Determinants of firm exit in Vietnam

Using panel data from 10 provinces and cities in Vietnam and applying the logistic regression method, this study finds that total asset and leverage have positive impacts on firm exit wh

TRAN THI LAM

MASTER OF ARTS IN DEVELOPMENT ECONOMICS

HO CHI MINH CITY, MAY 2017

VIETNAM – THE NETHERLANDS PROGRAMME FOR M.A IN DEVELOPMENT ECONOMICS

DETERMINANTS OF FIRM EXIT IN VIETNAM

A thesis submitted in partial fulfilment of requirements for the degree of MASTER OF ARTS IN DEVELOPMENT ECONOMICS

By

TRAN THI LAM

Academic Supervisor:

DR TRUONG DANG THUY

HO CHI MINH CITY, MAY 2017

TABLE OF CONTENTS

ACKNOWLEDGEMENT 1

ABSTRACT 2

LIST OF FIGURE AND TABLES 3

CHAPTER 1: INTRODUCTION 4

1.1 Problem Statement 4

1.2 Research objectives 6

1.3 Research questions 6

1.4 Scope of research 7

CHAPTER 2: LITERATURE REVIEW 8

2.1 Literature Review 8

2.1.1 Firm exit behavior 8

2.1.2 Determinants of the decision to exit 9

2.2 Review of empirical studies on firm survival/exit 17

2.2.1 Approaches of analyzing firm exit and survival 17

2.2.2 Emprical analyses of firm survival 18

2.2.3 Emprical analyses of firm Exit 20

CHAPTER 3: RESEARCH METHODOLOGY 23

3.1 Conceptual framework and the econometric model 23

3.1.1 Conceptual framework 23

3.1.2 Theoretical reviews 24

3.1.3 The econometric model 25

3.2 Data and variables 27

CHAPTER 4: RESEARCH RESULT 29

4.1 Descriptive Statistic 31

4.2 Regression results 34

4.2.1 Bivariate analysis 34

4.2.2 Multivariate analysis 36

4.2.3 Multicollinearty analysis 37

4.2.4 Results of random-effect logistic regressions……… …38

4.2.5 Marginal effects 41

CHAPTER 5: CONCLUSION……… 43

5.1 Conclusion 43

5.2 Policy Implications 44

5.3 Thesis limitations and suggestion for further researches 44

REPERENCE 46

APPENDICES 56

Firstly, I would like to express my sincere gratitude to my advisor Dr TRUONG DANG THUY for his enthusiastic assistance He has not only several insightful comments based on his immense knowledge helps me to solve all my problems, but also encourages me to finish my thesis

I would like to express my thanks to my friends in MDE class 19, and 20, especially Duy Chinh (class 19), Do Luat (class 20), Duy Lap (class 20), Nguyen Thai Duong (class 20) and one special friend who have given their limited time to help me solved the difficulties in the process

I also would like to send love to my family and my close friends for always being beside

me, spiritually encouraging me and letting me know that I am not alone in all difficult situations

Finally, my special thanks also to my husband and my baby who help me to have a strong motivation to finish my thesis

This paper examines the determinants of firm exit in Vietnam using SME data from 2005 to

2011 Using panel data from 10 provinces and cities in Vietnam and applying the logistic regression method, this study finds that total asset and leverage have positive impacts on firm exit while the size, age, investment and total gross profit negatively affect firm exit

Keywords: Small and medium enterprise, total asset, firm exit, firm size, firm age, debt leverage,

investment, total gross profit, random effects logistic regression

LIST OF FIGURE AND TABLES

Figure 1.1: Number of registered enterprises and stop business over period 2007-2015

Figure 3.1: Conceptual framework

Table 3.1: List of surveyed province/city

Table 3.2: Descriptive variables

Table 4.1: Descriptive Statistic

Table 4.2: The number of survey exit firms in each city/province

Table 4.3: Overview of firm exit in Viet Nam

Table 4.4: Firm exit in province/city

Table 4.5 A comparison between Firm non-exit and Firm exit in term of variables

Table 4.6: Covariance matrix

Table 4.7: VIF index

Table 4.8 Results of three Random – effect logistic regressions

Table 4.9 Marginal effect results

CHAPTER 1: INTRODUCTION

1.1 Problem Statement

The stable development and growth of firms are key factors what influence directly the socioeconomic development On the other hand, firm’s activities affect almost all respects of economy and society such as rate of unemployment, national budget, activity of trade, and other macroeconomic indicators

The role of small and medium enterprises (SMEs) in economic growth has been recognized through many studies in the world recently An article by Joshua and Quartey (2010) describes that SMEs play important roles to an economic development as creating efficient and productive jobs, the seeds of large corporations and the fuel maintaining the national machine In advanced economies, the number of firms in the SME sector, which has substantial labors, is larger than the multinational one (Mullineux, 1997) In addition, Feeney and Riding (1997) reveal that governments in most countries have conducted several policies to encourage the development of SMEs Whereas, the growth of SMEs promotes the process of redistribution of both inter and intra-regional division of the firm and they also become a countervailing force again the influence of large-scale corporations There is approximately 91% of the enterprises is Small, Medium and

Micro Enterprises (SMMEs) in South Africa (Hassbroeck 1996, Berry et al 2002) They create

about 61% jobs for labor source and also represent for 52%- 57% of GDP (CSS 1998, Ntsika 1999,

Gumede 2000, Berry et al 2002) In additional, Small and medium enterprises also account over

90% of the private business sector and play a crucial role in contributing GDP in most African countries (UNIDO 1999)

