University of Groningen Trade credit in the rice market of ... · Dao Thi Ba ), my aunties (Y Uong,...

University of Groningen

Trade credit in the rice market of the Mekong Delta in VietnamNguyen, Lam Thu Uyen

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite fromit. Please check the document version below.

Document VersionPublisher's PDF, also known as Version of record

Publication date:2011

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):Nguyen, L. T. U. (2011). Trade credit in the rice market of the Mekong Delta in Vietnam. University ofGroningen, SOM research school.

CopyrightOther than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of theauthor(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons thenumber of authors shown on this cover page is limited to 10 maximum.

Download date: 06-11-2020

https://www.rug.nl/research/portal/en/publications/trade-credit-in-the-rice-market-of-the-mekong-delta-in-vietnam(b700839b-2506-4671-b3fa-67efc59f03c1).html

https://www.rug.nl/research/portal/en/publications/trade-credit-in-the-rice-market-of-the-mekong-delta-in-vietnam(b700839b-2506-4671-b3fa-67efc59f03c1).html

Nguyen Lam Thu Uyen

Trade Credit in the Rice Market of

the Mekong Delta in Vietnam

Theses in Economics and Business

ii

Publisher: University of Groningen, Groningen, The Netherlands

Printed by: Ipskamp Drukkers B.V., The Netherlands

ISBN: 978-90-367-4863-6

978-90-367-4864-3

©2011 Nguyen Lam Thu Uyen

All rights reserved. No part of this publication may be reproduced, stored in

a retrieval system of any nature, or transmitted in any form or by any means,

electronic, mechanical, now know or hereafter invented, including

photocopying or recording without prior written permission of the

publisher.

Trade Credit in the Rice Market

of the Mekong Delta in Vietnam

Proefschrift

ter verkrijging van het doctoraat in de

Economie en Bedrijfskunde

aan de Rijksuniversiteit Groningen

op gezag van de

Rector Magnificus, dr. E. Sterken,

in het openbaar te verdedigen op

donderdag 21 april 2011

om 13.15 uur

door

Nguyen Lam Thu Uyen

geboren op 17 september 1978

te Soc Trang, Vietnam

iv

Promotores: Prof. dr. B.W. Lensink

Prof. dr. C.L.M. Hermes

Copromotor: Dr. C.H.M. Lutz

Beoordelingscommissie : Prof. dr. Christopher Woodruff

Prof. dr. Arjun. Bedi

Prof. dr. J. Van De Meer- Kooistra

Acknowledgements

This dissertation would not have been possible without the help and support I have

received from colleagues, friends and family. In this section, I express my sincere

gratitude to them.

I would like to express my deepest gratitude to my supervisors, Prof.

Robert Lensink, Prof. Niels Hermes and Dr. Clemens Lutz. As an academic

supervisor of NPT project - a corporation between Faculty of Economics and

Business, University of Groningen and School of Economics and Business

Administration (SEBA), Cantho University- Robert offered me the chance to write

this PhD dissertation. He was also very helpful with his useful feedback and

support while I was working on this project. My deepest gratitude also goes to

Professor Niels Hermes. Niels helped me to find joy in doing research into trade

credit issues since I worked on my master thesis, and I could not have wished for a

better coach to guide me through the process of writing a dissertation. His

accessibility, enthusiasm and open-mindedness made it a real pleasure to work

with him. In addition, I am very grateful to Dr. Clemens Lutz. With a lot of effort,

he provided me invaluable assistance and support during the five years it took me

to write this dissertation.

Furthermore, I greatly appreciate the efforts and useful comments of the

members of the reading committee consisting of Prof. Christopher Woodruff, Prof.

Arjun Bedi and Prof. Van De Meer- Kooistra.

There are so many people who contributed by providing me both

emotional and practical support. I greatly appreciated the assistance from several

members of the International Relations Office, especially Anita Veltmaat, Gonny

Lakerveld, Wiebe Zijlstra, Erik Haarbrink. My thanks also go to all staff members

of SOM, the Center for Development Studies for the arrangements during my

vi

study and especially to Mr. Arthur De Boer for the instructions in the layout of this

book. I would like to express my deep gratitude to the SEBA staffs that arranged

necessary conditions for me to complete this dissertation i.e. the official procedures

and data collection (Dr. Mai Van Nam, Mr. Trung Tin, and Ms. Truc. Lien, Ms.

Linh Giang). Also, my great thanks goes to the staffs of statistic bureaus in

TienGiang, Angiang, VinhLong, CanTho, HauGiang and SocTrang provinces for

providing the opportunity to conduct the large number of enterprise interviews in

2007.

Next, I would like to thank my colleagues and friends providing me with

both emotional and practical support. My thanks go to Thong, Tra and Scott for

spending time discussing on trade credit issue. In addition, my special thanks go to

Tu for being my paranymphs and Xuan for her great help while I was in hospital in

2007. I thank Aljar, Binh, Tu, Dut, Khuong, Tran, Tri, Kadek, Ilko, chi Phuong and

other friends who have supported me but it is not possible to mention each one

individually.

I want to warmly acknowledge my parents in law (Nguyen Van Lanh and

Dao Thi Ba), my aunties (Y Uong, Y La, K. Yin), my uncles (Tia Xiem and Cu

Phen), my cousins (My Linh, My Phuong, My Tran, Kathy, Stijn and Kristiof) for

their emotional support. During the 5 years, they made my life in Groningen easy

and enjoyable.

To end with, I would like to express my deepest thankfulness to my

parents (Nguyen Tan Nhan and Lam Thi Lien), my husband (Nguyen Minh Phung)

and my sister (Nguyen Lam Phuong Thao and Huynh Nguyen Huy Hoang) for their

constant love, support and encouragement. By always standing behind me,

providing both practical and mental support, they have considerably contributed to

the completion of this thesis.

Uyen

Groningen, March 2011

Table of Contents

Chapter 1 Introduction 1

1.1 Introduction 1

1.2 Aim of the Study 4

1.3 Structure of the Thesis and Summary of the Main Findings 6

1.4 Limitations and Suggestions for Further Research 10

Chapter 2 Trade Credit Supply and Firm Sales Growth 13

2.1 Introduction 13

2.2 Review of Literature 14

2.2.1 Literature on Firm Growth 14

2.2.2 Trade Credit in Marketing Theory 22

2.3 The Data, Empirical Strategy 25

2.3.1 The Sample 25

2.3.2 Empirical Strategy 27

2.3.3 Empirical Measures 28

2.4 Empirical Results 30

2.4.1 Trade Credit Provision 30

2.5 Conclusions 34

Chapter 3 A Survey in the Rice Market of the Mekong Delta37

ii

3.1 Introduction 37

3.2 A Brief Description of the Rice Market of the Mekong Delta 38

3.2.1 Overview of the Paddy/Rice Production 38

3.2.2 Characteristics of Rice Traders in the Mekong Delta 40

3.3 Description of Data Collection 43

3.3.1 The Pilot Survey 43

3.3.2 The Official Survey 44

3.4 Descriptive Statistics of the Sample 47

3.4.1 Overview of the Sample 47

3.4.2 Trade Credit Provision 49

3.4.3 Firm Competitiveness 52

3.4.4 Financial Constraints 55

3.5 Summary 57

Chapter 4 Trade Credit Supply and Firm Competitiveness 59

4.1 Introduction 59

4.2 Trade Credit Supply and Firm Competitiveness: A Survey 61

4.3 The Vietnamese Rice Sector 65

The Survey 66

4.4 Methodology and Data 68

4.4.1 The Regression Model 68

Type of Firm Dummy Variables 72

4.4.2 Descriptive Statistics 73

4.5 Empirical Results and Discussion 77

4.5.1 Firm Competitiveness and Trade Credit Supply 78

4.5.2 Customer Bargaining Power and The Role of Perceived

Competition 80

iii

4.6 Summary and Conclusions 82

Chapter 5 Trade Credit Supply and Customer Characteristics

89

5.1 Introduction 89

5.2 Related Literature 90

5.3 Hypothesis and Methodology 93

5.3.1 Hypothesis Development 93

5.3.2 Empirical Methodology 94

5.3.3 Descriptive Statistics 99

5.4 Estimation Results and Discussion 104

5.4.1 The Impacts of Customers’ Characteristics 105

5.4.2 The Effects of Market Characteristics 106

5.4.3 The Effects of Supplier Characteristics 107

5.5 Conclusion 112

Chapter 6 Trade Credit Supply in Different Market Segments

113

6.1 Introduction 113

6.2 Empirical Methodology 114

6.3 Why May Trade Credit Differ Across Segments? 118

6.3.1 Market Context Characteristics and Trade Credit Supply 118

6.3.2 Structural Differences Across Market Segments 119

Market Structure Across Market Segments 121

Access to External Finance 123

Customer Characteristics 124

6.4 Estimation Results 126

6.4.1 Differences Between Market Segments: General Overview 126

iv

6.4.2 A Close Look at Each Segment 127

Wholesale-Miller Segment 127

Wholesaler Segment 128

Miller Segment 130

Retailer Segment 130

6.5 Summary of the Results 131

v

List of Tables

Table 2.1: Summary of studies on determinants of firm growth 18

Table 2.2: Sample firms by industry 26

Table 2.3: Descriptive statistics 29

Table 2.4: Results of the fixed-effect models 32

Table 2.5: Results from the one step system GMM estimator 33

Table 2.6: Results from the random effect and the OLS

regressions 34

Table 3.1: Distribution of rice firms according to main business

activities and firm size (measured by total assets) 48

Table 3.2: Size and age of the surveyed rice firms 48

Table 3.3: Descriptive statistics of profitable variables 49

Table 3.4: Distribution of the surveyed rice firms providing credit 50

Table 3.5: Experience with defaulting by clients 51

Table 3.6 Trade credit received 51

Table 3.7: Firm competiveness measures 53

Table 3.8: Mean and median of competitive measures across

different market segments 55

Table 3.9: The importance of financial constraints 56

Table 3.10: Firms using bank loans in 2006 57

Table 4.1: Descriptive statistics 76

Table 4.2: Correlation matrix 77

Table 4.3: Tobit regressions of the relationship between trade

credit supply and competitiveness 84

Table 4.4: Tobit regressions of the relationship between trade

credit supply and competitiveness 86

Table 5.1: Variable definition 101

Table 5.2: Summary statistic of the sample 103

Table 5.3: Correlation matrix 104

Table 5.4: The impact of customer characteristics on trade

vi

credit provision 110

Table 6.1: Mean Differences across different Market

Segments 120

Table 6.2: Types of customers 124

Table 6.3: Determinants of trade credit supply in the four

market segments: w-millers, wholesalers, millers and

retailers 134

List of Figures

Figure 3.1: Marketing channel of the rice market in The

Mekong Delta 37

Chapter 1

Introduction

1.1 Introduction

Since Vietnam conducted economic reforms in 1986, it has transformed itself from

a centrally planned economy to a market-oriented economy. One of the most

important features of the economic reform was the legalization of the private

sector. Consequently, the private sector has grown dramatically during the last 20

years. In particular, the number of private firms has increased rapidly, and most of

them are small and medium-sized enterprises (SMEs). For example, Nguyen et al

(2008) showed that 99 per cent of private firms in Vietnam in 2004 were SMEs.

These SMEs face many challenges, however. The major barrier is the limited

access to capital for future growth (Rand, 2007; Nguyen and Ramachandran, 2006;

Sakai and Takada, 2000).

The financial sector in Vietnam is characterized by an immature financial

market and a banking system in which state-owned commercial banks still play a

dominant role. The official Vietnamese stock market was launched in July 2001

with only two individual stocks on the first trading day. The list of companies

increased to 32 enterprises at the end of 2005 and 620 enterprises in 2010. Yet,

according to HCMC Institute for Development Studies, in 2010, Vietnam has

500,000 enterprises of which SMEs account for 97 per cent (Saigon Time 10,

2010). SMEs in Vietnam have not had access to the stock market, and therefore

they had to rely on the banking system for their capital needs. The four main state-

owned commercial banks accounted for 73 per cent of total bank assets and

provided 72 per cent of total loans in the economy in 2001. However, half of the

Trade Credit in The Rice Market of The Mekong Delta in Vietnam

2

state-owned commercial bank loans (45 per cent) are channeled to state-owned

enterprises (SOEs) (Asian Development Bank, 2002). Consequently, SMEs, and

especially private SMEs, find it very difficult to access bank loans though bank

financing is among the most important sources of external financing for SMEs.

While facing financial constraints is an important problem of Vietnamese

firms, Vietnamese suppliers provide much trade credit to their trading partners.

Trade credit occurs when a supplier delivers the goods or services to a customer at

t=0 and allows the customer to delay payments for a certain period. Thus, the

customer pays for these goods or services at t=1. We observe that in general trade

credit provision is popular in Vietnam (McMillan and Woodruff, 1999), but also in

the case of the rice market of the Mekong Delta (Luu, 2003). Given the fact that

Vietnamese SMEs as well as rice firms have faced financial constraints, we find

Vietnam to be an interesting case to study in terms of the underlying motivations

for firms to supply trade credit.

Existing studies on trade credit supply concentrate on two different

views. According to the finance theory, suppliers know their trading partners better

than banks and other financial institutions. Therefore, the information asymmetry

between lenders and borrowers is less of an issue for suppliers than for banks and

financial institutions. As a result, firms facing bank-borrowing constraints tend to

turn to suppliers for credit. Accordingly, suppliers with good access to external

finance may be willing to supply credit.

The marketing theory of trade credit stresses the importance of trade

credit provision as a marketing device to support sales. This category of studies

particularly focuses on the impact of market characteristics, i.e. market competition

on trade credit supply. It argues that when firms have difficulties raising sales due

to fierce competition, they will supply trade credit. Several explanations for this

have been suggested. In fact, trade credit enables suppliers to allow their clients to

verify product quality before making any payments. Consequently, trade credit can

Chapter1: Introduction

3

work as a signal or guarantee of product quality. Also, by granting credit, suppliers

provide financial help to clients to finance their purchases. Indeed, trade credit is a

short-term finance source that can be used to attract clients, especially financially

constrained customers.

The literature that examines access to external finance as the main

determinant of trade credit supply is extensive. Yet, the number of empirical

studies investigating the role of trade credit supply as a marketing device is

relatively limited. In addition, recent studies show that the finance theory may not

provide a full explanation of trade credit supply. In fact, it is shown that small,

young and financially constrained firms seem to provide more credit, especially in

dealing with large clients (Van Horen, 2005, 2007). This poses an interesting

research question: what are the factors that drive firms to supply trade credit? To

contribute to an answer to the question, this dissertation concentrates on studying

the factors that determine trade credit supply for Vietnamese firms.

This dissertation will focus on a particular industry – the rice industry in the

Mekong Delta of Vietnam – for several reasons. First, rice is one of the main

commodities in Vietnam, especially in the Mekong Delta. In fact, during the last

decade Vietnam has been the second largest rice exporter in the world. In

particular, the Mekong Delta has been Vietnam’s rice bowl, producing about 50

per cent of the country’s total rice output and providing more than 90 per cent of

the volume of Vietnam’s rice export. Consequently, the rice market of the Mekong

Delta has been a large market that consists of different market segments. The four

main market segments of this particular market are wholesale-millers (w-millers),

millers, wholesalers and retailers. Wholesale-millers process paddy into rice and

sell rice as their final products, while millers only provide milling services and will

charge a milling fee. Wholesalers distribute milled rice from wholesale-millers to

retailers, who sell rice to the final consumer. According to Luu (2003), market

competition varies substantially across these four market segments. Thus, the rice


4

market is an interesting case for studying the impact of market competition on

trade credit provision, which is one of the main targets of this thesis.

Second, concentrating on one industry in one country allows us to measure

firm competitiveness and market competition more carefully. Third, the rice

market contains a few large rice exporters and state-owned enterprises, and a large

number of private firms that are mostly SMEs. As mentioned above, a shortage of

capital is one of the greatest difficulties Vietnamese SMEs meet. Rice firms,

however, provide a lot of credit to their trading partners (Luu, 2003). The rice

market is therefore interesting and suitable to investigate the factors that cause

firms to supply credit.

1.2 Aim of the Study

The ultimate objective of this thesis is to contribute to answering the following

research question: what are the factors that persuade firms to provide trade credit to

their clients? In order to answer this research question, the thesis will be examining

the following issues.

Trade Credit Supply and Sales Growth

According to the marketing theory, suppliers provide credit with the aim of

promoting sales. However, whether trade credit provision indeed enhances sales

growth is still unknown, since no empirical study has investigated the impact of

trade credit on sales growth.

Trade Credit Supply and Firm Competitiveness

We will be taking the marketing view to investigate the relation between market

competition and trade credit supply. It argues that firms facing fierce competition


5

find it difficult to increase sales. Therefore, they will provide more trade credit to

support sales. The impact of market competition on trade credit provision,

however, is still an unresolved issue in the literature. Empirical studies on this

issue are rare and produce conflicting results. On the one hand, Fisman and Raturi,

(2004) find a positive effect of market competition on trade credit provision. On

the other hand, McMillan and Woodruff (1999) show a negative influence of

market competition on trade credit provision. These studies use similar approaches

and apply a cross-country and/or cross-industry analysis with a single or a dummy

variable measuring competition and/or market structures. Yet, it is difficult for a

dummy variable to capture the differences in the market structures and/or the

degree of market competition across different countries and/or industries. In

addition, these studies may not control for differences in product characteristics

and/or industry-specific characteristic that influence the use of trade credit

(Summer and Wilson, 2003; Giannetti et al, 2009). For example, Giannetti et al

(2009) find that service firms are more willing to offer credit than manufacturers.

With the aim of examining the relation between market competition and trade

credit supply, we focus on one industry in one country to eliminate various

products/industry characteristics that may affect trade credit provision. Therefore,

we are able to focus on the market structure in one industry as the major factor

influencing trade credit supply. At the same time, we plan to use several variables

to carefully measure different factors driving a firm’s competitiveness. This

approach is expected to provide us a better insight into the impact of market

competition on trade credit supply.

Trade Credit and Customer Characteristics

Trade credit may work as a marketing device to attract and keep customers to

generate sales (Long et al, 1993; Fisman and Raturi, 2004). Yet, suppliers do not


6

provide credit to all customers. For example, some studies suggest that firms are

motivated to give more credit to large clients with strong bargaining power (Van

Horen, 2007; Fabbri and Klapper, 2009). Other studies find that suppliers are more

willing to offer credit to credit-worthy customers (McMillan and Woodruff, 1999).

This poses an interesting question: what determines the amount of trade credit

supplied to customers? One of the objectives of this thesis is to study how

individual customer characteristics affect the amount of trade credit provided at the

transaction level. Using our own survey, we obtain details on specific trade

relations between suppliers and customers. Consequently, we are able to

investigate which characteristics provide customers with better access to trade

credit.

Trade Credit Supply and Market Segments

The discussion of the data in chapter 3 shows that trade credit supply varies

significantly across the main market segments; the substantial differences in trade

credit supply across market segments are confirmed in chapters 4 and 5. Therefore,

one of the targets of this project is to understand these differences across market

segments such as w-millers (processing firms), wholesalers, retailers, and millers

(milling-service firms).

1.3 Structure of the Thesis and Summary of the Main Findings1

Chapter 2 aims at examining whether the view of trade credit as a marketing device

is relevant to the case of Vietnamese firms. The chapter concentrates on

1 The thesis is a collection of papers. There is some overlap. Those who have read the previous

chapters can skip the section 4.3, section 5.2 and section 6.2 since these matters have also been dealt

with in chapter 3, section 4.2 and section 5.3 respectively.


7

investigating whether trade credit supply enhances sales growth. It uses a panel of

Vietnamese listed firms in the country’s two stock markets: Ho Chi Minh Stock

Exchange and Hanoi Stock Exchange, for the period 2004-2007. Although several

theoretical studies suggest that trade credit functions as a marketing device to

support sales, our study is one of the first to empirically examine the impact of

trade credit supply on firm sales growth. The significant effect of trade credit

supply on sales growth does indeed reveal the necessity of investigating the role of

trade credit as a marketing device, which is one of the main issues of the empirical

studies presented in chapters 4, 5, and 6.

Chapter 3 discusses the rice market in the Mekong Delta. It describes the

survey used to collect the data for the empirical studies presented in chapters 4, 5,

and 6. Next, the chapter discusses the data and provides an in-depth overview of

the rice market of the Mekong Delta. This is important in terms of providing an

overall picture of the rice industry and insights into the causes and nature of the use

of trade credit by rice traders in the Mekong Delta.

The focus of chapter 4 is the relation between firm competitiveness and

trade credit supply. Existing empirical evidence on this issue is based on cross-

industry and/or cross-country analyses. We apply a different approach in this

thesis. We concentrate on one industry, namely the rice industry in the Mekong

Delta, to study the effect of firm competitiveness on trade credit supply. By

following this approach, we expect to remove the effect of the differences in

product characteristics, the various industries (Giannetti et al, 2009; Summer and

Wilson, 2003), and the differences in the development in financial markets banking

system (Fisman and Love, 2003), and legal system (Lu, 2009) on trade credit

supply in different countries. Indeed, these factors may influence trade credit

supply differently, and thus they influence the relation between trade credit supply

and firm competitiveness. Moreover, this approach allows us to apply Porter’s five

forces model, which suggests that firm competitiveness is influenced by five


8

factors in five distinct dimensions (Porter 1979, 1980, 1985, and 2008). We plan to

explicitly investigate the different dimensions that drive firm competitiveness and

their impact on the supply of trade credit. We concentrate on two different

dimensions of competitiveness: the competitive pressure from other suppliers

(rivals), and bargaining power vis-à-vis the customer.

We find strong evidence that competition causes suppliers to grant credit

to their trading partners to support sales and avoid customer switching. In addition,

concerning a firm’s bargaining power vis-à-vis clients, the results show that

suppliers also offer more credit in dealing with clients that have strong bargaining

power. The reason is that the threat of switching to another supplier is higher for a

client with strong bargaining power, and this provides suppliers with incentives to

grant credit to keep the customer. In fact, this chapter adds convincing evidence for

the relation between trade credit supply and market competition. It shows that

suppliers may be highly motivated to use credit to enhance their sales.

Chapter 5 uses transaction level data on suppliers and customers involved

in specific trade relations, and proceeds to identify the customer characteristics that

improve access to suppliers’ trade credit sources. Empirical analyses on this topic

are rare, and the available studies leave out important variables like customer

bargaining power and market competition. In chapter 5, we cover a broader range

of variables that reflect customer characteristics. We include several proxies

measuring customer creditworthiness, characteristics of trading relations and

market contexts. In line with our expectations, the study reports a larger amount of

credit offered to creditworthy clients. In addition, suppliers are very much inclined

to give more credit to large clients and/or clients with strong bargaining power.

The empirical evidence in this chapter suggests that suppliers use trade credit to

build trade relations with creditworthy clients and high potential customers, i.e.

large buyers. In doing so, suppliers expect to reduce the risk of client default and

increase their present and future sales.


9

We observe significant results for the market segment dummy variables in

all scenarios in chapters 4 and 5. This evidence indicates that there are structural

differences in trade credit supply across the four main market segments. Therefore,

chapter 6 aims at understanding the differences in trade credit behavior across

different market segments. We investigate the differences in determinants of trade

credit supply across market segments. The empirical findings show that the market

segment context significantly influences trade credit supply. In the segments of the

wholesale-miller and wholesaler, the marketing role of trade credit is among the

main motivations driving suppliers to grant credit in these segments. Second, in the

market segment of retailers and millers, the set of variables representing the

marketing theory does not show significant results. Thus, the marketing role may

not be among the major factors driving trade credit supply in the miller and retailer

segments. Third, the information advantage theory of trade credit, which mentions

that suppliers grant credit since suppliers know their trading partners much better

than banks and financial institution, also appears to be a relevant determinant. Rice

firms appear to be willing to offer more credit to customers with whom they have

close relations because of better information. Yet, information advantage theory

appears to be much more relevant in the segment of millers, in which suppliers

select to grant credit to closely related customers. This is different from the other

segments, where traders are willing to grant credit to large customers/customers

with strong bargaining power to generate sales. Intuitively, we provide a number of

explanations for these results. First, the small value of transactions may decrease

the attractiveness of the trade credit option to clients. At the same time, the cost of

collecting information concerning clients may be too high compared to the value of

transactions. This discourages suppliers to employ credit as a marketing instrument

to support sales in these segments.

In general, this thesis has brought a better understanding of the role of

trade credit supply, especially concerning the relation between firm


10

competitiveness and trade credit supply on the one hand. On the other hand, this

project also provides fruitful information on the impact of customer characteristics

and market contexts on trade credit supply.

A remarkable result of our study is that even though we concentrate on one

industry, major differences in trade credit supply still exist across the four main

market segments. This study shows that trade credit supply in different markets

depends on different combinations of market context characteristics like the

customer bargaining power, the product, the main business activity, the banking

system, etc. One of our main findings is that the role of trade credit supply may

differ even among the different segments of one industry. Therefore, this study

implies that not only country and industry characteristics but also market segment

specificity should be taken into account when studying the role of trade credit.

1.4 Limitations and Suggestions for Further Research

This study is subject to several limitations. For example, the survey data gave us

numerous useful insights that may help improve our understanding of the role of

trade credit. However, we realize that data collected through surveys may be

biased. For example, a significant number of the participants in the interviews are

retailers, who may have a low average level of education. This may lead to a lower

reliability of their answers, since they may not understand all survey questions. Or

perhaps respondents have incentives to give incorrect answers. To deal with these

difficulties, in 2006 we conducted a pilot survey to test our questionnaires and

make them right for rice traders. In addition, we employed the staff of statistic

bureaus in each province, as they are well-trained interviewers. They were capable

and willing to further illustrate and clarify questions. Therefore, we think that we

have done everything we could to reduce biases to a minimum.

This thesis focuses on one industry in one country to gain an insight into the

impact of market structure on trade credit supply. It is interesting that the


11

determinants of trade credit supply differ in different market contexts; we find that

the main determinants of trade credit differ throughout the four market segments.

Firms in the processing and wholesale sector appear to have stronger incentives to

use credit to generate sales than firms in the retail and service sector. This fact

offers directions for further research, investigating the role of trade credit as a

marketing tool to enhance sales growth in different sectors.

We also find strong evidence that competition and buyer bargaining power

are associated with trade credit supply. However, very little theory has been

developed on this issue. Another direction for further research may therefore be to

develop a theory to explain the impact of market structure, including competition

and buyer bargaining power, on trade credit supply.

Chapter 2

Trade Credit Supply and Firm Sales

Growth

2.1 Introduction

Trade credit has been used extensively both in developed and developing countries.

Several theoretical and empirical studies argue that suppliers employ trade credit as

a marketing device to generate sales. Yet, no study has investigated whether trade

credit can actually enhance firm sales. Therefore, this study aims at investigating if

trade credit can help to enhance sales growth. We employ a firm-level data set of

the listed firms at the Ho Chi Minh Stock Exchange and the Hanoi Stock Exchange

to examine the impact of trade credit provision on firm sales growth. In addition,

we investigate whether the impact of trade credit provision on firm sales varies

across different types of firms: i.e. young firms and old firms.

Vietnam is an interesting case to examine the impact of trade credit

provision on firm sales for several reasons. Vietnam is a transition economy with

an inefficient banking system and infant financial markets. Vietnamese enterprises

face several administrative difficulties in formal and informal credit markets

(Rand, 2004). According to Rand, 14 to 25 per cent of Vietnamese enterprises are

credit-constrained. These firms would increase their debt holdings 40 to 115 per

cent if borrowing constraints were relaxed. Yet, although facing financial

constraints, Vietnamese suppliers provide a lot of trade credit to their trading

partners (McMillan and Woodruff, 1999; Luu, 2003). In our sample, 97 per cent of


14

the firms provide credit to their customers. We find it interesting to study the

underlying motivations to supply credit in these cases.

