Disappearing dividends: the case of Thai listed firms

Southern Cross University
[email protected]
Theses
2004
Disappearing dividends: the case of Thai listed
firms
Malinee Ronapat
Southern Cross University
Publication details
Ronapat, M 2004, 'Disappearing dividends: the case of Thai listed firms', DBA thesis, Southern Cross University, Lismore, NSW.
Copyright M Ronapat 2004
[email protected] is an electronic repository administered by Southern Cross University Library. Its goal is to capture and preserve the intellectual
output of Southern Cross University authors and researchers, and to increase visibility and impact through open access to researchers around the
world. For further information please contact [email protected]
DBA Thesis: Malinee Ronapat
DISAPPEARING DIVIDENDS: THE CASE OF THAI LISTED FIRMS
MALINEE RONAPAT
BBA (Finance and Banking), Assumption University
MSc (International Financial Markets), University of Southampton
A THESIS SUBMITTED TO PARTCIALLY FULFILL THE REQUIREMENTS
FOR THE DEGREE OF DOCTOR OF BUSINESS ADMINISTRATION
JUNE 2004
21
DBA Thesis: Malinee Ronapat
Certificate
I certify that the substance of this thesis has not been submitted for any other degree
and is not currently being submitted for any degree.
I also certify that to the best of my knowledge any help received in preparing this
thesis, and all the sources used, have been acknowledged in this thesis.
………………………………………..
Malinee Ronapat
22
DBA Thesis: Malinee Ronapat
ACKNOWLEDGEMENTS
I owe a debt of gratitude to a number of people who helped, supported and encouraged me
during my candidature for this degree.
I am very grateful to my supervisor, Associate Professor Dr. Michael David Evans, who
inspired and encouraged me. This thesis could not have been completed without the help,
suggestions, guidance, patience and wisdom he has given to me whilst preparing this
thesis.
I also acknowledge Dr. Mark Manning for his support with statistical techniques and
suggestions.
The staff of the International Office and Graduate College of Management are also
sincerely thanked, particularly Sue White, Heike Kerber and Rosemary Graham.
Lastly, I would like to express deepest gratitude to my parents, Narong and Maria
Ronapat, for providing me with the priceless opportunity of undertaking my DBA degree
at Southern Cross University (Australia) and for their patience, endless support, love and
caring throughout.
23
DBA Thesis: Malinee Ronapat
ABSTRACT
The Stock Exchange of Thailand (SET) is an important source of funds for firms and
provides opportunities for investors. However, the economic boom of 1990-1996, the
Asian Economic Crisis and the recession of 1997-2002 have affected the performance of
firms listed at SET. The dividend policies of listed firms have also been influenced by these
fluctuations in the business cycle.
This study investigates the phenomenon of disappearing dividends in the developing capital
market of Thailand. It adopts a similar methodology to Fama and French (2001) by
classifying listed firms in line with changes in their dividend polices over the period 1990
to 2002. More specifically, the study explores the characteristics of firms which pay
dividends, non-payers, former payers and firms which have never paid dividends. These
characteristics include profitability, investment opportunities and firm size. The analysis
uses firm characteristics for predicting the dividend policies of listed firms. Changes in
firm characteristics and the propensity to pay dividends are identified in this process.
The analysis suggests that firms which pay dividends tend to be large and highly profitable,
although they possess low investment opportunities. The study also suggests that the
characteristics of firms which paid dividends changed slightly before the crisis of 1997 and
changed markedly during the crisis. However, after the crisis (1998-2002) the
characteristics of firms are similar to those observed before the crisis. This result is
attributed to the fact that some firms have resumed paying dividends after briefly ceasing
this payment during the crisis. More importantly, when firm characteristics are held
constant, the propensity to pay dividends of listed firms declined slightly before the crisis
and declined strongly after the crisis. Consequently, the majority of new firms and many
mature firms do not pay dividends.
The findings of this study are consistent with the results of Fama and French (2001),
particularly with regard to the characteristics of firms and changes in the propensity to pay
dividends. However, this study extends the knowledge on the phenomenon of disappearing
dividends by focussing on a developing economy, Thailand. Finally,
24
DBA Thesis: Malinee Ronapat
this study suggests that investors should consider the characteristics of firms, changes in
these characteristics and the propensity to pay dividends when identifying opportunities for
investment.
Key Words: dividends, capital market, investment
25
DBA Thesis: Malinee Ronapat
GLOSSARY OF ABBREVIATIONS
At
ADB
AEC
AMEX
ANOVA
APT
AUD
BEt
BOI
BOT
BSE
CAPM
DDM
Et
EMH
EPS
FTSE
GDP
GNP
IMF
IPO
Lt
MEt
MOE
MOF
MUA
NASDAQ
NESDB
NPL
NYSE
P/BV
P/E
SEC
SET
SIMS
SP
SPSS
S&P’s
THB
TRIS
UK
US
USD
Vt
Yt
Total Assets
Asian Development Bank
Asian Economic Crisis
The American Stock Exchange
Analysis of Variance
Arbitrage Pricing Model
Australian Dollar
Book Common Equity
Board of Investment (Thailand)
Bank of Thailand
Bangkok Stock Exchange
Capital Asset Pricing Model
Dividend Discount Model
Earnings Before Interest but After Tax
Efficient Market Hypothesis
Earnings per Share
London Stock Exchange
Gross Domestic Product
Gross National Product
International Monetary Fund
Initial Public Offering
Total Liabilities
Market Value of Common Equity
Ministry of Education
Ministry of Finance
Ministry of University Affairs
National Association of Securities Dealers Automated Quotation System
National Economic and Social Development Board
Non-Performing Loan
The New York Stock Exchange
Price to Book Value
Price to Earning
Security Exchange Commission
The Stock Exchange of Thailand
The Stock Exchange of Thailand’s Information Management System
Trading Suspended
Statistical Package for Social Sciences
Standard and Poor’s Rating Services
Thai Baht
Thai Rating Information Service
United Kingdom of Great Britain and North Ireland
United States of America
American Dollar
Market Value of Firm
After-Tax Earnings to Common Stock
26
DBA Thesis: Malinee Ronapat
TABLE OF CONTENTS
Certificate .......................................................................................................................... ii
Acknowledgements .......................................................................................................... iii
Abstract ............................................................................................................................ iv
Glossary of Abbreviations............................................................................................... vi
Tables of Contents........................................................................................................... vii
List of Tables .................................................................................................................... xi
List of Figures................................................................................................................. xiii
CHAPTER 1 – INTRODUCTION TO THE STUDY
1.1
Introduction........................................................................................................ 1
1.2
Background to the Research ............................................................................. 3
1.3
Research Issues................................................................................................... 5
1.3.1 Research Problem ................................................................................... 5
1.3.2 Research Questions................................................................................. 5
1.3.3 Research Hypotheses .............................................................................. 6
1.3.4 Research Objectives.............................................................................. 10
1.4
Justification for the Study ............................................................................... 10
1.4.1 Disappearing Dividends and Investors ............................................... 11
1.4.2 Regulators.............................................................................................. 11
1.4.3 The Research Community.................................................................... 12
1.5
Data and Methodology Overview ................................................................... 12
1.5.1 Key Variables Used in the Study ......................................................... 12
1.5.2 Research Model and Data Analysis Procedure .................................. 16
1.5.3 Testing Periods and Data Sampling .................................................... 18
1.6
Structure of the Thesis..................................................................................... 18
1.7
Conclusion ........................................................................................................ 20
CHAPTER 2 – COUNTRY REVIEW (THAILAND)
2.1
Introduction...................................................................................................... 21
2.2
Country Profile................................................................................................. 23
2.2.1 Demographic ......................................................................................... 24
2.2.2 Social Highlights ................................................................................... 27
2.2.3 Economic Highlights............................................................................. 27
2.2.4 Trading Environment in Thailand ...................................................... 30
2.3
Asian Economic Crisis..................................................................................... 30
2.3.1 Asian Economic Crisis Review ............................................................ 31
2.3.1.1 Financial Sector Weaknesses ............................................................ 31
2.3.1.2 Disequilibrium of Balance of Payment ............................................ 32
2.3.1.3 Contagion............................................................................................ 33
2.3.2 Impact of the Crisis............................................................................... 33
2.3.2.1 Financial Impact ................................................................................ 33
2.3.2.2 Political Impact .................................................................................. 35
2.3.2.3 Social Impact ...................................................................................... 35
2.4
Thai Sovereign Rating (by Fitch, Moody’s and S&P) .................................. 35
2.4.1 Thailand Sovereign Credit Rating Before the Crisis......................... 35
2.4.2 Thailand Sovereign Credit Rating During the Crisis........................ 36
27
DBA Thesis: Malinee Ronapat
2.4.3 Recent Sovereign Credit Rating .......................................................... 36
2.5
2.6
Thai Capital Market ........................................................................................ 37
2.5.1 History of the Stock Exchange of Thailand........................................ 37
2.5.2 Establishment of the Bangkok Stock Exchange................................. 38
2.5.3 Establishment of the Stock Exchange of Thailand ............................ 38
2.5.4 Roles of the Stock Exchange of Thailand ........................................... 38
2.5.5 Regulatory framework of the Stock Exchange of Thailand.............. 39
2.5.6 The Performance of SET...................................................................... 41
2.5.7 SET Vision 2003 .................................................................................... 48
2.5.8 Listing..................................................................................................... 48
2.5.9 Disclosure Procedure ............................................................................ 51
2.5.10 Material Information............................................................................ 52
2.5.11 Stock Information ................................................................................. 52
2.5.12 Information Dissemination .................................................................. 56
2.5.13 Investor Protection................................................................................ 57
2.5.14 Systems Reliability ................................................................................ 57
Conclusion ........................................................................................................ 58
CHAPTER 3 – LITERATURE REVIEW
3.1
Introduction...................................................................................................... 59
3.2
Background on Financial Investment ............................................................ 61
3.2.1 Definition and Objective of Financial Investment ............................. 61
3.3
Capital Market and Efficient Market Hypothesis ........................................ 63
3.3.1 Capital Market and Efficient Market Hypothesis ............................. 63
3.3.2 Efficient Market Hypothesis (EMH)................................................... 71
3.4
The Importance of Dividends ......................................................................... 76
3.4.1 Dividends and Stock Valuations .......................................................... 76
3.4.2 Components of Dividend Policy and
Dividend Payment Procedures............................................................ 78
3.4.3 The Relevance of Dividend Policy ....................................................... 81
3.4.4 Dividend Policy and Share Price Movement ...................................... 85
3.4.5 Other Forms of Dividends.................................................................... 90
3.4.6 Factors Affecting Dividend Policy....................................................... 91
3.4.7 Types of Dividend Policies.................................................................... 93
3.4.8 Firm’s Dividend Payment Decision..................................................... 93
3.5
The Phenomenon of Disappearing Dividends in Capital Markets.............. 95
3.5.1 The Global Phenomenon of Disappearing Dividends........................ 95
3.5.2 The Disappearance of Dividends in Thailand .................................. 103
3.5.3 Thailand in the Light of Research on Capital Market .................... 105
3.6
Research Gaps, Issues and Model Development......................................... 105
3.6.1 Research Gaps and Issues .................................................................. 106
3.6.1.1 Research Gaps.................................................................................. 106
3.6.1.2 Research Issues................................................................................. 107
3.6.2 Development of the Research Model................................................. 112
3.7
Conclusion ...................................................................................................... 113
28
DBA Thesis: Malinee Ronapat
CHAPTER 4 – METHODOLOGY
4.1
Introduction.................................................................................................... 114
4.2
Problem Discovery and Definition ............................................................... 116
4.3
Research Design ............................................................................................. 117
4.3.1 Purpose of the Study........................................................................... 117
4.3.2 Selecting the Type of Research Design.............................................. 118
4.3.3 Selecting the Research Technique ..................................................... 119
4.4
Data Classification and Methodology .......................................................... 121
4.4.1 Dependent and Independent Variables............................................. 122
4.4.2 Data Availability and Period of Study .............................................. 123
4.4.3 Proposed Methodology ....................................................................... 127
4.4.3.1 The Incident of Dividend Payment Among Thai Firms ............... 127
4.4.3.2 The Characteristics of Dividend Payers and Non-Payers ............. 130
4.4.3.2.1 The Effects of Profitability, Investment
Opportunities and Firm Size ...................................................................... 136
4.4.3.3 Changing Propensity to Pay Dividends ......................................... 139
4.5
Plan for Data Analysis ................................................................................... 141
4.5.1 Data Processing ................................................................................... 142
4.5.2 Validity................................................................................................. 142
4.6
Conclusion ...................................................................................................... 142
CHAPTER 5 – ANALYSIS OF DATA
5.1
Introduction.................................................................................................... 144
5.2
Descriptive Statistics...................................................................................... 146
5.2.1 Descriptive Statistics Definitions ....................................................... 146
5.2.2 A Description of the Sample............................................................... 146
5.3
An Initial Analysis of Descriptive Statistics................................................. 148
5.4
Time Trends in Cash Dividends ................................................................... 148
5.4.1 Payers and Their Dividend Yield ...................................................... 153
5.4.2 Dividend Pattern ................................................................................. 158
5.5
Characteristics of the Dividend Payers........................................................ 166
5.5.1 Firm Characteristics Descriptive Statistics Approach .................... 166
5.5.1.1 Profitability....................................................................................... 166
5.5.1.2 Investment Opportunities ............................................................... 175
5.5.1.3 Size..................................................................................................... 178
5.5.2 ANOVA Test........................................................................................ 182
5.5.3 Review of Descriptive Statistic Results ............................................. 189
5.5.4 Confirmation from Logit Regressions .............................................. 191
5.5.5 Hypotheses Testing Using Logit Regression (Average Coefficient) 196
5.5.6 Hypotheses Testing Using Logit Regression (Group Coefficient) .. 199
5.6
The Propensity to Pay Dividends (Descriptive Statistics Evidence).......... 204
5.7
Changing Characteristics and Propensity to Pay Dividends
(Logit Regressions)......................................................................................... 208
5.7.1 Regression Estimates .......................................................................... 208
5.8
Conclusion ...................................................................................................... 213
CHAPTER 6 – CONCLUSIONS AND IMPLICATIONS
6.1
Introduction.................................................................................................... 214
29
DBA Thesis: Malinee Ronapat
6.2
6.3
6.4
6.5
6.6
Conclusions..................................................................................................... 216
6.2.1 Conclusions – Research Hypotheses and Research Questions ....... 216
6.2.1.1 Conclusions Relating to the Chapters ............................................ 216
6.2.1.2 Conclusions Relating to the Testing the Model............................. 219
Implications .................................................................................................... 230
6.3.1 Implication for Theory ....................................................................... 230
6.3.2 Implication for Policy ......................................................................... 231
6.3.3 Implication for Investment Practice.................................................. 232
Contribution of the Study ............................................................................. 233
Limitation of the Study.................................................................................. 234
Suggestions for Further Research ................................................................ 235
REFERENCES.............................................................................................................. 237
APPENDICES ............................................................................................................... 250
30
DBA Thesis: Malinee Ronapat
LIST OF TABLES
Table 2.1:
Table 2.2:
Table 2.3:
Table 2.4:
Table 2.5:
Table 2.6:
Table 2.7:
Table 2.8:
Table 2.9:
Table 2.10:
Table 2.11:
Table 2.12:
Table 2.13:
Table 3.1:
Table 3.2:
Table 3.3:
Table 3.4:
Table 3.5:
Table 4.1:
Table 4.2:
Table 5.1:
Table 5.2:
Table 5.3:
Table 5.4:
Table 5.5:
Table 5.6:
Table 5.7:
Table 5.8:
Table 5.9:
Table 5.10:
Table 5.11:
Table 5.12:
Table 5.13:
Table 5.14:
Table 5.15:
Table 5.16:
Table 5.17:
Table 5.18:
The Share of Bank Lending to Property Sector ................................................. 32
Balance of Payment in US Dollar Billion (1985-2000)..................................... 32
Economic Growth Rates of Selected East Asian Countries 1996-1999 ............ 33
Exchange Rate (Per USD) ................................................................................. 34
Stock Market Collapsed (1997 to 1998) ............................................................ 34
Number of Companies Obtaining Securities Business Licenses ....................... 40
Selected Foreign Stock Exchange Indices ......................................................... 46
Percentage of Buying and Selling Value ........................................................... 47
Market Capitalization of Selected Foreign Stock Exchanges............................ 49
Number of Listed Companies of Selected Foreign Stock Exchanges ............... 50
Financial Statement Disclosure Deadlines......................................................... 52
Overview of Stock Exchange of Thailand ......................................................... 54
Turnover Value of Selected Foreign Stock Exchanges ..................................... 56
Key Differences between Debt and Equity Capital ........................................... 64
Characteristics of Preferred Stock and Common Stock..................................... 66
Advantages and Disadvantages of Common Shares.......................................... 67
Advantages and Disadvantages of Preference Shares ....................................... 67
Decision Rules for Investors and Analysts ........................................................ 75
Number of Samples in this Study .................................................................... 124
Comparison of Different Ratios Applied in Different Stages of
this Research .................................................................................................... 138
Counts and Percent of SET Firms in Different Dividend Groups ................... 149
Number of Sample Firms in Each Industry ..................................................... 150
Number of Dividend Payers in Each Industry ................................................. 151
Percentage of Payers to Total Number of Listed Firms in
Each Industry ................................................................................................... 152
Annual Dividend Yield by Range.................................................................... 155
Percentage of Payers in Two Ranges of Dividend Yield................................. 155
Average Dividend Yield, Average Weighted Dividend Yield, Average
Percentage Changes in Security Prices and Percentage of Market
Capitalization of Payers of Total Firms ........................................................... 157
What Happens in Year t to SET Firms That Do and
Do not Pay Dividends in Year t-1.................................................................... 163
Average Values of Six Characteristic Ratios................................................... 167
Summary of the Characteristics of the Listed Firms in
Thailand from Descriptive Statistics Approach............................................... 180
Percentage of Aggregate Values Accounted for by Firms Paying
Dividends ......................................................................................................... 181
ANOVA F-statistic Test of Characteristics from 1991 to 2002 ...................... 183
Summary of the Payers’ Characteristics Confirmation ................................... 184
ANOVA Test for Payers, Former Payers and Never Paid Firms .................... 186
Summary of the Characteristics of Payers, Former Payers and
Never Paid Firms ............................................................................................. 188
Logit Regression to Explain which Firms Pay Dividends ............................... 192
Summary of Hypothesis Testing Results (Average Coefficient)..................... 199
Logit Regression Results (Controlling Two Investment Variables)................ 200
31
DBA Thesis: Malinee Ronapat
Table 5.19:
Table 5.20:
Table 5.21:
Table 5.22:
Table 6.1:
Table 6.2:
Table 6.3:
Table 6.4:
Table 6.5:
Table 6.6:
Logit Regression Results (Controlling One Investment Variable) .................. 201
Summary of Hypothesis Testing Summary (Group Coefficient) .................... 203
Hosmer and Lemeshow Test (Goodness of Fit) .............................................. 204
Logit Regression Estimates of the Effect of Changing Firm
Characteristics and Decline in the Propensity to Pay Dividends..................... 210
Results of Hypotheses Testing Following Fama and French (2001)
Method of Averaging Coefficient (Profitability)............................................. 220
Results of Hypothesis Testing Using Group Coefficient
(Profitability).................................................................................................... 221
Results of Hypotheses Testing Following Fama and French (2001)
Method of Averaging Coefficient (Investment Opportunities) ....................... 222
Results of Hypotheses Testing Using Group Coefficient
(Investment Opportunities) .............................................................................. 222
Results of Hypotheses Testing Following Fama and French (2001)
Method of Average Coefficient ....................................................................... 224
Results of Hypotheses Testing Using Group Coefficient
(Size) ................................................................................................................ 224
32
DBA Thesis: Malinee Ronapat
Figure 1.1:
Figure 1.2:
Figure 1.3:
Figure 1.4:
Figure 2.1:
Figure 2.2:
Figure 2.3:
Figure 2.4:
Figure 2.5:
Figure 2.6:
Figure 2.7:
Figure 2.8:
Figure 2.9:
Figure 2.10:
Figure 2.11:
Figure 2.12:
Figure 2.13:
Figure 2.14:
Figure 3.1:
Figure 3.2:
Figure 3.3:
Figure 3.4:
Figure 3.5:
Figure 3.6:
Figure 3.7:
Figure 3.8:
Figure 3.9:
Figure 3.10:
Figure 3.11:
Figure 4.1:
Figure 4.2:
Figure 4.3:
Figure 4.4:
Figure 4.5:
Figure 4.6:
Figure 4.7:
Figure 4.8:
Figure 4.9:
Figure 5.1:
Figure 5.2:
Figure 5.3:
Figure 5.4:
Figure 5.5:
LIST OF FIGURES
The Structure of Chapter One ............................................................................. 2
Hypotheses Developed........................................................................................ 9
Categories of Firms by Fama and French (2001) ............................................. 14
Structure of the Study ....................................................................................... 19
The Structure of Chapter Two .......................................................................... 22
Map of Thailand and Geographical Position .................................................... 23
Number of Population from 1997 to 2002........................................................ 24
Minimum Daily Wage in THAI Baht ............................................................... 25
Level of Education 1999................................................................................... 26
Real GDP Growth (%) from 1999 – 2003 ........................................................ 28
Inflation (%) from 1999 to 2003....................................................................... 29
Official Reserve in Billion AUD from 1999 to 2003 ....................................... 29
Prime Minimum Loan Rate (%) from 1999 to 2003......................................... 30
Current Sovereign Credit Rating of Thailand by Three Raters ........................ 37
Regulatory Framework of SET......................................................................... 39
Number of Companies Obtaining Securities Business Licenses ...................... 41
Stock Exchange of Thailand (SET)’s Performance from 1995 to 2003 ........... 45
Net Trading Value Since 1995.......................................................................... 47
The Structure of Chapter Three ........................................................................ 60
Type of Information in the Capital Markets ..................................................... 73
Removal of Mispricings.................................................................................... 75
Dividend-Retention-Financing Trade-Offs....................................................... 79
Sequence of Dividend Payment Dates.............................................................. 81
Two Ways of Raising Cash for Firm ................................................................ 83
Number of Listed Firms Which Paid Dividends............................................. 103
Percentage of Dividend Payers to Total Listed Firms .................................... 104
Research Questions and Hypotheses .............................................................. 108
Research Questions in Relation to Hypotheses .............................................. 111
Model for this Research .................................................................................. 112
The Structure of Chapter Four ........................................................................ 115
Links Between the Research Problem, Research Questions,
Research Hypotheses and Methodology......................................................... 116
Alternatives for the Basic Research Method .................................................. 120
Defining Independent and Dependent Variables ............................................ 122
Process of Counts and Percentages of SET Firms in SIMS in Time t............ 127
Payers and Non-payers in Time t-1 ................................................................ 130
Average Firm Profitability, Investment Opportunities and Size .................... 135
How Logit Regressions is Used for Explaining the Dividend
Policies of Firms ............................................................................................. 139
Logit Regression Estimates of the Effect of Changing Firm
Characteristics and Decline in the Propensity to Pay Dividends.................... 141
The Structure of Chapter Five......................................................................... 145
Annual Average Dividend Yields and the Pattern in the Yield of
The Weighted Average Dividends.................................................................. 154
Trends in the Dividend Yield 1990-2002 ....................................................... 156
The Number of SET Firms in Each Group ..................................................... 158
Percent of All SET Firms in Different Dividend Groups ............................... 159
33
DBA Thesis: Malinee Ronapat
Figure 5.6:
Figure 5.7:
Figure 6.1:
Figure 6.2:
Hypotheses Revisiting .................................................................................... 195
Percentage of Payers Among Firms With (1) Positive Earnings
(2) Negative Earnings (3) Earning Above Investment, and
(4) Earnings Below Investment ...................................................................... 207
The Structure of Chapter Six .......................................................................... 215
Relationships of Logit Regression, ANOVA and Descriptive
Statistics Findings ........................................................................................... 230
34
DBA Thesis: Malinee Ronapat
TO MY PARENTS, WHO INSTALLED IN ME THE
IMPORTANCE OF EDUCATION AND HARDWORK.
35
DBA Thesis: Malinee Ronapat
C
HAPTER
1.1
1
INTRODUCTION TO THE STUDY
INTRODUCTION
This chapter is an introduction to this study. The objective of this chapter is to establish
foundations to the following chapters and to provide an overall picture of the study. The
chapter is presented in eight sections as indicated by figure 1.1. Section 1.1 contains a
general overview and the structure of this study. Section 1.2 discusses the research
background. Section 1.3 introduces the research problem and other research issues
including research questions and research hypotheses. Section 1.4 is a justification of the
study. Section 1.5 briefly discusses the research methodology and introduces the procedure
for identifying key variables, collecting data, data sampling, developing the research model
and analysing the data. Chapter 4 will discuss this methodology in more detail. Section 1.6
presents the limitations of the study. Section 1.7 outlines the content each chapter of the
thesis and finally section 1.8 is a conclusion to this chapter.
36
DBA Thesis: Malinee Ronapat
Figure 1.1: The Structure of Chapter One
1.1
1.2
Introduction
Background to the Research
1.3
Research Issues
1.3.1 Research Problem
1.3.3 Research Hypotheses
1.3.2 Research Questions
1.3.4 Research Objectives
1.4
Justification for the Study
1.4.1 Disappearing Dividends
and Investors
1.4.3
The Research
Community
1.5
1.5.1
1.4.2 Regulators
Data and Methodology
Key Variables Used in
the Study
1.5.2
Research Model and
Data Analysis Procedure
1.5.3 Testing Periods and
Data Sampling
1.6
Structure of the Study
1.7
Conclusion
Source: Developed for this research
37
DBA Thesis: Malinee Ronapat
1.2
BACKGROUND TO THE STUDY
One of the basic requirements of a developed financial system is the existence of a formal
capital market where investors can buy and sell securities (Melicher and Norton 2000).
When firms wish to obtain finance, they have two alternatives: borrowing and thereby
increasing their debt capital, or issuing stocks (preferred or common) and increasing their
equity capital (Brealey and Myers 1991, 2000; Fisher and Jordan 1991; Gitman 2000;
Melicher and Norton 2000; Petty et al. 2000; Gitman, Juchau and Flanagan 2002).
Firms normally choose to issue stocks because they have no maturity date, and no cash
outflow is associated with their redemption (Fisher and Jordan 1991; Gitman 2000;
Melicher and Norton 2000; Petty et al. 2000; Brealey and Myers 1991; Gitman, Juchau and
Flanagan 2002). Investors who buy a firm’s stocks are called shareholders, or stockholders
and they are considered the owners of the firm (Melicher and Norton 2000; Petty et al.
2000). However, investors desire a financial return for the role they have played in the
success, or failure of the firm’s operations (Petty et al. 2000). Therefore, firms normally
pay dividends for common and preferred stockholders when they have positive earnings
(with the recommendation by the Board of Directors). Stockholders view dividend
payments as a return for holding the stocks of firms (Brealey and Myers 1991, 2000; Fisher
and Jordan 1991; Gitman 2000; Melicher and Norton 2000; Petty et al. 2000; Gitman,
Juchau and Flanagan 2002).
