Firm-specific Factors Affecting on cost of equity in Vietnam

This thesis investigates the relationship between specific-factors and the cost of equity capital in the Vietnam stock market. The approach applied by Pomerleano and Zhang (1991) has been used to estimate the cost of equity. The sample consists of 79 listed firms spread across both HoSE and HNX. For a period of three years, namely, 2009 to 2012 have been collected data to test the relationship between cost of equity capital and the level of firm-specific factors including market risk (BETA), the size of company (SIZE), Book to market ratio (BM), Debt to equity ratio (DE), Asset turnover ratio (ATO), Inventory to total assets (INV). Panel data and econometric testing model are implemented in the paper in order to get the most precise results. After running regressions, the final results show that BETA, SIZE, BM, ATO, INV are positive related to the cost of equity, while DE shows a negative association. Most coefficients confirm the expected relationship, documented by previous studies. For exception results between DE, SIZE and cost of equity, the paper explains basing on some unique characteristics of Vietnam market. Finally, some conclusions and recommendations for further research are made in order to capture some limitation of this paper as well as understand deeply these relationships accompanied by difference of Vietnam stock market.

FIRM-SPECIFIC FACTORS AFFECTING

COST OF EQUITY IN VIETNAM

HoChiMinh city, Vietnam

2013

TABLE OF CONTENTS

LIST OF TABLES iv

LIST OF FIGURES v

ABSTRACT vi

CHAPTER I: INTRODUCTION 1

1.1 Background and motivation of the study 1

1.2 Research Objectives 3

1.3 Outline of the study 4

CHAPTER II: OVERVIEW OF VIETNAM STOCK MARKET 5

CHAPTER III:LITERATURE REVIEW 8

3.1 Cost of equity and estimation models 8

3.1.1 Cost of equity 8

3.1.2 Cost of equity estimation models 9

3.1.3 Alternative methodology for estimation of cost of equity capital 12

3.2 Firm-level factors 13

3.2.1 Beta 13

3.2.2 Size 14

3.2.3 Debt to Equity ratio (DE) 15

3.2.4 Book to market ratio (BM) 15

3.2.5 Assets turnover ratio 16

3.2.6 Inventory to total assets 17

CHAPTER IV: DATA & METHODOLOGY 18

4.1 Data 18

4.2 Methodology 20

4.2.1 Estimating cost of equity 20

4.2.2 Hypothesis 20

4.3 Econometrics Testing 21

4.3.1 Multicolinearity Test 21

4.3.2 Auto-correlation Test 22

4.3.3 Fixed or Random Effects Model: 23

4.3.4 Heteroskedasticity Test : 23

CHAPTER V: EMPIRICAL RESULTS 26

5.1 Data description 26

5.2 Econometric tests 28

5.2.1 Multicolinearity Test 28

5.2.2 Auto-correlation Test 28

5.2.3 Fixed or Random Effect Model ( Hausman Test) 29

5.2.4 Heteroskedasticity Test 30

5.3 Empirical Results 31

5.3.1 Beta and Cost of equity 31

5.3.2 Size and Cost of Equity 32

5.3.3 Debt to Equity and Cost of equity 32

5.3.4 Book to Market and cost of equity 34

5.3.5 Total Asset Turnover and Cost of Equity 35

5.3.6 Inventory to Total Asset and Cost of Equity 36

CHAPTER VI: CONCLUSION 37

6.1 Conclusion 37

6.2 Suggestions for further research 39

6.3 Research limitation 39

LIST OF REFERENCES 41

APPENDIX 45

LIST OF TABLES

Table 1 : Hypothesizes 21

Table 2: Description Data 26

Table 3: Correlation 28

Table 4 : Hausman Test 29

Table 5: Final Regression Result on firm-specific factors and cost of equity 31

LIST OF FIGURES

Figure 1: The performance of VN-Index and trading volume from 2000-2013 7

Figure 2 : The research framework 25

Figure 3 : Mean of Cost of Equity and Debt to Equity Ratio (2009-2012) 33

Figure 4: Mean of Cost of Equity and Book to Market Ratio (2009-2012) 34

ABSTRACT

This thesis investigates the relationship between specific-factors and the cost of equity capital in the Vietnam stock market The approach applied by Pomerleano and Zhang (1991) has been used to estimate the cost of equity The sample consists of 79 listed firms spread across both HoSE and HNX For a period of three years, namely, 2009 to 2012 have been collected data to test the relationship between cost of equity capital and the level of firm-specific factors including market risk (BETA), the size of company (SIZE), Book to market ratio (BM), Debt to equity ratio (DE), Asset turnover ratio (ATO), Inventory to total assets (INV) Panel data and econometric testing model are implemented in the paper in order to get the most precise results After running regressions, the final results show that BETA, SIZE,