Viet Nam has changed from a centralized planned economy to a socialist-oriented market economy after “Đổi Mới” reform period since mid-1980 Undergoing a period of formation and development, the Vietnam economy continues to grow and get many substantial achievements A private ownership sector contributes a vital role for Vietnam economic growth According to Vietnam General Statistics Office, Vietnam attains around 7% GDP growth over the period 2000 to

2005 and continues to grow at 7.01 percent from 2005 to 2010 The process of developing keeps in

stable until now, especially SMEs sector Small and medium enterprises (SMEs) have an essential role to play in motivating growth, generating jobs and contributing to poverty reduction The

contribution of SMEs is a major tax resource for Vietnam budget annually Furthermore, according

to the report of the Vietnam Chamber of Commerce and Industry (VCCI), we have 543,963 enterprises in 2011, about nearly 513,000 enterprises in 2015, but there is nearly 97 percent of small and medium firm, mainly private businesses The important role of SME is increasing thanks to not only contributing significantly to the Gross domestic product (GDP), reducing poverty, enhancing security society, but also creating more than one million new jobs per year

According to Ministry of Science and Technology, there are 692 thousand enterprises registered business in the period 2007 – 2015 However, there are too many enterprises not enough power to survive in globalized market and international integration and move out of market every year According to General Statistics Office of Vietnam, more and more firms exit and stop activity

in annual reports (2015: 80,900 firms; 2014: 67,800 firms; 2013: 70,500 firms and 2012: 63,500 firms) Figure 1.1 shows that the number of firms that ceased operations is increasing in the recent years

Figure 1.1: Number of registered enterprises and stop business over period 2007-2015

(Source: Administrative Department of Business Registration - Ministry of Science and Technology)

Therefore, this is a crucial problem requires a reaction from government official and policy makers to decrease the number of firm stopping business every year

In recently, the determinant impacts the ratio of birth - death enterprises and issue of firm survival in the developed countries have been attracting a lot of attention of scholars around the world (Parker 2004, Strotmann 2006) In addition, the issue about entrepreneurial activities, which has mentioned in empirical literatures recently, is a crucial factor affects to speed of economic progress (Stel et al 2005) Some empirical studies focus on relationship between characteristics of firm, policy, an environment, the process of international trade liberalization and rate of firm exit Most of studies focus on the topic of firm exit is conducted at the firm or industry level (Hannan and Carroll 1992, Sarkar et al 2006) They study on effect of the determinants on exit probability of small and medium enterprises (SMEs) Many other empirical studies also reveal that the perseverance of small firms has an important role in economic development in a context of increasing import competition from low-cost nations (Colantone et al 2014)

However, in both the theoretical and practical sides, most of the articles are empirically carried out on foreign firms in developed countries and studies about survival of enterprise are less mention in developing countries (Parker 2004) There are a few researches studied directly on exit issue in developing countries, especially in Viet Nam

Thus, from the obtained results of this research will contribute to understand deeply about the determinants influence on firm exit in Viet Nam as practical part They can make policies priority and influence debates on firm’s activity to reduce firm exit

1.2 Research objectives

The main objective of this study is to identify factors influence on the decision of firm exit Using the SME data from 2005 to 2011, this study applies a logit model with the decision to exit as the dependent variable and total asset, firm age, total gross profit, firm size, investment and debt leverage as explanatory variables

1.3 Research questions

This research examines how the determinants affect the exit probability of Small and Medium Enterprise in Vietnam To identify and understand the dimension clearly, the research is developed based on following questions:

 What factors influence probability of firm exit?

1.4 Scope of research

The study will examine the relationship between firm exit and related determinants using the panel data of 2005-2011 This is a full-scale survey covering small and medium-scale enterprises in Viet Nam (SMEs) in the rural and urban area of Vietnam and conducted by collaboration between the Institute of Labour Studies and Social Affairs (ILSSA) in the Ministry of Labour, Invalids and Social Affairs (MOLISA) and Department of Economics, University of Copenhagen with funding from DANIDA in the period from 2005 to 2011 To collect data by choosing the key questions in the main questionnaire for every surveyed year, we have the panel data to run the logistic model

CHAPTER 2: LITERATURE REVIEW

2.1 Literature Review

2.1.1 Firm exit behavior

Understanding increasingly about exit behavior and reasons for exit out market help managers making a comfortable decision for their survival and increasing their evaluation efficiency with insights into the reasons for the failure of a product (Witteloostuijn 1998, Sheppard

1994, Dixit et al 2007) In addition, thanks to the exit process, the manager can learn about their actual essential productivity levels (Yang and Temple 2012)

Because of the significant importance of exit decision, we are necessary to consider the reasons why firm decided to exit which help me understanding about the decision and its possible drivers forced firm to exit

The firm decides to either survive or decline and exit depends in many respects and their evaluation and expectation According to Jovanovic (1982), firm’s productivity is a key determinant which manager bases on to decide whether to start up a new company or exit the market Harada (2007) produces the main reasons for exit of small firm in Japan Among these reasons, despairing perception for further business accounts for 38%, followed by aging of the manager (20%) and the injury or sickness of the manager (15%) He presents more detail about despairing perception of further business which indicated through the reduction of sales (71%) and having a deficit (only 9%) Therefore, the diminution in sale plays an important role He supposes that exit also sometimes happens because owner desires to have an easier life or to take a position doing work or

to begin a new business According to several researchers, each company acts basing on the aim of profit maximization, and whenever the profit (or realized/expected profit) declines below some threshold, they would exit the market (Das and Das 1996, Klepper 1996, Frank 1988, Ghemawat and Nalebuff 1985, Jovanovic 1982) Investigating a set of data from Botswana, Kenya, Malawi, Swaziland, and Zimbabwe, paper was researched by Liedholm et al (1994, 1178) indicates that the reasons of company exit are not only business reasons but also related to personal reasons He finds that only about one-half of rural firms in these countries shut their business because of business failures and approximately one-quarter firms do it due to some individual reasons (e.g retirement or poor health) while the rest decides to close owing to the availability of better business opportunities

or obligation imposed by the government Other result was proposed by Hopenhayan (1992) shows

that firm decides to exit when it experiences series of unfavorable productivity shocks The companies having a least productive are easy to exit because of lacking competition and not enough capacity to remain their business, while the most efficient ones try to expand their activities to increase sale and profit

