Factors affecting price to earnings ratios of listed companies on ho chi minh city stock market period 2008 2016

The author expects that the obtained empirical model of factors affecting earning P/E ratios for the case of listed stocks on Ho Chi Minh City Stock Market can be a good reference model

BANKING UNIVERSITY HO CHI MINH CITY

NGUYEN MINH NGUYET

FACTORS AFFECTING PRICE-TO-EARNINGS RATIOS OF LISTED COMPANIES ON HO CHI MINH CITY STOCK MARKET:

PERIOD 2008 - 2016

GRADUATION DISSERTATION MAJOR: FINANCE - BANKING

CODE: 7340201

HO CHI MINH CITY, 2018

MINISTRY OF EDUCATION AND TRAINING STATE BANK OF VIETNAM

BANKING UNIVERSITY HO CHI MINH CITY

NGUYEN MINH NGUYET

FACTORS AFFECTING PRICE-TO-EARNINGS RATIOS OF LISTED COMPANIES ON HO CHI MINH CITY STOCK MARKET:

PERIOD 2008-2016

GRADUATION DISSERTATION MAJOR: FINANCE - BANKING

CODE: 7340201

SUPERVISOR

DR NGUYEN TRAN PHUC

HO CHI MINH CITY, 2018

The empirical results demonstrate that dividend payout ratios, growth rate, beta and return on equity have effect on P/E ratios and it is possible to invest in low P/E ratios to beat the market Moreover, the research effect will be better with more factors employed

DECLARATION OF AUTHENTICITY

I declare that this dissertation is my original work, gathered and utilized especially to fulfil the purposes and objectives of this study, and has not been previously submitted to any other university for a higher degree I have mentioned all people who were significant facilitators of the work

Ho Chi Minh City, May, 2018

ACKNOWLEDGEMENTS

First of all, I want to send my gratitude and respect to Doctor Nguyen Tran Phuc for his sharing, understandings and considerate supports to give me useful recommendations and guidances during my study

Finally, best regards to my lecturers, my friends, my classmates and my beloved BUH for their sharing and supports during my Bachelor program

Ho Chi Minh City, May, 2018

TABLE OF CONTENTS

ABSTRACT i

ACKNOWLEDGEMENTS iii

TABLE OF CONTENTS iv

LIST OF ABBREVIATIONS vii

LIST OF TABLES viii

LIST OF FIGURES ix

Chapter 1 : INTRODUCTION 1

1.1 THE NECESSITY OF THE THESIS 1

1.2 RESEARCH OBJECTIVES 4

1.3 RESEARCH QUESTIONS 5

1.4 THE RESEARCH SUBJECT AND SCOPE OF THE STUDY 5

1.4.1 The research subject 5

1.4.2 Scope of the study 5

1.5 RESEARCH METHOD 5

1.6 RESEARCH CONTRIBUTION 6

1.6.1 The scientific contribution 6

1.6.2 The practical contribution 6

1.7 RESEARCH STRUCTURE 6

Chapter 2 : LITERATURE REVIEWS 8

2.1 Introduction 8

2.2 Theoretical background 8

2.2.1 The concept of P/E ratio 8

2.2.2 Use of P/E ratio for stock selection 9

2.2.3 Factors affecting P/E ratios 10

2.3 Previous empirical studies 15

2.3.1 Previous empirical studies in the international context 15

2.3.2 Previous empirical studies in the Vietnamese context 22

2.4 The position of this dissertation 25

2.5 Chapter conclusion 26

Chapter 3 : RESEARCH METHOD AND DATA 27

3.1 Introduction 27

3.2 Estimation model 27

3.3 Measurement of variables and hypothesis 28

3.3.1 Measurement of variables 28

3.3.2 Hypothesis 30

3.4 Source of data 33

3.5 Steps of data analysis and estimation 33

3.5.1 Descriptive statistics 33

3.5.2 Multicollinearity Test 33

3.5.3 Residual Diagnostics 34

3.5.4 Selection of estimation method 34

3.5.5 Test of the obtained empirical model 35

3.6 Chapter conclusion 35

Chapter 4 EMPIRICAL RESULTS AND DISCUSSIONS 36

4.1 Introduction 36

4.2 Descriptive statistics 36

4.3 Test of multicollinearity 42

4.4 Residual Diagnostics 42

4.5 Selection of estimation method 43

4.5.1 Pooled OLS 43

4.5.2 Random Effect Model (REM) 44

4.5.3 Fixed Effect Model (FEM) 45

4.5.4 Redundant Fixed Effect-Likelihood Ratio 46

4.5.5 Correlated Random Effects Hausman Test 47

4.6 Test of the obtained empirical model 51

4.7 Chapter conclusion 52

Chapter 5 : CONCLUSIONS AND RECOMMENDATIONS 53

5.1 Conclusion 53

5.2 Recommendations 54

5.3 Limitations of this dissertation and new approaches in the future 55

5.3.1 Limitations 55

5.3.2 Future studies 55

REFERENCE 57

APPENDIX 60

APPENDIX 1: The descriptive statistics for all variables 60

APPENDIX 2: The covariance of variables 60

APPENDIX 3: Breusch-Godfrey Serial Correlation LM Test 60

APPENDIX 4: Heteroskedasticity Test: White's result for empirical model61 APPENDIX 5: Pooled OLS's result for empirical model 61