BM, ATO, INV are positive related to the cost of equity, while DE shows a negative association Most coefficients confirm the expected relationship, documented by previous studies For exception results between DE, SIZE and cost of equity, the paper explains basing on some unique characteristics of Vietnam market Finally, some conclusions and recommendations for further research are made in order to capture some limitation of this paper as well as understand deeply these relationships accompanied by difference of Vietnam stock market

1

CHAPTER I

INTRODUCTION

In order to increase the total value of economic unit, it is not only making the right decisions and selecting the optimal solutions for investment funds but also estimating accurate cost of capital, which is major issue in the financial literature A company's securities typically include both debt and equity, so determining a firm’s cost of capital requires one to calculate both the cost of debt and the cost of equity However, the cost of equity is more challenged to calculate because equity does not pay a set return to its investors

The cost of equity is the return that stockholders require for their investment in a company There are many reasons to explain why equity cost is important Firstly, cost of equity represents to the compensation that the market demands in exchange for owning the asset and bearing the risk of ownership Secondly, the cost requires the company to maintain a share price that is theoretically satisfactory to investors Thirdly, it is also a major criterion in determining the investment priorities and the optimal capital structures Hence, the practice of estimating precise equity cost is a necessity

Eichengreen (2001) suggests that firm-level data should be used to capture the effects of institutional development on the cost of equity capital In Vietnam stock market, the paper conducted by Anh (2012) also presents that there is

2

not only the systematic risk but also the liquidity risk strongly affect the rate of return, so using CAPM in Vietnam is not stable To continue this subject, this study would investigate whether and how some firm-specific variables such as market risk (BETA), the size of company (SIZE), Book to market ratio (BM), Debt to equity ratio (DE), Asset turnover ratio (ATO), Inventory to total assets (INV) to impact on cost of equity in the emerging market, Vietnam is in particular There have been conducted relatively large numbers of research about the relationship between firm-specific factors and the cost of equity And empirical results of the paper’s findings will make the literature of the subject richer

In the present, there are many models using for estimating cost of equity: the capital asset pricing model (CAPM) (Sharpe,1964; Lintner,1965); Dividend Growth Model (Gordon, 1962); the three-factor model (Fama and French,1993) the international asset pricing model (IAPM) (Bekaert and Harvey, 1995), and so on Each of these models have the effectiveness needed for estimating the cost of equity capital in one or more specific market in a period of time; and their efficiency is reduced and replaced with each other over time Although these different models or techniques provide estimates of ‘reasonable’ magnitude, the search for a model that

is easy to use and provides accurate and stable estimates over time is not over yet (Fama and French, 1997; Madanoglu and Olsen, 2005) Moreover, Vietnam is an emerging market and the stock exchange is too young (established in 2000) with lack of historical data and consistent accounting standards Although the above techniques have been applied successful in developed countries, such limitations cause some issues when they are used in emerging markets like Vietnam Instead, this study introduces an alternative estimation model applied by Pomerleano and Zhang (1999) to estimate ex-post measures of cost of equity capital at firm-level based on historical realized returns The disadvantage of ex-ante measures, the limitation of the above methods as well as the reasons why using the approach applied Pomerleano and Zhang (1999) would be considered in the literature review

