Báo cáo khoa học nông nghiệp " SMALL-MEDIUM ENTERPRISES IN THE LIVESTOCK FEED SECTOR IN VIETNAM ( VOLUME I )" potx

INSTITUTE OF POLICY AND STRATEGY FOR AGRICULTURE AND RURAL DEVELOPMENT CENTER FOR AGRICULTURAL POLICY --- CARD Project 030/06 VIE: Developing a strategy for enhancing the competitiven

INSTITUTE OF POLICY AND STRATEGY FOR AGRICULTURE AND RURAL DEVELOPMENT

CENTER FOR AGRICULTURAL POLICY

-

CARD Project 030/06 VIE: Developing a strategy for enhancing the

competitiveness of rural small and medium enterprises in the agro-food

chain: the case of animal feed

SMALL-MEDIUM ENTERPRISES IN THE LIVESTOCK FEED SECTOR IN VIETNAM

VOLUME I: Livestock feed production Pham Thi Lien Phuong 1 , Nguyen Thi Thinh 1 , Donna Brennan 2 , Sally Marsh 2 , Bui

This report provides a current overview of the Vietnamese livestock feed production sector Besides outlining the environment and infrastructure within which feed mills are working, an important part of the report was to provide information on how feed mills operate along their supply chain, describing relationships with their material suppliers and output customers Data was collected during mid-2008 using a survey of 62 feed mills situated throughout the country Information collected included that concerning the production and business activities of the different sized feed mills, their distribution channels, and their technology levels

The focus of this report is on comparing SMEs and large enterprises with regards to the production and business activities of enterprises such as material input use, storage, product types, quality control, and type of customer These activities give indications of how SMEs compete with larger feed mill enterprises The mills have been categorised in the following way in the analyses: small mills are those producing less than 10,000 tonnes per annum; medium mills are those producing from 10,000 to less than 60,000 tonnes per annum, and large mills are those producing 60,000 or more tonnes per annum

In this study we have explored aspects of competitiveness in the feed mill sector Competitiveness in the sector is likely to involve more than cost efficiencies as a result of economies of scale Competitiveness in the sector is also associated with quality aspects

of feed (and perceptions of quality), services provided along with the sale of feed, and procurement and distribution channels used by the mills

We found statistical evidence in this study that cost of production was inversely related to scale, with small enterprises having significantly higher costs of production per kg of output than medium enterprises which had higher costs of production than large enterprises This alone is not necessarily an indicator of greater inefficiency of SMEs For example, we found that SMEs were more focused on production of concentrate than larger mills Concentrate production has higher raw material input costs per kg of output,

so cost of production per kg of total output would necessarily be higher for those producing more concentrate

We found no statistical evidence to indicate that small enterprises paid more for the key raw material inputs used in production of feed Analyses of prices paid by location also indicated that there were no significant differences between the northern and the southern regions of Vietnam Our findings indicated that raw material costs made up about 80% of the total cost of production in feedmills An analysis of costs other than raw material costs showed that large mills had significantly lower unit costs than small mills We also found other indicators, outlined in detail in the report, to suggest that the lower cost of production experienced by large feedmills may reflect greater efficiency

Our data indicate that small mills (producing less than 10,000 tonnes per annum) are likely to struggle to remain competitive The data indicate that they face significantly higher costs, and sell some feeds at significantly lower prices, resulting on average in a significantly lower profit Anecdotally, this is supported by reports of small mills ceasing business, and our observations when conducting the survey of many previously listed small mills no longer in business However, our results support the idea that medium-size mills (producing between 10,000 and 60,000 tonnes per annum) are remaining competitive, and have costs and product mix and prices similar to large mills

Our results show that the supply/distribution chains operating for SMEs and large firms in the livestock feed sector are quite different SMEs source their raw materials and distribute their products differently to large feedmills They deal more with small householders and traders both for the procurement of raw materials and in the distribution

of their products Large mills are more dependent on imports (e.g maize) to meet their raw material requirements, whereas SMEs are more likely to be able to source sufficient domestic supplies It is likely that small-medium feedmills do provide more direct support

to other small-medium enterprises operating in the livestock sector

Quality control operates at a lower level in domestic mills compared to foreign-owned mills and mills operated by joint venture partners, although protein content (as nominated

on the feed label) was not statistically lower in feeds produced by domestic mills compared to foreign mills However, the fact that advanced international standards for quality control such as ISO and HACCP are only applied by foreign-owned and joint venture companies, indicates better quality control processes for both raw materials and output products from foreign/large companies These mills are also more likely than domestic enterprises to have a quality control laboratory, to have separate production lines, to own automatic cleaning systems and to use least-cost feed ration software

Based on the findings of this study we make the following policy recommendations for policy makers:

• Focus is needed on quality control If long term food safety and export potential is

to be reached, the Government of Vietnam needs to address quality control issues associated with the production of livestock feed products

• Restrictions on the movement of goods due to irregular police conduct need to be addressed

• Invest in domestic production of raw materials used in feed production

• Credit support for SMEs in the sector should be expanded

• Support and expand the role of the Vietnamese Animal Feed Association (VAFA)

• Investigate the possible ways the government may provide price support for raw materials and livestock feed outputs

We suggest the following recommendations for SMEs operating in the sector:

• Small mills need to increase their scale of operation

• Quality control standards need to be raised

• Continue to explore and exploit niche market opportunities

• Consider advantages from diversification and/or a cooperative structure

• Support a strong role for the VAFA

ACKNOWLEDGEMENTS

The authors gratefully acknowledge funding for this research work from AusAID’s CARD Program for Project 030/06 VIE: Developing a strategy for enhancing the competitiveness of rural small and medium enterprises in the agro-food chain: the case of animal feed

The authors acknowledge contributions made to the research work from CAP staff other than the listed authors, including: Pham Tuyet Mai, Tran Cong Thang, Nguyen Ngoc Que, Nguyen Do Anh Tuan, Nguyen Anh Phong and Nguyen Le Hoa

The authors also acknowledge helpful discussions with Mr Le Van Lich (VAFA), Mr Tran Cong Xuan (VPA), Ms Bui Thi Oanh (MARD) and Mr La Van Kinh (South VAAS)

as well as participants in the two stakeholder workshops held in December 2009 in Hanoi and January 2010 in HCMC Advice on technical issues was received from the Vietnamese experts mentioned above, and also from Dr Johanna Pluske (livestock economics consultant) and Professor John Pluske (animal nutritionist from Murdoch University in Western Australia.)

Feedback on draft versions of this report was provided by Dr Johanna Pluske and we highly appreciate her useful and critical comments

Finally, we acknowledge and thank feed mill survey participants for their willingness to support the survey work and give details associated with their businesses