However, dividends are normally taxed at a higher rate than interest payments (Brealey and
Myer 1991; Fisher and Jordan 1991; Gitman 2000; Melicher and Norton 2000; Gitman,
Juchau and Flanagan 2002). Therefore, it is possible that borrowing is cheaper than issuing
equities. Despite this fact, some firms still pay dividends and this is a puzzle (Black 1976).
Fama and French (1997) and other researchers examined the dividends and debt decisions
of some US firms and showed that there is a relationship between the firm’s dividend
38
DBA Thesis: Malinee Ronapat
decisions and investment, profitability and size (Fama and French 1995, 1997, 1998, 1999,
2001; La Porta et al. 2000; Wetherilt and Weeken 2002)
A research paper by Fama and French (2001) showed that the percentage of firms paying
dividends fell from 66.5 percent in 1978 to 20.8 percent in 1999 for three US stock markets
(AMEX, NYSE and NASDAQ). Fama and French (2001) attempted to explain whether the
fall in the percentage of dividend payers is due to the decreasing propensity to pay
dividends, or the changing characteristics of listed firms on the US stock markets (Fama
and French 2001). Their methods are simplistic but meaningful. Fama and French (2001)
concluded that the changing propensity to pay dividends and the characteristics of the firms
are both important factors which explain the decline in the percentage of dividend payers in
the US stock markets.
In Thailand, the Stock Exchange of Thailand’s (SET) summary statistics indicate that the
percentage of firms paying dividends fell from 84.2 percent of listed firms in 1990, to 46.4
percent in 2002 (SET 1990, 2002a). It is not known whether the decline in the percentage
of payers is due to the change in the propensity to pay, or other characteristics of firms. A
formal study which uses logit regression to analyse dividend patterns has not been
conducted in Thailand. The researcher has observed that there is a gap in the literature on
disappearing dividends and the percentage of payers in emerging capital markets such as
Thailand. An analysis of this topic will provide benefits for investors, regulators,
researchers, and contribute to the body of knowledge on disappearing dividends.
39
DBA Thesis: Malinee Ronapat
1.3
RESEARCH ISSUES
This section describes the research problem, issues, questions, hypotheses and objectives of
this study.
1.3.1
RESEARCH PROBLEM
The research problem for this study is defined as:
‘How do the characteristics of publicly traded firms and their propensity to pay
dividends influence the payment of dividends and the reasons for their
disappearance from the stock market of Thailand?’
In this study, an apparent disappearance of dividends is explained by (1) changes in the
characteristics of different dividend groups and (2) the propensity to pay dividends of
firms. The study examines whether the disappearance of dividends in Thailand results from
changes in firms’ characteristics or/and their propensity to pay dividends over time.
Dividends are the only actual cash flow which investors in capital markets receive without
selling their shares. Therefore, the disappearance of dividends will affect shareholders,
investors, regulators, analysts and firms which make investment decisions.
1.3.2 RESEARCH QUESTIONS
Research questions are the questions to which researchers seek answers for in a research
project (Black 1999). These questions must be researchable and answerable (Ticehurst and
Veal 2000). Five research questions are developed for solving the research problem of this
study. These questions are defined below:
40
DBA Thesis: Malinee Ronapat
Research question #1
: What is the background and performance of the
Stock Exchange of Thailand and its listed firms?
Research question #2
: How have researchers investigated dividends
and their patterns throughout the world?
Research question #3
: What is the appropriate methodology for
collecting and analysing data?
Research question #4
: Does the data from Thailand support the model?
Research question #4.1
: What are the characteristics of dividend payers?
Research question #4.2
: Do these characteristics change over time?
Research question #4.3
: How has the propensity to pay dividends of firms
changed over time?
Research question #5
: How could a model be implemented?
1.3.3 RESEARCH HYPOTHESES
After formulating a research question, one or a set of research hypotheses which relate to
the research question are developed. It is rarely possible to test whether the theory holds
when faced with empirical evidence (Clover and Balsley 1984). Instead, one is more likely
to find that ‘a hypothesis which relates to a limited fact of the theory, will be deduced from
the theory and submitted to a searching enquiry’ (Bryman and Cramer 1999, p. 3).
A research hypothesis is defined as ‘a logically conjectured relationship between two or
more variables expressed in the form of a testable statement’ (Cavana, Delahaye and
Sekaran 2001, p. 98). By testing the hypotheses and conjectured relationships, it is
expected that solutions ‘can be found to correct, or answer the research problem
41
DBA Thesis: Malinee Ronapat
encountered’ (Cavana, Delahaye and Sekaran 2001, p. 98). This process is useful ‘when a
researcher is seeking to prove or disprove the research questions by using statistical
analysis’ (Ticehurst and Veal 1997, p. 37). The hypotheses relating to the questions of this
study are described below:
Step A: Sample listed firms 1991 to 1996 (before the Asian Economic Crisis)
Hypothesis # 1A
: Does the dividend payment action depend on the profitability of
the listed firms?
Hypothesis # 2B
: Does the dividend payment action depend on the investment
opportunities of the listed firms?
Hypothesis # 3C
: Does the dividend payment action depend on the size of the listed
firms?
Step B: Sample listed firms 1997 to 2002 (During and after the Asian Economic
Crisis)
Hypothesis # 1A
: Does the dividend payment action depend on the profitability of
the listed firms?
Hypothesis # 2B
: Does the dividend payment action depend on the investment
opportunities of the listed firms?
Hypothesis # 3C
: Does the dividend payment action depend on the size of the listed
firms?
42
DBA Thesis: Malinee Ronapat
Step C: Sample listed firms 1991 to 2002 (Before, during and after the Asian
Economic Crisis)
Hypothesis # 1A
: Does the dividend payment action depend on the profitability of
the listed firms?
Hypothesis # 2B
: Does the dividend payment action depend on the investment
opportunities of the listed firms?
Hypothesis # 3C
: Does the dividend payment action depend on the size of the listed
firms?
Figure 1.2 lists the research hypotheses of steps A, B and C.
43
DBA Thesis: Malinee Ronapat
Figure 1.2: Hypotheses Developed
CHARACTERISTICS OF DIVIDEND PAYERS BEFORE THE ASIAN ECONOMIC CRISIS (Step A)
(1) PROFITABILITY
Ho
:
Ha
:
There is no significant relationship between the profitability of the firms and the probability
that the listed firms will pay dividends before the Asian Economic Crisis.
There is a significant relationship between the profitability of the firms and the probability
that the listed firms will pay dividends before the Asian Economic Crisis.
(2) SIZE
Ho
:
Ha
:
There is no significant relationship between the size of the firms and the probability that the
listed firms will pay dividends before the Asian Economic Crisis.
There is a significant relationship between the size of the firms and the probability that the
listed firms will pay dividends before the Asian Economic Crisis.
(3) INVESTMENT OPPORTUNITIES
Ho
:
Ha
:
There is no significant relationship between the investment opportunities of the firms and the
probability that the listed firms will pay dividends before the Asian Economic Crisis.
There is a significant relationship between the investment opportunities of the firms and the
probability that the listed firms will pay dividends before the Asian Economic Crisis.
CHARACTERISTICS OF THE DIVIDEND PAYERS DURING AND AFTER THE ASIAN ECONOMIC CRISIS (Step B)
(1) PROFITABILITY
Ho
:
Ha
:
There is no significant relationship between the profitability of the firms and the probability
that the listed firms will pay dividends after the Asian Economic Crisis.
There is a significant relationship between the profitability of the firms and the probability
that the listed firms will pay dividends after the Asian Economic Crisis.
(2) SIZE
Ho
:
Ha
:
There is no significant relationship between the size of the firms and the probability that the
listed firms will pay dividends after the Asian Economic Crisis.
There is a significant relationship between the size of the firms and the probability that the
listed firms will pay dividends after the Asian Economic Crisis.
(3) INVESTMENT OPPORTUNITIES
Ho
:
Ha
:
There is no significant relationship between the investment opportunities of the firms and the
probability that the listed firms will pay dividends after the Asian Economic Crisis.
There is a significant relationship between the investment opportunities of the firms and the
probability that the listed firms will pay dividends after the Asian Economic Crisis.
CHARACTERISTICS OF THE DIVIDEND PAYERS FOR THE WHOLE PERIOD (Step C)
(1) PROFITABILITY
Ho
:
Ha
:
There is no significant relationship between the profitability of the firms and the probability
that the listed firms will pay dividends for the whole period of investigation.
There is a significant relationship between the profitability of the firms and the probability
that the listed firms will pay dividends the whole period of investigation.
(2) SIZE
Ho
:
Ha
:
There is no significant relationship between the size of the firms and the probability that the
listed firms will pay dividends the whole period of investigation.
There is a significant relationship between the size of the firms and the probability that the
listed firms will pay dividends the whole period of investigation.
(3) INVESTMENT OPPORTUNITIES
Ho
:
Ha
:
There is no significant relationship between the investment opportunities of the firms and the
probability that the listed firms will pay dividends the whole period of investigation.
There is a significant relationship between the investment opportunities of the firms and the
probability that the listed firms will pay dividends the whole period of investigation.
Source: Developed for this research
44
DBA Thesis: Malinee Ronapat
1.3.4
RESEARCH OBJECTIVES
Based on the research background and research issues described earlier, this research has
eight objectives:
x
To classify listed firms in the Thai capital market into different dividend groups
according to their dividend actions in time t (dividend payers, former payers and never
paid).
x
To determine the dividend pattern and explain the phenomenon of disappearing
dividends in the Thai capital market.
x
To determine the characteristics and changes in these characteristics of listed firms
in different dividend groups (dividend payers, former payers and never paid firms) over
time.
x
To determine the relationship between each characteristic and the dividend payment
action of the listed firms.
x
To determine the propensity to pay dividends of listed firms in Thailand.
x
To generate a formula to predict the percentage of dividend payers in the market.
x
To provide an early warning of the likelihood of the disappearance, or the
reappearance of dividends.
x
To draw inferences on the importance of institutional and financial market issues
for payers of dividends.
1.4
JUSTIFICATION FOR THE STUDY
Research on the phenomenon of disappearing dividends is relatively recent in the US
because Fama and French discussed the issue in 2001. Furthermore, Fama and French’s
45
DBA Thesis: Malinee Ronapat
(2001) investigation focused on a developed country, the US. There is an absence of
academic research in the Thai context. Most of the studies on dividends in the emerging
markets have concentrated on dividend policy and the behavior of stock markets (La Porta
et al. 2000; Aivazian, Booth and Cleary 2003). This is the first Thai study on the
‘disappearing dividends’ phenomenon. This research will provide new evidence on the
phenomenon and will benefit investors and regulators.
1.4.1 DISAPPEARING DIVIDENDS AND INVESTORS
Investors would be aware of trends in the payment of dividends in the Thai stock market
between 1990 and 2002. They will be able to make investment selections between firms
(dividends payers, former-payers and never paid firms) because they will be aware of the
different characteristics and propensity to pay of these firms (Fama and French 2001).
Institutional investors include Banks, Mutual funds, Pension Funds and Brokers. These
institutions have portfolios with values in the millions of Thai baht. This research applies
fundamental analysis which involves analysing stocks from published sources, such as
financial statements. Shifting from payers to non-payers (former payers or never paid
firms) or non-payers to payers, depending on the return, should be done by the institutional
investors. They will be able to give accurate analysis and forecasts of expected stock prices
and returns in the market, thus, reducing the level of risk in their portfolios.
1.4.2 REGULATORS
Regulators need to ensure the investment climate is fair for all market participants.
Regulators such as the Securities and Exchange Commission (SEC) and the Stock
Exchange of Thailand (SET) should be aware of the publicly available information. SET
and SEC need to be aware of changes in the payment of dividends and the regulations
imposed on the payment of dividends, such as taxation. They also need to be aware of the
changing characteristics and the propensity to pay dividends of the newly listed firms. They
should consider announcing more fundamental analysis ratios and summary statistics for
investor education and protection.
46
DBA Thesis: Malinee Ronapat
1.4.3 THE RESEARCH COMMUNITY
This research aims to contribute to the body of knowledge on ‘disappearing dividends’ in
the emerging stock market in Thailand, where the ‘disappearing dividends’ is a new topic
with a lack of literature. This will raise the level of understanding on the phenomenon of
disappearing dividends, firm characteristics, the measurement of changes in firm
characteristics, the propensity to pay dividends in the developed market (the US) and the
emerging market of Thailand.
1.5
DATA AND METHODOLOGY OVERVIEW
This section is an introductory overview of the methodology used in this study. The
research methodology will be discussed in more detail in chapter 4. The methodology was
identified after exploring the empirical research papers relating to the disappearing
dividend, dividend omissions, Efficient Market Hypotheses (EMH) and several capital
market research papers. This section is divided into 3 subsections. Subsection 1.5.1 shows
the key variables used in this research. Subsection 1.5.2 briefly summarises the research
models in this study and the procedure for data analysis. Subsection 1.5.3 presents the
sources of data, testing periods and data sampling.
1.5.1 KEY VARIABLES USED IN THE STUDY
This research is an example of causal research as it ‘identifies the cause-and-effect
relationships among variables’ (Zikmund, 2000, p. 56). This paper investigates the history
of the payment of dividends (secondary data technique) of listed firms in the Thai capital
market and applies summary statistics and logit regression to explore the decline in the
percentage of payers of dividends.
Secondary data will be used as the principal research technique of this study (Fama and
MacBeth 1973; Fama and French 1988, 1995, 1997, 1998, 1999, 2000, 2001, 2002; La
Porta et al. 2000; DeAngelo, DeAngelo and Skinner 2002; Aivazian, Booth and Cleary
2003). Numerical data on dividends payments are required to complete this study. This data
47
DBA Thesis: Malinee Ronapat
is available in secondary form (Fama and MacBeth 1973; Fama and French 1988, 1995,
1997, 1998, 1999, 2001; La Porta et al. 2000; DeAngelo, DeAngelo and Skinner 2002;
Aivazian, Booth and Cleary 2003).
The research commences with a review of the literature on the phenomenon of
disappearing dividends which was discussed by Fama and French (2001). Later, the
discussion branches out to include relevant research papers that discuss methodologies and
the findings of Fama and French (2001). The researcher’s choice of factors, variables and
methodology is motivated by previous empirical findings (Fama and MacBeth 1973; Fama
and French 1988, 1995, 1997, 1998, 1999, 2000, 2001; La Porta et al. 2000; DeAngelo,
DeAngelo and Skinner 2002; Aivazian, Booth and Cleary 2003).
The research methodology has two main components. The first component involves the use
of summary statistics, or descriptive statistics such as means, frequencies and standard
deviation. These statistics will be used to group listed firms (figure 1.3) and identify the
phenomenon of disappearing dividends in the Thai stock market. This technique has been
used in developed countries and has been suggested by a number of researchers (Fama and
MacBeth 1973; Fama and French 1988, 1995, 1997, 1998, 1999, 2000, 2001; DeAngelo,
DeAngelo and Skinner 2002; Aizazian, Booth and Cleary 2003). The second component is
the application of logit regression which is used for classifying the characteristics of
dividend payers and non-payers (former-payers and never paid firms) as well as identifying
changes in the characteristics and the propensity to pay dividends of listed firms in
Thailand. Logit regression has been used by many researchers including Fama and
MacBeth (1973), Fama and French (1988), (1995), (1997), (1998), (1999), (2001),
DeAngelo, DeAngelo and Skinner (2002), and Aizazian, Booth and Cleary (2003).
As mentioned earlier, listed firms will be classified into four groups, depending on their
dividend action in time t (present time). Payers are firms that pay cash dividends in time t
or the present time. Non-payers are firms that do not pay cash dividends in time t, or the
present time. Non-payers are divided into (1) former payers which are firms that do not
pay at the present time, but paid in a previous year and (2) never payers which are firms
that have never paid dividends (Fama and French 2001).
48
DBA Thesis: Malinee Ronapat
Figure 1.3: Categories of Firms or Dividend Groups by Fama and French (2001)
All Firms
Pay Cash Dividends
Do not Pay Cash Dividends
Formerly Paid
Have Never Paid
Source: Fama and French (2001)
The initial discussion of the characteristics of dividend payers focuses on the findings of
the summary statistics. Payers and non-payers differ in terms of profitability, investment
opportunities and size. The evidence from the summary statistics will be confirmed by
conducting logit regressions (Fama and French 2001). Each characteristic will be measured
by financial ratios. The financial ratios include:
For Profitability, the indicators are Et/At and Yt/BEt
Et/At
is the ratio of aggregate earnings before interest to aggregate assets
Yt/BEt
is the aggregate common stock earnings over aggregate book equity
Where:
Et
: Earnings before interest but after taxes for fiscal year t
At
: Assets at the end of fiscal year t
Yt
: After-tax earnings to common stock for fiscal year t
BEt
: Book common equity at the end of fiscal year t
For Investment opportunities, the indicators are dAt/At and Vt/At
49
DBA Thesis: Malinee Ronapat
dAt/At
is the firm’s rate of growth of assets
Vt/At is the ratio of the aggregate market value to the aggregate book value of assets
Where:
Vt
: Total market value at the end of fiscal year t
At
: Total assets at the end of fiscal year t
dAt
: The change of total assets from year t-1
For Size, the indicator is At and Lt/At
At
is the total assets of the firm in time t
Lt/At
is the ratio of aggregate liability to the aggregate value of a firm’s assets
Where:
At
: Total assets at the end of fiscal year t
Lt
: Total liabilities at the end of fiscal year t
As analysis of data and summary statistics is expected to achieve outcomes which are
consistent with the literature:
x
Profitabilityinvestment opportunities and size are factors in the decision to pay
dividends (Fama and French 2001).
x
Dividend payers tend to be large and profitable firms with lower investment
opportunities (Fama and French 2001).
x
Firms that have never paid dividends tend to be smaller and less profitable than
dividend payers, but they have greater investment opportunities (Fama and French 2001).
x
Former dividend payers have low earnings and few investment opportunities (Fama
and French 2001).
1.5.2 RESEARCH MODEL AND DATA ANALYSIS PROCEDURE
Fama and French (2001) suggested that logit regression should be used to confirm the
characteristics of dividend payers. Three characteristics of the firms will be analysed:
50
DBA Thesis: Malinee Ronapat
Profitability, this characteristic will be confirmed with Et/At
Where:
Et/At
is the ratio of aggregate earnings before interest to aggregate assets
Investment opportunities this characteristic will be confirmed with Vt/At anddAt/At
Where:
dAt/At
is the firm’s rate of growth of assets
Vt/At is the ratio of the aggregate market value to the aggregate book value of assets
The Size characteristic will be confirmed with SETt
Where:
SETt is the percentage of SET (Stock Exchange of Thailand) firms with the same or
smaller rates of market capitalisation.
This size measure is meant to simplify any effects of the growth in typical firm size
through time (Fama and French 2001).
The logit regression analysis will determine:
x
The roles of profitability, investment opportunities and size and in the decision to
pay dividends (Fama and French 2001).
x
Larger firms are more likely to pay dividends (Fama and French 2001).
x
More profitable firms are more likely to pay dividends (Fama and French 2001).
x
Firms with more investments are less likely to pay dividends (Fama and French
2001).
The analysis of the propensity to pay dividends will apply the quantitative evidence and is
expected to indicate:
51
DBA Thesis: Malinee Ronapat
x
That dividends are less likely to be paid by firms with characteristics such as
positive earnings and earnings in excess of investment (Fama and French 2001).
x
Firms are less likely to pay dividends, whatever their characteristics (Fama and
French 2001).
The effects of changing firm characteristics and the propensity to pay on the percentage of
firms that pay dividends will be measured. The following methodologies will be adopted:
x
Estimating the probability that firms with given characteristics (profitability,
investment opportunities and size) paid dividends during 1991-2002 (Fama and French
2001).
x
Apply the probabilities (estimated from logit regression) from the base period
(1991-1996) for estimating the expected percentage of dividend payers for each year from
1991 to 2002 (Fama and French 2001).
x
Use the difference between the expected percentage of payers and the actual
percentage to measure the change in the propensity to pay dividends. A positive difference
between the expected and the actual percentage of payers implies a decline in the
propensity to pay (Fama and French 2001).
x
Use the changes in the expected percentage of payers over time to present the
changes in the characteristics of the listed firms (Fama and French 2001).
1.5.3 TESTING PERIODS AND DATA SAMPLING
The annual data on dividend payments by the companies listed at SET will be used for
conducting a yearly analysis of disappearing dividends between 1990 and 2002. The data is
available for the period 1990 to 2002. Information is required for calculating financial
ratios such as: annual dividend yields, asset, liabilities, market capitalization, earnings
before interest and taxes, earnings before extra ordinary items, taxes and interest expenses
between 1990 and 2002. However, the dAt/At ratio, which is the growth rate of asset will
require data for time t and t-1. Consequently, the period of analysis will be shortened to
1991 to 2002.
52
DBA Thesis: Malinee Ronapat
Sample observations used in this study were collected from all listed firms between 1990
and 2002. This data is available on SIMS (the Stock Exchange of Thailand Information
Management System). Sample observations include every sector in the Stock Exchange of
Thailand.
1.6
STRUCTURE OF THE THESIS
This study is presented in 6 chapters as illustrated in figure 1.4. A brief summary of each
chapter is set out below.
Figure 1.4: Structure of the Study
Chapter 1:
Introduction to the Study
Chapter 2:
Country Review
Chapter 3:
Review of Literature
Chapter 4:
Research Methodology
Chapter 5:
Analysis of Data
Chapter 6:
Conclusions and Implications
Source: Developed for this research
Chapter 1 is an introduction to the study. This chapter contains background to the study,
defines the research problem and develops a list of questions and hypotheses for answering
the research question. Chapter 1 also justifies the research, its methodology, provides
definitions of key terms, and discusses its limitations.
53
DBA Thesis: Malinee Ronapat
Chapter 2 is a background discussion on Thailand. This chapter provides an overview of
the economy of Thailand, the Asian Economic Crisis (in 1997), the Thai capital market, the
Securities Exchange of Thailand (SET) and the sovereign credit rating of Thailand.
Chapter 3 is a review of research papers relating to the parent disciplines of financial
investment, capital markets and dividend policy. This chapter commences with background
on finance, investment, risk and returns. The discussion also focuses on the components of
Thailand’s capital market and the dividend policies of firms. Emphasis is placed on the
issue of the phenomenon of disappearing dividends in the stock markets of developed
economies. Chapter 3 concludes with a summary of the gaps in the knowledge and the
limitations of the relevant papers on the topic of disappearing dividends. The discussion
encourages the investigation of this phenomenon in developing countries, especially in the
case of Thailand.
Chapter 4 discusses the methodology used for extracting and analysing data, and testing
the stated hypotheses. Definitions and measurements of key variables are provided and a
model is designed for this study. This chapter concludes with a plan for the analysis of data
which will be conducted in Chapter 5..
Chapter 5 discusses the findings of this study. The findings are analysed using summary
statistics and the results are confirmed by ANOVA and logit regression. The findings are
analysed in a pattern which is consistent with the methodology described in Chapter 4.
Finally, Chapter 6 discusses the implications of the findings which were presented in
Chapter 5. It also provides answers to the research question and hypotheses which were
developed in Chapter 1. The chapter concludes with a discussion on the limitations of the
study and identifies several areas for future research.
54
DBA Thesis: Malinee Ronapat
1.7
CONCLUSION
This chapter introduced the main features of this research study, including its background,
research issues, questions, justification, an overview of the methodology, definitions and its
limitations. Chapter 2 will present background on Thailand and its the Stock Exchange
(SET) and highlight the performance of the Thai economic and financial environments.
55
DBA Thesis: Malinee Ronapat
C
HAPTER
2.1
2
THAILAND: COUNTRY REVIEW
INTRODUCTION
This chapter provides background information on Thailand, with emphasis on its economy,
geography populations, capital markets and the Asian Economic Crisis of 1997. The
development of Thailand’s capital markets and their performance are also discussed to
focus the direction of this study.
This chapter is presented in four sections. Section 2.1 is an introduction to the chapter.
Section 2.2 is a brief outline of the geography, demography and economy of Thailand.
Section 2.3 discusses the causes and impacts of the Asian Economic Crisis on the economy
of Thailand and its capital markets. Section 2.4 focuses on the trends in the sovereign credit
rating given to Thailand by three international rating companies since 1989. Section 2.5
concentrates on Thailand’s capital markets and the Stock Exchange of Thailand (SET). In
particular, this section traces the history, establishment, performance and operation of SET.
Finally, section 2.6 is a brief conclusion to the chapter.
56
DBA Thesis: Malinee Ronapat
Figure 2.1: The Structure of Chapter 2
2.1
2.2
Introduction
Country Profile
2.2.1 Demographic
2.2.3 Economic Highlights
2.2.2 Social Highlights
2.2.4 Trading Environment
2.3
2.4
Asian Economic Crisis
Thai Sovereign Rating (by Fitch, Moody’s, S&P)
2.5
Thai Capital Market
2.5.1
History of SET
2.5.8
Listing
2.5.2
Establishment of BSE
2.5.9
Disclosure Procedure
2.5.3 Establishment of SET
2.5.10 Material Information
2.5.4 Roles of SET
2.5.11 Stock Information
2.5.5
Regulatory Framework
2.5.12 Information Dissemination
2.5.6
SET Performance
2.5.13 Investor Protection
2.5.7
SET Vision 2003
2.5.14 Systems Reliability
2.6
Conclusion
Source: Developed for this research
57
DBA Thesis: Malinee Ronapat
2.2
COUNTRY PROFILE
Figure 2.2: Map of Thailand and Geographical Position
Figure removed due to copyright restrictions
Source: BOI (2004)
The Kingdom of Thailand is situated in the heart of Southeast Asia (see Figure 2.2).
Thailand's total area of 514,000 sq. km. includes a land area of 507,171 sq. km. and 2,230
sq. km. of water territory. The country shares borders with Myanmar (Burma), Laos,
Cambodia, and Malaysia. (Pringvanich 2002). The name of the country was ‘Siam’ until
1949. On May 11, 1949, an official proclamation declared that the country would
henceforth be known as ‘Thailand’ (BOI 2004).
Thailand is a warm and humid tropical country. The climate is monsoonal with a
pronounced rainy season which lasts from May to September and a relatively dry season
from August to April. Temperatures are highest in March and April and lowest in
December and January. The average temperature is 28.1o C (BOI 2004).
Thailand is divided into four natural regions (1) The North; (2) The Central Plain; (3) The
Northeast and (4) The South. The North is a mountainous region comprised of natural
forests, ridges and deep narrow alluvial valleys. Central Thailand is a lush, fertile valley. It
is the richest and most extensive rice-producing area in the country. Bangkok, the capital of
58
DBA Thesis: Malinee Ronapat
Thailand, is located in this region. The Northeast region is arid, characterised by rolling
terrain and undulating hills. This region often suffers from climatic extremes such as floods
and droughts. The Southern region is hilly to mountainous, with thick virgin forests and
rich deposits of minerals and ores. This region is a centre for the production of rubber and
cultivation of tropical crops (BOI 2004).