3

Vietnam is one of the emerging markets which have small young stock markets, so Vietnam market has some unique characteristics which influence on the relationships between cost of equity and firm-specific factors mentioned above The question is here whether theories as well as results of previous studies regarding the relationships are still accurate in Vietnam stock markets or not For example, the financial leverage ratio or Debt to Equity (DE) of Vietnamese firms is 1.1x on average (2009-2012) and higher than mean of the DE in other neighbor markets, 0.98x (Feng et al., 2013); however, investors still believe that whether there is a negative relationship between cost of equity and Debt to Equity ratio in Vietnam market Similarly, such questions are also made to other factors Hence, the answers would be in the finding section of this study

This paper would contribute on the literature as well as the real stock exchange by several ways First, the results give clearly understanding cost of equity related to firm-specific factors for investors to make accurate decisions in Vietnam stock market In addition, this paper supports reliable results in conclusions for the future studies to investigate deeper the relationship between equity costs and factors

in emerging markets in general or Vietnam in particular

The purpose of this thesis would be

Indicating whether and how firm-specific factors relate to cost of equity in Vietnam stock market

Investigating empirical results compared with arguments of the previous papers

Recommendations for the future studies regarding the relationship between cost of equity and firm-specific factors

4

The study is laid out as follows

Chapter I: INTRODUCTION briefly mentions the dissertation objectives, important contents and methodology conducted in the dissertation

Chapter II: OVERVIEW OF VIETNAM STOCK MARKET includes the background, the organization and operation and the main features of Vietnam stock market index

Chapter III: LITERATURE REVIEW focuses on the theoretical, estimating cost of equity models and empirical review of the relationships between firm-specific factors and cost of equity

Chapter IV: DATA & METHODOLOGY describes the data characteristics, the way used for gathering figures of different sources and research methodology employed in this dissertation The chapter provides the test process including econometric assumptions

Chapter V: EMPIRICAL RESULTS illustrate data, the regression results

of the study, analysis and compare arguments obtained from the empirical tests

Chapter VI: CONCLUSION part summaries the outcome and provides some limitations and recommendations for further studies

5

CHAPTER II

OVERVIEW OF VIETNAM STOCK MARKET

In order to have an overview of Vietnam stock market, this chapter provides information including history, development, and some highlight periods

The government did aware the advantages of open market and competitive business environment, so the economic reform namely “Doi Moi” was introduced in 1986 Since early 1990s, an official stock market was established by Vietnam government with the purpose of creating a more robust market economy After four years of delays, the State Securities Commission of Vietnam (SSC) center was established in 1996 as a remarkable achievement in creating a public security market SSC plays an important role as the highest governmental body responsible for organization and operation regulation of Vietnam securities market As regards international aspect, SSC has cooperated with foreign countries and international financial institutions to complete legal framework and leads Vietnam to an effective operation capital market For conducting an appropriate model in Vietnam stock market, SSC has studied the stock market development of other Asian countries such as China, Thailand and Taiwan thoroughly For instance, Vietnam has followed the rules of foreign investment and ownership of these Asian markets

As a rule, a developed official trading center of stock market should be over the operation of an informal or OTC market in advance However, the practice did not happen in Vietnam before the announcement of the official stock market in

2000, which is the result of the government’s awareness of its necessity in developing the capital market in particular and the whole economy in general In

6

2000, the first official stock exchange market – Security Trading Centre was launched in Ho Chi Minh City (HOSTC), marking the new stage of Vietnam stock market development HOSTC is a non-profit organization under the direct administration of the State Securities Commissions with the three main functions which are organizing, managing and operating the securities trading system, conducting examinations and supervision of the trading activities and performing and settling the securities transactions In the first trading days, HOSTC begins with only two listed companies with a total capitalization of 16.87 million, namely the Refrigeration Electrical Engineering Joint Stock Company (REE), and Saigon Cable and Telecommunication Material Joint Stock Company (SACOM) The market’s growth by number of listed companies so far has been higher With the purpose of opening more capital mobilization channels for the government as well as supporting the relationship between the fiscal policies and the monetary system, in

2005, Hanoi Stock Exchange (HNX) was established and started trading on 14/07/2005 Until now, there are more than 600 listed stocks Compared with the other Asian capital markets, Vietnam stock market can be seen as the smallest and the most illiquid market Vietnam stock market has considerably low but steady growth Recently, Vietnam stock market witnessed the most rapid development in