LIST OF ABBREVIATIONS

ANOVA Analysis of Variance

ASEAN Association of South East Asian Nations

CAP Center for Agricultural Policy

CARD Collaboration for Agriculture and Rural Development

DLP Department of Livestock Production

FAO Food and Agriculture Organization

GDP Gross Domestic Product

GMP Good Management Practice

HACCP Hazard Analysis Critical Control Points

ISO International Organization for Standardization

MARD Ministry of Agriculture and Rural Development

MCP Mono Calcium Phosphate

NSD No Significant Difference

SD Standard Deviation

SME Small Medium Enterprise

SOE State Owned Enterprise

VAFA Vietnamese Animal Feed Association

VAT Value Added Tax

VBARD Vietnamese Bank for Agriculture and Rural Development

VND Vietnamese Dong

TABLE OF CONTENTS

EXECUTIVE SUMMARY i

LIST OF FIGURES vi

LIST OF TABLES vii

1 INTRODUCTION 9

1.1 Background 9

1.1.1 The livestock sector in Vietnam 9

1.1.2 Small medium enterprises (SMEs) in Vietnam 9

1.1.3 SMEs in the livestock feed sector in Vietnam 10

1.2 Objectives 10

1.3 Research questions 11

1.4 Methodology 11

1.4.1 Background research 11

1.4.2 Survey design and sites 12

1.4.3 Sampling design 12

1.4.4 Data collection and processing 13

1.5 Limitations and structure of the report 15

1.5.1 Limitations 15

1.5.2 Structure 15

2 RECENT CHANGES IN THE VIETNAMESE FEED INDUSTRY 16

3 BACKGROUND INFORMATION ON FEED ENTERPRISES IN VIETNAM 21

3.1 General characteristics 21

3.2 Labor use and wages 22

3.3 Infrastructure 25

4 COST STRUCTURE AND RAW MATERIAL USE 28

4.1 Cost structure of feed enterprises 28

4.2 Procurement of raw materials 29

4.2.1 Prices paid for raw materials and percentage of raw material from different sources 29

4.2.2 Providers of raw materials 33

4.2.3 Payment method for input purchase 34

5 FACTORY OUTPUT 36

5.1 Output types 36

5.2 Profit 39

5.3 Market share and competition 40

5.4 Nutrient content and additives use 40

6 SUPPLY CHAIN 42

6.1 Distribution channels – transportation distances 42

6.2 Distribution channels - type of customer 42

6.3 Overview of input supply and output distribution channels for different scale enterprises 45

6.3.1 Supply sources and distribution channels for large feed mills 45

6.3.2 Supply sources and distribution channels for medium feed mills 46

6.3.3 Supply sources and distribution channels for small feed mills 47

6.3.4 Summary of input supply and distribution channels 47

6.4 Payment methods 48

6.4.1 Payment methods for complete feed 48

6.4.2 Payment methods for concentrate feed 49

6.5 Services 50

6.5.1 Services to agents 50

6.5.2 Service to livestock producers 51

6.6 Determination of livestock feed prices 52

6.6.1 Setting of feed prices 52

6.6.2 Commission rates 53

6.6.3 Selling price changes made by enterprises during 2007 54

6.7 Product quality control 56

6.7.1 Certification and testing 56

6.7.2 Processing method 58

6.7.3 Product formulation 59

6.7.4 Post-production and storage 59

6.8 Location, information sources and restrictions on the movement of goods 62

6.8.1 Location of firms 62

6.8.2 Restrictions faced by mills on the movement of goods 63

6.8.3 Sources of information for feed mill enterprises 64

7 OTHER ISSUES 66

7.1 Credit 66

7.2 Profitability and investment 67

7.3 Issues, opportunities and constraints 69

8 SUMMARY OF KEY FINDINGS AND IMPLICATIONS FOR POLICY 72

8.1 Main findings 72

8.1.1 Costs of production 72

8.1.2 Revenue and production activities 73

8.1.3 Profitability 74

8.1.4 Procurement (and storage) of raw material inputs 74

8.1.5 Distribution of livestock feed products 75

8.1.6 Quality control 75

8.1.7 Services to clients 76

8.1.8 Constraints on production 76

8.1.9 Opportunities 77

8.2 Evidence of returns to scale 77

8.3 Strategies currently being used by SMEs to compete in the livestock feed sector 78

8.3.1 Location 78

8.3.2 Product mix 78

8.3.3 Commission rate on concentrate feed 79

8.3.4 Discounted prices 79

8.3.5 Supply chain differences 79

8.3.6 Payment in advance for inputs and provision of credit for output sales 79

8.3.7 Services provided to agents and farm households 80

8.4 Role played by SMEs in the livestock feed sector 80

8.5 Policy recommendations 80

8.5.1 For policy makers 80

8.5.2 For SMEs 82

APPENDIX 84

REFERENCES 88

LIST OF FIGURES

Figure 1 Number of feed enterprises, total industrial feed production (plus premix) vs

capacity by region in 2006 17

Figure 2 Vietnam’s import value of animal feed and inputs for feed processing, 2001 - 2008 19

Figure 3 Prices of some raw feed inputs in Vietnam during 2007-2008 20

Figure 4 Prices of raw materials and pig complete feed for Proconco company in 2007 20 Figure 5 Distribution of feed enterprises by production scale in north and south Vietnam (%) 21

Figure 6 Labor size of feed enterprises by foreign and domestic ownership 23

Figure 7 Labor size of feed enterprises by production scale 24

Figure 8 Average salary for labor of feed enterprises by production scale 24

Figure 9 Raw material purchase costs per ton of output, by ownership type and production scale 29

Figure 10 Composition of energy rich ingredients as percentage of total energy inputs used 30

Figure 11 Composition of protein rich ingredients as a percentage of total protein inputs used 30

Figure 12 Inputs used to produce one tonne of feed output by production scale 33

Figure 13 Payment methods used for raw material purchases, by production scale 35

Figure 14 Percentage of firms paying VAT for raw material inputs, by production scale .35

Figure 15 Percentage of firms producing complete and concentrate feed 37

Figure 16 Supply sources and distribution channels for large feedmills in Vietnam 45

Figure 17 Supply sources and distribution channels for medium feedmills in Vietnam 46

Figure 18 Supply sources and distribution channels for small feedmills in Vietnam 47

Figure 19 Percentage of feed mills and retail agents fixing the selling price, by production scale 53

Figure 20 Commission on factory gate prices obtained by agents (wholesale and retail agents combined), for product from small, medium and large mills 54

Figure 21 Percentage change in the factory gate price of the main product during the quarters 2, 3 and 4 and the total change in 2006/2007, by production scale 55

Figure 22 Laboratories used for testing raw materials and products by foreign and domestic mills and by large, medium and small mills 57

Figure 23 Methods used by mills for cleaning between product batches 59

Figure 24 Total days and percentage of the storage time of the main product by mills, agents and others by production scale 61

Figure 25 Percentage of main product reaching expiry date 61

Figure 26 Percentage of feed enterprises by production size nominating various characteristics about factory location 62

Figure 27 Percentage of feed enterprises indicating the most important restrictions affecting the movement of goods 63

Figure 28 Reasons given by firms for not being able to obtain more funds 66

Figure 29 Feed sectors expected by feed mill enterprises to experience the greatest future growth, by scale of enterprise 69

Figure 30 Percentage of feed mill enterprises that said that government should provide support in the following areas, by enterprise scale 71

LIST OF TABLES

Table 1 The number of enterprises listed as available for the survey, the number of enterprises surveyed in the project and the percentage of enterprises in each province

surveyed with respect to the project total 13

Table 2 For each ownership category: the total number of mills in the original sample; the number in each classification and total number providing consistent information for data analysis; and the percentage of the original sample that provided consistent information 14

Table 3 Industrial feed production in Vietnam during the period 2000 – 2008 16

Table 4 Domestic production of the main inputs for feed production and estimated demand for feed input (1000 tonnes) 18

Table 5 Quantity of imported feed inputs to Vietnam in 2006 (‘000t) 18

Table 6 Actual and designed output of feed mills, and percent utilization of designed output in 2007, by scale and region 22

Table 7 Average revenue from activities of feed enterprises in 2007 22

Table 8 Land area owned/leased and percentage of land being used by the feed enterprises, by scale and region 25

Table 9 Distribution of enterprises owning or renting land and location (percent) 26

Table 10 Percentage of enterprises having storage equipment, by scale and region 26

Table 11 Average capacity (tonnes) of storage facilities 26

Table 12 Period, volume and place of material storage of feed enterprises by region and scale 27

Table 13 Cost of production (‘000 VND per kg output) and share of total cost % 28

Table 14 Comparison of raw material input prices by scale, source and location 31

Table 15 Percentage of raw material purchases from different sources by production scale 32

Table 16 Percentage of raw material inputs purchased from different suppliers for each production scale of feed enterprises 34

Table 17 Percentage of firms producing each type of animal feed, by production scale and region 36

Table 18 Percentage of firms producing complete and concentrate feed for livestock types 38

Table 19 ANOVA analysis of pig and chicken feed prices (‘000VND per kg): mean, standard deviation and price difference by region and production scale 38

Table 20 Profit (million VND) by production scale 39

Table 21 Cost and profit for firms grouped according to relative importance of concentrate production 40