2.2.1
DEMOGRAPHIC
The population of Thailand is diverse and includes Thais, Chinese, Malays, Cambodians,
Vietnamese, Indians, and others. Approximately 30 percent of the total population is under
the age of 15. With a growth rate of 1.2 to 1.4 percent per year the population is projected
to exceed 70 million by 2010 (BOT 2000, 2002; BOI 2004).
Figure 2.3: Number of Population from 1997 to 2003
Year
Population (Million)
1997
60.8
1998
61.5
1999
61.8
2000
62.4
2001
62.4
Population (Million)
Million
People
2002
62.5
2003
63.8
Population (M illion)
64.5
63.8
64
63.5
63
62.5
62
61.5
61
61.5
62.4
62.4
62.5
2000
2001
2002
61.8
60.8
60.5
60
59.5
59
1997
1998
1999
2003
Y ear
Source: BOI (2004); BOT (2000, 2003)
59
DBA Thesis: Malinee Ronapat
The size of the workforce exceeds 32.9 million and the majorities are under 30 years of
age. About 800,000 people join the workforce annually. The minimum wage in Thailand is
currently 162 baht per day (6 AUD at the rate of 27 Baht/1AUD) in Bangkok, and 130 to
140 baht (4.8–5.2 AUD) in the provinces. Although labour is not the cheapest in the region,
Thailand's work force is among the most cost-efficient in the world (BOT 2002). Figure 2.4
below presents the minimum daily wage.
Figure 2.4: Minimum Daily Wage in THAI baht
Provinces
Minimum Daily Wage (THB)
Phuket
Bangkok
Chonburi
Chaing Mai
The rest of the country
168.00
165.00
146.00
143.00
133.00
Minimum Daily Wage (THB)
180
160
140
120
100
80
60
40
20
0
168
165
146
Phuket
Minimum Daily Wage
(THB)
Bangkok
143
133
Chonburi Chaing Mai The rest of
the country
Source: BOI (2004)
In recent years, the Government has made education an important priority. Indeed,
education will be important in the 21st century because the development of the nation’s
60
DBA Thesis: Malinee Ronapat
human resources is the highest priority of the Eighth National Economic and Social
Development Plan (1997-2002) (BOT 2002; BOI 2004).
Figure 2.5: Level of Education in 1999
Level of Education
Number of graduates
Higher than Bachelor Degree
30,719
Bachelor's Degree
238,256
Lower than Bachelor's Degree
11,193
Number of graduates
Number of graduates
Number of Graduates
300000
238256
250000
200000
150000
100000
50000
30719
11193
0
Higher than
Bachelor Degree
Bachelor's Degree
Lower than
Bachelor's Degree
Education Level
Sources: BOT (2002); BOI (2004)
The average rate of the literacy rate in Thailand has risen since the plan was implemented
and stood at 96 percent for the total population in 2003. At present, 97.5 percent of males
and 94.6 percent of females are literate (BOI 2004).
61
DBA Thesis: Malinee Ronapat
2.2.2
SOCIAL HIGHLIGHTS
Buddhism is the national religion and faith of 95 percent of the population. However,
Islam, Christianity, Hinduism and other creeds are also present in Thailand (BOT 2002;
BOI 2004).
The official language is Thai. English is a compulsory subject in public schools and it is
widely spoken and understood, in Bangkok and the major cities. Chinese and Malay are
also popular (BOT 2002; BOI 2004).
Thailand is governed by a constitutional monarchy and an elected parliament. Thailand was
an absolute monarchy until 1932. In 1932 a revolution was staged by a small group of
disaffected civil servants and military men. Since this revolution, Thailand’s monarchs
have ruled under a constitution and their powers are theoretically no greater than those of
European monarchs (BOI 2004). The current Prime Minister is Dr Thaksin Shinnawatra (he
was elected in January 2001). The country is divided into 76 provinces, each administered
by an appointed governor. The provinces are sub-divided into districts, sub-districts,
tambons, and villages. The Bangkok Metropolitan area is administrated by an elected
governor (BOI 2004).
The Thai monarchy is revered and regarded as the central, unifying element of the nation.
King Bhumibol was born on December 5, 1927 and this day in December is now the Thai
National Day. He ascended the throne on June 9, 1946.
2.2.3
ECONOMIC HIGHLIGHTS
More than a decade of strong sustained growth concluded in 1996 when GDP growth
slowed to 5.5 percent. The performance of the economics deteriorated until August 1997
when Thailand committed to a US$17.2 billion (or AUD 26.7 billion at THB27/AUD, July
16, 2003) rescue package which was arranged by the International Monetary Fund (IMF).
62
DBA Thesis: Malinee Ronapat
In mid-2000 Thailand began to repay its loans from the IMF. Under this program of
recovery, Thailand enhanced its revenue streams, substantially reduced spending,
implemented a range of finance sector reforms and initiated a program to privatise many of
its 60 state-owned enterprises (BOI 2004).
Thailand has entered the 21st century on a firm economic footing. In 1999, the economy
began to rebound from the crisis of 1997 and GDP expanded by 4.2 percent. This was a
marked turnaround from the 10% contraction in 1998. Steady growth continued through
2000 and 4.3 percent grow was recorded in this year (BOT 1999, 2000, 2001, 2002; BOI
2004).
Thailand is not immune to fluctuations in the global economy. Weaknesses in the global
economy reduced the growth of GDP to 1.8 percent in 2001. However, solid domestic
consumption and rising private investment in late 2001 and 2002 helped raise GDP by 4.9
percent in 2002. The National Economic and Social Development Board (NESDB) forecast
that this level of growth would be sustained throughout 2003. Indeed, many analysts
believe that Thailand could be experiencing a sustained recovery for the first time since the
economic crisis of 1997 (BOT 1999, 2000, 2001, 2002; BOI 2004).
Figure 2.6: Real GDP Growth (%) from 1999 – 2003
Re al GDP Growth (%)
Real GDP Growth
(%)
6
Year
1999
4.2
4
2000
4.3
2001
1.8
2002
4.9
2003
4.5
GDP (%)
5
4.2
4.9
4.3
3
Real GDP Growth (%)
4.5
1.8
2
1
0
1999
2000
2001
Year
2002
2003
Source: BOT (1999, 2000, 2001, 2002); BOI (2004)
63
DBA Thesis: Malinee Ronapat
Figure 2.7: Inflation (%) from 1999 to 2003
Inflation (%)
Inflation
(%)
1999
0.3
2000
1.6
2001
1.6
2002
0.7
2003
1.8
2
Inflation (%)
Year
1.6
Inflation (%)
1.8
1.6
1.5
1
0.7
0.3
0.5
0
1999
2000
2001
Year
2002
2003
Source: BOT (1999, 2000, 2001, 2002); BOI (2004)
Responsible macroeconomic policies have kept inflation under control since 1999. The
Bank of Thailand (BOT) estimates that inflation for 2004 will be about 2 percent.
Thailand’s official reserves remained strong at US$38 billion at the end of 2003 (see
Figure 2.8). The BOT has pledged to keep reserves strong through 2004.
Figure 2.8: Official Reserve in Billion AUD from 1999 to 2003
Year
Official
Reserve
(Billion
AUD)*
1999
54.0
2000
50.8
2001
51.3
2002
60.5
2003
56.3
O fficial Re se rve (Billion AUD)
Billion AUD
62
60
58
56
54
52
50
48
46
44
Official Reserve
(Billion AUD)
60.5
56.3
54
1999
50.8
51.3
2000
2001
Ye ar
2002
2003
Source: BOT (1999, 2000, 2001, 2002); BOI (2004)
Note: *The spot exchange rate at July 16, 2003 at THB27/AUD
64
DBA Thesis: Malinee Ronapat
Figure 2.9: Prime Minimum Loan Rate (%) from 1999 to 2003
Year
Prime Minimum Loan Rate (%) *
1999
8.25 - 8.50
2000
7.50 - 8.25
2001
7.00 - 7.50
2002
6.50 - 7.00
2003
6.50 – 6.75
Source: BOT (1999, 2000, 2001, 2002); BOI (2004)
Note: * Average prime minimum loan rate for five largest commercial banks
The average prime minimum loan rate (2001-2003) has been below 7.5 percent since 2001,
thereby ensuring that long-term investment (houses and cars) has become attractive.
2.2.4
TRADING ENVIRONMENT IN THAILAND
On the external front, export growth has traditionally been the major driver of the Thai
economy and has contributed to the diversification of the country's industrial
structure. Manufactured exports have expanded strongly and have accounted for 80
percent of total exports in the past few years. Thailand is now a major exporter of
computers and parts, textiles, gems and jewelry, electronics and automotive products, in
addition to agricultural products (BOI 2004).
2.3
ASIAN ECONOMIC CRISIS
The Asian Economic Crisis (AEC) affected most of South East Asia although it
commenced in Thailand. Thailand’s economic problems resulted from its fixed exchange
rate, the collapse of the property market, over expansion of the financial sector and a
decline in the level of exports. It is generally accepted that the Asian Economic Crisis
65
DBA Thesis: Malinee Ronapat
commenced on 2 July 1997 when the Bank of Thailand (BOT) announced that the baht
would be floated. The decision, heavily influenced the value of the Philippine peso,
Malaysian ringgit and Indonesian rupiah and the region’s financial markets collapsed (BOT
1997, 1998; Radelet and Schs 1998; Chua, Sim and Tham 1999; Rahman, Yui and Foon
2003).
2.3.1
ASIAN ECONOMIC CRISIS REVIEW
Goldstein (1998) claimed there were three main causes of the Asian Economic Crisis: (1)
weaknesses in the financial sector (2) dis-equilibrium in the balance of payments and (3)
contagion from Thailand (Rahman, Yui and Foon 2003).
2.3.1.1
FINANCIAL SECTOR WEAKNESSES
A boom in credit and property markets resulted in over excessive lending by the banks to
the property sector. Table 2.1 indicates that before the crisis around 30 to 50 percent of
Southeast Asian banks lent to the property sector. However, with the onset of the crisis,
property prices fell sharply and the level of Non-Performing Loans (NPL) of the region’s
banks rose to as high as 35 percent in some cases. Non-Performing Loans (NPL) are
defined as ‘loans currently in default for three months or longer regardless of whether they
are secured or unsecured’ (TRIS 2001). In addition, borrowing by the banks and corporate
short-term and long-term borrowing was in foreign currency. Some of the banks and firms
had too little capital to repay their loans and the financials sector collapsed (BOT 1997,
1998; Radelet and Schs 1998; Chua, Sim and Tham 1999; Rahman, Yui and Foon 2003).
66
DBA Thesis: Malinee Ronapat
Table 2.1: The Share of Bank Lending to Property Sector
Hong Kong
Singapore
Thailand
Malaysia
Indonesia
South Korea
The Philippines
End of 1997
40-55
30-40
30-40
30-40
25-30
15-25
15-20
Source: Goldstein (1998); ADB (2000a); Rahman, Yui and Foon (2003, p. 496)
2.3.1.2
DISEQUILIBRIUM OF BALANCE OF PAYMENT
Many of the Asian countries facing economic problems had experienced a prolonged
period of disequilibrium in their balance of payments because they imported more than
they exported. Large deficits in the balance of payments were an important factor leading
to the exchange rate crisis (Rahman, Yui and Foon 2003). Table 2.2 shows trends in the
balance of payments for seven Asian countries before and after crisis.
Table 2.2: Balance of Payment in US Dollar Billion (1985-2000)
Country
Hong Kong
Singapore
Thailand
Malaysia
Indonesia
South Korea
The Philippines
1990
6.3
3.1
-7.3
-0.9
-3
-2
-2.7
1991
5.7
4.9
-7.6
-4.2
-4.3
-8.3
-1
1992
5.4
5.9
-6.3
2.2
-2.8
-3.9
-1
Pre-Crisis and Crisis
1993 1994 1995 1996
8.2
1.6 -6.1 -2.2
4.2 11.4 14.4 14.5
-6.4 -8.1
-14
-15
-3 -4.5 -8.5 -4.6
-2.1 -2.8 -6.4 -7.7
1 -3.9 -8.5
-23
-3
-3
-2
-4
1997
-6.8
15
-3
-4.8
-4.9
-8.2
-4.4
1998
2.3
17.6
14.2
9.4
4.1
40.6
1.3
Post-Crisis
1999 2000
6.4
5.3
17.2 17.2
11.4 11.6
12.5
8.4
5
4.4
25.1
6.5
5.2
3.2
Source: ADB (2000b); Lim (2000); Rahman, Yui and Foon (2003, p. 497)
67
DBA Thesis: Malinee Ronapat
2.3.1.3
CONTAGION
Economists and scholars believed that the value of the currencies of Thailand’s neighbours
could not be sustained if there was a large depreciation of the baht. Panic amongst domestic
and foreign investors resulted in strong depreciation of the Philippine peso and Indonesia
rupiah (Radelet and Schs 1998; Chua, Sim and Tham 1999; Rahman, Yui and Foon 2003)
2.3.2
IMPACT OF THE CRISIS
The Asian Economic Crisis has affected most of the region’s countries financially,
politically and socially.
2.3.2.1
FINANCIAL IMPACT
Every country in East and South East Asia faced recession in 1998 (Lim 2000). Table 2.3
shows the rate of economic growth for six Asian countries from 1996 to 1999.
Table 2.3: Economic Growth Rates of Selected East Asian Countries 1996-1999
Hong Kong
Singapore
Thailand
Malaysia
Indonesia
South Korea
1996
4.5
7.5
5.9
10
7.8
6.8
Period
1997
1998
5
-5.3
8.4
0.5
-1.7
-10.2
7.3
-7.4
4.7
-13
5
-6.7
1999
3.1
5.4
4.2
5.8
0.3
10.7
Source: ADB (2000c), Rahman, Yui and Foon (2003, p. 497)
The fixed exchange rate system collapsed resulting in an extreme exchange rate crisis in
68
DBA Thesis: Malinee Ronapat
the region. The Thai baht depreciated by 54%, the Malaysian ringgit by 40%, the Korean
won by 48% and the Indonesian rupiah declined by more than 83% (see Table 2.4). These
declines in the exchange rate are measured against the United State dollar (USD).
Table 2.4: Exchange Rate (per USD)
Crisis Countries
Singapore (S$)
Thailand (Baht)
Malaysia (Ringgit)
Indonesia (Rupiah)
South Korea (Won)
Jun-97
1.4264
24.318
2.5157
2427.9
887.03
Lowest Rate (Period)
1.7566 (August 1998)
52.551 (January 1998)
4.1941 (August 1998)
13995.9 (July 1998)
1693.65 (January 1998)
Depreciation (%)
-19%
-48%
-40%
-83%
-48%
Source: ADB (2000a); Lim (2000); Rahman, Yui and Foon (2003, p. 498)
The stock markets of these economies collapsed after the depreciation of their exchange
rates. Decline in the net asset value of banks and other financial institutions also resulted
from the increase in Non-Performing Loans that was discussed earlier. The withdrawal of
funds and investments from the region resulted in a skew of funds from USD 92.8 billion in
1996 to negative USD 12 billion in 1999 (Radelet and Sachs 1998; Rahman, Yui and Foon
2003). Table 2.5 contains the highest (expressed as 100%) and lowest points that were
recorded for the stock market indices of six Asian countries during the Asian Economic
Crisis.
Table 2.5: Stock Market Collapsed (1997 to 1998)
Crisis Countries
Hong Kong (Hang Seng)
Singapore (STI)
Thailand (SET)
Malaysia (KL Composite)
Indonesia (Jarkarta SE Composite)
The Philippines (Manila Composite)
Highest Pre-Crisis
Aug-98
Feb-97
Jan-97
Feb-97
Jul-97
Feb-97
Lowest Point
43% (September 1998)
40% (September 1998)
25% (September 1998)
35% (September 1998)
37% (October 1998)
35% (September 1998)
Source: ADB (2000a); Lim (2000); Rahman, Yui and Foon (2003, p. 498)
69
DBA Thesis: Malinee Ronapat
2.3.2.2
POLITICAL IMPACT
The level of political instability was relatively high during the Asian Economic Crisis and
many policy responses were unsuccessful in restoring economic growth. Political
instability was particularly high in Thailand under the government led by Charvalit
Yongchaiyudh and in Indonesia under President Suharto and President Habibie.
2.3.2.3
SOCIAL IMPACT
Conditions in the region’s labour market were also depressed during the crisis. High levels
of unemployment lowered the incomes of households and the standards of living. Social
externalities such as high levels of crime were also recorded. (Rahman, Yui and Foon
2003).
The next section will discuss the sovereign ratings given to Thailand before, during and
after the crisis. Later, more detail on the Thai capital market including its establishment,
performance and expansion from 1975 to the present day will be discussed.
2.4
THAI SOVEREIGN RATING (BY FITCH, MOODY’S, S&P)
2.4.1
THAILAND SOVEREIGN CREDIT RATING BEFORE THE CRISIS
Thailand was initially rated in 1989 by two well-known credit agencies namely Standard
and Poor (S&P) and Moody’s. A credit rating is a measurement of a country’s risk of
defaulting on its loans, compared to that of other countries. It also measures the ability of a
country to repay its debts over time (BOT 2000, TRIS 2001). Thailand was initially rated,
A- by S&P and A2 by Moody’s in the investment grade. These ratings changed little before
the Asian crisis in late 1997. The highest rating that has been granted by S&P was A and
70
DBA Thesis: Malinee Ronapat
A2 by Moody’s during the economic boom in 1995 and 1996. However, the credit upgrade
in those years was not attractive for Thai society because the Government and private
sectors rarely borrowed from foreign countries. Also foreign investors came to Thailand for
financial and direct investment due to the economic boom in those years (TRIS 2001).
2.4.2
THAILAND SOVEREIGN CREDIT RATING DURING THE CRISIS
In 1997, the two raters often downgraded Thailand’s sovereign rating in response to a
downgrading by other raters. S&P downgraded Thailand’s rating three times during a five
month period from A level to BBB- on 9 January 1998. Moody’s downgraded Thailand’s
rating four times in nine months from A2 to Ba1 on 21 December 1997. Ba1 is considered a
Speculative, or Non-investment Grade (TRIS 2001).
The announcement of the credit downgradings affected business activities directly and
indirectly. The Thai money market was especially affected, because investors consider the
credit rating of a country when making decisions. As the credit rating was downgraded, this
made investors think that Thailand had become more risky and this increased the cost of
moving funds among countries (so called capital mobilisation). Consequently, total
investments in Thailand decreased considerably in 1998. Total capital outflows from the
private sector were 645,096 billion baht (or AUD 24 billion at the rate of 27 Baht/AUD), or
AUD 12 billion higher than 1997 (TRIS 2001).
The credit downgrading also affected the political climate atmosphere in Thailand. It gave
the Government a weaker image and enhanced the level of political instability. However,
an upgrading of the credit rating has a positive effect on the Thai economy and this is
manifested in the value of the Thai baht and the number of investments at SET (TRIS
2001).
2.4.3
RECENT SOVEREIGN CREDIT RATING
Three international credit raters have recently rated Thailand differently. Moody’s rated
Thailand at Baa3, S&P at BBB- and Fitch, at BBB-. All these grades are speculative, or
71
DBA Thesis: Malinee Ronapat
non-investment types of sovereign credit ratings.
The current Sovereign Credit Rating of Thailand (2002) is shown in Figure 2.10 below:
Figure 2.10: Current Sovereign Credit Rating of Thailand by Three Raters
Raters
Sovereign Credit Rating (December 2002)
Moody's
Baa3
S&P
BBB-
Fitch
BBB-
Sources: TRIS (2002)
2.5
THAI CAPITAL MARKET
2.5.1
HISTORY OF THE STOCK EXCHANGE OF THAILAND
The creation of Thailand's first officially sanctioned and regulated securities market was
initially proposed as part of the Second National Economic and Social Development Plan
(1967-1971). The plan stressed that the market's most important role would be to mobilise
funds to support Thailand's industrialisation and economic development (SET 2002a).
The development of a modern capital market in Thailand can be divided into two phases.
The first stage involved ‘The Bangkok Stock Exchange’ which was privately owned while
the second stage involved the establishment of ‘The Securities Exchange of Thailand’ (SET
2002b).
2.5.2
ESTABLISHMENT OF THE BANGKOK STOCK EXCHANGE
Thailand’s stock market was established in July 1962, when a private group established an
organised stock exchange as a limited partnership. The name of this institution was
72
DBA Thesis: Malinee Ronapat
changed to the ‘Bangkok Stock Exchange Co., Ltd.’ (BSE) in 1963. BSE was initially
rather inactive because annual turnover consisted of only 160 million baht in 1968 and 114
million baht in 1969. Trading volumes fell sharply to 28 million baht in 1971 and BSE
ceased operations in the early 1970s. It is generally agreed that BSE failed to succeed
because of a lack of official government support and a limited level of investor
understanding of the equity market.
2.5.3
ESTABLISHMENT OF THE STOCK EXCHANGE OF THAILAND
In 1972 the Government took a step toward building a formal capital market by amending
the ‘Announcement of the Executive Council No. 58 on the Control of Commercial
Undertakings Affecting Public Safety and Welfare’. By 1975 the basic legislative
framework was in place and on April 30 1975, ‘The Securities Exchange of Thailand’
officially started trading. On January 1 1991 its name was formally changed to ‘The Stock
Exchange of Thailand’ (SET) (SET 2002c).
2.5.4
ROLES OF THE STOCK EXCHANGE OF THAILAND
SET's primary roles are:
x
To serve as a centre for the trading of listed securities and to provide the essential
systems needed to facilitate securities trading (SET 2002c).
x
To undertake any business relating to the Securities Exchange, such as a clearing
house, securities depository center, securities registrar, or similar activities (SET 2002c).
x
To undertake any other business approved by the Securities Exchange
Commission (SEC) (SET 2002c).
2.5.5
REGULATORY FRAMEWORK OF THE STOCK EXCHANGE OF
THAILAND
73
DBA Thesis: Malinee Ronapat
Figure 2.11: Regulatory Framework of SET
Source: SET (2002c)
SET’s Board of Governors is comprised of a maximum of eleven people. Its President is
appointed by the Board and is an ex-officio member of the Board. The Board is also
responsible for formulating SET’s policies and supervising the Exchange's operations.
However rules and regulations prescribed by the Board must also be approved by SEC
(SET 2002c).
Only member companies of SET are authorised to buy, or sell securities on the Exchange.
Firms which have obtained a securities licence from the Ministry of Finance (following a
recommendation by the SEC) to engage in the securities business, as stock brokers, may
apply for membership of SET. Membership status is obtained once approval is granted by
SET’s Board of Governors. At present, SET has 28 member firms. Besides brokers, there
are sub-brokers which, like brokers, take orders from investors. But sub-brokers are not
members of SET and they have to route trading orders to a broker. The broker then
transmits these orders to the SET trading system (SET 2002c).
74
DBA Thesis: Malinee Ronapat
Table 2.6: Number of Companies Obtaining Securities Business Licenses
Number of Companies
1997
1998
1999
2000
2001
2002
% increase
(since 1997)
Securities Brokerage
41
42
42
43
43
43
5%
Securities Dealing
59
59
59
59
60
63
7%
Investment Advisory Service
39
39
39
39
39
39
0%
Securities Underwriting
58
58
59
60
60
63
9%
Mutual Fund Management
14
14
14
14
14
14
0%
Private Fund Management
19
18
17
22
24
27
42%
Type of Securities Business
Source: SEC (2003)
Table 2.6 indicates that the number of companies in the securities brokerage business
increased from 41 firms in 1997 to 43 firms in 2002, representing growth of 5 percent. The
securities brokerage business includes brokerage firms and sub-brokerage firms. Securities
dealing expanded by 7 percent, from 59 firms in 1997 to 63 firms in 2002. The securities
dealing business includes finance and securities companies and commercial banks.
However, the number of firms that offer investment advisory services has been stable since
1997. The number of securities underwriting firms increased by 9 percent from 58 in 1997
to 63 firms in 2002. The securities underwriting business includes securities companies,
finance and securities companies, and commercial banks. The mutual fund management
business has been stable at 14 licenced firms since 1997. Private funds management has
shown the largest expansion at 42 percent. With only 19 firms in 1997, the number of
private funds management firms, which includes securities companies, finance and
securities companies, fund management companies, commercial banks and insurance
companies, increased to 27 licensed firms in 2002. Figure 2.12 shows the number of
securities brokerage firms, securities dealing firms, private fund management firms, and
securities underwriting firms over the period 1997 to 2002 (SEC 2001, 2002, 2003).
75
DBA Thesis: Malinee Ronapat
Figure 2.12: Number of Companies Obtained Securities Business Licenses
Number of companies obtained securities business licenses
Securities Brokerage
Securities Dealing
Securities Underwriting
Private Fund M anagement
70
63 63
59 58
Number of companies
60
59 58
60 60
59 60
59 59
50
42
41
43
42
43
43
40
27
30
24
22
19
20
18
17
10
0
1997
1998
1999
2000
2001
2002
Year
Source: SEC (2003)
76
Chapter 5: Analysis of Data and Findings
2.5.6
THE PERFORMANCE OF SET
This section discusses the conditions in the capital market from 1990 to 2002. In 1990,
the performance of the capital market was strong in the first six months and the SET
index rose from 879.2 to 1,143.8. However, after the crisis in the Persian Gulf, the index
declined to 544.3 at the end of 1990. The trading volume was, therefore, low in the last
six months of 1990. The main factors contributing to the low level of activity in the
capital market were domestic political instability and as stated above, the Persian Gulf
crisis, which resulted in strong increases in oil prices and interest rates, as well as the
collapse of overseas capital markets (BOT 1990; SET 1990).
In 1991, the level of trade at SET recovered and closed at 711.4 in December 1991,
which was 16.1 percent higher than in 1990. The volume of trading averaged 3,237
million baht per day, which is almost 30 percent higher than 1990. The volume of trade
rose in the first quarter of 1991 and there was some fluctuation later in 1991. Indeed, the
volume of trade in 1991 rose 25.6 percent above that recorded in 1990 and the index
stood at 893 points. The factor contributing to the unstable index was the political unrest
in May (referred to as ‘Black May’) which involved an overthrown of the government by
the military. However, the market experienced a rapid increase in the volume of trade
after September 1992. This favourable result was due to the implementation of corrective
policies by government, the activities of listed companies and the impact of legal action
against individuals and firms which were manipulating the market. These measures
enhanced the level of confidence and stability both in the short and long term (BOT 1991,
1992; SET 1991, 1992).
Trade fluctuated widely in 1993 and closed at 1682.9 (the highest point in the history) in
December, up by almost 90 percent over the figure recorded in 1992. The level of
fluctuation can be attributed to negative and positive factors. In early 1993, one financial
institution faced problems with liquidity and ceased operating and 30 individuals were
caught manipulating prices by SEC and SET in April 1993. Therefore, trade at the market
suffered slightly but was boosted in May by lower interest rates and higher capital inflow
(BOT 1993; SET 1993).