APPENDIX 6: Random Effect Mode's result for empirical model 62

APPENDIX 7: Fixed Effect Mode's result for empirical model 63

APPENDIX 8: Fixed Effect-Likelihood Test' result for empirical model 63

APPENDIX 9: Hausman Test's result for empirical model 63

APPENDIX 10: Data for estimating P/E from the empirical model 65

APPENDIX 10: Data for estimating mean of ‘higher than normal’ P/E ratio stocks and ‘lower than normal’ 66

LIST OF ABBREVIATIONS

CAPM Capital asset pricing model

GDP Gross domestic product

HOSE Ho Chi Minh stock exchange

NYSE New York Stock Exchange

OLS Ordinary Least squares

R Required rate of return

S&P 500 The Standard & Poor‟s 500 market index

LIST OF TABLES

Figure 3.1 : Factors affecting P/E raios……….………30 Figure 4.1: The value of average Dividend Payout Ratio and P/E ratio on 2008 -

2016 period.……… 37 Figure 4.2: The value of average Growth Rate and P/E ratio on 2008 -2016 period……… 38 Figure 4.3: The value of average ROE and P/E ratio on 2008 -2016 period

……… 39 Figure 4.4: The value of average Dividend Payout Ratio and P/E ratio on 2008 -

2016 period……….40

LIST OF FIGURES

Table 2.1: Summary of international prior studies 18

Table 2.2: Summary of Vietnam prior study 23

Table 2.3: Summary of independent variables influencing on P/E ratios 24

Table 3.1: Variables measurement in the author's analysis ………30

Table 4.1: The descriptive statistics for all variables 36

Table 4.2: The covariance of variables ……….41

Table 4.3: Breusch-Godfrey Serial Correlation LM Test ……….42

Table 4.4: Heteroskedasticity Test: White's result for empirical model………42

Table 4.5: Pooled OLS 's result for empirical model ………43

Table 4.6: Random Effect Mode 's result for empirical model 44

Table 4.7: Fixed Effect Mode's result for empirical model 45

Table 4.8: Fixed Effect-Likelihood Tesst' result for empirical model 46

Table 4.9: Hausman Test's result for empirical model 47

Table 4.10: Summary of research’s results (5% significance) 49

Table 4.11: Summary of research’s results (5% significance) 50

Table 4.12: ‘Higher than normal’ P/E ratio stocks and ‘lower than normal’ P/E ratio stocks in 2016 Error! Bookmark not defined. Table 4.13: Mean of ‘higher than normal’ P/E ratio stocks and ‘lower than normal’ P/E ratio stocks in 2016 52

Chapter 1 : INTRODUCTION 1.1 THE NECESSITY OF THE THESIS

Security analysis and asset selection represents a crucial stage of the investment process The general approach in security analysis and asset selection is

to identify underpriced and overpriced stocks Apart from fundamental analysis based on applying appropriate stock valuation model to estimate the intrinsic value

of the stock, the price-earnings (P/E) ratio or earnings multiplier approach is also used as an alternative approach by security analysts and investors in seeking worthwhile stocks

The P/E ratio of a stock is referred to as the ratio of its market price per share

to its earnings per share In reality, this ratio is widely used measure of the expected performance of firms (Anderson and Brooks (2005a and 2005b)) The P/E ratio indicates the price that investors are willing to pay for each dollar of earnings Investors may value the earnings of one firm differently from that of the other, thus P/E ratios differ across different firms, across different industries, and even across different markets However, the P/E ratios may imply different stories and different market perspectives A high P/E ratio stock may be considered as an overpriced stock or it may reflect investors‟ expectation about the high growth On the other hand, a low P/E ratio stock may reflect investors‟ pessimism about the future growth of the firm‟s earnings or it is may be an underpriced stock

In the practice of investment activities, investors attempt to find ways to generate higher expected returns while maintaining the same level of risk Over the past decades, many investors have not believed in the hypothesis of market efficiency Apart from different types of technical analysis, the P/E approach in security analysis has been conducted in this perspective with an aim of obtaining abnormal returns Many investors follow the principle of price-earning effect by investing in low P/E ratio stocks with a view that these stocks are undervalued This

investment strategy seems to be consolidated by the findings of many significant empirical studies For example, Nicholson (1960) provides evidence that low P/E ratio stocks on an average yield higher returns than those of high P/E stocks Such difference is known as the value premium The results of Basu‟s test (1977) of the investment strategy of selecting low P/E stock also suggest that investment in low P/E stocks may provide positive abnormal return Basu‟s methodology is as follows: (i) Ranking 650 stocks in accordance with their P/E ratios; (ii) setting up five equal-sized P/E ranked portfolios from the stocks; (iii) calculating monthly rate

of returns of each portfolio over the subsequent year; (iv) running a regression of the monthly excess returns of each portfolio against the corresponding excess market index return The results of Basu‟s test indicate that portfolios containing low P/E ratio stocks provided alphas or higher risk-adjusted returns as compared to those P/E ratios

However, the P/E ratio reflects investors‟ expectation about the growth potential of a stock as well as the risk involved (Bodie, Kane and Marcus, 2014) Low P/E ratios may be due to its low growth potential of due to its high risk involved P/E ratios may also differ across industries From this perspective, an investment in a low P/E ratio stock may not be always a guarantee for abnormal returns To this point, a question may come out in one‟s mind is: what are relevant approaches for identifying “really” low P/E ratio stocks or “relatively” low P/E ratio stocks? One approach that can be used for identifying “relatively” low P/E ratio stocks is to conduct a regression of firms‟ P/E ratio on their determining factors The obtained empirical model is then used to estimate a firm‟s „normal‟ P/E ratio Based on the identified a firm‟s „normal‟ P/E ratio, investors can identify whether the firm‟s stock is over or under priced However, this approach may not be relevant for individual investors in general due to its technical natures