Table 22 Protein content of main products and share of firms using additives, by ownership type 41

Table 23 Average transportation distances by region and production scale 43

Table 24 Amount and share of complete feed sold to different customer types by small, medium and large feed mill enterprises 43

Table 25 Amount and share of concentrate feed sold to different customer types by small, medium and large feed mill enterprises 44

Table 26 Payment methods used by purchasers of complete feed from enterprises by different scale 49

Table 27 Payment methods used by purchasers of concentrate feed from enterprises by different scale 50

Table 28 Percentage of feedmill companies providing different service to agents 51

Table 29 Percentage of feedmill companies providing different services to livestock

producers 52

Table 30 Percentage of feed enterprises fixing the selling price for wholesale agents/ traders, by production scale 52

Table 31 Main reasons for price change of feed products (percentage of foreign, domestic and total mills) 55

Table 32 Percentage of firms having formal certification status 56

Table 33 Percentage of mills with quality control laboratories and percentage of testing done in those laboratories 57

Table 34 Percentage of firms that complete various tests of raw feed materials and products 58

Table 35 Expiry period (days) nominated by mills for products, by region 60

Table 36 Mean rating of importance of location characteristics 63

Table 37 Most important information sources for feed enterprises, by scale 65

Table 38 Credit information of enterprises by scale 66

Table 39 Percentage of all loans by sources, average loan amount and length of loan, by scale of enterprise 67

Table 40 Percentage of loans for feed/raw materials, buildings/equipment and other purposes from various loan sources, by scale of enterprise 68

Table 41 Main reasons nominated by feed mills for change in profit in 2007 and 2005 68 Table 42 Expectations of feed mill enterprises of the domestic supply for raw material in the future, by enterprise scale 70

1 INTRODUCTION

1.1 Background

1.1.1 The livestock sector in Vietnam

Global livestock feed prices have generally been rising and this trend seems likely to continue into the future This increase is due to two main reasons: firstly, a higher demand for livestock feed results from a higher demand for meat and related products; and secondly, an increasing demand for maize and other raw materials for making biofuel products competes with the use of these raw materials for livestock feed (Pluske, 2007)

In Vietnam, the livestock industry has played an increasing role in total agricultural GDP, with its share ranging from 22.6 to 25.5% annually during the period 2001 to 2007, and estimated to increase to between 25 and 26% in 2008 (Livestock Department, 2008) The livestock industry has grown quickly and spontaneously, and a lack of regulation has led

to a number of problems: there are many small and dispersed producers who have problems of low yield, exacerbated by disease outbreaks; and a rapidly growing commercial scale sector with little environmental oversight Further, feed costs contribute

to about 75% of the total production costs of livestock-producing households

In recent times, one of the hottest issues facing the livestock industry has been the increasing trend of feed prices The large dependence on imports of raw material ingredients such as soybean and maize, high import taxes, and low domestic yield of these inputs have been considered as the causes of the high livestock feed prices Industrial feed prices in Vietnam are around 10-15% higher than other countries in the region, such as Thailand and China It is estimated that Vietnam imports about 20-30% by volume of raw materials used for livestock feed production, which accounts for 45% of the total value of raw ingredients (www.mard.gov.vn) High livestock feed prices directly affect producers as they result in higher production costs, especially when the prices of livestock products can not increase sufficiently to cover the increased costs

1.1.2 Small medium enterprises (SMEs)1 in Vietnam

Showing impressive internal strength, the number of Vietnamese enterprises has grown

by fifteen times in the past nine years (1999-2008) (Online policy dialogue of development pathway of Vietnamese SMEs, 2010) Accounting for 97% of more than

400 thousand enterprises operating in Vietnam's economy, small and medium enterprises are recognized as a driving force for the development of Vietnam over time However, SMEs have not contributed as much as expected, due to the recent economic crisis and global recession as well as inherent shortcomings in the business environment

in Vietnam and the internal weakness of many SMEs

Difficulties such as lack of capital, difficult access to land for business premises, outdated technology, weak competitiveness, the level of corporate governance restrictions and difficulties in obtaining market access have been considered as main obstacles for SMEs The Vietnamese Government has recently announced a series of policies on macroeconomic stability, including capital injection and removal of administrative

1 SMEs are independent production and business establishments, which make business registration according to the current law provisions, each with registered capital not exceeding VND 10 billion or annual labor not exceeding 300 people On the basis of the concrete socio-economic situation of each branch or locality, in the course of implementing the support measures and programs, both or either of the above-mentioned criteria on capital and labor may be applied in a flexible manner (Article 3, Government Decree 90/2001/ND-CP , dated 23 November 2001)

procedures for businesses so as to create a facilitating environment for enterprises However it is unclear by how much Vietnamese businesses have benefited from this support The number of small and medium enterprises in Vietnam is still increasing, despite some forecasts that 80% of SMEs are having difficulties, and 20% will eventually disappear (Online policy dialogue of development pathway of Vietnamese SMEs, 2010)

In the context of global economic difficulties and international integration, assessing the competitiveness of SMEs in terms of their position, strengths, weaknesses, opportunities and challenges becomes increasingly urgent These assessments will be an important basis for further state support for the development of SMEs in the future

1.1.3 SMEs in the livestock feed sector in Vietnam

In 2008, large feed mills including CP Group, Cargill and Proconco, made up nearly 70%

of the domestic livestock feed market in Vietnam Vietnamese farmers nominate large mills as the “price makers” Large mills are also criticised as being a catalyst for the increasing feed prices in Vietnam due to their dominant place in the feed market There are open questions about the role of small and medium enterprises (SMEs) in the livestock feed sector, and whether they can compete with large enterprises given the current market structure

There have been different discussions on the classification of SMEs enterprises in general, as well as livestock feed SMEs in particular The classification in Vietnamese feed enterprises is usually based on the design capacity of the mill, rather than the number

of employees, or shareholders’ funds or fixed assets, or capital, all of which are used in some ASEAN countries However, this indicator of design capacity may be used differently by agencies such as MARD, VAFA and enterprises themselves to differentiate between small, medium and large enterprises This issue of classification of enterprise size is still very much open to discussion

As yet, there are very few studies on the economic efficiency of feed mills in Vietnam, and their distribution channels Recent data from the Vietnamese Animal Feed Association (VAFA), dated to February 2009, suggested that there were around 40 domestic trading and feed-producing enterprises that were at risk of production closure or bankruptcy due to costs incurred when import prices increased rapidly There has been little research on the competitiveness of the small and medium mills, compared to the larger ones and especially in terms of their use of input materials and how these relate to output The aim of this study is to fill this gap by providing information on the production and business activities of the different sized feed mills, their distribution channels, and their technology levels This information is important to help policymakers build suitable policies for the livestock industry in the years to come

1.2 Objectives

The main aim of this livestock feed mill survey is to provide a quantitative assessment of the factors affecting the competitiveness of the animal feed industry in Vietnam To do that, the following objectives have been addressed:

• to give an assessment of mill characteristics and production costs, output and prices by types of ownership, regions and size of feed mills;

• to find out the nature of information and product flow, quality control standards and processes for livestock feed mills; and

• to assess the competitiveness of SMEs compared to large feed mills

1.3 Research questions

During the scoping activities carried out for the study, it became apparent that small mills were often regarded as those that usually used cheap, low-protein materials (sometimes with contaminants), lacked adequate quality control and had poor equipment, and therefore sold their product at a relatively low price Their main buyers were thought to

be small producers and agents, especially those living in remote areas However, some people argued that Vietnamese SMEs could survive and compete in niche markets, as do similar feed mills in Thailand

Therefore, in this study, the aim was to gain a deeper insight into feed production and trading generally, especially for both small and large scale mills

The research therefore focused on answering the following questions:

• Are economies of scale evident in the livestock feed sector in Vietnam?

• How different is production and trading between large feed mills and SMEs in terms of material input use, storage, product types, quality control, types of customers and services offered to customers?

• Are the raw material procurement and output distribution channels used by SMEs and larger feed mills different?