144
Chapter 5: Analysis of Data and Findings
In 1994, the SET index fluctuated slightly and fell to 1,360.09 points, a decline of 19.2
percent. Although the performance of listed firms improved, the economy expanded
strongly. Listed firms were permitted for the first time to open branches overseas,
although this did not overcome negative issues such as domestic political instability and
volatility in the international money market (weakening of the U.S. dollar against the Yen
and Deutsche Mark in June 1994).
The Mexican financial crisis and the associated outflow of capital from foreign investors
resulted in a decline of 5.8 percent in the SET index in 1995. In 1996, the capital market
closed at 831.57 points, which is 35.1 percent less than the 1995 close. This result can be
attributed to the poor performance of some of the listed firms and political uncertainty
(BOT 1995, 1996; SET 1995, 1996).
In 1997, the capital market fell sharply and closed at 372.69 points, or 55.2 percent below
the 1996 close. The daily volume of trade declined by almost 30 percent from the
previous year. This decline resulted from the slowdown in economic growth, operational
losses of the listed firms, the closure of 52 financial institutions, strong depreciation in
the exchange rate and a high level of financial instability due to the Asian Economic
Crisis. Many firms were delisted, or received a ‘C’ posting which signals the possibility
of a delisting (BOT 1997; SET 1997).
The SET index continued to decline in 1998 due to the unfavorable economic outlook,
low domestic liquidity and the Asian currency crisis. However, the market improved
slightly during the last four months of 1998 due to a decline in domestic interest rates,
greater stability of the Thai baht, and the transparency of the Comprehensive Financial
Restructuring Scheme which was introduced by SEC on 14 August. Due to the high
operational losses which were reported by nearly every listed firm, SET changed the
listing criteria for delisting in 1998. The previous criteria stated that every listed firm
must have a specific amount of tangible assets totalling at least 50 percent of its total
assets and should report net profits for 3 consecutive years. The new delisting rule used
the benchmark of non-negative shareholder equity. Some firms were transferred to the
145
Chapter 5: Analysis of Data and Findings
restructuring (rehabilitation) sector, rather than delisted as would be the case under old
rule (BOT 1998; SET 1998).
In 1999, SET closed at 481.92 points, almost 40 percent higher than the previous year.
The trading volume improved due to low interest rates, low stock prices, and a return to
economic growth. However, problems with non-performing loans were still unresolved
and a rise in crude oil prices ensured that economic growth remained relatively slow
(BOT 1999; SET 1999).
In 2000, the stock market closed at 269.19 points, which was 44.1 percent below the
1999 close. The trading volume also declined by 37.2 percent. The Bank of Thailand
identified the positive factors as low domestic interest rates, less inflationary pressure,
progress in resolving the problems of financial institutions and a decline in the value of
NPLs. The bank claimed that the negative factors still out weighed the positive factors.
The negative factors included an increase in public debt, high oil prices in the world
market, and the slowdown of the global economy (BOT 2000; SET 2000a, 2000b).
In 2001 the index fell to 265.22 in November. It was affected by the terrorist attack in the
US, high domestic public debt and higher NPLs. However, the index closed at 303.85
points in December which is slightly higher than the 2001 close (BOT 2001, 2002; SET
2001a, 2001b)
In 2002, the index closed at 356.48 points, nearly 20 percent higher than the figure
reached at the start of the year. However, the 2002 result was superior than the
performance of other markets in the region. This outcome was due to high performance
of listed firms, new and more diversified investments, high economic growth and low
interest rates. However, the index was affected by war in Iraq (BOT 2001, 2002; SET
2001a, 2002a). Figure 2.13 traces the movement of the SET index from 1995 to 2002.
146
Chapter 5: Analysis of Data and Findings
Figure 2.13: Stock Exchange of Thailand (SET)’s Performance from 1995 to 2003
SET index
SET index
1400
1200
1281
SET index
1000
832
800
772
600
482
400
373
356
269
200
356
304
0
1995
1996
1997
1998
1999
2000
2001
2002
2003
Year
Source: SET (2002c)
The SET index stood at 1,281 points in 1995, two years before the Asian economic crisis.
As stated above the index started to decline in 1996 and fell to its lowest level in 2000
(269 points). However, the index has risen since 2001 and now stood at 772.15 points in
2003 although it was still well below 1995 levels (SET 2002a; SET 2003).
It is important to compare the performance of Thailand’s capital market with the markets
of a number of developed and developing countries. Table 2.7 compares SET with the
exchange indices from selected countries between 1997 and 2002 (the ending date of the
study period) (SET 2002a).
147
Chapter 5: Analysis of Data and Findings
Table 2.7: Selected Foreign Stock Exchange Indices
Index
Exchange
Country
USA
1997
1998
1999
2000
2001
2002
7,908.25
9,181.43
11,497.13
10,786.85
10,021.50
8,341.63
684.61
688.99
876.97
897.75
847.61
824.38
Industrials-30
Dow Jones
AMEX-Composite
NASDAQComposite
AMEX
NASDAQ
1,570.35
2,192.69
4,069.31
2,470.52
1,950.40
1,335.51
NYSE-Composite
NYSE
5,405.19
6,299.93
6,876.10
6,945.57
6,236.39
5,000.00
FTSE-100
London
UK
5,135.50
5,882.60
6,930.20
6,222.46
5,217.35
3,940.36
Nikkei-225
Tokyo
15,258.74
13,842.17
18,934.34
13,785.69
10,542.62
8,578.95
Hang Seng
Hong Kong
Japan
Hong
Kong
10,722.76
10,048.56
16,962.10
15,095.53
11,397.21
9,321.29
Straits Times
Singapore
Singapore
1,507.65
1,392.73
2,479.58
1,926.83
1,623.60
1,341.03
All Ordinaries
Sydney
Australia
2,579.50
2,697.00
3,108.80
3,154.70
3,359.90
2,975.50
Weighted Price
Taiwan
8,187.27
6,418.43
8,448.84
4,743.94
5,551.24
4,452.45
KLSE-Composite
Taipei
Kuala
Lumpur
Malaysia
594.44
586.13
812.33
679.64
696.09
646.32
SET
Bangkok
Thailand
372.69
355.81
481.92
269.19
303.85
356.48
PSE-Composite
Manila
Philippines
1,869.23
1,968.78
2,142.97
1,494.50
1,168.08
1,018.41
JSX-Composite
Jakarta
Indonesia
401.71
398.04
676.92
416.32
392.04
424.95
Source: SEC (2003)
SET was at its highest in 1999 (481.92 points) and lowest in 2000 (269.19 points). Table
2.7 indicates that SET rose to 356.48 in 2002. The SET index continues to be small when
compared with the US, UK, Japan, Hong Kong and Taiwan. The indices with similar
values to SET include the KLSE-Composite (Malaysia) and the JSX-Composite
(Indonesia) (SEC 2001, 2002a, 2003).
There are different types of investors in the Thai capital market. Each type of investor
invests different proportions in the capital market, and accounts for a varying percentage
of the total trading value. Table 2.3 shows the percentage of buying and selling which is
attributed to three types of investors namely, (1) local institutional investors, (2) local
non-institutional investors, and (3) foreign investors.
148
Chapter 5: Analysis of Data and Findings
Table 2.8: Percentage of Buying and Selling Value
1997
100
9.94
46.81
43.25
Percentage of Buying and Selling Value (%)
Local Institutional Investors
Local Non-institutional Investors (Retail)
Foreign Investors
1998
100
5.64
59.74
34.62
1999
100
4.90
65.69
29.41
2000
100
5.69
62.12
32.19
2001
100
3.95
77.43
18.62
2002
100
5.34
72.09
22.57
Source: SET (2002b)
Note: Percentage of buying and selling value is equal to the sum of buying value and selling value divided
by the total market buying and selling value multiplied by 100.
The trading value of the three groups of investors is illustrated in Figure 2.14. This figure
presents the net trading value (buying and selling) of these investors between 1995 and
2001.
Figure 2.14: Net Trading Value Since 1995
Local institute
retail
foeign
Net Trading Value
80000
55437
60000
47302
Million Baht
40000
34016
30227
20000
-756
13377
3680
6963
-538
6006
0
-3239
-20000
-17056
-22453
-32984
-40000
-2872-3134 -948
-26987
-6426
-33068
-46546
-60000
1995
1996
1997
1998
1999
2000
2001
Source: SET (2002b), SEC (2003)
Note: ( - ) = negative value/selling more than buying
After viewing Figure 2.14 and Table 2.8 it can be seen that the net trading value of
foreign investors and retail investors have mirrored each other since 1995. This may be
due to the different information obtained and the investment objectives of these investors.
149
Chapter 5: Analysis of Data and Findings
Furthermore, the most important investors in the Thai capital market since 1995 have
been retailers and foreign investors. These two groups accounted for more than 90% of
buying and selling in the market.
2.5.7
SET VISION 2003
The SET vision for year 2003 is:
‘To be one of the most attractive capital markets in Asia by providing quality products,
representing the Thai economy truthfully, having effective risk management tools, and complying
with international standards of enforcement and corporate governance (SET 2002b)’.
Achieving the goals of SET’s vision will bring greater benefits to participants in the
market. Investors will be able to reduce their investment risks and receive better
protection of their rights. Listed companies will see an overall improvement in quality
and will be able to mobilise funds more easily at lower costs. In addition, members will
be more fairly regulated and able to generate higher business volumes (SET 2002b).
2.5.8
LISTING
As a company expands its business, additional funding is usually required to support this
expansion and further investment. Offering securities to the public is an alternative
method of finance, with lower funding costs than borrowing from financial institutions. A
public offering must be approved by SEC before the securities can be released on the
primary market. To add flexibility to the capital market and create a channel to enhance
the liquidity of public companies, these securities may be traded in the secondary market
at SET. SET plays a vital role in helping companies to raise capital and this is beneficial
not only for the company but also the country's long-term economic development. Table
2.9 shows the market capitalisation of selected foreign stock exchanges in million of US
dollars from 1997 to 2002 (SET 2002b).
150
Chapter 5: Analysis of Data and Findings
Table 2.9: Market Capitalization of Selected Foreign Stock Exchanges
Country
Millions of US Dollars
Exchange
1997
1998
152,512
126,307
n.a.
86,775
60,292
44,769
NASDAQ
1,737,510
2,524,373
5,204,620
3,578,593
2,896,856
1,994,494
NYSE
8,879,631
10,271,900
11,440,767
11,442,380
11,026,518
9,015,167
UK
London
2,068,246
2,297,651
2,954,816
2,576,990
2,149,501
1,785,199
Japan
Tokyo
2,085,370
2,439,549
4,455,348
3,157,220
2,264,528
2,069,299
Hong Kong
Hong Kong
413,323
343,567
609,090
623,400
506,073
463,055
Singapore
Singapore
104,438
94,987
192,983
152,830
115,688
99,807
Australia
Sydney
295,766
328,929
427,835
372,790
375,131
380,087
Taiwan
287,813
260,015
375,991
247,600
292,621
261,211
Malaysia
Taipei
Kuala
Lumpur
93,230
95,562
139,908
113,160
118,981
125,778
Thailand
Bangkok
22,886
34,252
57,177
29,220
35,943
45,504
Philippines
Manila
31,261
35,297
48,080
25,980
21,245
18,507
Indonesia
Jakarta
29,050
22,078
64,045
26,810
22,998
30,067
USA
AMEX
1999
2000
2001
2002
Source: SEC (2002a)
Capitalisation of the Thai market was at its lowest in 1997 (22,886 million USD), during
the Asian economic crisis. The highest level of market capitalisation for Thailand
occurred in 1999 (57,177 million USD). As stated earlier, the value of market
capitalisation fell to 29,220 million USD in 2001, but recovered to 45,504 USD in 2002.
It appears that a period of sustained expansion of Thailand’s capital market is finally
occurring after the Asian economic crisis. However, the rate of capitalisation in the Thai
market remains small when compared with the US, UK and Japan. The markets of the
Philippines and Indonesia possess a similar rate of capitalisation as Thailand (SET
2002b).
SET has placed emphasis on the quality of its listed companies by improving the
standards of its listing rules and regulations. A listed company must comply with SET's
listing requirements prior to obtaining listing status (SET 2002b). Listed companies must
publicly disclose all materials that are important for investors' decision making. SET has
implemented a full disclosure policy, which permits investors to obtain accurate,
151
Chapter 5: Analysis of Data and Findings
adequate and timely information. This helps to ensure market transparency and integrity.
Disclosure occurs by facsimile and on-line through the SET Information System (SET
2002b).
Table 2.10 presents the number of listed companies on selected foreign stock exchanges
from 1997 to 2002. The number of listed companies in Thailand reached its highest in
1997 and fell to its lowest in 2000. The most recent number of companies listed on
Thailand’s exchange is 398 firms and this is still below the number attained in 1997 (SET
2002b).
Table 2.10: Number of Listed Companies of Selected Foreign Stock Exchanges
Exchange
Number of Companies
Country
1997
1998
1999
2000
2001
2002
771
711
n.a.
693
605
572
NASDAQ
5,487
5,010
4,829
4,726
4,128
3,649
NYSE
2,626
2,669
2,592
2,468
2,400
2,366
AMEX
USA
London
UK
2,991
2,920
2,791
2,929
2,891
2,824
Tokyo
Japan
1,865
1,890
1,932
2,096
2,141
2,153
Hong Kong
Hong Kong
658
680
701
790
867
978
Singapore
Singapore
294
295
317
377
386
385
Sydney
Australia
1,219
1,222
1,287
1,409
1,410
1,421
Taipei
Taiwan
404
437
462
531
586
640
Kuala Lumpur
Malaysia
703
731
752
790
807
860
Bangkok
Thailand
431
418
392
381
382
398
Manila
Philippines
221
221
226
230
232
235
Jakarta
Indonesia
282
287
276
286
315
330
Source: SEC (2002a)
Table 2.10 indicates that the number of companies listed on the Thai stock market since
1997 has fallen. This phenomenon is also evident in the US and UK. However, stock
markets, such as Australia, Hong Kong, and Japan, have continued to increase since 1997
(see Table 2.10). The fall in the number of listed companies in Thailand, the US and UK
may be due to the bursting of the Asian investment bubble and the hi-tech boom (SEC
2002a).
152
Chapter 5: Analysis of Data and Findings
2.5.9
DISCLOSURE PROCEDURE
To ensure equal access in the SET, any disclosure of material information must be made
by the listed company at least one hour prior to the commencement of each trading
session or after the close of the day's trading. There are two trading sessions, the morning
trading session 10:00 a.m.-12:30 p.m. and the afternoon trading session 2:30-5:00 p.m..
Therefore, listed companies must submit their information:
1.
before 9.00 a.m. for the morning trading session (SET 2002a)
2.
between 12:30 and 1:30 p.m. for the afternoon trading session or (SET 2002a)
3.
anytime from after 5:00 p.m. and before 9:00 a.m. on the morning of the next
day of trading (SET 2002a).
To facilitate instantaneous disclosure, SET has established guidelines for determining the
of information which requires prompt disclosure as follows:
1.
Information which is likely to have a significant effect on the market price of
the firm's securities (SET 2002a).
2.
Information which is likely to be important for investment decisions (SET
2002a).
3.
Information which is likely to affect the interests of shareholders (SET 2002a).
2.5.10
MATERIAL INFORMATION
Vital information, which may affect the stock prices, investment decisions, or investor
interests, must be disclosed within a specific period. This information includes the
company’s report and financial statements (audited and unaudited). Vital information is
important for both technical and fundamental analyses (this will be explained later)
especially ratio and stock valuation analysis (SET 2002b). This information should be
presented on time to SET and SEC.
153
Chapter 5: Analysis of Data and Findings
Table 2.11: Financial Statement Disclosure Deadlines
1. Financial Statements
Unreviewed and unaudited financial statements
Reviewed quarterly financial statements
Audited annual or semiannual financial statements
2. Annual Report
3. Disclosure report of additional information
4. Information on any operational and financial results
which is price-sensitive and/or affects shareholder
interests
Within 30 days of the end of the accounting period.
Within 45 days of the end of each quarter.
Within 3 months of the end of the accounting period.
Within 110 days of the end of the accounting period.
Within 3 months of the end of the accounting period.
At least one hour prior to the trading session.
Sources: SET (2002b)
2.5.11
STOCK INFORMATION
The Department of Investors Relations and the Stock Exchange of Thailand (2003)
advise investors to analyse stock information before making any investment decision in
the stock market.
x
Price: The stock price is an important factor for investors, as buying and selling
affects the level of supply and demand of a particular stock and its price. Changes in
stock price reflect the amount of money involved and the investor’s decision on whether
to buy, sell, or hold a particularly stock. When valuing a stock the price has to be
considered, together with other performance variables such as earnings per share and
dividends (SET 2002b).
x
Price-Earnings Ratio (P/E Ratio): This ratio is a measure of the stock's
fundamental value. P/E Ratio is calculated by dividing the Close Price (P) with Earnings
per share (E).
P/E = Price / Earning per share
A stock with a low P/E ratio is preferable to one with a high P/E. In brief, a low P/E stock
has higher earnings potential, or is cheaper than a high P/E stock, when its earnings
potential is considered (SET 2002b).
154
Chapter 5: Analysis of Data and Findings
x
Dividends: Dividends as described by SEC, refer to the proportion of a firm’s
profits which are distributed to shareholders (common, preferred or investment unit) and
depend on the right of the holders. A preferred share’s dividend is normally fixed as a
percentage of par, while a common share’s dividend (or investment unit) depends on the
firm’s success over a fiscal year. The Board of Directors of a firm announce the dividend
payment to common shareholders, yearly, semi-annually or quarterly. The dividend to
common shareholders can be a cash dividend, or stock dividend (both concepts will be
discussed later). However, SEC stated that the stock dividend in Thailand is not as
popular as the cash dividend and firms do not want to pay stock dividends because of the
tax burden. SEC is now considering legislative changes in relation to stock dividends
(SEC 2002a).
x
Dividend Yield: The rate of dividend return is shown as a percentage. A stock
with a high dividend yield is more attractive because a higher rate of return is received in
the form of dividends.
Dividend Yield = (Dividend * 100) / Earning per share
x
Trading Volume: The trading volume, or liquidity of a stock is important
because a high trading volume indicate that investors can buy/sell a stock easily as there
are many buyers and sellers (SET 2002b).
x
Financial Analysis: This refers to an analysis of a company's growth potential,
stability, financial and management strengths and profit potential for its investors.
Financial analysis is a complex exercise which requires strong investment knowledge
(SET 2002b).
In addition to market conditions and factors specific to individual stocks, a myriad of
variables influence the stock market and the movement of prices. Investors can follow
these issues in various media reports. Table 2.12 presents an overview of the important
indicators in the Stock Exchange of Thailand (SET) from 1997 to 2002 (SET 2002b).
155
Chapter 5: Analysis of Data and Findings
Table 2.12: Overview of Stock Exchange of Thailand
1997
372.69
1998
355.81
1999
481.92
2000
269.19
2001
303.85
2002
356.48
929,598
3,764
1,133,344
855,169
3,505
1,268,199
1,609,787
6,571
2,193,067
923,697
3,740
1,279,224
1,577,758
6,440
1,607,310
2,047,442
8,357
1,986,236
Number of Listed Companies
431
418
392
381
382
389
Number of Listed Securities
P/E Ratio (times)
P/BV Ratio (times)
Dividend Yield (%)
529
6.59
0.89
6.04
494
10.04
1.05
1.34
450
14.70
1.72
0.61
438
5.52
1.11
1.78
449
4.92
1.29
2.06
471
6.98
1.36
2.72
SET Index
Total Turnover (B mil.)
Daily Average Turnover (B mil.)
Market Capitalization (B mil.)
Source: SET (2002b)
As stated earlier, the performance of the Stock Exchange of Thailand (SET) has
fluctuated since 1997. Total turnover and daily average turnover have increased since
1997, while the number of listed companies and securities has declined since 1997 (SET
2002b).
The Table 2.12 presents trends in a range of SET’s performance indicators since 1997,
including the P/E and P/BV ratios and the Dividend Yield. As stated earlier, P/E is the
price to earning ratio, P/BV is the price to book value ratio and the dividend yield is the
percentage of dividends per share to earnings per share. The price-book value ratio sets
the current share quote in relation to its book value per unit as stated by the balance sheet.
The higher this Figure, the better are a stock’s investment potential according to market
participants (SET 2002b).
The P/E ratio was at its highest in 1995. Consequently, it appears that two years before
the Asian economic crisis investors believed that Thai stocks were over-valued (SET
2002b). The P/E ratio was 6.59 in 1997 and rose to 14.7 in 1999. However, the P/E ratio
fell to 5.52 in 2002 and to 4.92 in 2001.
A stock with a high Dividend Yield is more attractive to some investors because they
receive a higher rate of return in the form of dividends. Dividend yields were at their
156
Chapter 5: Analysis of Data and Findings
highest in 1995. Consequently, listed firms paid higher dividends before the Asian
economic crisis. The dividend yield reached its lowest level (0.89) in 1997. Table 2.13
presents the turnover value of selected foreign stock exchanges in millions of US dollars
from 1997 to 2002 (SET 2002b).
Table 2.13: Turnover Value of Selected Foreign Stock Exchanges
Country
Millions of US Dollars
Exchange
1997
1998
1999
2000
2001
2002
143,230
287,929
n.a.
945,391
817,042
642,181
NASDAQ
4,481,682
5,518,946
10,466,613
19,798,799
11,000,223
7,533,875
NYSE
5,777,602
7,317,949
8,945,205
11,060,041
10,489,324
10,311,156
UK
London
1,989,489
2,887,990
3,399,349
4,558,661
4,550,503
4,006,732
Japan
Tokyo
896,055
750,831
1,675,641
2,315,502
1,544,662
1,565,824
Hong Kong
Hong Kong
453,657
206,153
230,032
376,664
241,011
194,425
Singapore
Singapore
74,138
58,510
107,406
95,152
71,771
62,770
Australia
Sydney
171,004
163,054
198,195
226,485
244,099
295,649
Taiwan
1,308,634
895,986
913,610
986,269
544,597
632,666
Malaysia
Taipei
Kuala
Lumpur
145,688
26,840
42,431
52,868
21,323
32,038
Thailand
Bangkok
28,796
20,976
37,246
21,119
31,034
41,292
Philippines
Manila
20,349
10,148
19,716
8,526
3,148
3,107
Indonesia
Jakarta
42,605
10,637
20,033
15,109
9,527
13,050
USA
AMEX
Source: SEC (2003)
As stated earlier, the turnover value of the Thai capital market fluctuated over the 6 years
between 1997 and 2002. It was at its lowest in 1998, (20,976 million USD). The highest
market capitalisation turnover for Thailand occurred in 2002 (41,292 million USD).
However, the turnover value of the Thai market is small compared to that of the US, UK,
Japan, Hong Kong and Taiwan. However, the level of turnover in the capital markets of
the Philippines, Indonesia and Thailand were similar between 1997 and 2002 (SET
2002b).
157
Chapter 5: Analysis of Data and Findings
2.5.12
INFORMATION DISSEMINATION
Listed company and trading information is distributed through various channels and they
are very important sources of information for investors and aid in enhancing the level of
accessibility for investors. The three most important sources of information are as
follows:
SETinfo: All data and information are updated in the SET database (SIMS), and
disseminated via an electronic system known as, SETINFO. A comprehensive range of
information alternatives is provided on SETinfo (SET 2001b).
Intranet: Investors can access the SET website: http://www.set.or.th to obtain
comprehensive information (SET 2001b).
Mass media and publications: Information on SET’s members and listed companies,
while its educational programs on the stock market are also available on the media and in
its publications (SET 2001b).
2.5.13
INVESTOR PROTECTION
The Stock Exchange of Thailand recognises the crucial role investor protection plays in
enhancing investor confidence and contributing to market growth. Such protection is best
understood as a combination of different, but closely integrated measures including, but
not limited to, market regulation and enforcement, trading and settlement system
reliability, information disclosure and equal accessibility.
2.5.14
SYSTEMS RELIABILITY
158
Chapter 5: Analysis of Data and Findings
The SET trading system has been fully-computerised since 1991, thereby expanding the
capacity for trade and improving efficiency and transparency. The trading system has the
capacity to handle 600,000 orders daily (SET 2001b).
2.6
CONCLUSION
This chapter has provided a brief background on Thailand, its economy, sovereign credit
ratings and capital markets. It has, therefore, provided a political, economic and
institutional setting for this study. In next Chapter reviews the literature on dividends and
the dividend disappearance phenomenon.
159
Chapter 5: Analysis of Data and Findings
C
HAPTER
3.1
3
REVIEW OF LITERATURE
INTRODUCTION
Chapter 2 discussed the development and role of Thailand’s capital market. This
chapter reviews the literature on the issue of disappearing dividends, firm’s
characteristics, ratio analysis, efficient market hypothesis (EMH), and value and growth
stocks. It also identifies gaps in the body of knowledge and aids in developing an
effective model for explaining the phenomenon of disappearing dividends.
The chapter is presented in seven sections as shown by Figure 3.1. Section 3.2
introduces the topic of financial investment, its definitions and the concepts of finance
and investment. Section 3.3 introduces the capital market and efficient market
hypothesis (EMH). Section 3.4 discusses the importance of dividends, the dividend
payment procedure, the components of dividends, share price movements, the diverse
types of dividends, factors affecting dividend policy, types of dividend policies, the
firm’s dividend payment decision and company analysis. Section 3.5 integrates the
topics of capital markets and dividends by reviewing the global experiences on the
phenomenon of disappearing dividends. The discussion also focuses on recent
experiences in Thailand. Section 3.6 identifies gaps in earlier research and develops a
model for this study. Finally, section 3.7 is a brief conclusion to this chapter.
160
Chapter 5: Analysis of Data and Findings
Figure 3.1: The Structure of Chapter Three
3.1 Introduction
3.2 Background to Financial Investment
3.3 Capital Market and Efficient Market Hypothesis
3.3.1 Capital Market, Risk and Return
3.3.2 Efficient Market Hypothesis
3.4 The Importance of Dividends
3.4.1 Dividend Payment Procedure
3.4.2 Components of Dividend Policy
3.4.3 The Relevance of Dividend Policy
3.4.4 Dividend Policy and Share
Price Movement
3.4.5 Other Forms of Dividends
3.4.7 Types of Dividend Policies
3.4.6 Factors Affecting Payment
Decisions
3.4.9 Company Analysis
3.4.8 Firms Dividend Payment
Decisions
3.5 Disappearing Dividends in Capital Market
3.5.1 Disappearing Dividends
3.5.2 Disappearing Dividends in Thailand
3.5.3 Asian Stock Markets
3.5.4 Thai Capital Market Research
3.6 Research Gaps and Model Development
3.6.1 Research Gaps and Issues
3.6.2 Research Model Development
3.7 Conclusion
Source: Developed for this research.
161
Chapter 5: Analysis of Data and Findings
3.2
BACKGROUND ON FINANCIAL INVESTMENT
This study will investigate the phenomenon of disappearing dividends and relate this
issue to the characteristics of listed firms and their propensity to pay dividends.