On the other hand, Colantone and Sleuwaegen (2010) suppose that each entrepreneur will evaluates the basis of risk-reward and assess the outside choices regarding to scrap value of companies and market wages to have decision about staying in or leaving out the market Following Baggs (2005), the firm’s prospects for profits, which is affected not only by the characteristics of the firm itself but also the characteristics of the business environment, is a key determinant to remain or exit in an industry Frazer (2005) shows other reasons made the enterprises closed as personal reason or because of the closure order from the government

According to several suppositions, incumbent enterprises will consider to continue operation

or exit belong to their health In general, they will be less likely to exit when performance is stronger and more likely to stop operations when prior performance is poor (McCann and Vroom 2010) In the view of Yang and Temple (2012), the firms consider about moving out the market when they expect sufficiently future operating profits and exit will be their least costly option Many surveys on enterprise survival report that the probability to survive of efficient firms is higher than the probability of inefficient ones (Baggs 2005) According to Kovenock et al (1997), the exit

is a decision made by the owner or manager of firms regarding the use of inherited arrangement of capital and he separates exit into three types, including strategic reallocation, restructuring, and corporate form and structure

2.1.2 Determinants of the decision to exit

There are considerable a number of studies investigate the firm exit and survival, not only in developed countries but also in developing countries recently (Roberts et al 1996a, Das et al 1997, Frazer 2005) Determinants influence firm on exit and survival and the duration of one survival in developed countries have been interesting numerous researchers and scholars in recent years The exit issue is also concerned with a common part of the landscape of business (Geroski 1995)

Recently, the crucial determinants which effect on the ratio of business dissolution are being examined more and more in body of literature which used in modeling to express structural motivation, barriers to exit and firm features According to Baggs (2005), firm’s profit prospect is a

fundamental factor affecting the company’s decision whether exit or remain in an industry Both the characteristics of competitive industry or environment and the characteristics belong to the company influence on this prospect All most papers regularly use explanatory variables on their researches that have been used in the literature are expressed below:

FIRM SIZE

In the relevant literature, there are many approaches to find the factors affecting firm exit or firm survival Building on the existing literature, Audretsch and Mahmood (1995) report that age and firm’s size, capital intensity, the innovation rate at the industry level and new firms with new branches are key determinants affect the likelihood of survival firm Firm size is also explored closely connected with analysis of industry dynamics (Baldwin and Gorecki 1989, Mata and Machado 1996, Fariñas and Moreno 2000) Audretsch and Mahmood (1995) find that size, innovation rate, and price-cost margin influence the decision to exit Mata et al (1995) also argue that size is a crucial factor of chances for survival The other considerable empirical papers reveal a difference in the relationship between exit and firm’s size

Regarding to their production capacity and fixed assets, Yang and Temple (2012) divide firms into three size types: small, medium, and large Size is estimated by total labors of the firm in each year (Dimara et al 2008) Liedholm and Mead (1999) investigate a data set of micro firms (fewer than 10 workers) Other research finds that large and small firms faced various competitive conditions due to operating on markets of a dissimilar scope and applying different technologies (e.g Audretsch et al 1999)

The majority of researches show a negative relationship between firm size and ratio of firm exit such as Segarra and Callejón (2002), Mahmood (2000), Audretch et al (2000), Baldwin (1995), Audretch and Mahmood (1995), Wagner (1994), Dunne and Hughes (1994), Hall (1987) In contrary, other researchers divided enterprises into different types of size to find the relationship between firm exit variable and size variable Lieberman (1990) reveals a positive relationship between small firm size and firm exit probability Most papers also show a positive correlation between size of small firm and exit probabilities and vice versa Other researchers also report that firm size is negatively correlated with probability of company exit (Lieberman 1990, Cohen and Klepper 1996, Yang and Temple 2012, Lieberman 1990, Frazer 2005, Liedholm and Mead 1999) Studying in a completely different way, a large number of researchers suppose that the company at

the lowest production level often has a tendency to the exit behavior On the other hand, if the exit

is the result of a market selection process, the inefficient firm will be high probability of exit (Blanchard et al 2012)

Following the previous authors, we will investigate the reasons to explain why size variable effect on the firm’s exit behavior According to Cohen and Klepper (1996), the bigger enterprises tend to invest in Research and Development (R & D) that tries to explore methods to improve existing products, and to develop new ones In addition, they are supplied with necessary amounts

of physical, financial, human and other resources that protect company from failure and enhance the possibility of exploiting scale economies and dealing with outside shocks Thus, bigger enterprises are also likely to survive in shakedown process

A negative correlation between firm size and the probability of exit can be explained by several reasons Firstly, the company with higher level of size is easier to make the closure decision

if their level of production is at minimum efficient scales Secondly, because larger firms are easier

to access to capital markets and have higher possibility to employ qualified and skilled workers, thus they have a higher capability of being survival in comparison with small firms (Ferragina et al 2012)

