benefits from trading with foreigners. Consistent with this idea, previous research documents
that international trade openness mediates the relationship between culture and economic
outcomes. Stulz and Williamson (2003) find that a country’s openness to international trade
moderates the relationship between religion and a country’s protection of creditor rights. Eun et
al. (2015) suggest that the influence of domestic culture on stock price co-movements is tempered
by trade and financial openness. In the same vein, we predict that corporate decision on trade
credit extension would be less influenced by domestic culture as a country’s trade openness
increases.
When trading with foreigners adhering to different cultures, suppliers may need to alter
their trade credit policies. For instance, suppliers in collectivist countries are unlikely to sell
9
In a related paper, Fabbri and Menichini (2010) model sellers as having an advantage in both repossessing goods
and controlling moral hazard. One of their model’s predictions is that trade credit provision decreases with the extent
of creditor protection.
9
products to customers belonging to the same social networks, making it harder to collect
information about customers’ creditworthiness and to trigger collective punishment of delinquent
customers. Suppliers in high power distance countries could export products to foreign markets
characterized by fewer inequalities, thus weakening their incentives to use trade credit for price
discrimination purposes. Moreover, these suppliers may face greater global competitive pressures
that undermine their ability of price discriminate. Suppliers in high masculinity countries may lose
their edge foreign banks because geographic distance may undermine their ability to collect
information about customers as well as repossess products sold to defaulting customers. Hence,
we predict that as an economy becomes more open to international trade, domestic culture has less
of an influence on trade credit provision. This leads to our fifth hypothesis:
Hypothesis 5 (H5): Ceteris paribus, the influence of national culture on trade credit provision
is attenuated in countries that are more open to international trade.
3. Sample, Variables, and Descriptive Statistics
3.1.
Sample
To construct our sample, we merge firm financial data from Compustat Global and
Compustat North America with the cultural indices of Hofstede (2001) and macroeconomic and
financial development indicators from World Development Indicators (World Bank, 2012) and
Beck et al. (2000). We require that firms have non-missing values for Hofstede (2001)’s cultural
indices in our baseline model (described below). We further exclude observations before 1992 due
to limited country coverage, firms with missing Standard Industrial Classification (SIC) codes,
and firms in the financial industry (SIC between 6000 and 6999). Finally, we restrict our sample
to countries that have no less than 15 unique firms. These filters yield a sample of 335,405 firmyear observations from 51 countries over the period 1992 to 2012.
3.2.
Variables
Definitions and sources for all variables used in the main analysis and the robustness tests
are summarized in the Appendix. We winsorize all firm-level variables at 1st and 99th percentiles
to mitigate the influence of outliers. All time-variant independent variables are lagged one year
relative to the dependent variable to alleviate simultaneity concerns.
3.2.1. Trade credit
10
While Compustat separates accounts receivable due to trade from other receivables, it does
not do so for accounts payable. Our primary measure of trade credit therefore focuses on the
volume of trade credit extended by suppliers as reported in Compustat.10 Specifically, we measure
the trade credit extended by suppliers as 100 times the ratio of trade receivables to total sales
(Receivables/sales), where trade receivables is the amount (net of applicable reserves) owed by
customers for goods and services sold in the ordinary course of business. The ratio of receivables
to sales indicates the percentage of sales made through credit. Multiplying Receivables/sales by
3.6 gives the number of days that suppliers are willing to extend credit assuming that all buyers
receive 100% credit.
To test the sensitivity of our results to the measure of trade credit, in robustness tests
(Section 5) we replace Receivables/sales with the ratio of trade receivables to total assets
(Receivables/assets) and with the ratio of accounts payable to assets (Payables/ assets).
3.2.2. National culture
Following recent literature on culture and finance (e.g., Chui and Kwok, 2008; Chui et al.,
2010; Zheng et al., 2012), we capture national culture using Hofstede (2001)’s four cultural
dimensions, namely, individualism/collectivism (IDV), power distance (PDI), uncertainty
avoidance (UAI), and masculinity/femininity (MAS). We construct a collectivism index (CLT) as
100 minus Hofstede’s IDV. A higher CLT thus indicates greater emphasis on collectivist values.
Hofstede’s framework is arguably the most influential of cultural classifications, due to “its clarity,
parsimony, and resonance with managers” (Kirkman et al. (2006: 285-6).
Hofstede (2001)’s cultural indexes are based on survey data collected between 1967 and
1973.11 The data used to construct the cultural variables are thus from an earlier period than the
data used to construct the trade credit variables. While this reduces concerns of reverse causality,
it raises concerns about whether the data are outdated. Williamson (2000) and North (1991) show,
however, that culture changes on the order of centuries. In addition, Hofstede (2001) argues that
10
We would ideally measure trade credit for each bilateral relationship, collecting information on the suppliers that
sell on credit and the customers that purchase on credit, since the amount of trade credit is determined simultaneously
by suppliers’ willingness to extend and customers’ willingness to use trade credit (Klapper et al., 2012; Petersen and
Rajan, 1997). However, we are not aware of any database providing such detailed data on trade credit transactions for
a large cross-country sample.
11
The surveys comprise approximately 117,000 questionnaires from over 88,000 employees of IBM subsidiaries in
more than 70 countries.
11
the cultural indexes, which measure relative rather than absolute differences in culture, are fairly
stable over time, as factors such as new technologies tend to influence all countries and thus do
not change the relative differences between countries much. Using data from the World Values
Surveys, Inglehart and Baker (2000) further suggest that different societies move on parallel
trajectories. Nonetheless, in our main analyses (Section 4) we test whether our findings continue
to hold when using the updated Hofstede indexes of Tang and Koveos (2008) (CLT_TK, PDI_TK,
UAI_TK, and MAS_TK), which are based on changing economic environments.
3.2.3. Control variables
Following prior literature on trade credit (e.g., Giannetti et al., 2011; Love et al., 2007;
Petersen and Rajan, 1997), we control for firm- and country-level characteristics related to trade
credit to isolate the effect of national culture on trade credit. At the firm level, we include
Log(assets), the natural logarithm of total assets in $US millions; Profits/sales, the ratio of income
before extraordinary items to total sales; Cash/assets, the ratio of cash and short-term investments
to total assets; Fixed assets/assets, the ratio of total (net) property, plant, and equipment to total
assets; Sales growth, the growth rate of total sales; and Gross profit margin/sales, the ratio of total
sales less cost of goods sold to total sales. In addition, we include industry dummy variables based
on the Fama-French 48-industry classification to account for differences that result from variation
in the nature of product and market structure. All firm-level control variables are constructed from
Compustat. At the country level, we include GDP per capita, the natural logarithm of GDP per
capita from World Development Indicators (World Bank, 2012) in constant 2000 U.S. dollars, and
Private Credit from Beck et al. (2000), the private credit by deposit money banks divided by GDP.
Finally, we add year dummies in all regressions.
3.3.
Descriptive statistics
Tables 1 and 2 report descriptive statistics for Receivables/sales, Hofstede’s cultural
indexes, and the firm- and country-level control variables by country and for full sample,
respectively. Firms located outside the U.S. account for approximately 70% of the sample
observations. The average fraction of sales made on credit is 19.34% with a standard deviation of
17.16%. Receivables/sales varies widely across countries, from 9.71% (Estonia) to 34.30%
(Greece). Fig. 1 illustrates the geographical coverage of the sample—purple shading indicates
trade credit provision, with darker purple indicating a larger volume of trade credit extended in a
12
given country. The figure shows that firms in Mediterranean and Asian countries tend to extend
more trade credit, which includes the top ten countries with regards to trade credit volume, namely,
Greece, Italy, Morocco, Spain, France, Malaysia, India, China, Singapore, and the Philippines.
***Insert Tables 1 & 2 about here***
***Insert Figure 1 about Here***
Table 3 reports Pearson pairwise correlation coefficients for all variables in the main
regressions. We find that our primary proxy for trade credit provision, Receivables/sales, is
positively correlated with CLT, PDI, UAI, and MAS at the 1% level. These results provide initial
support for our four hypotheses. Moreover, Receivables/sales is correlated with most of the control
variables at the 1% level, confirming their relevance for trade credit.
***Insert Table 3 about Here***
4. Empirical Analysis
4.1.
The relation between national culture on trade credit provision
To investigate the impact of culture on trade credit, we use pooled OLS with standard errors
adjusted for heteroskedasticity and clustered at the firm level to reduce concerns about within-firm
correlation.
Table 4 presents regressions of trade credit. Our baseline model, Model (1), reports
estimates of regressing Receivables/sales on the firm- and country-level control variables. The
coefficient on Log(assets) is negative and significant at the 1% level, suggesting that smaller firms
tend to offer more sales on credit. Fixed assets/assets loads negatively at the 1% significance level,
consistent with Giannetti et al. (2011). Coupled with the previous finding, this result is inconsistent
with the redistributive view of trade credit whereby creditworthy suppliers redistribute funds to
less creditworthy customers (Meltzer, 1960; Schwartz, 1974). The coefficients on Profits/sales,
Cash/assets, and Sales growth are negative and significant at the 1% level, suggesting that firms
that generate and hold less internal cash or that face a decline in sales tend to extend more credit,
in line with the idea that firms in financial trouble may extend more trade credit to preserve sales
and the notion that customers are reluctant to repay financially troubled suppliers (Petersen and
13
Rajan (1997).12 The coefficient on Gross profit margin/sales is positive and significant at the 1%
level, consistent with the notion that firms with a larger gross profit margin have greater incentive
to sell an additional unit, if necessary, on credit. In line with Demirgüç-Kunt and Maksimovic
(2001), we find that GDP per capita is negatively related to trade credit provision, significant at
the 1% level, indicating that firms in less developed countries tend to sell more on credit. The
coefficient on Private Credit is positive at the 1% significance level, suggesting that firms in
countries with more developed financial markets provide more trade credit. This supports the
finding that trade credit is complementary to bank credit (Biais and Gollier, 1997; Demirgüç-Kunt
and Maksimovic (2001). The adjusted R2 of Model (1) is 0.142, similar to that in prior literature
for the U.S. (e.g., Giannetti et al., 2011; Petersen and Rajan, 1997), which indicates that our
baseline model is valid.
Following Chui and Kwok (2008) and Zheng et al. (2012), we first introduce the four
cultural indexes (CLT, PDI, UAI, and MAS) separately in Models (2) through (5). In Model (2),
we find that CLT is positively correlated with Receivables/sales at the 1% significance level,
consistent with the prediction in H1 that firms located in more collectivist countries extend more
trade credit. The effect of CLT is also economically sizable: holding the other explanatory variables
constant at their means, a one standard deviation increase in CLT would increase trade receivables
by about 2.8%, or alternatively lengthen the time suppliers are willing to extend credit by 10 days
(0.104×27.10×3.6). Further, compared with Model (1), the adjusted R2 is 7.7% higher, at 0.153,
after including CLT. These results support the view that firms extend more trade credit in
collectivist countries where firms can more easily share information on customer creditworthiness
and implement collective punishment on delinquent customers.
In Model (3), the coefficient on PDI is positive and significant at the 1% level, consistent
with the prediction in H2 that firms in high power distance countries extend more trade credit.