This chapter analyzes the impact of trade credit provision on firm sales

growth. The chapter is organized as follows. Section 2.2 provides an overview of

the relevant literature. Section 2.3 continues with the data and empirical strategy,

while section 2.4 presents the discussion of the empirical results. Finally, section

2.5 concludes the chapter.

2.2 Review of Literature

2.2.1 Literature on Firm Growth

Much attention has been devoted to examining the determinants of firm

growth in the literature. Previous studies have focused primarily on initial firm-

specific conditions such as firm size and age, innovation, financial structure,

ownership structure, and human capital.

Concerning the relationship between firm size and firm growth, Gibrat’s

theory (1931) suggests that the expected growth rate of a given firm does not

depend on its size. The principle behind Gibrat’s “law of proportional effects” is

that on average a firm’s expected growth in the next period is proportional to its

current size, implying that the expected growth rate is independent from its size.

Yet, several subsequent empirical studies demonstrate a significant relation

between firm size and growth. Although some studies report positive impacts

(Prais, 1974; Singh and Whittington, 1975; Wijewardena and Tibbits, 1999), the

negative relationship between firm size and growth has been reported by most

recent studies. According to Beck et al (2006), small firms in countries with more

developed financial institutions grow faster in comparison with large firms. The

reason is that information asymmetry is an important problem for small firms with

potentially profitable growth opportunities that want access to finance. In countries

Chapter2: Trade Credit Supply and Firm Sales Growth

15

with developed credit markets, this problem is less serious and financial constraints

are far more relaxed, so that small firms can achieve their growth potential. For

example, Dunne and Hughes (1994) find a negative impact of size on growth for

the case of listed UK manufacturing firms; Bottazzi and Secchi (2003) use data of

the Italian manufacturing firms and show a negative influence of size on firm

growth. Goddard et al (2002) use data from quoted Japanese manufacturing firms

and show a similar result: a negative effect. Yashuda (2005) also points at the

negative effect of size on firm growth for Japanese manufacturing firms.

McPherson (1996) finds that small Southern African firms grow faster. Calvo

(2006) shows the same results for Spanish manufacturing firms. Brown et al (2005)

find that small firms have higher growth rates in Romania.

The effect of a firm’s age on growth has also been discussed widely.

According to Jovanovic’s model (1982), firms learn about their efficiency through

production activities. Consequently, if output is a decreasing convex function of

managerial inefficiency, Jovanovic concludes that young firms grow faster than

older firms. The negative and significant relation between firm age and growth is

also found to be prominent in empirical studies: (Evans (1987a, b) uses data from

US manufacturing firms and shows that young firms grow faster; Variyam and

Kraybill (1992) point at a negative influence of age on firm growth in the case of

US manufacturing and services firms; Geroski and Gugrle (2004) show a negative

correlation between firm age and growth for large European companies; Brown et

al (2005) employ data from small businesses in Romania and find a negative effect

of age on firm growth; Wijewardena and Tibbits (1999) show similar results

regarding the negative impact of age on firm growth for small Australian

companies; Yasuda (2005) finds that young firms experience higher growth rates

than older firms in the case of Japanese manufacturers.

Several theoretical papers stress the important role of innovation for firms

that want to expand sales and market shares. For example, Hey and Kamshad


16

(1994) stress that investment in product innovation is the main strategy for

expansion – the reason being that investment in innovation (R&D) may lead to

technical changes, causing (partly) economic growth (Griliches and Lichtenberg,

1984; Solow, 1957). However, empirical studies find it hard to prove the impact of

innovation on sales growth. For example, Freel (2000) uses firm-level data from

228 US manufacturing firms and does not find that innovators are experiencing

higher growth rates. Similarly, Bottazzi et al (2001) do not find a significant effect

of innovation and technology on sales growth. Only a few studies manage to

demonstrate a positive relationship between innovation and sales growth. Geroski

and Toker (1996) employ firm-level data from UK firms and show that innovation

has a significant positive impact on sale growth. Another notable study that

stresses the positive effects of innovation on sales growth is Roper (1997).

Next, turning to the financial structure, several studies focus on the

relationship between firm expansion and financial performance. These studies

stress that credit constraint problems negatively affect firm growth. Indeed,

financial constraints may hinder investments in potentially profitable projects and

influence firm growth negatively. For example, Brown, Earle and Lup (2005) show

that credit constraint relaxation (measured by access to bank loans) has a positive

effect on firm sales growth. Bottazzi et al (2008) also find significant effects of

financial performance on firm growth even though the magnitude of the coefficient

is quite small. Musso and Schiavo (2008) prove that access to external finance

funds has a positive effect on firm growth in term of sales and employment

numbers.

Ownership structure is also argued to affect growth rates. According to

Ehrlich et al. (1994), owners invest capital in a firm to maximize the value of the

firm. However, managers may pursue private objectives. The degree to which

manager’s private objectives deviate from the firm profit maximization objective

may be related to firm ownership structure. In particular, this study shows that state


17

ownership can effect firm growth negatively in the long run. When firms are state-

owned, managers may pursue more easily their own goals rather than the

objectives of the state, since the government may not be able to monitor managers

properly anyway. Backer and Sleuwaegen (2003) indicate that foreign ownership is

an additional source of firm heterogeneity affecting firm productivity. Beck et al.

(2005) find that foreign-owned firms experience faster growth while state-owned

companies seem to experience lower growth rates.

Human capital also appears to be a factor that contributes to firm growth.

Education and training can influence knowledge, skills, motivation, and the ability

to provide suggestions to short- and long-term business plans that affect firm

productivity growth. In fact, Almus (2002) shows the positive effects of human

capital on growth in the case of German companies. McPherson (1996) also

emphasizes the positive and significant influences of human capital level on the

growth of micro and small businesses in five Southern African countries.

Final determinants for firm growth as proposed by some studies refer to the

manager: the manager’s age, sex, and competencies. For example, Ibrahim and

Goodwin (1986) show that a manager’s managerial skills are key factors that

contribute to the growth of small firms. Similar findings were described by Huck

and McEwen (1991): they identify entrepreneurs’ competencies in management,

planning and budgeting, and marketing as most crucial to the successful operation

of a small business.

Although several studies investigate the determinants of firm sales growth,

not one investigates the impact of trade credit provision on firm sales growth rates.

This is remarkable since several theoretical studies argue that trade credit actually

works as a marketing device to promote sales for suppliers.

18

Tab

le 2

.1:

Sum

mar

y o

f st

udie

s on d

eter

min

ants

of

firm

gro

wth

No.

Au

thor

V

ari

ab

les

C

ou

ntr

y/d

ata

O

utc

om

es

1.

Alm

us

(2002)

Gro

wth

rat

es i

n

emplo

ym

ent

Ger

man

fir

ms

Posi

tive

effe

ct o

f hum

an c

apit

al o

n

gro

wth

2

Bott

azzi

et

al (

2001)

Gro

wth

in s

ales

T

he

worl

d t

op 1

50

phar

mac

euti

cal

firm

s

No s

ignif

ican

t ef

fect

of

innovat

ion o

n

firm

gro

wth

3.

Bott

azzi

and S

ecch

i

(2003)

Gro

wth

in n

et a

sset

s T

he

Ital

ian m

anufa

cturi

ng

firm

s

Neg

ativ

e im

pac

t of

firm

siz

e on g

row

th

4

Bott

azzi

et

al (

2006)

Gro

wth

in t

ota

l sa

les

A p

anel

of

Ital

ian

man

ufa

cturi

ng a

nd s

ervic

e

firm

s.

Posi

tive

effe

ct o

f fi

nan

cial

per

form

ance

on f

irm

gro

wth

5

Bro

wn e

t al

(2005)

Gro

wth

in e

mplo

ym

ent

and s

ales

Sm

all

busi

nes

s in

Rom

ania

N

egat

ive

impac

t of

firm

siz

e on g

row

th

Neg

ativ

e im

pac

t of

firm

age

on g

row

th

Posi

tive

effe

ct o

f cr

edit

const

rain

ts

rela

xin

g o

n f

irm

gro

wth

.

1

9

No.

Au

thor

V

ari

ab

les

C

ou

ntr

y/d

ata

O

utc

om

es

6

Cal

vo (

2006)

Gro

wth

in e

mplo

ym

ent

Span

ish m

anufa

cturi

ng f

irm

s

in t

he

per

iod 1

990-2

000

Neg

ativ

e im

pac

t of

firm

siz

e on g

row

th

7

De

Bac

ker

and

Sle

uw

aegen

(2003)

Gro

wth

in l

abor

pro

duct

ivit

y t

hat

is

def

ined

as

the

rati

o o

f

val

ue

added

to t

he

num

ber

of

emplo

yee

s

22.4

52 B

elgia

n

man

ufa

cturi

ng f

irm

s fo

r th

e

per

iod 1

990-1

995

Posi

tive

effe

ct o

f fo

reig

n o

wner

ship

on

gro

wth

.

8

Dunne

and H

ughes

(1994)

Annual

gro

wth

in n

et

asse

ts

The

UK

fir

ms

in t

he

per

iod

1975-1

985

Neg

ativ

e im

pac

t of

firm

siz

e on g

row

th

9

Ehrl

ich e

t al

(1994)

Gro

wth

in t

ota

l to

n-

kil

om

eter

per

form

ed

by t

he

airl

ines

23 i

nte

rnat

ional

air

lines

.

Neg

ativ

e im

pac

t of

stat

e-o

wned

shar

es

on f

irm

gro

wth

.

10

Evan

(1987 a

, b)

Gro

wth

in e

mplo

ym

ent

U

S m

anufa

cturi

ng f

irm

s N

egat

ive

impac

t of

firm

age

on g

row

th

20

No.

Au

thor

V

ari

ab

les

C

ou

ntr

y/d

ata

O

utc

om

es

11

Fre

el (

2000)

Gro

wth

in s

ales

and

emplo

ym

ent

228 U

S m

anufa

cturi

ng f

irm

s N

o s

ignif

ican

t ef

fect

of

innovat

ion o

n

firm

gro

wth

12

Ger

osk

i an

d T

oker

(1996)

Gro

wth

in t

urn

over

The

UK

fir

ms

Posi

tive

effe

ct o

f in

novat

ion o

n f

irm

gro

wth

13

Ger

osk

i an

d G

ugle

r

(2004)

The

log o

f th

e num

ber

of

emplo

yee

s

Euro

pea

n c

om

pan

ies

firm

wit

h m

ore

100 e

mplo

yee

s

Neg

ativ

e im

pac

t of

firm

age

on g

row

th

for

the

case

14

Goddar

d e

t al

(2002)

Annual

gro

wth

rat

es i

n

tota

l as

sets

Japan

ese

man

ufa

cturi

ng f

irm

s N

egat

ive

impac

t of

firm

siz

e on g

row

th

15

McP

her

son (

1996)

Gro

wth

in t

he

num

ber

of

work

ers

Mic

ro a

nd s

mal

l en

terp

rise

s

in s

outh

ern A

fric

a

Neg

ativ

e im

pac

t of

firm

siz

e on g

row

th

Posi

tive

effe

ct o

f hum

an c

apit

al o

n

gro

wth

16

Sin

gh a

nd

Whit

tingto

n (

1975)

Annual

gro

wth

in n

et

asse

ts

Quote

d c

om

pan

ies

in t

he

UK

P

osi

tive

impac

t of

firm

siz

e on g

row

th

2

1

No.

Au

thor

V

ari

ab

les

C

ou

ntr

y/d

ata

O

utc

om

es

17

Var

iyam

and

Kra

ybil

l (1

992)

Gro

wth

in e

mplo

ym

ent

US

man

ufa

cturi

ng a

nd

serv

ice

firm

s

Neg

ativ

e im

pac

t of

firm

age

on g

row

th

18

Wij

ewar

den

a an

d

Tib

bit

s (1

999)

Gro

wth

in t

ota

l as

sets

S

mal

l A

ust

rali

an c

om

pan

ies

Neg

ativ

e im

pac

t of

firm

age

on g

row

th

19

Yas

huda

(2005)

Gro

wth

rat

es i

n t

ota

l

asse

ts

Japan

ese

man

ufa

cturi

ng f

irm

s N

egat

ive

impac

t of

firm

siz

e on g

row

th

Neg

ativ

e im

pac

t of

firm

age

on g

row

th


22

2.2.2 Trade Credit in Marketing Theory

Trade credit is used extensively in trade transactions among firms all over the

world. Trade credit occurs when a supplier delivers the goods or services to a

customer at t=0 and allows the customer to delay payments for the goods for a

certain period. Thus, the customer pays for these goods or services at t=1. A

common idea that runs through several studies on trade credit supply is that the

main purpose of granting credit is to increase sales (Nadiri, 1969; Blazenko and

Vandezande, 2003; Long et al., 1993).

Theoretical Studies

Nadiri (1969) was the first to develop a neoclassical model in which firms consider

the cost of providing trade credit as a type of selling costs to influence product

demand. In his theoretical framework, firms are assumed to select output prices,

the costs of providing trade credit (opportunity costs of investing in trade credit)

and the volume of output in order to maximize net profit. The outcome of this

model is that trade credit provision is positively related to the level of sales as well

as profit margins. This study shows that firms use trade credit as a type of selling

instrument to increase sales and profitability. Therefore, firms that spend more on

trade credit provision will gain a higher level of sales and profitability.

Sharing this view, Blazenko and Vandezande (2003) extend Nadiri’s

theory by including the cost of bad debts caused by trade credit provision. In

particular, they assume that the higher the prices of products, the larger the amount

of trade credit needed, the more likely it is that the trader is not paid by clients –

and therefore the higher the costs of bad debts caused by losses from client default.

In this theoretical framework, firms seek to maximize profit margins while the

marginal cost increases. This model arrives at a relation between trade credit


23

provision and profit margins that can be either positive or negative, depending on

the price elasticity of product demand. For example, when the marginal cost

increases, firms should increase selling prices; thus, firms with very high price

elasticity of product demand will experience a large drop in sales. Consequently,

firms with a high price elasticity of products may attempt to support revenues by

using the two following measures at the same time: (1) controlling the degree of

increasing selling prices strictly, and (2) providing additional trade credit. In this

case, while trade credit use increases, the profit margin may decrease since firms

manage to keep the increase of selling price to a minimum. This outcome adds to

the prediction of Nadiri (1969) that the relation between trade credit provision and

profit margins can also be negative, depending on the price elasticity of product

demand. In short, according to Blazenko and Vandezande (2003), under the

condition of facing an increase in marginal costs, trade credit can be used to

encourage sales and profitability. Yet, if a firm’s price elasticity of product demand

is too high, trade credit supply may not help to increase the firm’s profitability

though it may support sales.

In this stream of research, several theoretical studies attempt to go one step

further to explain how trade credit functions to support sales. For example, Long et

al (1993) show that trade credit may be like a guarantee of product quality to

customers. In particular, they explain how trade credit can function as a marketing

device to promote sales. By providing trade credit to buyers, sellers allow their

clients to verify the quality of a product before making any payment. When the

quality is poor, buyers can return products without making payments.

Consequently, it is only profitable for sellers that produce high-quality products to

grant credit to customers, while producers of low quality will choose not to provide

credit. Long et al (1993) suggest that as a result trade credit can work as a product

guarantee as well as being a sign of product quality to increase product demand. In


24

particular, this product guarantee can promote sales of firms that produce high-

quality products and have not yet gained a well-established reputation of product

quality.

Hyndman and Serio (2006) have developed an alternative theoretical model

that explains how trade credit works as a marketing device to support sales. In this

theoretical framework, the authors consider the number of competitors as the main

factor that drives market competition. At the same time, they assume that the seller

can decide to sell on cash or on credit under the constraints of incurring monitoring

costs if they sell on credit. The outcome of this study implies that when there is a

single supplier, the seller does not need to attract customers and thus they only sell

cash. However, increasing the number of suppliers first induces suppliers to grant

credit to develop a trading relation with customers in order to generate sales. Yet,

when the number of suppliers increases too much, customers find it easy to switch

to another supplier, and suppliers will decrease the amount of trade credit provided.

The reason is that it is difficult for suppliers to enforce the repayments from

customers. Indeed, this study implies that trade credit can be used to encourage

sales. However, firms in a monopolistic market do not need to use trade credit to

support sales. At the same time, it is very costly to provide credit in highly

competitive markets, and consequently firms may be less willing to provide credit

to support sales when facing fierce competition.

Furthermore, several studies argue that suppliers provide credit to encourage

customers to order large quantities. These studies concentrate on discussing the

optimal economic quantity order under conditions of delayed payment. For

example, Rachamadugu (1989), Chung (1998), and Huang (2007) have developed

theoretical models to determine the optimal economic order quantity when delayed

payment is permitted. In these theoretical models, firms attempt to determine the

optimal economic order quantity in order to minimize the total annual costs, which

include (1) the annual costs of placing orders, (2) the annual costs of stock holding,


25

and (3) the annual costs of opportunity capital. In other words, the objective of the

firm is to maximize the net present value of the shareholder wealth by minimizing

the net present value of the cash outflows resulting from the total annual costs. The

solution of these models shows that the optimal order quantity is an increasing

function of the permitted delay in payment.

To conclude, several theoretical studies argue that trade credit works as a

marketing device to stimulate sales growth. However, we do not know of any

empirical study that investigates whether trade credit provision enhances firm sales

growth. In this chapter, we examine the impact of trade credit supply on firm sales

growth rates to gain an insight into whether trade credit provision can help to

support firm sales growth.

2.3 The Data, Empirical Strategy

2.3.1 The Sample

The dataset for this study is a panel for the period 2004-2007. We derive the basic

sample from the official websites of the Ho Chi Minh Stock Exchange and the

Hanoi Stock Exchange, as well as annual reports and websites of listed companies.

In total, 320 firms are listed at the two stock trading centers in September 2008. In

our sample we do not include firms in the financial sector since they have different

accounting rules. In addition, we select only listed firms that released annual

reports at least in the last three years. This means that we exclude several firms that

were listed in 2008 since these firms released annual reports only in the last two

years. The variable of interest is a measure of the sales growth rate, and we need to

observe at least two observations of sale growth rate per firm. Due to these

restrictions, our sample was reduced from 320 firms to 260 firms that were listed in

the period 2000-2007.


26

The sample represents large manufacturing Vietnamese firms: in total 260

firms in 20 industries. The number of firms per industry is shown in table 2.1. In

our sample, 252 of the 260 firms (97 per cent of the sample) provide credit to

customers. The average proportion of account receivables in total assets is about 18

per cent (see table 2.2)

Table 2.2: Sample firms by industry

Ord.

no

Industry Number of

Firms

1 Basic metals 1

2 Chemical and all allied products 6

3 Clothing 4

4 Commercial machinery and computer equipment 12

5 Construction 80

6 Electronic and other electrical equipment products 12

7 Food and drink 39

8 Furniture and fixtures 7

9 Gasoline, diesel and electricity wholesale 9

10 Medical products 5

11 Mining and quarrying of nonmetallic minerals 2

12 Oil mining services 1

13 Printing and publishing 16

14 Rubber and miscellaneous products 18

15 Services (hotel and entertainment services) 7

16 Stone, clay, glass and concrete products 18

17 Textile mill products 1

18 Tobacco products 1

19 Transportation equipment products 6

20 Transportation service 15

Total 260


27

2.3.2 Empirical Strategy

The main research objective of this study is to investigate whether the size of trade

credit supply enhances firm sales growth. If trade credit functions as a marketing

device that influences product demand, then firms that provide more credit will

achieve higher sale growth rates.

In this section, we explore the impact of trade credit supply on firm sales

growth rates. At the same time, we also control for the influence of other relevant

factors such as firm age, firm size, human capital, ownership structure and leverage

on firm growth rates. The regression is based on a panel of Vietnamese listed firms

during the period 2004-2007.

The estimated equation is the following:

SAGROWit = αXit + βTDPROit + εit (2.1)

In equation (2.1) the dependent variable SAGROWit is defined as the annual growth

rate in firm sales at time t in the four-year period 2004-2007. The Xit includes a

constant term and characteristics of the individual firm such as firm size, firm age,

ownership structure, financial structure, and human capital. The core variable of

interest is a measure of trade credit provision. We employ two popular measures of

trade credit supply in the existing literature: (1) TDPRO1it – measured by the ratio

of account receivables to firm total assets (Deloof and Van Overfelt, 2009); (2)

TDPRO2 – measured by the ratio of account receivables to firm sales (Giannetti et

al, 2009).

We use a fixed-effects analysis to estimate the impact of trade credit

provision. The fixed-effect estimator offers several advantages. First of all, it

disentangles the impact of trade credit provision from the observed attributes that

affect the outcome variables. Second, it controls for relevant unobserved firm and

industry characteristics that affect sales growth and do not change over time.


28

Yet, one of potential criticisms of our study is the potential endogeneity of

trade credit supply variable since causality may run in both directions – from trade

credit provision to sales growth and vice versa. For example, it is possible that

firms having low sales growth in the past may offer more credit to archive higher

sales growth rates or firms with high sales growth may provide more trade credit.

Another concern to consider are unobserved firm characteristics that can influence

both trade credit supply and firm growth. For example, firm competitiveness may

be correlated with trade credit and sales growth. However, based on the data we

have, we are not able to have a good instrumental variable measuring firm

competiveness to solve this problem. In response, we employ the method of GMM

system (Arellano and Bover,1995; Blundell and Bond, 1998), which is expected to

overcome the problem of reverse causation and omitted variable bias mentioned

above.

2.3.3 Empirical Measures

Control Variables

The relations of firm size and firm age to firm growth have been studied

extensively in the literature. Consequently, we control for the effects of firm size

and firm age in our study. FSIZE is measured as the log of firm sales, while FAGE

is the age of a listed firm, defined as the number of years since a firm has been

established. To control for the impact of the financial structure on firm growth, we

introduce leverage, LEV, measured as the ratio of total debt to the value of equity

as our proxy for capital structure. This variable reflects the degree to which a firm

utilizes borrowed money. Furthermore, in order to capture the effects of human

capital on firm growth rates, we employ the ratio of the number of employees with

a university degree (bachelors/masters/doctors) to the total number of employees,

DEGREE, as our measure of human capital (Brown et al,. 2005). Another


29

important factor that can contribute to firm growth rates is ownership structure. We

include the variable STSHARE, defined as the percentage of a firm’s shares held by

the state. We also use FRSHARE, measured as the percentages of a firm’s shares

held by foreign investors. Besides, we include the profitable variable, PRORATIO,

measured as the ratio of total profit from the main business activities before tax

(excludes profit from financial activities) to total sales. Finally, trade credit

obtained, TDOBi, is measured as the ratio of account payables to total assets,

measuring the proportion of total assets that firms financed with trade credit from

their suppliers. Due to the data restriction, we cannot include a measure of

innovation. Yet, in the previous discussion, it was suggested that characteristics of

firm managers do influence the growth of small business. However, our data only

includes large firms. Also, characteristics of a board of directors in the sample have

not changed much during 2004-2007. Therefore, the fixed-effect regression is

removed from the variables that measure the characteristics of a board of directors,

such as sex, and education; these variables have not been included in our model.

Table 2.3: Descriptive statistics

Variables Mean Max Min Std Obs

SAGROW 0.366 5.77 -0.944 0.86 662

TDPRO1 0.17 1.00 0 0.16 779

TDPRO2 0.185 0.71 0 0.14 776

PRORATIO 0.089 0.698 -0.201 0.0740 778

LOGSALES 12.104 16.461 7.703 1.316 784

STSHARES 0.300 0.85 0 0.232 871

FRSHARES 0.065 0.882 0 0.155 868

AGE 12.71 53 1 10.10 883

DEGREE 0.237 0.942 0.012 0.179 871

LEV 2.37 39.82 0.022 3.52 772

TDOB 0.156 0.822 0.0001 0.142 773


30

2.4 Empirical Results

2.4.1 Trade Credit Provision

Results of the fixed-effects and the one step system GMM models for determinants

of sales growth are presented in table 2.4 and table 2.5. We start by estimating the

general model, which includes all our possible variables.

Our discussion first concerns the impact of trade credit provision on firm

sales growth, since this is the main focus of our study. Indeed, trade credit

provision has a significant and positive impact on sales growth (table 2.4, columns

1 and 2; table 2.5). The estimated coefficient indicates that firms that offer more

trade credit do realize higher sales growth rates. Consequently, the empirical

evidence of a positive association between trade credit provision and firm sale

growth supports the view of trade credit as a marketing instrument.

In the fixed effect regression we also use the interaction term of the trade

credit provision variable, TDPRO, and firm age, FAGE (table 2.4, columns 3 and

4). This allows us to test whether the impact of trade credit provision on sales

growth is different for young firms that have not yet established their reputation of

product quality as compared to old firms with a reputation. The negative

coefficients of the interaction terms between FAGE and TDPRO confirm that the

impact of trade credit provision is stronger for young firms. The reason may be that

young firms do not have an established reputation, and therefore the need of testing

quality before making any payment is higher in their case. When young firms

provide trade credit, clients can return the inferior products without incurring any

costs. This is like a guarantee, which is very important to clients of new suppliers.

Thus, trade credit supply affects sales growth strongly.

In general, the fixed-effect and the system GMM estimators show that

trade credit can help to enhance sales growth rates. In addition, the fixed effect


31

model indicates that trade credit effect appears to be more important to young

firms than to older firms. This finding is consistent with Long et al (1993), which

suggest that trade credit functions as a marketing device to support sales by

offering a guarantee and/or to signal product quality.

As a robustness test, the random effect and OLS regressions also provide

similar results for the positive and significant impact of trade credit supply on sale

growth. However, we do not find a significant impact of the interaction term

between trade credit supply TDPRO and firm age AGE on firm sales growth.

Control Variables

Our next concern is the effect of control variables on firm sales growth. First, we

do not find support for the Gibrat law, since large firms in our sample appear to

experience higher sales growth rates than small firms. In addition, we find

supportive evidence that young firms also realize higher sales growth rates than old

firms. The data also suggests that highly profitable firms experience higher sale

growth rates than firms that are less profitable. However, our findings do not

confirm that firms that utilize a higher ratio of debts to equity reach a higher sales

growth rate.

Moreover, our measure of human capital shows no statistically significant

correlation with firm sale growth. However, we find supportive evidence that

ownership structure affects firm sale growth rates. In fact, the estimated results in

this study review the negative and significant effects of the proportion of shares

hold by the states and the proportion of shares hold by foreign investors on firm

sales growth.

In short, our findings suggest that in our sample firm size, firm age, trade

credit provision, and firm profitability are the main determinants of firm sales

growth.


32

Table 2.4: Results of the fixed-effect models

Variables (1)

(2) (3)

(4)

TDPRO1 1.049

(2.01)**

3.38

(3.030)***

TDPRO2

1.269

(2.84)***

3.440

(2.65)***

TDPRO1*FAGE

-1.276

(-2.79)***

TDPRO2*FAGE

-1.034

(-1.99)**

FAGE -0.809

(-3.71)***

-0.699

(-3.21)***

-0.595

(-2.58)***

-0.502

(-2.14)**

FSIZE 0.768

(3.68)***

0.753

(3.66)***

0.753

(3.67)***

0.744

(3.63)***

PRORATIO 2.078

(2.49)**

1.874

(2.31)**

2.066

(2.44)**

1.903

(2.32)**

LEV 0.018

(1.08)

0.019

(1.13)

0.013

(0.72)

0.018

(1.03)

DEGREE 0.932

(0.90)

0.980

(1.00)

0.981

(0.98)

0.943

(0.96)

FORSHARE -0.032

(-0.14)

0.119

(0.53)

-0.074

(-0.34)

0.074

(0.33)

STSHARE 0.586

(0.73)

0.452

(0.57)

0.365

(0.50)

0.395

(0.53)

TDOB 0.12

(0.27)

-0.044

(-0.09)

0.373

(0.86)

0.066

(0.15)

Conts -8.01

(-3.61)***

-8.067

(-3.67)***

-8.173

(-3.78)***

-8.354

(-3.84)***

R SQUARE 0.2641 0.2702 0.2844 0.2811

Observations 660 660 660 660

The fixed-effect regressions of sales growth on trade credit provision, trade credit obtained,

firm age, firm size, employee degrees, and leverage and ownership structure. Standard

errors are robust and have been adjusted for cluster effects. The T-statistics are given in the

parentheses above. *, ** and *** denote a significant level at 10, 5 and 1 per cent,

respectively.