Changes in the characteristics of the listed firms and their propensity to pay dividends
will be discussed in relation to Thailand. This section commences with a discussion of
definitions and the concepts of finance, assets and markets. The final subsection links
the capital market with the efficient market hypothesis.
3.2.1 DEFINITION AND OBJECTIVE OF FINANCIAL INVESTMENT
Finance is ‘the study of how individuals, institutions, and businesses acquire, spend,
and manage their financial resources’ (Melicher and Norton 2000, p. 3). Finance has its
origins in economics and accounting. Economists apply the supply-and-demand
framework, while accountants use a record-keeping mechanism to record revenues,
expenses and profitability (Case and Fair 1996; Melicher and Norton 2000).
The interaction, or mixture of intermediaries, markets, instruments, policy makers, and
regulations which aid the flow of funds from savings to investments is called the
financial system (Melicher and Norton 2000). There are three basic requirements for a
modern financial system (Melicher and Norton 2000).
1. An efficient monetary system for creating and transferring money. The
financial system is an efficient medium for exchanging and paying for goods
and services and measuring prices (Melicher and Norton 2000).
2. Capital formation is the channelling of savings into investment. It takes place
whenever resources are used to produce buildings and equipment for the
production of goods for consumers, or producers (Melicher and Norton 2000).
3. Financial markets are where people buy and sell, or transfer measures of (or
claims to) wealth, such as real estate, or financial assets (stocks, bonds). They
encourage investment by providing the means for savers to quickly and easily
162
Chapter 5: Analysis of Data and Findings
convert their claims into cash when needed (Reilly and Norton 1999; Melicher
and Norton 2000).
This study will focus on financial markets because they are important for modern
business. Financial markets help corporations raise money by selling different types of
financial assets with different maturities (Brealey and Myers 1991, 2000). There are
two types of assets, namely: (1) Real assets which include the ownership of land,
buildings, machinery, inventory and even precious metals; and (2) Financial assets,
which are claims against the income and assets (wealth) of those who issued them.
Examples of financial assets include securities, bonds and options (Brealey and Myers
1991, 2000; Reilly and Norton 1999; Melicher and Norton 2000).
Liquidity is one of many properties that distinguish real assets from financial assets.
Liquidity refers to the ease of converting an asset into money, with minimal costs of
exchange (Fisher and Jordan 1991). Real assets are generally less liquid than financial
assets because they are adapted to a specific use and yield benefits only in limited
circumstances (Fisher and Jordan 1991). Furthermore, returns on real assets are difficult
to measure and the markets for these assets are not broad, ready and active (Fisher and
Jordan 1991; Haugen 1995; Melicher and Norton 2000). This study will concentrate on
investment in financial assets.
There are two key financial markets (Brealey and Myers 1991, 2000; Gitman, Juchau
and Flanagan 2002) which are classified by maturities (either short-term or long-term).
x
Money markets are markets where financial assets, such as debt instruments of
1 year, or less, (hereafter called short-term securities) are traded. (Brealey and
Myers 1991, 2000; Madura 2000; Gitman, Juchau and Flanagan 2002).
x
Capital markets are markets for financial assets such as debt securities (notes
and bonds) and debt instruments (mortgages and stocks) which mature after a
period of time which is greater than one year (hereafter called long-term
securities) (Brealey and Myers 1991, 2000; Gitman 2000; Gitman, Juchau and
Flanagan 2002).
163
Chapter 5: Analysis of Data and Findings
Many financial assets do not require the existence of a formal financial market. Only
securities and debt instruments are marketed and traded in financial markets (Melicher
and Norton 2000).
This study will focus on capital markets because they help firms raise long-term funds
by issuing securities to the market (so called flotation or public offering). According to
Fisher and Jordan (1991, p. 3), ‘capital markets help firms grow, jobs are provided,
families prosper, and the economy grows’.
3.3
CAPITAL MARKET AND EFFICIENT MARKET HYPOTHESIS
This section commences with a definition of the term ‘capital’ (debt and equity capital).
Capital markets, the place where different types of assets are bought and sold will then
be outlined. A variety of types of investments will be identified and the concepts of risk
and return will be defined. In the latter stage of this section, the efficient market
hypothesis (EMH) is discussed to identify the variety in the types of efficient markets,
as defined by Fama (1970).
3.3.1 CAPITAL MARKET, INVESTMENT, RISK AND RETURN
The term capital refers to the ‘long-term funds of the firm’ (Gitman, Juchau and
Flanagan 2002, p. 37). Therefore, when analysing a firm’s balance sheet, all the
liabilities, or financing items, excluding short-term liabilities, are classified as longterm capital (Melicher and Norton 2000; Gitman, Juchau and Flanagan 2002).
Long-term capital consists of debt capital and equity capital. Debt capital ‘includes all
long-term borrowing incurred by the firm, including bonds’ (Gitman, Juchau and
Flanagan 2002, p. 37). Equity capital ‘consists of long-term funds provided by the
164
Chapter 5: Analysis of Data and Findings
firm’s owners, the shareholders’ (Gitman, Juchau and Flanagan 2002, p. 37). There are
two firms of equity capital, namely, internal and external raised capital. Capital is
raised internally by spending a firm’s retained earnings, while capital is raised
externally by selling ordinary (common), or preferred shares to the public (Gitman,
Juchau and Flanagan 2002). The key differences between debt and equity capital are
summarised in Table 3.1.
Table 3.1: Key Differences Between Debt and Equity Capital
Type of capital
Debt
No
Voice of Management
Senior to equity
Claims on Income and Assets
Stated
Maturity
Interest deduction
Tax Treatment
Source: Gitman, Juchau and Flanagan (2002), p. 37
Equity
Yes
Subordinate to debt
None
No deduction
Equity capital is necessary for a firm’s growth because all firms must initially be
financed with equity capital. There are four key differences between debt and equity
capital (see Table 3.1), which ensure that equity capital is the preferred source of
finance of a firm’s managers and more interesting as a topic of study (Reilly and Brown
2000; Gitman, Juchau and Flanagan 2002). The key differences are as follows:
Voice of Management (or voting rights): Holders of equity capital (shareholders) are
owners of the firm. They have voting rights which determine the make up of a firm’s
board of directors and other special issues, while holders of debt capital (lenders) do not
have these rights (Brealey and Myers 1991, 2000; Melicher and Norton 2000; Gitman
2000; Gitman, Juchau and Flanagan 2002).
Claims on Income and Assets: Shareholders have claims on the income and assets of a
firm after interest has been paid to creditors or lenders. The remaining funds, or
residual amount can be distributed as dividends and/or retained by the firm. The
shareholders’ claims on a firm’s assets are repaid if the firm fails and its assets are sold
on the market. However, the claims of shareholders are of lesser value than those of a
165
Chapter 5: Analysis of Data and Findings
firm’s creditors. This is because shareholders expect greater returns from dividends
and/or growth in share (stock) prices (Brealey and Myers 1991, 2000; Melicher and
Norton 2000; Gitman, Juchau and Flanagan 2002).
Maturity: equity capital is a perpetual form of financing because it does not mature,
and no repayment is required. Debt capital, however, has a maturity date (Brealey and
Myers 1991, 2000; Melicher and Norton 2000; Gitman, Juchau and Flanagan 2002).
Tax Treatment: interest payments from debt capital is tax-deductible, while dividend
payments from equity capital are taxable (Brealey and Myers 1991, 2000; Melicher and
Norton 2000; Gitman, Juchau and Flanagan 2002).
In summary, equity capital does not mature or involve repayments and its holders have
voting rights. However, dividend payments for equity capital are not tax-deductible.
The level of dividends also depend on a firm’s annual profits and debt capital is repaid
before equity capital if a firm fails (Brealey and Myers 1991, 2000; Melicher and
Norton 2000; Gitman, Juchau and Flanagan 2002).
Firms can raise equity capital externally by issuing ordinary or preferred shares
(Brealey and Myers 1991, 2000; Melicher and Norton 2000; Gitman, Juchau and
Flanagan 2002).
(1) Ordinary Shares
Holders of ordinary shares are owners of a corporation which stock share in the
company’s success and problems (Reilly and Brown 2000). Common shareholders can
vote to select a firm’s board of directors, and help decide the outcome of major issues
facing the firm, such as changes to its corporate charter and potential mergers with
other companies (Melicher and Norton 2000; Christensen and Jordan 2001; Gitman,
Juchau and Flanagan 2002). Common shareholders have a claim on a firm’s remaining
profits after the holders of all other classes of debt and equity securities have received
their interest payments or returns (Melicher and Norton 2000; Christensen and Jordan
2001; Gitman, Juchau and Flanagan 2002). In the event of liquidation creditors,
166
Chapter 5: Analysis of Data and Findings
bondholders, and preferred stock holders must be paid in full before common
stockholders receive any proceeds from the firm (Melicher and Norton 2000;
Christensen and Jordan 2001; Gitman, Juchau and Flanagan 2002). However, owners of
a firm’s common stock are entitled to receive dividend payments.
Common stock can be divided into two groups, Class A which does not have voting
rights and Class B which has voting rights. Par Value is used mainly for accounting
purposes and some legal needs. It has little relationship to the current price of the shares
(Melicher and Norton 2000).
(2) Preferred Shares
A preferred share is an equity security with a senior claim on a firm’s earnings and
assets (Melicher and Norton 2000). In the event of liquidation, the claims of preferred
shareholders are satisfied before common shareholders receive any proceeds from the
firm (Melicher and Norton 2000; Christensen and Jordan 2001; Gitman, Juchau and
Flanagan 2002). Preferred shareholders receive fixed dividends, which are paid before
common shareholders’ receive their dividends (Melicher and Norton 2000). This
dividend is specified as a percentage of par value, or a fixed amount per year (Melicher
and Norton 2000). Unlike common stock, the par value of preferred stock is important,
because dividends are expressed as a percentage of par and the par value represents the
holder’s claim on corporate assets in the event of liquidation (Gitman, Juchau and
Flanagan 2002; Christensen and Jordan 2001; Melicher and Norton 2000). Table 3.2
presents the variety of claims and characteristics of preferred and common stocks.
Table 3.2: Characteristics of Preferred Stock and Common Stock
Voting
Priority for
Priority claim
Higher
Higher
right
dividend
in bankruptcy
risk
Expected return
D
D
D
Common stock
D
Preferred stock
D
Source: Developed from this research
167
Chapter 5: Analysis of Data and Findings
Table 3.3 indicates that there are a number of advantages and disadvantages of raising
new funds by issuing common shares, debt and preference shares (Petty et al. 2000).
Table 3.3: Advantages and Disadvantages of Common Shares
Advantages to the firm
Disadvantages to the firm
(1) The firm is not legally obligated to
pay dividends. Thus, during financial
distress, there is no need to have cash
outflow from dividends, while there
must be with debt and preference
shares.
(2) Common shares have no maturity
date, thus, no cash outflow associated
with their redemption.
(3) By issuing common shares, the firm
is increasing its financial base and its
future borrowing capacity. By issuing
debt, the firm is cutting its ability to
borrow.
(1) Issuing debt is tax-deductible while
dividends on common shares are not.
(2) The issue costs on new equity tend
to be larger than those on new debt
(3) Common shares are issued with
voting rights. Therefore, issuing new
common shares may result in a change
in the ownership and control of the
firm.
Source: Petty et al. (2000)
Table 3.4 indicates that there are advantages and disadvantages of issuing preference
shares over debt and common shares (Petty et al 2000).
Table 3.4: Advantages and Disadvantages of Preference Shares
Advantages to the firm
Disadvantages to the firm
(1) Preferred shares do not have
default risk to the issuer arising from
the non-payment of dividends.
(1) Debt interest is tax deductible
while preferred shares dividend is not,
thus, increasing cost of borrowing for
the firm.
(2) Too much issuance of preferred
shares can lead to the omission of
dividend by the firm.
(2) Preferred shares carry dividends
limited to a fixed amount. They do
not participate in excess earnings like
common shares.
(3) No dilution of the ownership
control.
(4) Preferred shares get higher
dividend compared to debtholders
but lower than common shareholders.
(5) Preferred shares increase the
firm's equity base and enable the firm
to borrow more in the future.
Source: Petty et al. (2000)
168
Chapter 5: Analysis of Data and Findings
Capital equities are traded in capital markets, or securities markets (Melicher and
Norton 2000; Gitman, Juchau and Flanagan 2002).
Securities markets are classified by the types of securities bought and sold through
them (Fisher and Jordon 1991). The broadest classification depends on whether the
securities are new issues, or are already owned by someone (Fisher and Jordon 1991;
Melicher and Norton 2000). Securities markets can be classified into (1) a primary
market and (2) a secondary market according to the types of securities which are
traded (Fisher and Jordan 1991; Melicher and Norton 2000; Gitman, Juchau and
Flanagan 2002).
(1)
Primary Market
The Primary market is the market for newly issued securities. When firms wish to
raise capital the initial sale of newly issued debt, or equity securities is called a
Flotation (Melicher and Norton 2000). The initial sale of equities to the public is called
an Initial Public Offering (IPO) (Melicher and Norton 2000). To raise money, firms
often use the services of investment bankers. The main activity of investment bankers
is to market securities and to act as intermediaries between corporations and the general
public when corporations wish to raise capital (Fisher and Jordan 1991; Melicher and
Norton 2000; Petty et al. 2000).
(2)
Secondary Market
The Secondary market is the market for securities already bought, sold, or owned by
investors (Fisher and Jordon 1991). Selling securities to investors would be difficult if
there was no easy way to profit from their holdings, or no way to sell them for cash
(Melicher and Norton 2000). The secondary market provides liquidity and
marketability for investors. According to Melicher and Norton (2000), the benefits of
the secondary market are as follows:
169
Chapter 5: Analysis of Data and Findings
x
The secondary market provides liquidity.
x
The secondary market allows investors to shift their assets into different
securities and markets.
x
The secondary market provides information pricing.
x
The secondary market provides a means for evaluating a firm’s
management and for managers to determine how investors are interpreting
their actions.
The securities market is beneficial for issuers and investors (people who invest their
money in capital markets) (Fisher and Jordan 1991).
The Government and firms are issuers of securities. Securities markets help issuers to
raise funds. They also ensure savings are converted to capital and directed to productive
investments. Governments also borrow to fund their activities. In short, the market
helps to efficiently transfer funds from surplus to deficit sectors (Fisher and Jordon
1991; Reilly and Norton 1999).
Investors also benefit from the activities of securities markets. As mentioned earlier,
securities markets provide investors with liquidity to convert assets to cash. Investors
could not easily resell, or buy securities without a liquid market and they would be
hesitant to acquire them. This would reduce the quantity of funds which are available to
the finance industry and the government (Fisher and Jordon 1991; Reilly and Norton
1999). ‘Those who own securities must be assured of a fast, fair, orderly and open
system of purchase and sale at known prices’ (Fisher and Jordon 1991, p. 17).
Investors provide the primary source of capital for business (Brigham, Gapenski and
Ehrhardt 1999). Fisher and Jordan (1991, p. 2) defined investment as ‘a commitment of
funds with the expectation of a positive rate of return.’ Investors include stockholders,
bondholders and lenders such as banks (Brigham, Gapenski and Ehrhardt 1999).
Brigham, Gapenski and Ehrhardt (1999, p. 39) suggested ‘if the investment is properly
undertaken, the return will be commensurate with the risk the investor assumes’.
170
Chapter 5: Analysis of Data and Findings
Further, investors ‘are interested in a good rate of return, earned on a consistent basis
for a relatively long period of time’ (Fisher and Jordan 1991, p. 2).
The Fundamentals of Risk and Return
Investors are concerned with two key factors (1) the level of risk and (2) the rate of
return.
Risk is ‘the chance of financial loss, or the variability of returns associated with a given
asset’ (Gitman, Juchau and Flanagan 2002, p. 196).
Return is ‘the total gain or loss experienced on an investment over a given period of
time. The level of return is calculated by dividing the asset’s change in value plus any
cash distributions during the period by its beginning-of-period investment value’
(Gitman, Juchau and Flanagan 2002, p. 196).
Investors have different feelings about risk. There are ‘three basic risk preference
behaviours: risk averse, risk indifferent and risk seeking’ (Gitman, Juchau and Flanagan
2002, p. 197).
Risk-Indifferent Investors feel that ‘no change in return would be required for an
increase in risk’ (Gitman, Juchau and Flanagan 2002, p. 197).
Risk-Averse Investors feel that ‘an increase in return would be required for an
increase in risk’ (Gitman, Juchau and Flanagan 2002, p. 197).
Risk-Seeking Investors feel that ‘a decrease in return would be required for an
increase in risk’ (Gitman, Juchau and Flanagan 2002, p. 197).
Most investors are risk-averse. For a given increase in the level of risk, an increase in
the rate of return is desired. Higher returns will compensate investors for holding high
risk assets. Therefore, different classes of investments are associated with different
levels of risk and rates of return (Reilly and Norton 1999; Gitman, Juchau and Flanagan
2002).
171
Chapter 5: Analysis of Data and Findings
Investment Categories
Investment is viewed as ‘a single asset held in isolation, or a portfolio which is a
collection, or group, of assets’ (Gitman, Juchau and Flanagan 2002, p. 196). When
making investment decisions, investors make selections from various investment
categories, including: Debt Instruments, Institutional Deposits and Contracts,
Government Debt Securities, Private Issues, International Bond Investing, Equities
Real Estate, Options and Futures (Brealey and Myers 1991, 2000; Gitman, Juchau and
Flanagan 2002).
3.3.2
EFFICIENT MARKET HYPOTHESIS (EMH)
As stated earlier, markets which exchange securities are liquid so that firms can obtain
finance (Gitman, Juchau and Flanagan 2002). ‘A market is efficient if current market
prices fully reflect available information’ (Fama 1970, p. 383). Information should be
publically and privately available to investors in the market place. When all available
information is freely available, there are no transaction costs, market participants are
price takers and a perfect capital market is said to exist (Fama and Miller 1972, Fama
1976, Fama 1998). Fama and Miller (1972) added that:
‘Such a market has a very desirable feature, in particular, at any point in time market
prices of securities provide accurate signals for resource allocation; that is, firms can
make production-investment decisions, and consumers (investors) can choose among the
securities that represent ownership of firms’ activities under the presumption that security
prices at any time ‘fully reflect’ all available information. A market in which prices fully
reflect available information is termed efficient’ (Fama and Miller 1972, p. 335).
In an efficient market, it is impossible to receive above-average returns and there are no
reasons to believe that current prices are too high or low (Christensen and Jordan 2001).
Academics and investors have attempted to measure the performance of the market’s
price mechanism and to apply descriptions to the market. The accepted description is
efficient-markets theory (EMT) or, the efficient market hypotheses (EMH) of Fama
172
Chapter 5: Analysis of Data and Findings
(1970). EMH has been accepted as a description of the world’s major stock markets
such as the New York Stock Exchange (Firth 1977) and the London Stock Exchange
(Firth 1977; Fisher and Jordan 1991; Haugen 1995; Petty et al. 2000; Christensen and
Jordan 2001; Gitman, Juchau and Flanagan 2002).
EMH states that a market is efficient with respect to a given information set, if it is
impossible to make profits by trading on the basis of this information set (Jensen 1978).
Fama (1970) divided EMH into three categories of market efficiency depending on the
type of information which is available: (1) Weak, (2) Semi-strong, and (3) Strong
(Fama 1970).
Weak market efficiency is present if past prices cannot be used for forecasting future
prices (Fama 1970; Jensen 1978; Sutcliffe 1997). This implies that technical analysis is
ineffective (Fama 1970; Jensen 1978; Sutcliffe 1997). Technical Analysis is the study
of an investor’s psychology by observing his or her behaviour (Fama 1970; Jensen
1978; Sutcliffe 1997). This technique is widely used by stock brokers, speculators and
investment advisors (Sutcliffe 1993, 1997). See figure 3.2.
In a semi-strong market, public information is ineffective for forecasting futures prices
(Fama 1970; Jensen 1978; Sutcliffe 1997). This implies that fundamental and technical
analysis is inappropriate (Fama 1970; Jensen 1978; Sutcliffe 1997). ‘Fundamental
analysis is applied when a security analyst uses public information (e.g. published
accounts, interviews with the company management and public statements by the
company) to value, predict or forecast the prices of stock (Fisher and Jordan 1991, p.
612). Therefore, trading techniques which are past pricing and published information
are not helpful for investors that wish to increase their returns in a semi-strong market
(Fama 1976 1998). See figure 3.2.
When the level of efficiency is strong, all information (both public and private) is
ineffective for forecasting future prices (Fama 1970; Jensen 1978; Sutcliffe 1997). This
also implies that the use of a firm’s internal information (insider trading) is not
profitable (Fisher and Jordan 1991). See figure 3.2.
173
Chapter 5: Analysis of Data and Findings
Fundamental research says that ‘prices reflect all relevant information and that
changes in prices are unpredictable’ (Brealey and Myers 2000, p. 358). However
technical research says that ‘current prices reflect past information and prices, and that
future price changes cannot be predicted from past prices’ (Brealey and Myers 2000, p.
358).
If the capital market is not efficient, technical and fundamental analysis could be
profitable. However, if the capital market is at least weakly efficient, technical analysis
will be ineffective, but fundamental analysis can still be useful. If the efficiency of a
capital market is semi-strong, fundamental analysis will also be ineffective (Sutcliffe
1993). Fundamental analysis can therefore, be applied in markets with less than semistrong efficiency, while technical analysis can be applied only when the level of
efficiency is less than weak (Fama 1976). This supports the view that fundamental
analysis is effective in a wider range of markets than technical analysis (Fama 1976,
1998). If public information is regularly updated and readily available to investors,
fundamental analysis can be very useful in investment strategies (Fama 1976, 1998).
Figure 3.2: Type of Information in the Capital Markets
Past Prices
Public Information
Private
Information
All Information
Source: Jensen (1978); Sutcliffe (1997)
174
Chapter 5: Analysis of Data and Findings
After assessing the information which is available to the public, fundamental investors
and analysts normally estimate the intrinsic value of a stock. This estimate is equivalent
to its equilibrium share price after trading in the current market (Fisher and Jordan
1991). ‘Intrinsic value, or fair value, is the present value of the investment’s expected
future cash flows, discounted at the investor’s required rate of return’ (Petty et al. 2000,
p. 263). Future cash flows can be estimated using publicly available information.
Therefore, the intrinsic value of a security is equivalent to its investor value (Petty et al.
2000).
After the investor has determined the intrinsic value of a security, this can be compared
with the security’s current market value (market price) (Petty et al. 2000). If the
intrinsic value is not equal to the current market value (price), the security is mispriced
(Firth 1977). A mispriced security can be categorised in three ways. If the intrinsic
value is greater than the market value, the security is underpriced (Petty et al. 1991).
If the intrinsic value is lower than the market value, the security is overpriced (Petty et
al. 1991). If the intrinsic value is equal to the market value, the market is efficient
(Petty et al. 2000) because price reflects all the available information. Over time, the
forces of supply and demand tend to adjust the price of a mispriced share towards its
intrinsic value (Fisher and Jordan 1991). Investors and analysts sell overpriced stocks,
thereby driving their prices towards their intrinsic value. The same scenario applies to
underpriced stocks (Fisher and Jordan 1991).
Figure 3.3 indicates that prices of overpriced and underpriced stocks move towards
their intrinsic value as times passes.
175
Chapter 5: Analysis of Data and Findings
Figure 3.3: Removal of Mispricings
Overpriced
Share price
Intrinsic
Value
Intrinsic
Value
Underpriced
Time
Source: Sutcliffe (1997)
The decision rules of investors and analysts are summarised below:
Investors will buy a stock if the intrinsic value is greater (>) than the current price, or
will sell a stock if the intrinsic value is less (<) than the current price. Table 3.5
illustrates the decision rules for investors and analysts.
Table 3.5: Decision Rules for Investors and Analysts
Decision rules
Intrinsic Value > Market Value
Buy
Sell
D
Intrinsic Value < Market Value
D
Source: Developed for this research
In practice, no stock market is perfectly efficient in the terms described above (Firth
1977). Complete knowledge about a firm is not publicly available, and if it was, it is
unlikely that it would be interpreted accurately enough to reflect the value of the share
price (Firth 1977). Although there are shortcomings with EMH, researchers have
considerable scope for measuring the performance of the stock market and the
relevance of efficient-markets theory (Fama 1970; Firth 1977).
176
Chapter 5: Analysis of Data and Findings
3.4
THE IMPORTANCE OF DIVIDENDS
As discussed earlier, firms issue equity capital to obtain funds from investors in equity
markets. However, investors require a reasonable return on the funds they lend to firms.
This leads to a discussion of dividends and dividend policy.
3.4.1 DIVIDENDS AND STOCK VALUATIONS
The profits of successful firms can be reinvested, used to acquire securities, retire debt
or distributed to shareholders as dividends (Bringham, Gapenski and Ehrhardt 1999).
Dividends are a cash flow to shareholders of firms (Gitman 2000; Gitman, Juchau and
Flanagan 2002). They are considered a return for holding stocks and provide a
method for determining the value of a share (Gitman 2000; Gitman, Juchau and
Flanagan 2002).
The Dividend Discount Model (DDM) is used for determining the value of shares in
equity markets (Reilly and Brown 2000). Dividends are the only direct cash flows from
firms to investors (Brealey and Myers 2000; Gitman, Juchau, Flanagan 2002). The
concept of discounted dividend is expressed as:
PV (stock) =
PV (expected future dividends)
The reduced form of the Dividend Discount Model (DDM) is known as the Gordon
model.
PV (stock) =
Expected Dividend / (Discount rate - Growth)
This model is useful for discussing the valuation of stable or mature firms, where an
assumption of relative constant growth and a stable discount rate for the long term is
appropriate (Reilly and Brown 2000). Therefore, investigating the dividend
phenomenon will contribute to a knowledge of the stock’s return and its valuation.
177
Chapter 5: Analysis of Data and Findings
Investors may buy stocks and expect to receive capital gains in addition to dividends
(Brealey and Myers 2000). Consequently, the cash flow from investing in stocks is in
two forms (1) cash dividends and (2) capital gains or losses. The current price of a
stock is equal to the sum of future dividends and the expected future price of the stock
discounted to the present (Brealey and Myers 2000). Therefore;
PV (stock) =
PV (Future dividend + Expected Future stock price)
Although firms prefer issuing common stocks rather than preferred stocks, investors are
aware that common shares are more risky than preference shares (Brealey and Myers
2000).
Common stocks are more difficult to value than other securities for three reasons:
Firstly, the cash flows of common stocks are undertaken because the value of
dividends and expected future stock prices, are not known in advance (Christensen and
Jordan 2001).
Secondly, common stocks do not mature, therefore they have an indefinite life span
(Christensen and Jordan 2001).
Thirdly, the required market rate of return is not easy to observe (Christensen and
Jordan 2001).
Two main classes of techniques have been used for valuing stocks:
(1) Discounted Cash-Flow Valuation Techniques: the value of stocks depends on the
present value of several cash flow items, such as dividends, the operating cash flow, the
free cash flow, the Dividend Discount Model, the present value of operating cash flows
and the present value of free cash flows (Reilly and Brown 2000).