Apart from the empirical studies investigating whether portfolios consisting

of low P/E ratio stocks provide better than average returns, a number of recent studies focus on examining the factors affecting the P/E ratios (Loughlin (1996),

Sushil Kumar Bhadu and Warne (2009), Shamsuddin, Hiller (2004), Muhammad, Azam, (2010)) However, these studies do not explicitly point to the possibility of using the obtained models as a tool for identifying “relatively” low P/E ratio stocks

In the Vietnamese context, the country‟s stock market is still a young one, though experiencing almost twenty years of development The market has grown significantly and rapidly in both the number of stocks listed and value of market capitalization However, making investment decisions based on "herd mentality" rather than professional strategies is still popular among Vietnamese investors, especially individual investors The “herd mentality” or “crowd psychology” was clearly seen when the country‟s stock market was hit by the global financial crisis Nguyen Kim Tung (2008) According to Nguyen Kim Tung (2008), about two-third of total investors on Vietnamese stock market were individual Vietnamese individual investors tend to possess limited financial knowledge and have no skills

of portfolio management They are likely to imitate the crowd‟s investment decisions without their own careful judgement and consideration For example, there was a warning in 2007 that the prices of Vietnamese stocks were higher than their intrinsic value by 30%, but individual investors continued making buy orders When the market was on the way of plummeting since March 2007, local investors appeared to be in a terribly panicking mood, having sold out stocks as fast as possible

The “herb mentality” phenomenon actually has contributed to creating instable situations on the Vietnamese stock market As a result, the stock prices may diverge considerably from their intrinsic value This reality suggests the need of improving investment knowledge and skills for individual investors In this course, developing analytical investment tools such as a sound empirical model of determinants of P/E ratios would be conducive for individual investors in conducting security analyses and making rational investment decisions

In sum, the literature on P/E ratios seems to suggest that low P/E stocks, or low price multiple stocks, are considered to be undervalued Thus, investors can form an

investment strategy based on this price-earning effect In other words, constructing

a portfolio consisting of stocks with low multiples can help investors obtain abnormal returns This idea would give rise to a motivation to conduct a research to test whether it is a good sense for investors to make investment in low P/E ratio stocks on Vietnamese stock exchange But before conducting this investigation, it is also very important to uncover a sound approach for identifying low P/E stocks

Therefore the topic of this dissertation is “Factors affecting price-to-earnings ratios

of listed companies on Ho Chi Minh City Stock Market: Period 2008-2016” The

author expects that the obtained empirical model of factors affecting earning (P/E) ratios for the case of listed stocks on Ho Chi Minh City Stock Market can be a good reference model for identifying “normal” P/E ratios, which in turn serve as benchmarks for identifying over or under valued firms when comparing with their actual P/E ratios using the actual market data and the data from the firms‟ financial statements

price-to-1.2 RESEARCH OBJECTIVES

The objective of this dissertation is to examine the factors affecting P/E ratios of companies listed on Ho Chi Minh City Stock Market using a multiple regression model The obtained regression model is then expected to be used as a reference tool for investors in identifying a firm‟s „normal‟ P/E ratio By comparing the „normal‟ P/E ratio with that actual P/E ratio computed based on the market data and the data from the firm‟s financial statements, investors may find whether the firm is over or under priced In other words, the obtained model may help investors identify worthwhile stocks for investment In a sense, the research approach adopted

in this dissertation is based on a cross sectional analysis of P/E ratios of listed companies on Ho Chi Minh City stock market

1.4 THE RESEARCH SUBJECT AND SCOPE OF THE STUDY

1.4.1 The research subject

The subject of this dissertation is to examine the factors affecting earnings ratios of listed companies on Ho Chi Minh City Market for the period 2008-2016

price-to-1.4.2 Scope of the study

This research is conducted by using data collected from a sample of 123 companies listed on Ho Chi Minh City for the period 2008-2016 Finance companies and commercial banks are not included in the sample due to their specific business characteristics Furthermore, the sample only includes those listed companies that pay dividends regularly

1.5 RESEARCH METHOD

This dissertation mainly adopts an approach of quantitative analysis In order

to answer the first research question, based on the identified theoretical framework and the previous empirical studies, the author identifies relevant factors as determinants of P/E ratios to put into a regression multivariate model The identified regression model is then estimated by methods that are suitable for panel data such as Pooled OLS model, Fixed Effect Model (FEM) and Random Effect Model (REM) Different tests are also conducted for selecting the most appropriate estimated model and for the robustness of the estimated model

To answer the second question, the author conducts the following steps to test whether the obtained empirical model can be used as a reference tool for selecting worthwhile stocks The „normal” or „adjusted‟ P/E ratio for each company

in the sample will be estimated based on the obtained empirical model Based on these „normal‟ P/E ratios estimated at year-end 2016, stocks will be classified into two groups The first group consists of those stocks having the actual P/E ratios lower than their „normal‟ P/E ratios The other group consists of those stocks having the actual P/E ratios higher than their „normal‟ P/E ratios The author will assume

an investment horizon of one year from the end of 2015 to the end of 2016 If the average return of the „lower than normal‟ P/E ratio stocks is higher than that of the

„higher than normal‟ P/E ratio stocks, it can be to a certain extent to conclude that the obtained model is a relatively effective model for selecting worthwhile stocks for investing

1.6 RESEARCH CONTRIBUTION

1.6.1 The scientific contribution

This dissertation contributes to providing empirical evidence on the factors effecting P/E ratio of companies listed on Ho Chi Minh City Stock Market