• How do domestic SMEs compete in the sector against larger foreign-owned mills?

• Is there any evidence of prices for raw material imports being higher than domestic prices for raw material inputs?

• Is there an opportunity for Vietnamese SMEs to compete in niche markets? (e.g

smaller mills targeting more remote areas)?

• What are the constraints facing SMEs operating in the livestock feed sector in

Pluske’s (2007) report titled “A Desktop Review of the Animal Feed Sector at a Global Scale” was used as background material to context the survey The report concluded that

it is possible that SMEs will have a role to play in the feed sector in developing countries, especially for those in remote areas where it is not profitable for the larger companies to operate However, appropriate policies will need to be developed to ensure that the presence of these SMEs is beneficial for society as a whole and not just one sector within

it As a consequence policy makers will need to have adequate knowledge of the relevant supply chains and other institutional arrangements involved in decision making in the feed and animal sectors

To gain an understanding of the current issues facing feed production enterprises and farms, in August 2007, researchers associated with this project engaged in field tours and visited a number of senior industry leaders These activities reiterated the infrastructure and output differences between the large animal feed firms and the SMEs Further, discussions with animal producers and industry leaders provided insight into determinants

of demand for feed

1.4.2 Survey design and sites

The questionnaire for animal feed manufacturers was designed on the basis of all stakeholders’ ideas, pre-tested and finalized after critical discussion The field survey was

a face-to-face interview with a senior staff member of each feed mill Each interview generally took between two and three hours to complete In some instances enumerators needed to return to the enterprise to either complete or clarify the data collection Feed mills in seven provinces across the country were selected to take part in the survey, which took place between May and June 2008

To obtain an overview of animal feed production patterns and feed uses, a survey of pig and chicken producers was also designed as part of this project It was conducted during November and December 2008 in 6 of the same 7 provinces These data are presented and discussed in Volume II

To cover a broad cross-section of animal feed enterprises, the field work was undertaken

in three ecological regions, in provinces that are the biggest producers of livestock feed Feed mills were surveyed in Ha Noi, Ha Tay and Hung Yen provinces located in the Red River Delta; Binh Duong and Dong Nai in the South East; and Tien Giang and Long An

in the Mekong River Delta The three regions had previously been identified in the project proposal, while six out of seven provinces were agreed upon by the project team and CARD, based as being representative of their regions Hung Yen was added because of a lack of enterprises in Ha Noi, resulting from recent changes in this province compared to

2006 data (on which the original selection of provinces was made) and because many of the companies listed in 2006 had gone out of business

1.4.3 Sampling design

Based on a Department of Livestock Production list of feed enterprises current for 2006,

as well as the limitations of the project budget, we initially planned to survey a total of 70 feed mills and agents in the six provinces identified in Section 1.4.2 The sampling strategy, to ensure a representative selection of small, medium and larger mills, was designed as follows:

• A list of 241 feed mills operating in 2006 (including their location and capacity) was obtained from the Department of Livestock Production

• A stratified size classification criteria of mills was specified with mills divided into 5 classes based on output capacity (tonnes/year): (1) <5,000; (2): >=5,000 and <10,000; (3): >=10,000 and < 20,000; (4): >=20,000 and <80,000; and (5):

• In the case that one of the 5 classes was under represented in a province (for example, the class with capacity of more than 80,000 tonnes), mills from other province were included to make up the required number For example, Hung Yen province (in the Red River Delta) has similar characteristics with Ha Noi and Ha Tay and so could be used as a substitute province

• Mills from the list were randomly selected until the desired number was achieved However, many mills on the list had gone out of business and could not be replaced with equivalent mills and hence only 62 mills were used in the study The final sample distribution from the seven provinces is shown in Table 1

Table 1 The number of enterprises listed as available for the survey, the number of

enterprises surveyed in the project and the percentage of enterprises in each

province surveyed with respect to the project total

available

Enterprises surveyed

Percentage surveyed

1.4.4 Data collection and processing

Before each survey, secondary data and information were collected to obtain an overall background of livestock production and feed use in a particular province During the survey, our experienced enumerators worked under close supervision and were randomly checked by the project team The information provided by the respondent sometimes was not consistent between different sections of the questionnaire because of the reluctance of some enterprises to provide certain information After completion of the survey, data entry was conducted using the Microsoft Access program The data entry template was designed and run by the project team at the survey time The final dataset was converted

to Stata™ format, with notes attached for specific cases so it was easier for doing data analysis at the later stage

As mentioned above, some data provided by companies was not very consistent Data checking and cleaning were done very carefully to ensure that matching sections of data were consistent For example, we compared the stated quantity of raw material inputs against product output and in a few cases where they were really inconsistent and we could not resolve why, we rejected the questionnaire from the sample One of the other data problems we encountered was the sensitivity over cost of production questions In order to deal with this we provided respondents with the opportunity to answer cost of production in absolute or percentage terms Then, because we had detailed information on raw material input costs we could use the percentage cost of production data to infer total

costs of production in cases where they did not answer directly In cases where they did

answer the total raw material cost we used a comparison between the stated costs of

production and the calculated raw material input cost as part of the data checking process

Ultimately, although the total sample surveyed was 62 mills, only data from 44 mills were

accepted for analysis of scale and cost of production

Table 2 describes the total original sample size for each ownership status In addition, the

mills that provided consistent production data were able to be classified by size The scale

classification used, based on the animal feed production data of mills in 2007, was as

follows: small - less than 10,000 tonnes/year; medium - from 10,000 to less than 60,000

tonnes/year, and the large group are those with production of 60,000 tonnes or more

Preliminary data analysis indicated that this size classification provided some distinction

between the large and medium-small mills, and the number of small and medium mills

was roughly equivalent The classification of medium and small firms highlighted the

very small mills which made up half the small-medium mills in the survey

Around 70% of the total sample was accepted for providing consistent cost of production

and scale data while the rest was omitted (Table 2) Particularly, nearly half of those

companies in the registered private foreign group were rejected during the data cleaning

process due to either non-response or inconsistency in response on the cost of production

data In this report all the analyses of region and ownership status was done with the total

sample of 62, but data from only 44 mills was used when analyses were based on the

scale classification

Table 2 For each ownership category: the total number of mills in the original

sample; the number in each classification and total number providing consistent

information for data analysis; and the percentage of the original sample that

provided consistent information

Original sample By scale classification a

% of original sample

By ownership status Small Medium Large Total

1.5 Limitations and structure of the report

1.5.1 Limitations

One of the main limitations of this study was difficulties in approaching enterprises and collecting some of the information In particular, most companies were not willing to provide detailed business information related to production and revenue Like other enterprise surveys, the results of this survey depended on external factors that were out of control of the project team Though the sampling method was logical, substantial adjustments had to be made to the actual survey depending on the availability of current mills, as well as respondents’ willingness to answer questions As a result, data cleaning and analyses were difficult and time consuming for our team, as we endeavoured to ensure that information in different sections of the survey was consistent

Secondly, limited enumerators and difficulties in approaching the leaders of enterprises interrupted the survey in different provinces This difficulty together with the dispersed distribution of the 62 feed mills located in 7 provinces and 3 ecological regions of Vietnam meant that survey implementation took longer than anticipated

Another key constraint was the necessity to omit inconsistent data and observations from the analyses A larger sample would have produced results with greater validity and reliability but due to the limited budget, combined with the broad content requirement of the study, it was not possible to extend the sample size

1.5.2 Structure

This report consists of seven additional sections Section 2 is an overview of the most recent changes in the Vietnamese feed industry, including a discussion of production, and price fluctuations of different feed types and raw materials The results of the feed mill survey, including data on general information, production patterns, raw material use, feed output and supply chain as well as other relevant issues (such as transportation, communication, credit, profitability and investment, opportunities, constraints) are presented from Section 3 to 7 A summary of the key findings and policy recommendations for the feed industry in general and SMEs in particular are discussed in the final section