Further, holding the other explanatory variables constant at their means, a one standard deviation
increase in PDI would increase trade receivables by 3.4%, which is equivalent to an increase of 12
receivable collection days. Relative to the baseline model, the adjusted R2 in Model (3) is 11.3%
higher, at 0.158, after including PDI. These findings are consistent with firms using trade credit to
12
To check whether our results are driven by distressed firms, that is, firms with negative sales growth or negative
profits, in untabulated tests we re-estimate Table 4 excluding distressed firms and find that the coefficients on the
cultural indices exhibit the expected signs and are significant at the 1% level in Models (2) through (6).
14
price discriminate where the markets are characterized by large inequalities between rich (cash)
and poor (credit) customers.
In Model (4), we find that UAI loads positively at the 1% significance level, consistent with
the prediction in H3 that firms located in countries with greater uncertainty avoidance extend more
trade credit. Increasing UAI by one standard deviation from its mean holding all other explanatory
variables constant at their means would increase trade receivables by 1.6%, which is equivalent to
an increase of 6 receivable collection days. Relative to the baseline model, adding UAI increases
the adjusted R2 by 4.9%, to 0.149. These results are consistent with firms in countries with greater
uncertainty avoidance making a larger proportion of sales on credit to provide customers an
implicit warranty.
In Model (5), the coefficient on MAS is positive and significant at the 1% level, consistent
with the prediction in H4 that firms in more masculine countries extend more trade credit.
Economically, holding all other explanatory variables at their means, a one standard deviation
increase in MAS would increase trade receivables by 1%, which is equivalent to customers having
an additional 4 days to repay their credit. Introducing MAS increases the adjusted R2 by 2.1%, to
0.145, compared with the baseline model. We interpret these results as consistent with high
masculinity countries observing greater use of trade credit, as trade credit more effectively
mitigates the increased incentives for borrower opportunism in such countries than traditional bank
credit.
Next, given the correlations among the four cultural dimensions (Table 3), one may be
concerned that the results for one cultural dimension reflect the influence of another correlated
dimension on trade credit. To address this issue, in Model (6) we present the results of a horserace regression that includes all four cultural dimensions at the same time. We continue to find
that the coefficients on CLT, PDI, UAI, and MAS are positive and significant at the 1% level,
consistent with our previous findings when introducing these measures sequentially. The adjusted
R2 of 0.162 is higher than in the previous models.
To alleviate concerns that the relationship between national culture and trade credit may
not be linear, we follow Zheng et al. (2012) to construct cultural dummy variables (DCLT, DPDI,
DUAI, and DMAS) that take the value of one if a country’s score for a given dimension is above
the median based on the 51 countries in our sample. Model (7) reports the results. We find
15
significant evidence that firms in countries characterized by above-median collectivism, power
distance, uncertainty avoidance, and masculinity have a higher proportion of sales on credit, in
line with our previous results.
Since the survey data used to derive Hofstede’s cultural indexes were collected between
1967 and 1973, one may raise questions about the validity of the cultural values. To address this
concern, we use updated individualism/collectivism, power distance, uncertainty avoidance, and
masculinity/femininity scores from Tang and Koveos (2008). The results are presented in Model
(8). We continue to find that the four cultural variables load significantly at the 1% level, with the
same signs as in Model (6).
***Insert Table 4 about here***
Taken together, the results in Table 4 suggest that collectivism, power distance, uncertainty
avoidance, and masculinity have statistically significant explanatory power in predicting the
provision of trade credit. In particular, firms located in countries with higher CLT, PDI, UAI, and
MAS tend to extend more credit through receivables.
Below, we test the robustness of main results (Models (2) through (5) of Table 4) to our
choice of trade credit measure, potential endogeneity problems, and our choice of estimation
method and sample composition.
4.1.1 Alternative measures of trade credit
In Table 5, we test the sensitivity of our results to the choice of trade credit measure. We
first replace Receivables/sales with 100 times trade receivables scaled by book value of total assets,
Receivables/assets, and re-estimate Models (2) through (5) of Table 4. The results are reported in
Models (1) through (4) in Table 5. In addition, we examine whether our results on the relationship
between national culture and trade credit hold if we take customers’ perspective and regress
Payables/assets, that is, 100 times the ratio of accounts payable to book value of total assets
(Petersen and Rajan (1997)),13 on CLT, PDI, UAI, and MAS. The results are reported in Models (5)
through (8). In this second set of tests we include the same set of control variables as in Table 4,
except that we replace Gross profit margin/sales with Finished inv./total inv., the percentage of
inventory that is finished goods, to control for liquidation costs to credit suppliers; a higher
13
We also capture trade credit using 100 times the ratio of accounts payable to total sales. In untabulated results, we
find that the relationship between the four cultural dimensions and trade credit continues to hold.
16
percentage of finished goods indicates larger liquidation costs. Table 5 shows that our main
evidence on the effect of national culture on trade credit is virtually unchanged: we continue to
find a positive effect of collectivism, power distance, uncertainty avoidance, and masculinity on
the use of trade credit.
***Insert Table 5 about here***
4.1.2. Endogeneity
The next concern we address is potential endogeneity problems, which plague empirical
corporate finance studies. In the context of this paper, endogeneity problems could result from our
inability to randomly assign cultural levels to firms and observe their use of trade credit. Roberts
and Whited (2013) list three sources of endogeneity: omitted explanatory variables, simultaneity
bias, and measurement errors. Having already shown that our main findings are not sensitive to
using alternative measures of trade credit and culture, below we focus on omitted explanatory
variables and simultaneity bias.
A firm’s choice between requiring cash on (or before) delivery and selling on credit is
influenced by the contracting environment of the country it is located in (Demirgüç-Kunt and
Maksimovic, 2001). Aggarwal and Goodell (2009) similarly suggest that the efficiency of
enforcing incomplete contacts depends on a country’s economic, legal, political, social and cultural
environments. Despite a large reduction in sample size (to 293,972 firm-year observations from
45 countries) and potential multicollinearity resulting from the inclusion of a large set of
institutional factors, in Table 6 we add such factors given their potential correlation with trade
credit provision, in which case their omission could bias our estimates of the cultural indices. In
particular, we include Common Law, a dummy variable from La Porta et al. (2008) indicating
whether a country’s legal origin is English, to capture the general legal environment; Creditor
Rights from Djankov et al. (2007) to quantify the power of secured lenders in the bankruptcy
process granted by a country’s laws and regulations; Information Sharing, a dummy from Djankov
et al. (2007) indicating the existence of either a public registry or a private bureau in a given
country-year, to control for the information advantage of suppliers over other creditors; Law and
Order, from International Country Risk Guide (2008), to measure the overall efficiency of a
country’s legal system; Political Rights Index, from Freedom House (2014), to capture a country’s
level of political freedom; Trust from Barro and McCleary (2003), defined as the percentage of
17
participants in a country who respond that most people can be trusted based on the World Values
Survey (WVS), to capture a country’s social capital; and Cath from Barro and McCleary (2003),
defined as the percentage of a country’s population adhering to the Catholic faith, to capture a
country’s religious factors.
We find that Common Law loads significantly negatively, suggesting that firms in common
law countries have lower receivables to sales ratios. The coefficient on Creditor Rights is negative
and significant (except for Model (2)), suggesting that trade credit provision is higher in countries
with weak creditor protection. This result is consistent with Burkart and Ellingsen (2004), who
argue that suppliers’ advantage in controlling moral hazard problems relative to banks subsides
with stronger legal creditor protection. Information Sharing enters all regressions negatively at the
1% level, suggesting that firms use more trade credit when information asymmetry about
borrowers is more severe. This is consistent with the argument in Demirgüç-Kunt and Maksimovic
(2001) that when suppliers have proprietary information on customers, it could be optimal for
suppliers to borrow and redistribute credit to customers. Trust tends to load significantly negatively,
suggesting that firms use less trade credit in countries where people trust each other more. One
possible explanation for this result is that suppliers are less likely to use trade credit as an implicit
warranty in the presence of more trusting customers. The coefficient on Cath is positive and
significant at the 1% level in Model (1), which is consistent with Catholics’ preference for trade
credit and historical hostility to bank (interest-bearing) loans(Stulz and Williamson, 2003). More
importantly, our previous findings on the relation between national culture and trade credit
provision remain unchanged, reducing concerns that our results are driven by the omission of
country-level institutional factors.
***Insert Table 6 about here***
Turning to simultaneity bias, this could occur if trade credit and culture are determined in
equilibrium, with trade credit policies influencing a country’s culture. However, as we discuss in
Section 3.2.2, cultural values are relatively stable over time, with changes taking place on the order
of centuries (Williamson, 2000) North (1991). In addition, Licht et al. (2005) argue that culture
influences the design of formal institutions (such as political and legal rules) such that they are
compatible with prevailing cultural norms, which works to further stabilize cultural norms. Hence,
it is not likely that trade credit decisions at the firm level drive national culture. Further, from an
18
empirical perspective, Hofstede’s (2001) cultural indexes were derived from a sample period that
precedes our trade credit provision and firm characteristic sample. Nevertheless, to mitigate
concerns about possible simultaneity bias, we follow Gorodnichenko and Roland (2011) and
Zheng et al. (2013) and instrument Hofstede’s (2001) individualism/collectivism dimension (CLT)
using the genetic distance (Genetic Distance_CLT) between a given country and the U.S. (the most
individualistic country in our sample). Specifically, we use the population-weighted FST distance
from Spolaore and Wacziarg (2009), where FST distance captures the probability that two alleles
(a particular form taken by a gene) at a given locus selected at random from two populations will
be different. A higher FST distance indicates more genetic separation between two populations.
Similarly, we instrument for PDI, UAI, and MAS using Genetic Distance_PDI, Genetic
Distance_UAI, and Genetic Distance_MAS, which are given as the FST distance between a given
country and Malaysia (the country with largest power distance), Greece (the country with the
strongest uncertainty avoidance), and Japan (the country with the highest score along the
masculinity dimension), respectively. Since cultural and genetic transmission from one generation
to the next occur together, and since corporate trade credit decisions are not likely to influence
genetic variation, our instruments are theoretically relevant and exogenous.
We present the pooled IV regression results in Table 7. Models (1) through (4) use Genetic
Distance_CLT, Genetic Distance_PDI, Genetic Distance_UAI, and Genetic Distance_MAS to
instrument CLT, PDI, UAI, and MAS, respectively. To check the relevance of our IVs, we perform
an F-test of the excluded variable in the first-stage regressions using the null hypothesis that the
instrument (Genetic Distance_CLT, Genetic Distance_PDI, Genetic Distance_UAI, and Genetic
Distance_MAS) does not explain the variation in the corresponding cultural index (CLT, PDI, UAI,
and MAS, respectively). In each model we reject this null hypothesis at the 1% level. More
importantly, we find that our previous results on the relationship between the four cultural
dimensions and trade credit provision remain unaffected.
***Insert Table 7 about here***
4.1.3. Alternative sample composition and estimation methods
In our last set of tests we check whether our results on the relation between national culture
and trade credit are sensitive to sample composition and the choice of estimation method. First,
since U.S firms account for 30% of our sample, we examine whether our cross-country evidence
19
is driven by the dominance of U.S. firms. To do so, we employ the country-year means of all
control variables. Models (1) to (4) of Table 8 report results of pooled OLS regressions using 910
country-year observations with standard errors adjusted for heteroskedasticity. We also run WLS
regressions, where the weights are equal to the inverse of the number of firm-year observations in
each country, using the full sample in Models (5) through (8).14 Second, since Receivables/sales
is bounded at zero, we alternatively use Tobit estimation; the results are in Models (9) to (12).