1.6.2 The practical contribution

The obtained empirical model can be used as a reference model for investors

in analyzing and picking worthwhile stocks in their portfolio

Chapter 3 describes the adopted regression multivariate model, reports the measurements of variables and sources of data, presents the hypotheses on the effects that the identified factors may have on the P/E ratios, and discusses about the used estimation methods as well as the relevant tests In this chapter, the procedure that is employed to test whether the obtained model is a suitable model for identifying worthwhile stocks for investment is also described

Chapter 4 presents mainly the estimated results obtained from the regression analysis of the model as described in chapter 3, together with the results of relevant fit tests of model This chapter also provides relevant discussions of the findings and makes comparisons with those finding of previous studies The results of testing of the effectiveness of the obtained empirical model in identifying worthwhile stocks for investing are also reported in this chapter

Chapter 5 draws conclusions and presents recommendations to investors It also discusses the limitations of the dissertation and recommends an agenda for possible future research

Chapter 2 : LITERATURE REVIEWS 2.1 Introduction

The objective of chapter 2 is to provide a literature review on P/E ratios and impacting factors from the theoretical perspective as well as from the empirical perspective On one hand, the review of the literature helps to make it clear that the P/E ratio approach is an essential approach in security analysis for the purpose of security selection On the other hand, it helps argue for the necessity of conducting this research to provide an empirical model of factors affecting P/E ratios of stocks listed on Ho Chi Minh City Stock Market, which can be used as a reference tool for security analysis and security picking The review of literature also provides arguments as well as an available analytical framework for formulating the relevant regression model

The remainder of Chapter 2 is structured into three sections Section 2.2 presents the theoretical background, including an introduction of the concept of P/E ratio and related theories on the factors affecting P/E ratios Section 2.3 reviews the empirical literature on the factors affecting P/E ratios in the international context as well as in the Vietnamese context The necessity of conducting this research is discussed at the end of this section Section 2.4 provides a conclusion for the chapter

2.2 Theoretical background

2.2.1.The concept of P/E ratio

Price per earnings ratio (“P/E” ratio) is one of the most commonly used key indicators of the desirability of particular stocks for investment by many value fund managers In general, P/E ratio is simply the ratio between the market price and the equity‟s earnings per share (see equation (2.1))

Where:

P: the general market price per share

EPS: the common stock annual earnings per share

The P/E ratio described as above is sometimes also referred to as the price multiple, which means the amounts of dollars that investors are willing to pay for each dollar of the firm‟s earning For example, if a stock is currently traded at a P/E ratio of twenty-five, its current market price is twenty-five times higher that the earning per share of that stock In other words, investors are assumed to be willing

to pay 20 dollars for each dollar of earning Correspondingly, if the P/E ratios of the two firms differ, it means that the market values the earnings differently from that

of the other

2.2.2 Use of P/E ratio for stock selection

P/E ratio is one of the most widely used indicators of firms‟ performance How is it used in practice for evaluating and selecting stocks? In general, investors can formulate an investment strategy based on price-earning effect The principle is that investors would construct a portfolio consisting of low P/E ratio stocks only in order to obtain abnormal returns This belief is consolidated by the research results

of Nicholson (1960) and the seminal test results provided by Basu (1977) Nicholson (1960) provides evidence that low P/E ratio stocks on an average yield higher returns than those of high P/E stocks Such difference is known as the value premium Basu (1977) also indicates that portfolios containing low P/E ratio stocks provided higher alphas or higher risk adjusted returns These results suggest that investing in low P/E ratio stocks can provide future positive abnormal returns However, as indicated in the subsequent section, P/E ratio reflects investors‟ expectation about the growth potential of a stock as well as the risk involved (Bodie, Kane and Marcus, 2014) Low P/E ratios may be therefore due to its low growth potential of due to its high risk involved P/E ratios may also differ across industries From this perspective, an investment in a low P/E ratio stock may not result in an abnormal return To this point, a question may come out in one‟s mind is: what are relevant approaches for identifying “really” low P/E ratio stocks or

“relatively” low P/E ratio stocks?

Several approaches for identifying low P/E ratio stocks are documented in the literature (Bodie, Kane and Marcus, 2014) First, investors often determine the expected P/E ratio for the next year and decide if a stock in undervalued or overvalued This expected P/E ratio is mostly based on the expected earnings growth and considered as a „justified‟ P/E ratio for that share If the current P/E ratio is lower than that expected P/E ratio, the stock is considered to be currently undervalued and vice-versa There is an issue of this approach is that it requires investors to estimate accurately the expected price as well as the expected earnings The second approach is empirical in nature Investors would conduct a cross sectional analysis of P/E ratios by collecting market data and data from firms‟ financial statements and regression firms‟ P/E ratio on their determining factors The obtained empirical model is then used to estimate a firm‟s „normal‟ P/E ratio Based on the identified a firm‟s „normal‟ P/E ratio, investors can identify whether the firm‟s stock is over or under priced If a firm‟s actual P/E ratio is higher than its

„normal‟ P/E ratio, then the firm‟s stock is overvalued In this case, investors may consider selling this stock if it is currently hold in their portfolios In contrast, if a firm‟s actual P/E ratio is lower than its „normal‟ P/E ratio, then the firm‟s stock is undervalued In this case, investors may consider including this stock in their portfolios

There are several issues with this approach First, the investors should have a good understanding of possible factors affecting P/E ratios The investors should also possess considerably good skills in conducting econometric analyses Apart from these, the task of collecting and processing data is also time consuming Thus, this approach may be relevant for institutional investors, but may not be suitable for individual investors, especially for those participating in young stock markets in developing countries like Vietnam The next subsection will review theories on the possible determinants of P/E ratios