FIRM AGE

Another determinant is concentrated in the papers recently is a firm age variable Dunne et

al (1988), Thompson (2005) finds that the pattern of failure is systematically correlated with the firm age and the failure rate of non-failing enterprise declines with age Everything else constant, the firm’s probability of exit decreases with firm age (Agarwal 1997, Agarwal and Gort 1996, Audretsch 1991, Audretsch and Mahmood 1995, Olley and Pakes 1992) When researching on firm activity, the empirical studies examine the impact of age on exit and propose the ratio of exit diminishing with one (Agarwal 1997; Agarwal and Gort 1996; Wagner 1994; Mata and Portugal 1994; Mahmood 1995; Audretch 1991, 1994, 1995; Dunne et al 1988) Other studies also report a different relationship between the probability of exit and age of firm They reveal an inverted U-shaped relationship between failure rate and age which can be predicted by the liability of

“adolescene” (Fichman and Levinthal, 1991; Bruderl and Schussler, 1990)

Firm Age is accounted from official year of a firms’ starting activity to year t (Ferragina et

al 2012) According to Dimara et al (2008), age also is calculated in years from the time of foundation for each enterprise

Older firms are harder to exit due to characteristics which help them in preventing the exit in the past (Ferragina et al 2012) While Dunne et.al (1988) and Disney et al (2003), in their empirical researches, show that younger firms have higher probability to exit because they are learning about their true productivity levels, they do not have much practical experience Moreover, older firms which have survived in the long time are recognized themselves in the market hence they might be better in surviving after adverse shock of given size (e.g through their goodwill, trademarks or the close connections to suppliers or to the capital market.)

PROFIT

Recent empirical studies have examined on the affection of Profit on the exit of enterprises, which determine whether to exit or continue to market rests essentially on the enterprise’s prospects for profits (Baggs 2005) The companies in both manufacturing and service sectors have a lower ratio of exit if the one has more profit, while lower profits encourage the decision to exit (Ferragina

et al 2012) Many researchers suppose that the firm with high profit has a negative influence on firm exit proportion However, other authors (Austin et al 1990, Evans and Siegfried 1994) find no connection between profitability and ratio of exit

Pérez and Castillejo (2008) also reveal that the longer or shorter survival times of enterprise might depend on the firm profitability On the one hand, it is expected that profitable company will have a better survival opportunities if high profit comes from market power From provided profit resources, companies can invest on assets by means of advertising and innovation that might alleviate firm survival On the other hand, high profitability might relate to shorter survival because

of some reasons Firstly, innovative activities which bring high returns also increase the risk of failure due to its uncertainty Secondly, profitable firms might have higher possibility to be involved in mergers or to be acquired (Pérez and Castillejo 2008)

Even though such this correlation would seem obvious to economics, the practical results are mixing Just like the conclusions from previous papers, we are expectation that the profit variable will negatively influence to probability of exit

Overall, the activity of the enterprise has a functions and goals differently in every time period However, the final target is to obtain earning, which decides to the business plan of the owner with expanding business operation in the competitive market condition If the firm’s profit does not cover all fees and costs to maintain the business operation, the company will fall into financial crisis Approach the model in perfect competition environment, enterprisers will consider plan to go out business operation if they do not control and cover their fixed cost Maximize the profitability of firm’s business is chosen if their profit is above their opportunity cost (Agarwal 1997)

INVESTMENT

Investment activity is understood as the sacrifice of financial resources in the present to obtain higher results for future investors There are different investment activities such as: development investment, commercial investment and financial investment For a business, the enhancement of competitiveness is achieved through a form of development investment which uses financial resources, physical resources, labor and intellectual resources to build and repair buildings and infrastructure, to train human resources to make regular expenditures associated with the operation of these assets in order to maintain, enhance and expand the production Other previous studies have explored the effect of the investment on firm exit, which is considered a proxy for expanding restructure, capability building and opportunities for investment in the business (Colantone and Sleuwagen 2010) Other researching at Japanese firms reveals that oversea investment and probability of firm survival have a negative relation (Kimura and Kiyota 2006)

If the business makes a profit in this financial year, they will have a tendency to set a plan for investment on building, equipment, research and development (R&D), workforce and other investments in next time Other articles basing on the standard economic theory report that an investment decision made by business owners could be treated as a risk with the ability of success

or failure (Rahaman 2009) Mistakes in investment can make a larger debt, thus causing serious consequences for the business Therefore, the investment activity can affect directly to development

of company and decide survival or exit of firm

Geroski (1995) and Colantone et al (2014) use an investment variable by accounting logarithm of the net spending on tangible assets over sales at the industry level However, Khandelwal (2010) defines the investment as standing for net investment per one employee at the

industry level In my study, I use total investments of the last survey is an explanatory variable in model We expect that my result shows an impact positively on exit from an investment variable as previous empirical studies

LEVERAGE

An important determinant in exit model is a leverage variable According to Bagg (2005), he presents that several authors in recent literatures have focused on debt leverage as crucial determinant in studying survival of enterprise in general and in the globalization context in particular

Several empirical articles give a positive relation between the likelihood of exit and leverage (Fotopoulos and Louri 2000, Tsionas 2006) Baggs et al (2005) show that higher debt leverage increases firm exit compare to firm with lower one level This means that the firm with higher leverage level is negative relationship with survive With the same result with other authors, Bags (2005) shows firms having higher leverage are more likely to exit However, less leveraged firms see a lower probability of survival

Like previous studies, Zingales (1998) also reports that the leverage is relevant to the ability survival of enterprises in competitive environment conditions is positive relationship which is likely

to described results in papers of Heiss et al and Huyghebaert et al (2004)