Third, to increase the homogeneity of our sample we restrict the sample to firms with SIC between
2000 and 4999 to examine whether our results continue to hold after removing non-industrial firms.
These results are presented in Models (13) through (16). Finally, we check whether our findings
on the relation between culture and trade credit are driven by the choice of sample period. To do
so, in Table 9 we split our main sample into two sub-periods. Models (1) through (4) report results
for the 1992 to 2002 period, while Models (5) through (8) report results for the 2003 to 2012 period.
Overall, we find that our main results continue to go through.
***Insert Tables 8 and 9 about here***
4.2.
The effect of international trade openness on the relation between national culture and
trade credit provision
In this subsection, we explore whether international trade openness shapes the relationship
between domestic culture and trade credit. We measure international trade openness using two
variables obtained from World Development Indicators: Trade/GDP, defined as the natural
logarithm of trade (the sum of exports and imports of goods and services) as a share of GDP, and
Export/GDP, computed as the natural logarithm of the ratio of exports of goods and services to
GDP.15
We re-estimate Models (2) to (5) of Table 4, after adding our proxy for international trade
openness and its interaction with the corresponding cultural dimension. We present the results in
Table 10. The proxy for international trade openness is Trade/GDP in Models (1) through (4) and
Export/GDP in Models (5) through (8).
***Insert Table 10 about here***
14
We also re-estimate Models (2) through (5) of Table 4 after excluding U.S. firms. The untabulated results show that
our main findings continue to hold.
15
We use the log values to reduce the influence of skewness in the original trade to GDP and export to GDP ratios.
Eun et al. (2015) employ the same transformation.
20
In Models (1) through (8), we continue to find that coefficients of four cultural dimensions are
positively and statistically significant at the 1% level, consistent with our previous findings. More
importantly, we find in all models that the interaction terms between the international trade
openness proxies and the cultural variables load negatively at the 1% level. These results support
the view that domestic culture becomes less important in influencing trade credit provision when
a country’s economy is more open to international trade.
5. Conclusion
In this paper we investigate the impact of national culture, as proxied by Hofstede’s four
cultural dimensions (collectivism/individualism, power distance, uncertainty avoidance, and
masculinity/femininity), on trade credit provision. Using 335,405 firm-year observations from 51
countries over the 1992 to 2012 period, and after controlling for firm- and country-level
determinants as well as industry and year effects, we find that firms extend more trade credit in
countries with higher collectivism, power distance, uncertainty avoidance, and masculinity scores.
These results are robust to using alternative measures of culture and trade credit, addressing
endogeneity concerns, and controlling for the choice of estimation method and sample composition.
We also find that international trade openness mediates the impact of culture on trade credit
provision.
The results are consistent with our predictions based on several theories of trade credit.
Because suppliers in collectivist countries are likely to share information about customer
creditworthiness and can rely on collective retribution against opportunistic customers, they are
willing to extend more trade credit than suppliers in individualist countries. In high power distance
countries, which are likely to be characterized by large inequalities between rich (cash) and poor
(credit) customers, suppliers have incentives to use trade credit to price discriminate. In high
uncertainty avoidance countries, where customers are more concerned about the quality of their
purchases, suppliers are more likely to offer implicit warranties through trade credit. And in high
masculinity countries, where customers are more likely to adopt opportunistic behaviors, trade
credit can be a more effective constraint on such behavior than bank credit.
Our results have several practical implications. First, multinational companies (MNCs) that
operate in culturally distinct markets are likely to have different trade credit policies in different
countries. For instance, MNCs operating in higher power distance and higher uncertainty
21
avoidance countries relative to their home market may be incentivized to extend more trade credit
to customers in these markets. These MNCs may thus need to raise additional funds to finance
their accounts receivable. Second, previous literature argues that during monetary contractions,
small firms’ substitution of trade credit for bank credit (Meltzer, 1960; Nilsen, 2002; Schwartz,
1974) can attenuate the negative impact of monetary contractions on economic activity. Because
national culture shapes suppliers’ willingness to extend trade credit to customers, it may affect this
substitution of trade credit for bank credit. Our results therefore suggest that, given the cultural
environment, policymakers develop policies that encourage trade credit provision to counteract
the negative effects of monetary contractions.
22
REFERENCES
Aggarwal, R., Goodell, J.W., 2009. Markets and institutions in financial intermediation: National
characteristics as determinants. Journal of Banking and Finance 33, 1770-1780.
Ahern, K.R., Daminelli, D., Fracassi, C., forthcoming. Lost in Translation? The effect of cultural
values on mergers around the world. Journal of Financial Economics.
Atanasova, C., 2007. Access to institutional finance and the use of trade credit. Financial
Management 36, 49-67.
Barro, R., McCleary, R.M., 2003. Religion and economic growth across countries. American
Sociological Review 68, 760-781.
Beck, T., Demirgüç-Kunt, A., Levine, R., 2000. A new database on financial development and
structure. World Bank Economic Review 14, 597-605.
Beck, T., Demirgüç-Kunt, A., Maksimovic, V., 2008. Financing patterns around the world: Are
small firms different? Journal of Financial Economics 89, 467-487.
Biais, B., Gollier, C., 1997. Trade credit and credit rationing. Review of Financial Studies 10, 903937.
Brennan, M.J., Maksimovic, V., Zechner, J., 1988. Vendor financing. Journal of Finance 43, 11271141.
Brick, I.E., Fung, W.K., 1984. Taxes and the theory of trade debt. Journal of Finance 39, 1169-1176.
Burkart, M., Ellingsen, T., 2004. In-kind finance: A theory of trade credit. American Economic
Review, 569-590.
Chui, A.C.W., Kwok, C.C.Y., 2008. National culture and life insurance consumption. Journal of
International Business Studies 39, 88-101.
Chui, A.C.W., Lloyd, A.E., Kwok, C.C.Y., 2002. The determination of capital structure: Is national
culture a missing piece to the puzzle? Journal of International Business Studies 33, 99-127.
Chui, A.C.W., Titman, S., Wei, K.C.J., 2010. Individualism and momentum around the world.
Journal of Finance 65, 361-392.
Cull, R., Xu, L.C., Zhu, T., 2009. Formal finance and trade credit during China's transition. Journal
of Financial Intermediation 18, 173-192.
Cunat, V., 2007. Trade credit: Suppliers as debt collectors and insurance providers. Review of
Financial Studies 20, 491-527.
Dass, N., Kale, J.R., Nanda, V., 2014. Trade credit, relationship-specific investment, and product
market power. Review of Finance.
23
Demirgüç-Kunt, A., Maksimovic, V., 2001. Firms as financial intermediaries: Evidence from trade
credit data. World Bank Policy Research Working Paper.
Desai, M.A., Foley, C.F., Hines Jr, J.R., 2012. Trade credit and taxes. National Bureau of Economic
Research.
Dittmar, A., Mahrt-Smith, J., Servaes, H., 2003. International corporate governance and corporate
cash holdings. Journal of Financial and Quantitative Analysis 38, 111-133.
Djankov, S., McLiesh, C., Shleifer, A., 2007. Private credit in 129 countries. Journal of Financial
Economics 84, 299-329.
El Ghoul, S., Guedhami, O., Kwok, C.C., Shao, L., 2014. National Culture and Profit Reinvestment:
Evidence from SMEs. Working Paper.
Emery, G., Nayar, N., 1998. Product quality and payment policy. Review of Quantitative Finance
and Accounting 10, 269-284.
Emery, G.W., 1987. An optimal financial response to variable demand. Journal of Financial and
Quantitative Analysis 22, 209-225.
Fabbri, D., Menichini, A.M.C., 2010. Trade credit, collateral liquidation, and borrowing
constraints. Journal of Financial Economics 96, 413-432.
Ferreira, M.A., Vilela, A.S., 2004. Why do firms hold cash? Evidence from EMU countries.
European Financial Management 10, 295-319.
Ferris, J.S., 1981. A transactions theory of trade credit use. Quarterly Journal of Economics 96, 242270.
Fisman, R., Love, I., 2003. Trade credit, financial intermediary development, and industry growth.
Journal of Finance 58, 353-374.
Frank, M., Maksimovic, V., 2005. Trade credit, collateral, and adverse selection. Unpublished
manuscript, University of Maryland.
Freedom House, 2014. Annual freedom in the world survey: Political rights and civil liberties
rating 1972-2014 & press freedom index 1994-2014. http://www.freedomhouse.org.
Ge, Y., Qiu, J., 2007. Financial development, bank discrimination and trade credit. Journal of
Banking & Finance 31, 513-530.
Giannetti, M., Burkart, M., Ellingsen, T., 2011. What you sell is what you lend? Explaining trade
credit contracts. Review of Financial Studies 24, 1261-1298.
Gorodnichenko, Y., Roland, G., 2011. Individualism, innovation, and long-run growth.
Proceedings of the National Academy of Sciences 108, 21316-21319.
24
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, A., 1994. Cultural beliefs and the organization of society: A historical and theoretical
reflection on. Journal of Political Economy 102, 912.
Greif, A., Milgrom, P., Weingast, B.R., 1994. Coordination, commitment, and enforcement: The
case of the merchant guild. Journal of Political Economy 102, 745-776.
Guiso, L., Sapienza, P., Zingales, L., 2006. Does culture affect economic outcomes? Journal of
Economic Perspectives 20, 23-48.
Han, S., Kang, T., Salter, S., Yoo, Y.K., 2010. A cross-country study on the effects of national
culture on earnings management. Journal of International Business Studies 41, 123-141.
Hofstede, G., 1984. Culture's consequences: International differences in work-related values.
SAGE Publications, Newbury Park, CA.
Hofstede, G., 2001. Culture's consequences: Comparing values, behaviors, institutions, and
organizations across nations. SAGE Publications, Thousand Oaks, CA.
Inglehart, R., Baker, W.E., 2000. Modernization, cultural change, and the persistence of traditional
values. American Sociological Review, 19-51.
Jain, N., 2001. Monitoring costs and trade credit. Quarterly Review of Economics and Finance 41,
89-110.
Johnson, S., McMillan, J., Woodruff, C., 2002. Courts and relational contracts. Journal of Law,
Economics, and Organization 18, 221-277.
Kalcheva, I., Lins, K.V., 2007. International evidence on cash holdings and expected managerial
agency problems. Review of Financial Studies 20, 1087-1112.
Khurana, I.K., Martin, X., Pereira, R., 2006. Financial development and the cash flow sensitivity
of cash. Journal of Financial and Quantitative Analysis 41, 787-808.
Kirkman, B.L., Lowe, K.B., Gibson, C.B., 2006. A quarter century of cultures consequences: A
review of empirical research incorporating Hofstede's cultural values framework. Journal of
International Business Studies 37, 285-320.
Klapper, L., Laeven, L., Rajan, R., 2012. Trade credit contracts. Review of Financial Studies 25,
838-867.