2 RECENT CHANGES IN THE VIETNAMESE FEED INDUSTRY

The livestock feed processing industry in Vietnam has developed rapidly since the 1990s

with the growth of the livestock industry The average growth rate of total feed

production has tended to rise since the year 2000, reaching 16.6% on average from 2000

(‘000t)

Total (‘000t)

Total (complete feed equivalent) (‘000t)

Growth rate (%)

Recognizing a potential profit in feed production, there has been an increasing number of

foreign and domestic companies setting up their businesses in Vietnam Further, due to

the reform policies of the Government as well as foreign and domestic investment

encouragement, multi-national feed companies, such as Cargill, CP, Proconco and Japfa,

have commenced production in Vietnam

Data from the Department of Livestock Production showed that there were 241 registered

feed mills3 in Vietnam in 2006, of which there were 33 foreign, 10 joint ventures and 198

domestic mills The total feed production (complete equivalent plus premix) in 2006 was

6612.4 thousand tonnes, accounting for around 75 per cent of the design capacity of

8803.9 thousand tonnes (Figure 1) The Red River Delta, had nearly half of the total

number of Vietnamese enterprises and had the largest capacity of the 8 regions with total

industrial production being 2427.1 thousand tonnes (Figure 1) The South East region had

the largest feed production (3274.5 thousand tonnes) with mills using almost all of their

capacity (Figure 1) The third largest production region was the Mekong River Delta

2 Growth rate slowed down as animal feed for poultry reduced by 25-30% during bird flu (Department of

Livestock Production 2006)

3 The total number of registered feed mills in Vietnam in 2008 is 225, which consists of 42 foreign, 12 joint

ventures and 171 domestic ones Red River Delta and South East region remain the two biggest feed

producing areas, accounting for 45.8% and 28.9% of the total national feed mills respectively

(771.6 thousand tonnes), followed by the North Central region The smallest feed producing region was the Central Highlands with two small mills producing about 200 tonnes The development of the livestock feed sector, with the existence of big feed enterprises from abroad, has gradually changed the traditional patterns of raising livestock and poultry in Vietnamese households Many households no longer use home produced raw feeds to raise their animals Instead they have increased their use of industrial feed in the Southern provinces, or mix industrial and raw feeds for producing livestock in the North

Delta

Nort

h C ent

ral R

egion

Sou

th Cent

ral Coas t

Cent

l Hi ghla

Industrial feed production Capacity No of enterprises

Source: Department of Livestock Production, 2006

Figure 1 Number of feed enterprises, total industrial feed production (plus premix)

vs capacity by region in 2006

Local input availability for feed production, particularly for protein rich ingredients, is limited compared to local demand for feed, with the deficit being met through imports The deficit was almost 3 million tonnes in 2006 which is almost double that of 2005 (Table 4)

It is estimated that more than 10 types of ingredients out of a total of 22 types is imported

by feed enterprises Data from Department of Livestock Production (2007) showed that the import quantity of materials for feed production in 2006 reached 3170.7 thousand tonnes (Table 5) with a value of 11.8 thousand billion VND and accounted for about 40%

of the 30.4 thousand billion VND of local industrial feed production value Import trends for some raw materials increased markedly in the two following years, especially for

soyabean, soybean cake, fishmeal, premix vitamin, and over 3,000 tonnes of feed materials has been imported in the first half of 2009 It is clear from the data in Table 5

that the Vietnamese animal feed sector depends more on imports for protein inputs than

for energy inputs In addition, the domestic industry has not yet produced materials for

premix and additives

Table 4 Domestic production of the main inputs for feed production and estimated

demand for feed input (1000 tonnes)

1 Broken rice, paddy and bran4 6084 6090

Total domestic output for

Source: Strategy for Livestock Development to 2020, Department of Livestock Production (2007)

Table 5 Quantity of imported feed inputs to Vietnam in 2006 (‘000t)

Feed input

of 2009

4 Broken rice is 3% equivalent to rice; paddy for feed production is equal to 3% of its production and bran is

equal to 11% The share for feed production of maize is 90%; cassava and potato 80% and soybean 33%

Imports of animal feed and materials for feed processing have been on an upward trend, especially in the last two years (Figure 2) In 2001, the import value was less than 200,000 thousand USD and it increased by 6 times in 2007 The estimated value for 2008 was over 1,700,000 thousand USD

Proconco, a large livestock feed joint-venture company, provided prices for some of the main feed inputs as well as for complete feed for pigs (Figure 4) Based on the upward trend in their prices during the year 2007, there appears to be a tendency for feed prices to

be adjusted as input prices go up (Figure 4) Except for the price of cassava which increased only by 10 percent, the prices of other ingredients rose considerably, especially soybean cake which was the most expensive with an increase in price of more than 50% compared to early 2007 The price of complete feed for pigs weighing between 30 and 60

kg, grew by around 30% during the 12 months Given the higher complete feed prices, it would seem that the changes in raw material input costs were passed on to the buyer

Fish meal 60% import protein Argentina soybean cake

Source: Department of Livestock Production, 2008

Figure 3 Prices of some raw feed inputs in Vietnam during 2007-2008

Maize Soybean cake Cassava Complete feed (30-60kg)

Source: Department of Livestock Production 5

Figure 4 Prices of raw materials and pig complete feed for Proconco company in

2007

5 Note: Proconco’s price (2008) (VND/kg, VAT is not included)

3 BACKGROUND INFORMATION ON FEED ENTERPRISES IN VIETNAM

3.1 General characteristics

Figure 5 shows the size classification of the surveyed feed mills in the north and south of Vietnam The size classification used here and throughout the report, unless otherwise stated is: large size is production greater or equal to 60,000 tonnes/year (n = 7), medium

is those producing between 10,000 and less than 60,000 tonnes/year (n = 20), and those producing less than 10,000 tonnes/year (n = 17) are classified as small enterprises In general, the feed production scale tends to be larger in the south than that in the north Half of the total mills in the north are small scale compared to around 27% in the south Meanwhile the share of medium and large enterprises is both higher in the south Around 45% of southern mills are medium-sized ones compared to 27% in the north

Small Medium Large

Figure 5 Distribution of feed enterprises by production scale in north and south Vietnam (%)

Table 6 shows the difference between actual output and design output in 2007 by size and region of surveyed feed mills as well as percent utilization of design output By region, the actual average output of small and medium mills is lower in the north compared to that in the north There are more small mills in the north than in the south (see Figure 5) Large mills in the north have relatively larger actual production than those in the south (78,429 tonnes versus 66,225 tonnes) However, overall, the average feed output of northern mills seems to be lower than southern ones (18,697 tonnes compared to 26,092 tonnes)

There are some similarities in the use of design capacity of mills at all size groups In both regions, enterprises by all size groups produce a little bit less than their design capacity, except for the case of medium mills in the north (22,400 tonnes of actual output versus 22,293 tonnes of designed output) A bigger gap is seen in large enterprises located

in the north where their actual output reached less than 80,000 tonnes while their designed

output is 110,000 tonnes Notably, the designed output is very much larger in the northern

large mills than in the southern large ones

The average capacity utilization is fully maximised in the northern medium mills while it

seems to be lowest in the large group (71.3%) In the south, the utilization rate increases

by production scale, ranging from 89.4% to 98.1%

Table 6 Actual and designed output of feed mills, and percent utilization of designed

output in 2007, by scale and region

Percent utilization (%)

Actual output (tonnes)

Designed output (tonnes)

Percent utilization (%)

Revenue from different activities of feed mills is very diverse among the three mill

groups (Table 7) Large mills depend completely on feed production activities while small

ones tend to diversify their business by being an agent or premix producer with the share

of revenue for these activities being 7.5% and 2.5% respectively Regarding production

activities only, production of concentrate feed is relatively more important for small mills

compared to medium and large mills, with less concentration on complete feed The

percentage of revenue from complete feed is over 50% for small mills while it is over

