Modern Economy, 2011, 2, 707-716
doi:10.4236/me.2011.25079 Published Online November 2011 (http://www.SciRP.org/journal/me)
Copyright © 2011 SciRes. ME
707
The Determinants of Trade Credit: Evidence from Indian
Manufacturing Firms
Rajendra R. Vaidya
Indira Gandhi Institute of Development Research, Mumbai, India
E-mail: vaidya@igidr.ac.in, vaidya_1962@yahoo.com
Received July 5, 2011; revised August 10, 2011; accepted August 20, 2011
Abstract
Trade credit (accounts receivable and accounts payable) is both an important source and use of funds for
manufacturing firms in India. This paper empirically investigates the determinants of trade credit in the In-
dian context. The empirical evidence presented suggests that strong evidence exists in support of an inven-
tory management motive for the existence of trade credit. Highly profitable firms both give and receive less
trade credit. Firms with greater access to bank credit offer less trade credit to their customers. On the other
hand, firms with higher bank loans receive more trade credit. Holdings of liquid assets have a positive influ-
ence on both accounts receivable and accounts payable.
Keywords: Trade Credit, Inventories
1. Introduction
Trade credit (measured by accounts receivable and ac-
counts payable in the balance sheet of a firm) is an ar-
rangement that allows firms to buy goods or services
without making an immediate payment. It thus allows the
separation of the exchange of goods and money over
time. It is well recognized that trade credit is likely to be
a very expensive source of credit1. Trade credit (with
respect to both the amounts and terms) varies substan-
tially across firms and industries and a substantial body
empirical research exists that attempts to explain this
variation.
Many theories have been put forward to explain the
existence of trade credit. Trade credit may be used as a
source of funds if raising capital th rough other sou rces is
more expensive. Price discrimination being illegal in
many countries, firms may choose to discriminate be-
tween buyers using trade credit. Some firms may choose
to make early payments to take advantage of discounts
while others may have an incentive to pay towards the
end of the credit period. Suppliers may have some fund-
ing advantage over banks in evaluating and controlling
credit risk. If suppliers are likely to interact much more
closely and more often with buyers compared to banks
then this is likely to give them a better idea of the busi-
ness prospects that the buyer faces. If the good supplied
cannot be resold by the buyer then the supplier could
hold of the threat of stopping supplies if payments are
not made in time. Suppliers may also have an advantage
over banks with respect to repossessing and reselling the
goods supplied in case of default. Trade c redit may arise
as a financial response to variable demand. Trade credit
can be seen an outcome of interaction between product
and financial markets which arises because it provides
the seller with an advantage in inventory management.
Sellers can reduce their finished good inventories by
offering trade credit. When business conditions are bad
(i.e. inventories pile up) firms may choose to postpone
payments for raw materials purchased. Trade credit may
also enable firms to lower transactions costs.
At an empirical level most studies relate accounts
payable and accounts receivable to various accounting
ratios and firm and industry characteristics. A few stud-
ies have attempted to examine variation s in the terms and
conditions of trade credit. Widely cited empirical stud ies
like Petersen and Raj an [1] and Ng, Smith and Smith [3]
have uncovered many empirical regularities but over-
whelming support or rejection for any particular theory
has as yet not been possible.
Trade credit has been generally recognized as an im-
portant component of corporate finance in many coun-
tries2. Recent data, from the Reserve Bank of India3,
shows that accounts receivable accounted for 10.86%
1Pertersen and Rajan [1] and Cunningham [2] report and effective
annualized interest rate upward s of 40%.
708 R. R. VAIDYA
and accounts payable accounted for 11.59% of total as-
sets/liabilities respectively in 2008 for a sample of large
public limited companies. The comparable figure of
short term bank credit was 10.75%. Evidently, in India,
trade credit is at least as important as bank credit. In
most advanced countries accounts receivables can be
easily collateralized. This makes it possible for firms to
obtain additional bank credit against their accounts re-
ceivables. Consequently, a firm providing trade credit
does not necessarily have to reduce its investment in
other avenues. In India banks have always been some-
what reluctant to lend against accounts receivable4. Bills
discounted accounted for less than one percent of total
credit advanced by Scheduled Commercial Banks in In-
dia as of March 20095. This institutional feature is lik ely
to have a significant impact on the determinants of trade
credit in India.
Unfortunately no systematic empirical evidence on the
determinants of trade credit in India is available. This
paper makes a small beginning in that direction. We do
not deal with the issue of terms and conditions of trade
credit due to lack of information in this regard in the In-
dian context. We estimate a model similar to Bougheas,
Mateut and Mizen [11] to study the determinants of trade
credit in India. It is found that trade credit arises essen-
tially as a financial response to variable demand and
variables suggested by other theories compliment this
basic explanation. In the next section a brief summary of
the existing theories of trade credit and empirical work
that seeks to explain inter firm differences is provided.
Section 3 outlines the empirical model u sed, data sour ces
and results. Section 4 concludes.
2. Theories of Trade Credit
Many reasons have been put forward to explain why
firms may offer or accept trade credit. We provide below
a short outline of the main arguments.