2006 in terms of listed companies, total capitalization and trading volumes From the end of 2006 to the beginning of 2007, VN-Index rapidly increased and hit the peak

in March, 2007

of 2002 to the end of 2007 The most rapidly development was seen in 2006 with the significant increase in number of new listed companies, amount of market capitalization, VN-Index and trading volumes

Overall, Vietnam stock market is still young and small with the typical characteristics of emerging market such as high volatile, illiquidity capital market, imperfect legal framework and irrationality of investors However, Vietnam stock market rapidly develops with huge potential, worthy to comprehensively investigate And some oversea theories are not consistent with empirical results in Vietnam, for example, the relationships between cost of equity and relative factors

3.1.1 Cost of equity :

The cost of equity capital can be defined as the expected rate of return of the current and prospective equity shareholders As a firm becomes riskier, the investors demand a higher return resulting in a higher cost of equity capital There are following main reasons explaining why equity cost is important Firstly, a firm's cost of equity represents the compensation demanded by the equity shareholders to bear the risks associated with the firm which in turn affects stock prices Second, a certain return is expected by common shareholders when their equity is invested in a company The equity holders' required rate of return is a cost from the company's perspective because the stock price would be fallen due to simply shareholders’ selling if that company does not deliver this expected return The cost of equity is seen as a cost which firms have to maintain a share price in order to gratify theoretically investors’ requirement Thirdly, determining the investment priorities and the optimal capital structure would not be feasible without estimating the cost of capital rate as well as cost of equity

9

3.1.2 Cost of equity estimation models:

Regarding the estimation of the cost of equity capital, there are two streams in the existing literature, namely ex-ante estimates based on analysts’ forecasts, and ex- post estimates based on historical realized returns (Mónica, 2007)

In addition, Erb et al (1996a) used sovereign credit rating for calculating cost of equity capital

Ex-ante estimates:

The basic idea of the ex-ante cost of capital models is using observable forward-looking data instead of historical return realizations (Gebhardt, Lee, and Swaminathan, 2001) The unobservable expected cost of equity capital is estimated from currently available analysts’ consensus forecasts about a firm’s future income and its market price (Klein, 2006) Since Ameer (2007) argues that the main assumption of these studies has been analysts’ forecasts of short-run and long-run earnings and dividend per share capture market expectations Moreover, lack of analysts’ forecast data for sufficiently large numbers of firms in the emerging markets further exacerbates the choice of firms for meaningful analysis The analysts overstate the long-term growth of earnings and dividends and objectivity are still major issues (Goedhart et al., 2002) Therefore, these ex-ante techniques have limitations when applied to emerging markets

Using sovereign credit rating:

Erb et al (1996a) estimated a time series cross-section regression and the model forces a linear relation with the negative slope coefficient between credit rating and expected returns This implies a higher credit rating associated with lower average returns However, he also argues that the model is not necessarily the best model for expected returns and volatility because of the nature of the problem that there is no way to verify the accuracy results until some of the developing countries

“emerge” into the databases In addition, such a method may be not appropriate with

10

the purpose of this paper which conducts specific cost of equity of each firm in a

country, Vietnam

Ex-post Estimation models:

Beside ex-ante estimates based on analysis’ forecasts in the valuation model, there are many ex-post measure models found on realized returns and dividend per share such as the dividend growth model (DGM (Gordon,1962); the three-factor model (Fama and French,1993); The international asset pricing model (IAPM) (Bekaert and Harvey,1995) ; Capital Asset Pricing Model or CAPM (Sharpe,1964; Lintner,1965); and the methods applied by Goedhart et al.(2002) and Pomerleano and Zhang (1999) The section would overview each model and emphasize below weakness of each model in order to recommend an alternative suitable model, the approach applied by Pomerleano and Zhang in 1999