Kusnadi, Y., Wei, K.C.J., 2011. The determinants of corporate cash management policies: Evidence
from around the world. Journal of Corporate Finance 17, 725-740.
25
Kwok, C.C., Tadesse, S., 2006. National culture and financial systems. Journal of International
Business Studies 37, 227-247.
La Porta, R., Lopez-de-Silanes, F., Shleifer, A., 2008. The economic consequences of legal origins.
Journal of Economic Literature 46, 285-332.
Landa, J.T., 1981. A theory of the ethnically homogeneous middleman group: An institutional
alternative to contract law. Journal of Legal Studies 10, 349-362.
Landa, J.T., 2008. The bioeconomics of homogeneous middleman groups as adaptive units:
Theory and empirical evidence viewed from a group selection framework. Journal of
Bioeconomics 10, 259-278.
Lee, Y.W., Stowe, J.D., 1993. Product risk, asymmetric information, and trade credit. Journal of
Financial and Quantitative Analysis 28, 285-300.
Licht, A.N., Goldschmidt, C., Schwartz, S.H., 2005. Culture, law, and corporate governance.
International Review of Law and Economics 25, 229-255.
Lins, K.V., Servaes, H., Tufano, P., 2010. What drives corporate liquidity? An international survey
of cash holdings and lines of credit. Journal of Financial Economics 98, 160-176.
Long, M.S., Malitz, I.B., Ravid, S.A., 1993. Trade credit, quality guarantees, and product
marketability. Financial Management, 117-127.
Longhofer, S.D., Santos, J.A., 2003. The paradox of priority. Financial Management, 69-81.
Lonner, W., Adamopoulos, J., 1997. Culture as antecedent of behavior, in: J. Berry, Poortinga, Y.,
Pandey, J. (Eds.), Handbook of Cross-Cultural Psychology. Allyn & Bacon, MA, pp. 44-82.
Love, I., Preve, L.A., Sarria-Allende, V., 2007. Trade credit and bank credit: Evidence from recent
financial crises. Journal of Financial Economics 83, 453-469.
McMillan, J., Woodruff, C., 1999. Interfirm relationships and informal credit in Vietnam.
Quarterly Journal of Economics 114, 1285-1320.
Meltzer, A.H., 1960. Mercantile credit, monetary policy, and size of firms. The Review of
Economics and Statistics 42, 429-437.
Mian, S.L., Smith, C.W., Jr., 1992. Accounts receivable management policy: Theory and evidence.
Journal of Finance 47, 169-200.
Molina, C.A., Preve, L.A., 2009. Trade receivables policy of distressed firms and its effect on the
costs of financial distress. Financial Management 38, 663-686.
Ng, C.K., Smith, J.K., Smith, R.L., 1999. Evidence on the determinants of credit terms used in
interfirm trade. Journal of Finance 54, 1109-1129.
26
Nilsen, J.H., 2002. Trade Credit and the Bank Lending Channel. Journal of Money, Credit and
Banking 34, 226-253.
North, D.C., 1990. Institutions, institutional change, and economic performance. Cambridge
University Press, Cambridge, UK.
North, D.C., 1991. Institutions. Journal of Economic Perspectives 5, 97-112.
Petersen, M.A., Rajan, R.G., 1997. Trade credit: Theories and evidence. Review of Financial
Studies 10, 661-691.
Pinkowitz, L., Stulz, R.M., Williamson, R., 2003. Do firms in countries with poor protection of
investor rights hold more cash? National Bureau of Economic Research, Working Paper No. 10188.
Pinkowitz, L., Stulz, R.M., Williamson, R., 2013. Is there a US high cash holdings puzzle after the
financial crisis? , Fisher College of Business Working Paper No. 2013-03-07
Roberts, M.R., Whited, T.M., 2013. Endogeneity in empirical corporate finance, in: Constantinides,
G.M., Harris, M., Stulz, R.M. (Eds.), Handbook of the Economics of Finance. Elsevier, Oxford, UK.
Schwartz, R.A., 1974. An economic model of trade credit. Journal of Financial and Quantitative
Analysis 9, 643-657.
Shao, L., Kwok, C.C., Guedhami, O., 2010. National culture and dividend policy. Journal of
International Business Studies 41, 1391-1414.
Shao, L., Kwok, C.C., Zhang, R., 2013. National culture and corporate investment. Journal of
International Business Studies 44, 745-763.
Smith, J.K., 1987. Trade credit and informational asymmetry. Journal of Finance 42, 863-872.
Spolaore, E., Wacziarg, R., 2009. The diffusion of development. Quarterly Journal of Economics
124, 469-529.
Stulz, R.M., Williamson, R., 2003. Culture, openness, and finance. Journal of Financial Economics
70, 313-349.
Tang, L., Koveos, P.E., 2008. A framework to update Hofstede's cultural value indices: Economic
dynamics and institutional stability. Journal of International Business Studies 39 1045-1063.
Tosi, H.L., Greckhamer, T., 2004. Culture and CEO compensation. Organization Science 15, 657670.
Weber, M., 1905 [2001]. The Protestant Ethic and the Spirit of Capitalism. Routledge Classic,
London.
Williamson, O.E., 2000. The new institutional economics: Taking stock, looking ahead. Journal of
Economic Literature 38, 595-613.
27
Wilner, B.S., 2000. The exploitation of relationships in financial distress: The case of trade credit.
Journal of Finance 55, 153-178.
World Bank, 2012. World development indicators, The World Bank, Washington, DC.
Zheng, X., El Ghoul, S., Guedhami, O., Kwok, C.C., 2012. National culture and corporate debt
maturity. Journal of Banking & Finance 36, 468-488.
Zheng, X., El Ghoul, S., Guedhami, O., Kwok, C.C., 2013. Collectivism and corruption in bank
lending. Journal of International Business Studies 44, 363-390.
28
Appendix
Variables
Definition
Panel A. Dependent Variables, Cultural Variables, and Openness
Receivables/sales
100 times the ratio of accounts receivable due to trade to total sales. Accounts receivable
from trade equals amounts on open account (net of applicable reserves) owed by customers
for goods and services sold in the ordinary course of business.
Receivables/ assets
100 times the ratio of accounts receivable due to trade to the book value of total assets.
Payables/ assets
100 times the ratio of accounts payable to the book value of total assets.
CLT
A cultural index of collectivism, which is equal to 100 minus Hofstede’s cultural index of
Individualism. A higher value of CLT indicates greater collectivism.
UAI
Hofstede’s cultural index on Uncertainty Avoidance.
PDI
Hofstede’s cultural index on Power Distance.
MAS
Hofstede’s cultural index on Masculinity.
Dummy variable equal to 1 if the country’s CLT is above median (based on the 52 countries
DCLT
included in the main sample).
Dummy variable equal to 1 if the country’s PDI is above median (based on the 60 countries
DPDI
included in the main sample).
Dummy variable equal to 1 if the country’s UAI is above median (based on the 68 countries
DUAI
included in the main sample).
Dummy variable equal to 1 if the country’s MAS is above median (based on the 53 countries
DMAS
included in the main sample).
CLT_TK
100 minus Tang and Koveos’ updated cultural index of Individualism
Sources
Compustat
As above
As above
Hofstede (2001)
As above
As above
As above
As above
As above
As above
As above
Export/GDP
Tang and Koveos
(2008)
As above
As above
As above
World Bank
national accounts
data, & OECD
National Accounts
data files.
As above
Profits/sales
Cash/assets
Fixed assets/assets
Sales growth
Computed
Compustat
As above
As above
As above
As above
UAI_TK
PDI_TK
MAS_TK
Trade/GDP
Tang and Koveos’ updated cultural index of Uncertainty Avoidance
Tang and Koveos’ updated cultural index of Power Distance
Tang and Koveos’ updated cultural index of Masculinity
The natural logarithm of trade (the sum of exports and imports of goods and services) as a
share of GDP.
The natural logarithm of exports of goods and services as a share of GDP, where exports of
goods and services represent the value of all goods and other market services provided to the
rest of the world. They include the value of merchandise, freight, insurance, transport, travel,
royalties, license fees, and other services, such as communication, construction, financial,
information, business, personal, and government services. They exclude compensation of
employees and investment income (formerly called factor services) and transfer payments.
Panel B. Firm-Level Control Variables
Log(assets)
The natural logarithm of total assets in $US millions.
Ratio of income before extraordinary items to total sales.
Ratio of cash and short-term investments to total assets.
Ratio of total (net) property plant and equipment to the book value of total assets.
Growth ratio in total sales in year t, defined as the ratio of the difference between total sale
in year t and total sales in year t-1 to total sales in year t-1.
Gross profit margin/sales
Ratio of the difference between total sales and cost of goods sold to total sales.
Finished inv/total inv
Percentage of inventory which is finished goods.
Panel C. Country-Level Control Variables and Variables for Robustness Tests
GDP per capita
The natural logarithm of GDP per capita in constant 2000 U.S. dollars.
Private Credit
Creditor rights
Information Sharing
Law and Order
Political Rights Index
Common Law
Private credit by deposit money banks divided by GDP.
Index of creditor rights, which measures four powers of secured lenders in bankruptcy
granted by a country’s laws and regulations. The index ranges from 0 (weak creditor rights)
to 4 (strong creditor rights).
Dummy variable equal to 1 if either a public registry or a private bureau operates in the
country in a given year, and 0 otherwise.
Assessment of the law and order tradition in the country. The index is time-varying and
measures the degree to which the citizens of a country are willing to rely on legal institutions
to adjudicate disputes. The range for the index is from 0 to 6, with higher values indicating
more reliance on the legal system
An index of political rights. The ratings are determined by a survey including ten political
rights questions, which are grouped into three subgroups regarding the electoral process,
political pluralism and participation, and the functioning of the government. This index is
time-varying and ranges from 1 (most free) to 7 (least free).
Dummy variable equal to 1 if a country’s legal origin is English law, and 0 if the legal origin
is French, German, or Scandinavian civil law.
from
As above
As above
World Development
Indicators (2012),
World Bank
Beck et al. (2000)
Djankov
et
al.
(2007)
As above
International
Country Risk Guide
(2008)
Freedom
(2014)
House
La Porta
(2008)
et
29
al.
Trust
Catholic
Genetic Distance_CLT
Genetic Distance_PDI
Genetic Distance_UAI
Genetic Distance_MAS
Percentage of participants in each country who answer that most people can be trusted for the
following question in the World values Survey (WVS): “Generally speaking, would you say
that most people can be trusted or that you need to be very careful in dealing with people.”
The values in WVS wave 1990, 1995, and 2000 are used for the periods 1990–1995, 1996–
1999, and 1999–2010, respectively.
Percentage of population adhering to the Catholic.
FST distance relative to the Venezuela (the most collectivistic country in our sample),
which is the probability that two alleles at a given locus selected at random from the
population of a given country and population of the Venezuela will be different
(based on dominant population of a country). A higher FST is associated with larger
genetic difference from Venezuela.
FST distance relative to Malaysia (the country with largest power distance in our
sample), which is the probability that two alleles at a given locus selected at random
from the population of a given country and population of the Malaysia will be
different (based on dominant population of a country). A higher FST is associated
with larger genetic difference from Malaysia.