(2) The Relative Valuation Techniques: the value of stocks depends on their current
prices relative to several significant variables. These variables include the
Price/Earnings Ratio (P/E), Price/Cash Flow Ratio (P/CF), Price/Book Value Ratio
(P/BV) and Price/Sales Ratio (P/S) (Reilly and Brown 2000).
178
Chapter 5: Analysis of Data and Findings
This study will investigate the phenomenon of disappearing dividends, which states that
dividends are no longer being paid by listed firms (Fama and French 2001). If this
phenomenon exists, it will raise the degree of complexity for valuing stocks when using
the discounting dividend technique and analysing ratios which are associated with
dividends (such as dividend yields and dividends per share).
If a firm does not wish to distribute its earnings as dividends, it holds them as retained
earnings which can be re-invested for upgrading a firm’s operations (Gitman, Juchau
and Flanagan 2002). If the firm requires finance, it can use the retained earnings,
(internal financing) or borrow from elsewhere (external financing) (Gitman, Juchau and
Flanagan 2002). Therefore, if a firm pays high dividends it is likely that it will need to
borrow more funds from external sources (Gitman 2000; Gitman, Juchau and Flanagan
2002).
As stated earlier, dividends are an observable cash flow (Wetherilt and Weeken 2002)
for investors. It is also possible to measure and analyse these flows.
3.4.2 COMPONENTS OF DIVIDEND POLICY AND DIVIDEND PAYMENT
PRODECURES
A firm’s dividend policy has two basic components:
(1)
A Dividend Payout Ratio which indicates the value of dividends paid relative to
the company’s earnings (such as 50% of the earnings or 60% of the earnings) (Petty et
al. 2000).
(2)
The Stability of Dividends is important because profits and investment
opportunities vary over time (Brigham, Gapenski and Ehrhardt 1999). Investors desire
certainty with future payments of dividends, but it is impossible to accurately estimate
the value of the dividends they will receive in any specific period in the future.
However, several indicators can help investors estimate the level and stability of a
firm’s dividends. These indicators help enhance the level of confidence of investors and
aid in their future decision making (Petty et al. 2000).
179
Chapter 5: Analysis of Data and Findings
x
Listed firms forecast the level of their earnings and dividends five to ten years
in advance, however these estimates are not available to the public. However,
analysts prepare similar forecasts which are available to the public (Brigham,
Gapenski and Ehrhardt 1999).
x
A Stable Dividend Policy commits a firm to pay the same value of dividends
each year (Brigham, Gapenski and Ehrhardt 1999). Firms with unstable
earnings may be unwilling to make such a commitment and investors would not
be aware of this decision by management, because it would not be detailed in
the firm’s annual report.
When determining a policy for dividends, a firm’s management faces a trade-off
between the satisfaction of shareholders and the required level of external financing
(Petty et al. 2000; Brigham, Gapenski and Ehrhardt 1999). If a firm makes large
dividend payments there will be less retained earnings (or profit retention) and possibly
a greater need to borrow from external sources (Petty et al. 2000). The reverse is true
when a firm pays a lower level of dividends. Figure 3.4 illustrates a firm’s trade-off
between large and small payments of dividends.
Figure 3.4: Dividend-Retention-Financing Trade-Offs
Given Investment Decisions Debt-Equity mix
Choices between
Large Dividends
Lower Profit
Retention
Heavy External
Equity Financing
Small Dividends
or
High Profit
Retention
Negligible External
Equity Financing
Source: Petty et al. (2000), p. 466
180
Chapter 5: Analysis of Data and Findings
The value of dividends which will be paid to shareholders is decided by the firm’s
board of directors. The firm’s past performance and history of dividend payments are
considered before the announcement of the next dividend payment to shareholders. The
dividend payment procedures include of the amount to be paid, the record date and the
date of payment (Gitman 2000; Gitman, Juchau and Flanagan 2002). These procedures
are discussed below:
Record Date
Firms usually pay dividends twice a year. Dividend policies vary between countries,
but generally firm’s will announce their dividend policies in the financial news media
(Gitman, Juchau and Flanagan 2002). Gitman, Juchau and Flanagan (2002, p. 450)
mentioned that the ‘date of dividend records is when all shareholders are notified that
they will receive their dividends on a specific date.
Ex-Dividend Date
According to Gitman, Juchau and Flanagan (2002) the ex-dividend date is a period
which commences five business days prior to the date of record. All shareholders who
sell their stocks during this period will not receive the current dividend. To be eligible
for a dividend payment, investors must buy shares before the ex-dividend date (Fisher
and Jordan 1991; Melicher and Norton 2000; Gitman, Juchau and Flanagan 2002).
Payment Date
The payment date is the date when dividends are distributed to shareholders (Gitman
2000; Petty et al. 2000; Gitman, Juchau and Flanagan 2002).
Figure 3.5 illustrates the sequence of payment dates from the announcement date to
payment date.
181
Chapter 5: Analysis of Data and Findings
Figure 3.5: Sequence of Dividend Payment Dates
Announcement
Date
Record Date
Declaration
Date
Payment Date
Ex-dividend
Date
Source: Petty et al. 2000, p. 493.
3.4.3 THE RELEVANCE OF DIVIDEND POLICY
The Residual Theory of Dividends suggests that the dividends paid by a firm should be
viewed as a residual amount remaining after all investment opportunities have been met
(Gitman, Juchau and Flanagan 2002).
When a firm makes long-term investment decisions, the cost of capital is an important
concept (Gitman, Juchau and Flanagan 2002). ‘The cost of capital is the rate of return a
firm must earn on its project investment to maintain the market value of its shares’
(Gitman, Juchau and Flanagan 2002; p. 392). The cost of capital is a benchmark for
justifying whether an investment should be undertaken. Projects with rates of return
greater than the cost of capital will increase the value of the firm and should be
accepted (Gitman 2000; Petty et al. 2000; Gitman, Juchau and Flanagan 2002).
Firms desire and attempt to maintain an optimal mix of debt and equity financing
(Gitman, Juchau and Flanagan 2002). The firm’s capital structure is comprised of three
sources of finance, namely, (1) long-term debt (2) preference share capital and (3)
common share capital (Gitman 2000; Gitman, Juchau and Flanagan 2002). The overall
cost of capital can be defined as:
ka
=
( w i x k i ) + ( w p x k p ) + ( w s x k r or n )
182
Chapter 5: Analysis of Data and Findings
where
ka
=
the cost of capital
wI
=
the proportion of long-term debt in the capital structure
wp
=
the proportion of preference share capital in the capital structure
ws
=
the proportion of common share capital in the capital structure
w s + w p + w I = 1 or 100%
kI
=
after-tax cost of long-term debt in the capital structure
kp
=
after-tax cost of preference share capital in the capital structure
k r or n
=
after-tax cost of common share capital in the capital structure
If the project’s return is greater than, or equal to the weighted marginal cost of new
finance, the firm should accept the investment project’ (Gitman, Juchau and Flanagan
2002, p. 412). However, if the returns from the investment are lower than the cost of
capital, the firm should distribute the earnings (as a dividend) to stockholders (Gitman
2000).
Dividend Relevance Arguments
The dividend relevance argument has been supported by Gordon (1962) and Lintner
(1962). These two scholars suggested that there is a direct, or positive correlation
between a firm’s dividend policy and the market value of this firm. Lintner (1962) and
Gordon (1962) suggested the bird-in-the-hand argument which says that investors are
generally risk-averse (investors try to minimize risk, or if the level of risk is high they
will expect a higher rate of return as compensation for consuming high levels of risk).
Therefore, the payment of dividends should reduce the level of uncertainty (risk) of
holding stocks. If the level of dividends is reduced, however, the level of uncertainty
will increase and this will place downward pressure on stock values (Gordon 1962;
Lintner 1962; Brigham, Gapenski and Ehrhardt 1999; Gitman, Juchau and Flanagan
2002).
183
Chapter 5: Analysis of Data and Findings
Dividend Irrelevance Argument
Academics and experienced practitioners perceive the relationship between dividends
and share prices in a variety of ways (Petty et al. 2000). Practitioners believe changes in
share prices result from announcements of dividends (Petty et al. 2000). However, the
dividend irrelevance argument has been suggested by Miller and Modigliani (1961).
These researchers claim, in a perfect world, the value of a firm is unaffected by the
distribution of dividends and is determined by the earnings ability of the firm and the
risk of holding assets (Miller and Modigliani 1961; Gitman, Juchau and Flanagan
2002). This argument suggests that firms can transfer value, and preserve the market
value of shares with a variety of combinations of assets, earnings and investment
opportunities (Brealey and Myers 2000). If the firm pays dividends, each share will be
worth less because new shares must be issued to finance the cash outflow. Investors can
obtain cash by selling their shares in the market (if the market is liquid or efficient)
(Miller and Modigliani 1961; Brealey and Myers 2000). Therefore, firms need not
worry about their dividend policy (Brealey and Myers 2000). Figure 3.6 shows the
relationship between new stockholder, existing stockholders and the firm’s trade-off
between cash and shares.
Figure 3.6: Two Ways of Raising Cash for Firm
Dividends Financed
by Stock Issue
Shares
New Stockholders
Cash
Firm
No Dividend, No
Stock Issue
New Stockholders
Cash
Shares
Cash
Old Stockholders
Old Stockholders
Source: Brealey and Myers (2000), p. 445.
184
Chapter 5: Analysis of Data and Findings
Dividend irrelevance theory has two preconditions or assumptions. Firstly, Miller and
Modigliani assume that investment and borrowing decisions have been undertaken and
will not affect the value of the dividend payment (Miller and Modigliani 1961; Petty et
al. 2000). Secondly, the scenario described by these researchers assumes that (1)
investors can buy and sell stocks without incurring transaction, or brokerage costs and,
(2) firms can issue shares without issuing costs and, (3) firms and investors are not
taxed and, (4) all information about the firm is available to the public and, (5) there is
no conflict of interest between management and stockholders and, (6) borrowing and
lending are unlimited and, (7) there is no problem with liquidity (Petty et al. 2000).
Given these assumptions, Miller and Modigliani suggested that there is no relationship
between a firm’s dividend policy and the value of its shares (Miller and Modigliani
1961; Petty et al. 2000).
An absence of taxes and brokerage costs (Brigham, Gapenski and Ehrhardt 1999) and
the assumptions made by Miller and Modigliani indicate that the dividend irrelevance
argument theory can be contested (Brigham, Gapenski and Ehrhardt 1999). Studies
have indicated that changes to dividends affect share prices, or are positively correlated
with share prices (James, Dodd and Kimpton 1985; Rappaport 1986; March and Merton
1987; Penman and Sougiannis 1998; Hurley and Johnson 1998; Yao 1997; Fama and
French 1999, 2000, 2001; James, Fama and French 2000, Wetherilt and Weeken 2002).
A higher payment of dividends is a positive signal for investors to buy the shares, while
a lower payment is a negative signal and suggests that the investor should sell the
shares (Gitman, Juchau and Flanagan 2002).
As discussed earlier, an increase in dividend payments is often followed by an increase
in the price of a firm’s stocks and this may indicate that investors prefer dividends to
capital gains (Brigham, Gapenski and Ehrhardt 1999). However, Miller and Modigliani
(1961) argued that an increase, or decrease in the payment of dividends is merely a
signal that a firm’s managers have forecast a positive, or negative outlook for profits. It
is suggested that the evidence termed ‘signaling of dividends’ is unclear (Brigham,
Gapenski and Ehrhardt 1999).
185
Chapter 5: Analysis of Data and Findings
The evidence on whether dividends are relevant, or irrelevant is mixed. It appears,
however, that investors possess a variety of views on the importance of dividends.
Some investors prefer high dividends, while others are satisfied with low dividends
(Brigham, Gapenski and Ehrhardt 1999). In addition, some investors prefer capital
gains rather than dividends. Overall it appears that investors prefer firms with a stable
and predictable dividend policy (Brigham, Gapenski and Ehrhardt 1999).
3.4.4 DIVIDEND POLICY AND SHARE PRICE MOVEMENT
As mentioned earlier, a firm’s dividend policy is likely to affect the value of its shares.
There are three views of the relationship between dividend policy and share prices: (1)
dividend policy is irrelevant; (2) high dividends increase share values and; (3) low
dividends increase share values (Brigham, Gapenski and Ehrhardt 1999; Gitman 2000).
(1)
Dividend policy is irrelevant: As discussed earlier, several analysts believe that
a firm’s dividend policy has no effect on its share price (Miller and Modigliani 1961;
Petty et al. 2000).
(2)
High dividends increase share values: Dividends are more predictable than
capital gains and managers can control dividends, although they cannot directly
influence the price of shares. Shareholders believe the risks associated with capital
gains are higher than those with dividends (Brigham, Gapenski and Ehrhardt 1999;
Brealey and Myer 2000; Gitman 2000). The value of dividends are therefore more
predictable than the size of capital gains. The ‘bird-in-the-hand’ theory is therefore
accepted (Linter 1956). Investors value a unit of expected dividends more highly than a
unit of expected capital gains (Brigham, Gapenski and Ehrhardt 1999; Gitman 2000).
Consequently, high dividends increase the value of shares (Brigham, Gapenski and
Ehrhardt 1999; Gitman 2000; Gitman, Juchau and Flanagan 2002).
(3)
Low dividends increase share values: Income in the forms of dividends and
capital gains is taxable. However, tax on dividend income is paid when the dividends
are received, while tax on capital gains can be deferred until the stock is sold (this is
186
Chapter 5: Analysis of Data and Findings
termed deferring the payment of taxes). Investors attempt to defer tax payments for as
long as possible (Petty et al. 2000). Therefore, if taxation is considered, investors prefer
capital gains rather than the payment of dividends (Brigham, Gapenski and Ehrhardt
1999; Gitman 2000; Gitman, Juchau and Flanagan 2002).
Corporate dividend policies have been the subject of intensive theoretical modelling
and empirical examination. There are many views on dividend payments. Large
payments can be harmful to investors and firms, although, some firms are willing to
pay high dividends (Petty et al. 2000). This phenomenon is termed the ‘dividend
puzzle’ and many scholars and practitioners has provided evidence to explain this
phenomenon (Petty et al. 2000). However, the conclusions of theoretical models are
often conflicting and lack empirical support when attempting to explain the dividend
policies of firms.
Policy mainly considers the likely effect of dividend payments on share prices. As
discussed earlier, there are two schools of thought and their conclusions are in conflict.
The relevance theories of Gordon (1962, 1963) and Lintner (1962) suggest that
dividends have positive and negative influence on stock prices. As stated earlier,
irrelevance theory states that a firm’s dividend policy does not affect the value of its
stocks. Recent models focus on corporate dividend policy and taxation and their impact
on the behaviour of investors. Some of these models have been formulated using full
information, information asymmetries and behavioral principles. Several recent models
are discussed below.
A. Models of Full Information and Tax Factor
Investors expect higher returns in the form of dividends due to increasing tax liabilities
(Frankfurter and Gong 1992, 1993). The shareholder’s pre-tax returns should therefore
be increased. Miller and Modigliani (1961) separated investors into a variety of
dividend tax clienteles. More recent research by these scholars (Miller 1977;
Modigliani 1982) found that the ‘clientele effect’ has a minor influence on the
composition of an investor’s portfolio. Masulis and Trueman (1988) found that
187
Chapter 5: Analysis of Data and Findings
investors have a variety of views on a firm’s investment and dividend policies.
Assuming that investors aim to maximize their after-tax income, Farrar and Selwyn
(1967) developed a model which suggests that dividends need not be paid, although
share repurchases should be used for distributing corporate earnings. Farrar and Selwyn
(1967) assumed that investors can decide whether they want to receive their returns in
the form of cash dividends or capital gains. These findings were also supported by
Brennan (1970) and Auerbach (1979a).
It has been suggested that investors will buy stocks which pay dividends and invest in
tax free securities if they are likely to be highly taxed (Miller and Scholes 1978;
Auerbach 1979b; Miller 1986). The returns from tax-free securities are preferred to
dividends which are highly taxed. However, this theory was rejected by DeAngelo and
Masulis (1980) and Fung and Theobald (1984). These researchers concluded that the
model is insufficient, incomplete and does not consider interest and liquidity.
B. Models of Information Asymmetries
B1. Signaling Models
A firm’s dividend policy is often used as a less costly way of convincing shareholders
that the firm is being well managed. Many researchers support this finding, including,
Laub (1976), Bar-Yosef and Kolodny (1976, 1978), Bhattacharya (1979, 1980), Talmor
(1981), Hakansson (1982), John and Williams (1985), Asquith and Mullins (1983),
Miller and Rock (1985), Bar-Yosef and Huffman (1986), Makhija and Thompson
(1986), Ambarish, John and Williams (1987), Ofer and Thakor (1987), Kumar (1988),
Kale and Noe (1990) and Rodriguez (1992).
B2. Agency Cost
The problems of agency costs are not new. Adam Smith (1937) and (Kindleberger
1984) found that it was not possible for shareholders to monitor a firm’s managers
because of the inefficiencies and costs involved. Agency problems result if a firm’s
188
Chapter 5: Analysis of Data and Findings
managers engage in high-risk-high-return projects and overlook projects with a positive
net present value. Easterbrook (1984) suggested that a large dividend payout could
reduce agency costs by forcing managers to seek funds from capital markets.
B3. The Free Cash Flow Hypothesis
The problem of agency costs was also considered by Berle and Means (1932). These
researchers concluded that the funds which remain after a firm has invested in
profitable areas were used inefficiently and raised the level of conflict between
managers and shareholders. Jensen (1986) developed a free cash flow hypothesis which
states that dividends and debt interest payments decrease the volume of free cash flows
which are available to managers. In short, dividends and interest repayments help
prevent managers from engaging in non-productive investments.
C. Behavioral Models
Several recent models have focused on the behavior of investors because it may
influence the dividend policies of firms. Shiller (1984) noted that the behaviour of
investors is influenced by societal norms and attitudes. The behavioral models initially
failed to gain acceptance but could be used to explain some issues in the area of
dividend policy. Shiller (1984) also mentioned that social pressures can lead to poor
judgment by investors or shareholders.
C1. Managerial Surveys
Lintner (1956) interviewed a number of corporate chief executive officers and chief
financial officers and found they are reluctant to reduce dividends, or adopt a policy of
stable dividends because this may provoke a negative response from investors. Darling
(1957), Turnovsky (1967), Fama and Babiak (1968), Baker, Farrelly and Edelman
(1985), Baker and Farrelly (1988), DeAngelo, DeAngelo and Skinner (1992) supported
Linter’s findings. DeAngelo, DeAngelo and Skinner (1992) also suggested that factors
such as regulatory constraints, investment magnitude, debt and firm size also affect a
189
Chapter 5: Analysis of Data and Findings
firm’s dividend policy. Managers consider current and expected earnings, a firm’s
dividend payment history, dividend stability, cash flows and investment opportunities
and the view of shareholders when determining the level of dividends (Harkins and
Walsh 1971). They also consider issues such as sustainability, current profitability,
expectations of future cash flows, and industry norms when determining their payout
(Frankfurter and Lane 1992).
C2. Theoretical Behavioral Models
Shefrin and Statman (1984) explained dividend preferences with the theory of selfcontrol and the descriptive theory of choice under uncertainty (Kahneman and Tversky,
1982). The theory of finance says that dividends and capital gains are perfect
substitutes. Thaler and Shefrin (1981) argued that dividends and capital gains are not
always perfect substitutes. The costs and payoffs of high risk investments are evaluated
separately because losses are more significant than gains (Thaler and Shefrin 1981).
Individuals have saving and investment goals, therefore, they will not sell parts of their
portfolio if they expect to receive dividends from the shares (Frankfurter and Wood
2001).
Variation in the results obtained by the above models is explained by differences in
modeling, the method of analysis, the variables and definitions used, the type of data
and the sample period (Watts 1973, 1976a, 1976b; Roll 1977, 1983; Miller and Scholes
1982; Frankfurter and Gong 1992, 1993; Watts 1973; Morgan 1980, 1982; Baker and
Farrelly 1988).
Black (1976) and Feldstein and Green (1983) concluded that dividend policy remains a
puzzle because the existing models lack empirical support. However, some general
conclusions can be drawn about dividend policies. Managers decrease the level of
dividend payments only when necessary (Myers, 1984; and DeAngelo, DeAngelo, and
Skinner, 1992). Shareholders must pay taxes on dividend income and research supports
the hypothesis that returns on dividend-paying stocks are increased to offset the tax
liability (Frankfurter and Gong 1992, 1993).
190
Chapter 5: Analysis of Data and Findings
Managers are reluctant to reduce dividend payments if a firm is suffering from financial
distress. Likewise, dividends are only increased if managers are confident that high
levels of profit can be sustained. Managers are also aware that shareholders expect
dividends to be paid and feel they deserve this payment (Myers 1984).
3.4.5
OTHER FORMS OF DIVIDENDS
Share dividends or bonus shares, share splits, share buybacks and share repurchases are
alternative forms of dividends which are paid to shareholders. Firms pay stock
dividends, or buy back shares as a replacement for, or a supplement to cash dividends
(Gitman, Juchau and Flanagan 2002).
Share Dividends
Share dividends, or a bonus share issue is a method of paying dividends by giving
shares to the existing shareholders. Share dividends are normally issued as a proportion
of current holdings, such as one bonus share for ten shares owned (Gitman 2000; Petty
et al 2000; Gitman, Juchau and Flanagan 2002).
Share Dividends: the Shareholder’s Viewpoint
A share dividend signals potential growth and higher confidence in a firm’s
current operations. This is because a firm needs to be confident that it can
generate enough earnings to pay dividends on the expanded, or additional shares
(Gitman 2000; Petty et al. 2000; Gitman, Juchau and Flanagan 2002).
Share Dividends: the Company’s Viewpoint
Share dividends are more costly than cash dividends because the firm needs to
pay dividends on the additional shares (Gitman, Juchau and Flanagan 2002).
However, share dividends are a good option if the firm is growing fast and needs
internal financing (Gitman, Juchau and Flanagan 2002). Share dividends are also
a way of giving shareholders a bonus without paying cash (Gitman, Juchau and
Flanagan 2002). Share dividends are appreciated by the shareholders as long as
191
Chapter 5: Analysis of Data and Findings
the internal funds are used to enhance the firm’s growth and not for paying
overdue bills (Gitman 2000; Gitman, Juchau and Flanagan 2002).
Share Buybacks or Share Repurchase
Share buybacks are also a form of dividend payment because excess cash, or
retained earnings are used by the firm to purchase its own shares. Share buybacks
increase shareholders value in three ways: (1) by reducing the number of shares
outstanding, and thereby raising Earnings Per Share (EPS); (2) giving a positive
signal that management believes the shares are undervalued and; (3) by
preventing an unfriendly takeover (Gitman 2000; Gitman, Juchau and Flanagan
2002).
This study will focus on cash dividends because they constitute actual cash which is
paid to shareholders. Cash dividends can significantly affect a firm’s financing because
they affect the volume of retained earnings (internal funds available) and the funds
which are available for investment (James, Dodd and Kimpton 1985; Melicher and
Norton 2000; Gitman, Juchau and Flanagan 2002).
3.4.6
FACTORS AFFECTING DIVIDEND POLICY
Legal Constraints
Firms only pay dividends when there are current, or accumulated past profits which are
available for distribution to shareholders. Firms are prevented from paying cash
dividends from its legal, or registered capital. The capital is required to register a firm
and must not be invested (James, Dodd and Kimpton 1985; Gitman, Juchau and
Flanagan 2002).
Internal Constraints
Profit is not cash, therefore firms with a high level of profitability may not possess a
high ability to pay dividends. Dividend payments are a drain on a firm’s liquid assets
(cash). The level of dividend payments, therefore depends on the internal constraints of
192
Chapter 5: Analysis of Data and Findings
a firm such as the liquidity of its assets (James, Dodd and Kimpton 1985; Grinblatt and
Titman 1998; Gitman, Juchau and Flanagan 2002).
Contractual Constraints
Firms with restrictive provisions in loan agreements may have difficulty paying
dividends. Generally, loan agreements prohibit the payment of cash dividends until a
certain level of earnings has been achieved. These provisions protect a firm’s creditors
from any losses in the case of insolvency (James, Dodd and Kimpton 1985; Grinblatt
and Titman 1998; Gitman, Juchau and Flanagan 2002).
Growth Prospects
Dividend policies may vary according to the stages of a firm’s growth. Growing firms
may need cash for additional investment projects. In addition, low growth firms may
also need cash for the replacement of existing assets. Therefore, a firm’s potential
growth can affect its dividend policies (James, Dodd and Kimpton 1985; Grinblatt and
Titman 1998; Gitman, Juchau and Flanagan 2002).
Owner Considerations
The goal of the firm is to maximize the shareholders’ wealth. Consequently, the views
of shareholders also influence a firm’s dividend policy. Issues which will affect the
views of shareholders include tax status, investment opportunities and the dilution of
ownership (James, Dodd and Kimpton 1985; Grinblatt and Titman 1998; Gitman,
Juchau and Flanagan 2002).
Market Considerations
Lintner (1956) suggested that the market’s response to a firm’s dividend policy is also
important for its success and is helpful in developing future dividend policies. As stated
earlier, shareholders often view the firm’s dividend payments as an indicator of its
future success (Grinblatt and Titman 1998; Gitman, Juchau and Flanagan 2002).
193
Chapter 5: Analysis of Data and Findings
3.4.7 TYPES OF DIVIDEND POLICIES
Gitman, Juchau and Flanagan (2002, p. 452) defined dividend policy as ‘the firm’s plan
of action to be followed whenever a decision concerning dividends must be made’.
Three dividend policies which are commonly used are described below.
Constant-Payout-Ratio Dividend Policy
The constant payout ratio is the most commonly used dividend policy and is calculated
by dividing the firm’s cash dividend per share by its earnings per share (EPS). The
constant payout ratio is used to ensure a specific percentage of earnings is paid to
owners in each dividend period (Grinblatt and Titman 1998; Gitman, Juchau and
Flanagan 2002).
Stable Dollar Dividend Per Share
This policy aims to maintain the value of dividend payments over time. A firm’s
management must be convinced that a specific dividend level can be maintained before
adopting such a policy (Brigham, Gapenski and Ehrhardt 1999).
Low-Regular-and-Extra Dividend Policy
Some firms establish a low-regular-and-extra dividend policy, which involves
distributing a low dividend payment on a regular basis and supplementing this with a
bonus when earnings are high (Gitman, Juchau and Flanagan 2002).
3.4.8 FIRM’S DIVIDEND PAYMENT DECISION
Lintner’s Model
Lintner (1956) conducted a series of interviews on the dividend policies of firms with
managers of a variety of firms. He found that dividends are determined by four stylised
facts:
x
‘Firms have long-run target dividend payout ratios. Mature firms have a high
payout ratio while growth firms have a low payout ratio’
194
Chapter 5: Analysis of Data and Findings
x
‘Managers focus more on dividend changes than on the absolute level’
x
‘Dividend changes follow shifts in long-run, sustainable earnings. Managers try
to smooth dividends’
x
‘Managers are reluctant to make dividend changes’ (Brealey and Myers 2000, p.