2.2.3 Factors affecting P/E ratios

In general, the P/E ratio of a stock is an indicator reflecting an interaction of

many different factors Both the components of the P/E ratio, price and earning, are jointly driven by the firm-specific factors, the industry specific factors as well as the macroeconomic factors However, based on the dividend discount model, one can derive clearly the fundamental factors that can have impacts on the P/E ratios

The dividend discount model, first published in the book "The Theory of Investment Value" by John Burr Williams (1938), represents a fundamental method

for valuing intrinsic value of a stock or business, exclusive of current market condition When purchasing a stock, investors expect two kinds of cash flows One

is the dividend received during the holding period and the other is the capital gain due to the increase in the price up to the end of the holding period In order to estimate the price of a stock, one may have to estimate the intrinsic value of the stock The intrinsic value of a stock is determined by the stream of dividends received in the future and the risk involved in that stream of dividend Equation (2.2) can be generally used for estimating the intrinsic value of any stock at a point

V: intrinsic value of the stock;

: dividend for the year;

R: discount rate, namely required rate of return;

t: the year for dividend payment

Assume the market is efficient, the share price should be equal to the intrinsic value

of the stock, then equation (2.2) becomes (2.3):

∑

Where :

: purchase price of the stock;

: dividend for the year;

R: discount rate, namely required rate of return;

t: the year for dividend payment

There are three versions of dividend discount model: (i) the zero growth rate model; (ii) the constant growth rate model; and (iii) the multiple growth model Assume the dividend grows steadily at the rate of g, the constant dividend growth model can be derived as follows:

∑

When the growth (g) is not fluctuate and R > g at the same time, then equation (2.5) will be:

Where

: purchase price of the stock;

: dividend at the purchase time;

: dividend for the year;

R: discount rate, namely required rate of return; g: the growth rate of dividend

Generally, many previous studies apply Dividend Discount Model to estimate P/E ratio Required rate of return and the growth rate of dividend that are significant measure to reflect the P/E ratios volatility

Equation (2.5) is popularly known as Gordon Model (Gordon, 1959)

By combining equation (2.1) and equation (2.5), the P/E ratio now can be derived as in equation (2.6):

⁄

The ratio D 1 /E 1 is actually the payout ratio, b Equation (2.6) can be written

as equation (2.7)

⁄

⁄ Where:

: purchase price of the stock;

: earnings per share (EPS) of the year after purchase;

: dividend for the year;

b: return on dividend;

R: discount rate, namely required rate of return;

g: the growth rate of dividend

Equation (2.7) indicates clearly four fundamental factors that affect the estimated P/E ratio First, the P/E positively depends on the payout ratio The higher the pyaour ratio is, the higher the P/E ratio is Second, the P/E ratio is also positively dependent on the expected growth rate (g) Thus, the higher the expected growth rate (g) is, the higher the P/E ratio is Finally, the P/E ratio is negatively dependent on the required rate of return (R) Accordingly, the higher the required rate of return (R) is, the lower the P/E ratio All the mentioned relationships are noted given the assumption of „other factors being equal However, the three mentioned determinants of P/E ratios are interrelated As discussed in the previous section, P/E ratios reflect investors‟ expectation about the growth potential of a stock, as well as the risk involved One may come up with such a question: Can firms increase their payout ratios to increase the market prices of their stocks? If the payout ratio is increased, the future growth prospects may be affected Thus, an increase in the payout ratio may not result in an increase in the stock price

Similarly, investors may raise such a question: Does rapid growth lead to an increase in the P/E ratio? The answer is not certain because rapid growth may affect the riskiness of earnings

Examining equation (2.7) closely also reveals different other possible factors affecting P/E ratios For example, as a determinant of P/E ratio, the growth rate of a company itself is determined by the product of its return on equity (ROE) and the retention ratio (1-b) The retention ratio is referred to as the percentage if the company‟s earning plowed back to the company As the return of equity is driven

by a range of factors, which are identified from the theoretical perspective as well as the empirical perspective Similarly, the required rate of return (R), an important determinant of P/E ratio, reflects investors‟ optimism and pessimism about the company‟s business perspective Investors may require higher rate of return (R) due

to their perception of higher business risk associated with the company‟s business Higher required rate of return (R) may also be generally due to an increase in in interest rates or inflation rates Thus, P/E ratios and interest rates are inversely related (Bordie, Kane and Marcus, 2014)

In the related empirical literature, equation (2.7) is explicitly or implicitly used for investigating the determinants of P/E ratios In addition to the three fundamental factors, several other firm-specific factors are also employed in the empirical model In general, the measurement of factors/independent variables is quite straightforward However, the required rate of return (R) is normally not measured directly, but proxied by a measure of risk

Two measures of risk are typically documented in the empirical literature One is the standard deviation of earnings and the other is the beta It appears that the beta may represent a better proxy for the required rate of return (R) due to their close relationship derived from the capital asset pricing model (CAPM) The CAPM

is used by investors to estimate the required rate of return for any risky asset Equivalently, this model is also used by managers to estimate the cost of equity for the purpose of making budgeting decisions The CAPM is built on the

groundbreaking paper of Harry Markowitz (1952), in which he describes how to select a portfolio using the mean-variance rule The CAPM is represented by equation (2.8)

( )