FST distance relative to Greece (the country with highest uncertainty avoidance in
our sample), which is the probability that two alleles at a given locus selected at
random from the population of a given country and population of the Greece will
be different (based on dominant population of a country). A higher FST is associated
with larger genetic difference from Greece.
FST distance relative to Japan (the most masculine country in our sample), which is
the probability that two alleles at a given locus selected at random from the
population of a given country and population of the Japan will be different (based
on dominant population of a country). A higher FST is associated with larger genetic
difference from Japan.
Barro and McCleary
(2003)
As above
Spolaore
and
Wacziarg (2009)
As above
As above
As above
30
Figure 1. Trade Credit around the World
Receivables/Sales
Receivables.sales
Receivables/Sales
9.7-12.6
12.7-15.5
15.6-16.3
16.4-16.5
16.6-16.7
16.8-17.0
17.1-17.4
17.5-17.8
17.9-18.7
18.8-19.6
19.7-20.3
20.4-22.4
22.5-23.3
23.4-27.7
27.8-34.0
31
Figure 2a. Trade credit and collectivism
Figure 2b. Trade credit and power distance
Figure 2c. Trade credit and uncertainty avoidance
Figure 2d. Trade credit and masculinity
32
# of unique firms
Receivables/sales
Log(assets)
Profits/sales
Cash/assets
Fixed
assets/assets
Sales growth
Gross profit
margin/sales
GDP per capita
Private Credit
CLT
PDI
UAI
MAS
Country
Argentina
Australia
Austria
Bangladesh
Belgium
Brazil
Bulgaria
Canada
Chile
China
Colombia
Czech Rep
Denmark
Estonia
Finland
France
Germany
Greece
Hong Kong
Hungary
India
Indonesia
Ireland
Israel
Italy
Japan
Korea, Rep.
Luxembourg
Malaysia
Mexico
Morocco
Netherlands
New Zealand
Norway
Pakistan
Peru
Philippines
Poland
Portugal
Russian Fed
Singapore
N
Table 1. Summary Statistics-by Country
813
11,580
1,099
206
1,383
3,749
77
11,328
241
23,344
318
178
1,483
167
1,814
8,356
9,424
1,986
2,411
200
24,504
3,394
525
1,958
3,128
47,485
6,620
301
8,121
1,351
484
1,936
825
1,260
2,065
861
1,196
1,943
707
1,488
5,409
72
1,687
127
59
147
372
15
1,789
126
2,437
40
30
182
18
155
936
958
234
336
28
2,495
355
79
280
331
3,691
992
40
979
126
54
215
125
238
247
80
150
357
76
210
690
16.33
16.29
17.28
18.91
18.81
19.62
20.53
19.71
21.74
23.66
18.01
11.19
17.14
9.71
16.5
23.4
16.67
34.3
20.89
17.92
24.5
16.77
17.25
21.92
32.2
22.18
18.75
14.42
26.84
16.39
32.21
16.8
15.57
18.43
11.54
13.94
22.97
20.01
22.62
12.39
23.45
5.46
3.11
5.92
3.99
5.83
6.08
5.48
4.76
6.07
5.21
6.44
6.1
5.15
4.31
5.68
5.51
5.27
5.32
5.41
5.54
3.35
4.59
5.59
4.77
6.28
6.1
5.94
6.77
4.2
6.87
4.66
6.2
4.24
4.94
3.86
4.93
4.48
3.86
6.18
6.83
4.33
-0.05
-4.16
-0.14
0.145
-0.11
-0.24
0.145
-0.94
-0.13
0.016
0.121
0.056
-0.18
0.085
-0.03
-0.13
-0.16
-0.04
-0.25
-0.07
-0.08
-0.12
-0.54
-0.4
-0.15
0.002
-0.02
-0.17
-0.05
0.016
0.093
-0.11
-0.77
-0.24
0.015
0.076
-0.39
-0.04
-0.02
-0.03
-0.08
0.07
0.24
0.13
0.11
0.13
0.11
0.06
0.14
0.08
0.18
0.08
0.07
0.15
0.11
0.13
0.15
0.15
0.09
0.23
0.1
0.07
0.11
0.18
0.25
0.11
0.17
0.14
0.13
0.13
0.09
0.09
0.1
0.11
0.17
0.08
0.07
0.12
0.1
0.06
0.09
0.19
0.49
0.31
0.32
0.5
0.3
0.4
0.42
0.42
0.43
0.34
0.45
0.63
0.32
0.44
0.29
0.18
0.25
0.36
0.26
0.48
0.35
0.4
0.32
0.2
0.26
0.3
0.36
0.36
0.35
0.48
0.28
0.27
0.37
0.29
0.46
0.51
0.38
0.34
0.36
0.5
0.28
0.21
0.74
0.16
0.36
0.12
0.25
0.21
0.38
0.12
0.26
0.22
0.1
0.16
0.26
0.09
0.11
0.15
0.07
0.3
0.22
0.33
0.29
0.25
0.24
0.11
0.04
0.18
0.27
0.17
0.19
0.14
0.13
0.38
0.35
0.27
0.16
0.26
0.28
0.03
0.43
0.22
0.373
-0.24
0.646
0.4
0.629
0.376
0.416
0.078
0.373
0.319
0.425
0.561
0.523
0.352
0.689
0.61
0.633
0.306
0.37
0.524
0.283
0.3
0.362
0.42
0.629
0.279
0.242
0.496
0.359
0.411
0.395
0.59
0.389
0.727
0.217
0.376
0.299
0.275
0.584
0.361
0.336
9.039
10.08
10.1
6.27
10.05
8.294
7.702
10.04
8.793
7.297
7.953
8.742
10.28
8.621
10.1
9.995
10.06
9.49
10.38
8.526
6.403
6.838
10.26
9.931
9.865
10.55
9.531
10.82
8.447
8.663
7.381
10.08
9.595
10.53
6.389
7.831
7.101
8.572
9.328
7.879
10.27
15.071
104.17
105.32
38.698
78.944
35.586
38.955
99.679
64.634
107
27.767
44.499
105.15
65.558
70.238
93.999
109.69
78.798
150.07
38.893
36.941
26.135
156.97
86.135
85.679
135.3
86.712
141.02
107.1
17.904
58.636
141.74
124.49
68.294
23.687
21.326
28.873
31.016
130.26
31.322
95.431
54
10
45
80
25
62
70
20
77
80
87
42
26
40
37
29
33
65
75
20
52
86
30
46
24
54
82
40
74
70
54
20
21
31
86
84
68
40
73
61
80
49
36
11
80
65
69
70
39
63
80
67
57
18
40
33
68
35
60
68
46
77
78
28
13
50
54
60
40
104
81
70
38
22
31
55
64
94
68
63
93
74
86
51
70
60
94
76
85
48
86
30
80
74
23
60
59
86
65
112
29
82
40
48
35
81
75
92
85
70
36
82
68
53
49
50
70
87
44
93
104
95
8
56
61
79
55
54
49
40
52
28
66
64
57
16
30
26
43
66
57
57
88
56
46
68
47
70
95
39
50
50
69
53
14
58
8
50
42
64
64
31
36
48
33
South Africa
2,520
313
16.94
4.82
-0.06
0.13
0.3
0.25
0.335
8.138
68.391
35
49
49
63
Spain
1,960
177
28.12
6.62
0.044
0.09
0.35
0.12
0.784
9.575
124.14
49
57
86
42
Sweden
3,126
481
17.46
4.33
-0.63
0.17
0.2
0.36
0.414
10.3
86.739
29
31
29
5
Switzerland
3,077
262
16.73
6.15
-0.19
0.16
0.34
0.14
0.617
10.47
156.86
32
34
58
70
Thailand
5,093
497
17.67
4.34
-0.03
0.1
0.41
0.14
0.289
7.741
104.79
80
64
64
34
Turkey
1,420
188
20.7
5.55
0.046
0.11
0.35
0.24
0.281
8.5
26.619
63
66
85
45
U.K.
20,761
2,437
17.55
4.74
-0.54
0.15
0.29
0.3
0.417
10.14
143.34
11
35
35
66
U.S.
100,519
11,914
15.44
4.93
-0.74
0.18
0.29
0.29
0.177
10.42
50.983
9
40
46
62
Venezuela
139
19
16.72
5.61
-0.02
0.07
0.51
0.23
0.349
8.526
12.115
88
81
76
73
Vietnam
1,072
250
16.81
3.51
0.085
0.14
0.3
0.3
0.239
6.531
94.897
80
70
30
40
Total
335,405
38,096
19.34
4.99
-0.47
0.16
0.31
0.25
0.282
9.538
85.235
36.95
52.49
55.24
62.4
This table reports averages by country of the dependent variable (Receivables/sales), firm- and country-level control variables, as well as Hofstede’s (2001) four cultural dimensions. The main
sample consists of 335,405 firm-year observations from 51 countries over the 1992 to 2012 period. Definitions and data sources for all variables are provided in the Appendix. All time-varying
variables are lagged by one year except for Receivables/sales, Receivables/assets, and Payables/ assets.
34
Table 2. Summary Statistics-Full Sample
Variable
N
Mean
Std. Dev.
Min
P25
Median
P50
Max
Receivables/sales
335,405
19.34
17.16
0.00
9.38
16.02
24.37
119.71
Receivables/assets
335,405
16.89
13.31
0.00
6.23
14.51
24.31
61.17
Payables/ assets
335,083
11.65
11.32
0.00
4.03
8.41
15.52
70.59
CLT
335,405
36.95
27.10
9.00
9.00
32.00
54.00
88.00
PDI
335,405
52.49
18.39
11.00
40.00
40.00
68.00
104.00
UAI
335,405
55.24
22.27
8.00
40.00
46.00
75.00
112.00
MAS
335,405
62.35
17.17
5.00
56.00
62.00
66.00
95.00
Log(assets)
335,405
4.99
2.15
-1.26
3.57
4.97
6.39
10.10
Profits/sales
335,405
-0.47
3.05
-29.56
-0.01
0.03
0.07
0.63
Cash/assets
335,405
0.16
0.18
0.00
0.03
0.09
0.21
0.91
Fixed assets/assets
335,405
0.31
0.23
0.00
0.12
0.27
0.46
0.92
Sales growth
335,405
0.25
0.88
-0.98
-0.03
0.08
0.25
6.65
Gross profit margin/sales
335,405
0.28
0.84
-7.55
0.19
0.31
0.51
1.00
Finished inv/total inv
240,952
0.33
0.33
0.00
0.00
0.25
0.57
1.03
GDP per capita
335,405
9.54
1.38
5.75
8.77
10.23
10.48
10.94
Private Credit
335,405
85.24
44.08
6.63
47.65
81.79
108.74
237.58
CLT_TK
305,945
27.01
29.29
0.00
0.00
23.00
40.00
94.00
PDI_TK
305,945
50.72
18.37
34.00
34.00
46.00
72.00
86.00
UAI_TK
305,945
32.26
22.24
12.00
12.00
26.00
42.00
88.00
MAS_TK
305,945
54.91
9.31
20.00
55.00
57.00
57.00
89.00
Creditor right
334,937
1.88
1.00
0.00
1.00
2.00
3.00
4.00
Information sharing
335,405
0.91
0.28
0.00
1.00
1.00
1.00
1.00
Law and Order
335,403
5.05
0.97
1.00
4.50
5.00
6.00
6.00
Political Rights
332,994
1.84
1.76
1.00
1.00
1.00
2.00
7.00
Common Law
334,731
0.59
0.49
0.00
0.00
1.00
1.00
1.00
Trust
294,621
0.38
0.10
0.03
0.36
0.36
0.43
0.66
Cath
332,994
0.17
0.22
0.00
0.01
0.13
0.18
0.94
Genetic Distance_CLT
335,405
0.06
0.05
0.00
0.00
0.04
0.12
0.18
Genetic Distance_PDI
335,405
0.10
0.03
0.00
0.10
0.12
0.12
0.19
Genetic Distance_UAI
335,405
0.06
0.04
0.00
0.02
0.05
0.11
0.15
Genetic Distance_MAS
335,405
0.09
0.05
0.00
0.05
0.12
0.13
0.22
This table reports the number of observations, mean, standard deviation, minimum, median, and maximum for the variables used in this paper.