443).
The findings of Lintner (1956) were published in DeAngelo, DeAngelo and Skinner
(1990). These findings were as follows: In times of financial difficulties, managers are
more likely to reduce dividend payments than cease paying them. In addition, managers
of firms with long dividend histories tend to avoid cuts, or omissions, to preserve the
favourable views of the firms which are held by investors (DeAngelo and DeAngelo
1990).
From these findings, Lintner developed a model for explaining the payment of
dividends:
DIV 1
=
Target Dividends
=
Target Ratio x EPS 1
=
Target Ratio x EPS 1–DIV 0
The dividend change would equal
DIV 1 – DIV 0
=
Target Change
When a firm decides to alter its target payout ratio (according to Lintner’s survey they
are reluctant to do this) the changes are formulated are follows:
DIV 1 – DIV 0
=
Adjustment Rate x Target Change
=
Adjustment Rate x (Target Ratio x EPS1 – DIV0)
Fama and Babiak (1968) suggested the findings of Lintner (1956), however, this study
found that the model could not explain all of the reasons that firms pay dividends.
Consequently, there is no universal approach that explains dividend policy. Managers
195
Chapter 5: Analysis of Data and Findings
may wish to maintain a long history of continuous dividend payments, strategic
considerations, and binding debt covenants (DeAngelo, DeAngelo and Skinner 1990).
3.5
THE PHENOMENON OF DISAPPEARING DIVIDENDS IN CAPITAL
MARKETS
3.5.1 THE GLOBAL PHENOMENON OF DISAPPEARING DIVIDENDS
In the US, dividends are generally taxed at a higher rate than capital gains. Therefore,
firms presume that dividends are less valuable than capital gains (Fama and French
2001). Consequently, firms that pay dividends are disadvantaged because they
experience higher costs of equity than firms that do not pay dividends, or finance their
activities with equity (Fama and French 2001). Even though dividends are a high cost
method of obtaining finance, some firms continue to pay dividends. Some firms claim
that they need to compensate and be honest to their loyal stockholders (Petty et al.
2000).
Fama and French (2001) investigated the issue of disappearing dividends in the US.
This study focused on the period from 1978 to 1999 and found that the percentage of
firms paying dividends in the US declined sharply after 1978. In 1973, 52.8 percent of
listed firms paid dividends and this figure rose to 66.5 percent in 1978. However, by
1999 only 20.8 percent of US listed firms paid dividends. Therefore, the percentage of
firms paying dividends decreased by more than 50 percent between 1973 and 1999.
Fama and French (2001) provided two explanations for this outcome. Firstly, the
characteristics of newly listed firms changed alter 1978. At present, many listed firms
experience characteristics such as low profitability, strong growth and have never made
dividend payments (Fama and French 2001). Secondly, Fama and French (2001)
studied the characteristics of listed firms over time and found that less firms appear to
pay dividends than in the past. Fama and French (2001) entitled this trend; ‘the
declining propensity to pay dividends’.
Fama and French (2001) used summary statistics (or descriptive statistics) such as,
frequencies and means, to categorise the firms listed (excluding financials and utilities
196
Chapter 5: Analysis of Data and Findings
firms) on three US stock markets (AMEX, NASDAQ and NYSE). Three characteristics
were found to affect the level of dividend payments, namely profitability, investment
opportunities and a firm’s size. Logit regression analysis was used to confirm the
significance of these factors.
Earlier researchers such as Tobin (1958), Gordon (1962), (1963), Higgin (1972)
Fazzari, Hubbard, and Petersen (1988), Fama and French (1997) and Aivazian, Booth
and Cleary (2003) focused on the relationship between internally generated funds,
profitability and investment. Ratios were used to assess the level of profitability,
investment, growth, size or the liquidity of firms to standardise the measurement of
these variables. Ratios which are often used by researchers include the debt ratio (total
liabilities divided by total assets), the M/B ratio (market to book value), ROA (net
income divided by total assets), ROE (net income divided by shareholders’ equity),
total assets, and the growth in assets (Fazzari, Hubbard, and Petersen 1988; DeAngelo
and DeAngelo 1990; Fama and French 1995, 1997; James, Fama and French 2000,
Aivazian, Booth and Cleary 2003).
In relation to a firm’s profitability, Fama and French (2001) used Et/At (the ratio of
aggregate earnings before interest to aggregate assets) and Yt/BEt (aggregate common
stock earnings over aggregate book equity). After calculating the values and the
average for the two ratios, the average values of ratios for payers of dividends, nonpayers and former payers were calculated using statistical software (Fama and French
2001). Fama and French (2001) found that the profitability of the firms (as measured by
Et/At and Yt/BEt ratios) varies with the type of firm (dividends payers, former payers
and non-payers). Firms that distribute dividends were found to have higher profitability
than non-payers (Fama and French 2001).
As mentioned earlier, the level of investment opportunities was seen as a factor which
influences the dividend policies of firms which are listed in the US. Fama and French
(2001) claimed that investment opportunities vary across dividend groups (payers, nonpayers: former-payers and never paid firms).
197
Chapter 5: Analysis of Data and Findings
In this way, the investment opportunities of firms listed on the Thai stock market will
be measured by using Vt/At (assets’ growth rates which is the ratio of the aggregate
market value to the aggregate book value of assets), dAt/At (the rate of change of total
assets, or the firms’ assets growth rate) and RDt/At (research and development or
R&D) expenditures following Fama and French’s research paper on ‘disappearing
dividends’. Summary statistics (descriptive statistics) and logit regression will be
used to analyse the significance of the investment opportunities of firms which are
listed in Thailand.
Fama and French (2001) concluded that firms which have not paid dividends also have
better growth opportunities than payers and former payers. Firms that have never paid
dividends also grow faster than those which have not paid dividends. Former payers of
dividends have poor investment opportunities, because their internal source of finance
has been depleted by the payment of dividends (Fama and French 2001).
The average assets size (At) and debt ratios (total liabilities to total assets) of firms
listed on the US stock markets has influenced firm size (Fama and French 2001).
Summary statistics were used to calculate the values of the average assets size ratio and
logit regression was applied to confirm this characteristic. Fama and French (2001)
found that dividends payers are much larger than former payers and non-payers, in the
case of US listed firms.
The propensity to pay dividends is measured by applying the average coefficient to the
value of each firm’s characteristic to form a formula on profitability, investment
opportunities and size (Fama and French 2001). A base period (or years) was set and
the ratios a variety of groups of firms were compared in relation to their propensity to
pay dividends in different time periods. This process of analysis and comparison was
undertaken with the aid of logit regression (Fama and French 2001). Fama and French
(2001) stated that listed firms have a lower propensity to pay dividends and this results
in lower payments and decline in the number of payers.
198
Chapter 5: Analysis of Data and Findings
To measure change in the characteristics of listed firms; the sample was divided into
four groups: (1) dividends payers, (2) non-payers (3) former-payers, and (4) never paid
(Fama and French 2001). The three important characteristics of listed firms, namely:
(1) profitability, (2) investment opportunities, and (3) size, were measured and a logit
regression analysis was used to test for any change in firm characteristics, or
divergence between the expected number of payers and the actual number of payers
(Fama and French 2001). Fama and French (2001) found that listed firms have changed
their characteristics over time and this has resulted in lower dividend payment in the
US stock markets.
Fama and French (2001) concluded that firms which have never paid dividends are
more profitable and experience stronger growth opportunities than former payers.
Consequently, it can be argued that low earnings, strong investment and growth
opportunities, and small size are characteristics of firms which have never paid
dividends (Fama and French 2001). In addition, former payers tend to be distressed
firms (Fama and French 2001). Finally, payers of dividends are large firms with high
earnings and low investment opportunities (Fama and French 2001). Fama and French
(2001) concluded that the characteristics and their propensity to pay dividends of listed
firms have changed. The change has resulted in a disappearance of dividends and a
reduction in the number of payers (Fama and French 2001).
Types of Listed Firms
Analysts have categorized listed firms into four groups according to their earnings.
These groups include: (1) growth, (2) defensive, (3) cyclical, and (4) speculative firms
(Reilly and Brown 2000).
Growth Firms
Growth firms generally experience ‘above-average growth in their sales and earnings’
(Reilly and Brown 2000). However, academics define this category as firms which
yield rates of return which are greater than the firm’s required rate of return or weighted
average cost of capital (WACC) (Reilly and Brown 2000). The sales and earnings of
199
Chapter 5: Analysis of Data and Findings
growth firms expand faster than firms with different trends in the value of their
earnings, although the risk associated with investing in these firms are similar. (Reilly
and Brown 2000).
Defensive Firms
The earnings of defensive firms are likely to be maintained during a downturn (Reilly
and Brown 2000). Defensive firms are also considered low risk investments. Public
utilities are an example of a defensive firm (Reilly and Brown 2000).
Cyclical Firms
The sales and earnings of cyclical firms are strongly affected by fluctuations in the
business cycle or life-cycle (Reilly and Brown 2000). This type of firm is regarded as a
high risk investment.
Speculative Firms
Speculative firms have high risk assets (Reilly and Brown 2000). However, this type of
firm may yield high returns for investors and compensate them for holding these assets
(Reilly and Brown 2000).
Analysts and investors have classified the listed stocks of firms into two groups (1)
value and (2) growth stocks. This classification is based on how the stocks generate
income, or earnings for investors (Reilly and Brown 2000).
Value Stocks appear to be undervalued for reasons other than their earnings, or growth
potential (Reilly and Brown 2000). Value stocks normally pay dividends and have low
P/E ratios (Petty et al. 2000; Reilly and Brown 2000).
Growth Stocks often experience rapid growth in their sales and earnings (examples
include Microsoft and Intel) (Reilly and Brown 2000). Investors in this type of stock
expect the price to appreciate (Petty et al. 2000). Firms with a high book-to-market
equity ratio (B/M), earnings to price ratio (E/P), or cash flow to price ratio (C/P) are
200
Chapter 5: Analysis of Data and Findings
classified as growth stocks (Lakonishok, Shleifer, and Vishny 1994; Fama and French
1995, 1998).
A paper by Fama and French (1998) suggested that value stocks tend to outperform
growth stocks in the capital markets around the world. This analysis also found that the
markets undervalue the stocks of distressed firms and overvalue successful firms
(growth stocks) (Fama and French 1998). These findings were consistent with the
trends in 13 developed capital markets including the US, UK, France, Germany, Italy,
the Netherlands, Belgium, Switzerland, Sweden, Australia, Hong Kong and Singapore.
Fama and French (2001) concluded that the characteristics of firms which pay
dividends tend to be similar to the characteristics of value stocks. However, if these
characteristics were held constant, the study found that the characteristics of listed firms
tended to change over time. Indeed, firm characteristics initially resembled those of
value stocks but later resembled growth stocks. Some firms also ceased paying
dividends. It was therefore concluded that the payment of dividends is becoming less
common in the US stock markets because the characteristics of listed firms are tending
to resemble those of growth stocks.
The relative performance of value stocks and growth stocks appears mixed throughout
the world’s capital markets. Sometimes, value stocks outperform growth stocks, while
sometimes, growth stocks tend to outperform value stocks (Reilly and Brown 2000).
There is no universal view on whether growth stocks perform better, or yield higher
profits than value stocks. Fama and French (2001)’s study change in the characteristics
of listed firms is a useful guide for investors around the world. Investors should be
aware that the characteristics of firms are changing and this will aid in improving the
quality of their investment decisions when choosing between value and growth stocks.
A study by Benito and Young (2001) supported the findings of Fama and French
(2001) in relation to the decline in the payment of dividends. This study used secondary
data from the UK stock market (FTSE) which was collected during the period 1974 to
1999. It was found that the proportion of firms which do not pay dividends rose from
201
Chapter 5: Analysis of Data and Findings
14.3 percent in 1974 to 25.2 percent in 1999 (Benito and Young 2001). The study also
found that the proportion of firms that were not paying dividends tended to increase
during recessions.
Benito and Young (2001) suggested that the increase in the proportion of firms omitting
dividends was largely due to an increase in the proportion of firms that have never paid
dividends. In addition, firms which paid high dividends more than doubled their
volume of payments to shareholders between 1977 and 1999 (Benito and Young 2001).
The study also found that low cash flows, high leverage and greater investment
opportunities are associated with an increase in the propensity to omit dividends in the
UK (Benito and Young 2001). This finding is consistent with the conclusions of Fama
and French (2001).
Benito and Young (2001) suggested that non-payers tend to be small firms with strong
investment opportunities. This indicates that firms which do not pay dividends have
positive prospects for growth and are not distressed. The decisions to avoid the
payment of dividends was regarded as evidence of financial distress. In short, Benito
and Young (2001) concluded that the characteristics of firms and the propensity to pay
dividends have a large impact on the payment of dividends in the UK (Benito and
Young 2001).
The two previous studies concentrated on the payment of dividends by firms listed on
stock markets of developed countries. Aivazian, Booth and Cleary (2003) investigated
eight developing countries including South Korea, India, Malaysia, Thailand,
Zimbabwe, Jordan, Pakistan and Turkey. This study used secondary data collected
between 1981 and 1990, and used descriptive statistics and t-statistics to compare these
emerging markets with the US. The findings indicate that firm profitability, growth and
investment opportunities affect the payment of dividends of emerging markets in a
similar fashion to the US (Aivazian, Booth and Cleary 2003). They added that a high
ROE (which represent a firm’s profitability) results in the payment of high dividends,
while high debt ratios are associated with a low payment of dividends. However the
‘sensitivity to these factors varies across countries and emerging market firms tend to
202
Chapter 5: Analysis of Data and Findings
pay higher dividends than the US firms and are more sensitive to some of the variables’
(Aivazian, Booth and Cleary 2003, p. 387). These researchers, therefore, concluded
that firms in emerging markets respond to the same variables as those in the US
(Aivazian, Booth and Cleary 2003). However, the study did not shed light on why the
payment of dividends is in decline (Aivazian, Booth and Cleary 2003).
Many researchers have attempted to explain the phenomenon of disappearing
dividends. Yardeni (2003) stated that the payment of dividends was important in the US
prior to 1982, but has become less important over the last 20 years (Yardeni 2003). The
researcher claimed that ‘if the US Congress eliminates the double taxation of dividends,
then dividends should matter more again’ (Yardeni 2003 p. 14).
Mr. Allan Greenspan (present Chairman of the Federal Reserve Bank, United States)
observed that shareholders’ obsession with earnings is a relatively new phenomenon:
‘Prior to the past several decades, earnings forecasts were not nearly so important a factor
in assessing the value of corporations. In fact, I do not recall price-to-earnings ratios as a
prominent statistic in the 1950s. Instead, investors tended to value stocks on the basis of
their dividend yields’ (Yardeni 2003, p. 15).
The phenomenon of disappearing dividends remains a puzzle (Black 1976; Fama and
French 2001, Benito and Young 2001; Aivazian, Booth and Cleary 2003). Fama and
French’s (2001) findings have added to the knowledge of dividends although more
research is needed to explain the pattern of dividends, payers’ characteristics and the
propensity to pay dividends in developing capital markets such as Thailand. This leads
to an investigation of the relationship between dividends and a given set of
characteristics of listed firms and other important market factors which may explain
why firms pay, or avoid paying dividends. As stated earlier, the analysis of Fama and
French (2001) focused on firms which are listed in the US and decline in the number of
payers and the propensity to pay dividends was evident. However, more analysis is
needed on whether this is occurring in foreign markets, (Ferris, Sen and Yui 2002)
particularly when the characteristics of firms and investors differ from those of the US.
The issue of whether decline in the payment of dividends is an ‘international
203
Chapter 5: Analysis of Data and Findings
phenomenon’ needs more research (Ferris, Sen and Yui 2002). This study will respond
to this need to this topic by focusing on the emerging capital markets such as Thailand.
3.5.2 THE DISAPPEARANCE OF DIVIDENDS IN THAILAND
The following trends were found after a preliminary investigation of dividends in
Thailand.
Figure 3.7: Number of Listed Firms Which Paid Dividends
Number of Payers and Non-Payers from 1990-2002
All firms
Payers
450
Number of samples
400
350
300
250
200
150
100
50
0
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002
Year
Source: SET (2002a, 2002b)
204
Chapter 5: Analysis of Data and Findings
Figure 3.8: Percentage of Dividend Payers to Total Listed Firms
Dividend trend in Thai capital market
Payers
Percent
Non-Payer
100.0
90.0
80.0
70.0
60.0
50.0
40.0
30.0
20.0
10.0
0.0
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002
Year
Source: SET (2002a, 2002b)
Figures 3.7 and 3.8 indicate that the phenomenon of ‘disappearing dividends is also
evident in Thailand although this market has not been studied in detail.
Both Thai and international investors in the capital market of Thailand lack confidence
in (1) the type of stocks to invest in and (2) the trading strategies and valuation
techniques which should be applied. This investigation of the phenomenon will
enhance the level of confidence in stock investments in Thailand by providing
information for investors on the type of stocks to invest, appropriate trading strategies
and the techniques which should be applied.
A common technique for valuing stocks is the Dividend Discount Model (DDM or
Gordon Model). Stock prices are calculated by discounting the expected cash dividend
payments back to the present time. Therefore, the most important component of DDM
is the level of the cash dividends during each period. However, if it is shown that the
payment of cash dividends is disappearing, DDM may not be as effective as it has been
in the past. Investors may need to apply, or develop more complex models to value
stocks. Conversely, if it is shown that the payment of cash dividends is not
205
Chapter 5: Analysis of Data and Findings
disappearing, DDM will continue to be supported as an accurate technique for valuing
dividends by investors and analysts.
3.5.3
THAILAND IN THE LIGHT OF RESEARCH ON CAPITAL MARKET
The Stock Exchange of Thailand (2003) intends to devote more funds to projects for
developing quality analysis and for research into the capital markets. This funding will
increase the quantity and quality of research in the securities field, enhance the level of
analysis and aid in establishing standards in the industry (SET 2003).
SET (2003) is encouraging Thai investors to focus more on improving their knowledge
of the following: the structure of capital markets, securities trading regulations, roles of
securities analysis in the capital market of Thailand, accessing information, industry
and company analysis, financial ratio analysis, stock evaluation, how to provide and
gather information, how to customise report presentations according to the users’ needs,
critical research examinations, investment instruments, analysis ethics, and finally the
Thai capital market development plan.
Financial support will be granted via the ‘Securities Research Support Project’ to
securities companies securities research on the industrial sectors. There are currently 12
companies which prepare research reports on 70 of the existing 600 securities which are
listed in Thailand (SET 2003). This represents only 11% of the existing listed
securities. Currently, most securities companies provide research reports on highly
capitalized and liquid shares. Therefore, research on trends in share prices, the number
of dividend payers, characteristics of payers, listed firms and the propensity to pay
dividends will benefit investors, firms, the Stock Exchange of Thailand (SET) and the
Securities Exchange Commission (SEC).
3.6
RESEARCH GAPS, ISSUES AND MODEL DEVELOPMENT
This section focuses on the research gaps which were identified in this review of
literature and will develop the research problems and hypotheses for this study. The
206
Chapter 5: Analysis of Data and Findings
research issues, which are composed of research problems and hypotheses, are derived
from the research gaps which were identified.
3.6.1
RESEARCH GAPS AND ISSUES
3.6.1.1 RESEARCH GAPS
This review has discussed how the characteristics of firms, the propensity to pay
dividends and financial ratios of listed firms affect the payment of dividends and has
shed light on the phenomenon of disappearing dividends. Although several researchers
of capital markets have commented on the phenomenon and the level of omission by
listed firms in a variety of markets and countries, there are some limitations in these
studies which are listed below:
x
The phenomenon of disappearing dividends was investigated by Fama and
French (2001). Research on this topic is limited and requires additional analysis.
x
Most of the literature on this topic has focused on the capital markets of
developed countries such as the US and the UK. There is a lack of empirical
evidence from emerging capital markets, such as Thailand.
x
The relevant studies have used data from non-financial and non-utility
(industrial) firms. Financial institutions were excluded because due to variation
in accounting styles and the large level of assets and liabilities of these firms.
This study will assess the financial sector.
The findings of earlier studies may not apply to all listed companies. As stated above,
the samples which were analysed were limited.
The gaps in the literature which will be examined include:
x
The need for additional research on disappearing dividends and the omission of
dividend payments in developing countries.
x
This will be the first analysis of the phenomenon of disappearing dividends or in
Thailand.
207
Chapter 5: Analysis of Data and Findings
x
This study will be the first analysis of all sectors in the capital market of
Thailand.
3.6.1.2 RESEARCH ISSUES
The research issues depend on the research gaps which have been identified by the
review of literature and research problem. After identifying the research issues, a list of
research questions and hypotheses will be developed for analysing these issues.
The Research Problem
The research problem is restated as follows:
‘How do the characteristics of publicly traded firms and their propensity to pay
dividends influence the payment of dividends and the reasons for their
disappearance from the stock market of Thailand?’
This study focuses on the characteristics of listed firms and their propensity to pay
dividends, therefore, the information outlined on firms is summarised by a set of
financial variables which are termed firm characteristics. The characteristics which
were suggested by Fama and French (2001) have been adopted for this study. These
characteristics include: (1) profitability (2) investment opportunities and (3) size. The
methodology to measure change in these characteristics and their impact and the
propensity to pay dividends will be discussed in more detail in chapter 4.
Research Questions and Hypotheses
Several research questions and hypotheses have been developed to answer the research
problem. The research questions and hypotheses are presented in figure 3.9.
208
Chapter 5: Analysis of Data and Findings
Figure 3.9: Research Questions and Hypotheses
RESEARCH
PROBLEM
‘How do the
characteristics
of publicly
traded firms and
RESEARCH HYPOTHESES
RESEARCH QUESTIONS
Q1: What is the background and
performance of the Stock Exchange of
Thailand and its listed firms?
their propensity
Q2: How have researchers investigated
to pay dividends
dividends and their patterns throughout
influence the
payment of
dividends and
the reasons for
their
the world?
Q3:
9 Research Hypotheses
What
is
methodology
for
the
appropriate
collecting
and
analysing data?
disappearance
Q4: Does the data from Thailand support
from the stock
the model?
market of
Thailand?’
Q5:
How
could
the
model
be
implemented?
Source: Developed for this research
The four research questions are answered in this study. In addition, research question 4
(Q4) contains four sub-questions (Q4.1, Q4.2 and Q4.3) are stated as follows:
Research Questions #4.1:
What are the characteristics of the dividend payers?
As stated earlier, Fama and French (2001) suggested that the characteristics of listed
firms have changed over time and the number of firms paying dividends in the US
stock markets has fallen. This study will consider the phenomenon of disappearing
dividends in Thailand by responding to hypotheses 1 to 3 and the impact of changes in
the characteristics of listed firms.
209
Chapter 5: Analysis of Data and Findings
Fama and French (2001) used summary statistics to analyse the characteristics of firms
listed on the US stock markets. These firms were divided into four groups: (1) Dividend
payers, (2) Non-payers, (3) Former payers and (4) Never-paid firms. Three
characteristics were identified to explain the pattern of dividends and actions, namely:
(1) profitability, (2) size and (3) investment opportunities. Logit regression analysis was
employed to test the significance of these characteristics. This research will investigate
whether the characteristics of firms listed in the developing economy of Thailand are
similar to those listed on the US capital markets. These hypotheses are stated as
follows:
Research Hypothesis # 1:
Ho :
There is no significant relationship between the profitability of listed firms
and the probability that these firms will pay dividends.
Ha :
There is a significant relationship between the profitability of listed firms
and the probability that these listed firms will pay dividends.
Research Hypothesis # 2:
Ho :
There is no significant relationship between the size of listed firms and the
probability that these firms will pay dividends.
Ha :
There is a significant relationship between the size of listed firms and the
probability that these firms will pay dividends.
Research Hypothesis # 3:
Ho :
There is no significant relationship between the investment opportunities
of listed firms and the probability that these firms will pay dividends.
Ha :
There is a significant relationship between the investment opportunities of
listed firms and the probability that these firms will pay dividends.
SET (2002) found that almost 70 percent of listed firms made negative earnings and
only 20 percent of listed firms paid dividends during the Asian economic crisis. There
could be a relationship between the Asian economic crisis, the characteristics of listed
firms and the number of dividend payers in Thailand’s capital market. Hypotheses # 1
210
Chapter 5: Analysis of Data and Findings
to 3 will be tested during three time periods, namely Step A, all firms from 1990 to
1996 (pre-crisis), Step B, all firms from 1997 to 2002 (post-crisis), Step C, all firms
from 1990 to 2002 (total outlook).
Figure 3.10 shows the linkage between each of the 9 research hypotheses and the 5
research questions.
Research Questions #4.2:
Do these characteristics change over time?
As stated earlier, Fama and French (2001) suggested that the characteristics of listed
firms have changed over time and the number of dividend payers in the US stock
markets has fallen. This study will consider the phenomenon of disappearing dividends
in Thailand and the impact of changes in the characteristics of listed firms by
responding to hypotheses 1 to 3.
Research Question #4.3:
How has the propensity to pay dividends of firms changed over time?
As stated earlier, Fama and French (2001) also found that many firms are not paying
dividends although they have sufficient earnings. This outcome is known as a lower
propensity to pay dividends. The propensity to pay dividends will be addressed by
responding to hypotheses 1 to 3. This will lead to an answer to Research Question 4.3.
211
Chapter 5: Analysis of Data and Findings
Figure 3.10: Research Questions in Relation to Hypotheses
CHARACTERISTICS OF DIVIDEND PAYERS BEFORE THE ASIAN ECONOMIC CRISIS (Step A)
(1) PROFITABILITY
Ho
:
Ha
:
There is no significant relationship between the profitability of the firms and the probability
that the listed firms will pay dividends before the Asian Economic Crisis.
There is a significant relationship between the profitability of the firms and the probability
that the listed firms will pay dividends before the Asian Economic Crisis.
(2) SIZE
Ho
:
Ha
:
There is no significant relationship between the size of the firms and the probability that the
listed firms will pay dividends before the Asian Economic Crisis.
There is a significant relationship between the size of the firms and the probability that the
listed firms will pay dividends before the Asian Economic Crisis.
(3) INVESTMENT OPPORTUNITIES
Ho
:
Ha
:
There is no significant relationship between the investment opportunities of the firms and the
probability that the listed firms will pay dividends before the Asian Economic Crisis.
There is a significant relationship between the investment opportunities of the firms and the
probability that the listed firms will pay dividends before the Asian Economic Crisis.
CHARACTERISTICS OF THE DIVIDEND PAYERS DURING AND AFTER THE ASIAN ECONOMIC CRISIS (Step B)
(1) PROFITABILITY
Ho
:
Ha
:
There is no significant relationship between the profitability of the firms and the probability
that the listed firms will pay dividends after the Asian Economic Crisis.