Where

: required rate of return

: risk free rate

: the systematic risk

: the market premium

As indicated in equation (2.8), the required rate of return of any stock is determined by three factors: (i) the pure time value of money measure by the risk-free interest rate, ; (ii) the reward for bearing systematic risk measured by the market risk premium, ; and (iii) the amount of systematic risk measured by beta Equation (2.8) suggests that the required rates of return on stocks differ due to different systematic risks measured or due to different betas

2.3.Previous empirical studies

This dissertation is contributed to the literature review on P/E ratios in two ways First, studies use obtained empirical model to test the ability to bring higher earning when the investors hold the low P/E stock Second, the are focus on

identifying and analyzing the factors effecting P/E ratios

2.3.1 Previous empirical studies in the international context

Use of P/E ratio for stock selection

Nicholson (1960, 1968) specifically studies price-earnings ratios of 100 listed companies on the New York Stock Exchange over a five-year period during 1939-1959 In this paper, an OLS multiple regression is employed to investigate the relationship between P/E ratios and firms return Specially, every five years,

Nicholson evaluates the portfolio and leads the result that the low P/E portfolio

evaluate 14.7 times the investor‟s initial investment.(Nicholson S.F 43-45) A few

years later, in 1968, he extended his study done in 1960 by including another 89

industrial stocks, totaling 189 stocks Depending on the size of the P/E ratio he

divided these 189 stocks into five different groups, and again, followed them over

five-year periods from 1937-1962 He found that the portfolio with P/E below 10

averaged 131% return compared to the portfolio with multiples above 20 averaging

only 71% over the first seven years (Nicholson S.F 105-109)

Basu (1977) - an economist – follows Nicholson‟s studies and identified that

low P/E ratio stocks had an approximately higher return than other stocks with high

ratios (Siegel 149-150) His findings uses the data sources of 1400 companies on

the New York Stock Exchange (NYSE) from 31st December 1956 to March 1971

First, the end of December would be the fiscal year-end of the company Second,

the firm was actually listed and traded on the NYSE, and third appropriate data

fiscal statements and investment were available

Factors affecting P/E ratios

P/E ratios plays an important roles in many perspectives, especially in

company valuation Abul and Shamsuddin, Hiller (2004)‟s study is supported to this

view after studying the time series by quarter, investigate the fundamentals of the

P/E ratio in the Australia stock market in period of 1/1984 – 3/2001 The research

result demonstrates that P/E ratio is an increase in function of the dividend payout

ratio

Bhadu and Warne (2009) analyze 243 companies data in Bombay stock

market from 2001-2007 evaluate the level of effect that each parameter of factors

have on P/E ratio in Indian capital market Moreover, the research performed the

result from the multiple regression process, which is concluded that the market price

volatility is a vital factor to determine P/E ratios

Using the data of 35 companies paying dividends on the Karachi stock

market during the period 2000-2008, Muhammad and Azam‟s work (2010) is

Comment [WU1]: New York Stock Exchange

employed that dividend payout ratio is positively correlated to the P/E ratio In this paper, an OLS multiple regression is employed to clarify the relationship between P/E ratio and firms return

In line of this study, other empirical evidence suggests a negative correlation between payout ratio and P/E ratios This opinion is supported by Johan & Filip (2007) showing that the result that the variables such as dividend yield, company size, financial leverage affecting P/E ratios It is find that the dividend has a negative impact on the P/E ratio means that the increase of this variable, the P/E ratio decreased Johan & Filip (2007) uses multiple regression to analyze the data sources on Swedish stock exchange over a period from 1998 to 2007

After that, Reilly et al (1983) studied the time series about the P/E relationship of S&P 500 for the period from 1963-1980 by using multiple regression The authors published the conclusion that the higher dividend payment, the higher P/E ratios and the higher P/E ratios

Gill (2003) studies the Indian market, taking the BSE 100 data from 1997 to

2001 to conclude that the higher earnings, the higher P/E ratios Companies with high and sustainable Earning Per Share (EPS) growth should be considered for investment

Loughlin (1996) conducts to studies for S&P 500 and comes up with conclusion that the expected return is positively correlated with the P/E Similarly, White (2000) uses S&P 500 data from 1926 to 1997 for applying multiple regression The result shows that the firm's earning growth is significantly related with P/E ratios

Afza (2012) studies the relationship between P/E ratios and other independent variables, including: dividend payments, financial leverage, market earnings, market price volatility, and firm size Based on an OLS regression for the aggregate data of 25 companies listed on the Karachi stock exchange for the period 2005-2009 This paper takes into account the volatility in the Pakistani stock market during the examined period

Puwanenthiren Premkanth (2013), employed a study of 30 companies listed

on the Colombo Stock Exchange from 2007 to 2011 Their finding shows that the dividend payout ratio has a positive and statistically significant impact on the P/E ratio

Graham and Dodd (1934) points out that the value of a stock is multified of its current earnings It depends on both macro (confidence on the stock) and micro factors (the property and history of the company) They considered that P/E ratios reflect the information on firms performance It can predict the future growth of the companies The average earnings must be accounted when valuing stock price, 16 times of the average earnings is the top price investors can afford

Penman (1996) comes up with conclusion that the low correlation between P/E ratios and observed long-term growth and finds a stronger relation using a more recent sample period (1968-1985) and a measure of earnings growth that corrects for dividends paid The correlation between portfolio-level P/E ratios and observed growth show higher correlations even nine years later

The new measure of stock valuation, which is developed by Azhar, Osman and Parinduri (2009), “it has 3 properties: symmetric, proportional and non-invariant A new technique presented for the empirical analysis of stock estimation and capital market ratios” (Gottwald, 2012, 22) Firstly, the long - term relationship between stock market returns and the P/E ratio is evaluated This evaluation demonstrates the impact of P/E ratios on individual stock Besides, it also examines how the relationships does between individual stock returns after controlling for the systematic risk in a CAPM regression This paper used database of the Standard & Poor´s index from 1872 to 2008 to conduct the survey