33

Table 2.5: Results from the one step system GMM estimator

Variables (1) (2)

L. salesgrowth 0.008

(0.18)

0.009

(0.12)

TDPRO1 1.537

(1.98)**

TDPRO2

3.256

(2.51)** FAGE -0.110

(-1.32)

-0.207

(-1.71)*

FSIZE 0.386

(3.04)***

0.631

(3.18)***

PRORATIO 1.993

(2.47)**

1.618

(1.69)*

LEV -0.014

(-0.60)

-0.019

(-0.68)

DEGREE -0.221

(-0.31)

-1.127

(-1.01)

FORSHARE -0.683

(-2.69)***

-0.857

(-2.40)**

STSHARE -0.337

(-1.46)

-0.145

(-0.41)

TDOB 0.404

(0.37)

-0.147

(-0.10)

Conts -4.401

(-3.33)***

-6.929

(-3.32)***

Obs. 406 406

Number of firms 257 257

Sargan test

P value 11.89 0.752

19.54

0.242

Wald test 27.70

0.002

19.88

0.030

The one step system GMM regression of sales growth on trade credit provision, trade credit

obtained, firm age, firm size, employee degrees, and leverage and ownership structure.

Standard errors are robust and have been adjusted for cluster effects. The Z-statistics are

given in the parentheses above. *, ** and *** denote a significant level at 10, 5 and 1 per

cent, respectively. 2

2 This result may be suffer from first order serial correlation


34

Table 2.6: Results from the random effect and the OLS regressions

Variables (1)

(2) (3)

(4)

TDPRO1 0.828

(2.81)***

0.674

(2.62)***

TDPRO2

0.622

(2041)**

0.455

(2.15)** FAGE -0.078

(1.98)**

-0.071

(-1.76)*

-0.064

(-1.75)*

-0.058

(-1.57)

FSIZE 0.091

(3.71)***

0.070

(2.76)***

0.061

(2.95)***

0.043

(2.03)**

PRORATIO 1.654

(3.07)***

1.429

(2.66)***

1.352

(3.21)***

1.156

(2.74)***

LEV 0.019

(1.10)

0.024

(1.38)

0.021

(1.09)

0.025

(1.44)

DEGREE 0.327

(2.32)**

0.430

(2.86)***

0.340

(2.59)***

0.422

(3.09)***

FORSHARE -0.129

(-0.96)

-0.038

(-0.33)

-0.108

(-0.80)

-0.028

(-0.24)

STSHARE -0.413

(-2.76)***

-0.430

(-2.82)***

-0.417

(-2.97)***

-0.429

(-3.04)***

TDOB -0.1027

(-0.34)

-0.130

(-0.45)

-0.058

(-0.21)

-0.063

(-0.25)

Conts -0.859

(-2.90)***

-0.601

(-1.92)**

-0.494

(-2.03)**

-0.261

(-1.05)

R SQUARE 0.137 0.126 0.108 0.093


The random effect regression (1) and (2), and the OLS regression (3) and (4) of sales

growth on trade credit provision, trade credit obtained, firm age, firm size, employee

degrees, and leverage and ownership structure. Standard errors are robust and have been

adjusted for cluster effects. The T-statistics are given in the parentheses above. *, ** and

*** denote significant level at 10, 5 and 1 per cent, respectively

2.5 Conclusions

In this chapter, we have used a panel of listed Vietnamese firms to identify the role

of trade credit supply in determining firm sales growth. Our major finding is a


35

positive and significant relationship between firm sales growth and trade credit

provision. In particular, we observe that the impact of trade credit provision on

firm sales growth is stronger for young firms than it is for old firms. This finding

supports the view that trade credit works as a marketing device to promote sales.

Chapter 3

A Survey in the Rice Market of the

Mekong Delta

3.1 Introduction

As mentioned in chapter 1, this thesis aims to investigate the role of trade credit

provision in the rice market of the Mekong Delta. The literature on trade credit

shows that data sets at firm-level and transaction-level are needed to study this

issue. In order to collect data, we conducted a survey in 2007: 626 rice firms

operating at the six main market segments of the rice market in the Mekong Delta

were interviewed. The survey was conducted in the six provinces of An Giang,

Tien Giang, Vinh Long, Can Tho, Hau Giang, and Soc Trang. This survey helped

us to construct a data set that includes several variables that are relevant to our

research.

This chapter describes the procedures of the survey, providing an overview

of the rice market in the Mekong Delta and some statistics on the sample. Section

3.2 will provide an overview of the rice market in the Mekong Delta; it will

describe the process of rice production and characteristics of the traders. Section

3.3 describes the data collection procedures. Next, section 3.4 presents some

descriptive statistics of the data collected, and finally section 3.5 provides a

summary of the main findings.


38

3.2 A Brief Description of the Rice Market of the Mekong Delta

3.2.1 Overview of the Paddy/Rice Production

The Mekong Delta is the largest delta in the south of Vietnam. With an area of

39,554 km2

(12 per cent of the entire country), the Mekong Delta has produced

about 20 million tons of paddies or over 5 million ton of rice in 2008. This region

has been considered as the rice bowl of Vietnam, as it accounts for 50 per cent of

the total of rice produced in the country and provides about 90 per cent of the rice

for export from Vietnam.

Details in the process of producing paddy/rice are described in figure 3.1.

In general, there are several phases in producing paddy/rice in the Mekong Delta.

First, farmers produce and dry the paddy immediately after harvesting. Next,

farmers sell the paddy to assemblers/gatherers or they sell directly to processing

firms; gatherers/assemblers are intermediaries who only purchase paddy from

farmers and transport to sell to processing firms without processing the paddy. In

the next phase the paddy is processed into rice. Processing firms check the standard

requirements on moisture. After checking, the dried paddy is stored in warehouses

while the paddy that is too moist is dried by dryers and/or by the sun. After this,

the paddy will be processed further: it will be milled into brown or white rice. In

the case of export rice, the brown rice is polished before packaging. The finished

product will be stored in the warehouse until it is sold on the market.

The flows of paddy/rice in the Mekong Delta are presented in the figure

3.1. Paddy/rice is transported to several marketing agents in the same province and

in other provinces in the Mekong Delta. Figure 3.1 shows that the key marketing

agents in the regional rice market are gatherers, w-millers, millers, wholesalers,

retailers, and SOEs.

Chapter 3: A Survey in The Rice Market of The Mekong Delta

39


40

3.2.2 Characteristics of Rice Traders in the Mekong Delta

Farmers

Farmers in the Mekong Delta generally cultivate relatively small irrigated farms

with three main cropping seasons: winter-spring (October-February), spring-

summer (February-May), and summer-autumn (June-September). Rice farmers in

the Mekong Delta have relatively small farms that are 1 to 3 hectares on average.

Farmers produce paddy and dry it immediately after harvesting. After storing some

paddy for home consumption and keeping seeds for the next production cycle,

most farmers sell the surplus to gatherers or processing firms i.e. SOEs and w-

millers. In the Mekong Delta, about 75 per cent of the total paddy production is

sold after harvesting. The proportion of marketable surplus is highest for farmers in

this region (Minot and Goletti, 1995).

Gatherers/Assemblers

Gatherers usually buy paddy from farmers and transport it to processing firms such

as w-millers and/or SOEs for immediate sale. They are not involved in any

processing activity. Gatherers often operate in a relatively small area of about 10

kilometers. Since they work in one area for a long time, they have a good

knowledge of the quality of the paddy from the farmers in the surrounding villages.

According to the 1995 IFPRI survey of rice farmers, gatherers account for more

than 95 per cent of the paddy purchased from farmers and transported and sold to

processing firms.

Processing Firms

Processing firms in the Mekong Delta include private millers and the milling

factories of state-owned enterprises. The difference in processing capacity among


41

traders at the milling sector is big. In Vietnam, there are about 620 large millers

with a capacity of 15 to 200 tons per day and 30,000 small and medium millers

(Huynh et al, 2009). Small millers have a producing capacity of 0.2 to 1 ton per

day, while the capacity of medium-scaled millers is 1 to 15 tons per day shift.

In this study, we refer to millers as processing firms that provide milling

services and charge a certain price per kg output. They are usually small millers

that own a small factory with a small processing capacity (less than 1 ton of rice

per day). They mainly provide milling services to farmers nearby for home

consumption. These small millers exist in almost every village in the Mekong

Delta. There are also a few large-scale millers/polishers in the Mekong Delta that

provide milling and polishing services to other rice exporters and/or w-millers.

W-millers (medium and large-scale private millers) have a capacity of at

least are larger than 1 ton of rice per day. They usually purchase paddy from

farmers/gatherers and process this paddy into rice. They also store and sell milled

rice as a final product to wholesalers and/or other w-millers. At the same time, w-

millers also sell a large amount of milled rice to SOEs and private rice exporters.

Millers/w-millers account for the milling and polishing of about 80 per cent of the

total rice in the region (www.stp.vn).

There are 11 SOEs (state-owned enterprises) in the Mekong Delta of

Vietnam. This are provincial food companies in each province. Each state-owned

enterprise usually has a few modern factories with modern and high capacity

machines to process high quality rice for exports. For example, the state-owned

food company in the province of Tien Giang has four factories with an average

production capacity of about 100 ton per day. Since SOEs have high capacity

equipment to process and store rice, they are the main rice exporters of the region.

They usually purchase milled rice from w-millers and polish rice to meet the

requirements of foreign traders. Rice exporters also purchase paddy from farmers

to produce rice; however, this accounts for a relatively small part of their exporting


42

volumes. They mill and polish about 20 per cent of the total of rice produced in the

region.

Although the government tries to remove external and internal restrictions

so that all rice firms are equal in exporting activities, the number of private firms

that export their products is still very small. Moreover, the total number of rice

exporters in the Mekong Delta remains small.

Wholesalers

Wholesalers play an important role in distributing rice for consumption in the

region. In fact, wholesalers act as intermediaries who distribute rice from w-millers

to final consumers in the Mekong Delta, by using rice retailers/small wholesalers

in the region. According to Minot and Goletti (1995), rice wholesalers trade about

2 million tons of rice to be consumed in the region.

Retailers

Retailers usually have shops at marketplaces. Retailers mainly purchase rice from

wholesalers located nearby, they store the rice and sell it to final consumers at local

markets.

Gatherers and exporters (SOEs) are excluded from the empirical analyses

in chapters 4, 5 and 6.3 We focus on the four market segments – w-millers, millers,

wholesalers, and retailers – since these traders only produce and trade rice at the

regional market. In these four segments, traders play a role that is different from

that of the processing firms, service firms, and wholesale and retail firms. Due to

3 Gatherers mainly transport and trade paddy, which is different from rice, the main product of other

market segments. Besides, SOEs concentrate on processing rice for exporting purposes and are

mainly involved in exporting activities. SOEs play a dominant role in long distance trading, e.g. on

the interregional and international market. The market of SOEs is very large since rice exporters may

not only compete with rice exporters inside the country but also with rice exporters from other

countries such as Thailand and India. As a consequence, it can be difficult to measure the market

structure in this market segment.


43

the differences in the main business activities, we notice that the degree of market

competition varies across these market segments (Luu, 2003).4 At the same time,

trade credit is popular in this particular market. Therefore, we find that the rice

market is an interesting case to investigate the impact of firm competitiveness on

trade credit supply.

3.3 Description of Data Collection

Our data collection procedures consist of two stages: a pilot survey was conducted

in 2006, and the survey was finalized in 2007.

3.3.1 The Pilot Survey

A pilot survey was carried out in June and July 2006 with the aim of field-testing

the questionnaires and survey procedures. The pilot survey took place in the four

provinces: Tien Giang, Vinh Long, Can Tho and Soc Trang in the Mekong Delta.

Participants of the pilot included a small group of agents involved in the rice

market of the Mekong Delta. They are two SOEs in Tien Giang and Vinh Long,

five millers, five wholesales, and five retailers and five gatherers in Can Tho and

Soc Trang. At this stage of the research, the quality of information and a good

insight into different businesses as well as business relationships in marketing was

the main issue.

The participants were selected through the network that had previously

been established. The interviewees were former students of the School of

Economics and Business Administration of Can Tho University and were at that

time working at the interviewed rice firms. The discussions with these interviewees

were open and helpful. The interviewees provided useful feedback on our measures

about firm competitiveness and trade credit use, and they allowed us to test the

4 See section 3.4.3.


44

validity of our questionnaires. After completing the pilot, much was revised on the

basis of the feedback we got from the interviewees, e.g. we rephrased several

questions to make them more accessible to rice traders. The interviewees also

suggested that we use some measures of competition that would be easier to apply

for the rice traders.

3.3.2 The Official Survey

The final survey was conducted in the summer of 2007. The key criterion to select

the provinces for our survey was that the provinces had to be the main rice

producers in the Mekong Delta: the number of rice traders in these provinces is

high, just like the volume of trade between rice traders.

The six selected provinces are: An Giang, Tien Giang, Vinh Long, Can

Tho, Hau Giang, and Soc Trang. An Giang was selected because recently this

province has been the largest rice producer in the Mekong Delta; for example, in

2006, when the Mekong Delta produced 18,229,200 tons paddy, An Giang was the

largest producer with a total amount of 2,923,200 tons.

The other five provinces in the survey are all large rice producers. One of

the advantages that these provinces have over others, is that they are located close

to one another and along national highway 1A; transport between these provinces

is convenient and can be via roads and waterways. Inter-province trade transactions

are large, and market integration between these provinces is expected to be high.

Traders in these provinces can easily exchange information on the rice market such

as prices, policy and other rice firms.

The Enterprise Survey

In Vietnam, nationwide enterprise surveys are conducted every five years to collect

information on every corporation. This enterprise survey collects data for the

government to evaluate the performance of current economic policies and to make


45

plans for the future economic development of the country. The latest enterprise

survey, which was announced by a Prime Minister’s Decision on August 15, 2006,

took place in July 2007.

Early July we learned about the enterprise survey. We found that it thought

it would be a great advantage if the staff of statistic bureaus in each province were

to conduct our survey together with the enterprise survey. Firstly, the interview

team appeared to be highly qualified; the staff of statistics bureaus has a lot of

experience with enterprise interviews. Secondly, the government had authorized

the statistic bureaus to do the interviews. That means that firms will be more

cooperative and will sooner make an appointment; this is important since it can be

very difficult to make appointments to interview a large number of firms. Thirdly,

the staff of the statistics bureaus has been working with these firms for quite some

time. Every year, firms in each province have to provide all kinds of reports such

as balance sheets, cash flow statements and profitability statements, since the

statistics bureau in each province has to prepare data for the province yearbook. As

a result, firms will provide relatively consistent answers since big differences in the

figures/numbers when compared to the previous year will lead to many questions.

In addition, firms also try to provide useful answers to other questions. Fourthly,

the two questionnaires of our survey and the enterprises survey contained several

similar questions. For example, both questionnaires include questions on

background information: main business activities, and accounting figures as total

sales, assets, resources of capital, bank loans, employees etc., and business results

for the previous year (2006). Therefore, it is convenient for staff of the statistics

bureaus in each province to conduct the two surveys in parallel: the enterprise

survey and our survey.


46

Survey Procedures

In June 2007, with the introduction from Can Tho University, we contacted

statistics bureaus in each of the six selected provinces. We proposed a cooperation

by conducting our survey together with the enterprises survey. Soon after reaching

an agreement with the statistics bureaus in 2007, a short training was organized to

explain our questionnaire to the staff of the bureaus. Our survey was conducted in

the period July to September 2007 and led to 626 interviews with rice firms in the

Mekong Delta.

Sample Selection

As mentioned, our sample mainly includes rice firms at the four main market

segments of the rice industry in the Mekong Delta. They are millers, w-millers,

wholesalers and retailers. In addition, we interviewed some rice exporters and

gatherers to enable us to better describe the rice market.

In each province, the statistics bureau prepared a list of rice firms from

different districts. The lists included registered rice firms that are millers, w-

millers, wholesalers and SOEs. Each list included all firms located in a particular

district, and the firms were listed in alphabetical order. We decided to select the

first firm of every five firms in the list (numbers 1, 5, 10 etc. of the lists) to take

part in the interview.

Rice retailers are not registered firms. They usually have shops at

marketplaces. These shops are often located in the main street of a local (district)

market. We picked the first shop out of every 10 shops for our interview.

In total, 626 rice firms in the Mekong Delta were interviewed within three

months. As the focus of the study is trade credit use among rice firms in the rice

market of the Mekong Delta, in our interviews we focused on important traders in

domestic markets as processing firms (w-millers/millers) and trading firms

(wholesalers and retailers).


47

The interviewers made an appointment and sent the questionnaires to the

interviewees one week prior to the interview. Next, the interviewers did interview

firm owners (managers) face to face. In each province, we attended the interviews

for two days to observe and/or check whether the interviewees had any questions

regarding the questionnaire and whether the interviewers could explain things

properly. At the same time, we also checked whether the finished questionnaires

were filled in properly. Early October 2007, we again visited each province for two

days, we worked together with the interviewers to check all the information of the

completed questionnaires and collected them.

3.4 Descriptive Statistics of the Sample

The core target of this project is to investigate the relevance of trade credit supply

as a marketing instrument. Existing studies on this topic mainly focus on the

impact of market characteristics on trade credit supply such as market competition,

bargaining power. We therefore concentrate on gathering information on market

context, i.e. firm competitiveness as a measure of market competition at the firm

level, a firm’s bargaining power. This section presents descriptive statistics of the

measures of trade credit supply: firm competitiveness. We also show statistic

figures of other control variables that may influence trade credit supply, such as

firm size, firm age, firm profitability, and the firm’s financial constraints.

3.4.1 Overview of the Sample

Size and Age of Firms In the Sample

The sample includes 158 w-millers, 183 millers, 101 wholesalers, and 145

retailers. It includes six local state-owned enterprises (exporters); the rest of the

sample is private firms. Categories of the survey firms are presented in table (3.1).

In our survey, the size of firms is measured by its total assets. On average,

the value of total assets of the firms in our sample is about 2.5 billion VND


48

(155,097 USD); the firm age is about eight years. Regarding the distribution of the

surveyed firms of the sample, table 3.1 reveals that firms with capital of less than 1

billion VND (62,039 USD) account for 70 per cent of the sample, while firms with

capital more than 1 billion VND make up 30 per cent of the sample.

Table 3.1: Distribution of rice firms according to main business activities

and firm size (measured by total assets)

Categories of the surveyed firms Number of firms Percentages

State-owned enterprises 6 1

Private firms 620 99

Types of firms

Retailers 145 23.3

Wholesalers 101 16.3

Millers 184 29.23

W-millers 158 25.08

Gatherers 17 2.88

Exporters 21 3.35

Size of firms

Less than 100 million VND 204 32.59

From 100 million to 1 billion VND 232 37.06

From 1 to 5 billion VND 131 20.93

More than 5 billion VND 59 9.42

Source: survey in 2007

Table 3.2: Size and age of the surveyed rice firms

Obs Max Mean Min St.dev

FAGE (years) 625 29 8.37 1 5.20

FSIZE(millions VND) 626 269,000 2,520 1 12,300

FSIZE at different

market segments

SOEs/private exporters 21 269,000 31,900 6,560 55,600

W-millers 158 73,300 4,220 18 8,580

Millers 184 6,110 669 15 807

Wholesalers 101 15,000 1,010 11 2,550

Retailers 145 410 35.3 1 55.5

Gatherers 17 2,500 615 170 605

Source: survey in 2007


49

Profitability

Table 3.3: Descriptive statistics of profitable variables

MARGIN Obs. Max Mean Min St.dev

The whole

sample

626 0.61 0.13 0 0.1183

W-millers 158 0.50 0.13 0 0.1289

Millers 184 0.61 0.23 0 0.1106

Wholesalers 101 0.55 0.06 0.003 0.0776

Retailers 145 0.28 0.07 0.003 0.0397

Source: own survey, conducted in 2007

We measure the firms’ profitability by using the price-cost margin.

5 On average,

the price-cost margin of firms in our sample is 0.13. The price-cost margin also

varies much across different market segments: it is higher for processing firms

(average of MARGIN is 0.23 for millers and 0.13 for w-millers) than it is for

trading firms (average of MARGIN is 0.06 for wholesalers and 0.07 for retailers)

(table 3.3).

3.4.2 Trade Credit Provision

Trade credit appears to be used extensively in this particular market. Overall, the

figures in table 3.4 show that 65 per cent of the firms in the sample (i.e. 410 of our

626 rice firms) grants trade credit to their clients. The average amount of trade

credit granted is about 25 per cent of total sales.

5 The price-cost margin is calculated by

ceaveragepri

tsiableaverageceaveragepri cosvar− . Details are

explained in chapter 4, pages 81


50

Table 3.4: Distribution of the surveyed rice firms providing credit

Total

firms

Credit supplying

firms

Amount of trade credit

supply (proportion of

credit sales (%)

Total sample

(%)

626 410

65.49

0.25

W-millers

(%)

158 145

91.77

0.40

Millers

(%)

184 55

28.98

0.13

Wholesalers

(%)

101 74

73.27

0.30

Retailers

(%)

145 101

69.65

0.19


Rice firms at different market segments appear to behave differently with

respect to trade credit provision. The differences in trade credit provision are large

and statistically significant across the four market segments. Trade credit occurs

more frequently in the segments of w-millers, wholesalers and retailers than with

millers. In fact, 92 per cent of w-millers in the sample provide trade credit to

customers, and about 40 per cent of revenues were made on delayed payments in

2006. The percentage of traders at the wholesale segment granting credit decreases

to 73 per cent, while the proportion of credit sales is 30 per cent of the total

revenues. Traders at the retail and miller segments seem to provide less credit to

their clients; on average, 70 per cent at the retail segment offer credit to trading

partners, and the percentage of sales made on credit is about 19 per cent. At the

same time, only 29 per cent of the millers grant credit to their customers, and about

13 per cent of their sales were made on delayed payments in 2006.

Experience with defaulting by clients

On average, 26 per cent of the traders in our sample confirm that they were not

been paid at least once in the last three year by their clients, who have receive

credit. Retailers are most likely to not being paid, since 41 per cent of retailers in


51

our sample mention that they have not been able to get their money back at least

once. This figure is 24 per cent for w-millers and wholesalers, while it is 19 per

cent in the case of millers. The maximum of times that firms in our sample have

not been paid in the last three year is six.

Table 3.5: Experience with defaulting by clients

Number of firms that were

not paid

Total

firms

Number of

firms that

were not paid

Number of

times firms

were not paid

Total sample

(%)

626 165

26.36

0.79

W-millers

(%)

158

38

24.05

0.59

Millers

(%)

184 35

19.02

0.63

Wholesalers

(%)

101 25

24.75

0.44

Retailers

(%)

145 60

41.37

1.57


Trade Credit Received

Table 3.6 Trade credit received

Trade credit obtained

(BUYDP)

Obs Max Mean Min St.dev

Total sample 626 1 0.13 0 0.2275

At market segment level

W-millers 132 0.80 0.14 0 0.2253

Wholesalers 99 1 0.10 0 0.1882

Retailers 145 1 0.13 0 0.3011


The firms in our sample also receive credit from their suppliers. On average, w-

millers purchase about 14 per cent of their total input on delayed payments. In the

case of retailers, 13 per cent of total inputs is purchased on credit, and at the


52

wholesale market this is 10 per cent. Millers are mainly involved in providing

milling services, they seldom purchase paddy/rice and therefore do not receive

credit from suppliers.

3.4.3 Firm Competitiveness

The data in this section reflect firm competitiveness in the rice market. According

to Porter’s five forces, firm competitiveness in an industry is driven by the

following factors: competition with existing rivals, bargaining power vis-à-vis

customers, bargaining power vis-à-vis suppliers, threats of nearby substitutes, and

barriers to new entrants.

In our survey, we use several indicators to measure firm competitiveness.

In this section, we show five indicators that measure the three following factors

that drive firm competitiveness: competition from existing rivals (PERCOMP,

NORIVALS), bargaining power vis-à-vis customer (BPWCUS, SALDEC), and

bargaining power vis-à-vis suppliers (BPWSUP). These factors are very likely to

affect trade credit supply.

Existing competition occurs between businesses in the rice market that

supply products to the same type of clients. We try to measure existing competition

by using two variables: PERCOMP and NORIVALS. PERCOMP is a scale variable

indicating the important role of market competition and its effects on the results of

their rice business, according to the interviewees: (0=no pressure through

competition at all; 1=little competition but not important; 2=competition is

relatively important; 3=competition is important; 4=competition is extremely

important). NORIVALS is estimated by the number of competitors a rice firm uses

as a reference to determine its prices. Although the variation of PERCOMP across

the four segments is relatively low, it is statistically significant. Table 3.7 shows

that on average w-millers consider competition to be more important to their

business activities than other rice traders do. (PERCOMP mean: w-millers 3.32;


53

wholesalers 2.9; millers 2.86; and retailers 2.83.) In general, w-millers check their

selling prices with more competitors to determine their own selling prices (five

competitors) than other rice traders do (three competitors).

Regarding a firm’s bargaining power vis-à-vis customers, the firm can be

more competitive if it has a high bargaining power over its customers with respect

to determining selling prices and contract terms. We use the two important

indicators: BPWCUS and SALDEC to estimate a firm’s bargaining power vis-à-vis

its customers. BPWCUS is estimated by the percentage of sales for which a firm

can set the selling prices without any negotiations with customers. SALDEC is

measured by the percentage sales would decrease if a firm sets its selling price at 1

per cent higher than the other suppliers. Our rice firms state that they can set

selling prices without negotiating with customers for about 44 per cent of total

sales. In addition, using a selling price that is 1 per cent higher than the prices of

other supplies leads to a decrease in sales of about 62 per cent. This shows that

competition in this particular market is fierce. Note that the Mekong Delta yields

about 4 million tons of exporting rice each year, which is about 90 per cent of the

exported rice of the entire country. Consequently, the number of traders is very

high in this region and they process and trade large volumes. To illustrate, the

average size of an order in the w-miller segment is about 94 tons.

Table 3.7: Firm competiveness measures

Obs. Max Mean Min Median St.dev

PERCOMP 626 4 3.04 0 3 0.9773

NORIVALS 599 20 3.99 0 3 3.4874

SALDEC 609 1 0.62 0 0.75 0.3720

BPWCUS 593 1 0.44 0 0.4 0.3728

BPWSUP 583 100 26 0 20 27



54

Moreover, there are large differences in a firm’s bargaining power vis-à-

vis its customers across different market segments. Retailers appear to have more

bargaining power vis-à-vis their clients than traders in other market segments. For

example, retailers can decide on the selling price of 65 per cent of their total sales

(table 3.8) while wholesalers can determine the selling price of 40 per cent of their

total sales. This figure is 43 per cent for millers and 40 per cent for the w-millers.