80% for both medium and large ones In terms of magnitude of revenue, large obtain 7

times and medium, 5 times more revenue for concentrate and complete feed than small

mills

Table 7 Average revenue from activities of feed enterprises in 2007

3.2 Labor use and wages

Figure 6 shows the distribution of labor numbers in domestic and foreign-owned

feedmills6 Around two thirds of total domestic enterprises have less than 100 employees

6 Foreign-owned mills include joint-ventures

while this number is less than 39% for foreign enterprises Meanwhile, about one quarter

of total foreign mills have above 300 employees7 compared to only 8.9% of domestic ones An analysis of labour productivity is presented in section 4.2

Less than 100 100-300 employees More than 300 employees

Figure 6 Labor size of feed enterprises by foreign and domestic ownership

Figure 6 gives some indication of enterprise scale for the whole sample of 62 mills, Employee numbers can be used as a proxy indication for scale (see footnote 7): with small enterprises having less than 100 employees, medium enterprises having 100 – 300 employees, and large enterprises having more than 300 employees Using these criteria, then for the whole sample of 62 mills, 58.7% are small enterprises, 28.6% are medium enterprises, and 12.7% are large enterprises This compares with our classification of 44 mills which gave 38.6% being small enterprises (producing less than 10,000 tonnes/year), 45.5% being medium enterprises (producing between 10,000 and 60,000 tonnes/year), and 15.9% being large enterprises (producing more than 60,000 tonnes/year) The classification of large firms is roughly similar; our classification of medium and small firms in this exercise was chosen to highlight the very small mills which made up half the small-medium mills in the survey

Figure 7 shows the number of employees for the firms in our categories of small, medium and large Only 14% of the large mills have more than 300 employees, perhaps reflecting the capital intensive nature of the production However, a high percentage of both the small and medium categories have less than 100 employees: 88% and 70% respectively

7 Vietnam classifies SMEs as those firms with less than 300 employees In the following we suggest that medium scale is more than 100 employees

87.5 70

12.5 30.0

Less than 100 100-300 employees More than 300 employees

Figure 7 Labor size of feed enterprises by production scale

Figure 8 shows the average salary that was paid to labourers in small, medium and large enterprises Office staff usually receive a higher salary than factory floor workers and the difference is considerable, especially in large mills Considering mills by production scale, in large mills, the salary of office staff is nearly double and that for workers is 1.5 times higher than that in small mills Generally, family laborers tend to have a lower salary than hired labor

Total Family Hired Total Family Hired Total Family Hired

Admin/Technic/Marketing Factory workers

Figure 8 Average salary for labor of feed enterprises by production scale

3.3 Infrastructure

Infrastructure of feed mills can be mainly listed as land, storage facilities, processing equipment The average land owned or leased by feed mills in general slightly larger for the northern mills than for the southern ones (Table 8) Also, nearly 95% of total land is being used by northern mills compared to only 82.7% by the southern mills

By scale of production, land area normally tends to rise with production level The share

of land being used by feed mills is also likely to increase with production scale This is true for mills in both regions, except for large mills located in the south that only use 59.5% of their total land for their business compared to 100% for large mills in the north

Table 8 Land area owned/leased and percentage of land being used by the feed enterprises, by scale and region

Of the large mills in the south that rent land, only one third are located in rural areas, while this proportion in the same group in the north is 75% It means that those large mills in the south have some certain disadvantages compared to northern ones in terms of rent cost, as rents are likely to be higher in urban areas Small and medium mills in the two regions, that are less likely to rent land for their business, often tend to locate more in the rural area rather than in the urban area, especially those in medium group (100% in the north and 61.5% in the south) This indicates that small and medium mills may have lower rental costs compared to large ones In addition, small enterprises may have some certain advantages in terms of locating near input supply areas and customers while they may face some difficulties in terms of infrastructure such as road network, information transfer, and electricity supply

8 Vietnamese SMEs often use their own land for their business and seem to encounter difficulties when renting land For those who rent land, there are often obstacles in clearing spaces and receiving compensation money (Le Xuan Ba et al 2007)

Table 9 Distribution of enterprises owning or renting land and location (percent)

North (n=27) South (n=35)

Location Land ownership Location Land ownership

Overall 18.2 81.8 54.6 45.5 45.5 54.5 33.3 66.7

Storage facilities were classified as either being open or closed facilities9 Only a

percentage of small and medium mills in the south report having open storage facilities

(Table 10) All northern mills report having only closed storage facilities, and 100% of all

mills, except for medium-sized mills in the south, report having closed storage facilities

The percentages of firms having chilled storage facilities was low with 11% of firms in

the north and 34% of firms in the south reporting having these

Table 10 Percentage of enterprises having storage equipment, by scale and region

In comparison to North Vietnam, enterprises in the south seem to have higher storage

capacity: combined open and closed storage being an average of 13,357 tonnes in the

south compared to 6,938 tonnes in the north (Table 11) However, small mills in the

south have a lower storage capacity then small mills in the north As would be expected,

the capacity of storage facilities tends to rise with production size

Table 11 Average capacity (tonnes) of storage facilities

Table 12 shows some indicators regarding storage of the two main raw material inputs

which are maize (energy-rich input) and soybean cake (protein-rich input) by region and

9 Open storage facilities include sheds without walls; closed storage facilities comprise sheds with walls,

warehouses, chilled storage, houses and silos

production scale The average storage period tends to be shorter in the north compared to the south, both for maize and soybean cake This may be related to the fact that storage capacity is larger in the south than the north, except for small mills (Table 11) The number of storage days is also larger for medium and larger enterprises

The average stored volume of raw materials is smaller in small and medium mills compared to large mills and is also lower in the south than in the north (which can also explain the longer storage period in the south) It is noticeable that large mills were storing a greater volume of soybean cake relative to maize (10,150 tonnes soybean cake versus 4,333 tonnes maize), whereas small and medium mills were storing relatively similar quantities of maize and soybean cake This is probably a purchasing strategy for imported soybean cake able to be used by large mills with greater storage capacity

Reliance on imports was calculated as the share of input quantity from import sources (both direct and indirect) of the total quantity purchased (Table 12) Generally, mills depend heavily on imports for soybean cake (protein input) rather than maize (energy input) which is often locally grown However, the import reliance for maize is relatively higher in northern mills compared to those in the south While small and medium mills source most of their maize from domestic sources, large mills import over a quarter of their total maize We examined the source of maize inputs by class and found that there was a significant difference between firms in the reliance on imports For soybean cake, medium and large mills mainly purchase imported sources, whereas only around 70% of total purchases by small mills are imported

Table 12 Period, volume and place of material storage of feed enterprises by region

and scale

Type of

input Indicator

North (n=27)

South (n=35)

Small (n=17)

Med (n=20)

Large (n=7)

2 Closed structure 100.0 72.7 66.7 91.7 100.0

Maize

With regards to processing equipment, more northern mills have automatic production systems than those in the south (25% versus 7.4%) Noticeably, all large and small mills

in the south use semi-automatic processing equipment and this was also the case for northern medium-sized mills (see Appendix, Table A1) Southern enterprises tend to use overseas technology more than those in the north (77.8% versus 59.3%) However, all northern large mills use foreign technology while only 37.5% of southern mills with same

size have this type of technology10 (see Appendix, Table A2)

10 This result may be due to data problems

4 COST STRUCTURE AND RAW MATERIAL USE

4.1 Cost structure of feed enterprises

Production costs are shown in Table 13 to be highest for the smallest firms, and this is significantly different at the 5% level Raw material costs are highest for the smallest firms although the large variance in costs means that there is no difference in mean feed input costs between small and medium firms Raw material costs are significantly cheaper for the large group at the 10% level, possibly because of their more efficient technology Costs of labour appear to be higher for smaller firms but the means are not significantly different according to our analysis of variance Similarly, per tonne costs of electricity, repairs, quality control, rent and management are not significantly different at the 10% level Although credit costs appear to be much larger for small firms, this mean was largely due to one firm that had very high credit costs, and the analysis of variance showed no significant difference in means at the 10% level Other costs, which include marketing and taxes, were significantly lower for the largest firms