Metzler [12] was possibly the first to point out that
large firms use trade credit instead of direct price reduc-
tions to push sales in periods when monetary conditions
were tight. Further, he argued that firms would accumu-
late liquid balances in periods of loose monetary policy
and utilize these to extend trade credit in periods when
monetary conditions were tight. These macroeconomic
implications of trade credit have been recently further
investigated by Guariglia and Mateut [13] and Mateut,
Bougheas and Mizen [14] who conclude that in the UK
trade credit increases in periods when monetary policy is
tight and ban k l e ndi n g fal l s.
Brennan, Maksimovic and Zechner [15] argue that if
the product market is non-competitive an d there exists an
adverse selection problem in credit markets then this
makes price discrimination through trade credit poten-
tially profitable. Imperfections in the product market
allow sellers to use trade credit to discriminate between
buyers who have different reservation prices. When the
credit characteristics of firms to whom the supplier (who
has market power in the product market) is attempting to
sell cannot be observed by him, trade credit makes it
possible to provide incentives for firms to self select.
“Good firms” might find it profitable to buy on a cash
basis or repay as soon as possible (given the high cost of
trade credit) while risky firms may find it advantageous
of buy on credit because other source of funds may be
even more costly for this firm. An empirical implication
that arises from the price discrimination arguments is
that more profitable firms are more likely to grant more
trade credit.
The possibility that sellers who have easier access to
the capital market may have an incentive to offer trade
credit to their buyers (who may not have access to capital
markets on the same terms) was first pointed out by
Schwartz [16]. The supplier’s greater ability to raise
funds is used to pass credit to their customers. If banks
are the main source of credit then th is suggests that firms
offering trade credit would borrow from banks and pass
this on as accounts receivable (on their books of accounts)
to the buyers. Biais and Gollier [17] have pointed out
that in a situation where banks are forced to ration credit
(which arises due to adverse selection), trade credit can
transmit a seller’s private information to banks. If the
seller is willing to offer trade credit to a firm this tells th e
banks that the supplier has private information regarding
this firm which makes it credit worthy. This would lead
to a reduction of credit rationing. In addition, Jain [18]
has argued that suppliers may have a monitoring advan-
tage over banks because in the course of their transac-
tions with the firm they have access to information which
banks may not.
2Rajan and Zingales [4] report that trade credit accounted for 17.8% o
f
total assets of American firms in the early part of 1990s. Kohler, Brit-
ton and Yates [5] report that in the U.K. 70% of total short term debt
extended to firms and 55% of total short term credit received by firms
was in the form of trade credit. Uesugi and Yamashiro [6] report that
trade credit (accounts payable) accounted for 15.5 of total assets in
Japan. Deloof and Jegers [7] report that in 1995 accounts receivable
formed 16% of total assets and accounts payable formed 12% of total
liabilities of Belgian non financial firms.
3These figures are based on a sample of 1558 large public limited com-
p
anies published in the Reserve Bank of India Bulle ti n March 2010 [8].
4See Report of the Working Group on Discounting of Bills by Banks
[9], for a detailed discussion.
5Basic Statistical Returns of Scheduled Commercial Banks in India,
Reserve Bank of India, Mumbai, Ta ble no. 1. 14. [10].
Burkart and Ellingson [19] argue that this monitoring
advantage arises because of an intrinsic difference be-
tween inputs and cash. Inputs cannot be as easily (if at all)
be diverted as cash. It is the fear of diversion of funds
C
opyright © 2011 SciRes. ME
709
R. R. VAIDYA
that induces banks to restrict lending. Trade credit be-
comes a means to overcome a moral hazard problem
created by this possibility. The fact that the firm has re-
ceived trade credit sig nals that the firm has bough t inputs
that cannot be diverted and this opens up the possibility
that returns from investing would be higher than the re-
turns from diverting funds. Thus if a bank observes that a
firm is receiving trade credit it may be willing to lend.
Consequently, firms whose investments are constrained
by their access to external funds, trade credit and bank
credit may be compliments. Firms whose investments are
not constrained by availability of external funds the fact
that a firm has/or has not received trade credit is of no
consequence, and, bank credit and trade credit may be
substitutes. Even though firms can use accounts receiv-
able as collateral there would always be a ceiling on the
amount a bank would lend through this channel. Burkart
and Ellingsen [19] argue that “firms that are credit
constrained but highly profitable abstain from investing
in receivables, leaving the extension of trade credit to
firms either have better access to funds or are constrained
and relatively unprofitable (pp. 570).” This conclusion
would be reinforced in a context where banks do not
accept account receivable as collateral.
Cunat [20] argues that firms offering trade credit may
have an advantage over banks in enforcing debt repay-
ment in a situation where it is difficult for the buyer to
find alternative suppliers and it is costly for the seller to
find alternative customers. This condition would be met
if the product in question has some technological speci-
ficity. This advantage arises because suppliers can
threaten buyers with stoppage of su pplies of the interme-
diate good which in turn would hit production. Suppliers
would be in a position to help buyers overcome tempo-
rary liquidity shocks by offering trade credit. Lee and
Stowe [21] point out that trade credit when offered
represents an implicit product guarantee of the products
quality. The buyer is able to verify the quality of the
product before making a payment. In the presence of
information asymmetry large discounts (inducements to
make quick payments) would covey information on
quality. Firms, whose products are of a lower quality,
other things being equal, would offer large discounts.
