Bank Credit, Trade Credit, and Profit: Evidence from Agricultural Firms in Vietnam

This paper investigates the relationships between bank credit and trade credit with profit of 130 agricultural firms listed on Vietnam’s stock exchanges in the period of 2008-2014. Using the GMM approach, the paper reveals inverted-U shaped (∩) relationships between bank credit and trade credit with profit. Specifically, the optimal threshold of bank credit and trade credit to total assets of the firms are 0.4173 and 0.2425, respectively. The findings mean that if the ratio of bank credit to total assets exceeds the benchmark of 0.4173, firms should consider restructuring debts to get them back to the benchmark. To do so, firms should withdraw from those business fields that are not of profession, in addition to liquiditizing unused assets to repay debts and not using short-term credit to invest in long-term projects. Firms may use of trade credit wisely when other sources of finance are lacking. In concrete, firms can increase trade credit use if the ratio of trade credit to total assets is below 0.2425. Yet, if this ratio goes beyond this benchmark, firms should get its back to this benchmark, e.g., keeping a suitable amount of inventory.


Listed Agricultural Firms in Vietnam
Agricultural firms which produce, process, trade agricultural products or supply inputs to agricultural (standard deviation of 31.28 percent). ROE of the firms seems to decline during that period, given the fact that their ROE in 2008 was 15.35 percent and dropped to a mere of 10.19 percent in 2014.
Bank credit and trade credit have remained principal sources of investment and business financing for the firms. In the period of 2008-2014, the average bank credit given to each firm is VND 270 billion per year, but there are firms that were denied access to bank credit. Growth rate of bank credit to the agricultural firms slowed down during this period because of their low profitability and policy revisions of commercial banks, in addition to prudent regulations imposed by the government on the banking system in response to the then economic risk and volatility. Despite that, the amount of bank credit given those firms had increased in the period of 2008-2014.
Beside bank credit, the firms also use trade credit in the form of deferred payment to input purchase (accounts payable) to develop business and make profit. The average of trade credit used by the firms is VND 104 billion per year, which is relatively low as compared to bank credit but increases rapidly (i.e., almost three times during the period of 2008-2014). The increase in trade credit use by the firms is mainly due to the demand on expansion that requires a larger amount of inputs and limited access to bank credit, as just clarified. Inputs of the firms are mainly agricultural products of which supplies and prices largely fluctuate, creating difficulties for them to get sufficient cash and other means of payments ready for that purpose. In that case, trade credit turns out to be an almost perfect solution for firms to have enough inputs for production in order to avoid interruption which has a very adverse consequence on the firms' profit.