The Dividend Growth Model was proposed by Gordon in 1962 Changes

in the dividend yield of firms have been used to measure the cost of equity capital at the country-level (Miller, 1999; Bekaert and Harvey, 2000; Henry, 2000) Ameer (2007) argues that many firms have moved away from paying cash dividends to buying back shares Firms are finding other ways to return cash to shareholders Madanoglu and Olsen (2005) argue that another significant issue in applying DGM lies in determining the specific dividend growth rate Especially emerging markets, the estimates of cost of capital based on dividend yield might not be robust (Ameer, 2007)

Fama and French’s (1993) three-factor model is an extension of the

CAPM-type structure It is a multi-factor model with postulates that factors other than market and the risk free rate impact stock prices Although Fama and French (1997) argue that the three-factor model explains more variation in the equity risk premium than the original CAPM model, it produces large standard errors

The international asset pricing model (IAPM) can give negative

estimates of the cost of equity capital, as reported in Daouk et al (2005) due to its nonlinear structure which it is mathematically challenging and not easy to use

11

The Goedhart et al (2002) method assumes that expected growth in the

firm-level earnings is equal to GDP growth rate plus expected inflation rate However, the major limitation of this method is that the assumption of the historical average of real GDP growth rate as a proxy for long-term corporate profit growth, it

is only accurate for mature firms which have profit growth tracks GDP While this paper requires estimations for whole Vietnam stock market including small and big firms

Capital Asset Pricing Model or CAPM (Sharpe,1964; Lintner,1965) is

one of the most well-known and accepted models calculating the cost of equity CAPM is used to estimate cost of equity CAPM was built from Harry Markowitz’s portfolio optimization model Although CAPM is actually very easy to apply, there are many barriers and many necessary assumptions when applying CAPM in Vietnam market

Firstly, CAPM was based on Markowitz portfolio theory, hence it has a same assumption, market return must consist of all risky assets of the whole economy such as stocks; real estate, gold, and so on Yet, these markets are not traded with published real information in Vietnam market As the result, it does not make sense if we apply CAPM to Vietnam and assume VN-Index as a benchmark portfolio It is not accurate to see VN index as representation for whole of the market because VN Index is not even a half of Vietnam market capitalization (data

of ETF) In addition, individual investors attract not only stock market but also other investment channels such as currency or gold Thus, it is hard to have correct beta value by using the VN Index as a benchmark

Secondly, VN Index is not able to represent to a liquidity and diversified market because it is small size and young (officially established since 2000) Moreover, VN Index annual returns change annually with very high volatility

Thirdly, Vietnam government bonds do have risk According to Moody country risk raking, Vietnam bonds are rated Ba3 which are below investment bonds The point of fact is that it is hard for Vietnam government to pay debt That means Vietnam government bonds are not suitably seen as risk free assets Besides,

12

it is not appropriate to make long-term assumption basing on the quality and length

of data on bond There three following main reasons why the original CAPM might

be expected to fail

3.1.3 Alternative methodology for estimation of cost of equity capital :

The previous papers investigating cost of equity such as Botosan and Plumlee (2004); Chen et al.(2004); Hail and Leuz (2006) have used analysts’ forecast data on earnings, dividend per share, return on equity and earnings growth

to estimate ex-ante cost of equity capital, but these techniques have limitations when applied to emerging markets Since Ameer (2007) argue that the main assumption of these studies has been analysts’ forecasts of short-run and long-run earnings and dividend per share capture market expectations Moreover, lack of analysts’ forecast data for sufficiently large numbers of firms in the emerging markets influences negatively on the choice of firms for meaningful analysis In his paper, he also mentions that from theoretical and methodological perspectives, ex post measures of cost of capital at the firm-level are better than ex post measures at the country-level obtained from aggregate stock returns using the CAPM, the APM and Fama and French’s (1997) three factors model The conclusion supports for the purpose of this study only wondering cost of equity at the firm-level

From the limitation of the other methods and the advantages of ex-post measures mentioned above, this paper would use the approach applied by

Pomerleano and Zhang (1999) to estimate ex-post measures of cost of equity

capital at firm-level In the research, they began with the relationship between equity prices and the implied growth of future nominal earnings is derived from the hypothesis that current equity price, Pt , equals the discounted present value of future earnings, Et+j , (j ≥ 1), with a discount factor rt