Table 2.1: Summery of international prior studies

Use of P/E ratio for stock selection

Nicholson (1960, 1968) Using OLS multiple

regression of 100 listed companies on the New York Stock Exchange over a five-year period during 1939-1959 and

189 stocks from 1937 to

1962

The low P/E portfolio evaluate 14.7 times the investor‟s initial investment (1939 – 1959) From 1937 to 1962, the portfolio with P/E below

10 averaged 131% return compared to the portfolio with multiples above 20 averaging only 71% over the first seven years Basu (1977) Using the data sources of

1400 companies on the New York Stock Exchange from 31st December 1956 to March

1971

Low P/E ratio stocks had

an approximately higher return than other stocks with high ratios

Factors affecting P/E ratios

Reilly et al (1983) Multiple regression of

S&P 500‟s data from

1963 to 1980

The higher dividend payout and , the higher P/E ratio P/E is negatively influenced with an increase

in the rate of business failure, risk-free income, inflation fluctuations and income fluctuations

Kane et al (1996) Multiple regression of

volatility, inflation and industrial production Real interest rates and dividend rates have the negative influence on P/E but not statistically significant

Loughlin (1996) Multiple regression of

S&P 500 data from 1926

to 1997

Expected dividends and profits are positively correlated with the P/E ratio

Treasury bill rates or bond rates of returns are inversely proportional to the P/E ratio

Vahid Faezinia (2012) Multiple regression in

market evidence of Tehran

Variable interest rates, dividend yields and inflation are inversely related to the coefficient P/E

Positive correlation between P/E ratio and variable growth rate, risk, debt ratio, ROE and company size

Dr Talat Afza (2012) The company is listed on

Karachi stock exchange for the period 2005 -

The dividend payout ratio

is the most important determinants of P/E ratios

2009 for aggregate analysis as

well as time series analysis Sushil Kumar Bhadu

and D.P Warne (2009)

Analysis of the 243 companies listed on the stock market

Block Bombay during 2001-2007

Dividend payment rates, financial leverage, market earnings, market price fluctuations, earnings growth and company size affect P/E ratios The volatility in market prices and company size are the

Dividend payout ratio has positive impact on with P/E ratios, which supports the conclusion of Amoako-Adu, Ben & Smith, Brian (2002)

Gill (2003) Using Pooled OLS in

Indian market research, taking the BSE 100 data for the period 1997-2001

High profit leads to higher P/E ratio

Companies with high and sustainable income should

be considered for investing‟s choice

Johan & Filip (2007) Multiple regression of The rate of dividends and

interest rates have an impact on the P/E ratio Similarly, the negative P/E ratios means that these

variables increase, the P/E ratios decreases

Afza (2012) OLS regression model

from 25 companies listed

on the Karachi stock exchange for the period 2005-2009

The dividend payout ratio

2011

The dividend payout ratio had a positive and statistically significant impact on the P/E ratio

2.3.2 Previous empirical studies in the Vietnamese context

Not only foreign studies, but also Vietnamese context have contributed to provide the empirical evidences on identifying factors influencing on P/E ratios Their findings form a robust foundation for both local and foreign investors to compare companies Consequently, have the better choice to invest in profitable stocks Some valuable researches are involved in P/E ratios and investment decision, such as:

Vu Phuc Thinh (2008) shows that investors can compare the relatively similar companies to estimate the stock valuation when using the comparative method with P/E ratios It also reflects the positively impact of market capitalization, dividend payout ratio on P/E ratios However, risks (Standard Deviation, Beta) , ROE and the growth rate have a negatively effect on P/E ratios Hua Quang Tri (2013) is using experimental results using regression of data tables and cross-sectional data for 2008-2012, find out that factors such as dividend payout ratios, Tobin's Q, Size of firms, fluctuation of the market price and financial

leverage ratios have positively influence on P/E ratios of non-financial companies listed on HOSE

Pham Hong Thai, Nguyen Thanh Den (2017) use Pooled OLS, FEM, REM

to study 191 non-financial listed companies in Vietnam during 2010-2015 It comes

up with conclusion that growth rate have positive impact on P/E ratios

Truong Dong Loc (2014) points out that payout ratios, earning per share, consumer price index have positive influence on P/E ratios The research uses multiple regression to analyze 20 listed companies on HOSE

Table 2.2: Summary of Vietnam prior study

Vu Phuc Thinh (2008) OLS regression model

from VnIndex from 2002

to 2008

Market capitalization, dividend payout ratio have positive impact on P/E ratios

Deviation, Beta) , ROE, and the growth rate have

an negative affection on P/E ratios

Hua Quang Tri (2013) Using multiple regression

(Pooled OLS, FEM, REM) to analyze non-financial firms listed on HOSE during the period

of 2008-2012

When the dividend payout ratio, Tobin‟s Q, the fluctuation of the market price increase, the P/E ratios also rise up The low financial leverage and the size of firms can lead to the rise

of P/E ratios

Pham Hong Thai,

Nguyen Thanh Den

(2017)

Using Pooled OLS, FEM, REM to study 191 non-financial listed companies in Vietnam during 2010-2015

Growth rate have positive impact on P/E ratios

Truong Dong Loc (2014) Using multiple regression

(Pooled OLS, FEM, REM) to analyze 20 listed companies on HOSE

Payout ratios, earning per share, consumer price index have positive influence on P/E ratios