Bank Credit and Firm Profit
Bank credit is a pivotal source of funds for firms to develop business and make profit. The presence of several banks with various policies concerning interest rate, loan term and collateral requirement enables firms to select the ones that best suit their profit target. This aspect even becomes more fruitful because banks possess professional skills in screening, monitoring and enforcing repayment that compel borrowers to use the credit given in a right way.
The MM theorems developed by Miller (1958, 1963) stress the role of tax deduction of bank credit that is meaningful for profit of firms. Since interest rate expenses are exempted from tax, profit would be enhanced for firms that borrow from banks since the deducted tax could be used to finance profitable investment projects and business activities. Moreover, firms use bank credit to invest in Research and Development (R&D) to upgrade technology in order to raise profit since updated technology enables firms to improve product quality, increase productivity and mitigate production cost (Hasan & Sheldon, 2016, p. 548). Firms that are denied access to credit must rely on backward technology, so both product quality and productivity are low. All this suppresses profit of firms.
In imperfect credit markets with a high degree of information asymmetry, limited liability and transaction cost, firms prefer internal funds thanks to its lower user cost (the pecking-order theory). Yet, internal funds may be insufficient to meet demand, so firms borrow from banks to finance business activities and ensure efficiency. Highly creditworthy firms that have created intimate relationships with banks are able to obtain credit at the right time with preferable interest rates and loan terms, allowing them to embark on profitable business opportunities. Meanwhile, given rich information and professional knowledge on a wide range of borrowers accumulated over a rather long period, banks can create business networks which link clients (i.e., depositors and borrowers) to help them improve profit via useful advices using that information and knowledge.
Despite its usefulness, if going beyond a certain limit, bank credit will be detrimental to profit of firms, as confirmed by researchers. According to Nkurunziza (2011, p. 465), if the economy falls into recession and lending activity turns out to be too risky, banks become more conscious of protecting themselves by strictly conforming to the Basel Core Principles and increasing risk provision. As a result, interest rates may go up, firms that have borrowed excessively will shoulder heavy unbearable burdens of repaying debts, so profit may drop. In transition economies, governments often pursue prudent policies of reforming financial and banking systems with a special intention of mitigating default risks and safeguarding commercial banks against NPLs and failures by imposing stringent regulations on capital adequacy, risk provision and deposit insurance. Because of that, banks may raise interest rates, suppressing profit of those firms that have heavily relied on bank credit due to increased user costs of capital. Empirically, Cheng et al. (2010Cheng et al. ( , p. 2502) uses a data set of 650 firms in China to prove that high leverage squeezes their profit.
In some circumstances, managerial weaknesses lead banks to unwise decisions, thus depriving them of the ability to closely monitor default risk that gives rise to loss. In order to recover the loss, banks tend to raise interest rates, thus shouldering firms with higher costs and suppressing their profit (Le, 2015, p. 49). Hardships resulting from external factors (e.g., financial reforms, financial crises or exchange rate fluctuations) are other reasons why banks employ tight lending policies and may raise interest rates that inflates borrowing costs and reduces profit of firms (Agostino & Trivieri, 2013, p. 133). Inspired by profit in the past, firms may have borrowed too much to invest in scale and scope expansions regardless of output market uncertainty, especially in the period of economic recession that can happen unexpectedly. If output market declines, it is hard for firms to repay a substantial amount of bank debt because of constraints in fetching funds for that purpose (Le & Huynh, 2014, p. 24). Consequently, profit goes down. This explains why bank credit adversely affects firm profit. The aforementioned arguments imply an inverted-U shaped (∩) relationship between bank credit and firm profit.