With the purpose of testing the relationship between firm-specific factors including market risk (Beta) and cost of equity, it is necessary to gain “crystal” measures cost of equity which use estimation models without following any model like CAPM, IAPM, and so on There are two appropriate methods namely Dividend Growth Model (DGM) and equation (4) However, DGM has some disadvantages mentioned above, therefore, the approach applied by Pomerleano and Zhang (1999) with equation (4) would be suitable and used in this paper

3.2.1 Beta:

Beta (β) indicates the sensitivity of the company to market risk factors Beta represents the market risk for an asset and is calculated as the statistical measure of volatility of a specific asset/investment relative to the movement of a market group The conventional approach for estimating beta of an investment is a regression of returns on investment against returns on a market index According to

14

the capital asset pricing model (CAPM), the cost of equity capital is positively related to a firm’s beta In addition, Botosan (1997) and Hail (2002), Sharpe (1964) find that a firm’s beta is positively associated with its expected stock return

On the other hand, Chui and Wei (1998) examine the relationship between expected stock returns and beta (β), book-to-market equity, and size in five Pacific Basin emerging markets: Hong Kong, Korea, Malaysia, Taiwan, and Thailand They find no evidence in supporting a positive relationship between

market beta and stock returns

3.2.2 Size:

One of the important interpretations in return, risk and asset pricing literature is Size There are various ways to measure the size of firm such as annual revenue, value added, and assets However, revenue measure tells nothing about the depth of the underlying activity and value added gives a precise measure of activity, but it is usually not publicly available for individual firms (Canbäck, Feb 2002) Wald (1999); Mackay and Phillips (2002) found that most effective proxy for firm size is the natural log of total assets Thus, this paper would calculate Size variable

by nature log of a firm’s total assets and test whether a negative relationship is still right in Vietnam stock market

In the finance literature, it has been documented a negative size effect in the return-risk relationship established by the CAPM (Banz, 1981) Moreover, Bloomfield and Michaely (2004) argue that large firms are expected to have higher returns and slightly less risk than small firms Bowen et al (2008) indicate that large companies attract more attention which can reduce the problem of asymmetric information and lowers the cost of equity capital A significant negative relationship between firm size and the cost of equity capital is found by Hail and Leuz, 2006; Chen et al., 2004 A negative relationship is expected between SIZE and cost of equity capital

However, Anh (2012) investigated the relationship size and returns of the Equities in Vietnam Stock Market He shows that the size variable can affect the equity costs of listed shares, the slope coefficient is above 0; which in turn implies

15

that there is the positive correlation between the size and the cost of equity and vice versa

3.2.3 Debt to Equity ratio (DE):

It is worth taking into account that Debt issue is significant factor influencing on firms’ performance in Vietnam market The Debt-to–Equity ratio recorded of 647 listed non-financial firms on Quarter II-2012 is quite high with 1.53x in average Vietnam market is seemed as one of countries which have the highest DE in the entire world This issue was warned many years ago, according to General Statistics Office (GSO, Vietnam) in 1999-2002, the Debt to Equity was 1.32x in 1999 and 1.96x in 2002 Especially in the period from 2007 to the present, Vietnamese companies have tended to use high financial leverage

DE is defined as a measure of a company's financial leverage calculated total debt-to-total market value of equity ratio at the end of year It is proved that the ratio relates directly to cost of equity capital of a firm

The high debt to equity means that debt is used much to finance increased operations, so the company likely to gain potentially more earnings Inversely, it is worth taking into account that shareholders are able to gain nothing due to bankruptcy The argument is that the tax benefit of debt diminishes beyond a certain point, and the additional financial risk outweighs the lower nominal cost of debt, thereby increasing the cost of equity capital, reflecting the increase in the financial risk of a firm

The Modigliani–Miller II theorem argues that a higher debt-to-equity ratio leads to a higher required return on equity, because of the higher risk involved for equity-holders in a company with debt Moreover, Botosan and Plumlee (2004) used Debt to Equity ratio variable (DE) to state that a positive relationship between