Regarding a firm’s bargaining power vis-à-vis its suppliers, theoretically it

is less attractive for a firm if its suppliers have strong bargaining power. We use

the variable BPWSUP, measured by the differences in selling prices that induce a

customer to move away from their regular supplier. This indicator aims to capture

the strength of a firm’s bargaining power compared to its suppliers in negotiating

contract terms and prices. Table 3.8 shows that w-millers are very careful with

input prices since a difference of 14 VND in purchasing price per kg paddy may

lead to w-millers switching to another supplier. This indicates that w-millers have a

stronger bargaining power vis-à-vis their suppliers (gatherers). In addition,

bargaining power vis-à-vis the supplier is better for wholesalers than it is for

retailers. A difference of 25 VND per kg rice will cause wholesalers to move to

another supplier, while a difference of 50 VND per kg rice would stimulate rice

retailers to behave similarly.

In short, competition appears to be important to rice traders in the Mekong

Delta. At the same time, the data shows significant differences in the

characteristics of market structure across the four main market segments. W-

millers face the strongest pressure from competition and have the least bargaining

power with customers; as processing firms, w-millers usually buy and sell very

large amounts and therefore their clients may have strong bargaining power. At the

same time, w-millers have strong bargaining power vis-à-vis their suppliers at their

input market. Wholesalers also experience a lot of competition and have a

relatively low bargaining power with their customers. Retailers have the strongest


55

bargaining power vis-à-vis customers; however, they have very little bargaining

power with their suppliers.

Table 3.8: Mean and median of competitive measures across different

market segments

W-millers Millers Wholesalers Retailers Sig.

PERCOMP ***

Mean 3.32 2.86 2.90 2.83

Median 4 3 3 3

NORIVALS ***

Mean 5.2 3.44 3.72 3.81

Median 4 3 3 3

BPWCUS ***

Mean 0.30 0.72 0.70 0.65

Median 0.15 0.80 0.85 0.75

SALDEC ***

Mean 0.72 0.43 0.40 0.34

Median 0.90 0.4 0.30 0.1

BPWSUP ***

Mean 14 18 25 50

Median 10 10 10 50


The one-way ANOVA test for the mean difference of firms’ competiveness measures

across the four different markets segments: w-millers, millers, wholesalers and retailers.

*** indicate significance at the 1 per cent level.

3.4.4 Financial Constraints

A lack of capital can negatively influence a firm’s incentives to provide trade

credit. With the aim of controlling for the impact of the lack of capital on trade

credit provision, we gather information on firms’ financial constraints and their

access to bank loans. First, we asked rice traders to evaluate the importance of their

shortage of capital on the results of their businesses (LACKCAP: 0=no lack of

capital at all; 1=lack of capital is not important; 2=lack of capital is relatively

important; 3=lack of capital is important; 4=lack of capital is extremely important).

According to Luu (2003), financial constraints can make life very difficult for rice


56

traders in the Mekong Delta; the figures in table 3.9 confirm this as the mean of

LACKCAP is 3.

Table 3.9: The importance of financial constraints

LACKCAP Obs Max Mean Min St.dev.

Sample 626 4 3.04 0 0.9904

W-millers 158 4 3.2 0 1.021

Millers 184 4 2.79 0 0.9246

Wholesalers 101 4 3.09 0 1.02

Retailers 145 4 3.15 0 0.9904


We also asked whether rice firms had any difficulty raising bank loans, and if they

had a loan in 2006. 21 per cent of our sample confirmed that they find it difficult to

apply for a bank loan, and 41 per cent of the rice traders in our sample had a bank

loan in 2006. W-millers used bank loans more frequently than traders in other

market segments; for example, 66 per cent of w-millers had a bank loan in 2006.

43.56 per cent of the traders in the wholesaler segment had a bank loan, while 37.5

per cent of the millers used bank loans in 2006. Rice retailers seldom use bank

credit, since only 9 per cent of them had a bank loan. A possible explanation for

this may be that w-millers and millers have factories that they can use as collateral

in applying for a bank loan. Wholesalers may obtain a bank loan by using their

large shops as collateral. Retailers do not have such means, and that explains why

they have fewer bank loans: retailers complained that financial constraints were

making business more difficult.


57

Table 3.10: Firms using bank loans in 2006

Number of firms Percentages

Sample 256 (626) 40.90

At segment level

W-millers 105(158) 66.46

Millers 68 (184) 37.50

Wholesalers 44(101) 43.56

Retailers 13(145) 8.97


3.5 Summary

This chapter describes the statistical data of our sample of respondents: rice firms

operating in the four main market segments of the rice industry. It sheds some light

on the way rice traders in the Mekong Delta work.

Our sample contains 626 rice firms, including four important types of

traders on the domestic rice market: w-millers, millers, wholesalers and retailers. It

shows that market competition is very fierce in this particular industry. It also

indicates that financial constraints are important obstacles to the rice traders of the

Mekong Delta. In addition, it shows that the use of trade credit, market structures

and several firm characteristics vary significantly across the four main segments of

the rice market. For example, we observe that there is a statistically significant

difference in firm size, firm profitability, competitiveness, and trade credit

behavior across these market segments.

Chapter 4

Trade Credit Supply and Firm

Competitiveness

4.1 Introduction

As mentioned in the introduction chapter, the main target of this thesis is to

investigate the motivations of trade credit supply in the rice market. The literature

on the determinants of trade credit supply is rather extensive. Research suggests

that trade credit is provided because suppliers have better information on

customers than banks do (Schwartz, 1974; Petersen and Rajan, 1997) and/or

because it reduces transaction costs (Ferris, 1981; Schwartz, 1974). Trade credit

supply may also be determined by the extent to which suppliers themselves have

access to external finance (Cook, 1999; Delannay and Weil, 2004; Demirguc-Kunt

and Maksimovic, 2002).

A relatively small number of papers focus on whether trade credit supply is

related to the competitiveness of firms. Most papers in this field rather discuss the

competitive environment in which firms are active. Some of these papers claim

that monopolistic suppliers may provide more trade credit, because they are better

able to enforce payments than suppliers in competitive environments (Petersen and

Rajan, 1997; McMillan and Woodruff, 1999). With strong competition, enforcing

repayment of trade credit may be more difficult, since customers can easily switch

to other suppliers. However, other papers claim that suppliers who are faced with

strong competition use trade credit as a means to sell (more) goods (Fisman and

Raturi, 2004). In a competitive environment, trade credit can be used as a tool to


60

reduce the likelihood of customers switching to another supplier. Both positions

are supported by some empirical evidence.

We revisit the relationship between firm competitiveness and the provision

of trade credit but we take a different approach by using data from a single industry

in one single country. This allows for controlling for industry-specific and country-

specific characteristics that may influence the relationship. In particular, we use

data from a large survey among different rice market participants in the

Vietnamese Mekong Delta. The survey contains information on over 600 millers,

w-millers, wholesalers, and retailers. We have chosen Vietnam, because trade

credit is an important source of finance, given that formal financial markets are

relatively underdeveloped. In our sample, 65 per cent of the firms report they

provide trade credit. We focus on one sector (rice) in order to control for the

impact of product characteristics on trade credit use (Wilson and Summers, 2002).

The rice sector plays a prominent role in Vietnam’s economy, especially so in the

Mekong River Delta.

Moreover, when analyzing the relationship between trade credit supply and

firm competitiveness, we take into account the fact that firm competitiveness is a

multi-dimensional phenomenon, determined by interactions between the firm and

its suppliers of inputs, customers and rivals (Porter, 2008). We measure firm

competitiveness in terms of the firm’s perceived competition from its rivals and its

bargaining power vis-à-vis its customers.

Finally, we analyze whether the relationship between some dimensions of

competitiveness and trade credit supply may be dependent on other dimensions of

competitiveness. In particular, we investigate whether the relationship between

customer bargaining power and trade credit supply of a firm depends on the extent

of perceived competition from its rivals.

Our empirical analysis documents that preventing customer switching is a

major reason why firms provide trade credit to their customers. We also show that

Chapter 4: Trade Credit Supply and Firm Competitiveness

61

the impact of customer switching on trade credit supply is stronger in more

competitive market environments.

This chapter is organized as follows. In section 4.2, we present a brief

review of the literature on the use of trade credit and firm competitiveness. In

section 4.3 we explain the country and industry context and show why the issue of

trade credit is of importance in this particular case. Section 4.4 provides a short

discussion of the survey we have used for this study. Section 4.5 then continues

with a discussion of the dataset and the methodology of measuring competition.

The empirical results are described in sections 4.6 and 4.7, after which we provide

a summary and conclusions in section 4.8.

4.2 Trade Credit Supply and Firm Competitiveness: A Survey

The literature focusing on the relationship between trade credit supply and

competitiveness is not very extensive. The available evidence is inconclusive with

respect to the nature of this relationship. Some papers explain the relationship by

pointing at the role of contract enforcement in case a customer defaults on repaying

the trade credit. These papers stress the problems of moral hazard which a trade

credit supplier may face, i.e. he/she may be confronted with the problem of non-

repayment by the customer. Non-repayment may stem from the fact that the

customer lacks the resources to pay, or it may be the result of a deliberate action by

the customer (strategic default) to capture the value of the credit received. The

latter may be prevented by using enforceable contracts. In the absence of such

contracts and/or mechanisms, the supplier may threaten to stop future deliveries.

Yet, the effectiveness of this threat depends on the extent to which the customer

needs future supplies, on the availability of alternative suppliers and/or on the

extent to which suppliers share information on defaulting clients. If alternative

supply is readily available (and assuming that customers need a continuous flow of

supplies for their business and that information sharing among suppliers is low or


62

absent), a customer may switch to a new supplier, which seriously reduces the

effectiveness of the threat of stopping ending deliveries. However, if the supplier is

a monopolist, it will be easier to effectively use the threat of non-delivery. This

explains why monopolistic suppliers may be more willing to provide trade credit

(Petersen and Rajan, 1997; Cunat, 2007; Giannetti et al, forthcoming).

Related to the above discussion, some papers stress the fact that

monopolistic suppliers in general have longer lasting relationships and contact their

clients more frequently than suppliers in competitive markets do. These longer

relationships provide suppliers with more information regarding the type of

customer they deal with. Due to the regular trade relationship they have with their

customers, suppliers have a better opportunity to investigate their customers’

creditworthiness and they are better able to monitor them. Therefore, long trade

relations help reduce information asymmetries, and thus the moral hazard problem.

This may again explain why monopolistic suppliers are more willing to provide

trade credit than suppliers in competitive markets (Petersen and Rajan, 1997; Jain,

2001; Atanasova and Wilson, 2003).

McMillan and Woodruff (1999) find empirical evidence for a negative

relationship between trade credit and competition, using information from trade

relations among small private firms in Vietnam based on survey data from 259

manufacturing firms. In particular, they find that suppliers provide trade credit

more frequently when customers have difficulties finding alternative suppliers,

when the supplier has information about the customer’s creditworthiness based on

long trading relations, and/or when the supplier is a member of a network of

suppliers in which information about customers is exchanged and which serves as a

way of sanctioning defaulting customers. Their conclusion is that suppliers only

provide trade credit to their trading partners if they can ensure that customers will

comply with the agreement to pay late or if they can enforce repayments.

A number of papers stress the importance of trade credit as an instrument

to improve the competitiveness of the supplier. When suppliers are confronted with


63

customers who can easily switch to new suppliers because of strong competition,

trade credit can be used as an instrument to prevent customers from switching. This

may work particularly well in cases where access to bank credit for working capital

purposes is low, making trade credit valuable to the customer. One of the few

papers investigating the positive relationship between trade credit and competition

is by Fisman and Raturi (2004). Using multi-industry firm level data from five

Sub-Saharan African countries, they show that firms that are active in more

competitive markets are more likely to obtain trade credit than firms that are active

in monopolistic markets.

Van Horen (2005) in her paper focuses on a related issue regarding the

relationship between trade credit and firm competitiveness, using data on trade

credit supply for firms from different industries in 42 different countries. She

argues that the bargaining power of the supplier vis-à-vis its customers may affect

decisions to provide trade credit. In particular, the weaker the bargaining power of

the supplier in its trade relations with customers, the higher the probability that

trade credit is supplied, since potential customers may only be willing to buy from

a supplier when trade credit is supplied. In the empirical analysis, Van Horen

shows that young and small firms, which are supposed to have relatively low

bargaining power vis-à-vis their clients, have to offer more trade credit in order to

lock in their customers. Thus, her results suggest that trade credit may be used by

suppliers as a competitiveness device.

In a recent study, Fabbri and Klapper (2008) explicitly focus on trade

credit supply and its relation to the extent of a firm’s market power. Using unique

data for about 2,500 Chinese firms in the manufacturing and service sectors,

including detailed information on the market power of firms in their input and

output markets, as well as details on the terms of trade credit, they show that firms

with weak market power extend more trade credit. Moreover, they show that firms

receiving trade credit from their suppliers of input are more likely to extend trade

credit to their customers, and that these firms tend to extend trade credit against


64

similar contract terms (“maturity matching”) more often when they face strong

competition in output markets and/or have strong market power in input markets.

On the basis of these results, Fabbri and Klapper (2008) argue that trade credit

plays an important role in the supply chain for financially constrained firms when

coping with competition.

To conclude this overview, the above discussion shows that the empirical

literature on trade credit and the competitiveness of firms is inconclusive and

focuses on different aspects of competitiveness. While MacMillan and Woodruff

(1999) focus on the importance of the market power of the supplier, others like

Fisman and Raturi (2004), Van Horen (2005) and Giannetti et al (forthcoming)

stress the importance of the customer’s bargaining power as a determinant of trade

credit supply. Moreover, all these studies use multi-industry and/or multi-country

datasets. Fabbri and Klapper (2008) are the first to look at the importance of both

market structure and competition in determining trade credit supply, using multi-

industry data.

We follow Fabbri and Klapper (2008) by focusing on market competition

and customer bargaining power when evaluating the relationship between trade

credit supply and firm competitiveness. However, in contrast to all previous work

on this subject, we use data from a single industry (rice) in a single country

(Vietnam). Using industry-specific data allows for controlling for the fact that

product characteristics may affect the supply of trade credit (Wilson and Summers,

2002). Moreover, using country-specific data allows for controlling for the

influence of the institutional setting, such as existing formal banking and law

systems, which have shown to be important determinants of trade credit supply

(Fisman and Love, 2003; Giannetti et al, 2008).


65

4.3 The Vietnamese Rice Sector

In this study we focus on Vietnam, since trade credit is considered to play an

important role in trade relations in this country (MacMillan and Woodruff, 1999).

In particular, we focus on the use of trade credit in the rice sector as rice is one of

the most important goods produced in this country. The Mekong Delta in the south

of Vietnam is an important rice producing region. Although the region covers only

12 per cent of the country’s total area, over the last number of years it has

accounted for approximately 50 per cent of total rice production and 90 per cent of

total rice exports.

Rice in the Mekong Delta is generally produced at small farms; a typical

farm producing rice has about 1-3 hectares of land. Rice farms produce paddy,

which is in most cases sold to gatherers. According to a 1995 survey by IFPRI,

gatherers purchase 95 per cent of the paddy produced by farmers (Minot and

Goletti, 2000). These gatherers then sell the paddy to milling firms. The millers

transform the paddy into rice. They may also polish the rice, which raises the

quality. In the Mekong Delta, millers account for about 80 per cent of the total

milling and polishing activities.6 After the rice has been produced and polished, the

millers sell it to wholesalers or directly to retailers, who then sell it to final

consumers in the domestic market. Next to millers that buy paddy and mill and sell

the rice themselves, there are also several millers that only provide milling

services.

The above described marketing channel focuses on rice trading for the

domestic market. Yet, a substantial part of the rice production is exported. The

marketing channel for exported rice is dominated by state-owned enterprises

(SOEs) and a few privately owned exporting firms. The SOEs and private

exporters may buy paddy directly from farms or from gatherers and mill and polish

the rice themselves. They account for 20 per cent of the total milling and polishing

6 Information obtained from the website www.stp.vn.


66

activities in the region. Alternatively, they bring the paddy to millers or buy milled

and polished rice from millers.

The Survey

The data used in this research comes from a survey held among 626 firms that are

active in the Vietnamese rice market in six different provinces in the Mekong

Delta. These provinces are Can Tho, An Giang, Vinh Long, Tien Giang, Hau

Giang and Soc Trang. The survey was held in 2007 and has been conducted by

experienced officials of the provincial statistical offices of the National Statistical

Department. The specific questions regarding trade credit were part of a

standardized nationwide enterprise survey these statistical offices carry out every

five years. This survey is used by the government to collect information in order to

evaluate the country’s economic performance and economic policies and to plan

for future economic developments. The latest enterprise survey took place in July

2007.

Our questionnaire was directed towards the four main types of firms that

are active in the rice market, i.e. the rice millers (including pure millers and

wholesale-millers who also sell rice), rice wholesalers, and rice retailers. For each

firm the questionnaire was filled out by the manager/owner. In each of the six

provinces, we used the list of officially registered firms to select the ones that

finally participated in our survey on trade credit. Based on this list, which was

ordered alphabetically using the names of the company, we selected the first of

every five firms (e.g. number 5, 10, 15, etc.) to join the survey.7 This approach

allowed us to select the millers, wholesalers and exporters for our survey.

However, rice retailers in general are not officially registered. Rice retailers usually

have small shops at local marketplaces in towns and cities. Retailers were selected

by visiting these local marketplaces and randomly picking every first one out of ten

7 This approach was used because it turned out to be too costly to have all firms participate in the

survey.


67

shops the officials of the provincial statistical offices of the National Statistical

Department came across. The gatherers officials came across at these markets were

also randomly asked to participate. Rice millers make up the largest part of our

sample (29 per cent); w-millers who also sell rice, and retailers each represent

around 25 per cent of the total sample; wholesalers account for 16 per cent of the

sample.

The survey included a number of questions related to the extent of

competitiveness of firms. All questions related to measuring competitiveness are

based on the manager/owner’s perception of competitiveness. So, in fact what we

measure is perceived competitiveness, rather than actual competitiveness, which is

mostly based on objective measures, such as market size in terms of shares in total

sales, assets, etc. We explicitly chose to focus on measuring perceptions because

objective measures of total market size in the context of local rice markets in

Vietnam are not available. Moreover, in order to understand what drives the

decision of a manager to supply trade credit, it is more important to look at

perceived competitiveness rather than competitiveness based on objective

measures.8

The importance of trade credit in the Vietnamese rice sector is

corroborated by our data. In total, 65 per cent of all firms in our sample (i.e. 410 of

626 firms) report that they provide trade credit. The importance of trade credit

varies for different types of firms. Of all rice millers who also sell rice, 92 per cent

report that they provide trade credit to their customers. For rice wholesalers and

retailers this is 73 per cent and 70 per cent, respectively; for rice millers, it is only

30 per cent.

When looking at the amount of trade credit provided, again millers who

also sell rice take the leading position. The amount of trade credit these firms

8 The downside of this approach is that perceptions can be biased and therefore over- or

underestimated. We have no reason to believe, however, that this bias is systematically related to

specific firm and/or owner characteristics.


68

provide equals 40 per cent of the total value of their sales. For wholesalers,

retailers and millers the numbers are 30, 19 and 12 per cent, respectively. Exporters

and gatherers provide trade credit equivalent to 28 and 33 per cent of their sales.

4.4 Methodology and Data

4.4.1 The Regression Model

In order to be able to analyze the relationship between trade credit and

competitiveness with the help of data for the Vietnamese rice sector, we estimate

the following regression model:

ininmimkiki TYPEFIRMCOMPTC εββββ ++++= ,,,0 (4.1)

In this model TC is the amount of trade credit supplied (i.e. delayed payments)

divided by total sales, COMP is a vector of k variables, measuring firm

competitiveness as perceived by individual trade credit supplying firms; FIRM is a

vector of m variables, measuring several firm-specific characteristics; i represents

firms. In the empirical analysis we use this simple model to investigate the impact

of different dimensions of competitiveness on the supply of trade credit separately.

Moreover, we investigate the relationship for different types of firms that are active

in the rice market. This is captured by the vector TYPE in equation (4.1). Our

dependent variable TC has a value between 0 and 100 per cent. Since it is a

censored variable, we use a Tobit model with two sides censoring when carrying

out the estimations in this paper. In the estimations, different measures of

competitiveness are included separately in the regression model along with the

firm-specific control and firm-type dummy variables, using data from the entire

sample of firms.


69

Measuring Firm Competitiveness

The literature on the competitiveness of firms, instigated by the work of Michael

Porter (1979, 1980, 1985, and 2008), puts forward the idea that analyses of

competitiveness should take into account the fact that it is a multidimensional

concept. In our analysis, we aim at explicitly addressing the different dimensions

of a firm’s competitiveness and their impact on the supply of trade credit. We

focus on two dimensions of competitiveness of firms: (1) the competitive pressure

from other suppliers (rivalry); and (2) its bargaining power vis-à-vis customers.9

Our vector COMP in equation (4.1) includes two proxies for a firm’s market power

relative to its rivals/competitors and two proxies for its bargaining power vis-à-vis

its customers. The measures we use in the analysis are defined as follows:

(1) PERCOMP is a variable that measures rivalry from other competitors in

the market as perceived by the firm owner. The competitive force from other

suppliers in the market is rated on a scale that runs from 0 (no competition) to 4

(severe competition).

(2) NORIVALS measures rivalry from other suppliers using the number of

competitors in the market the firm uses as a reference group when setting its own

selling price.

(3) BPWCUS is a measure of the bargaining power of the customer relative to

his/her supplier. It measures the percentage of total sales for which the firm can set

9 In Porter’s framework, three other dimensions are discussed: the bargaining power of a firm vis-à-

vis suppliers of inputs, the threat of new entrants, and the existence of substitutes (Porter, 1979, 1980,

and 1985). With respect to the case of the Vietnamese rice market, we have not included these three

dimensions in the analysis. First, the impact of a firm’s bargaining power vis-à-vis its own suppliers

of inputs on the supply of trade credit to its customers is indirectly taken into account by the fact that

we have information about the amount of trade credit the firm has received from its input suppliers.

The logic is that stronger bargaining power vis-à-vis suppliers of inputs increases trade credit

received, which means a firm has more resources to supply trade credit (see below). Second, the

threat of substitution is low as rice is the staple food in Vietnam, which is firmly rooted in

consumption habits. Third, as the rice market is highly competitive, and market entrance and exit of

competitors takes place on a regular basis, the threat of entrance cannot be easily separated from

market competitiveness measured in terms of rivalry from other suppliers.


70

the price without negotiating with the customer. Note that the higher the value of

this variable, the weaker the bargaining power of the customer.

(4) SALDEC is also a proxy for supplier’s bargaining power vis-à-vis its

customers. It is measured as the expected decrease in sales due to customer

switching if the firm were to increase its price by 50 VND per kg rice (i.e. on

average a one percentage change of the price of 1 kilogram rice).

As discussed in the earlier literature on trade credit supply and firm

competitiveness, the extent of competitive pressure due to the existence of rivals in

the market may affect trade credit supply in both ways. The nature of the

relationship between the two depends on whether preventing customer switching or

problems related to contract enforcement is the predominant underlying

determinant of trade credit supply decisions. If preventing customer switching is

the predominant factor, we expect a positive association between both variables of

competitive pressure (PERCOMP and NORIVALS) and the supply of trade credit;

if contract enforcement problems prevail, we expect a negative association. With

respect to bargaining power vis-à-vis the customer, the literature suggests that trade

credit supply increases if customers have a stronger bargaining power, again in

order to prevent them from switching to competitors. Given the definitions of both

variables measuring a firm’s bargaining power vis-à-vis its customers, we therefore

expect a negative association between BPWCUS and trade credit supply and a

positive association between SALDEC and trade credit supply.

Firm-Specific Control Variables

The vector FIRM in equation (4.1) consists of a number of firm-specific control

variables. The choice of these variables is based on the existing empirical literature

on the determinants of trade credit supply. This literature shows that the main

determinants are related to access to financial sources, access to information on the

credibility of customers, and past payment performance of customers regarding


71

trade credit received. On the basis of a survey of the literature, we have included

the following list of firm-specific control variables.

(5) AGE: Existing studies of the determinants of trade credit have shown that

the supply of trade credit is seriously and positively related to the availability of

financial sources. AGE is one of the variables we use to measure the availability of

financial sources to firms. Firm age is used in the literature as a measure of the

availability of credit, since older firms are believed to have a better reputation,

giving them better access to bank loans, and thus to more sources to supply trade

credit (Van Horen, 2005). AGE is measured by the logarithm of the number of

years since the firm was established.

(6) LACKCAP is a measure of the firm’s perceived access to financial sources,

ranging from 0 (no lack of capital at all) to 4 (lack of capital is an extremely

important problem). If a firm indicates it lacks financial sources, it is expected to

provide less trade credit.

(7) FSIZE is the logarithm of the value of total sales of the supplier, which is a

measure of its size. In general, larger firms have better access to sources to provide

trade credit to their clients.10

(8) DIFLOAN is a variable that measures the difficulty a firm has in obtaining

a bank loan. This variable takes the value 1 if the firm indicates it is difficult to

obtain a loan, and 0 otherwise. As with LACKCAP, a firm that indicates it is

difficult to obtain a loan is expected to supply less trade credit.

(9) MARGIN is a measure of the profit margin of a firm and is calculated as

i

ii

AP

AVCAP −. APi is the average selling price per ton of rice of firm i in 2006;

10 In other studies, this measure has also been used as a proxy for the bargaining power of the supplier

relative to its clients (Van Horen, 2005). Yet, the size of a firm is a very indirect measure of

bargaining power.


72

AVCi reflects the average variable costs of firm i in 2006.11

The more profitable a

firm is, the more trade credit it can supply.

(10) If a firm buys inputs for which payments can be delayed – i.e. it receives

trade credit – this increases the financial sources available to supply trade credit.

The variable BUYDP measures the percentage of inputs bought on credit and is

expected to correlate positively with trade credit supplied.

(11) The variable UNPAID measures the number of times a firm has had to deal

with customers who did not paid their trade bill in time (i.e. they did not repay the

trade credit they received) during the last three years. The impact of this variable

on the supply of trade credit is ambiguous. On the one hand, the more a firm has

been faced with non-paying clients in the past, the less it may be inclined to

provide trade credit now or in the future. On the other hand, the firm may feel it

has to provide more trade credit when confronted with non-paying clients, in order

to secure final payment of the trade.

(12) Several studies have shown that the extent to which a supplier has

information on the creditworthiness of its clients influences the amount of trade

credit given (Fisman and Raturi, 2004; McMillan and Woodruff, 1999). The better

the supplier is informed, the more willing he will be to provide trade credit. The

information is obtained by having frequent contact with clients. CONTACT is a

variable measuring the frequency of face to face and/ or phone contact between

supplier and customer, which ranges from 1 (=yearly contact) to 6 (=daily contact).

Type of Firm Dummy Variables

Finally, as mentioned earlier, we investigate the relationship for different types of

firms that are active in the rice market. Given the data we have, the analysis

11 The measurement of the average variable costs differs, depending on the specific market segment

in which a firm is active. For rice millers that also sell rice, the variable costs consist of the costs of

main materials, fuel and labor; for rice-trading firms (wholesalers and retailers) these costs reflect

main materials, transportation costs and labor; and for firms providing milling services the variable

costs consist of fuel and labor.


73

focuses on four different types of firms, i.e. rice milling firms performing milling

services only (MILLER); rice milling firms performing milling services as well as

rice trading (W-MILLERS); rice wholesalers (WHSALE); and rice retailers

(RETAIL). Although all four types are active on the rice market, they perform

different functions and activities. One important difference between firms is the lot

size per transaction. Bulk-breaking is important for retailers; they sell relatively

small amounts of rice to clients and they have to be located close to final

consumers. Competition for these firms is local. In contrast, in the wholesale

market distance becomes less important and competition takes place at the regional

or even national level. The value of wholesale trade transactions is usually much

larger than that of retail transactions. Due to these differences, price levels for

these types of firms differ, although price patterns may be strongly correlated (Lutz

et al, 2006).