Trade Credit and Firm Profit
Trade credit is a specific relationship in which suppliers provide a certain amount of credit to clients (trade credit receivers) by allowing them to defer the payment for the purchased good (Badu et al., 2012, p. 362). Given this attribute, trade credit helps firms tackle difficulties resulting from fund shortage, thereby pushing up their profit. The impact of trade credit on profit is of great concern of researchers since this is a common financing source for firms. Researchers ascertain that firms always have incentives to grant trade credit to customers, so if used wisely, trade credit will boost profit for the latter.
The financing advantage theory of trade credit contends that suppliers have an advantage over credit institutions with regard to evaluating and controlling the risks facing buyers. Indeed, the supplier obtains information needed for those activities at low costs via the normal course of business, visits to the buyer's premises and from other suppliers as business partners (Petersen & Rajan, 1997, p. 663).
The supplier is also better able to influence the buyer's behaviour since it may be in the nature of the good being supplied that there are only few economical alternative sources other than the supplier.
Then, the threat to cut off future supplies if the buyer acts in such a way that harms the possibility to repay is especially creditable as the buyer accounts for an unimportant portion of the supplier's sales. It is also effective for the supplier in salvaging value from the buyer's used assets. If the buyer defaults, the supplier can seize the good supplied to resell through her already established networks for selling goods. As a result, the cost of repossessing and reselling the good is much lower than that of a credit institution.
The theory of price discrimination argues that suppliers uses trade credit to discriminate price to allure customers. Buyers who rely on trade credit for funds are often credit rationed, so they constitute the most price elastic segment of the market (Petersen & Rajan, 1997, p. 664). In that case, trade credit is an effective means of price discrimination if suppliers lower the price of the good to entice higher demand (Brennan et al., 1988(Brennan et al., , p. 1128. Buyers with high profit margins who may be riskier find trade credit underpriced, thus being eager to use more of it. A supplier may apply price discrimination also because of the long-term concern of customers' survivals, especially when she has no potential substitutes for the customers. According to the transaction cost theory, the seller uses trade credit as an effective means to mitigate transaction costs. Trade credit may reduce the transaction costs of paying bills since the buyer can cumulate obligations and pay them once instead of every time when the good is delivered (Ferris, 1981, p. 244). Moreover, there may be a seasonality in the demand pattern for the good the seller supplies. In order to maintain smooth production cycles, the seller may have to build up large inventories, which incurs high costs of warehousing and financing them. By offering trade credit selectively, both across buyers and over time, the seller better manages its inventory position and transaction costs.
At the other end, there are reasons why a producer resorts to trade credit. First, trade credit enables the buyer to check out the quality of the good supplied, which is important where the information asymmetry about the quality of goods is widespread and cheating behaviour prevails. Second, trade credit conveys reliable information about the creditworthiness of the buyer to credit institutions, which helps improve her access to formal credit. Third, the financing theory argues that trade credit is a perfect complement to formal credit for those who need capital for urgent needs but are denied access to formal credit. Finally, trade credit allows the buyer to mitigate transaction cost and better handle risks. To make use of all the trade credit granting policies for profit maximization, trade credit receivers actively choose the most suitable deferred payment terms.
According to Burkart and Ellingsen (2004, p. 570), trade credit also enhances profit of firms since they tend to use the input purchased using trade credit in production instead of diverting it for some reasons.
Firstly, trade creditors have a better position to supervise the behaviour of trade credit receivers, so they can impose outright and timely sanctions if the latter diverts the purchased input, which adversely affects the ability to repay loans. This threat is highly effective, so trade credit receivers tend to use the purchased input in production. Secondly, due to its specificity, the input purchased using trade credit is less possibly used for any purpose other than production or converted into cash.
As a matter of fact, firms not only have to sell goods of high quality but have also to promptly respond to market uncertainty, which requires them to hold a certain amount of inputs in warehouses well equipped with proper facilities to avoid stockout loss as well as instantaneous deterioration. This is costly for agricultural firms since agricultural products are normally so bulky that they occupy a spacious place if stored. In that case, trade credit helps firms mitigate costs by shortening the storage duration and the quantity of inventories by wisely using orders that best suit demand (Tiwari et al., 2016, p. 155 Moreover, if well treated in terms of being granted a large amount of trade credit, firms tend to use them in less profitable activities. In spite of not facing severe information asymmetry like commercial banks, trade creditors may bear a substantial credit risk and confront challenges if granting too much trade credit since clients are normally subject to moral hazard-a problem that can go beyond the control of trade creditors due to limited skills and experience. All the aforementioned arguments would imply an inverted-U shaped (∩) relationship between trade credit and firm profit. Differently speaking, as the volume of trade credit increases from a minor level, it will enhance profit of firms. Yet, if going beyond a certain limit, trade credit will curtail firm profit.

Empirical Model
Based on the literature previously reviewed, this paper adopts the quadratic function that is popularly used by several studies (e.g., Banks et al., 1997;Yabu & Kessy, 2015) to investigate the non-linear relationships between bank credit and trade credit with profit of listed agricultural firms in Vietnam: In Model (1), it profit is profit of firm i in year t , measured by ROE (i.e., returns on equity).
ROE is used to measure profit of the firms due to its advantage over ROS (returns on sales) and ROA (returns on assets). Indeed, ROS is a poor proxy for profit of agricultural firms because the business of agricultural firms is strongly affected by seasonality, making their input prices and sales largely fluctuating. Moreover, ROA of listed firms substantially varies and is much dependent on the industry, meaning that it is only feasible to use ROA as a measure to compare ROA of the same firm over time or with that of a similar firm.
it Bankcredit is the ratio of bank credit to total assets of firm i in year t .