DE and cost of equity capital

3.2.4 Book to market ratio (BM):

The book to market ratio (BM) is an important valuation measure in order to find the value of a company by dividing the book value of a firm by its market value Botosan and Plumlee (2005) note that the cost of equity capital is

16

significantly related to book-to-market value Penman (1991) argues that low BM firms remain more profitable than high BM firms, or BM is expected to relate positively to cost of equity

However, Fama and French (1992) suggest that a firm’s stock returns and book-to-market value have a negative relationship In other words, a firm with a higher book-to-market value is expected to have better stock returns As a young stock market, the herd mentality primary influences on the way individual investors

to make decisions The market views that the weak current earnings of these high

BM firms (undervalued) as temporary and foresees profitability improving in the future, low cost of equity is also expected by shareholders In addition, a country’s economic risk has related and influenced to book to market (Erb et al.,1996b) The previous researchers found that economic risk rating of a country has negatively related to book to market That means when a country improves its economic risk rating, BM would decrease Moreover, Erb et al (1996b) also gave an important implication that there is a negative relationship between BM and cost of equity capital in the case that the valuation of stock reflects these upward improvements in economic fundamentals and capital flows to emerging markets related better economic fundamentals

3.2.5 Assets turnover ratio:

Assets turnover ratio, ATO, defined as total sales divided by average total assets, is used as an indicator of managerial agency costs to test influence of managerial agency costs on the cost of equity capital Singh and Nejadmalayeri (2004) documented that firm-level cost of equity capital is influenced by managerial efficiency in utilization of firm resources Ang et al (2000) and Singh and Davidson (2003) argue that the management’s ability to employ assets efficiently is assessed via asset turnover ratio assess Low asset turnover ratios mean that the company is not managing its assets wisely as well as inefficient utilization of assets They indicate that the assets are obsolete Low asset turnover ratios imply that the companies have operated below their full capacity Singh (2004) argues that firms with higher assets turnover ratio—reflecting lower managerial agency problem in

3.2.6 Inventory to total assets:

One ratio used to assess operational management and inventory turnover is inventory divided by total assets In general, a low inventory to total assets ratio is indicative of good performance and profitability In Vietnam, the number of manufacture firms is the highest in total listed stocks and because inventory influences directly on these firms’ performance, hence this is a factor worth taking into considering

Feng et al (2013) defined internal control as a process affected by an organization's structure, work and authority flows, people and management information systems, designed to help the organization accomplish specific goals or objectives He states that a high level of inventory also raises the internal control risks related to the proper measurement and recording of inventory and this is reflected in the rising cost of equity In addition, the quality and liquidity of current assets are largely dependent on the composition of current assets The lower the percentage of inventory to the current assets, the greater the liquidity of current assets and vice-versa, or it means that a low this ratio is better than high ratio (BOSE, 2006) A positive relationship between the level of inventory to total assets and cost of equity would be expected

18

CHAPTER IV

DATA & METHODOLOGY

This chapter describes the highlights of data, the way used for gathering figures of different sources; research methodology employed; and research framework In this dissertation provides the test process including econometric assumptions as well

In Vietnam, there are two main stocks exchanges namely Ho Chi Minh Stock Exchange (HOSE) starting trading on 27/08/2000 and Hanoi Stock Exchange (HNX) starting trading on 14/07/2005 Vietnam stock market can be divided into three main periods until now, the initial period is from 2000 to 2005, this is the start-up period and market movement is unstable The second period is from 2005

to 2008, a nonstandard return that only happens once in 60 years (Hua Chia Yee, CFA) did occur with Vietnam stock market in this period In addition, the beta is commonly estimated using data over two years, so we choose the rest period from

2009 to 2012 in order to minimize Beta error (Hua Chia Yee, CFA) Hence, the study would obtain the listing companies which had been non-stop trading and listed for at least two years because we use covariance with VN-index as the fundamental of calculating beta

According to that, the number of chosen stocks would be number of companies listed before 2007 and they are not de-listed during the period 2009-

2012 For each stock, the firm-specific variables (Total Asset, Asset turnover ratio, Debt to Equity ratio, Book to market, Inventory) would be collected in the end of the year for eliminating the effects of seasonal fluctuations