Per kilogram of output expenditure on inputs other than raw materials is shown in the third row of the table Small enterprises have costs of 2050 VND per kg of output, whereas large firms have costs of only 970 VND per kg of output These cost differences are significant at the 11% level Non-raw material costs of production were not significantly different between medium and large scale firms

Table 13 Cost of production (‘000 VND per kg output) and share of total cost %

Small (n=13) Med (n=20) Large (n=7)

Total Cost 8.42 6.34 5.38

s>m>l (5%)

Raw material costs 6.37 76% 5.05 80% 4.41 82%

s,m>l (10%) Non-raw material costs 2.05 24% 1.29 20% 0.97 18%

s>m,l (11%)

in the large one Labour costs range from 5% to 6% of total costs but there are no significant differences between firms Similarly, none of the other non-raw material costs are significantly different, except for the ‘other cost’ category Other costs, which include transportation, bags and taxes, are lower for large firms than for the smaller groups Small firms spend as much on credit as on labour (5%) but this high cost is not significantly

different from the larger scale producers, probably because of high variance in credit costs for the small group

4.2 Procurement of raw materials

The results in this section compare prices paid for raw materials from different sources, payment methods for raw materials, patterns of purchase, and storage of main raw materials for mills of different production scale

4.2.1 Prices paid for raw materials and percentage of raw material from

different sources

Raw material purchase costs per tonne of output for large, medium and small enterprises

by ownership type are shown in Figure 9 Generally, input costs per unit of output tend to decrease as production scale increases, for both foreign and domestic firms

However, foreign firms pay more than domestic firms for their raw materials per unit of output More specifically, for medium scale enterprises, the cost of raw material purchases paid by domestic companies is only 4 million VND per tonne of output while the cost for foreign companies is nearly 6 million VND/tonne of output This trend was also similar for large enterprises when comparing foreign and domestic firms

Dom es tic Foreign

Figure 9 Raw material purchase costs per ton of output, by ownership type and production scale

There are many types of raw materials which are used to produce animal feed but the share of inputs depends on whether complete or concentrated feed is made It should be noted that these inputs could also vary between mill size groups as a result of their different product outputs (e.g the ratio of concentrate to complete feed – see Table 6) Figure 10 shows the composition of energy ingredients (maize, cassava, bran, etc.) as the share of total energy inputs used The share of maize tends to increase with the scale of production (19.5% for small scale mills, 32.4% for medium ones and 39.5% for large ones) Maize also accounts for the largest share of energy inputs used by large mills, followed by bran (34.6%) On the contrary, for cassava, it seems to be lower in the large

group compared to the other two groups (20.9% versus 27.7% and 24.1%) Bran (from rice, maize, wheat) is used the most in the small group with 39.7% out of the total energy inputs while it is used least in the medium scale mill class (27.4%) Other inputs such as broken rice are mixed with other main ingredients at a higher share in the small and medium enterprises when compared to the large enterprises

The composition of protein rich ingredients is shown in Figure 11 Soybean cake is the most commonly used protein ingredient, and is more likely to be used by large mills Groundnut cake is used by small and medium mills (around 10%) but not by large mills The proportion of fish meal and meat meal tends to decrease by production scale, ranging between 3.9% and 9.9% for the former and 4.4% and 10.6% for the latter

19.5

32.4

39.5 24.1

Figure 10 Composition of energy rich ingredients as percentage of total energy inputs used

9.3

4.4 19.7

Figure 11 Composition of protein rich ingredients as a percentage of total protein inputs used

Feed inputs vary between enterprise classes when comparing the substitution level

between energy inputs and protein inputs While maize is the largest input for feed

production in the large and medium classes, bran is the predominant input used by the

small enterprises For protein materials, soybean cake is always the most important

ingredient in all classes, although the amount used as a percentage of total protein inputs

does vary between classes These results indicate a higher substitution level of energy

inputs (which are more likely to be sourced domestically) than protein inputs (which are

more likely to be imported)

Because input costs depend on the type of output, with concentrate costing more to

produce because of its higher protein content, it is difficult to draw a conclusion regarding

inefficiency from the average cost of raw material input referred to in Table 13 In Table

14 we look in detail at the nature of raw material purchases, in terms of prices paid for the

same input, sources of supply, and whether these are influenced by the scale of firm or

location The number of respondents who provided details on the price paid for inputs by

source was small (Table 14) Results indicated that there were no statistical differences

between the prices paid for any of the raw material inputs by scale of firm (Table 14a)

The statistical analysis may be affected by the small sample size, but even the magnitude

of the mean reported prices was small Analyses of prices paid by location also showed

that there was no significant difference between North and South Vietnam (Table 14b)

Generally, prices for energy rich inputs were around 3500 VND/kg, although cassava was

considerably cheaper The price of high protein inputs, soybean cake and fishmeal, were

at least double the price of energy ingredients The prices paid for imported versus

domestic raw materials was also analysed at the aggregate level, with both sources of

imported product pooled into one category (Table 14c) There was no significant

difference in prices paid for maize between imported and domestic sources The mean

price of soybean cake on the domestic market appeared to be different although the

analysis of variance showed no difference, probably due to the very large variance in the

price of imported product which possibly reflects quality variation Similarly, the mean

price of imported fishmeal appeared to be higher than for domestic sources but this

difference was not statistically significant

Table 14 Comparison of raw material input prices by scale, source and location

a Comparison of prices paid by scale and source for the main products and sources:

Maize: Domestic 3,750 (647) 3,942 (667) 4,050 (386) nsd 6 12 7

Maize: All sources 3,750 (647) 3,941 (667) 3,998 (380) nsd 6 12 8

Rice: Domestic 3,750 (1156) 3,636 (447) 4,128 (667) nsd 4 5 2

Cassava: Domestic 2,450 (659) 2,753 (482) 2,472 (664) nsd 7 9 6

Wheat bran: All sources 3,503 (831) 3,687 (1241) 3,917 (1152) nsd 8 4 9

Rice bran: All sources 3,309 (784) 3,166 (411) 3,161 (879) nsd 6 7 4

Soybean cake: All

sources 6,233 (1176) 6,901 (2292) 7,190 (2278) nsd 10 13 9

b Comparison of prices paid by location and source, main products and sources:

Maize: Domestic 4,100 (283) 3,828 (688) nsd 9 16

Cassava: Domestic 2,600 (418) 2,574 (634) nsd 5 17

Soybean cake: Avg price 6,415 (1,057) 7,052 (2,473) nsd 14 18

Fishmeal: Avg price 11,250 (957) 10,697 (3,364) nsd 4 16

c Comparison of prices paid by source, main imported products:

Table 15 shows the percentage of raw material inputs purchased from domestic and

imported sources Energy-rich inputs such as maize are mostly bought from local sources

however large mills are less likely to buy from these sources compared to the two other

smaller groups Locally produced bran is also sourced by all production scale groups, less

so by large mills, despite the price of local bran being higher than imported bran Soybean

cake, on the contrary, is mainly imported directly by mills (100% for large mills and

58.2% for medium ones) or from an imported domestic market source (71.4% for small

a “Domestic imported” means material purchased from a domestic supplier who has imported the material

It is known that the supply of protein rich materials locally produced in Vietnam is

insufficient to meet demand, especially for soybean cake Small and medium mills both

source some soybean cake produced in Vietnam: 28.5% and 4.6% respectively Fish meal

is the only main protein rich ingredient that comes largely from domestic sources

Secondary data from the Department of Livestock Production suggested that about two

thirds of fishmeal used for feed production comes from local sources

The analysis of raw material input prices indicated that there was little difference in the

price paid by different classes of firms This means that there must be another reason why

the average raw material input cost is higher for small firms Unless there is significant

wastage of materials on the part of small firms, the most likely explanation is that small

firms produce more concentrates and therefore have to pay more for inputs The

composition of output and the profitability by scale is discussed in a later section