Differences in size of transactions also occur between rice milling firms

that perform milling services only and rice milling firms that perform milling

services as well as rice trading. While the first in general have smaller transactions,

the second normally deal with larger volumes. Another difference between the

different types of firms along the rice marketing channel relates to differences in

market conditions, in particular market competition. Competition in retail markets

is usually much stronger than in wholesale markets. Similar differences exist for

the two types of millers: firms that only provide milling services are faced with

much more competition than millers that also sell the rice. These differences may

affect the extent to which different firms provide trade credit. Data on trade credit

supply provided by different firms as discussed in section 4.3.4 corroborates the

above statement.

4.4.2 Descriptive Statistics

Tables 4.1 and 4.2 provide information regarding the descriptive statistics and

correlations of the variables used in the analysis. Table 4.1 first of all shows that on


74

average the firms in our sample provide trade credit (see variable TC) equivalent to

25 per cent of their total sales. The amount of trade credit they receive themselves

is roughly half this percentage as the average value for BUYDP is 12 per cent.

Looking at the descriptive statistics of the measures of perceived competition from

rivals, the average value of PERCOMP of 3 (maximum is 4) suggests that

perceived competition is rather high. Moreover, firms state that on average they

focus on the pricing behavior of four rivals when setting their own selling price.

The descriptive statistics of BPWCUS and SALDEC suggest that the customer

bargaining power is rather strong. The average value of SALDEC is 60 per cent,

suggesting that a small rise of the selling price by the firm makes its customers

move to the competition. The average value of BPWCUS of 44 per cent suggests

that firms have to negotiate with the customer on the price for more than 50 per

cent of its total sales.

Table 4.2 provides some first indications of the existing associations

between variables in our dataset. First of all, especially rice millers who also sell

rice, and to lesser extent wholesalers, clearly provide the most trade credit; the

dummy variables W-MILLER and WHSALE show a relatively high correlation with

the trade credit supply variable TC. In contrast, millers, and to lesser extent

retailers, provide far less trade credit, which is borne out by the fact that the

dummy variables RETAIL and MILLER show a negative correlation with TC. Firm

size (FSIZE) correlates positively with TC, suggesting that larger firms provide

more trade credit. Correlation between BUYDP and TC is positive and relatively

high, which suggests that firms that get more trade credit themselves also provide

more trade credit to their customers.12

Most important to the research in this paper, however, is the fact that both

measures of perceived competition due to the existence of rivals in the market

12

The results for the other control variables reported in the table are less strong, as correlation

coefficients remain below 0.20. Therefore, we do not discuss them in this section but come back to

them, when appropriate, in the next section, in which we discuss the results of the econometric

investigation based on the regression model reported in equation (4.1).


75

(PERCOMP and NORIVALS) show a positive correlation with TC. Moreover, the

bargaining variables (BPWCUS and SALDEC) show a negative and positive

correlation with TC, respectively, although these are less strong. These results

strongly suggest that trade credit is supplied in order to prevent customers from

switching to competitors. We further discuss this issue in the next section, when

we present the results of the econometric investigation, using the regression model

specification reported in equation (4.1).

Table 4.1: Descriptive statistics

Variables Mean Max Min Standard dev. Obs

TC 0.251 1 0 0.255 626

WHSALE 0.161 1 0 0.368 626

W-MILLER 0.252 1 0 0.434 626

MILLER 0.294 1 0 0.455 626

RETAIL 0.232 1 0 0.422 626

EXPORT 0.033 1 0 0.180 626

GATHER 0.027 1 0 0.162 626

AGE 1.877 3.36 0 0.787 625

LACKCAP 3.038 4 0 0.990 625

FSIZE 20.32 28.37 15.61 2.322 625

DIFLOAN 0.209 1 0 0.407 626

MARGIN 0.134 0.61 0 0.118 626

BUYDP 0.127 1 0 0.227 567

UNPAID 0.784 6 0 1.621 624

CONTACT 3.856 5 0 3.855 624

PERCOMP 3.004 4 0 0.977 626

NORIVALS 3.988 20 0 3.487 599

BWPCUS 0.440 1 0 0.372 593

SALDEC 62.33 100 0 37.19 609

7

7

Tab

le 4

.2:

Corr

elat

ion m

atri

x

S

AL

DP

W

HS

A

LE

W-

MIL

L

MIL

LE

R

RE

TA

IL

AG

E

LA

CK

C

FS

IZE

D

IFL

O

MA

RG

B

UY

DP

U

NP

A

ID

CO

NT

A

C

PE

RC

O

NO

RI

V

BP

W

C

SA

L

DE

SA

LD

P

1.0

00

WH

SA

LE

0.1

64

1.0

00

W-M

ILL

0.2

94

-0

.221

1.0

00

MIL

LE

R

-0.3

26

-0

.239

-0.3

38

1.0

00

RE

TA

IL

-0.1

31

-0.2

51

-0.3

5

-0.3

84

1.0

00

AG

E

0.0

51

-0.0

08

-0.0

04

0.0

52

-0.0

54

1.0

00

LA

CK

C

0.0

47

0.0

35

0.1

40

-0.2

50

0.0

88

-0.0

35

1.0

00

FS

IZE

0.2

97

0.1

67

0.4

46

-0.3

67

-0.3

74

0.1

00

0.2

95

1.0

00

DIF

LO

-0

.179

0.0

24

-0.1

17

0.0

95

0.0

06

-0.0

41

-0.0

02

-0.0

38

1.0

00

MA

RG

-0

.094

-0.2

31

-0.0

73

0.6

12

-0.2

97

-0.0

69

-0.2

04

-0..

343

0.0

47

1.0

00

BU

YD

P

0.2

76

0.0

22

0.0

06

-0.2

94

0.2

71

-0.0

60

0.1

51

0.0

90

-0.1

49

-0.2

31

1.0

00

UN

PA

ID

0.1

75

0.0

23

-0.1

73

-0.1

50

0.2

96

0.1

16

0.1

28

-0.0

11

0.0

34

-0.1

54

0.2

15

1.0

00

CO

NT

AC

0.0

80

0.0

30

-0.1

17

-0.0

81

0.1

31

0.0

12

0.0

01

-0.0

10

-0.0

40

-0.1

03

-0.0

17

0.1

68

1.0

00

PE

RC

O

0.2

68

0.0

34

0.1

69

-0.1

60

-0.0

89

0.0

05

0.3

27

0.2

40

-0.0

50

-0.0

87

0.1

04

0.0

15

-0.0

68

1.0

00

NO

RIV

0.2

23

-0.0

11

0.1

95

-0.1

08

-0.0

48

-0.0

28

0.1

38

0.0

75

-0.1

73

0.1

023

0.1

67

-0.1

04

-0.2

03

0.2

57

1.0

00

BP

WC

-0

.169

-0.1

10

-0.2

11

-0.0

25

0.3

57

-0.0

01

0.0

10

-0.2

15

-0.1

28

-0.0

46

0.1

56

-0.0

38

0.0

08

-0.0

59

0.0

34

1.0

00

SA

LD

E

0.1

22

0.0

98

0.2

24

0.1

72

0.0

38

0.0

90

-0.0

91

0.2

12

0.1

52

0.1

51

-0.2

27

-0.1

35

0.0

84

0.1

13

-0.0

77

-0.0

30

1.0

0


78

4.5 Empirical Results and Discussion

4.5.1 Firm Competitiveness and Trade Credit Supply13

Table 4.3 shows the outcomes of the regressions in which the four competitiveness

variables have been added one by one to a base model specification that includes

all firm-specific and firm-type dummy variables.14

This base model specification is

presented in column 1. Thus, we have four models (presented in columns 2-5),

each including one variable measuring perceived firm competitiveness – the firm-

specific variables and a set of dummy variables referring to the different types of

firms in our dataset.

The results show that all four competitiveness measures are significantly

related to the supply of trade credit. The estimation results suggest the following.

First, we find strong evidence that increased competition from rivals in the market

is positively associated with increased supply of trade credit, as both PERCOMP

and NORIVALS are statistically significant. These results confirm that preventing

customer switching is one of the main reasons why firms to supply trade credit.

Second, we find evidence that the bargaining power of a firm vis-à-vis its

customers is associated with trade credit supply as the coefficient for BPWCUS is

negative and statistically significant, whereas the coefficient for SALDEC is

positive and statistically significant. These results again confirm that preventing

customer switching is an important concern for firms, leading them to grant trade

credit to their customers.

When we include all competitiveness variables in the empirical

specification at the same time (shown in column 6), the results remain the same,

13

We conduct a factor analysis to create new indicators reflecting firm competiveness from the

competitiveness variables used in this section. Applying the new indicators, empirical results on the

relation of trade credit supply and firm competitiveness is presented in appendix 1. 14 We left out a dummy for the millers (MILLER) in order to be able to estimate the models.


79

except for the fact that in this specification the coefficient for SALDEC becomes

insignificant. Overall, the estimation results strongly suggest that firm

competitiveness plays an important role when it comes to providing trade credit to

customers. In particular, the results suggest that firms provide trade credit to their

customers in order to reduce customer switching and to keep their market share.

In almost all cases the results for the control variables are as expected. The

variables that relate to access to financial sources have the expected sign and are

statistically significant in most cases, indicating the importance of financial access

as a determinant of the supply of trade credit. LACKCAP is always negative and

significantly related to trade credit supply in all six specifications, indicating that a

lack of capital reduces trade credit provided to customers. The same results are

found for the variable DIFLOAN, which is always negative and statistically

significant in all specifications, which shows that firms that have difficulty

obtaining a bank loan provide less trade credit. With respect to FSIZE, we find that

it is positive and statistically significant in five of the six specifications, indicating

that larger firms, which are supposed to have better access to finance, provide more

trade credit. The variable AGE always has a positive coefficient but is never

significant, suggesting that older firms, which are supposed to have better access to

finance, do not supply more trade credit. The variable MARGIN is positive and

significant in all six specifications, indicating that more profitable firms generally

provide more trade credit. The variable BUYDP is always positive and highly

significant in all specifications. Thus, firms receiving more trade credit themselves,

supply more trade credit to its own customers.

The coefficient of the variable UNPAID, which measures the number of

times a firm was confronted with customers who did not paid their trade bill in

time, is always positive and highly significant. This indicates that firms provide

more trade credit when faced with non-paying clients, which may be a rational

strategy in order to secure repayment in the future.


80

Finally, although the variable CONTACT does have a positive coefficient

as expected, it is almost never significant. This means that in our analysis we

cannot confirm the finding of other studies, which show that firms that contact

their clients more frequently and thus may be better informed about the

creditworthiness of their clients, are more willing to provide credit to customers.

Analysis of the firm dummy variables shows that on average wholesalers,

retailers and millers selling rice provide more trade credit than millers. Moreover,

retailers seem to provide less trade credit than wholesalers and w-millers selling

rice, as the coefficient of the retail dummy is considerably lower than for the other

two dummy variables. These results confirm our earlier discussion of differences

with respect to trade credit supply between different types of firms.

4.5.2 Customer Bargaining Power and The Role of Perceived

Competition

In the previous analysis the impact of perceived competition and the firm’s

bargaining power vis-à-vis its customers on the supply of trade credit have been

investigated separately. However, it can be argued that both dimensions of

competitiveness are, at least potentially, interrelated. In particular, a firm’s

bargaining power vis-à-vis its customers may be weaker if there are more

alternative suppliers in the market. In this situation, the threat of switching to

another supplier is more credible and a firm may then have more reasons to

provide trade credit to retain its sales to existing customers. We may therefore

hypothesize that at higher levels of perceived competition, the effect of bargaining

power on the supply of trade credit may be stronger, whereas at lower levels of

perceived competition the effect may be smaller or even absent.

We test this hypothesis as follows. We split the total sample of firms in

two sub-samples, one including firms with high values of perceived competition

and one including firms with low levels of perceived competition. We create the


81

two sub-samples using our two measures of perceived competition, PERCOMP

and NORIVALS. With respect to the variable PERCOMP firms are in the low

competition sub-sample if they report 0, 1 or 2 for this variable; they are in the

high competition sub-sample if they report 3 or 4 for this variable. With respect to

the variable NORIVALS firms are in the high competition sub-sample if they report

a number of rivals above the average number of rivals reported by all firms in the

sample (i.e. 3.9); they are in the low competition sub-sample if the reported

number of rivals is below the sample’s average.

Next, we run the baseline model specification as shown in column [1] of

table 4.3 for each of these sub-samples and include one of our two bargaining

power variables BPWCUS and SALDEC. Thus, we get eight model outcomes, i.e.

outcomes for the high and low PERCOMP samples using BPWCUS as the measure

for bargaining power (columns [1] and [2]); outcomes for the high and low

PERCOMP samples, using SALDEC as the bargaining power measure in the

regression (columns [3] and [4]); outcomes for the high and low NORIVALS

samples, using BPWCUS in the regression (columns [5] and [6]); and outcomes for

the high and low NORIVALS samples, using SALDEC as the bargaining power

measure (columns [7] and [8]).

The results of the eight models are presented in table 4.4. The results

provide the following general picture. When using PERCOMP to split the sample,

the results in columns [1] and [2] show that for the high and low PERCOMP

samples the coefficient of BPWCUS in both cases is indeed negative; it is

statistically significant only for the high PERCOMP sample. Yet, the size of the

coefficients for both sub-samples shown in columns [1] and [2] are not

significantly different from each other (-0.102 versus -0.092). This does not

support our hypothesis. A similar outcome can be observed when comparing the

results in columns [3] and [4] (high and low PERCOMP samples, using SALDEC

as our bargaining power measure) and [5] and [6] (high and low NORIVALS

samples, using BPWCUS as the bargaining power measure). Only when comparing


82

the outcomes for the NORVALS samples using SALDEC as our measure of

bargaining power, we find supportive evidence for the hypothesis. For these

samples, we observe a statistically significant positive coefficient for the high

NORIVALS sample, which is also larger and significantly different from the

coefficient found for the low NORIVALS sample (0.194 versus 0.051; see columns

[7 and [8]).

Overall, then, our results provide hardly any supportive evidence for the

hypothesis that at high levels of perceived competition, the effect of bargaining

power on the supply of trade credit is strong, whereas at low levels of perceived

competition the effect is small or even absent. These results indicate that the

different dimensions of competitiveness we have distinguished in this study

independently from each other are important determinants of trade credit supply. In

other words, high perceived competition as well as strong bargaining power of

customers may both provide strong incentives to firms to provide trade credit to

their customers

4.6 Summary and Conclusions

In this chapter we have investigated the relationship between trade credit supply

and firm competitiveness. We have looked at important dimensions of

competitiveness, i.e. the competitive pressure from other suppliers (rivalry) and a

firm’s bargaining power vis-à-vis its customers. To control for the influence of

country and industry effects, we have used data on trade credit supply for firms in

one industry in one country. In particular, we have used data from the rice market

of the Mekong Delta in Vietnam to study the relationship between competitiveness

and trade credit supply.

Overall, we find strong evidence for the fact that there is a relationship

between trade credit and firm competitiveness. First of all, stronger market

competition leads to a pressure to provide more trade credit to customers in order


83

to retain market shares and avoid customer switching. Second, when focusing on a

firm’s bargaining power vis-à-vis its clients, the results show that trade credit

supply is positively related to strong customer bargaining power. Again, this

supports the idea that firms provide trade credit in order to retain market shares and

avoid customer switching. Further elaborating on these outcomes we hypothesize

that customer bargaining power may be stronger if there are more alternative

suppliers in the market. The threat of switching to another supplier is more

credible, providing stronger incentives to provide trade credit in order to retain its

sales to existing customers. However, we do not find supportive evidence for this

hypothesis.

These results indicate that the different dimensions of competitiveness we

have distinguished in this study independently from each other are important

determinants of trade credit supply. Thus, high perceived competition and strong

bargaining power of customers both provide strong incentives to firms to provide

trade credit to their customers. Generally speaking, therefore, our results indicate

that studies measuring competitiveness as a one-dimensional concept fail to cover

relevant elements of competitiveness that may influence the provision of trade

credit.

Our research has added new evidence to the debate on the determinants of

trade credit. In particular, it makes a convincing case for relating trade credit

supply to the competitiveness of firms. Trade credit supply is used as an instrument

to keep customers from switching in the rice market in the Mekong River Delta of

Vietnam. Future research should verify to what extent our results are corroborated

in other market settings.

84

Tab

le 4

.3:

Tobit

reg

ress

ions

of

the

rela

tionsh

ip b

etw

een t

rade

cred

it s

upply

and c

om

pet

itiv

enes

s (d

epen

den

t var

iable

: per

centa

ge

of

tota

l

sale

s on d

elay

ed p

aym

ents

)

Var

iab

les

[1]

[2]

[3]

[4]

[5]

[6]

WH

SA

LE

0.3

54

(6.9

9)*

**

0.3

57

(7.1

6)*

**

0.3

70

(7.1

1)*

**

0.3

22

(6.2

0)*

**

0.3

49

(6.9

2)*

**

0.3

31

(6.3

2)*

**

W-M

ILL

ER

0.3

98

(8.0

2)*

**

0.3

81

(7.8

1)*

**

0.3

96

(7.8

1)*

**

0.3

74

(7.5

4)*

**

0.3

97

(8.0

3)*

**

0.3

59

(7.2

6)*

**

RE

TA

IL

0.1

78

(3.8

2)*

**

0.1

91

(3.9

5)*

**

0.1

81

(3.6

0)*

**

0.1

76

(3.5

7)*

**

0.1

97

(3.9

1)*

**

0.2

01

(3.9

8)*

**

AG

E

0.0

19

(1.0

8)

0.0

18

(1.0

3)

0.0

19

(1.0

9)

0.0

20

(1.1

8)

0.0

17

(0.9

8)

0.0

16

(0.9

3)

LA

CK

CA

P

-0.0

25

(-1

.73

)**

-0.0

48

(-3

.17

)***

-0.0

31

(-2

.17

)**

-0.0

17

(-1

.19

)*

-0.0

21

(-1

.49

)*

-0.0

46

(-2

.98

)***

FS

IZE

0

.01

6

(1.8

4)*

*

0.0

17

(1.9

1)*

*

0.0

19

(2.1

5)*

*

0.0

14

(1.6

1)*

0.0

13

(1.4

8)

0.0

17

(1.8

8)*

DIF

LO

AN

-0

.12

4

(-3

.42

)***

-0.1

24

(-3

.50

)***

-0.1

09

(-2

.97

)***

-0.1

41

(-3

.84

)***

-0.1

29

(-3

.56

)***

-0.1

30

(-3

.54

)***

MA

RG

IN

0.6

03

(3.9

0)*

**

0.6

07

(4.0

1)*

**

0.5

32

(3.3

0)*

**

0.5

40

(3.5

0)*

**

0.5

66

(3.3

8)*

**

0.4

70

(2.9

7)*

**

BU

YD

P

0.3

14

(5.1

3)*

**

0.2

97

(4.9

4)*

**

0.2

70

(4.3

9)*

**

0.3

23

(5.3

4)*

**

0.3

23

(5.3

1)*

**

0.2

74

(4.6

2)*

**

8

5

UN

PA

ID

0.0

46

(5.2

1)*

**

0.0

44

(5.0

8)*

**

0.0

51

(5.6

9)*

**

0.0

44

(5.0

2)*

**

0.0

46

(5.2

5)*

**

0.0

46

(5.3

6)*

**

CO

NT

AC

T

0.0

06

(0.5

5)

0.0

11

(0.9

4)

0.0

14

(1.2

0)

0.0

17

(1.3

7)

0.0

01

(0.0

8)

0.0

30

(2.3

8)*

*

PE

RC

OM

P

0.0

64

(4.3

2)*

**

0

.06

7

(4.2

9)*

**

NO

RIV

AL

S

0

.01

4

(3.3

6)*

**

0.0

13

(3.1

3)*

**

BP

WC

US

-0.0

93

(-2

.45

)**

-0

.08

1

(-2

.05

)**

SA

LD

EC

0

.08

0

(1.9

5)*

*

0.0

29

(0.6

9)

CO

NS

TA

NT

-0

.51

2

(-2

.72

)***

-0.6

49

(-3

.45

)***

-0.6

28

(-3

.31

)***

-0.4

71

(-2

.45

)**

-0.4

78

(-2

.56

)**

-0.7

39

(-3

.83

)***

Lo

g l

ikel

iho

od

-1

83

.2

-17

3.9

-1

64

.5

-17

2

-18

0.3

-1

42

.2

Pse

ud

o R

2

0.3

62

0

.39

5

0.4

12

0

.37

6

0.3

63

0

.46

9

Ob

serv

ati

on

s 5

26

5

26

5

02

5

09

5

22

4

85

86

Tab

le 4

.4:

Tobit

reg

ress

ions

of

the

rela

tionsh

ip b

etw

een t

rade

cred

it s

upply

and c

om

pet

itiv

enes

s (d

epen

den

t var

iable

: per

centa

ge

of

tota

l sa

les

on del

ayed

pay

men

ts)

Var

iab

les

[1]

= h

igh

PE

RC

OM

P

[2]

= l

ow

PE

RC

OM

P

[3]

= h

igh

PE

RC

OM

P

[4]

= l

ow

PE

RC

OM

P

[5]

= h

igh

NO

RIV

AL

S

[6]

= l

ow

NO

RIV

AL

S

[7]

= h

igh

NO

RIV

AL

S

[8]

= l

ow

NO

RIV

AL

S

WH

SA

LE

0.2

49

(4.2

2)*

**

0.6

28

(6.0

0)*

**

0.2

50

(4.3

3)*

**

0.7

51

(7.1

8)*

**

0.1

68

(2.3

4)*

**

0.4

09

(5.5

2)*

**

0.1

92

(2.6

5)*

**

0.4

58

(6.5

2)*

**

W-M

ILL

ER

0

.30

8

(5.5

9)*

**

0.5

52

(5.1

0)*

**

0.3

28

(6.0

3)*

**

0.6

24

(5.5

7)*

**

0.2

01

(2.8

4)*

**

0.4

91

(6.8

8)*

**

0.3

13

(3.0

2)*

**

0.5

25

(7.4

6)*

**

RE

TA

IL

0.0

81

(1.4

6)

0.6

00

(5.5

9)*

**

0.0

90

(1.5

9)*

0.6

93

(6.0

7)*

**

0.0

26

(0.3

6)

0.2

80

(4.1

4)*

**

0.1

20

(1.3

9)*

*

0.2

88

(4.3

0)*

**

AG

E

0.0

30

(1.5

4)*

-0.0

29

(-0

.87

)

0.0

22

(1.1

4)

-0.0

24

(-0

.70

)

-0.0

42

(1.5

9)

0.0

09

(0.4

0)*

0.0

35

(1.3

6)

0.0

08

(0.3

4)*

LA

CK

CA

P

-0.0

34

(-1

.87

)**

-0.0

52

(-2

.14

)**

-0.0

30

(-1

.69

)**

-0.0

64

(-2

.69

)**

-0.0

25

(-1

.02

)

-0.0

40

(-2

.01

)

-0.0

32

(-1

.35

)*

-0.0

41

(-2

.12

)

FS

IZE

0

.00

9

(0.8

4)

0.0

39

(2.3

5)*

*

0.0

07

(0.6

5)

0.0

47

(2.8

7)*

**

-0.0

06

(-0

.40

)

0.0

23

(2.0

2)

-0.0

04

(-0

.25

)

0.0

21

(1.8

5)

DIF

LO

AN

-0

.12

8

(-3

.06

)***

-0.1

58

(-2

.37

)**

-0.1

10

(-2

.70

)**

-0.1

70

(-2

.52

)**

-0.0

44

(-0

.56

)

-.0

13

3

(-3

.04

)*

-0.0

71

(-0

.90

)

-0.1

13

(-2

.65

)

MA

RG

IN

0.4

18

(2.3

4)*

*

1.4

01

(4.9

6)*

**

0.4

15

(2.3

1)*

*

1.4

53

(4.9

1)*

**

0.2

64

(1.0

8)*

0.6

50

(3.1

2)

0.2

78

(1.1

5)*

*

0.7

04

(3.4

0)*

BU

YD

P

0.3

86

(5.5

5)*

**

0.0

51

(0.5

0)

0.3

86

(5.5

4)*

**

0.0

60

(0.5

5)

0.4

95

(4.9

0)*

**

0.2

31

(3.0

2)*

**

0.4

83

(4.8

6)*

**

0.2

37

(3.0

4)*

**

8

7

UN

PA

ID

0.0

39

(3.8

3)*

**

0.0

65

(4.3

2)*

**

0.0

41

(3.9

9)*

**

0.0

73

(4.7

3)*

**

0.0

27

(1.5

7)*

*

0.0

56

(5.2

6)*

*

0.0

21

(1.1

7)*

*

.05

9

(5.4

6)*

**

CO

NT

AC

T

0.0

14

(1.0

9)

0.0

67

(1.8

0)*

0.0

05

(0.3

6)

-0.0

02

(-0

.07

)

0.0

04

(0.2

3)*

-0.0

36

(-1

.90

)

-0.0

13

(-0

.77

)

-0.0

08

(0.4

7)*

BP

WC

US

-0

.10

2

(-2

.32

)**

-0.0

92

(-1

.41

)

-0.1

14

(-1

.94

)**

-0.1

14

(-2

.26

)**

SA

LD

EC

0

.07

7

(1.5

9)

0.1

16

(1.6

3)

0.1

94

(2.5

3)*

**

0.0

51

(0.9

9)

CO

NS

TA

NT

-0

.20

6

(-0

.93

)

-1.4

14

(-3

.75

)***

-0.2

32

(-1

.06

)

-1.4

51

(-4

.08

)***

0.1

92

(0.5

8)

-0.7

57

(-3

.10

)**

0.0

55

(0.1

8)

-0.7

13

(-2

.99

)**

Lo

g

like

lih

oo

d

-12

3.5

-1

.19

-1

29

.3

-22

.8

-42

.5

-11

9.4

-4

2.6

-1

28

.2

Pse

ud

o R

2

0.3

58

0

.72

7

0.3

41

0

.71

0

0.3

68

0

.41

5

0.3

81

0

.39

3

Ob

serv

ati

on

s 3

72

1

37

3

79

1

43

1

68

3

41

1

69

3

53

Chapter 5

Trade Credit Supply and Customer

Characteristics

5.1 Introduction

In chapter 4 the analysis has been done by using variables at firm level, such as

competition, bargaining power, and profitability, to explain trade credit provision.

This implies that the analysis in the previous chapter can only provide empirical

evidence for variables explaining the average amount of trade credit given by a

supplier. Obviously, using firm level data does not allow us to make a direct link

between the individual user and the individual supplier of trade credit. Therefore,

this approach is broad and does not provide answers that relate to the selection

process: who do suppliers grant more trade credit to?

This chapter uses transaction level data and brings the analysis one step

further by examining the impact of customer characteristics on trade credit

provision. Regardless of the fact that a supplier provides little or a lot of credit, the

supplier does not treat all customers the same. In other words, suppliers usually

decide to grant more credit to some customers than to others. The main aim of this

chapter is to identify the characteristics that give some customers better access to

suppliers’ credit. In particular, we will concentrate at several issues that can be

measured much more precisely with transaction level data than the firm level data.

For example, the length of the trading relation, the number of visits to an individual

Trade credit in the rice market of the Mekong Delta in Vietnam

90

customer before granting credit, frequent contacts, the size of an order, the history

of payments, and the bargaining power of a customer.

Our design is straightforward. We aim to investigate to whom suppliers are

willing to provide more credit. At the same time, we control for several important

suppliers’ characteristics, e.g. market competition, lacking capital, firm size, and

age. This empirical setting is expected to provide an overall view of customer

characteristics that allow customers good access to suppliers’ credit sources.

This chapter contributes to trade credit literature along a few dimensions.