it
Bankcredit is squares of it bankcredit . As explained previously, coefficient 1  is supposed to be positive and 2  to be negative due to the presence of an inverted-U shaped (∩) relationship between bank credit with firm profit. it t Tradecredi is the ratio of accounts payable to total assets of firm i in year t and 2 it t tradecredi is squares of it t tradecredi . As expected, coefficient 3  is positive and 4  is negative because of the inverted U-shaped (∩) relationship between trade credit with firm profit.
However, according to other studies (e.g., de Haas & Peeters, 2006;Rahaman, 2011;Ferrando & Mulier, 2013;Nguyen & Tu, 2015), there are more determinants of profit of firms. Thus, the empirical model should be augmented to take account of those factors: is the ratio of equity to total assets of firm i in year 1  t . Equity that may have an advantage of low user cost is a preferred source of funds for firms, especially those in transition economies with less developed banking systems where the information asymmetry between firms and banks is largely widespread (de Haas & Peeters, 2006, p. 135;Rahaman, 2011, p. 712). Borrowers, including firms, in these economies also incur substantial transaction cost due to cumbersome procedures and immoral (even corrupt) bank staffs. This problem is pervasive for firms that are information opaque. Firms would also prefer equity over debt because it is not obligatory and they can flexibly adjust their dividend policy because in most cases investors are not really concerned with a firm's dividend policy since they can sell a portion of their portfolio of equities if they want cash (i.e., the so-called the dividend irrelevance theory). However, firms may sometimes misuse this valuable financing resource (e.g., investing in business fields that are not of their profession), resulting in loss that reduces profit of firms. Thus, coefficient 5 can be either positive or negative, depending on the economic and business environments in which firms operate and their attributes as well.
it Age is the number of years elapsed since the firm was established. This variable is an almost perfect proxy for experience-a factor that positively affects profit of firms (Rahaman, 2011, p. 709).
Therefore, coefficient 6  is expected to be positive. However, firms operating in saturate environments would become conservative to changes, resulting in a lack of creativeness and low profit.
This would mean that coefficient 6 Labour is the number of permanent labours of firm i in year t . Economies of scale argue that firms of larger size (proxied by the number of permanent labours) have an advantage over smaller ones regarding production costs since long-term average costs decline as output accumulates over time. This advantage brings about higher profit for those firms (Ferrando & Mulier, 2013, p. 3039 it oduction Pr takes a value of 1 for firms that produce, process or trade agricultural products and 0 for those supplying inputs to agricultural production. Coefficient 10  may be positive or negative. it  is the random error of the model.

Methodology
In order to test for the relationship between bank credit and trade credit with firm profit, we use a panel In order to tackle the problem of endogeneity, we use the Generalized Method of Moments (GMM) developed by Arellano and Bond (1991) to estimate Model (2) in order to make the estimation result unbiased and more statistically reliable as well, especially in the case of panel data. This estimation method requires instrumental variables for the two endogenous variables (i.e., it bankcredit and it t tradecredi ). In this case, we use time-lagged (one year) variables as instruments for these two endogenous variables as often done by those studies that utilize the GMM approach (e.g., Rahaman, 2011).