Figure 12 shows the efficiency of using raw material and labour to produce 1 tonne of animal feed for the different size enterprise groups The measurement of unit labour was shown to be statistically different between small and medium scale, with a smaller amount of labour needed in the medium group, 11 labourers to produce 1 tonne of output, compared to around 19 labourers for small enterprises This may reflect under utilisation

of labour on the part of small firms due to the scale of operation, or it may also reflect a greater reliance on more automated processes (that is, capital substitution) by large firms The measure of input quantity versus output quantity was found to be generally close to one indicating that the data is consistent within the different parts of the survey (Figure 12) Variation about one may reflect addition to, or draw down from storage However, results indicated that there were no statistical differences between the input quantity used

Medium (n=20)

Large (n=7)

tonne

0 4 8 12 16

20

labour

Input/1 tonne ouput labour/1 tonne ouput

Figure 12 Inputs used to produce one tonne of feed output by production scale

4.2.2 Providers of raw materials

In this study it was found that feed inputs in general were commonly provided by private processing businesses, followed by traders (Table 16) Farmers and traders seemed to have some role in providing materials to small and medium mills, but not to large ones State-owned companies were found to be the only provider of bran to large feed

enterprises

Considering each category of material input:

• Maize was mainly bought from private processing business who supplied about 80.5% of the grain to large mills, and over 50% to other mill groups Traders and farmers also provided maize to mills in the small and medium groups, but not to large mills

• Cassava: mills in the small and medium groups depended on traders and farmers as their main suppliers of cassava, while large firms bought 100% of their cassava requirements from private processing businesses

Table 16 Percentage of raw material inputs purchased from different suppliers for

each production scale of feed enterprises

Input

type Farmer Trader

Private processing business

State owned enterprise Other

• Soybean cake: Similar to maize, soybean cake was bought from three different sources (farmers, traders and private processing businesses), but private processing businesses dominated the market

In conclusion, the role of traders and farmers as direct suppliers to small and medium mills may be important in terms of supporting employment of farmers and traders This finding presents a significant difference in the supply chain between large and small/medium mills

4.2.3 Payment method for input purchase

Feed enterprises tended to indicate “pay at purchase” as their main method of payment for raw materials, but the proportion decreased from small scale to larger scale enterprises: 73%, 55% and 50% respectively (Figure 13) “Pay on credit” was selected as the second most common payment method ranging from 25% to over 30% The lower proportion of

“pay at purchase” and higher “pay on credit” by larger scale enterprises may be explained

by the fact that larger mills often have a closer relationship with suppliers, and higher prestige and possibly credit worthiness, compared to smaller enterprises

Only medium-scale firms purchased inputs using “pay in advance”: 15% of input purchases were paid in advance by medium-sized mills This method helps ensure a stable supply, especially when raw materials are scarce

Pay on credit

Other

Figure 13 Payment methods used for raw material purchases, by production scale

Figure 14 shows the percentage of firms having to pay VAT when purchasing raw materials The percentage of firms paying VAT increased with increasing production scale, for both energy and protein inputs, but less so for energy inputs Only around 70%

of small firms and 90% of medium firms paid VAT on energy inputs However, for both small and medium enterprises, the proportion of firms paying VAT on purchased protein inputs was generally higher than when buying energy inputs This is because the sources

of the former mainly come from outside Vietnam rather than locally produced sources All large mills have VAT receipt for 100% of their inputs, as their providers are often state enterprises or private processing businesses rather than farmers or traders

5 FACTORY OUTPUT

5.1 Output types

Pig feed was the most commonly produced feed type, being produced by 88.2% of total interviewed firms in the small group, and all mills in the medium and large groups (Table 17) The next most commonly produced product was chicken feed, followed by cattle feed, with a higher percentage of larger enterprises producing these two feed types (Table

17 Noticeably, there were a considerable percentage of firms producing other feed products such as premix, especially those mills in the large group where 71.4% of them were involved in this activity The proportion of mills producing fish feed was quite modest, but more medium mills were involved than large ones (25% versus 14.3%) Finally, it can be generally said that the larger the size of the enterprise, the more likely they are to be involved in producing a variety of feed products

Table 17 Percentage of firms producing each type of animal feed, by production

scale and region

Scale Region Pig Chicken Cattle Fish Other

When feed is separated into complete and concentrated feed, regardless of specific type of livestock, all large mills were involved in producing both complete and concentrate (Figure 15) In this study, all firms in the medium group were engaged in the manufacture

of complete feed and almost all (95%) produced concentrate feed (Figure 15) A lower share of small enterprises was making complete and concentrate feeds which is consistent with data presented in Table 7 which showed that small firms tend to diversify their business beyond production activities However more small enterprises produce concentrate feed than produce complete feed (82.4% versus 64.7%) This helps explain why raw material input costs are higher for small firms

Complete feed Concentrate feed

Figure 15 Percentage of firms producing complete and concentrate feed

When feed production is separated into complete and concentrate products in a more detailed way, generally there was a higher percentage of larger mills producing each type

of feed than for the small group (Table 18) While all large mills and almost medium ones were involved in producing feed for pigs, only around 59% and 82% of mills in the small group produced pig complete feed and pig concentrate feed respectively An even bigger production difference was seen in the case of chicken complete feed, with more or less 80% of medium and larger mills producing this output, compared to around 18% of mills

in the small mill class The pattern of cattle feed production was similar, with more large scale mills producing this product Small mills are only engaged in producing complete feed for cattle rather than concentrate feed There were no mills in the small and large groups producing concentrate fish feed, and also a relatively small share of these mills were involved in producing complete fish feed, compared to those in the medium group Data on feed prices are summarized in the Appendix(Table A3) Pig and chicken feeds were the most commonly produced items and therefore we have enough price data to analyze for differences between firm types These are reported in Table 19 In comparing between feed types at the aggregate level we found, as expected, that complete feed prices were significantly lower than concentrate prices, and this was significant at the 1% level There was no significant differences in the price of complete feed between species (comparing pig and chicken), nor for concentrate The premium on concentrate feed was higher for chicken, at 3250 VND per kg compared to 2680 for pig Comparing price data for pig complete feed, we found that prices were significantly higher in the South, and the price difference was 920 VND per kg, or 19% of the northern price The price charged by small firms was significantly lower than the price charged by medium scale firms, although the price by the large scale firms was not significantly different from either of these other groups The mean price discount for small firms over medium firm prices was

1400 VND per kg

Table 18 Percentage of firms producing complete and concentrate feed for livestock

Table 19 ANOVA analysis of pig and chicken feed prices (‘000VND per kg): mean,

standard deviation and price difference by region and production scale

Note: Group a significantly lower than b at 1% level; group c significantly lower than group d at

10% level; group e significantly different from group f at 10% level, n group not significantly

different from others

in terms of price discounts Because of product heterogeneity, and because the only cost

of production data we have is at the firm level, the only measure of efficiency we have is whole of operation profit These are compared in Table 20 The mean profit of the 15 small firms was -0.07 million VND That is, at the mean level they are making a loss, although variance was very high, with a coefficient of variation of 4, suggesting some did very poorly and others made a profit Despite the high variance in profits for the smallest group, the statistical analysis indicates that the profit of medium sized firms was significantly larger Whilst mean profit of the large firms was found to be similar to the medium firms in terms of means, variance, and is not statistically different; the test of means between small and large firms was only significant at the 23% level, due to the small sample from large firms

Table 20 Profit (million VND) by production scale

Mean Standard deviation N

Whilst the conclusion is that small firms have lower profits than their larger competitors,

it is useful to analyze how the profitability amongst small firms is affected by their product mix As was noted previously, high costs of production may be due to the greater emphasis on concentrates In Table 21, costs and profits per kg of output are shown for 3 groups of firms, classified by the share of concentrate in their total output Thirty one firms had less than 20% of total output as concentrate, 6 firms produced between 20% and 40% of output as concentrate, and 5 firms were focused largely on concentrate production with more than 80% of output as concentrate Analysis of variance on the data showed that the firms specializing in concentrate production had significantly higher costs

of production per kg than the other 2 groups (which were not statistically different from each other) However, there was no significant difference in the calculated profit per kg of output