First, existing studies on this topic use cross-industry data. Yet, the use of trade

credit is significantly influenced by the traded goods and/or specific industry

characteristics (Summer and Wilson, 2002; Giannetti et al, 2008). By focusing on

one industry in one country, we eliminate the impact from different industry and/or

products characteristics on trade credit provision. Accordingly, we are able to

measure the impact of customer characteristics on trade credit provision more

carefully. Second, studies on this topic using transaction level data are rare and

they omit some important characteristics that strongly affect trade credit provision

such as customer bargaining power and market competition. We use transaction

level data and cover more customer specific variables in our empirical analysis.

The remainder of this chapter is organized as follows. In section 5.2, we

present a brief review of the literature on trade credit and customer characteristics.

Section 5.3 introduces the hypothesis, the empirical strategy and data. We discuss

empirical results in section 5.4, and section 5.5 concludes the chapter.

5.2 Related Literature

In the literature, studies investigating the impact of customer characteristics on

trade credit provision are rare and can be divided into two categories. The first

category of studies uses firm level data and stresses the importance of customers’

bargaining power on trade credit provision. For example, Summer and Wilson

Chapter 5: Trade Credit Supply and Customer Characteristics

91

(2003) use surveyed data of small businesses in the UK and show that the size of

buyers can very much affect trade credit provision. They employ a dummy variable

that takes the value of 1 if firms agree with the statement that their markets are

dominated by large buyers as the main characteristic of customers. The authors

find that firms provide more credit if large clients dominate their markets. As a

conclusion, the authors argue that selling to large clients induces small suppliers to

provide more credit to attract large buyers.

Another study that focuses on the relation between trade credit provision

and customers’ bargaining power is Van Horen (2007). Using a dataset of 5,164

firms from 20 countries in Eastern Europe and Central Asia, Van Horen (2007)

measures the customers’ bargaining power by a dummy variable that takes the

value of 1 if the percentage of sales to the three largest customers is more than 20

per cent. This study proves that the customers’ market power has a positive impact

on trade credit provision. The author argues that customers with strong market

power would demand to purchase goods on credit to increase their surplus.

Accordingly, suppliers that sell to large firms with high market power are inclined

to extend more credit.

Fabbri and Klapper (2008) use data of 2,500 Chinese small and medium

enterprises and investigate the impact of customer bargaining power and market

competition on trade credit provision. The authors use a dummy variable that takes

the value of 1 if the percentage of total sales accounted by the largest customer is

larger than 5 per cent and 0 otherwise, as an indicator of customer bargaining

power. This study also shows a positive effect of customers’ bargaining power on

trade credit provision.

The second category of studies employs transaction level data and

emphasizes the effect of customer creditworthiness on trade credit provision: e.g.

McMillan and Woodruff (1999), and Aaronson et al (2004). McMillan and

Woodruff (1999) use a sample of 259 Vietnamese firms and analyze trade relations

between each firm and two specific customers. The study investigates how the


92

customers’ ability to switch to another supplier and characteristics of trading

relations influence a supplier’s willingness to provide credit. The authors show that

if it is difficult to find an alternative supplier for the traded goods, which can be

interpreted as low market competition, suppliers are more willing to provide credit.

The major finding is a negative relation between market competition and trade

credit provision, which contrasts with the positive relation found in other studies

(Fisman and Raturi, 2004; Van Horen, 2005). In addition, the authors also

document that more trade credit will be granted to a customer if his suppliers

receive more information on customers through business networks or trading

relationships.

Aaronson et al. (2004) use surveyed data from their own survey and study

trade relations of small businesses in Chicago. This study examines the impact of

the ability to collect information about customers, with a special focus on

geographic distance and ethnic ties on trade credit, using data at the transaction

level. The study shows that working with a nearby supplier allows customers to

receive more credit from suppliers. Moreover, it also indicates that Hispanic firms

are more likely to get more credit from firms owned by other Hispanics. The

authors argue that suppliers may know more about customers located nearby

and/or belonging to the same ethnic group.

In general, customers’ bargaining power and creditworthiness are found to

be important factors with respect to trade credit provision. Yet, existing studies on

this issue employ cross-industry analysis and do not control for the impact of the

transacted goods and industry characteristics that affect trade credit provision

strongly (Wilson and Summer, 2002; Giannetti, Burkart and Ellingsen, 2009).

Second, these studies omit several important factors that may influence trade credit

provision. For example, McMillan and Woodruff (1999), and Aaronson et al.,

(2004) do not control for customers’ bargaining power, which significantly

influences trade credit provision.


93

The empirical analysis in this chapter will cover all above mentioned

variables that affect trade credit supply. For example, the size of the orders,

customers’ bargaining power, the length of a trading relation with his suppliers, the

number of times a customer does not pay suppliers in time, the frequency of

contact and orders, etcetera, will be included. Consequently, we expect to cover a

broader range of customer specific characteristics than previous studies have.

5.3 Hypothesis and Methodology

5.3.1 Hypothesis Development

Customer creditworthiness can be a significant issue that suppliers need to take

into account when considering whether or not to provide credit to a customer.

Suppliers may prefer to grant credit to customers they know well, so that they can

be sure that clients will repay. Consequently, we hypothesize that firms may supply

more credit to customers they know better through (a) visiting the customer’s

business before providing credit, (b) the length of the trading relation, and (c) the

customer’s habit of repaying in time.

H1a: Suppliers grant more credit to customers who they have visited before.

H1b: Firms grant more credit to customers with whom they have a longer trading

relation.

H1c: Firms grant less credit to customers that in the past were unable to repay in

time.

Several studies also suggest that suppliers offer trade credit to build a

relation with large customers to generate sales. Therefore, firms provide more

credit to customers who order large quantities since they are high- potential

customers. It may also be the case that large customers have high bargaining power

and demand more credit from suppliers. Consequently, suppliers have to grant


94

more credit to large buyers/buyers with high bargaining power (Summer and

Wilson, 2002; Van Horen, 2007).

H2a: Firms are more willing to provide credit to clients that order large

quantities.

H2b: Firms are more likely to provide credit to customers that have a high

bargaining power.

5.3.2 Empirical Methodology

As mentioned earlier, this chapter aims to provide an insight into the suppliers’

selection process as to who they will grant credit. Consequently, it examines the

link between trade credit supply and customer characteristics based on data

collected through our questionnaire. The data contains detailed information about

trade credit at the transaction level, characteristics of trading relations among

suppliers and specific customers, customer characteristics and some relevant

supplier characteristics.

We estimate the following regression model:

DPCUS ij = α + βCustomercharacteristicij + δ Xi + ε ij (5.1)

In our specifications, the dependent variable is the proportion of the payments that

customer j is allowed to delay in making payments to supplier i. Customer

characteristics are the core explanatory variables. Important measures of

customers’ characteristics are employed to capture the impact of clients’

characteristics on trade credit supply. These measures will be explained in the next

section. X is a vector of control variables including supplier characteristics and

market characteristics that may affect supplier incentives to offer credit to

customers. In general, the following set of variables will be employed in our study.


95

Empirical Measures

Our dependent variable DPCUS15

ij, being the proportion of the payment made after

the delivery of goods by customer j to supplier i, always falls between 0 and 1.

Therefore, we treat this variable as a censored variable and we employ a standard

Tobit model in our study.

Information On Customer Characteristics and Trading Relations

Suppliers may be willing to provide more credit to a customer that they know

better and consider to be a creditworthy client to ensure repayment. The first set of

variables attempts to reflect customer and trade relation characteristics that can

affect a supplier’s incentives to provide credit to a specific customer.

(1) PAYLATEij: the numbers of times customer j has not paid supplier i on time

during the last three years.

(2) LOGLENGREij: the natural log of the length of the trading relation in days

between supplier i and customer j.

(3) VISITij: the number of times supplier i visited the shop of customer j.

(4) CONTACTij: we ask each supplier how often customer j usually contacts

supplier i, either by calling or by actually visiting, to ask about selling prices.

Frequent contact with the supplier may provide more information about the

customer’s business and leads to more trade credit to the customer (Fisman and

Raturi, 2004). In this chapter, we use CONTACTij as a scale variable ranging from

1 to 6. This variable measures the frequency of face to face and/or phone contact

between a supplier and a specific customer:, which ranges from ( 1=yearly;

2=every six months; 3=every three months; 4=monthly; 5=weekly; and 6=daily).

15

We do not define the time gap between delivery and payment in the definition of dependent

variable since it is a common practice that rice suppliers receive the payment of one order when the

next order is delivered (see McMillan and Woodruff, 1999)


96

The decision to grant credit to an individual customer may also be

influenced by the size of the customer’s orders and his bargaining power. Some

theoretical studies suggest that suppliers provide trade credit to encourage

customers to order in large quantities. An increasing in the proportion of credit

sales (the proportion of sales on delayed payments) will result in larger order sizes

(Yung, Fu Huang, 2007; Rachamadugu, 1989). Therefore, we include variable

IMCUSij to study whether suppliers prefer to provide credit to customers that place

large orders.

(5) IMCUSij is expected to measure the size of customer j’s average order. It is

measured by the ratio of the average order from customer j to supplier j’s average

monthly revenue.

It may also be the case that customers with high bargaining power demand

to purchase on credit. Accordingly, suppliers selling to clients with high bargaining

power might grant more credit to them (see Van Horen, 2007; Fabbri and Klapper,

2008; Summer and Wilson, 2002).

(6) SUPOWij measures supplier i’s bargaining power vis-à-vis customer j. It is

a dummy variable equal to 1 if supplier i can set the selling prices and contract

terms without negotiating in the trade relation with customer j, and 0 otherwise.

The Measurement of Supplier Characteristics

The decision to provide credit to customers is also influenced by various supplier

characteristics. Our second set of independent variables reflects some general

supplier characteristics that affect a supplier’s willingness to grant credit to

customers. First, some theoretical studies suggest that firms with good access to

external financial markets and/or excess credit at low interest rates are more

willing to provide credit to their trading partners (Schwartz, 1974; Smith, 1988).

Empirical studies use firm size and firm age as proxies for access to external

finance. Large and old firms may have better access to external financial markets,


97

and therefore they may be more capable of offering credit to trading partners (Van

Horen, 2005; Fisman and Raturi, 2004). In contrast, firms that face credit

constraints may not be able to provide trade credit. We pick up the effects of being

credit-constrained by using LACKCAPi in our empirical analysis. Second, some

studies also show the impact of supplier profitability on suppliers’ incentives to

provide credit to customers. For example, Petersen and Rajan (1999) show that

profitable firms are more willing to provide credit to customers. Third,

experiencing default in the past may also affect supplier incentives to grant credit.

We control for the supplier characteristics mentioned above by the following

variables.

(7) LACKCAPi is a five-scale variable measuring the importance of a lack of

capital estimated by the firms’ owners. This value can be 0=no lack of capital at

all; 1=lack of capital is not a problem; 2=lack of capital is somewhat problematic;

3=lack of capital is an important problem; and 4=lack of capital is extremely

important.

(8) FSIZE: the natural logarithm of capital is used as a proxy for firm size;

several empirical studies on this topic show the impact of firm size on trade credit

provision (Van Horen, 2005; Fisman and Raturi, 2004).

(9) LOGAGEi: the natural log of firm age, measured in years from the year a

firm has been established until 2007.

(10) BEUNPAIDi: the fact of being unpaid by clients in the past can strongly

affect sellers’ incentives to provide credit to customers. Therefore, we include

BEUNPAIDi, measured by the number of times a firm has not been paid by all of

its customers in the last three years of our study.

(11) MARGINi: we measure a firm’s profitability by estimating the price cost

margins.

i

ii

iAP

AVCAPMARGIN

−=


98

where APi is the average selling price per ton rice of firm i in 2006, and AVCi is the

average variable cost of firm i in 2006. The average variable cost is defined as the

total costs including the purchasing price of rice/paddy as a main material, fuel

costs, transportation costs, and labor cost.

The Measurements of Market Circumstances

The third set of variables in our study is a proxy for the characteristics of the

product market in which suppliers operate. We attempt to pick up the effect of

market competition perceived by using the following measure.

(12) PERCOMPi: a five-scale variable reflects the importance of competition as

perceived by a supplier. This value ranges from 0 to 4 (0=no competition at all;

1=competition is not important; 2=competition is relatively important;

3=competition is important; and 4=competition is extremely important).

In addition, several of the above characteristics may differ substantially

across market segments, which may affect trade credit provision. Therefore, we

include a dummy variable for each market segment to control for the market

segment characteristics:

(13) WHSALEi is a dummy variable that takes the value of 1 if a firm is a

wholesaler and 0 if otherwise.

(14) W-MILLERi is a dummy variable that takes the value of 1 if a firm is a

miller that sells rice and 0 if otherwise.

(15) MILLERi is a dummy variable that takes the value of 1 if the firm’s main

business activity is to provide milling services and is not involved in trading, and 0

if otherwise.

(16) RETAILi is a dummy variable that takes the value of 1 if a firm is a retailer

and 0 if otherwise.


99

5.3.3 Descriptive Statistics

The data used in this chapter comes from the 2007 rice firm survey in the rice

market of the Mekong Delta. These firms were drawn from the lists of rice firms

provided by the provincial statistic bureau in each province. The data includes rice

firms that operate at the four main market segments of the rice market in our

quantitative analysis: 158 w-millers, 100 wholesalers, 145 retailers, and 184

millers.

The core of our survey was to collect information on trade credit provision

and firm trade relations with two specific customers. We asked rice firms for

details of two trade relations with two specific customers, of whom one is one of

their larger clients while the other is a smaller client. The reason for this selection

method was that one of our targets was to measure the impact of customer

bargaining power and customer size on trade credit provision. For each of the two

specific relationships, firms provided information on the proportion of the

transaction value that was usually paid on delayed payments. Regarding the length

of the credit period, rice firms mentioned that rice suppliers usually allow clients to

settle payments for one order when the next order is delivered. In addition,

suppliers do not explicitly charge any interest on the credit sales.

Table 5.1 presents the definitions of variables and table 5.2 provides

descriptive statistics of trade credit and the above mentioned variables. Trade credit

varies considerably from one customer to another, ranging from 0 to 100 per cent

of all bills. Among the 1252 trade relations we investigated, 58 per cent of firms

provided trade credit, and the average proportion of delayed payments was 37 per

cent.

Trade credit provision also differs a lot across different market segments.

Trade credit appears to be most popular with w-millers. 87 per cent of the trade

relations of w-millers in the sample involve in trade credit. Moreover, our

wholesalers provide more credit to clients than our millers and retailers. About 71


100

per cent of the trade relations with rice wholesalers in the sample dealt with trade

credit provision. At the retail segment, the proportion of the trade relations

decreased to 33 per cent, and only 26 per cent of millers in the sample granted

credit to clients.

Another important part of our questionnaires deals with customer

characteristics. We observe that about 49 per cent of the surveyed firms say that

they visit customers at least once before giving them credit. Besides, the customers

of the rice firms appear to contact their suppliers on a monthly basis (mean=4). In

our sample, only 2 per cent of the surveyed firms say that the interviewed

customers used to pay late. This figure shows that the risk of being paid late is

relatively low in this industry.

As mentioned in the previous section, we measure a customer’s bargaining

power through SUPOWij (a dummy variable that is 1 if supplier i can set the selling

price and contract terms without any negotiation with customer j). About 25 per

cent of the surveyed rice firms in our sample say that they have a strong bargaining

power and that they set selling prices without negotiating.

The size of a customer is measured by IMCUSij, the ratio of customer j’s

average order to supplier i’s average total monthly sales. When the value of

IMCUSij is equal to 1, an average order by the client j equals the average sales of

one month of supplier i. In our sample, the average value of IMCUSij is 0.45, which

indicates that the average orders from interviewed clients account for two weeks

sales. This figure is high since several firms purchase in bulk. It is quite usually

that rice exporters sign contracts with foreign importers. In the next step,

SOEs/exporters sign contracts with w-millers, and it takes w-millers several

months to gather and process enough rice for a contract with an exporter. This

explains why the maximum value of IMCUSij is 7.54, which means that the average

order from the customer j equals the sales of 7.5 months of supplier i.


101

Table 5.1: Variable definition

Variables Definition

Measures of Trade Credit

DPCUSij

The proportion of the delayed payment, which is paid after the

delivery of goods by customer j to supplier i.

Measures of Customers’ Characteristics

IMCUSij

The ratio of an average order from customer j to his supplier i

‘s average monthly sales

SUPOWij A dummy variable takes the value of 1 if supplier i can set

the selling price in the trade relation with customer j, without

experiencing negotiation and 0 otherwise.

LOGLENGREij Natural log of the length of the relationship between supplier

i and customer j (in days)

VISITij The number of times supplier i visits customer j’s shop

before granting credit to the customer.

CONTACTij a scale variable ranging 1 to 6, showing how often a firm

contact with the specific customer, (1=yearly contact), (2=

contact every half year), (3= quarterly contact), (4= monthly

contact), (5 = weekly contact) (6= daily contact)

PAYLATEij The numbers of times customer j has not paid supplier i on

time and asked for extra extension during the last three years.

Measures of Suppliers’ Characteristics

LOGCAPi

As a measure of firm size, natural logarithm of a firm’s

capital

LOGAGEi Natural logarithm of a firm’s age, the period in years since a

firm was established to 2007


102

Variables Definition

LACKCAPi A five scale variable measuring the importance of lacking

capital. This value ranges from: (0 = totally no lacking

capital) (1 = lacking capital is not a problem) (2 = lacking

capital is a little problematic) (3 = lacking capital is an

important problem) (4 = lacking capital is an extremely

important problem).

MARGINi MARGINi =

i

ii

AP

AVCAP − where

APi: the average selling price per ton rice of the firm i in 2006

AVCi: the average variable cost of firm i in 2006.

BEUNPAIDi The number of times that supplier i has been unpaid by its

customers in the last three years

Measures of Market Circumstance

PERCOMPi A five scale variable showing the importance of competition.

This value ranges from (0 = totally no competition), (1=

competition is not important), (2 = competition is relatively

important), (3= competition is important), and (4= competition

is an extremely important issue)

WHSALEi A dummy variable takes the value of 1 if supplier i is a

wholesaler and 0 otherwise

W-MILLERi A dummy variable takes the value of 1 if supplier i is a miller

that sells rice and 0 otherwise

MILLER A dummy variable takes the value of 1 supplier i is a miller

that mainly provides milling service and do not involve much

in any trading activity. Otherwise, this variable takes 0.

RETAILi A dummy variable takes the value of 1 if supplier i is a retailer

and 0 otherwise

1

03

Tab

le 5

.2:

Corr

elat

ion m

atri

x

Dpscus

Imcus

Supow

Contact

Visit

Lengre

Paylate

Lackcap

Fsize

Beunpaid

Margin

Logage

Percomp

Mill-seller

Wholsa

Miller

Retail

Dp

scu

s 1

.00

Imcu

s 0

.28

1

.00

Su

po

w

-0.2

2

-0.2

2

1.0

0

Co

nta

ct

0.0

5

-0.2

3

0.1

0

1.0

0

Vis

it

0.2

3

0.1

4

-0.0

5

0.1

8

1.0

0

Len

gre

0

.07

-0

.04

0

.01

0

.19

0

.04

1

.00

Pay

late

-0

.04

-0

.04

0

.04

-0

.01

0

.07

-0

.05

1

.00

Lac

kca

p

0.0

5

0.0

9

0.0

8

0.0

6

0.1

2

-0.0

9

-0.0

8

1.0

0

Fsi

ze

0.3

1

0.2

6

-0.3

9

-0.1

2

0.2

0

-0.0

3

-0.0

2

0.1

6

1.0

0

Beu

np

aid

0

.17

0

.06

0

.10

0

.12

0

.09

0

.00

-0

.02

0

.10

-0

.15

1

.00

Mar

gin

0

.06

0

.02

-0

.07

0

.06

-0

.07

0

.01

0

.10

0

.04

0

.11

-0

.04

1

.00

Lo

gage

-0.1

1

-0.0

4

0.0

9

0.0

8

0.0

0

0.1

8

0.0

5

-0.0

9

-0.1

0

0.0

2

-0.0

6

1.0

0

Per

com

0

.24

0

.20

-0

.03

-0

.06

0

.16

-0

.02

-0

.07

0

.36

0

.27

0

.03

0

.06

-0

.08

1

.00

W-m

ille

r 0

.35

0

.48

0

.16

0

.11

0

.17

0

.05

0

.03

0

.08

0

.50

0

.06

0

.08

0

.06

0

.18

1

.00

Wh

sale

0

.11

-0

.03

-0

.03

0

.03

0

.07

0

.11

-0

.04

0

.04

0

.01

-0

.04

-0

.08

-0

.04

-0

.02

-0

.25

1

.00

Mil

ler

-0.2

6

-0.2

0

-0.3

1

-0.1

3

-0.2

0

-0.1

0

-0.0

1

-0.1

7

0.0

5

-0.1

5

0.0

7

0.0

2

-0.1

0

-0.3

6

-0.2

9

1.0

0

Ret

ail

-0

.18

-0

.22

0

.51

0

.17

-0

.08

0

.01

0

.07

0

.08

-0

.70

0

.27

-0

.11

0

.06

-0

.10

-0

.32

-0

.26

-0

.37

1

.00


104

Table 5.3: Summary statistic of the sample

Obs Mean Min. Max Std.

Measures of trade credit

DPCUSij 1174 0.372 0 1 0.419

Measures of customers’

characteristics

IMCUSij 1135 0.4501 0.00000315 7.537 1.055

SUPOWij 1153 0.251 0 1 0.434

CONTACTij 1170 3.814 0 5 1.365

VISITij 1108 0.921 0 6 1.244

LOGLENGREij 1162 3.304 0 6.897 1.463

PAYLATEij 1172 0.014 0 1 0.116

Measures of suppliers’

characteristics

LACKCAPi 1172 3.039249 0 4 0.983

FSIZE (Logcapi) 1172 19.291 13.815 25.018 2.135

BEUNPAIDi 1170 0.791453 0 6 1.619

MARGINi 1172 0.1313579 0 0.610 0.117

LOGAGEi 1172 1.875233 0 3.367 0.785

Measures of market

circumstances

PERCOMPi 1172 2.982935 0 4 0.976

W-MILLERi 1174 0.170 0 1 0.376

WHSALEi 1174 0.269 0 1 0.444

MILLERi 1174 0.313 0 1 0.464

RETAILi 1174 0.247 0 1 0.431


105

5.4 Estimation Results and Discussion

5.4.1 The Impacts of Customers’ Characteristics

We start by estimating the specific to general approach by adding variables one by

one. Next, we include all our possible variables in our model. Thus, we are able to

test the stability of the remaining variables. We also adjust for cluster effects and

we standardize the coefficients in our estimations.

First, we will focus on the effects of customer characteristics and

characteristics of trading relations on trade credit provision. The empirical results

in table 5.4 provide support for the first hypothesis that better information about a

customer results in more trade credit being offered to the customer. First, we find a

significant and positive effect of VISITij on the proportion of payment made after

delivery. The estimated coefficients in table 5.4 show that the level of trade credit

provided to a customer increases with 7 per cent if a supplier visits the customer’s

shop once (β=0.071, t=4.03). These visits may be a way for suppliers to get to

know the customer’s shop and business. Therefore, if rice suppliers visit a

customer several times, the customer will be able to get more trade credit. Second,

we observe that a longer trading relation is significantly and positively associated

with trade credit offered. The length of the trade relation can reveal more private

information about the client’s creditworthiness and the business. In addition,

frequent contact tells the supplier more about the customer. We also observe that

frequent contact enables customers to get more trade credit from suppliers. In

general, the estimation results show that a customer will receive more credit from

his suppliers if the supplier has better information on the customer’s business.

Another important characteristic that influences trade credit supply is

concerned with the customer’s creditworthiness. We proxy for creditworthiness by

the number of times a customer did not pay in time during the last three years.

Unlike our expectation, we do not find a significant and negative impact of


106

PAYLATEij on the proportion of trade credit provision. This empirical result

implies that asking for extra time during a credit period does not affect the amount

of trade credit received negatively.

The empirical findings in table 5.4 reveal the importance of the impact of

customer bargaining power on trade credit supply. As discussed above, we proxy

for this by using two variables: IMCUSij and SUPOWij. In addition, we find a

significant and positive relation between the variable IMCUSij and the proportion

of payments made after delivery. This evidence indicates that suppliers prefer to

provide more credit to buyers/clients who order relatively large quantities. In

addition, we observe a negative and significant impact of SUPOWij on the

proportion of the delayed payments made by a customer. The estimation

coefficient shows that if a customer cannot participate in the negotiating process

for better selling prices, the proportion of delayed payments will decrease with 40

per cent (β=-0.399, t=-6.46). This finding confirms that a customer who has little

bargaining power is less likely to obtain credit. In general, the above empirical

findings provide strong evidence for our second hypothesis that rice firms provide

more trade credit to large buyers and/or buyers with high bargaining power. By

doing so, rice firms in the Mekong Delta expect to build up a trading relationship

with large clients, attract and lock in large buyers to support present and future

sales. These empirical findings are consistent with the view of trade credit as a

marketing device to promote sales for rice firms in the sample.

5.4.2 The Effects of Market Characteristics

The estimation results in table 5.4 show that suppliers who perceive a higher

degree of competition, PERCOMPi, are more willing to provide credit. This

supports the view that firms use credit as a marketing instrument to promote sales.

In addition, we observe that rice traders at different market segments

behave differently with respect to trade credit. The amount of trade credit changes

enormously through different market segments. We find significant and positive


107

effects of W-MILLERSi, WHSALEi and RETAILi on the proportion of delayed

payments. The regressions show that on average w-millers are likely to grant 54

per cent more credit than millers do (β=0.544, t=6.75). At the same time,

wholesalers provide about 55 per cent more credit than millers (β=0.554, t=6.90)

while retailers provide about 39 per cent more credit than millers (β=0.394,

t=3.90).

5.4.3 The Effects of Supplier Characteristics

Finally, we examine some relevant suppliers’ characteristics that are likely to be

associated with trade credit. We find that large rice firms appear to grant more

credit to clients than small firms. The reason may be that large rice firms (large

capital) have better access to bank loans since they can use their factories or

warehouses as collateral.

In line with theoretical predictions we find that a lack of capital,

LACKCAPi, has a negative and significant effect on trade credit provision. In our

sample, rice firms that lack capital provide 8 per cent less trade credit to their

customers (β=-0.087, t=-3.36). In addition, we investigate the relation between

suppliers’ profit margins and trade credit provision. Profitable firms may have

more available cash flows and thus are able to provide more trade credit to clients

(Petersen and Rajan, 1999). However, this is not confirmed in our study.

In an attempt to measure the default risk incurred by a supplier, we include

UNPAIDi: the number of times that the supplier was not paid by clients in the last

three years. Contrary to our prediction, we observe a significant and positive

association between UNPAIDi and the proportion of delayed payments. This

evidence shows that rice firms that experienced defaults by customers in the past

continue to provide trade credit. Yet, we observe that the risk of not being paid is

rather small. The average number of times that firms were not paid in the last three

years is 0.78. In case a customer is unable to pay his debts in time, suppliers can

offer credit to help these customers survive so that the customer can pay back later.


108

At the same time, suppliers can generate future returns from current trade relations.

This also explains why customers who do not pay in time and ask for an extension

of the credit period do not receive less trade credit, since we note no significant

impact of PAYLATEij on trade credit supply.