19

In general, panel data models are more “efficient” than pooling sections (Michael Hauser, Financial Econometrics, 1292) Panel data analysis is a method of studying an exacting subject within multiple sites, periodically observed over a defined time frame With the combination of time series with cross-section contributes on investigating the dynamics of change with short time (Gujarati, 2003; 638-640)

cross-The raw data including stock price for calculating Beta, Total asset, Debt

to Equity ratio, Book-to-Market ratio, Asset turnover ratio, Inventory to Total Assets would be accumulated from Bloomberg’s source, other financial websites such as vietstock.vn, vndirect.com.vn It is necessary to check whether data collected is the most reliable or not After gathering the numbers, ratios of 10 firms are chosen randomly from three different sources mentioned above and compared with firms’ actually financial statements The result shows that Bloomberg is the

most verifiable source

As mentioned in the Literature Review section, firm-specific factors would be calculated with some ways below

DE = Total debt-to-total market value of equity ratio at the end of year

BM = The book value per share divided by market price

All amassed figures are put into panel data table

20

4.2.1 Estimating cost of equity:

The following model of Pomerleano and Zhang (1999):

RE=

Where:

RE is ex-post cost of equity capital

P/E is the price to earnings ratio

g is constant rate equals moving average previous 5years in growth rate

of earning per share

4.2.2 Hypothesis:

The purpose of this paper is investigation the question: How do

firm-specific factors affect to the cost of equity capital in Vietnam market?

To get the final equitation, we implement the following steps:

Size) The change of Adjusted R-Square, p-value would be estimated in order to

scrutinize whether adding Size variable into the model 1 is more validity

Similarly, we put other factors into model and check changes in p-value

and Adjusted R-Square in order to check whether those factors have relationship

with cost of equity And finally Equation 6: Re=f(Beta, Size, BM, DE,ATO,INV)

RE it = β 0 + β 1 *Beta + β 2 *Size it + β 3 * DE it + β 4 * BM it + β 5 *ATO it + β 6 *INV it + ε i

21

β 1 ; β 2 ; β 3 ; β 4; β 5; β 6 : respectively the slope coefficients of the market risk_ Beta;

Size; Debt to Equity Ratio; Book to market; Assets turnover ratio and Inventory to Total assets variables;

ε i : the error term

Table 1 : Hypothesizes

Hypothesis

In order to verify validity of the results, some econometric tests should

be implemented After conducting the model, Eview 6.0 software would be used for following tests

4.3.1 Multicolinearity Test :

Multicolinearity occurs when two or more explanatory variables in the model are correlated and provide redundant information about the response There are two main consequences of high multicolinearity including increased standard error of estimates of the coefficient β’s and often confusing as well as misleading results

One of the easiest methods determining multicolinearity is computing correlations between pairs of variables If some results are close to 1 or -1, there is a problem with multicolinearity

22

Solutions to multicolinearity:

Leonor Ayyangar (2007) recommended removing all but one of the highly correlated variables from the analysis We does not retain redundant variables since they measure the same characteristic All highly correlated variables can be deleted and re-introduced in the model in the form of an interaction or as a composite variable If possible, increase the sample size Increasing sample size will decrease the standard errors, which is greatly inflated by the presence of highly inter-correlated variables The study expects that there is no high correlation among variable

4.3.2 Auto-correlation Test:

Regression models using time series data often have the assumption of uncorrelated or independent errors for time series data But it is often not appropriate Usually the errors in time series data exhibit serial correlation, E(εiεj≠0) Such error terms are said to be auto-correlated

The best known formal test for auto-correlation is the Durbin-Watson statistic It is based on the assumption that the errors in the regression model are generated by a first-order auto-regressive process observed at equally spaced time periods,

The test statistic is: d = where e i = yi-ŷi and yi and ŷi are

respectively; the observed and predicted value of response variable for individual i

d becomes smaller as the serial correlations increase Upper and lower critical

values, dU and dL have been tabulated for different value of k (the number of explanatory variables) and n, in this study k=6 and n= 316