The main sample consists of 335,405 firm-year observations from 51 countries over the 1992 to 2012 period. Variable definitions are provided
in the Appendix.
35
GDP per capita
Gross profit margin/sales
Sales growth
Fixed assets/assets
Cash/assets
Profits/sales
Log(assets)
MAS
UAI
PDI
CLT
Payables/ assets
Receivables/assets
Receivables/sales
Variables
Table 3. Correlation
Receivables/assets
0.4573*
1
Payables/ assets
0.0515*
0.4403*
1
CLT
0.1490*
0.0339*
0.0346*
1
PDI
0.1606*
0.0140*
0.0229*
0.7881*
1
UAI
0.0470*
0.1438*
0.1019*
0.2005*
-0.0251*
1
MAS
0.0375*
0.0980*
0.0983*
0.0185*
-0.0570*
0.4122*
1
Log(assets)
-0.0827* -0.0973* -0.1053*
0.0625*
-0.0375*
0.2437*
0.1548*
1
Profits/sales
-0.0525*
0.1567*
-0.0213*
0.1365*
0.1111*
0.0541*
0.0220*
0.2200*
1
Cash/assets
0.0072*
-0.1192* -0.0987* -0.0800* -0.0986* -0.0316*
0.0605*
-0.1835* -0.2247*
1
Fixed assets/assets
-0.1907* -0.3725* -0.1983*
0.1085*
0.0948*
0.0050*
-0.0259*
0.1896*
0.0410*
-0.3942*
1
Sales growth
-0.0046* -0.0614* -0.0184* -0.0648* -0.0324* -0.0928* -0.0601* -0.1104* -0.0567*
0.1098*
-0.0364*
1
Gross profit
margin/sales
-0.0162*
0.0847*
-0.0599*
0.0600*
0.0356*
0.0368*
-0.0313*
0.1459*
0.6341*
-0.1807*
0.0308*
-0.0487*
1
GDP per capita
-0.1059*
0.0589*
0.0327*
-0.6118* -0.7312*
0.2660*
0.2794*
0.1877*
-0.0743*
0.1474*
-0.1347* -0.0215* -0.0262*
1
Private Credit
0.0796*
0.0794*
0.0869*
0.2034*
-0.0508*
0.2266*
0.3519*
0.1630*
0.0064*
0.0603*
-0.0513* -0.0551*
0.0338*
0.2642*
This table presents Pearson pairwise correlation coefficients between all variables in the main analysis. The main sample consists of 335,405 firm-year observations from 51 countries over the 1992
to 2012 period. Variable definitions are provided in the Appendix. All time-varying variables are lagged by one year except for Receivables/sales, Receivables/assets, and Payables/ assets. *
indicates that the correlation is significant at the 1% level or better.
36
Table 4. The Influence of Culture on Trade Receivables
(1)
(2)
(3)
(4)
(5)
(6)
Hofstede's
VARIABLES Baseline
CLT
PDI
UAI
MAS
four
dimensions
CLT
0.104***
0.026***
(27.883)
(5.232)
PDI
0.185***
0.145***
(32.060)
(20.820)
UAI
0.070***
0.039***
(20.441)
(9.537)
MAS
0.057***
0.019***
(14.828)
(4.440)
Log(assets)
-0.122*** -0.259*** -0.221*** -0.238*** -0.157*** -0.311***
(-3.241) (-6.867) (-5.925) (-6.267) (-4.184)
(-8.226)
Profits/sales
-0.442*** -0.501*** -0.497*** -0.451*** -0.463*** -0.513***
(-12.092) (-13.608) (-13.521) (-12.305) (-12.641)
(-13.856)
Cash/assets
-10.084*** -11.366*** -10.950*** -10.052*** -10.402*** -11.169***
(-25.559) (-28.679) (-27.826) (-25.665) (-26.244)
(-28.280)
Fixed
-15.314*** -16.288*** -15.690*** -15.507*** -15.509*** -16.023***
assets/assets
(-38.283) (-40.495) (-39.519) (-38.792) (-38.734)
(-40.032)
Sales growth -0.259*** -0.151*** -0.165*** -0.173*** -0.229*** -0.100**
(-5.075) (-2.961) (-3.244) (-3.393) (-4.491)
(-1.973)
Gross profit
0.332*** 0.420*** 0.457*** 0.343*** 0.436***
0.492***
margin/sales
(2.799)
(3.522)
(3.833)
(2.893)
(3.648)
(4.100)
GDP per capita -2.140*** -0.769*** -0.310*** -2.408*** -2.270***
Private Credit
Constant
(-31.210)
0.049***
(32.772)
39.304***
(33.631)
(-9.013)
0.027***
(17.628)
26.202***
(20.357)
(-3.519)
0.040***
(27.589)
13.620***
(9.616)
(-34.903)
0.045***
(29.677)
39.516***
(34.206)
(-32.415)
0.043***
(28.036)
37.778***
(32.467)
(7)
VARs
DCLT
Cultural
VARs
dummies
(8)
Alternative
cultural
measures
0.026***
(2.667)
0.120***
(11.841)
0.102***
(16.889)
0.052***
(7.735)
-0.259***
(-6.631)
-0.458***
(-12.403)
-9.946***
(-25.374)
2.943*** CLT_TK
(12.175)
DPDI
3.344*** PDI_TK
(12.088)
DUAI
2.812*** UAI_TK
(13.017)
DMAS
0.481*** MAS_TK
(2.576)
Log(assets)
-0.309*** Log(assets)
(-8.138)
Profits/sales
-0.507*** Profits/sales
(-13.721)
Cash/assets
-10.994*** Cash/assets
(-27.799)
Fixed
Fixed
-16.019***
-15.347***
assets/assets
assets/assets
(-39.831)
(-36.783)
Sales growth
-0.117** Sales growth
0.007
(-2.298)
(0.130)
Gross profit
Gross profit
0.471***
0.186
margin/sales
margin/sales
(3.917)
(1.567)
GDP per
-0.557*** GDP per capita -0.763***
0.450**
capita
(-5.482)
(-7.110)
(2.139)
0.032*** Private Credit 0.033*** Private Credit 0.016***
(20.253)
(21.372)
(9.450)
15.535*** Constant
26.343*** Constant
6.714***
(10.802)
(18.279)
(2.701)
Year Dum
YES
YES
YES
YES
YES
YES
Year Dum
YES Year Dum
YES
FF48 Dum
YES
YES
YES
YES
YES
YES
FF48 Dum
YES FF48 Dum
YES
Obs.
335,405 335,405 335,405 335,405 335,405
335,405 Obs.
335,405 Obs.
305,945
Adj. R-sq
0.142
0.153
0.158
0.149
0.145
0.162
Adj. R-sq
0.161 Adj R-sq
0.160
This table reports results of regressing trade credit on measures of culture as well as firm- and country-level controls. The dependent variable is
Receivables/sales, defined as 100 times the ratio of trade receivables to total sales. All variables are defined in the Appendix. Year and industry
dummies based on the Fama-French 48-industry classification are included in all regressions but are unreported for brevity. All parameters are
estimated using pooled OLS with standard errors adjusted for heteroskedasticity and clustered at the firm level. Z-statistics (in parentheses) are
reported beneath each coefficient estimate. Significance at the 10%, 5%, and 1% level is indicated by *, **, and ***, respectively.
37
Table 5. Alternative Dependent Variables
(1)
VARIABLES
CLT
(2)
(3)
Receivables/assets
CLT
0.080***
(18.588)
UAI
MAS
Profits/sales
Cash/assets
Fixed assets/assets
Sales growth
Gross profit margin/sales
GDP per capita
Private Credit
Constant
-0.668***
(-24.501)
0.579***
(42.716)
-21.242***
(-77.268)
-21.681***
(-74.898)
-0.525***
(-19.204)
-0.595***
(-12.037)
1.200***
(19.015)
0.018***
(15.591)
9.203***
(9.167)
-0.755***
(-27.679)
0.593***
(43.769)
-20.831***
(-76.239)
-21.736***
(-75.799)
-0.469***
(-17.220)
-0.636***
(-12.856)
0.106**
(2.393)
0.017***
(14.979)
20.554***
(27.345)
(6)
(7)
Payables/ assets
0.038***
(9.534)
UAI
0.054***
(16.227)
-0.658***
(-24.142)
0.583***
(43.264)
-21.168***
(-76.538)
-21.703***
(-74.905)
-0.537***
(-19.637)
-0.551***
(-11.320)
0.286***
(6.384)
0.016***
(13.516)
18.874***
(24.450)
0.047***
(18.571)
MAS
Log(assets)
Profits/sales
Cash/assets
Fixed assets/assets
Sales growth
Finished inv/total inv
GDP per capita
Private Credit
Constant
(8)
0.021***
(7.540)
PDI
0.079***
(30.485)
-0.705***
(-25.864)
0.568***
(42.060)
-21.620***
(-79.004)
-22.090***
(-76.310)
-0.502***
(-18.375)
-0.597***
(-12.113)
1.213***
(19.920)
0.009***
(7.560)
12.625***
(14.303)
(5)
VARIABLES
0.061***
(20.870)
PDI
Log(assets)
(4)
-0.474***
(-14.739)
-0.259***
(-8.172)
-11.967***
(-38.777)
-13.270***
(-42.158)
-0.001
(-0.016)
0.111
(0.701)
0.298***
(5.302)
0.018***
(14.876)
11.182***
(14.952)
-0.462***
(-14.358)
-0.256***
(-8.098)
-11.804***
(-38.646)
-13.182***
(-42.154)
-0.003
(-0.075)
0.103
(0.651)
0.412***
(6.852)
0.020***
(17.198)
8.436***
(9.673)
-0.542***
(-16.848)
-0.253***
(-8.002)
-11.913***
(-39.252)
-12.767***
(-41.348)
0.056
(1.447)
0.619***
(3.956)
-0.220***
(-4.908)
0.022***
(19.260)
14.201***
(21.926)
0.058***
(18.176)
-0.488***
(-15.214)
-0.254***
(-8.075)
-11.830***
(-38.791)
-13.321***
(-42.625)
0.022
(0.574)
0.713***
(4.564)
-0.100**
(-2.277)
0.016***
(13.813)
12.005***
(18.075)
Year Dum
YES
YES
YES
YES
Year Dum
YES
YES
YES
YES
FF48 Dum
YES
YES
YES
YES
FF48 Dum
YES
YES
YES
YES
# of Obs
335,405
335,405
335,405
335,405
# of Obs
240,921
240,921
240,921
240,921
Adj. R-sq
0.344
0.343
0.352
0.342
Adj. R-sq
0.173
0.174
0.181
0.180
This table reports results of regressing trade credit on measures of culture as well as firm- and country-level controls. In Models (1) through (4), the dependent variable is Receivables/assets, defined as
100 times the ratio of trade receivables to the book value of total assets. In Models (5) through (8), the dependent variable is Payables/assets, defined as 100 times the ratio of accounts payable to the
book value of total assets. All variables are defined in the Appendix. Year and industry dummies based on the Fama-French 48-industry classification are included in all regressions but are unreported
for brevity. All parameters are estimated using pooled OLS with standard errors adjusted for heteroskedasticity and clustered at the firm level. Z-statistics (in parentheses) are reported beneath each
coefficient estimate. Significance at the 10%, 5%, and 1% level is indicated by *, **, and ***, respectively.