Although there are variety studies about the relation between P/E ratio and identified factors or the factors affecting the investment decision, none of them have both views about the relationship between P/E ratios and investor decisions Then, this dissertation will discuss and make clear the mentioned studies above

Table 2.3: Summary of independent variables influencing on P/E ratios

Variables Abbreviation Expected

2.4 The position of this dissertation

To the author‟s best knowledge, there are only a few studies conducted for

investigating the factors impacting the P/E ratios of Vietnamese stocks However,

most of the studies under review are not published yet This dissertation is expected

Shamsuddin, Hiller (2004) Muhammad, Azam, (2010)

Vu Phuc Thinh (2008) Hua Quang Tri (2013) Truong Dong Loc (2014)

- Kane et al (1996)

Vahid Faezinia (2012) Afza (2012)

Puwanenthiren Premkanth (2013)

Dividend

Growth Rate

GROWTH + Reilly et al (1983)

Loughlin (1996) Muhammad, Azam, (2010) Pham Hong Thai, Nguyen Thanh Den (2017)

- Vu Phuc Thinh (2008)

Comment [TN3]: Thêm chữ “most of” vì có một

vài nghiên cứu đã đăng báo

to contribute to the literature on P/E ratios in two ways First, it provides empirical evidence on the factors effecting P/E ratios for the case of listed companies on Ho Chi Minh City Stock Market Second, the obtained empirical model can be used as

a reference model for investors in analyzing and picking worthwhile stocks in their portfolio

2.5 Chapter conclusion

This chapter presents a review of literature on the P/E ratios and impacting factors From the theoretical perspective, it is clear that P/E ratio approach or earnings multiplier approach is an essential approach in security analysis for the purpose of security selection The dividend discount model indicates three fundamental determinants of P/E ratios, namely: the payout ratio, the growth rate and the required rate of return Examining the dividend discount model closely also reveals other different factors that can affect P/E ratios In the international context, empirical studies on P/E ratios can be grouped into two research strands One strand

is conducted to test the hypothesis that investment in low P/E ratio stocks provides higher returns than that in high P/E ratio stocks The other strand consists of studies investigating the determinants of P/E ratios In the Vietnamese context, very rare empirical studies are found With an aim of providing a reference analytical tool for Vietnamese investors in security analysis and asset selection, this dissertation is conducted to provide an empirical model of factors affecting the P/E ratios of companies listed on Ho Chi Minh City Stock Market Correspondingly, this dissertation contributes to providing additional evidence on the determinants of the P/E ratios using a sample of stocks listed on Ho Chi Minh City Stock Market

Chapter 3 : RESEARCH METHOD AND DATA 3.1 Introduction

The objective of this chapter is to present method and data used to answer the research question This dissertation is used multiple regression ( Pooled OLS, Fixed Effect Model, Random Effect Model) from database of 123 listed companies

on HOSE on the period 2008 – 2016 Besides, the author build up empirical model that investment in low P/E ratio stocks provides higher returns than that in high P/E ratio stocks by using the comparison between mean of overvalued stock return and undervalued ones Stock return is calculated by using the stock price at the beginning and at the end of 2016

3.2 Estimation model

Based on literature review in the chapter 2, it is can clear that there are 4 variables affecting P/E ratios There are Dividend Payout Ratio, Dividend Growth Rate, Beta, Return on Equity (ROE)

In this dissertation, the dividend growth rate is calculated by the author following DDM Beta comes from the stock price database Dividend payout ratio and ROE are taken from financial statement of commercial banks

This section also presents the estimation model used in this study This estimable model identifies the effects of dividend payout ratio, dividend growth rate, beta, return on equity factors on P/E ratios And, the model is specified as follows:

Y it = β 0 + β 1 X 1it + β 2 X 2it + β 3 X 3it + β 4 X 4it

Where

Yit is the P/E ratio i at time t;

β0 is intercept;

β1…4 are regression coefficients

X1…4 are independent explanatory variables

The classification of variables for this empirical dissertation is based especially on previous studies, particularly, the author‟s approach follows to Afza (2012) Specially, there are several identified variables selected for this study including 1 dependent variable – Price-to-earnings (P/E) ratio and 4 independent variables Independent variables are Dividend Payout Ratio (Payout) , Dividend Growth Rate (g) , Beta, Return on Equity (ROE)

3.3 Measurement of variables and hypothesis

3.3.1 Measurement of variables

After chapter 2, it can be clearly seen that which factors affecting P/E ratios

In this thesis, the value of P/E ratio has influenced to investors decision which is calculated by the author following the study of Afza (2012) with P/E ratio as dependent variable and explanatory variables are measured from financial statement

of companies and HOSE

Dividend Payout Ratio

In order to measure this variable, Dividend Payout Ratio is the percentage of

a company's earnings paid out to investors as cash dividends

The part of the earnings not paid to investors is left for investment to provide for future earnings growth Investors seeking high current income and limited capital growth prefer companies with high Dividend payout ratio

Dividend Growth Rate

The dividend growth rate is a component of the Dividend Discount Model (DDM) which values a stock on the basis of expected dividends, discounting them

to their present value and determining if a stock trades over or under its fair value

Beta

The beta factor (β) is a systematic risk measure It represents the correlation between the level of risk of an individual asset or a portfolio over the level of the overall market risk

The author calculate beta by collecting the stock price at the end of days in each year The author believe that using daily return will help increase statistical capacity and be provided more observations for estimating beta This dissertation uses the Microsoft Excel scope function to calculate beta in each year

Table 3.1: variables’ measurement in the author’s analysis

P/E