An Empirical study by Lang et al (1994) points a definition of leverage is debt divided by the total asset In this paper, we use lagged leverage which is accounted by total debt per total assets

in last year (t-1) and expect to have a significant positive relationship with a probability of firm exit

The firm’s manager always considers a decision in tradeoff advantage between the equity and cost debt Stockholders and bondholders also can happen to several conflicts when the firm makes a choice between investment and financing policies The reason for higher risk is that firm using high debt leverage has to payment a large interest amount Thus, they reduce the probability

of survival (Lang et al 1994) The level of debt or the decision about financing instrument is affected by many factors such as operating risk, firm size, the company's assets structure, and the rate at which retentions are generated (Krugman and Paul 1980) When the firm has a high leverage, firm’s capital's self-control ability is lessened because it relies so much on debt Zingales and Luigi (1998) reveal that enterprises with higher leverage have a tendency for less investment on

the exogenous shock Thus, the firms can be active to control their finance, which contributes to reduce the exit ability The company will reduce the probability of survival in case of a high level

of debt because of high interest payments On the paper which published in 2000, Fotopoulos and Louri said that the rate of debt to total assets is more likely to probability of exit

ASSET

Baggs (2005) tests hypothesis about relation between firm survival, firm exit and several variables: levels of employment, total revenue, total assets, debt leverage and tariffs for all enterprises He finds that the important result that asset will effect positively on ratio of exit if sales are hold remain constant Macus (2016) also comments to total asset as size variable in his research when he focuses on determinants effecting on firms’ exit rate The firm size impacts negatively on the likelihood of exit, said Tsionas (2006) He used the firm size variable is the logarithm of the firm’s total asset

OTHER DETERMINANTS

RESEARCH AND DEVELOPMENT (R&D)

Research and development (R&D) is considered as a special spending regarding firm’s sunk cost It sets a barrier helps company less likely to exit Doi (1999) reveals that (R&D) impacts negatively to firm exit In contrary with above result, Audretch et al (2000) and Segarra and Callejón (2002) find the positive relation between exit and (R&D) Audretch et al (2000) also show that R&D on high specialization industries is more increasing exit rate because it relates to bigger uncertainty level Besides, Mahmood (2000) also uses the R&D variable, which is collected data from varies across industries, to measure an influence level on firm exit

SUNK COSTS

The enterprise, which operates in several industries having a large level of sunk cost in total asset, has a tendency to less probability of exit than other firms Because the characteristics of asset are durability and specificity create a barrier when business owner considers cost to exit Other evidence is provided by Fotopoulos and Louri in 2000 reveals that sunk cost variable and probability of exit exist a negative relationship

CAPITAL REQUIREMENT

There are many papers in the world provide evidence about negative relationship between ratio of exit and capital intensity such as: Audretch et al (2000), Austin and Rosenbaum (1990), Dunne and Roberts (1991), and MacDonald (1986) The capital requirement in one manufacture is desirable to make decreasing exit because enterprises having a large capital level are more responsibility with their resources It makes a barrier to exit when owner consider a decision about your business activity

MARKET CONCENTRATION

Flynn (1991) and Doi (1999) show a hypothesis about relationship between market concentration and likelihood of firm closure which has a negative relation The main reason is because enterprises operate in highly concentrated markets usually are applied monopoly policies from this market segment which do not use for other firms

INDUSTRY GROWTH

There are varied different results from researches in the world when they study about relation between industry growth and firm exit The downtrend of market is expected to increase probability of exit This result is consistent with conclusions are provided by Audretch et al (2000), Dunne and Roberts (1991), Segarra and Callejón (2002), Austin and Rosenbaum (1990), and Doi (1999) Several papers present the opposite result such as: Evans and Siegfried (1993), Shapiro (1983) and MacDonald (1986) A few other researches reveal no relationship between exit and industry growth

ECONOMIES OF SCALE

The minimum efficient scale (MES) is a lowest production point which one company needs

to achieve to produce a product at a competitive price compare to competitors When they operate below at MES level, they fall in disadvantageous situation compare to other companies in the same industry because it takes more costs to produce goods Thus, MES is an important determinant to decrease exit ratio Some articles give a same argument such as: Audretch et al (2000), Mahmood (2000), Wagner (1994) and Dunne and Hughes (1994)

FIRM PRODUCTIVITY

Production capacity decides efficiency and effectiveness of enterprise Thus, it impacts on possibility of survival This result is provided by several researchers when they studied in enterprises product at lowest-productive level such as: Jovanovic (1982), Melitz (2002) and Ericson and Pakes (1995)

2.2 Review of empirical studies on firm survival/exit

2.2.1 Approaches of analyzing firm exit and survival

Researchers approach issues about exit and survival follow many different perspectives Many of the works on the firm survival or firm exit have analysis at macro level Some other studies examine the effect of determinants on exit at the industry level They collected data from manufacturing industries, services industries or focus on several important industries at their countries or across countries

Colantone and Sleuwaegen (2010) put a lot of emphasis on ratio exit of firm at the twelve manufacturing each country in eight European countries, during the period 1997 -2003 Kneller and McGowan (2012) find the effects of tax policy on the enterprise entry and exit which was conducted at 19 OECD countries over the time-span 1998 - 2005

On the contrary, other scholars only focus on exit at the firm or industry, which exit is considered as dependent variable in country level Baggs (2005) considers both firms survive, and exit based on firm information within 15 years, from 1984 to 1998 at manufacturing firms which is supplied by Statistic Canada Acs and Audretsch (1989) show that a ratio of return at the industry level is likely positive to probability of entry and exit There are many empirical researches recently present a relation positively between ratio of exit at year (t) and ratio of entry at year (t-1) at same industry such as: Backer and Sleuwaegen (2003), Mata and Portugal (1995), Dunne et al (1988), Siegfried and Evans (1994) Colantone and Sleuwaegen (2010) also put a lot of emphasis on ratio exit and entry of the firm at the industry - country level Other articles also focus on studying on probability of enterprise exit at the industry level, such as: Brian & Vroom (2014), Blanchard et al (2013), Dunne (1988), Ferraginaa et al (2011), Dimara et al., (2007)