From a transactions cost perspective, a supplier can
reduce inventory carrying costs if the buyer’s costs of
holding inventories are lower. Emery [22] argues that
trade credit arises as a financial response to variable de-
mand. Consider a situation where a firm experiences a
sudden dip in demand. The firm has two choices. Either
to accumulate costly inventories (which may or may not
be sold in later periods) or offer trade credit to its cus-
tomers who may be finance constrained. There clearly
exists a trade-off between holding inventories and offer-
ing trade credit. For trade credit to be a mutually benefi-
cial arrangement the firm offering trade credit must have
an advantage in bearing the financial cost (of the dip in
demand) but must be at a disadvantage in terms of the
operational cost for holding higher finished goods in-
ventories. The firm that accepts trade credit gains from
the fact that implicitly he receives a lower price (if the
payment is made within the stipulated period) and the
seller gains because of lower inventory costs. Bougheas,
Mateut and Mizen [11] incorporate this basic idea in a
formal two period model which incorporates the trade-
off between inventories and trade credit under conditions
of stochastic demand. Using this model they derive em-
pirically testable propositions with respect to accounts
payable and accounts receivable and their relationship
with changes in costs of inventories, profitability, risk
profile, liquidity position of firms and bank loans. They
show that:
a) firms with higher stock of inventories would have
lower accounts receivables and accounts payables.
b) profitability will be positively related to both ac-
counts payable and accounts receivable.
c) The relationship of accounts receivable and ac-
counts payable with riskiness of a firm and its liquidity
position is indeterminate.
d) Accounts receivables wo uld be positively related to
bank loans i.e. they are compliments. Accounts payable
can either be positively or negatively related to bank
loans.
The empirical literature has unearthed quite a few ro-
bust relationships between extent of trade credit offered
and received and various firm characteristics. A large
variety of variables measuring various firm characteris-
tics have been used to explain inter firm variations in
both accounts receivable and accounts payable.
A large number of papers [Petersen and Rajan [1],
Deloof and Jegers [7], Miwa and Ramseyer [23] and
Bougheas et al. [11], among others] report a positive
relationship between accounts payable and accounts re-
ceivables and size (usually measured by total assets or
log of total assets). Size is typically interpreted as re-
flecting the credit worthiness of the firm. Thus, larger
firms are seen to both receive and give more trade credit.
Profitability according to Petersen and Rajan [1] could
reflect a number of firm characteristics. Net profit could
be taken as a proxy for internal cash generation, and thus
one would expect profitable firms to extend more trade
credit. Surprisingly, they report a negative relationship
between net profits and accounts receivable. Gross prof-
its on the other hand would be an indicator of the incen-
tives to sell. If firms have the ability to discriminate be-
tween buyers through the use of trade credit (leading to
higher gross margins) then higher the gross profit the
Copyright © 2011 SciRes. ME
710 R. R. VAIDYA
higher the incentive to sell. They report a positive rela-
tionship between gross profits and accounts receivable.
Net profits are found to be negatively related to accounts
payable. As the firm’s ability to generate internal funds
increases its tendency to buy on credit decreases. Given
that trade credit is extremely expensive this is as ex-
pected. Deloof and Jegers [7] also report a negative rela-
tionship between net profits and accounts payable.
Bougheas et al. [11] find that profitability is positively
related to both accounts receivable and accounts payable.
This finding is interpreted as extra profit being channeled
to accounts receivable and more profitable firms being
more credit worthy receive more credit from their sup-
pliers.
Petersen and Rajan [1], report that firms who can se-
cure enough credit from institutional sources have lower
accounts payable and point to the possibility that trade
credit is a substitute for credit from financial institu tions.
Other papers like, Kohler, Britton and Yates [5] and Nil-
sen [24] using different data sets and time periods come
to a similar conclusion. Deloof and Jegers [7] using data
on Belgian firms provide persuasive evidence that short
term bank credit is a substitute for accounts payable. On
the other hand, Demirguc-Kunt and Maskimovic [25], in
a cross country setting empirically demonstrate that trade
credit is a compliment to lendin g by financial institu tions.
Cunningham [2] finds that for medium wealth firms (i.e.
those firms whose investment is less likely to be con-
strained by availability of external funs) trade credit and
bank credit are substitutes and for low wealth firms
(firms whose investments are more likely to be finance
constrained) trade credit and bank credit are compliments.
This paper provides strong support for the arguments put
forward by Burkart and Ellingson [19]. Bougheas et al.
[11] find that accounts receivable are compliments to
bank loans and accounts payab le are substitutes for bank
loans. This they argue is clearly indicative of the fact that
trade credit is more expensive than bank loan s and fits in
nicely with the pecking order hypothesis. Thus firms
who can borrow from banks are seen to pass on bank
credit to their buyers on the one hand and they take less
credit on the other.
Inventories have not been used as explanatory variable
in empirical studies of trade credit very often. Petersen
and Rajan [1] relate the ratio of finished goods invento-
ries to total inventories in the regression analysis with
respect to accounts payables and find a strong negative
relationship between the two. They argue that the ratio of
finished goods inventories to total inventories reflects the
“supplier’s advantage in liquidating the borrowers as-
sets”. If the ratio of finished goods inventories to total
inventories is large this reflects a lowering of the sup-
plier’s advantage in repossessing and selling supplied
goods because the buyer has transformed the raw mate-
rial supplied into fin ished goods. Both banks and suppli-
ers may face the same level of difficulty in selling re-
possessed finished goods. Thus accounts payable of
firms with a high ratio of finished goods inventories to
total inventories turn out to be lower. Cunat [20] uses
inventories as an explanatory variable while explaining
accounts payable of firms. He finds a significant and
positive relationsh ip. He argues th at acco unts payab le are
higher for firms with higher inventories because invento-
ries act as collateral. Bougheas et al. [11] relate finished
and semi finished goods inventories to both accounts
receivable and accounts payable. They find a strong
negative relationship between inventories and accounts
receivables. They interpret this as providing strong evi-
dence that firms use trade credit (i.e. allow buyers to
delay payment) to increase sales and thus reduce inven-
tories. Inventories turn out to be insignificant when re-
lated to accounts payab le.