11

0

Tab

le 5

.4:

The

impac

t of

cust

om

er c

har

acte

rist

ics

on t

rade

cred

it p

rovis

ion:

C

oef

fici

ents

(1)

Coef

fici

ents

(2)

Coef

fici

ents

(3)

Coef

fici

ents

(4)

Coef

fici

ents

(5)

Coef

fici

ents

(6)

Coef

fici

ents

(7)

SU

PO

Wij

-0

.451

(-7.9

7)*

**

-0

.399

(-6.4

6)*

**

IMC

US

ij

0.0

51

(4.0

7)*

**

0.0

51

(2.8

3)*

**

LO

GL

EN

GR

Eij

0.0

44

(3.0

0)*

**

0.0

34

(2.4

3)*

*

VIS

ITij

`

0.0

84

(4.7

2)*

**

0.0

71

(4.0

3)*

**

CO

NT

AC

Tij

0.0

51

(3.1

0)*

**

0.0

33

(1.8

9)*

*

PA

YL

AT

Eij

-0

.076

(-0.4

8)

-0.1

16

(-0.8

8)

FS

IZE

i 0.0

63

(3.6

5)*

**

0.0

95

(5.4

8)*

**

0.0

84

(4.8

4)*

**

0.0

83

(4.6

7)*

**

0.0

84

(4.8

3)*

**

0.0

86

(4.8

9)*

**

0.0

73

(4.1

9)*

**

BE

UN

PA

IDi

0.0

88

(6.1

2)*

**

0.0

91

(6.3

5)*

**

0.0

92

(6.1

4)*

**

0.1

02

(7.3

7)*

**

0.0

86

(5.8

3)*

**

0.0

90

(6.0

7)*

**

0.0

97

(7.0

1)*

**

LO

GA

GE

i -0

.026

(-0.9

5)

-0.0

42

(-1.4

8)

-0.0

67

(-2.3

2)*

*

-0.0

53

(-1.8

4)*

-0.0

55

(-1.9

7)*

*

-0.0

50

(-1.7

6)*

-0.0

44

(-1.5

0)

LA

CK

CA

Pi

-0.0

82

(-3.0

3)*

**

-0.0

94

(-3.3

8)*

**

-0.0

84

(-2.9

7)*

**

-0.0

85

(-3.1

5)*

**

-0.0

92

(-3.3

5)*

**

-0.0

90

(-3.1

7)*

**

-0.0

87

(-3.3

6)*

**

MA

RG

INi

0.2

15

(1.0

0)

0.2

37

(1.0

6)

0.2

42

(1.1

0)

0.3

60

(1.6

0)

0.2

04

(0.9

3)

0.2

48

(1.1

1)

0.2

55

(1.1

6)

PE

RC

OM

Pi

0.1

24

0.1

20

0.1

21

0.1

02

0.1

34

0.1

20

0.1

21

Ch

ap

ter

5:

Tra

de

Cre

dit

Su

pply

an

d C

ust

om

er C

ha

ract

eris

tics

1

11

(4.7

7)*

**

(4.5

1)*

**

(4.5

6)*

**

(3.9

0)*

**

(5.2

1)*

**

(4.5

1)*

**

(4.7

3)*

**

WH

SA

LE

i 0.6

83

(9.0

7)*

**

0.6

11

(7.8

6)*

**

0.6

09

(7.8

0)*

**

0.5

47

(6.8

9)*

**

0.6

30

(7.9

9)*

**

0.6

41

(8.2

6)*

**

0.5

54

(6.9

0)*

**

W-M

ILL

ER

i 0.7

16

(9.9

0)*

**

0.5

75

(7.4

7)*

**

0.6

24

(8.4

0)*

**

0.5

50

(7.1

1)*

**

0.6

39

(8.6

0)*

**

0.6

33

(8.5

0)*

**

0.5

44

(6.7

5)*

**

RE

TA

ILi

0.4

71

(4.9

4)*

**

0.3

33

(3.4

1)*

**

0.3

08

(3.0

8)*

**

0.2

38

(2.4

0)*

*

0.2

99

(2.9

8)*

**

0.3

26

(3.2

8)*

**

0.3

94

(3.9

0)*

**

Const

-1

.581

(-4.8

3)*

**

-2.1

7

(-6.5

2)*

**

-2.0

92964

(-6.3

2)*

**

-1.9

19

(-5.6

3)*

**

-2.1

69

(-6.5

1)*

**

-2.0

01

(-6.0

2)*

**

-1.9

63

(-5.8

1)*

**

Log l

ikel

ihood

-830.5

82

-851.1

39

-879.4

40

-829.8

45

-879.0

31

-891.0

03

-756.8

35

Pse

udoR

2

0.2

27

0.1

97

0.1

88

0.2

01

0.1

92

0.1

833

0.2

48

Obse

rvat

ions

1149

1131

1160

1106

1168

1170

1067


112

5.5 Conclusion

In this chapter, we use transaction level data from the survey rice market of the

Mekong Delta and continue to identify the customer characteristics that make

access to supplier credit source possible. We cover a broader range of variables

than previous studies, since these studies either concentrate on customer

creditworthiness or customer bargaining power, whereas we jointly use these

variables in our analysis. In line with our expectations, the empirical findings

indicate that suppliers do indeed have incentives to provide more credit to larger

clients and/or clients with strong bargaining power. At the same time, they are

willing to grant credit to creditworthy clients. This supports the view of the

marketing theory on trade credit, i.e. suppliers seem to provide trade credit to build

trading relations with large buyers to generate present and future sales.

Chapter 6

Trade Credit Supply in Different

Market Segments

6.1 Introduction

One of the starting points of this research project was to concentrate on one

industry in one country to eliminate the various effects on trade credit supply

caused by the differences in the development of financial institutions, and the

banking and legal systems across different countries; for example, Lu (2009)

shows that the legal system has a positive effect on the provision of trade credit.

This approach also allows us to eliminate the effect of the differences in product

characteristics and/or industry characteristics on trade credit supply; for example,

Giannetti et al (2009) show that firms producing standardized products grant less

credit than service firms. Therefore, we focus on market structure (firm

competitiveness) in one market as the major market characteristic affecting trade

credit supply.

Yet, when focusing on one industry with a homogeneous product, we still

observe that trade credit supply differs significantly across the four main market

segments: w- millers, wholesalers, retailers and millers. The coefficients of the

segment dummy variables in the previous chapters show that w-millers and

wholesalers grant much more trade credit to trading partners than millers and

retailers do. This raises an interesting question: what factors cause the large


114

differences in trade credit supply across market segments? Do traders who are

active in different market segments behave differently with respect to trade credit

provision?

This chapter tries to explain why trade credit supply varies across market

segments. We examine the determinants of trade credit supply for each market

segment.

The chapter is organized as follows: section 6.2 continues with the

empirical methodology. Section 6.3 describes the structural differences across

market segments that can affect trade credit provision in each segment. Section 6.4

introduces the empirical results on determinants of trade credit supply. Section 6.5

concludes the chapter.

6.2 Empirical Methodology

In this chapter, we employ the transaction-level dataset used in chapter 5, and

investigate determinants of trade credit supply for each segment of the rice market.

Also, we look at the differences in determinants of trade credit supply across

market segments and aim at gaining insights into the differences in trade credit

behavior across different market segments. We estimate equation (6.1) separately

for the four groups of firms at the four market segments: w-millers, millers,

wholesalers and retailers.

DPCUS ij = α + βCustomercharacteristicij + δ Suppliercharacteristici + ε ij

(6.1)

Chapter 6: Trade Credit in Different Market Segments

115

Empirical Measures16

Our dependent variable, DPCUSij,, is the proportion of the payment made after the

delivery of goods by customer j to supplier i. Since it always falls between 0 and 1,

we treat this variable as a censored variable and employ a standard Tobit model.

Customer Characteristics and Trading Relations

Suppliers are expected to provide more credit to a customer whom they know to

ensure repayment. Therefore, the following independent variables should reflect

customer creditworthiness and trade relation characteristics, which may affect a

supplier’s incentives to provide credit to a specific customer.

(1) PAYLATEij: the number of times customer j has not paid supplier i on time

in the past.

(2) LOGLENGREij: the natural log of the length of the trading relation in days

between supplier i and customer j.

(3) VISITij: the number of times supplier i has visited the shop of customer j in

the year before granting credit to customer j.

(4) CONTACTij: we also ask each supplier how often customer j usually

contacts supplier i, either by calling or visiting the supplier, to ask about selling

prices. Frequent contact with his/her supplier may provide more information

about a customer’s business and can result in more trade credit being provided

to the customer (Fisman and Raturi, 2004). In this study, we employ

CONTACTij as a scale variable ranging from 1 to 6. This variable measures the

frequency of face to face and/or phone contact between a supplier and a

specific customer:, which ranges from ( 1=yearly; 2=every six months;

3=every three months; 4=monthly; 5=weekly; and 6=daily).

16

The thesis is a collection of papers. There is some overlap. Those who have read the previous

chapters can skip this section since this matter has also been dealt with in section 5.3.


116

(5) According to the marketing theory on trade credit (see Van Horen, 2007;

Fabbri and Klapper, 2008; Hyndman and Serio, 2010), suppliers operating in a

more competitive market are highly motivated to grant credit to support sales. In

addition, the decision to grant credit to an individual customer may be influenced

by the customer order sizes and bargaining power. Several studies suggest that

suppliers provide trade credit to encourage customers to order in large quantities

and/or to establish trade relations with large clients. An increase in the proportion

of sales on delayed payments will result in larger optimal order sizes (Huang,

2007; Rachamadugu, 1989). Moreover, large customers may gain more bargaining

power over suppliers and may therefore demand to purchase on credit to increase

their surplus. Accordingly, suppliers may grant more credit to large clients and/or

clients with strong bargaining power. Therefore, the variables, PERCOMPi,

IMCUSij, SUPOWij, picking up the size and bargaining power of clients, are main

proxies for the marketing theory in our study.

(6) IMCUSij is a measure of the size of customer j’s average order. It is

measured by the ratio of average orders from customer j to the supplier i’s average

monthly revenue.

(7) PERCOMPi: this five-scale variable reflects the importance of competition

perceived by a supplier. This value ranges from 0=no competition at all,

1=competition is not important, 2=competition is relatively important,

3=competition is important, to 4=competition is an extremely important issue.

(8) SUPOWij measures supplier i’s bargaining power vis-à-vis customer j. It is

a dummy variable that equals 1 if supplier i can set the selling prices and contract

terms without any negotiations with customer j, and 0 otherwise (SUPOWij).

Supplier Characteristics

The choice of providing credit to customers is also influenced by various supplier

characteristics. The second set of variables measures some general supplier


117

characteristics that influence the willingness to grant credit. First, firms with good

access to external financial markets are more willing to provide credit to their

trading partners (Schwartz, 1974; Smith, 1987). In contrast, firms that face credit

constraints may not be able to provide trade credit. In our empirical analysis we

pick up the effects of being credit-constrained with LACKCAPi. Firm size and firm

age may be important determinants of trade credit provision in several studies. The

reasons may be that large and old firms have better access to external financial

markets and therefore they may be more capable of offering credit to their trading

partners (Van Horen, 2005; Fisman and Raturi, 2004). Second, some studies reveal

the impact of suppliers’ profitability on suppliers’ incentives to provide credit to

customers (Petersen and Rajan, 1999). We control for the supplier characteristics

mentioned above by the following variables.

(9) FSIZEi: natural logarithm of capital, used as a proxy for firm size; several

empirical studies on this topic reveal the impact of firm size on trade credit

provision (Van Horen, 2005; Fisman and Raturi, 2004).

(10) LOGAGEi: natural log of firm age, measured in years from the moment a

firm was established until 2007.

(11) LACKCAPi: a five-scale variable measuring the importance of lacking

capital estimated by firms’ owners. This value ranges from 0=no lack of capital at

all, 1=lack of capital is not a problem, 2=lack of capital is a little problematic,

3=lack of capital is an important problem, to 4=lack of capital is an extremely

important problem.

(12) BEUNPAIDi: being unpaid by clients in the past could negatively affect

sellers’ incentives to provide credit to customers. Therefore, we include

BEUNPAIDi measured by the number of times that a firm has been unpaid by all of

its customers in the last three years.

(13) MARGINi: we measure a firm’s profitability by estimating the price cost

margins.


118

MARGINi = i

ii

AP

AVCAP −

where

APi = the average selling price per ton of rice of firm i in 2006, and AVCi = the

average variable cost of firm i in 2006. Average variable cost is defined as the total

costs including the purchasing rice/paddy cost as main material, fuel cost,

transportation cost and labor cost.

6.3 Why May Trade Credit Differ Across Segments?

6.3.1 Market Context Characteristics and Trade Credit Supply

According to existing studies such as Summer and Wilson (2003), Giannetti et al.

(2009) and Hyndman and Serio (2010), suppliers differ in their willingness to

extend trade credit for several reasons such as product characteristics

(homogeneous versus differentiated products), business activity (service firms and

non-service firms); market structures (competition, buyers’ market power); intra-

industry trade (suppliers and customers operating in the same section may provide

more information about clients).

Since the characteristics mentioned above vary across different industries

and firms in different industries rely on different combinations of those industry

characteristics, trade credit supply behavior may vary substantially across different

industries. In other words, the context of industry and/or market may induce firms

in different industries/markets behave differently with respect to trade credit

supply. For example, Giannetti et al (2009) finds that service firms and firms

producing differentiated products are willing to grant more credit while firm

producing standardized products offer less credit. It also shows that industry

market structure concerning seller market power vis-à-vis buyer bargaining power

influences trade credit supply significantly.


119

6.3.2 Structural Differences Across Market Segments

In our study, we observed that structural differences exist even across different

market segments of the rice industry. These differences concern market structure,

customer characteristics and access to external finance, which are supposed to

influence trade credit supply strongly (Wilson and Summer, 2002; Giannetti et al,

2009; and Hyndman and Serio, 2010). The ANOVA results (table 6.1) show that

the mean differences of the variables concerning the three aspects mentioned above

are substantial and significant across the four market segments (w-millers,

wholesalers, retailers and millers). For example, 67 per cent of w-millers used bank

loans in 2006, and the average amount of bank loans was 859 million VND while

only 9 per cent of retailers had bank loans, with an average loan of 2.33 million.

Also, in previous chapters we observe that trade credit supply differs

substantially throughout the four market segments. For example, while 61 per cent

of transactions in the w-millers segment involve trade credit activities, only 20 per

cent of transactions in the miller segment do. The reason may be that the different

contexts of different market segments may induce traders in different segments

behave differently with respect to trade credit supply. This provides us incentives

to extend the analysis further in this chapter on the differences in determinants of

trade credit supply across market. In the remainder of this section, we discuss the

differences of a few variables reflecting market structure, customer characteristics

and access to external finance between the four segments.


120

Table 6.1: Mean Differences across different Market Segments

Variables W-miller Wholesaler Retailer Miller Sig

Trade credit supply

DPCUSij 0.61 0.48 0.26 0.20 ***

ADPMENTSi17

0.156 0 0 0 ***

Access to external finance

BANKLOANSi18

(million VND)

840 351 127 2.334 ***

LACKCAPi 3.20 3.10 3.20 2.80 **

Customer characteristics

TRANSACTION VALUEij

(million VND)

777 95.8 0.801 1.534

***

PAYLATEij 0.006 0.06 0.028 0.011 **

Supplier characteristics

FSIZEi 21.289 19.346 16.672 19.609 ***

MARGINi 0.151 0.117 0.109 0.141 **

Market structure

IMCUSij 1.517 0.482 0.072 .119 ***

SUPOWij 0.147 0.213 0.617 0.066 ***

PERCOMi 3.32 2.90 2.83 2.86 ***

NORIVALSi 5.20 3.72 3.82 3.44 ***

The one-way ANOVA test for the mean difference of firms across the four different

markets segments: w-millers, millers, wholesalers and retailers . *, ** and *** denote

significant level at 10, 5 and 1 per cent, respectively.

17 The proportion of advanced payments from customers to total sales. It usually happens when rice

exporters sign a contract and pay their suppliers in advance. 18 The total amount of bank loans on December 31, 2006.


121

Market Structure Across Market Segments

Market structure appears to vary significantly across the four market segments. In

particular, the degree of competition among existing rivals and customer

bargaining power appears to be less intense throughout the segments: w-miller,

wholesaler, miller and retailer.

Wholesale-Millers

Competition pressure appears to be strongest in the w-miller segment. First, w-

millers consider competition to be a more important difficulty in directing their

business (PERCOM: 3.20). Second, w-millers seem to have the highest number of

rivals compared with other private traders (5.20). Third, w-millers seem to have the

lowest bargaining power vis-à-vis customers compared with other private traders.

Indeed, the figures for IMCUSij show that the ratio of an average order from the

interviewed clients to the average monthly sales is 1.5 in the w-miller segment.

This figure indicates that w-millers’ clients usually order in bulk since the average

order is equal to 1.5 times the monthly revenues. The reason may be that 30 per

cent of the w-millers’ clients are exporters: very large firms that purchase in large

quantities. In fact, it usually takes w-millers several months to process the required

amount of rice for a contract with a rice exporter. Therefore, customers of w-

millers appear to have a strong bargaining power when it comes to negotiating

contract terms and prices. This is confirmed by the variable SUPOWij (table 6.1), in

that w-millers use negotiating procedures to set contract terms and selling prices

for 85 per cent of their sales. In other words, w-millers’ clients have such a strong

bargaining power that w-millers are able to set selling price without negotiating for

only 15 per cent of their sales. Fourth, w-millers may find it easier to collect

information on clients as well as to enforce repayments. The reason may be that 99

per cent of the clients in the w-miller segment are wholesalers (54 per cent), w-

millers (15 per cent), and rice exporters (30 per cent). These clients are registered


122

firms with some shops, factories that can be used as collateral to enforce

repayment.

Wholesalers

In the wholesale segment, competition pressure among existing rivals also appears

to be high. For example, wholesalers seem to pay more attention to the effect of

fierce competition on their profitability than millers and retailers (PERCOMP:

2.90). The average number of competitors in this segment is 3.7, which is higher

than for retailers. At the same time, a firm’s bargaining power vis-à-vis customers

is relatively weak in this segment. A large proportion of the wholesalers’ clients

are wholesalers, i.e. relatively large firms (67 per cent). Also, the IMCUSij variable

suggests that an average order is about 50 per cent of the average monthly sales of

wholesalers. This indicates that wholesalers’ clients often order in bulk. The

SUPOWij variable also confirms that wholesalers possess relatively little

bargaining power and have to negotiate to set the price and contract terms for about

78.7 per cent of their revenues. In short, these figures show that wholesalers have

relatively little bargaining power and face a lot of pressure through competition.

Millers

Facing a lot of competition appears to be a significant problem in the miller

segment. Millers seem to have little bargaining power vis-à-vis customers since a

large proportion of clients are w-millers (17 per cent), wholesalers (47 per cent)

and exporters (2 per cent), which are large and formal firms. This is confirmed by

the variable SUPOWij, which indicates that millers negotiate to set prices for about

93.4 per cent of their revenues.

Unlike other private rice traders, millers only provide milling services and

charge a milling fee, which is a small fraction of the total value of rice. In fact, the

average milling fee is about 50 VND/kg, which is about 1 per cent of the average

price of rice. Thus, the average value of the transaction is relatively low (see table

6.1). The low value of transactions may influence the demand for trade credit


123

negatively, especially since many of the millers’ clients are large firms. The reason

is that clients may not need trade credit to finance the low value transactions.

Retailers

Retailers appear to face less intense pressure through competition than traders in

other market segments. Moreover, retailers seem to have the strongest bargaining

power vis-à-vis customers as compared to other private traders. In marketplaces

retailers sell to a large number of final consumers who purchase only small

quantities. On average, IMCUS indicates that the proportion of sales accounted for

by a customer is only 7 per cent of the average monthly sales. SUPOWij also

suggests that retailers set their own selling prices without any negotiation for about

62 per cent of their total sales. This figure is the highest in this market segment,

which confirms the strongest bargaining power vis-à-vis customers of retailers.

Access to External Finance

Table 6.1 shows that access to external financing varies significantly for rice

traders at different market segments. For example, among private traders, w-

millers have the best access to bank loans. On average, 67 per cent of w-millers

used bank loans in 2006, and the average amount of bank loans was 859 million

VND. At the same time, only 9 per cent of retailers had a bank loan, with the

average loan amounting to 2.33 million VND. This difference may be explained by

the fact that w-millers are usually large and formal firms that own valuable assets

which they can use as collateral. In contrast, retailers are often small and

unregistered firms; it is therefore difficult for them to apply for a bank loan.

In addition, w-millers are the only private traders with access to advance

payments from state-owned enterprises involved in large-scale rice export. Table

6.1 shows that w-millers sometimes receive advance payments from rice exporters;

the average amount of these payments is 15.6 per cent of total firm sales. In fact,

rice exporters usually sign very large exporting contracts with foreign importers. In


124

order to guarantee the timely supply of sufficient rice and the required quality, rice

exporters often pay w-millers in advance to collect rice for them. It can also

happen that during the harvesting season, the government provides free credit to

state-owned enterprises to purchase and store rice to maintain the price in the

regional market above the floor price.

Customer Characteristics

Customer characteristics appear to differ fundamentally across market segments.

For example, in the w-miller segment, trade relations are mostly among large firms

since 99 per cent of the w-millers’ clients are exporters (30 per cent), w-millers (15

per cent) and wholesalers (54 per cent). These clients are registered firms and have

very large shops and/or factories. Also, in this segment clients often order in bulk.

For example, the average value of transactions is about 777 million VND.

Consequently, w-millers may find it easier and less costly to gather information

about this type of client and to enforce repayment.

Table 6.2: Types of customers

Supplie

rs

Customers

W-

miller

Export

er

Wholesal

er

Retailer Consu

mer

Registered

firm

Non-

registered

firms or

consumer

W-

miller

(%)

47

15

94

30

170

54

4

1

0

0

301

99

4

1

Miller

(%) 64

17

6

2

173

47

14

4

110

30

243

66

14

34 Wholes

aler

(%)

0

0

0

0

134

67

64

32

2

1

134

67

66

33

Retailer

(%)

0

0

0

0

0

0

38

13

252

87

0

0

290

100

Source: own survey in 2007


125

In the wholesaler segment, about 67 per cent of customers are other

wholesalers: registered firms with relative large shops. Yet, 32 per cent of the

clients are retailers: informal firms with very small shops and no collateral of any

real value. As a consequence, it is relatively more difficult for wholesalers to

collect information about clients and enforce repayments than w-millers.

In the miller segment, clients can be wholesalers (47 per cent), w-millers

(17 per cent) and exporters (2 per cent), which means that they are also relatively

large firms. However, a large proportion of clients is a farmer from the neighboring

area (34 per cent) – it may not be that costly to collect information about these

clients as most of them live nearby. Yet, it may not be easy for millers to enforce

repayment since farmers may not have valuable assets they can use as collateral.

In the retailer segment, trade relations are mainly between retailers and

final consumers. Again, to collect information and enforce repayments is probably

most complicated and costly for retailers since 87 per cent of retailers’ customers

are individual consumers who purchase in small quantities in marketplaces (see

table 6.2).

In addition, the size of customer orders varies significantly across the

different market segments. Table 6.1 shows that the average value of transactions

is 777 million VND in the w-miller segment and 95.8 million VND in the

wholesale segment, 0.8 million VND in the retail segment and 1.5 million VND in

the miller segment. The low value of transactions in the retail and miller segments

can influence trade credit supply for several reasons. First, the cost of searching for

information about customer creditworthiness and contract enforcement is high

compared to the value of the transaction. Second, the low value of transactions in

the retail and miller segment can also influence trade credit demand, since buyers

do not need trade credit to finance their purchases.

In short, we observe that there are some structural differences across the

four main market segments. The differences refer to access to market structure,

external finance and customer characteristics, which may influence trade credit


126

supply significantly. This provides us with the incentive to investigate whether or

not traders in different segments behave differently with respect to trade credit

supply.

6.4 Estimation Results

6.4.1 Differences Between Market Segments: General Overview

The set of independent variables that proxy for the marketing view of trade credit

(PERCOMPi, SUPOWij, IMCUSij) shows more significant results for the w-miller

(table 6.3, column 1) and the wholesaler segments (column 2) than the market

segments of retailers (column 3) and millers (column 4).

In addition, LACKCAPi, measuring access to external finance, shows

robust and significant results for the segments of w-millers, wholesalers, and

millers. Nevertheless, a lack of capital seems to be less important to trade credit

provision for w-millers than for wholesalers and millers. The magnitude of the

coefficients is smaller for w-millers (β=-0.051, t=-1.77) than for wholesalers (β=-

.113, t=-2.35) and millers (β=-0.232, t=-3.16). In fact, w-millers have the best

access to external finance compared to other private rice traders. Furthermore,

LACKCAPi has no significant impact on trade credit supply in the retailer segment.

The reason may be that retailers are very small shops and have limited access to

bank loans: only 9 per cent of the retailers had a bank loan in 2006. In fact, a lack

of capital may be a common problem to most retailers. As a result, this may not be

one of the main factors that determines trade credit supply in this segment.

The set of variables measuring the information on customer

creditworthiness also shows remarkable differences across different market

segments. For example, CONTACTij only shows significant results in the w-miller

and retailer segments. Yet, VISITij is only relevant in the wholesaler and miller


127

segments. In fact, w-millers sell to very large and formal firms. In this case,

frequent contact is expected to provide one with

sufficient information to make the credit-granting decision. Unlike the case of w-

millers, 33 per cent of wholesalers’ clients are retailers in marketplaces (very small

and informal firms), and 34 per cent of millers’ customers are farmers located in

the neighborhood. Apparently, visiting shops (houses) leads to more information

than contact by telephone. This explains why visits appear to influence trade credit

supply significantly and positively in the wholesaler and miller segments. The

retail segment in particular is different from the other segments. In the other market

segments, clients are rice firms that may have frequent contact to exchange

information about prices, exporting quotas, clients, etcetera. In the retail segment,

individual consumers contact only to purchase. Therefore, when there is more

contact, purchases will be more frequent and the quantity per order will be smaller.

Accordingly, consumers who contact frequently do not need trade credit. This may

explain why frequent contact is negatively associated with trade credit supply in

the retailer segment.

The above outcome indicates that there are major differences in trade

credit determinants across market segments. In order to understand the differences,

the next section will look closer at each segment.

6.4.2 A Close Look at Each Segment

Wholesale-Miller Segment

The most outstanding feature of trade credit supply determinants for the w-miller

segment is the significant results of the set of variables that represents the

marketing theory on trade credit (PERCOMPi, SUPOWij, IMCUSij). First, we

observe a significant and positive effect of PERCOMPi on trade credit supply. W-

millers perceive a higher degree of market competition causing them to grant more

credit to their trading partners. Second, w-millers have incentives to grant more


128

credit to large clients and clients with strong bargaining power. The coefficient of

SUPOWij in table 6.3 shows that customers receive about 30 per cent less credit if

they do not have bargaining power and do not negotiate prices and contract terms.

At the same time, a significant and positive association between trade credit supply

and IMCUSij indicates that w-millers provide more credit to large clients. The

explanation may be that large clients and clients with high bargaining power are

usually the clients with a high potential to generate sales. Therefore, w-millers may

grant more credit in order to develop a trading relation with them. These empirical

findings suggest that the marketing theory, which explains trade credit as a mean to

encourage sales, is among the motives that determine trade credit supply in the w-

miller segment.

In fact, it is worth recalling that the transaction value in the w-millers

segment is very large. As a result, trade credit may be an attractive option for

clients to finance their purchases. Accordingly, customers may demand much trade

credit, making trade credit an efficient means for establishing trade relations with

large clients.

As mentioned in the previous section, w-millers may have less difficulty in

gathering information about clients and enforcing repayment, since many of their

clients are relatively large firms that own valuable collateral such as factories and

shops. Yet, w-millers appear to face a relatively lower risk of default. In our

sample, only 0.6 per cent of the w-millers’ transactions were paid late during the

last three years. It may explain why the set of variables measuring information

about customer creditworthiness shows less relevant results with respect to w-

millers. In particular, we do not find a significant effect of the variables VISITij,

LOGLENGREij,, PAYLATEij on trade credit supply.