A large number of researches recently access enterprises operation through survive and exit They also use the different estimation models to evaluate the influence of determinants on firm survival or firm exit

Almost articles approach the effect of explanatory variables on the rate of firm survival through using hazard model such as: Cefis and Marsili (2005), Agarwal and Gort (2002), Pérez et

al (2008), Musso and Schiavo (2008), Görg and Strobl (2000)

The same with papers studied on firm survival, many other research articles around the world use hazard models to study relationships between business and non-business factors and the rate of firm exit (Dimara et al 2008, Ferragina et al 2012, Yang and Temple 2012) However, A large number of recent papers have used probit regression (Blanchard et al 2012, Frazer 2005, Baggs 2005).Specially, Tsionas et al (2006) use both logit model and probit model to examine the relation between technical inefficiency (TI) and firm exit

2.2.2 Emprical analyses of firm survival

Because of the important of enterprises in social and economic development, so there are many discussions of firm survival base on its determinants and other characteristics not belong to firms

Cefis and Marsili (2005) imply proportional hazards Cox model with databases was surveyed by the Central Bureau of Statistics in Netherlands Company’s survival time is used as crucial variable, which was excluded observes happened exit event Independent variables include innovation, firm size, firm age, firm growth, and industrial classification The same results with previous papers in the literature part, they confirm that age, size, and growth rate are more likely to survive for longer In addition, when they controlled for the influence of other variables such as age, size, sale’s growth, and character of technology, innovation variable affect positively the survival probability

Besides that, Agarwal and Gort (2002) use three main determinants including: net investment, disparity of initial endowments’ quality, and learning by firm to reduce cost, to increase return, and to improve productivity They also imply the overall hazard-rate function to focus on simultaneously effect of independent variables, the variation in the probability of survival is illustrated through three main determinants of firms, such as net investment, disparity of initial

endowments’ quality, and learning by firm to reduce cost, to increase return, and to improve productivity Like with previous authors, they use the Cox proportional-hazards regression model which was applicant mostly in survival analysis They detect the positive correlation between technology intensity and probability of survival, while ratio of high capital labor and firm survival

is contrary

Moreover, Pérez et al (2008) investigate the determinants of firm survival by using the manufacturing firms data which was collected every year by Ministry of Industry in Spain The hazard model is implied to find the impact of hypotheses on survival based on the enterprise’s Resource-Based Theory This study describes that large, highly-productive and conducting R&D firms are significantly better possibility of survival than other firms at a 1% significance level They also indicate that the firm operates in factors with high and low innovation levels are significantly more difficult than in the intermediate level ones The result is found by Audretsch (1995) and Segarra and Callejon (2002) is the relatively unusual relationship between survival chance and age, and the ratio of exit raises in the first period of a firm, then reduces and increase later

In 2008, Muss and Schiavo implied a new approach explored the influences of financial constraints on firm survival and development In the line with previous researches, Muss and Schiavo used proportional hazards form with panel data over the time period from 1996 to 2004 They only focused on manufacturing firms in French They found that the financial constraint is positive relation with the likelihood of firm exiting

Another recently study, Görg and Strobl (2000) focus on Multinational companies (MNCs), using firm level data for Irish manufacturing industries for time survey from 1973 to 1996, study the effect of multinationals on firm survival using a Cox proportional hazard model They investigate that MNCs influence positively in firm’s survival chance through technology spillovers, but contrary through the crowding out of rivals They do not detect any influence of MNCs presence on domestic low tech companies and foreign companies in high tech sectors, while they find a negative impact of MNCs on the survival of foreign firms at low tech industries

A large number of papers use panel data of enterprises in varied country to find out the factors affected to firm survival such as: Agarwal and Audretsch (2001), Mata and Portugal (2002), Mata, Portugal and Giumaraes (1995), Audretsch and Mahmood (1995), Dunne et al (1989) They use the data at the firm level, and most explanatory variables have an influence negatively on ratio

of firm survival such as: number of labor, the number of businesses that own their factories; a competitive advantage of ownership; level of diversity; ability of learning by doing and management experience When they researched at the industrial level, the variables as process of increasing industry, industry’s life cycle, development of technology have positive relationship with firm of survival

2.2.3 Emprical analyses of firm Exit

Consistent with many previous papers, Colantone and Sleuwaegen (2010) put a lot of emphasis of international trade on ratio exit and entry of the firm at the industry- country level This rate is determined as the rate of the firm sum of shut down or birth at survey year over the sum of surveyed firms in the given period Furthermore, he also uses other variables to explain the changing annual firm exit rate of the manufacturing sector, such as investment, capital intensity, total factor productivity (TFP), index of change in trade, etc This survey is conducted in eight European countries The author researches on two key dependent variables including the rate of industry-level firm exit and rate of firm entry based on using the least Squares estimation They also use the lagged entry and lagged exit rate in their regression model As expected, exit is more likely with previous entry Many other research’s results in the papers by Mata & Portugal (1994), Siegfried & Evans (1994), Caves (1998), Dunne et al (1988) show that entry and exit have a positive relationship On the contrary, this rate has the negative relation with a variable which presents forward changing comparative advantage By contrast, it is not significant with capital sensitivity and multi-factor productivity The variation in the trade openness at the first lag has significant relation positively in firm exit, while the second lag is insignificant This result is consistent with previous papers show an effect negatively between trade openness and rate of firm survival (Biernard et al 2006a)