Result and Discussions
The statistical summary of the independent variables of Model (2) is shown in Table 2. The results for the test of the relationship between bank credit and trade credit with profit of the firms are revealed in Table 3. Columns 2, 3 and 4 of Table 3 present the results of RE, FE, and GMM estimation methods, respectively. Despite being unable to tackle the issue of endogeneity, since RE and FE are common estimation methods for a panel data set, we present their results as a reference.
Prior to conducting the regression of Model (2) using the GMM method, we perform a check for multicorrelation between the independent variables described in Table 2. The result shows that all the coefficients between the independent variables are much smaller than 0.8, implying no multicorrelation problem in the empirical model (Yazdanfar & Ohman, 2015). We have also been concerned with the problem of heteroskedasticity, but this problem is automatically corrected by the GMM itself (Cragg, 1983;Wooldridge, 2001). Sargan test shown in column 4 of Table 3 confirms the robustness of the instrumental variables as exogenous variables (i.e., being not correlated with the model's error).
Moreover, Wald test rejects H 0 hypothesis (i.e., all coefficients of the independent variables being equal to zero). Differently speaking, all the coefficients of the independent variables of Model 2 have explanatory powers.  The value of 1  , 2  given in Table 3 and Expression (4)  between trade credit and profit of the firms. Using similar approach to bank credit, we can estimate the optimal ratio of trade credit to total assets of the firms that is 0.2425. This means that if the ratio of trade credit to total assets is below 0.2425, trade credit will boost profit of the firms and the effect reverses if the ratio goes beyond that benchmark.
It is interesting that 1 3    , implying the more important role of trade credit to profit of the firms, other things being equal. This finding is quite understandable, since during the period of 2008-2014 banks had used contracted lending policies (especially to those firms that are financially unsound) while the government curtails the growth rate of credit given to the economy. Differently speaking, this is an advocate of the reality of credit crunch emerging as a consequence of prudent policies maintained by banks and the government of Vietnam.
In addition to bank credit and trade credit, firms also use equity to exploit the advantage of lower user costs. In the period of economic recession, equity positively affects profit of the firms as coefficient 5  of ) 1 (  t i equity has positive value at significance level of 5 percent. This is evidence of the fact that the firms had used equity to maintain and develop business as the access to bank credit had been squeezed. Note. (***), (**) and (*) corrensponds to 1 percent, 5 percent and 10 percent significance levels, respectively.
Source: Estimated out of the data set.
As seen from has a negative value at a significance level of 5 percent, divulging the instability of profit of the firms. Finally, firms that supply inputs to agricultural production seem to have higher profit as compared to those that produce, process or trade agricultural products since coefficient 10  of it production has a negative value at a significance level of 10 percent.

Conclusion
This paper uses a data set of 130 Vietnamese listed agricultural firms in the period of 2008-2014 to figure out the relationship between bank credit and trade credit with their profit. Using the GMM approach, the paper reveals inverted-U shaped (∩) relationships between bank credit and trade credit with profit of the firms. According to the estimates, the optimal threshold of bank credit to total assets of firms is 0.4173 and that for trade credit is 0.2425. Moreover, trade credit seems to be more important 2  2  than bank credit in terms of boosting profit of the firms. However, since the data used in this paper covers the period of 2008-2014, the results of this paper should be taken with a caution.
The paper also shows that equity plays an positive role to profit of the firms as bank credit growth rate tends to decline because of the economic downturn and the financial crisis in the period of 2008-2014, in addition to managerial weaknesses of Vietnam's commercial banks. Age is inversely related to profit of the firms, which is consistent with the practice that a number of agricultural firms have operated quite long but lacked creativeness. Finally, firms that supply inputs to agricultural production enjoy higher profit than their counterparts since their output markets are more stable since farmers who produce an overwhelming portion of agricultural products always need a lot of inputs.
The findings of this paper would mean that if the ratio of bank credit to total assets exceeds the benchmark of 0.4173, firms should consider restructuring debts to get them back to the benchmark. To do so, firms should withdraw from those business fields that are not of profession, in addition to liquiditizing unused assets to repay debts and not using short-term credit to invest in long-term projects.
Moreover, firms should use of trade credit wisely when other sources of finance are lacking. In concrete, firms can increase trade credit use if the ratio of trade credit to total assets is below 0.2425.
Yet, if this ratio goes beyond this benchmark, firms should get its back to this benchmark, e.g., keeping a suitable amount of inventory. Firms need to use efficiently equity to take advantage of lower user cost of capital as compared to bank credit and trade credit that are adversely affected by asymmetric information and transaction cost.
However, it is not without limitation because of the limitation of the data used. Further research may be necessary to cover longer period after 2014 because Viet Nam has signed (multi-) bilateral trade agreement, e.g., EVFTA, CPTPP so that the increased trade flows among the trading partners with commitment of tariff reduction could influence the concerning results of this paper.