A firms holding of liquid assets (cash and other short
term securities) has been used as a determinant of trade
credit in a number of papers. Van Horne [26] has argued
that firms adopt what is called the matching approach to
finance i.e. finance short term needs with short term fi-
nance. If such an approach is actually followed by firms
then accounts payable should have a positive relationship
with holding of liquid assets. Deloof and Jegers [7] find
that liquid assets have no influence on accounts payable
of Belgian firms. Cunat [20] reports a negative influence
of liquid assets on accounts payable. Cunat further shows
that when liquid assets fall, this is accompan ied by a rise
in accounts payable. This finding is interpreted as an
adjustment in accounts payable whenever there is an
unexpected liquidity shock. Bougheas et al. [11] use liq-
uid assets as an explanatory variable for both accounts
payable and accounts receivable. The holding of liquid
assets is assumed to have a direct relation to the cost of
extending trade credit bu t theoretically the expected sign
for this variable remains indeterminate. They report that
liquid assets have a negative and significant influence on
accounts receivable and a positive and significant influ-
ence on accounts payable.
3. The Empirical Model and Estimation
Results
3.1. The Empirical Model
A model very similar to Bougheas et al. [11] is estimated.
They explain trade credit extended (accounts receivable
divided by sales) and trade credit received (accounts
payable divided by sales) by the same set of explanatory
variables. The difference between accounts receivable
C
opyright © 2011 SciRes. ME
711
R. R. VAIDYA
and accounts payable (net trade credit) is considered, in
this paper, as an additional dependent variable which
shows whether the firm is a net receiver (if this variable
has a negative sign) or net giver of trade credit (if this
variable has a positive sign). The importance of this
variable becomes obvious once it is recognized that firms
typically are a part of a credit chain both receiving and
offering trade credit. The same set of dependent variables
is used to explain this difference as well. The main dif-
ference (apart from the fact that an additional variable,
net trade credit is considered) in the model estimated in
this paper and Bougheas et al. [11] lies in the treatment
of inventories. Bougheas et al. [11] define inventories as
the level of finished good s and work in progr ess invento-
ries while our data allows us to segregate inventories into
finished goods inventories on the one hand and semi fin-
ished goods and raw materials on the other. Finished
goods inventories are more likely to influence accounts
receivable (AR6) while semi finished goods and raw ma-
terial inventories are more likely to influence accounts
payable (AP7). By including them separately into the
analysis helps in isolating the influence of variable de-
mand (for the firm’s product) on accounts receivable and
accounts payable. Following Cunat [20] we include the
level of collateralizable assets (ratio of fixed assets to
total assets) as an explanatory variable. Firms having
higher collateralizable assets are expected to have easier
access to other sources of credit (including banks) and
thus would use less trade credit. Profitability (profits
before depreciation interest and taxes) divided by sales,
size (log of total assets), liquid assets8 and short term
bank loans are another standard explanatory variables
that we include9. The estimated equations take the fol-
lowing form10.
ARi,t/Salesi,t = αi + β1Stocksit/Salesit + β2Sizeit
+ β3Co llateralit + β4Prof itsit/salesit
+ β5 liquid assetsit/salesit
+ β6 short term bank loans it/salesit + eit
APi,t/Salesi,t = αi + γ1Stocksit/Salesit + γ2Sizeit
+ γ3Collateralit + γ4Profitsit/salesit
+ γ5 liquid assetsit/salesit
+ γ6 short term bank loansit/salesit + uit
(ARi,t – APi,t)/Salesi,t = αi + τ1Stocksit/Salesit + τ2Sizeit
+ τ3Collateralit + τ4Profitsit/salesit
+ τ5 liquid assetsit/salesit
+ τ6 short term bank loansit/salesit
+ νit
In two other specifications we replace stocksit by fin-
ished goods inventories and semi finished goods inven-
tories plus ra w m a t e ri al s inve n t ori e s.
αi, is a firm specific effect, βi, γi and τi are the coeffi-
cients and eit, uit and νit are the idiosyncratic error terms.
The equations are estimated using a first difference
GMM approach which controls for firm specific time
invariant effects and for possible endogeniety of regres-
sors11. Lags of all the independent variables are used as
instruments. Time dummies are included in all the re-
gressions.
We use data from the PROWESS database provided
by the Center for Monitoring the Indian Economy
(CMIE), Mumbai. This data base contains accounting
details of a very large number firms operating in India.
The data we use pertains to the 14 year period between
1993 and 2006. From this data base we chose firms
which met the following criteria.
(a) Firms with at least five years of continuous data.
(b) Firms whose ratio of manufacturing sales to total
sales was in excess of 75 percent for at least half th e
years for which data were available. This was done
to drop firms who had diversified into non manu-
facturing activities.