Viewing in another side, Huyghebaert et al (2004) used hazard model to emphasis on the effect of competition, debt leverage, and characteristics of financial market on the probability of startups’ firm exit This research also reveals the correlation between competition ability and the exit

of entrepreneurial start-ups Data and information for this analysis were collected from 235 entrepreneurial startups in Belgium over the years 1992–1999 The research has also shown the negative significant effect between employees and exit in the year following start-up The firm size

is a good sign for firm exit but debt leverage is not significant

By another study, Baggs (2005) suggests that exist the interaction in empirically modeling exit/ survival between exit/survival and explanatory variables such as age, size, rate of firm entry, and operation efficiency Using a unique data set of Canadian manufacturing firms at the firm-level control variables, he estimates an equation by using a probit technique to find the effect between exit/ survival of firm and firm-level control variables, especially Canadian tariff reductions issue when Canada agreed on Canada-US Free Trade Agreement They found that lager firms in term of labor and sales have an ability to survive better than other firm It is likely with result showed by

Gu et al (2002) However, lager firms in terms of assets have a lower probability of survival By other hand, if author held sales constant, the enterprises with higher assets increase ratio of exit If author estimated without sales as independent variable, we have a result inversely They also find that probability of firm survival has a significantly negative relation on leverage variable and highly positive correlation with age They also investigate that enterprises with higher leverage are more likely to exit than enterprises with lower one levels

Add a study was conducted in European countries by Colantone et.al (2014) also uses the method of least squares to fit a model to their data with Dummy Variables, divide into small companies and large companies group, over time-span 1997-2003 The dependent variable in this empirical model is the industry level exit rate which is affected by the change in import competition condition In particular, they find the same result with previous studies Large enterprises with larger production scale increase the probability of exit when global competition environment is higher level The empirical also give the positive relation between exit and previous entry This result is showed by De Backer and Sleuwaegen (2003), Mata and Portugal (1994), Dunne et al (1988) They also investigated a positively relationship between exit and lagged sectoral TFP growth, but only true for large firms (Malerba 2007) Additionally, the effect of import competition variable on firm’s exit is heterogeneous Dunne et al (1994) point that large enterprises operate in large scale production increase exit if import competition was higher but the exit of small firm is not impacted by imports factor in low-cost nations However, the relation between exit of small firms and import competition in developed economies countries is vice versa

Yang and Temple (2012) use labor Productivity as the indicator of firm performance Similar to the empirical model in previous papers, they use hazard rate model to evaluate effect of determinants on firm exit by using data collected all firms belong to manufacturing sector in China

with different ownership types at one province With an unbalanced panel of 3992 firms during the period 1987 - 1996, they classify firms as large, medium or small company as China’s State Planning committees classify The used features in his model include ownership, age, size, enterprise performance, financial constraints, and some factors relate to social policy In this analysis, they use the logarithm of enterprise age to apply in their model Running to the regressions

by two ownership types, they reveal a result interestingly contrast between the state owned enterprises (SOEs) and small and collectively owned enterprises (COEs) Regression result reveals that age of an enterprise in two ownership types impact declining significantly with probability of exit Besides, Yang and Temple also investigate that the impact of the reform dummy on change in rate of exit is fairly lower for COEs than SOEs In addition, a decrease for SOEs is higher the probability of exit than COE

Frazer (2005) applied a probit model to predict rate of firm exit by approached linear probability model With data was collected from companies operating in manufacturing factors, he evaluated firm exit by using fixed effects estimator to refer whether less enterprises with less production are more or less probable to exit than other enterprises The paper also provided the factors which influence firm exit include firm age, firm old, capital, and export operation The estimation gives results that the capital-sensitive businesses are more likely to exit The line with expect at previous researches, the bigger firms and older firm decrease with exit rate but in general the age variable is negative relationship with exit variable Contrary to results from other researches, state-owned enterprises have tendency more exit than private enterprises In addition, companies that switched from more exporting activity to not exporting activity are higher probability of exit

CHAPTER 3: RESEARCH METHODOLOGY

3.1 Conceptual framework and the econometric model

3.1.1 Conceptual framework

Basing mentioned in the literature review, a conceptual framework is built as in Figure 3.1

The empirical research indicates a large number of factors which affect on firm exit They are collected from a large number of previous studies and papers in the world

Figure 3.1: Conceptual framework

In my study, I just only focus on several determinants including: firm size, firm age, investment, profit, leverage and asset The main reason is that we do not have enough information about other determinants to collect enough data In addition, some factors are not mentioned in main questionnaires which was used to survey of small and medium scale manufacturing enterprises (SMEs) in Vietnam

EXIT

INVEST MENT

PROFIT

LEVERA GE

ASSET

R&D

SUNK COSTS CAPITA

L REQUI MENT

MARKET CONCEN TRATIO N

INDUST

RY GROWT H

ECONOM IES OF SCALE

FIRM PRODUC TIVITY

FIRM SIZE

FIRM AGE

3.1.2 Theoretical reviews

According to Random Utility Theory by Marschak (1960), the exit behavior can be descripted through utility of enterprise which is determined by deterministic components and random components

In this way, the total utility of a firm (i) exit is equal the sum of the both utility components

as below:

𝑈1𝑖 = 𝑉1𝑖 + 𝑒1𝑖

Where V1i can be approximated by a linear function of exit in the vector of Xi and 𝑉1𝑖 can

be approximated by a linear function of non-exit in the vector of 𝑋𝑖 and the population utility weights for each attribute in the vector βi: 𝑉1𝑖 = 𝛽1𝑖 * 𝑋1𝑖