38
Table 6. Additional Controls-Trust, Religion, Legal, and Political Institutions
VARIABLES
CLT
(1)
CLT
0.115***
(16.168)
PDI
(2)
PDI
(3)
UAI
0.177***
(19.068)
UAI
0.052***
(7.786)
MAS
Log(assets)
Profits/sales
Cash/assets
Fixed assets/assets
Sales growth
Gross profit margin/sales
GDP per capita
Private Credit
Additional Controls
Common Law
Creditor Rights
Information Sharing
Law and Order
Political Rights Index
Trust
Cath
Constant
(4)
MAS
-0.208***
(-5.115)
-0.552***
(-12.168)
-10.207***
(-24.729)
-16.205***
(-37.435)
-0.108*
(-1.914)
0.424***
(3.032)
-0.121
(-0.762)
0.046***
(21.663)
-0.231***
(-5.662)
-0.544***
(-12.032)
-10.052***
(-24.398)
-15.729***
(-36.660)
-0.112**
(-1.974)
0.412***
(2.947)
-0.348**
(-2.274)
0.040***
(18.943)
-0.244***
(-5.978)
-0.530***
(-11.700)
-10.028***
(-24.262)
-15.679***
(-36.315)
-0.117**
(-2.064)
0.367***
(2.615)
-0.824***
(-5.401)
0.044***
(20.635)
0.050***
(12.014)
-0.239***
(-5.845)
-0.536***
(-11.834)
-10.103***
(-24.405)
-15.639***
(-36.257)
-0.109*
(-1.918)
0.387***
(2.762)
-1.217***
(-7.934)
0.041***
(19.065)
-0.534*
(-1.679)
-0.663***
(-6.058)
-5.848***
(-12.938)
-0.401***
(-2.687)
-0.226***
(-2.888)
-2.164*
(-1.769)
2.447***
(3.770)
28.787***
(13.875)
-2.095***
(-9.052)
0.088
(0.734)
-2.342***
(-5.045)
-0.203
(-1.397)
-0.456***
(-5.615)
-1.825
(-1.498)
0.271
(0.454)
21.608***
(9.768)
-2.170***
(-6.775)
-0.308***
(-2.683)
-5.682***
(-11.896)
-0.508***
(-3.416)
0.572***
(6.282)
-4.806***
(-3.774)
0.009
(0.016)
37.341***
(18.454)
-4.115***
(-21.835)
-0.606***
(-5.527)
-5.400***
(-11.665)
-0.116
(-0.773)
0.173**
(2.231)
-9.595***
(-8.841)
-0.365
(-0.618)
43.314***
(24.651)
Year Dum
YES
YES
YES
YES
FF48 Dum
YES
YES
YES
YES
Observations
293,972
293,972
293,972
293,972
Adj. R-squared
0.169
0.169
0.165
0.165
This table presents results of regressions that re-estimate Models (2) to (5) of Table 4 by adding potential omitted variables to control for formal
and informal institutional environment. The dependent variable is Receivables/sales, defined as 100 times the ratio of trade receivables to total
sales. All variables are defined in the Appendix. Year and industry dummies based on the Fama-French 48-industry classification are included
in all regressions but are unreported for brevity. All parameters are estimated using pooled OLS with standard errors adjusted for
heteroskedasticity and clustered at the firm level. Z-statistics (in parentheses) are reported beneath each coefficient estimate. Significance at the
10%, 5%, and 1% level is indicated by *, **, and ***, respectively.
39
Table 7. Instrumental Variables Regression of Culture on Trade Credit
VARIABLES
CLT
PDI
UAI
(1)
CLT
0.111***
(21.790)
(2)
PDI
(3)
UAI
(4)
MAS
0.101***
(9.274)
0.272***
(16.718)
MAS
0.183***
(22.066)
Log(assets)
-0.269***
-0.176***
-0.574***
-0.235***
(-7.096)
(-4.712)
(-11.869)
(-6.103)
Profits/sales
-0.506***
-0.472***
-0.478***
-0.511***
(-13.686)
(-12.853)
(-12.800)
(-13.822)
Cash/assets
-11.461***
-10.556***
-9.961***
-11.100***
(-28.726)
(-26.848)
(-24.743)
(-27.607)
Fixed assets/assets
-16.360***
-15.519***
-16.065***
-15.937***
(-40.423)
(-39.097)
(-37.714)
(-39.371)
Sales growth
-0.143***
-0.208***
0.077
-0.163***
(-2.794)
(-4.061)
(1.402)
(-3.190)
Gross profit margin/sales
0.426***
0.400***
0.376***
0.664***
(3.571)
(3.360)
(3.127)
(5.454)
GDP per capita
-0.667***
-1.141***
-3.184***
-2.552***
(-7.005)
(-9.057)
(-33.609)
(-35.143)
Private Credit
0.026***
0.044***
0.032***
0.028***
(15.612)
(30.223)
(19.038)
(16.403)
Constant
25.233***
25.286***
40.132***
34.437***
(18.583)
(13.108)
(34.021)
(29.195)
Year Dum
YES
YES
YES
YES
FF48 Dum
YES
YES
YES
YES
Observations
335,405
335,405
335,405
335,405
Adj. R-squared
0.153
0.155
0.0909
0.132
F-test of excluded instruments
0.000***
0.000***
0.000***
0.000***
The table presents IV estimation results of regressing trade credit on measures of culture as well as firmand country-level controls. The dependent variable is Receivables/sales, defined as 100 times the ratio of
trade receivables to total sales. The instruments in the first-stage regressions are Genetic Distance_CLT,
Genetic Distance_PDI, Genetic Distance_UAI, and Genetic Distance_MAS for CLT, PDI, UAI, and MAS,
respectively, in Models (1) through (4). All controls used in the second stage are controlled for in the firststage regression. All variables are defined in the Appendix. Year and industry dummies based on the FamaFrench 48-industry classification are included in all regressions but unreported for brevity. All parameters
are estimated using pooled OLS with standard errors adjusted for heteroskedasticity and clustered at the
firm level. Z-statistics (in parentheses) are reported beneath each coefficient estimate. Significance at the
10%, 5%, and 1% level is indicated by *, **, and ***, respectively.
40
Table 8. Robustness Checks-Alternative Estimation Methods and Sample Compositions
(1)
VARIABLES
CLT
PDI
UAI
MAS
Log(assets)
Profits/sales
Cash/assets
Fixed
assets/assets
(2)
(3)
4)
(5)
(6)
(7)
(8)
Pooled country-year sample
WLS
0.048**
0.125***
(2.547)
(28.958)
0.163***
0.249***
(8.975)
(32.091)
0.046***
0.112***
(3.162)
(26.253)
0.052***
0.147***
(3.809)
(24.818)
0.709*
0.184
0.495
0.716*
0.092** 0.099** 0.115** 0.116**
(1.797)
(0.510)
(1.214)
(1.843)
(1.993)
(2.145)
(2.494)
(2.511)
-0.051
-0.263
0.264
0.101
-0.329*** -0.324*** -0.320*** -0.321***
(-0.114) (-0.607)
(0.595)
(0.228)
(-8.998) (-8.881) (-8.801) (-8.822)
-12.927 -17.811** -3.488
-12.416 -7.119*** -7.018*** -6.668*** -6.816***
(-1.543) (-2.307) (-0.458) (-1.546) (-18.121) (-17.873) (-17.068) (-17.386)
(9)
(10)
(11)
(12)
(13)
Tobit
0.107***
(28.561)
(14)
(15)
Industrial Firms
(16)
0.114***
(27.130)
0.189***
(32.441)
0.186***
(29.293)
0.071***
(20.662)
0.076***
(19.766)
0.058***
(14.786)
-0.216*** -0.176*** -0.194*** -0.111*** -0.212*** -0.152*** -0.194***
(-5.627) (-4.633) (-4.998) (-2.888)
(-4.863) (-3.525) (-4.392)
-0.450*** -0.445*** -0.397*** -0.409*** -0.481*** -0.476*** -0.417***
(-11.275) (-11.177) (-9.981) (-10.296) (-10.052) (-9.984) (-8.799)
-12.049*** -11.600*** -10.671*** -11.027*** -10.788*** -10.380*** -9.204***
(-29.079) (-28.211) (-26.078) (-26.632) (-22.191) (-21.484) (-19.125)
0.087***
(20.724)
-0.123***
(-2.839)
-0.448***
(-9.440)
-9.729***
(-20.055)
-26.824*** -22.671***-25.250*** -25.400*** -12.963*** -12.720*** -12.778*** -12.772*** -16.588*** -15.960*** -15.772*** -15.769*** -17.936*** -17.046*** -16.784***-16.517***
Sales growth
(-5.921)
0.179
(0.180)
(-5.342)
0.302
(0.350)
(-5.873)
0.240
(0.249)
(-5.858)
0.183
(0.181)
Gross profit
margin/sales
-0.073
1.326
-0.026
0.608
(-28.099) (-27.649) (-27.766) (-27.709) (-40.295) (-39.294) (-38.560) (-38.490) (-34.988) (-33.827) (-33.064) (-32.527)
0.174*** 0.169*** 0.180*** 0.164*** -0.202*** -0.218*** -0.226*** -0.284*** -0.257*** -0.276*** -0.259*** -0.302***
(2.949)
(2.863)
(3.058)
(2.793)
(-3.770) (-4.064) (-4.201) (-5.289)
(-3.646) (-3.916) (-3.681) (-4.285)
0.322*** 0.319*** 0.313*** 0.319*** 0.581*** 0.617*** 0.499*** 0.597***
0.368**
0.414***
0.242
0.422***
(-0.051)
(0.987)
(-0.019)
(0.423)
(2.656)
(2.630)
(2.581)
(2.631)
(4.380)
(4.652)
(3.778)
(4.480)
(2.408)
(2.714)
(1.598)
(2.754)
GDP per capita -1.066*** 0.056 -1.779*** -1.755*** -0.911*** -0.079 -3.039*** -2.866*** -0.768*** -0.316*** -2.456*** -2.313*** -0.512*** -0.170* -2.314*** -2.158***
(-2.818)
(0.158)
(-5.136) (-5.104)
(-9.771) (-0.730) (-38.320) (-36.282) (-8.944) (-3.563) (-35.189) (-32.684) (-5.439) (-1.750) (-30.282) (-27.780)
Private Credit 0.027*** 0.031*** 0.034*** 0.028*** 0.025*** 0.043*** 0.037*** 0.036*** 0.028*** 0.041*** 0.046*** 0.043*** 0.036*** 0.051*** 0.058*** 0.052***
(3.565)
(4.805)
(4.498)
(3.546)
(15.127) (26.874) (22.548) (22.067) (17.550) (27.715) (29.878) (28.263) (20.405) (30.409) (33.001) (29.718)
Constant
30.647*
13.084 38.120** 34.707** 26.307*** 7.708*** 43.439*** 37.786*** 25.866*** 13.186*** 39.600*** 37.847*** 18.908*** 8.047*** 32.985*** 30.374***
(1.780)
(0.788)
(2.289)
(2.076)
(14.664) (3.796) (25.626) (22.292) (19.954)
(9.239) (34.000) (32.271) (18.482) (6.704) (38.564) (36.267)
Year Effect
YES
YES
YES
YES
YES
YES
YES
YES
YES
YES
YES
YES
YES
YES
YES
YES
FF48 Effect
YES
YES
YES
YES
YES
YES
YES
YES
YES
YES
YES
YES
YES
YES
YES
YES
Observations
910
910
910
910
335,405 335,405 335,405 335,405 335,405 335,405 335,405 335,405
213,120 213,120 213,120 213,120
Adj.(Pseudo)
0.308
0.396
0.311
0.312
0.180
0.180
0.181
0.178
0.0198
0.0204
0.0191
0.0185
0.145
0.149
0.139
0.136
R-sq
This table reports robustness tests on the relation between Hofstede’s (2001)[?] four cultural dimensions (CLT, PDI, UAI, and MAS) and trade credit by re-estimating Models (2) to (5) of Table 4 using
different estimation methods and subsamples. The dependent variable is Receivables/sales, defined as 100 times the ratio of trade receivables to total sales. All variables used are defined in the Appendix.