(c) Firms with a positive net worth for at least half the
number of years for which data were available. This
was done to drop firms in financial distress.
(d) Firms with accounts payable and accounts receiv-
able in excess of their total assets were not chosen.
This was again done with a view to excluding dis-
tressed firms.
(e) Firms needed to be in the private sector. All firms
owned by the central and state governments were
dropped.
These filters yielded an unbalanced panel of 1522
firms with an average of 10.66 years observations for
each firm. The descriptive statistics are provided in Ta-
ble 1. In general, the mean and medians of accounts re-
ceivable are far larger than accounts payable. This is also
reflected by the fact that net trade credit has a positive
mean and median. The firms in our sample thus, on av-
erage, give more trade credit than they receive.
6AR is measured by sundry debtors as reported in the PROWESS data-
base.
7AP is measured by sundry creditors as reported in the PROWESS
database.
8Liquid assets considered are cash, bank balances and marketable in-
vestment.
9Except size and collateralizable assets all other variables (both de-
p
endent and independent) a re normalized by sales.
10Bougheas et al. [11] include a variable measuring likelihood of com-
p
any failure which we do not.
3.2. Estimation Results
Tables 2, 3 and 4 report the empirical results with re-
spect to accounts receivable, accounts payable and the
ifference between accounts receivable and accounts d
11The estimation is performed in stata using the xtabond2 command
developed by Roodman [27].
Copyright © 2011 SciRes. ME
R. R. VAIDYA
Copyright © 2011 SciRes. ME
712
Table 1. Summary statistics (mean, standard deviation and median).
Bottom 25% Middle 50% Top 25% Whole sample
Accounts Receivable/sales Mean 0.357 0.213 0.175 0.240
Std. Dev 3.620 1.496 0.172 2.101
Median 0.184 0.165 0.144 0.162
Accounts Payable/sales Mean 0.210 0.157 0.157 0.170
Std. Dev 0.496 0.299 0.306 0.361
Median 0.120 0.116 0.125 0.120
(Accounts Receivable-Accounts payable)/sales Mean 0.147 0.057 0.018 0.070
Std. Dev 3.496 1.456 0.302 2.036
Median 0.050 0.042 0.018 0.038
Inventories/sales Mean 0.193 0.127 0.113 0.140
Std. Dev 0.846 0.262 0.218 0.476
Median 0.079 0.079 0.079 0.079
Finished good inventories/sales Mean 0.119 0.086 0.081 0.093
Std. Dev 0.438 0.230 0.195 0.290
Median 0.041 0.045 0.052 0.046
Raw material inventories/sales Mean 0.245 0.127 0.107 0.152
Std. Dev 2.614 0.317 0.176 1.331
Median 0.086 0.090 0.079 0.086
Fixed assets/Total ass ets Mean 0.637 0.636 0.670 0.645
Std. Dev 0.348 0.279 0.264 0.295
Median 0.601 0.631 0.670 0.634
Profit/sales Mean 0.035 0.101 0.159 0.099
Std. Dev 4.508 1.544 0.158 2.508
Median 0.087 0.120 0.149 0.121
Liquid Assets/sales Mean 0.363 0.088 0.100 0.160
Std. Dev 6.708 0.697 0.342 3.399
Median 0.028 0.027 0.043 0.031
Bank loans/sales Mean 0.351 0.205 0.154 0.229
Std. Dev 3.722 2.335 0.319 2.495
Median 0.105 0.120 0.105 0.113
Size Mean 2.023 3.822 5.916 3.894
Std. Dev 0.662 0.670 0.931 1.566
Median 2.111 3.805 5.716 3.803
Note: Firms are separated into size categories by using a dummy variable which takes the value of 1 in a given year if the firms total assets are in the top 25,
middle 50 and bottom 25 pe rcentile of the dist r ibution of total assets of all the firms in that year.
payable (net trade credit) respectively. Column 1 refers
to the specification wh ere total inven tories are used as an
independent variable and column 2 refers to the specifi-
cation where inventories are bifurcated into finished
goods and raw material inventories12.
The inventory to sales ratio is negatively (the coeffi-
cient is significant at 5%) related to accounts receivable.
When inventories are split into finished goods invento-
ries and raw material and semi finished inventories the
coefficient on finished goods inventories has a negative
sign and is significant at 1%. The coefficient of raw ma-
terial inventories turns out to be positive but insignificant.
713
R. R. VAIDYA
Table 2. Accounts receivable.
1 2
Inventories/sales –0.715**
(0.312)
Finished good inventor ie s/s al es –0.915***
(0.306)
Raw material inventories/sales 0.032
(0.041)
Fixed assets/total assets 0.622*
(0.376) 0.209
(0.481)
Profit/sales –0.455**
(0.195) –0.557***
(0.202)
Liquid assets/sales 0.833***
(0.013) 0.823***
(0.016)
Bank loans/sales –0.034
(0.022) –0.049**
(0.023)
Size 1.427***
(0.530) 1.013
(0.663)
No. of observations 11609 11609
m1(p) 0.00 0.00
m2 (p) 0. 281 0.231
Hansen/Sargan 0.958 0.988
Test statistics and standard errors (in parentheses) are asymptotically robust
to heteroscedasticity. (m2) is a test for first order serial correlation in levels,
asymptotically distributed as N(0,1) under the null of no serial correlation.