Wholesaler Segment

The set of variables representing the marketing theory on trade credit shows

significant results in the wholesale segment (PERCOMPi, SUPOWij, IMCUSij,).


129

The positive and significant impact of PERCOMPi indicates that wholesalers grant

more credit if they perceive considerable market competition. Furthermore,

wholesalers seem to offer credit to large customers (IMCUSij,: β=0.015, t=1.92)

and customers with high bargaining power (SUPOWij: β=-0.251, t=-2.38). This is

consistent with the marketing theory, which suggests that firms give credit to

attract and lock in high potential clients to generate sales now and in the future.

Unlike the w-miller segment, wholesalers may face a higher risk of default

since their clients include relatively many small and unregistered firms (33 per

cent). Consequently, wholesalers have more incentives to invest in assessing a

client’s creditworthiness than w-millers. First, it is shown in table 6.3 that if

wholesalers visit a customer’s shop before granting credit, they will provide more

trade credit. The location and the shop of retailers/wholesalers may reveal

information about the success of a rice retailer. Second, the significant and positive

coefficient of LENGREij shows that wholesalers are willing to grant more credit to

customers who have been in a trade relation for a longer period. Such a relation

may reveal a lot of information concerning customers and may lead to a closer

relation that facilitates trade credit. Third, wholesalers also seem to provide less

credit to clients who have not been able to pay them on time in the last three years.

The variable PAYLATEij only shows significant results for the segment of

wholesalers. In this segment, customers who asked for an extension of the credit

period in the last three years receive about 27 per cent less credit (β=-0.268, t=-

2.02). The reason may be that the risk of default may be more real for wholesalers:

wholesalers’ customers include 33 per cent of all retailers, with very small shops or

seats at marketplaces and who own no valuable assets to use as collateral. Also,

retailers are usually non-registered. As a consequence, wholesalers may have

difficulty enforcing repayments with such clients. This may explain why

wholesalers are more careful in assessing customer creditworthiness than w-

millers.


130

Miller Segment

We observe that the set of variables that represent the marketing theory is less

relevant in the miller segment than for w-millers and wholesalers. In particular, the

significant and positive coefficient for PERCOMPi indicates that pressure through

high competition is relevant to millers supplying credit. Yet, we find no significant

effect of IMCUSij and SUPOWij on trade credit provision. This suggests that millers

do not have strong incentives to grant credit to large customers or customers with

high bargaining power, which is different from the case of w-millers and

wholesalers.

A special aspect of the miller segment is that the milling fee is only about 1

per cent of the total value of rice and the average value of transactions is very small

(see table 6.1). This suggests that it is relatively easy for large clients like w-millers

and wholesalers to pay bills immediately. Consequently, large clients may not need

trade credit to finance their purchases. Accordingly, millers do not supply trade

credit to large clients and/or clients with a lot of bargaining power. In fact, large

clients in the miller segment may care more about product quality than about the

possibility of trade credit. This explains why millers provide relatively little trade

credit, which they do not seem to grant to large clients. In contrast, millers grant

credit to clients that have been good trading partners. We observe that in this

segment VISITij is among the relevant determinants of trade credit supply.

Retailer Segment

The results in table 6.3 show that the set of variables representing the marketing

theory also shows less significant results for the retail segment. In column 4 we

note a negative and significant association for SUPOWij. Yet, PERCOMPi and

IMCUSij do not have a significant effect on trade credit supply. Retailers do not

seem to grant more credit when they perceive market competition to be fierce;


131

retailers do not to offer credit to large clients either; they only seem to grant credit

when their clients have bargaining power and demand credit.

Retailers sell to a large number of final consumers who purchase small

quantities at the marketplace. The value of transactions is low in this segment and

therefore clients may not need trade credit to finance their purchases. Only when

retailers have large clients like restaurants and these large clients demand trade

credit, retailers will grant credit to them. This may explain the negative and

significant impact of SUPOWij. on trade credit supply. Also, it is difficult and

costly for retailers to gather information about their customers’ creditworthiness.

This may be one of the problems that prevents retailers from using trade credit

frequently, even when pressure through competition is high. In fact, it is costly to

collect information on clients and to enforce repayment – especially compared to

the low value of the transactions.

6.5 Summary of the Results

In this chapter, we aimed at gaining a better insight into the causes of the

differences in trade credit supply across the various market segments. We find that

structural differences in market competition and customer characteristics are

among the factors that lead to these large differences. In particular, in the market

segments where there are inter-firm trade relations (clients are rice firms) and the

average value of transactions is large, trade credit supply is higher. In segments

where there is trade between firms and individual traders and consumers and the

transaction value is small, trade credit supply becomes less popular.

In the w-miller segment, trade relations are between relatively large firms

and the value of transactions is high, making trade credit an attractive option for

clients. W-millers also seem to have little difficulty in assessing customer

creditworthiness and enforcing repayments. Also, w-millers have good access to

bank loans. This explains why trade credit is used most extensively in this market


132

segment. Concerning the determinants of trade credit supply, the set of variables

representing the marketing theory is among the most relevant determinants of trade

credit supply. The results of these variables suggest that suppliers grant credit to

attract high potential clients and support sales.

In the wholesaler segment, trade relations are inter-firms relations, though

a relatively a lot of clients are retailers, with small and informal firms. Therefore,

wholesalers may find it more difficult to collect information on clients and enforce

repayment. Also, the average transaction value of this segment, which may

influence the demand of trade credit, is lower than that of w-millers. Trade credit

supply is less popular in this segment than in the w-miller segment. With respect to

trade credit determinants, the set of variables that represent the marketing theory

comprises the relevant factors driving trade credit supply in the wholesaler

segment. Suppliers tend to offer more credit to large clients with high bargaining

power. They also provide more credit when they perceive market competition to be

higher. Yet, due to difficulties in enforcing repayment, wholesalers appear to invest

more in assessing customer creditworthiness than w-millers.

Millers are special because they are service firms. Millers charge a milling

fee, which is about 1 per cent of the value of rice. As a result, the value of

transactions is low in this segment, making trade credit a less attractive option.

This may explain why the marketing theory shows less relevant determinants of

trade credit supply in the miller segment than in the segments of w-millers and

wholesalers. In particular, millers do to grant no credit to large clients and/or

clients with strong bargaining power. The reason may be that large clients do not

need trade credit to finance these small transactions. This is in contrast with the

case of the w-miller segment, in which high value transactions make trade credit

attractive to clients. Therefore, trade credit may function less efficiently in building

trade relations in this market segment. This may explain why trade credit supply is

relatively limited in the miller segment. Besides, many of the millers’ clients are

consumers and/or farmers in the neighborhood, and therefore millers may have


133

difficulty enforcing repayment by such clients. Millers appear to grant credit to

customers with whom they have close trade relations.

In the retailer segment, most trade relations are with final consumers, who

purchase small quantities at local markets. First, final consumers do not need trade

credit to finance these small purchases and therefore the demand for trade credit is

low in this segment. Second, it may be difficult for retailers to gather information

about clients and enforce repayments, e.g. the cost of collecting information and

enforcing repayment may be too high compared to the value of transactions. This

may explain why trade credit is used less frequently in this market segment as

compared to the wholesalers and retailers. The variables representing marketing

theory also show fewer relevant results in this segment. Only when retailers have

large clients like restaurants that demand trade credit, they may provide credit.

In short, this chapter suggests that the market segment specificity such as

the type of clients (firms or individual consumers), the value of transactions, the

nature of the business activity (service firms or non-service firms) are important

factors that influence the determinants and/or the role of trade credit supply.

Table 6.3: Determinants of trade credit supply in the four market segments:

w-millers, wholesalers, millers and retailers

Independent variables

Coefficient

(1)

W-miller

(2)

Wholesalers

(3)

Retailers

(4)

Millers

Marketing theory of trade

credit

SUPOWij

-0.297

(-4.07)***

-0.251

(-2.38)**

-0.539

(-4.41)***

-0.165

(-0.48)

IMCUSij 0.026

(2.26)**

0.015

(1.92)*

0.305

(1.34)

0.142

(0.64)

PERCOMPi 0.091

(2.78)***

0.188

(3.37)***

-0.009

(-0.20)

0.384

(4.70)***

Customer creditworthiness

LOGLENGREij 0.005

(0.35)

0.068

(2.05)**

0.012

(0.29)

0.084

(1.96)**

PAYLATEij 0.102

(1.49)

-0.268

(-2.02)**

-0.221

(-0.13)

CONTACTij 0.056

(3.31)***

-0.004

(-0.13)

-0.170

(-2.39)**

0.031

(0.47)

VISITij -0.009

(-0.42)

0.057

(1.82)*

0.04

(1.10)

0.419

(5.60)***

Supplier characteristics FSIZEi 0.001

(0.07)

0.039

(1.14)

0.310

(5.86)***

0.079

(1.12)

BEUNPAIDi 0.076

(4.72)***

0.055

(1.20)

0.099

(4.39)***

0.314

(6.63)***

LOGAGEi -0.061

(-1.81)*

-0.052

(-0.70)

-0.143

(-2.15)**

0.038

(0.39)

LACKCAPi -0.051

(-1.77)*

-0.113

(-2.35)**

-0.016

(-0.30)

-0.232

(-3.16)***

MARGINi -0.036

(-0.16)

-0.364

(-0.51)

0.729

(1.12)

0.676

(1.17)

Const 0.335

(0.68)

-0.695

(-1.01)

-4.32

(-4.67)***

-3.35

(-2.57)***

Log likelihood -104.698 -136.228 187.82 -187.669

PseudoR2 0.285 0.1646 0.2033 0.2466


Table 6.3 shows the results of equation (6.1) for each of the four groups: w-millers,

wholesalers, millers, and retailers. The standard errors are robust and adjusted for cluster

effects. The T-statistics are given in parentheses above. *, ** and *** denote significant

levels at 10, 5 and 1 per cent, respectively.

135

References:

Aaronson, Daniel and Bostic, Raphael, Huck, Paul and Townsend, Robert, 2004,

“Supplier Relationships and Small Business Use of Trade Credit”, Journal

of Urban Economics, 55, 1, 46-67

Almus, M., 2002, “What characterizes a fast-growing firm?” Applied Economics,

34, 1497-1508

Atanasova, Christina V. and Wilson, Nicholas, 2003, “Bank Borrowing Constraints

and the Demand for Trade Credit: Evidence from Panel Data”, Managerial

and Decision Economics, Vol. 24, No. 6/7, 2003, 503-514

Backer, Koen & Sleuwaegen, Leo, 2003, “ Foreign Ownership and Productivity

Dynamics”, Economic Letters, 79, 2, 177-83

Beck, T. Demirguc-Kunt, A., and V. Maksimovic, 2006, “The Influence of

Financial and Legal Institutions on Firm Size”, Journal of Banking and

Finance, 2995-3015

Blazenko, George W and Vandezande, Kirk, 2003, “The Product Differentiation

Hypothesis for Corporate Trade Credit”, Managerial Decision Economics,

24, 457-469

Bottazzi, G., Dosi, G., Lippi, M., Pammolli, F., Riccaboni, M., 2001, “Innovation

and Corporate Growth in the Evolution of the Drug Industry”, International

Journal of Industrial Organization 19, 1161-1187.

Bottazzi, G. and A. Secchi, 2003, “A Stochastic Model of Firm Growth”, Physica

A, 324, 213-219

136

Bottazzi, G., Secchi, A. and Tamagni, F, 2008, “Productivity, Profitability and

Financial Performance” Industrial and Corporate Change, 17, 711-751

Brown, J David & Earle, John S & Lup, Dana, 2005, What Makes Small Firms

Growth? Finance, Human Capital, Technical Assistence, and The Business

Environment In Romania, Economic Development and Cultural Change,

University of Chicago Press, 54, 1, 33-70,

Calvo, J.L., 2006, "Testing Gibrat's Law for Small, Young, and Innovating Firms",

Small Business Economics, 26, 2, 117-123

Chung, K.J., 1998, “A theorem on the determination of economic order quantity

under conditions of permissible delay in payments”, Computers and

Operations Research, 25, 49-52

Cook, Lisa D., 1999, “Trade Credit and Bank Finance: Financing Small Firms in

Russia”, Journal of Business Venturing, 14, 5-6, 493-518

Cunat, Vicente, 2007, “Trade Credit: Suppliers as Debt Collectors and Insurance

Providers”, Review of Financial Studies, 20, 2, 491-527

Delannay, Anne-France and Weill, Laurent, 2004, “The Determinants of Trade

Credit in Transition Countries,” Economics of Planning, 37, 3-4, 173-193

Demirguc-Kunt and Maksimovic, Vojislav, 2002, Firms as Financial

Intermediaries: Evidence from Trade Credit Data, World Bank, Unpublished

Working Paper

Dunne, P. and A. Hughes, 1994, “Age, Size, Growth and Survival: UK

companies in the1980s”, Journal of Industrial Economics, 42, 2, 115-140.

137

Ebben, Jay and Johnson, Alec, 2005, “Bootstrapping in Small Firms: An Empirical

Analysis of Change over Time”, Journal of Business Venturing, 21, 851-

865.

Ehrlich, Isaac & Gallais-Hamonno, Georges, Liu, Zhiqiang and Lutter, Randall

1994, “Productivity Growth and Firm Ownership: An Analytical and

Empirical Investigation”, Journal of Political Economy, University of

Chicago Press, 102, 5,1006-38

Evans, D.S., 1987a, “Tests of Alternative Theories of Firm Growth”, Journal of

Political Economy, 95, 4, 657-674.

Evans, D.S., 1987b, “ The Relationship between Firm Growth, Size and Age:

Estimates for 100 Manufacturing Industries”. Journal of Industrial

Economics, 35, 567-581.

Fabbri, Daniela and Klapper, Leora F., 2008, Market Power and the Matching of

Trade Credit Terms, World Policy Research Working Paper, 4754,

Washington DC: World Bank

Ferris, J. Stephen, 1981, “A Transaction Theory of Trade Credit Use”, Quarterly

Journal of Economics, 96, 247-270

Fisman, Raymond and Love, Inessa, 2002, “Trade Credit, Financial Intermediary

Development, and Industry Growth”, Journal of Finance, 58, 353-374

Fisman, Raymond and Raturi, Mayank, 2004, “Does Competition Encourage

Cooperation? Evidence from Trade Credit Relationships” Review of

Economics and Statistics, 86, 1, 345-352.

Freel, M.S., 2000, “Do Small Innovating Firms Outperform Non-Innovators?”

Small Business Economics 14, 195-210.

138

Geroski, P. A. and Toker, S., 1996, “The turnover of market leaders in UK

manufacturing industry 1979-86”, International Journal of Industrial

Organization 14, 141-158.

Geroski, P. A., K. Gugler, 2004, “Corporate Growth Convergence in Europe”,

Oxford Economic Papers, 56, 597-620

Giannetti, Mariassunta, Burkart, Mike and Tore Ellingsen, forthcoming, “What

You Sell is What You Lend? Explaining Trade Credit Contracts”, Review of

Financial Studies.

Gibrat R, 1932, Les inegalites economiques , Paris, Sirey

Griliches, Zvi and Lichtenberg Frank R., 1984, R&D and Productivity Growth at

Industry Level: Is There Still A Relationship? University of Chicago Press,

465-502

Goddard, J., J. Wilson and P. Blandon, 2002,” Panel tests of Gibrat’s Law for

Japanese Manufacturing”, International Journal of Industrial Organization,

20, 415-433

Hay, M., K. Kamshad, 1994, Small Firm Growth: Intentions, Implementation and

Impediments, Business Strategy Review, 5, 3, 49-68

Huang, Y. F, 2007, “Optimal Retailer’s Replenishment Decisions in The EPQ

Model Under Two Levels of Trade Credit Policy”, European Journal of

Operational Research, 176, 1577-1591

Huck, J.F. and McEwen, T., 1991, “Competencies needed for Small Business

Success: Perceptions of Jamaican Entrepreneurs”, Journal of Small

Business Management, 29, 4, 90-93.

139

Hyndman, Kyle and Serio, Giovanni, 2010, Competition and Inter-firm Credit:

Theory and Evidence from Firm-level data in Indonesia”, Journal of

Development Economics, 93, 88-108

Ibrahim, B. Ahmed and James R. Goodwin, 1986, “Perceived Causes of Success in

Small Business”, American Journal of Small Business, 11, 2, 41-50

Jain, Neelam, 2001, “Monitoring Costs and Trade Credit,” The Quarterly Review

of Economics and Finance, 41, 89-110.

Jovanovic, B.,1982, “Selection and the evolution of industry”, Econometrica, 50,

649-670.

Long M.S., Malitz I.B. and Ravid S.A.,1993, “Trade Credit, Quality Guarantees,

and Product Marketability”, Financial Management 22, 117-127

Lu, Yi, 2009, Trade credit and Legal System: Evidence from International Data,

Working Paper, School of Business, University of Hongkong, Hongkong

Lutz, Clemens, Praagman, Kees and Luu Thanh Duc Hai, 2006, “Rice Market

Integration in the Mekong River Delta,” Economics of Transition, 14, 3, 517-

546

Lutz, Clemens, Kemp, Ron and Dijkstra S. Gerhard, 2010, “SMEs Perceptions

Regarding Strategic and Structural Entry Barriers”, Small Business

Economics, 35, 1, 19-33.

Luu, Thanh Duc Hai, 2003, The Organization of The Liberalized rice market in

Vietnam, Center of Development Studies, University of Groningen.

McMillan, John and Woodruff, Christopher, 1999, “Interfirm Relationship and

Informal Credit in Vietnam”, Quarterly Journal of Economics, 114, 1285-

1230

140

McPherson, M.A., 2006, “Growth of Micro and Small Enterprises in Southern

Africa”, Journal of Development Economics, Vol.48, 53-277

Minot, Nicholas and Goletti, Francesco, 2000, Rice Market Liberalization and

Poverty in Vietnam, Research Report No. 114, Washington DC: IFPRI

Musso, Patrick and Schiavo, Stefano, 2008, “ The Impact of Financial Constraints

on Firm Survival and Growth”, Journal of Evolutionary Economics, 18, 2,

135-149

Nadiri. M.I, 1969, “The Determinants of Trade Credit in The US Total

Manufacturing Sector”, Econometrica, 37, 408-423

Nguyen, Ngoc Anh, Pham, Ngoc Quang, Nguyen, Chuc Dinh and Nguyen, Nhat

Duc, 2007, “Innovation and Exports in Vietnam’s SME Sector”, The

European Journal of Development Research, 20, 262-280

Nguyen, Tran Dinh Khoi and Ramachandran, Neelakantan, 2006, “Capital

Structure in Small and Medium-Sized Enterprises: The Case of Vietnam”,

ASEAN Economic Bulletin, 23, 2, 192-211

Petersen, Mitchell and Rajan, Raghuram G., 1997, “Trade Credit: Theory and

Evidence” The Review of Financial Studies, 10, 3, 661-691

Porter, Michael E., 1979, “How Competitive Forces Shape Strategy”, Harvard

Business Review 57, 2, 137-145

Porter, Michael, E., 1980, Competitive Strategy: Techniques for Analyzing

Industries and Competitors. New York: Free Press, 1980

Porter, Michael E., 1985, The Competitive Advantage: Creating and Sustaining

Superior Performance. New York: Free Press, 1985

141

Porter, Michael E., 2008, “The Five Competitive Forces That Shape Strategy”,

Harvard Business Review, 86, 1

Prais, S. J.,1974, A New Look at the Growth of Industrial Concentration, Oxford

Economic Papers 26 , 2, 273-288

Rachamadugu, R., 1989, “Effects of Delayed Payments (Trade credit) on Order

Quantity”, Journal of the Operational Research Society, 805-813

Rand, John, 2007, “Credit Constraints and Determinants of The Costs of Capital in

Vietnamese Manufacturing”, Small Business Economics, 29, 1, 1-13

Roper, S., 1997, “Product Innovation and Small Business Growth: A Comparison

of The Strategies of German, UK and Irish Companies”, Small Business

Economics 9, 523-537

Sakai, H and Takada, N, 2000, Developing Small and Medium Scale Enterprises in

Vietnam, NRI Paper No 13, Nomura Research Institute

Smith J.K.,1987, “Trade Credit and Informational Asymmetry”, Journal of

Finance 42, 863-872

Schwartz, Robert A., 1974, “An Economic Model of Trade Credit”, Journal of

Financial and Quantitative Analysis, 9, 643-57

Singh, A. and Whittington, 1975, “The Size and Growth of Firms”, Review of

Economic Studies,42, 1, 15-26

Solow, 1957, “Technical Change and the Aggregate Production Function”, Review

of Economics and Statistics, 39, 312-20

Summer, Barbara and Wilson, Nicholas, 2003, “Trade Credit and Customer

Relationship”, Managerial and Decision Economics, 24, 6-7, 439-455

142

Van Auken, Howard, 2005, “Differences in the Usage of Bootstrap Financing

among Technology-Based versus Nontechnology-Based Firms”, Journal of

Small Business Management, 43, 1, 93-103.

Van Horen, Neeltje, 2005, Trade Credit as a Competitiveness Tool: Evidence from

Developing Countries, Working paper, University of Amsterdam,

Amsterdam

Van Horen, Neeltje, 2007, Customer Market Power and The Provision of Trade

Credit: Evidence from Eastern Europe and Central Asia, Working paper,

University of Amsterdam, Amsterdam

Variyam, J. N. and D. S. Kraybill, 1992, “Empirical evidence on determinants of

firm growth”, Economics Letters, 38, 4, 31-36

Wilson, Nicholas, and Summers, Barbara 2002, “Trade Credit Terms Offered by

Small Firms”, Journal of Business Finance and Accounting, 29, 3, 317-35

Winborg, Joakim and Landstrom, Hans, 2000, “Financial Bootstrapping in Small

Business: Examining Small Business Managers’ Resource Acquisition

Behaviors”, Journal of Business Venturing, 16, 235-254.

Wijewardena, Hema and Tibbits, Garry E., 1999, “Factors Contributing To The

Growth of Small Manufacturing Firms: Data from Australia”, Journal of

Small Business Management, April, 88-95

Yasuda, T.,2005, “Firm Growth, Size, Age and Behavior in Japanese

Manufacturing” Small Business Economics, 24, 1-15

143

Appendix 1:

Factor Analysis on measures of perceived competition and bargaining

power

Factor analysis

Factor analysis/correlation Number of obs = 568

Method: principal factors Retained factors = 2

Rotation: (unrotated) Number of params = 6

Factor loadings (pattern matrix) and unique variances

Variable Factor1 Factor2 Uniqueness

BPWCUS -0.408 0.190 0.798

PERCOMP 0.3468 0.2838 0.799

NORIVALS 0.167 0.379 0.829

SALDEC 0.445 -0.189 0.766

Factor Eigenvalue Difference Proportion Cumulative

Factor1 0.512 0.216 1.293 1.293

Factor2 0.296 0.448 0.747 2.040

Factor3 -0.152 0.108 -0.384 1.656

Factor4 -0.260 0.000 -0.656 1.000

LR test: independent vs. saturated: chi2(6) = 118.26 Prob>chi2 = 0.0000

Rotation

144

Factor analysis/correlation Number of obs = 568

Method: principal factors Retained factors = 2

Rotation: orthogonal varimax Number of params= 6

Factor Variance Difference Proportion Cumulative

Factor1 0.459 0.110 1.159 1.159

Factor2 0.349 . 0.882 2.040

Rotated factor loading and unique variances:

Variable Factor1 Factor2 Uniqueness

BPWCUS -0.448 -0.037 0.798

PERCOMP 0.1602 0.419 0.799

NORIVALS -0.044 0.412 0.829

SALDEC 0.480 0.057 0.766

Based on the rotated factor loading, we name the two variables:

Factor1= CUSPOW (Customer bargaining power vis-à-vis a supplier, formed by

BPWCUS and SALDEC, same sign with SALDEC)

Factor2= COMP (Competition from existing rivals, formed by PERCOMP and

NORIVALS)

145

Table 4. 5:Tobit regressions of the relationship between trade credit supply

and competitiveness (dependent variable: percentage of total sales on

delayed payments)

Variables Coefficients

WHSALE 0.271

(5.87)*** W_MILLER 0.261

(6.94)*** RETAIL 0.199

(3.58)*** AGE 0.002

(0.90) LACKCAP -0.052

(-3.69)*** FSIZE 0.037

(3.85)*** DIFLOAN -0.103

(-3.01)*** MARGIN 0.204

(1.49) BUYDP 0.294

(5.09)*** UNPAID 0.046

(5.68)*** CONTACT 0.033

(2.74)*** CUSPOW 0.055

(2.10)** COMP 0.153

(5.64)*** CONSTANT -0.773

(-3.73)*** Log likelihood -154.886 Pseudo R

2 0.438

Observations 522

147

Samenvatting

De meeste studies over het gebruik van handelskrediet maken gebruik van data van

bedrijven in ontwikkelde landen. Deze studie bestudeert het gebruik van

handelskrediet door bedrijven die actief zijn in verschillende onderdelen van de

supply chain in de rijstsector in de Mekong Delta van Vietnam. Het proefschrift

beoogt een bijdrage te leveren aan de inzichten met betrekking tot de vraag

waarom handelskrediet wordt verstrekt en welke rol het verstrekken van

handelskrediet speelt in het creëren van waarde voor het kredietverlenende bedrijf.

Het grootste deel van het onderzoek is gebaseerd op de resultaten van een

uitgebreide survey onder ruim 600 bedrijven. Het onderzoek bestaat uit vier

empirische studies. De eerste studie onderzoekt de relatie tussen het verstrekken

van handelskrediet en de groei van de afzet van het kredietverstrekkende bedrijf,

waarbij gebruik wordt gemaakt van publiek beschikbare financiële informatie van

beursgenoteerde Vietnamese ondernemingen. De studie laat zien dat er een

positieve correlatie bestaat tussen het verlenen van handelskrediet en de groei van

de afzet. De tweede studie maakt gebruik van de primaire data op basis van de

survey en onderzoekt de relatie tussen het verlenen van handelskrediet in de

rijstsector en de mate waarin een bedrijf onderhevig is aan concurrentie. De

resultaten van deze studie suggereren dat bedrijven die in sterkere mate onderhevig

zijn aan concurrentie meer handelskrediet verlenen. Een mogelijke interpretatie

van dit resultaat is dat bedrijven kredietverlening gebruiken om hun positie ten

opzichte van de concurrentie te behouden dan wel te versterken. In de derde studie

wordt dit resultaat verder onderzocht. Aan de hand van informatie op het niveau

148

van handelstransacties wordt aangetoond dat bedrijven geneigd zijn meer krediet te

verlenen aan grotere en bekendere cliënten. Dit kan worden gezien als een verdere

ondersteuning van de hypothese dat handelskrediet gebruikt wordt als een

instrument om de verkoop te stimuleren. Ten slotte wordt onderzocht in hoeverre

specifieke marktomstandigheden (zoals de mate van concurrentie, de

onderhandelingspositie van handelspartners in onderlinge transacties,

klantkenmerken, en de waarde van de verhandelde goederen) verschillend zijn

voor verschillende onderdelen van de supply chain in de rijstsector van Vietnam.

Uit de analyse blijkt dat marktomstandigheden inderdaad sterk verschillen per

onderdeel van de supply chain en dat dit gevolgen heeft voor het gebruik van

handelskrediet. Toekomstige studies naar het belang en het gebruik van

handelskrediet dienen daarom rekening te houden met de specifieke

marktomstandigheden waarin dit plaatsvindt.

University of Groningen Trade credit in the rice market of ... · Dao Thi Ba ), my aunties (Y Uong,...

Documents

Transcript of University of Groningen Trade credit in the rice market of ... · Dao Thi Ba ), my aunties (Y Uong,...