In additional, 𝑒1𝑖 is a random utility component

Similarly, the total utility of a firm i with non- exit:

𝑈0𝑖 = 𝑉0𝑖 + 𝑒0𝑖

Where, 𝑉0𝑖 can be approximated by a linear function of non-exit in the vector of Xi and the population utility weights for each attribute in the vector βi: 𝑉0𝑖 = 𝛽0𝑖 𝑋0𝑖 In additional, 𝑒0𝑖 is a random utility component

The probability that a firm can be expressed as the probability that the utility associated with exit is higher than utility of non-exit:

Pr (exit) = 1

3.1.3 The econometric model

A repeat enterprise is a firm still in operation this survey year, which was still activity last surveyed These enterprises will be categorized as type 1 enterprises (Code 0) And in those surveys, if the enterprise is not presently operating and there are no concrete plans to start the business again in the near future or if the firm has been formally declared bankrupt is considered an exit And the answer,

in this case, is “exit” If the firm is still in business, the answer, in this case, is “not - exit” The response to that circumstance is coded into a binary variable which means 1 in the case of “exit” and 0 otherwise

The baseline estimating equation is as follows:

Such models are usually estimated by specialized methods, such as logit and probit Many papers choose logit model because it has mathematical simplicity (Gujara, 2003) In the thesis, a logit model is employed

The logit model:

𝒚𝒊𝒕* = 𝜷𝟏 + 𝜷𝟐 *𝒙𝟐𝒊𝒕+ 𝜷𝟑* 𝒙𝟑𝒊𝒕+ ……… + 𝜷𝒌* 𝒙𝒌𝒊𝒕+ 𝒖𝒕

The above expression is generalized to the case in which there are k independent variables (𝑥2, 𝑥3, 𝑥𝑘) each firm i, and t equal year (2005, 2007, 2009, 2011)

Where 𝑦𝑖𝑡* is unobservable variable, but we have 𝑦𝑖𝑡= 0 if yit*<0 and 𝑦𝑖𝑡 = 1 if 𝑦𝑖𝑡* >= 0

non-exit), or the log of the odds Ratio that 𝒚𝒊𝒕=1 is a linear function of the explanatory variables

3.2 Data and variables

The main data for my paper is based on firm-level exit data which is the result of surveys conducted annually by collaboration between the Institute of Labour Studies and Social Affairs (ILSSA) in the Ministry of Labour, Invalids and Social Affairs (MOLISA) and Department of Economics, University of Copenhagen with funding from DANIDA To collect data, investigators use the key survey instrument is the so-called main questionnaire for a firm interview The survey aims at collecting information on the actual development of SMEs in rural and urban area Data from Vietnamese firms is currently available for the years 2005 to 2011 This survey was conducted

in the cities and provinces following code in bellowing table

Table 3.1: List of surveyed province/city

No Province/city Code Freq Percent Cum

Source: Author’s calculation

The variables in this paper mention about the exit, total gross profit, debt leverage, investment, firm size, total asset, firm age and other factor is province Table 3.2.2 defines variables

Table 3.2: Definitions of variables Variable/symbol in data Descriptive

Exit (exit)- dependent

variable

The firm exit is not presently operating and there are no concrete plans to start the business again in the near future Dummy variable is equal “1” if firm exit is equal “0” if otherwise

Total gorss profit (tgp) Total gross profit equal Value of production/ manufactured output minus

(Total indirect costs and Value of raw materials used) minus (Total wage bill, including allowances and other labor costs such as social and health insurance, training, recruitment, etc.) in the last year (1,000,000 VND) Debt leverage (lvr) Total short-term debt over total assets in year t-1

Investment (i) Investments during the past two years, then add up all the investments

made (1,000,000 VND) Total asset (aset) Total asset in last survey year (end-year) (1,000,000,000 VND)

Firm size (size) Total Number of regular full-time labor force in year t-1

Firm age (old) Number of year since establishment up for survey year (t)

Province (prv - D1) Province or city of the main production facility

Year (year - D2) Survey year include 2005,2007, 2009, 2011

Enterprise number (id) Enterprise number For “repeat” enterprises make sure that the number of

the enterprise corresponds with the number given in the last survey

The firm is considered as exit out market when it is last surveyed enterprises no longer in existence

We use the question “Has the enterprise since the last survey and Been closed down for a year or more?” to find result If they answered “yes” is equivalent to code “1”, this enterprise is exit If they

answered “No” is equivalent to code “0”, this enterprise is operating

CHAPTER 4: RESEARCH RESULT

4.1 Descriptive Statistic

Table 4.1: Descriptive Statistic

Total gross profit (mil

VND)

Source: Author’s calculation

Table 4.1 presents the descriptive statistic of the panel data, including 10 provinces and cities in the 2005-2011 period

Regarding descriptive analysis, we use six main explanatory variables to explore the exit issue, including total gross profit, firm size, investment, total asset, firm age and leverage

As can be shown in the above table, the huge difference in the value of these variables exposes the large differences among enterprises

There is a major difference between the largest total gross profit of enterprise and the smallest one Inside the firms has a big profit, we also have many firms operation ineffectively, minimize is near (-451) million dongs per year, the maximum is 236.824 billion dong per year This figure is not high, but it affects directly on financial position of the business, is an important condition for ensuring the solvency of the business If enterprises do effective business, high profit, the strong solvency, firms can repay all matured liabilities and vice versa

The mean of the total asset is about 3.34 billion dong This figure reaches the maximum at

1044 billion dong and minimum at nearly one million dong