The country-level regressions in Models (1) to (4) control for year dummies and the proportion of firms in each industry based on the Fama-French 48-industry classification. Year and industry dummies
based on the Fama-French 48-industry classification are included in Models (5) through (16). All year and industry controls are unreported for brevity. Models (1) to (4) report pooled OLS results using
910 country-year observations with standard errors adjusted for heteroskedasticity. Models (5) to (8) present WLS results of a regression of Receivables/sales on the cultural indices and firm- and countrylevel control variables, with weights equal to the inverse of the number of firm-year observations in each country. Models (9) to (12) present regression results using Tobit estimation, with a pseudo Rsquared reported instead of an adjusted R-squared. Models (13)–(16) use the subsample of industrial firms with SIC code between 2000 and 4999. All standard errors in Models (5) through (16) are
adjusted for heteroskedasticity and clustered at the firm level. Z-statistics (in parentheses) are reported beneath each coefficient estimate. Significance at the 10%, 5%, and 1% level is indicated by *, **,
and ***, respectively.
41
Table 9. Robustness Checks-Subsample Periods
(1)
VARIABLES
CLT
(2)
(3)
4)
(5)
1992-2002
(7)
(8)
2003-2012
0.089***
0.101***
(14.031)
PDI
(6)
(25.819)
0.195***
0.178***
(21.940)
UAI
(28.433)
0.066***
0.067***
(12.392)
MAS
(17.767)
0.034***
0.059***
(5.762)
-0.053
(-1.080)
(-0.694)
(-0.558)
(1.150)
(-9.441)
(-8.667)
(-8.737)
(-7.612)
Profits/sales
-0.568***
-0.592***
-0.546***
-0.578***
-0.478***
-0.463***
-0.419***
-0.421***
(-8.627)
(-8.963)
(-8.303)
(-8.742)
(-11.049)
(-10.736)
(-9.764)
(-9.796)
Cash/assets
-8.073***
-7.792***
-7.691***
-7.587***
-13.626***
-13.197***
-11.808***
-12.510***
(-15.638)
(-15.141)
(-14.965)
(-14.724)
(-27.266)
(-26.648)
(-23.980)
(-25.102)
-15.792***
-15.246***
-15.509***
-15.628***
-16.734***
-16.109***
-15.725***
-15.707***
(-29.621)
(-28.683)
(-29.015)
(-29.377)
(-35.021)
(-34.233)
(-33.270)
(-33.166)
-0.004
0.021
0.004
-0.037
-0.264***
-0.291***
-0.313***
-0.358***
Fixed assets/assets
Sales growth
-0.034
-0.028
0.056
(14.103)
Log(assets)
-0.407***
-0.372***
-0.379***
-0.331***
(-0.063)
(0.296)
(0.061)
(-0.518)
(-3.798)
(-4.185)
(-4.501)
(-5.146)
Gross profit margin/sales
0.563***
0.703***
0.485***
0.638***
0.394***
0.380**
0.320**
0.360**
(3.102)
(3.853)
(2.683)
(3.456)
(2.636)
(2.555)
(2.159)
(2.420)
GDP per capita
-1.370***
-0.794***
-2.881***
-2.782***
-0.564***
-0.057
-2.153***
-1.958***
(-9.363)
(-5.899)
(-27.443)
(-25.855)
(-6.391)
(-0.609)
(-29.205)
(-26.327)
Private Credit
0.038***
0.046***
0.049***
0.055***
0.021***
0.033***
0.042***
0.035***
(15.892)
(24.121)
(21.908)
(24.396)
(9.960)
(16.221)
(20.698)
(16.950)
30.692***
16.916***
43.196***
42.416***
24.475***
11.967***
37.518***
36.134***
Constant
(15.409)
(8.329)
(24.976)
(24.193)
(18.180)
(7.786)
(31.137)
(29.850)
Year Dum
YES
YES
YES
YES
YES
YES
YES
YES
FF48 Dum
YES
YES
YES
YES
YES
YES
YES
YES
132,933
132,933
132,933
132,933
202,472
202,472
202,472
202,472
0.181
0.187
0.181
0.177
0.144
0.149
0.138
0.134
Observations
Adj. R-sq
This table reports robustness tests for our findings on the relation between four cultural dimensions (CLT, PDI, UAI, and MAS) and trade credit
provision by re-estimating Models (2) to (5) of Table 4 using two different subsample periods. The dependent variable is Receivables/sales,
defined as 100 times the ratio of trade receivables to total sales. All variables are defined in the Appendix. Year and industry dummies based on
the Fama-French 48-industry classification are included in all regressions but are unreported for brevity. All parameters are estimated using
pooled OLS with standard errors adjusted for heteroskedasticity and clustered at the firm level. Z-statistics (in parentheses) are reported beneath
each coefficient estimate. Significance at the 10%, 5%, and 1% level is indicated by *, **, and ***, respectively.
42
Table 10. Openness, Culture and Trade Credit
VARIABLES
CLT
(1)
(2)
(3)
0.197***
(11.686)
PDI
0.663***
(19.714)
0.452***
(27.163)
0.592***
(21.393)
0.834***
6.089***
5.958***
9.708***
(3.609)
(13.961)
(24.416)
(21.709)
-0.022***
-0.104***
-0.095***
-0.125***
(-5.413)
(-14.067)
(-22.878)
(-18.461)
Interaction:
Export/GDP*Culture
Profits/sales
Cash/assets
Fixed assets/assets
Sales growth
Gross profit
margin/sales
GDP per capita
Private Credit
Constant
(7)
(8)
0.404***
(29.223)
Export/GDP
Log(assets)
(6)
0.594***
(22.609)
MAS
Interaction:
Trade/GDP*Culture
(5)
0.186***
(14.390)
UAI
Trade/GDP
(4)
0.475***
(20.456)
0.844***
(4.168)
5.972***
(15.552)
6.033***
(25.892)
8.904***
(20.556)
-0.103***
(-15.466)
-0.305***
(-7.982)
-0.498***
(-13.554)
-10.826***
(-27.632)
-15.581***
(-39.256)
-0.131***
(-2.583)
-0.101***
(-24.543)
-0.229***
(-6.060)
-0.491***
(-13.345)
-10.678***
(-27.192)
-16.044***
(-40.120)
-0.123**
(-2.428)
-0.114***
(-16.945)
-0.151***
(-4.024)
-0.509***
(-13.837)
-10.865***
(-27.508)
-16.200***
(-40.258)
-0.169***
(-3.309)
0.407***
(3.406)
0.685***
(6.701)
0.029***
(17.501)
-16.231***
(-7.222)
0.383***
(3.203)
-2.412***
(-33.489)
0.023***
(13.160)
21.771***
(15.741)
0.487***
(4.059)
-2.231***
(-30.821)
0.013***
(6.881)
8.725***
(4.803)
-0.290***
(-7.593)
-0.500***
(-13.544)
-11.327***
(-28.476)
-16.286***
(-40.472)
-0.140***
(-2.744)
-0.303***
(-7.925)
-0.497***
(-13.510)
-10.877***
(-27.734)
-15.586***
(-39.259)
-0.136***
(-2.674)
-0.229***
(-6.058)
-0.488***
(-13.267)
-10.612***
(-27.030)
-16.009***
(-40.031)
-0.127**
(-2.500)
-0.150***
(-3.996)
-0.512***
(-13.902)
-10.847***
(-27.456)
-16.199***
(-40.266)
-0.165***
(-3.250)
-0.024***
(-6.255)
-0.291***
(-7.636)
-0.499***
(-13.541)
-11.300***
(-28.403)
-16.284***
(-40.465)
-0.139***
(-2.720)
0.413***
(3.435)
-0.630***
(-7.120)
0.023***
(12.803)
22.344***
(13.982)
0.408***
(3.412)
0.616***
(5.958)
0.030***
(18.185)
-20.010***
(-7.556)
0.373***
(3.125)
-2.424***
(-33.989)
0.023***
(13.557)
18.020***
(12.046)
0.504***
(4.190)
-2.252***
(-31.470)
0.013***
(6.833)
-0.098
(-0.047)
0.412***
(3.422)
-0.611***
(-6.946)
0.023***
(12.646)
22.754***
(15.622)
FF48
YES
YES
YES
YES
YES
YES
YES
YES
Year
YES
YES
YES
YES
YES
YES
YES
YES
Observations
335,405
335,405
335,405
335,405
335,405
335,405
335,405
335,405
Adjust R-squared
0.154
0.161
0.158
0.152
0.154
0.162
0.159
0.152
This table presents regression results of how openness, measured by Trade/GDP and Export/GDP, influences the relationship between
culture and trade credit. The dependent variable is Receivables/sales, defined as 100 times the ratio of accounts receivable due to trade to total
sales. Trade/GDP is defined as the natural logarithm of trade (the sum of exports and imports of goods and services) as a share of GDP. Export/GDP
is defined as the natural logarithm of exports of goods and services as a share of GDP, All other variables are defined in the Appendix. Year
dummies and industry dummies based on Fama-French 48-industry classification are included in all regressions yet unreported for brevity. All
parameters are estimated by pooled OLS regression with standard errors adjusted for heteroskedasticity and clustered at the firm level. Z-statistics
(in parenthesis) are reported beneath each coefficient estimate Significance at the level of 10%, 5%, and 1% is indicated by *, **, ***, respectively.
43