The Hansen/Sargan test is a test of over identifying restrictions distributed
as chi-square under the null of instrument validity. Both equations are esti-
mated using a GMM first difference specification. The instruments include
first and second lags of Inventory/sales, Finished good inventories/sales,
Raw material inventories/sales, Fixed assets/total assets, Profit/sales, Liquid
assets/sales, Bank loans/sales and Size. Time dummies were always in-
cluded as regressors and instruments. *, ** , and *** indicate significance at
10%, 5% and 1% respectively.
This indicates that firms with lower finished goods in-
ventories have higher accounts receivable and thus firms
offer more trade credit to boost sales and lower finished
goods inventories. Inventory management is thus an im-
portant motive for firms to offer trade credit to other
firms.
Profits have a negative coefficient which is significant
in both specifications. Profitable firms thus do not offer
higher trade credit. This finding is contrary to what is
found in the literature where generally speaking a posi-
Table 3. Accounts payable.
1 2
Inventory/sales 0.103***
(0.061)
Finished good inventor ie s/s al es 0.256***
(0.049)
Raw material inventories/sales 0.104***
(0.010)
Fixed assets/total assets –0.171
(0.131) –0.073
(0.118)
Profit/sales –0.079**
(0.034) –0.025
(0.027)
Liquid assets/sales 0.033***
(0.002) 0.034***
(0.001)
Bank loans/sales 0.024**
(0.011) 0.027**
(0.011)
Size –0.070
(0.172) –0.218
(0.158)
No. of observations 11609 11609
m1(p) 0.084 0.269
m2 (p) 0.333 0.374
Hansen/Sargan 0.410 0.870
Test statistics and standard errors (in parentheses) are asymptotically robust
to heteroscedasticity. (m2) is a test for first order serial correlation in levels,
asymptotically distributed as N(0,1) under the null of no serial correlation.
The Hansen/Sargan test is a test of over identifying restrictions distributed
as chi-square under the null of instrument validity. Both equations are esti-
mated using a GMM first difference specification. The instruments include
first and second lags of Inventory/sales, Finished good inventories/sales,
Raw material inventories/sales, Fixed assets/total assets, Profit/sales, Liquid
assets/sales, Bank loans/sales and Size. Time dummies were always in-
cluded as regressors and instruments. *, ** , and *** indicate significance at
10%, 5% and 1% respectively.
tive and significant coefficient is common [Petersen and
Rajan [1] and Bougheas et al. [11]. The result is con-
sistent with the Burk art and Ellingsen [19] argument that
profitable but finance constrained firms would prefer not
to offer trade credit. The relevance of this argument is
strengthened by the finding of a number of papers that
investment by firms in India is finance constrained13. The
negative coefficient also calls in question the relevance
of the price discrimination motive for offering trade
credit.
The coefficient of bank loans is negative and signifi-
cant in specification 2. Bougheas et al. [11] report a
positive and significant coefficient for bank loans. Bank
loans and accounts receivable turn out to be substitutes.
The fact that banks do not accept account receivables as
collateral could be driving this result. Clearly those firms
having access to bank finance do not pass this on as ac-
counts receivable to their customers.
12Following Bougheas et al. [11] an interaction term between size and
inventories was tried to reflect the influence of size on costs of holding
inventories. The coefficient for this interaction variable turned out to be
insignificant in all specifications and was dropped. A dummy variable
representing membership of industrial group was introduced and this
turned out to be insignificant in all specifications. In the initial phases
of the empirical investigation industry dummies of 32 NIC 2 digit level
industries were created. Very few of these dummies turned out to be
significant in any of the specifications. These were subsequently
dropped in later specifications.
13See Athey and Lumas [28], Ganesk-Kumar, Sen and Vaidya [29],
Lensink, Remco and Gangopadhay [30] amoung others.
Liquid assets have a positive and significant coeffi-
cient in both the specifications. This again is contrary to
Copyright © 2011 SciRes. ME
714 R. R. VAIDYA
Table 4. Accounts receivable-accounts payable.
1 2
Inventory/sales –0.354**
(0.172)
Finished good inventor ie s/s al es –1.185***
(0.287)
Raw material inventories/sales –0.065**
(0.038)
Fixed assets/total assets 0.713
(0.448) 0.249
(0.548)
Profit/sales –0.0.401*
(0.223) –0.558***
(0.202)
Liquid assets/sales 0.795***
(0.015) 0.790***
(0.015)
Bank loans/sales –0.056**
(0.023) –0.074***
(0.021)
Size 1.375**
(0.664) 1.039
(0.782)
No. of observations 11609 11609
m1(p) 0.377 0.352
m2 (p) 0. 661 0.128
Hansen/Sargan 0.997 0.998
Test statistics and standard errors (in parentheses) are asymptotically robust
to heteroscedasticity. (m2) is a test for first order serial correlation in levels,
asymptotically distributed as N(0,1) under the null of no serial correlation.
The Hansen/Sargan test is a test of over identifying restrictions distributed
as chi-square under the null of instrument validity. Both equations are esti-
mated using a GMM first difference specification. The instruments include
first and second lags of Inventory/sales, Finished good inventories/sales,
Raw material inventories/sales, Fixed assets/total assets, Profit/sales, Liquid
assets/sales, Bank loans/sales and Size. Time dummies were always in-
cluded as regressors and instruments. *, **, and *** indicate significance at
10%, 5% and 1% respectively.
what Petersen and Rajan [1] and Bougheas et al. [11]
report. The coefficient of collateralizable assets turns ou t
to be positive and significant at the 10 percent level in
specification 1 and insignificant in specification 2. Size
turns out to be positive and significant at 10 percent in
specification 1 and insignificant in specification 2. This
again is contrary to the general finding that large firms
offer more trade credit.
As regards accounts payable the coefficient of inven-
tories has a positive sign but is significant only at the 10
percent level in specification 1. When we bifurcate in-
ventories both finished goods and raw material invento-
ries have a positive and significant sign with the coeffi-
cient of finished goods inventories being much larger.
When firms pile up both types of inventories they take
more trade credit. Trade credit is thus offered to firms
who encounter a negative shock to sales. In this case too
profits turn out to be negative and significant. This is
contrary to what Bougheas et al. [11] find. More profit-
able firms thus neither offer nor take more trade credit.
Liquid assets have a positive and significan t sign and th is
time this result in line with Bougheas et al. [11]. This
turns out to be consistent with the Van Horne [26] view
of a matching approach to finance. Moreover possession
of liquid assets could signal an ability to pay back on
time.
The coefficient of bank loans has a positive and sig-
nificant sign. This is consistent with the Burkart and El-
lingsen [19] view that bank credit and trade credit would
be compliments for firms who are likely to face binding
finance constraints. This again is contrary to Bougheas et
al. [11] findings. Size does not turn out to be significant
in any of the specifications.
Turning to results for net trade credit (accounts re-
ceivable-accounts payable) the coefficient of total in-
ventories is negative and significant at 5 percent. Once
the inventories are bifurcated both finished goods and
raw material inventories have a negative and significant
coefficient. What is important is that the finished goods
inventories have a much larger coefficient. Thus it is
finished good inventories that are more influential in
determining net trade credit given. The coefficients of
profits and bank loans have a negative and significant
sign while liquid assets have a positive and significant
sign. Size turns out to be insignificant in both specifica-
tions.
4. Conclusions
The empirical evidence presented suggests that in the
Indian context strong evidence exists in support of an
inventory management motive for offering trade credit.
Firms attempt to increase sales and lower finished goods
inventories by offering trade credit both on a gross and
net basis. When inventories of finished goods and semi
finished goods and raw materials rise firms tend to post-
pone payments to their supplier and this shows up on
their books of accounts as higher accounts payable. This
is likely to help firms tide over negative shocks to sales.
Thus trade credit in general can be seen to arise as a fi-
nancial response to variable demand for their finished
goods. Highly profitable firms are found to both give (on
both net and gross basis) and receive less trade credit.
There could be many underlying results for this finding.
Firstly more profitable firms may not face a major prob-
lem with respect to variability of demand for their prod-
uct. The need to offer trade credit for inventory man-
agement is thus smaller. Moreover the need to accept
trade credit for such firms would also be lower, as in-
ventories would rarely he high. Secondly, as argued by
Burkart and Ellingsen [19], profitable but finance con-
strained firms would prefer not to offer trade credit. The
C
opyright © 2011 SciRes. ME
715
R. R. VAIDYA
fact that the coefficients of profitability are negative,
price discrimination does not seem to be a motive for the
existence of trade credit in India.
Firm’s holdings of liquid assets have a positive influ-
ence on accounts receivable and accounts payable and
net trade credit. Firms with greater access to bank credit
offer less trade credit to their customers. Firms with
more access to bank funds do not pass them on to their
buyers as accounts receivable. On the other hand, firms
with higher bank loans receive more trade credit. The
empirical results on the determinants of trade credit in
India are very different from those for advanced coun-
tries.
5. Acknowledgements
I would like to than k Dr. S. Chandrashekhar and Dr. Na-
veen Srinivasan for helpful discussions and Mr. Ankush
Agarwal for help with the GMM estimation. Mr. Ashish
Singh provided efficient research assistance.
6. References
[1] M. A. Petersen and R. G. Rajan, “Trade Credit: Theories
and Evidence,” The Review of Financial Studies, Vol. 10,
No. 3, 1997, pp. 661-691.
doi:10.1093/rfs/10.3.661
[2] R. Cunningham, “Trade Credit and Credit Rationing in
Canadian Firms,” Bank of Canada Working Paper 49, 2004.
[3] C. Ng, J. Smith and R. Smith, “Evidence on the Deter-
minants of Credit Terms in Inter firm Trade,” Journal of
Finance, Vol. 54, No. 6, 1999, pp. 1109-1129.
doi:10.1111/0022-1082.00138
[4] R. G. Rajan and L. Zingales, “What Do We Know About
Capital St ructure? Some Evide nce from International Data,”
Journal of Finance, Vol . 50 , No. 5, 1 99 5, p p. 1 4 21 -1 460.
doi:10.2307/2329322
[5] M. Kohler, E. Britton and T. Yates, “Trade Credit and the
Monetary Transmission Mechanism,” Bank of England
Working Paper ISSN 1368-5562, 2000.
[6] I. Uesugi and G. M. Yamashiro, “How Trade Credit Dif-
fers from Loans: Evidence from Japanese Trading Com-
panies,” RIETI Discussion Paper Series 04-E-028, 2004.
[7] M. Deloof and M. Jegers, “Trade Credit, Corporate
Groups and the Financing of Belgian Firms,” Journal of
Business Finance and Accounting, Vol. 26, No.7, 1999,
pp. 945-966. doi:10.1111/1468-5957.00281
[8] Reserve Bank of India, “Reserve Bank of India Bulletin,”
Mumbai, March 2010.
[9] Reserve Bank of India, “Report of the Working Group on
Discounting of Bills by Banks,” Mumbai, 2000.
[10] Reser ve Bank of India, “Basic Statistical Returns of Sche-
duled Commercial Banks in India,” Mumbai, 2010.
[11] S. Bougheas, S. Mateut and P. Mizen, “Corporate Trade
Cre d i t and Inventories: New Evi dence of a Trade-Off f rom
Accounts Payable and Receivable,” Journal of Banking
and Finance, Vo l. 33, No. 2 , 2009, pp. 300-307 .
doi:10.1016/j.jbankfin.2008.07.019
[12] A. H. Metzler, “Mercantile Credit, Monetary Policy and
Size of Firms,” The Review of Economics and Statistics,
Vol. 42, No.4, 1960, pp. 429-437. doi:10.2307/1925692
[13] A. Guariglia and S. Mateut, “Credit Channel, Trade Credit
Cha nn el and Inventory Investment: Evidence fro m a Panel
of U.K. Firms,” Journal of Banking and Finance, Vol. 30,
No. 10, 2006, pp. 2835-2856.
doi:10.1016/j.jbankfin.2005.11.002
[14] S. Mateut, S. Bougheas and P. Mizen, “Trade Credit,
Bank Lending and Monetary Policy Transmission,” Euro-
pean Economic Review, Vol. 50, No. 3, 2006, pp. 603-
629. doi:10.1016/j.euroecorev.2005.01.002
[15] M. J. Brennan, V. Maksimovic and J. Zechner, “Vendor
Financing,” Journal of Finance, Vol. 43, No. 5, 1988, pp.
1127-1141.
[16] R. Schwartz, “An Economic Model of Trade Credit,” Jour-
nal of Financial and Quantitative Analysis, Vol. 9, No. 4,
1974, pp. 643-657. doi:10.2307/2329765
[17] B. Biais and C. Gollier, “Trade Credit and Credit Ration-
ing,” Review of Financial Studies, Vol. 10, No. 4, 1997,
pp. 903-937. doi:10.1093/rfs/10.4.903
[18] N. Jain, “Monitoring Costs and Trade Credit,” Quarterly
Review of Economics and Finance, Vol. 41, No. 1, 2001,
pp. 89-110. doi:10.1016/S1062-9769(00)00063-6
[19] M. Burkart and T. Ellingson, “In-Kind Finance: A Theory
of Trade Credit,” American Economic Review, Vol. 94,
No. 3, 2004, pp. 569-590.
doi:10.1257/0002828041464579
[20] V. Cunat, “Trade Credit: Suppliers as Debt Collectors
and Insurance Providers,” The Review of Financial Stud-
ies, Vol. 20, No. 2, 2007, pp. 491-527.
doi:10.1093/rfs/hhl015
[21] Y. W. Lee and J. D. Stowe, “Product Risk, Asymmetric
Information and Trade Credit,” The Journal of Financial
and Quantitative Analysis, Vol. 28, No. 2, 1993, pp. 285-
300. doi:10.2307/2331291
[22] G. Emery, “An Optimal Financial Response to Variable
Demand,” Journal of Finance and Quantitative Analysis,
Vol. 22, No. 2, 1987, pp. 209-225. doi:10.2307/2330713
[23] Y. Miwa and J. M. Ramseyer, “Trade Credit, Bank Loans
and Monitoring: Evidence from Japan,” CARF Working
Paper F-054, 2005.
[24] J. Nilsen, “Trade Credit and the Bank Lending Channel,”
Journal of Money Credit and Banking, Vol. 34, No. 1,
2002, pp. 226-253. doi:10.1353/mcb.2002.0032
[25] A. Demirguc-Kunt and V. Maksimovic, “Firms as Finan-
cial Intermediaries: Evidence from Trade Credit Data,”
World Bank Research Working Paper No. 2696, 2001.
[26] J. Van Horne, “Financial Management and Policy,” Pren-
tice Hall, UK, 1995.
[27] D. Roodman, “Xtabond2:Stat Module to Extend Xtabond
Dynamic Panel Data Estimator,” Center for Global De-
Copyright © 2011 SciRes. ME
R. R. VAIDYA
Copyright © 2011 SciRes. ME
716
velopment, Washington, 2005.
[28] M. J. Athey and P. S. Laumas, “Internal Funds and Cor-
porate Investments in India,” Journal of Development
Economics, Vol. 45, No. 2, 1994, pp. 287-303.
doi:10.1016/0304-3878(94)90034-5
[29] A. Ganesh_Kumar, K. Sen and R. R. Vaidya, “Outward
Orientation, Investment and Finance Constraints: A Study
of Indian Firms,” Journal of Development Studies, Vol.
37, No. 4, 2001, pp. 133-149.
doi:10.1080/00220380412331322071
[30] R. Lensink, R. van der Molen and S. Gangopadhay,
“Business Groups, Financing Constraints and Investment:
The Case of India,” Journal of Development Studies, Vol.
40, No. 2, 2003, pp. 93-119.
doi:10.1080/00220380412331293787