There is a significant relationship between the profitability of the firms and the probability
that the listed firms will pay dividends after the Asian Economic Crisis.
(2) SIZE
Ho
:
Ha
:
There is no significant relationship between the size of the firms and the probability that the
listed firms will pay dividends after the Asian Economic Crisis.
There is a significant relationship between the size of the firms and the probability that the
listed firms will pay dividends after the Asian Economic Crisis.
(3) INVESTMENT OPPORTUNITIES
Ho
:
Ha
:
There is no significant relationship between the investment opportunities of the firms and the
probability that the listed firms will pay dividends after the Asian Economic Crisis.
There is a significant relationship between the investment opportunities of the firms and the
probability that the listed firms will pay dividends after the Asian Economic Crisis.
CHARACTERISTICS OF THE DIVIDEND PAYERS FOR THE WHOLE PERIOD (Step C)
(1) PROFITABILITY
Ho
:
Ha
:
There is no significant relationship between the profitability of the firms and the probability
that the listed firms will pay dividends for the whole period of investigation.
There is a significant relationship between the profitability of the firms and the probability
that the listed firms will pay dividends the whole period of investigation.
(2) SIZE
Ho
:
Ha
:
There is no significant relationship between the size of the firms and the probability that the
listed firms will pay dividends the whole period of investigation.
There is a significant relationship between the size of the firms and the probability that the
listed firms will pay dividends the whole period of investigation.
(3) INVESTMENT OPPORTUNITIES
Ho
:
Ha
:
There is no significant relationship between the investment opportunities of the firms and the
probability that the listed firms will pay dividends the whole period of investigation.
There is a significant relationship between the investment opportunities of the firms and the
probability that the listed firms will pay dividends the whole period of investigation.
Source: Developed for this research
212
Chapter 5: Analysis of Data and Findings
3.6.2
DEVELOPMENT OF THE RESEARCH MODEL
Figure 3.11 summarises the concepts which will be investigated and their linkage to the
payment of dividends in the Thailand’s capital market and the proposed methodology
for this study.
Figure 3.11: Model for this Research
Summary Statistics
Logit Regression
Number of Payers,
Non-Payers and
Former Payers
Cash Dividends
Characteristics
(Profitability,
Investment Opportunity
and Size)
Characteristics (Profitability,
Investment Opportunity and Size)
(Confirmed)
Changes in
Characteristics
Changes in Characteristics
(Confirmed)
Changes in Propensity
to Pay Dividends
Changes in Propensity to Pay
Dividends
(Confirmed)
Source: Developed for this research
As illustrated by figure 3.11, the phenomenon of disappearing dividends in Thailand
will be explained by summary statistics and logistic regression. The characteristics of
listed firms, their propensity to pay dividends and the concentration of dividends will
firstly be identified by the summary statistics and confirmed by logit regression
analysis. The significance of each factor will be analysed with logit regression analysis
and this will answer the research hypotheses, questions and problems which were
developed for this study.
213
Chapter 5: Analysis of Data and Findings
The characteristics of firms, the propensity to pay dividends and the concentration of
dividends and market factors are discussed in chapter 4. All terms and definitions will
also be defined for this research in chapter 4. A discussion of summary statistics and
logistic regression analysis and why they represent the best methodology for this study
will also be presented in Chapter 4.
3.7
CONCLUSION
This chapter reviewed the key theoretical issues associated with finance, capital market,
investment, dividend policies and the phenomenon of disappearing dividends. This
review of literature has shed light on the issues associated with the phenomenon,
efficient market hypothesis, financial ratio analysis, stock valuation and value and
growth stocks. The discussion also focused on the existence of the phenomenon in the
capital markets of the developed economies and indicated that the issue could be
present in Thailand. Finally, a model was developed to illustrate how the research
problem will be answered. The methodology of this study will be discussed in detail in
Chapter 4.
214
Chapter 5: Analysis of Data and Findings
C
HAPTER
4.1
4
RESEARCH METHODOLOGY
INTRODUCTION
Chapter 3 reviewed the literature on the phenomenon of disappearing dividends, the
characteristics of listed firms, the propensity to pay dividends, dividend policy and capital
market theory. This chapter discusses the methodology and data used for investigating
whether the phenomenon of disappearing dividends is present in Thailand. This
methodology will investigate the characteristics of listed firms, changes in these
characteristics over time and the propensity to pay dividends of firms.
Figure 4.1 illustrates the methodology of this study and the structure of this chapter.
Section 4.2 discusses the research problem and how it was defined. The research design
of this study is explained in section 4.3. The techniques used for categorizing and
confirming data are presented in section 4.4. This section also presents the model which
has been developed for this study. Section 4.5 presents a plan for analysing the data.
Finally, section 4.6 is a conclusion to this chapter.
215
Chapter 5: Analysis of Data and Findings
Figure 4.1: The Structure of Chapter Four
4.1 Introduction
4.2 Problem Discovery and Definition
4.3 Research Design
4.3.1
Purpose of the Study
4.3.2
Selecting
Design
the
4.3.3 Selecting the Research
Technique
Research
4.4 Data Classification and Methodology
4.4.1
Data Availability and
Period of Study
4.4.2
Summary Statistics
4.4.3
Logit Regression
4.5 Plan for Data Analysis
4.6 Conclusion
Source: Developed for this research
216
Chapter 5: Analysis of Data and Findings
4.2
PROBLEM DISCOVERY AND DEFINITION
The first step in a business research project is to define the problem. ‘Problem definition
is a specific area that will be clarified by answering a number of research questions or
hypotheses’ (Zikmund 2000, p. 88). The nature of the research problem will therefore
influence important parts of the research, including the research design, selection of
samples, data gathering, data analysis and evaluation (Black 1999; Zikmund 2000).
Figure 4.2 presents the research problem and issues of this study, the research hypotheses
and methodology developed for responding to the research problems and questions.
Figure 4.2: Links between the Research Problem, Research Questions, Research Hypotheses and
Methodology
RESEARCH
PROBLEM
‘How do the
characteristics
of publicly
traded firms
and their
propensity to
pay dividends
influence the
payment of
dividends and
the reasons
for their
disappearance
from the
stock market
of Thailand?’
RESEARCH
HYPOTHESES
RESEARCH QUESTIONS
RESEARCH
METHODOLOGY
RESEARCH DESIGN
Q1: What is the background and
performance of the Stock Exchange
SAMPLING DESIGN
of Thailand and its listed firms?
Q2:
How
have
researchers
investigated dividends and their
patterns throughout the world?
Q3:
What
is
the
9 Research
Hypotheses
DATA COLLECTION
METHODS
appropriate
methodology for collecting and
analysing data?
DATA ANALYSIS
TECHNIQUES
Q4: Does the data from Thailand
support the model?
Q5:
How
could
a
model
be
CONCLUSIONS AND
REPORTS
implemented?
Source: Developed for this research
217
Chapter 5: Analysis of Data and Findings
4.3
RESEARCH DESIGN
The research design of a study is a plan, road map, or framework specifying the methods
and procedures for collecting and analysing the information which is required for
answering a research problem (Emory 1995; Zikmund 2000). The determination of
appropriate sources of information, the design technique and sampling methodology of a
research project are carried out in the research design process (Zikmund 2000).
4.3.1
PURPOSE OF THE STUDY
There are two main forms of business research, namely, basic research and applied
research (Zikmund 2000; Cavana, Delahaye and Sekaran 2001).
Basic Research
Basic research is conducted for the purpose of theory development. It involves building a
body of knowledge which is refined by attempting to explain, or understand the problems
occurring in organisations. It does not directly involve solving a particular problem
(Zikmund 2000; Cavana, Delahaye and Sekaran 2001).
Applied Research
Applied research is undertaken for applying or testing the existence of a theory and for
evaluating how well this theory describes, or solves current business problems (Zikmund
2000; Cavana, Delahaye and Sekaran 2001).
It is possible that organisations will apply knowledge from basic research to solve
specific problems. Therefore, basic and applied research may be adopted for solving a
business research problem (Cavana, Delahaye and Sekaran 2001). Basic and applied
research need to be ‘undertaken in a scientific manner’ (Cavana, Delahaye and Sekaran
2001, p. 13). However, applied research is more popular than basic research because it
often involves a shorter period of investigation and lower costs than basic research
(Cavana, Delahaye and Sekaran 2001).
218
Chapter 5: Analysis of Data and Findings
This study is an example of applied research which will answer questions to specific
problems or situations. It will examine whether dividends are disappearing from the Thai
capital market and considers changes in the characteristics of listed firms and their
propensity to pay dividends.
4.3.2
SELECTING THE TYPE OF RESEARCH DESIGN
Business research can be classified on the basis of technique or function (Zikmund 2000).
The nature of the research problem will determine the research design. According to
Zikmund (2000), there are three types of research designs, namely: (1) exploratory
research, (2) descriptive research, and (3) causal research.
Exploratory research is the initial research which is conducted to clarify and define the
nature of a problem (Zikmund 2000). This research is conducted when the researchers
lacks a clear idea of the problems that will be met during the course of the study (Emory
and Cooper 1991). Exploratory research aids researchers or organisations to develop the
concepts more clearly, establish priorities and improve the research design especially
when the area of investigation is new, or vague (Emory and Cooper 1991). Exploratory
research also helps to save time and money. Researchers and managers are often urged to
obtain prompt answers to research questions (Emory and Cooper 1991).
Descriptive research seeks to identify and describe, but not to explain a phenomenon
(Ticehurst and Veal 2000). It helps in determining answers to who, what, when, where
and how questions (Zikmund 2000). Descriptive studies are based on a previous
understanding or knowledge of the nature of the research problem or phenomenon
(Zikmund 2000).
The goal of Causal research is to identify, explain, understand and predict the
relationship between the cause and effect of variables (dependent and independent
variables) (Zikmund 2000).
219
Chapter 5: Analysis of Data and Findings
Emory and Cooper (1991) discussed John Stuart Mill’s method of agreement, which
states ‘when two or more cases of a given phenomenon have one and only one condition
in common, that condition may be regarded as the cause (or effect) of the phenomenon’
(Emory and Cooper 1991, p. 151). In short, causal research attempts to prove when one
action is undertaken, another action will follow.
After considering the research problem which was stated in section 4.2, the research
questions and hypotheses will investigate the relationship between dividends and (1) the
characteristics of listed firms and (2) the propensity to pay dividends of firms. The
relationships which are identified will aid in determining whether dividends are
disappearing from Thailand’s capital market. These relationships will be of a cause-andeffect type. They will indicate whether changes in firm characteristics and the propensity
to pay dividends of listed firms have led to a disappearance of dividends from Thailand’s
capital market. An analysis of these relationships is an example of ‘causal research’ and
this study is therefore classified as casual research.
4.3.3
SELECTING THE RESEARCH TECHNIQUE
A research design (exploratory, descriptive and causal research) is conducted with the aid
of research methods and techniques. These techniques include surveys, experiments,
observation and the analysis of secondary data (Zikmund 2000).
220
Chapter 5: Analysis of Data and Findings
Figure 4.3: Alternatives for the Basic Research Method
Problem Discovery and Definition
Selection of Basic Research Method
Survey:
(1) Interview
(2) Questionnaire
Experiment
Observation
Secondary
Data
Source : Emory and Cooper (1991); Zikmund (2000), p. 55
A Survey is commonly used for collecting primary data. There are two main techniques
for conducting a survey, namely, interviews and questionnaires. The most important part
of developing a survey is preparing a questionnaire, or questions which will be asked at
an interview. The process includes preparing questions, determining the list of questions,
and designing the format and structure of the questionnaire interview. The wording of the
survey questions (avoiding the use of jargon) is very important because it will ensure the
response is accurate, unbiased and useful (Clover and Balsley 1984; Ticehurst and Veal
2000; Zikmund 2000).
An experiment is an investigation of a cause-and-effect relationship. Experiments deal
with variables, which are controlled, whilst others are not controlled (Zikmund 2000).
An observation is an unobtrusive technique (Emory 1995) which records, or gathers
information on the objects, or behaviour which is being studied, without intervening with
these objects, or people or letting them know that they are being observed or recorded
(Ticehurst and Veal 2000, Zikmund 2000).
221
Chapter 5: Analysis of Data and Findings
The study of secondary, or historical data, involves the use of data which was collected
by earlier researchers for a reason which may be related to the current research topic. The
secondary data technique also includes the use of historical data which is generally
available to the public. Secondary data techniques can be applied in all types of research
(exploratory, descriptive and causal research) (Clover and Balsley 1984; Emory and
Cooper 1991; Zikmund 2000).
The data which is needed for investigating the phenomenon of disappearing dividends in
Thailand’s capital market exits is secondary form. Useful secondary data is contained in
the financial statements of listed firms, SET’s annual report, BOT’s annual report and the
financial press. This information includes cash dividend payments, earnings, assets,
liabilities, interest expenses, earnings before interest and taxes, interest rates, market
capitalisation rates, and finally, data on the volume of trade. An analysis of recording
data is an appropriate technique for this study because it focuses on the cause-effect
relationship between dividends, their disappearance, the characteristics of firms and the
propensity to pay dividends. All information required for the analysis is readily available
in secondary form and at minimal cost. Therefore, the secondary data approach was
considered appropriate research method for this study.
4.4
DATA CLASSIFICATION AND METHODOLOGY
This section discusses the availability of data, the period of study and the techniques used
for explaining the phenomenon of disappearing dividends. Subsection 4.4.1 identifies the
dependent and independent variables, listed firms are categorized into several groups and
the three characteristics of these groups are also discussed. Subsection 4.4.2 and 4.4.3
discuss the summary statistics and logit regression in detail.
222
Chapter 5: Analysis of Data and Findings
4.4.1
DEPENDENT AND INDEPENDENT VARIABLES
The data used in this study measures cash dividend payments, the characteristics of firms
and their propensity to pay dividends. Zikmund (2000, p. 91) defines a variable as
‘anything that may make differences in numerical values’. The term dependent and
independent variables are frequently used in statistical analysis to describe cause-effect
relationships (Ticehurst and Veal 2000). This section identifies and defines the variables
that are used in this study. Emory and Cooper’s (1991) illustration of how to define
dependent and independent variables is summarised in figure 4.4.
Figure 4.4: Defining Independent and Dependent Variables
Independent Variable
Dependent Variable
Presumed Cause…..
Presumed Effect…..
Stimulus…..
Response…..
Predicted from…..
Predicted to…..
Antecedent…..
Consequence…..
Manipulated…..
Measured Outcome…..
Source: Emory and Cooper (1991)
The independent variable is the presumed cause, predicted from, manipulated, antecedent
and it stimulates the dependent variable. The dependent variable is the presumed effect,
predicted to, measured outcome, consequence, and it responds to the independent
variable (Emory and Cooper 1991; Page and Meyer 2000).
223
Chapter 5: Analysis of Data and Findings
Dependent Variable
A dependent variable ‘is a criterion, or a variable that is to be predicted or explained’
(Zikmund 2000, p. 91). Therefore, in this study the dependent variable is the percentage
of dividend payers in the Thai capital market.
Independent Variable(s)
An independent variable ‘is a variable that is expected to influence the dependent
variable’ (Zikmund 2000, p. 92). In this study, the independent variables are the
characteristics of listed firms (profitability, investment opportunities and sizes) and their
propensity to pay dividends.
4.4.2
DATA AVAILABILITY AND PERIOD OF STUDY
The data used in this study is obtained from the 1998, 1999, 2000, 2001 and 2002 Listed
Company Info (CD-ROMs) which is prepared by the Department of Investor Relations
and Communication (The Stock Exchange of Thailand), SIMS (The Stock Exchange of
Thailand Information Management System), and SET’s annual reports from 1990 to
2002. SIMS provides trading and company information of listed firms in Thailand in a
Microsoft Excel compatible format. SET’s records of trading activities and company data
is available on CD-ROM from 1990 till the present. The year 1990 is the starting period
for this study because SET’s computerised data base was only established in 1990.
Consequently, the data which was collected before 1990 is incomplete and incompatible
due to a different Securities Exchange Act applying to the disclosure of information for
listed firms. After the repealing of the SEC Act, B.E. 2517 (1974) and the SEC Act II,
B.E. 2527 (1984), the SEC Act B.E. 2535 (1992) is the most recent securities and
exchange act. Although adjustments have been made to this act (the SEC Act B.E. II
2542 (1999), III 2546 (2003)), it still applies to all listed firms from 1992 onward. In
addition, the data for this study was collected in mid 2003. Consequently, the most recent
information which was available was found on the annual financial statements for 2002.
224
Chapter 5: Analysis of Data and Findings
As a result, 2002 was chosen as the final year for the focus period of this study. Therefore
the sample period for this study will be 1990 – 2002.
As discussed earlier, Thailand has been affected by the Asian Economic Crisis of 1997,
and this study will examine the effect of this crisis. The time frame of this study is
divided into three periods, or steps; Step A: pre-crisis; Step B: post-crisis and; Step C: the
whole period. The ‘pre-crisis’ period covers data from 1990 to 1996 (seven years). The
‘post-crisis’ period lasts from 1997 to 2002 (six years). The ‘whole period’ covers data
from 1990 to 2002 (thirteen years). Differences in the relevant variables will be analysed
over these periods.
The data used for this study will also be analysed according to the time frames which
were discussed above.
As a result:
Step A, represents firms from 1990 to 1996 (pre-crisis)
Step B, represents firms from 1997 to 2002 (post-crisis)
Step C, represents firms from 1990 to 2002 (all times, total outlook)
The analysis of descriptive statistics and logistic regression will also use this
classification.
Given the study’s time frame, the number of firms in the sample is indicated by table 4.1.
Table 4.1: Number of Samples in this Study
Counts of Sample
SET Firms
All Firms
1990
209
1991
228
1992
254
1993
288
1994
328
1995
344
1996
378
1997
393
1998
388
1999
385
2000
368
2001
376
2002
386
Source: SET (2002b)
225
Chapter 5: Analysis of Data and Findings
Firms that are included in the sample must provide full information on their financial
statements (income statement and balance sheet) and trading information. The sample
contains financial data of firm (1990 to 2002) by calendar year on the following:
Total assets, average stock prices, shares outstanding, income before extraordinary items,
interest expenses, dividend yields, dividend per share, preferred dividends, preferred
stock value par value at the end of year t.
All firms in the sample have shareholder’s equity, total liabilities and common equity.
However, the figures for total assets in year t and t-1 permit the calculation of changes to
total assets as a ratio. This ratio represents the investment opportunity of a firm. Other
items used for time t.
A number of firms have provided information on balance sheet deferred taxes,
investment tax credits and income statement deferred taxes. However, this information is
not required in this study.
The firms included in the sample are found on SET’s database (SIMS) at the end of each
fiscal year. This ensures that the stocks of these firms are publicly traded in the market. A
firm must also have market equity data for December of year t, which includes average
prices, shares outstanding and the market capitalization rate.
The sample firms are categorized into 31 sectors, according to the nature of business;
(1) Agribusiness (2) Banking (3) Commerce (4) Communication (5) Construction,
building materials and furniture (6) Electronic components (7) Electrical and computer
parts (8) Energy (9) Entertainment (10) Finance (11) Food and Beverage (12) Health
Care Services (13) Hotels (14) Insurance (15) Jewelry (16) Mining (17) Packaging (18)
Pharmaceutical (19) Printing (20) Professional Services (21) Property (22) Pulp and
Paper (23) Textiles (24) Transportation (25) Vehicles (26) Warehouse (27) Household
226
Chapter 5: Analysis of Data and Findings
(28) Machinery (29) Chemicals (30) Others (31) Rehabilitation. The listed firms include
Thai and foreign firms.
Classification of Dividend Groups
As stated earlier, the dependent variable is the percentage of dividend payers from 1990
to 2002. Therefore, it is necessary to define the term ‘dividend payer’ or ‘payer’ so that
each firm can be classified in an appropriate group. A firm which is listed at SET is
classified as:
Payer, in a calendar year t, if the firm had positive dividends per share (Fama and French
2001).
Non-payer, if the firm has not paid dividends in time t (Fama and French 2001).
An all firms listing includes payers and non-payers, where payers pay dividends in year
t and non-payers do not (Fama and French 2001).
Former payer is a firm that does not pay in year t but paid dividends in year t-1
(previous year) or in a previous year (Fama and French 2001).
Never paid is a firm that has not paid dividends in time t-1 and t or in a previous year
(Fama and French 2001).
New list, if the firm is added SET’s database between January and December of t (Fama
and French 2001).
227
Chapter 5: Analysis of Data and Findings
Dividends are observed at the end of each year (31st December). This is an identical time
period as the factors which determine firm characteristics and the propensity to pay
dividends (profitability, investment opportunities and size). The study period commences
after the full financial statements of the listed firms have been published and recorded on
SIMS (The Stock Exchange of Thailand’s Information Management System).
As stated earlier, the number of firms in each category will be summed and calculated as
a percentage of total N, with a table of summary statistics, while t represents each
calendar year from January 1990 to December 2002. Figure 4.5 illustrates the process for
calculating the percentage of payers, non-payers, never paid firms and former payers.
Figure 4.5: Process of Counts and Percentages of SET Firms in SIMS in Time t
Listed Firms in Time t
All firms (N) (Seasoned
and New lists)
Payers (n, % of N)
Non Payers (n, % of N)
Never Paid (n, % of N)
Former Payers (n, % of N)
Source: Developed for this research
4.4.3 PROPOSED METHODOLOGY
4.4.3.1 THE INCIDENCE OF DIVIDEND PAYMENT BY THAI FIRMS
This study will identify whether the percentage of listed firms which pay dividends is
decreasing in Thailand.
228
Chapter 5: Analysis of Data and Findings
As stated earlier, the pattern of dividend behaviour in Thailand will be observed by
grouping the listed firms into the following: (1) dividend payers, (2) non-payers, (3)
former payers, or (4) never paid firms. New lists which commenced trading in time t will
be identified as newly listed firms and will be grouped as, newly listed firms which pay
dividends or newly listed firms which do not pay dividends (Fama and French 2001).
The summary statistics, or descriptive statistical techniques, are subgroup statistics for
the variables (Tabachnick and Fidell 2001). The descriptive statistics include the sum,
number of cases, mean, median, grouped median, standard error of the mean, minimum
value, maximum value, the range, variable value of the last category of the grouped
variable, standard deviation, variance, kurtosis, standard error of kurtosis, skewness,
standard error of skewness, percentage of total sum, percentage of total N, percentage of
sum in, percentage of N in, geometric mean and the harmonic mean (Saengkaew 1996;
Bryman and Cramer 1999, 2001; Tabachnick and Fidell 2001). Frequencies and means
are the descriptive statistics which are used most frequently. Frequencies include the sum
and percentage of each variable, while the mean is the average value of a total
population, or samples of variables (Ticehurst and Veal 2000).
The dividend policies of firms will also be considered by determining whether firms
continue, or cease paying dividends. The study will investigate whether the payers in
time t continue to pay dividends in subsequent years, or whether they cease paying and
become non-payers. In addition, the study will observe whether non-payers of period t
commence distributing dividends in the next period, or continue being non-payers (Fama
and French 2001). In short, the study uses descriptive statistics to summarise the
behaviour of firms which are categorised as payers and non-payers in time t-1 and the
current period (t). Both payers and non-payers in time t-1 are classified according to
whether they pay, or do not pay dividends in time t. These firms will continue to pay,
stop paying, continue to be a non-payer, start paying or will be delisted from SET.
229
Chapter 5: Analysis of Data and Findings
The dividend policy of payers (at time t-1) and non-payers (at time t-1) in time t are
expressed as follows:
Continue to pay: Firms that continue paying dividends in time t and t-1 (Fama and
French 2001).
Start paying: Firms that commence paying dividends in time t but were non-payers in
t-1 (Fama and French 2001).
Stop paying: Firms that were payers in time t-1 but cease paying dividends in time t
(Fama and French 2001).
Continue to be non-payer: Firms that were non-payers in time t-1 and continue to be
non-payers of dividends in the current period (t) (Fama and French 2001).
Delisted firms: Firms that were payers, or non-payers in time t-1 and were delisted in
time t (Fama and French 2001).
Figure 4.6 classifies the behaviour of listed firms by their dividend policy at time (t-1)
and time t.
230
Chapter 5: Analysis of Data and Findings
Figure 4.6: Payers and Non-payers in Time t-1
All firms (N)
Payers (t-1)
Non Payers (t-1)
Continue to Pay (t)
Start to Pay (t)
Stop Paying (t)
Do Not Pay (t)
Merge (t)
Merge (t)
Delist (t)
Delist (t)
Source: Developed for this research
The percentage of firms in each category will be calculated for each year between 1990
and 2002 using summary statistical methods (descriptive statistics approach). The method
for presenting the number of listed firms in each group will be similar to the analysis of
Fama and French (2001) (see table 4.1).
4.4.3.2 THE CHARACTERISTICS OF DIVIDEND PAYERS AND NON-PAYERS
A firm’s dividend policy is related to its profitability, investment opportunities and size
(Lintner 1963; James, Dodd and Kimpton 1985; Rappaport 1986; Fama and French 1988,
1995, 1997, 1999, 2000, 2001; Fisher and Jordan 1991; Penman and Sougiannis 1998;
Haugen 1997; Hurley and Johnson 1997; Yao 1997; Bringham, Gapenski and Ehrhardt
1999; Brealey and Myers 2000; Reilly and Brown 2000; Melicher and Norton 2000;
Petty et al. 2000; Gitman, Juchau and Flanagan 2002; Wetherilt and Weeken 2002). This
study categorises the characteristics of the sample firms with the aid of several financial
231
Chapter 5: Analysis of Data and Findings
variables (Fama and French 2001). As stated earlier, this analysis will observe whether
dividend payers, non-payers, former payers, and never paid firms differ according to (1)
profitability, (2) investment opportunities, and (3) size. The results will be compared over
the thirteen year period (1990-2002).
The characteristics are measured in the following ways:
Profitability: Two ratios are adopted as indicators of a firm’s profitability namely: the
ratio of aggregate earnings before interest to aggregate assets (Et/At) and the ratio of
aggregate common stock earnings over aggregate book equity (Yt/BEt). The components
of these ratios are discussed below:
For Profitability, the indicator is Et/At
Et/At is the ratio of aggregate earnings before interest to aggregate assets, or the amount
of earnings before interest generated by a single dollar of assets (aggregate asset) (Fama
and French 2001).
Where:
Et
: Earnings before interest for fiscal year t
Earnings Before Interest (Et) = Earnings Before Extraordinary Items + Interest Expense
+ Income Statement Deferred Taxes (if available)
At
: Assets at the end of fiscal year t
Total Assets (A t) = Short-term and Long-term assets
For Profitability, the additional indicator is Yt/BEt
232
Chapter 5: Analysis of Data and Findings
Yt/BEt is the ratio of aggregate common stock earnings divided by the aggregate book
equity, or the value of earnings from common stock which is generated by a single dollar
of the firm’s book equity (Fama and French 2001).
Where:
Yt
: After-tax earnings to common stock for fiscal year t
Earnings Available for Common Stock (Yt) = Earnings Before Extraordinary Items –
Preferred Dividends (If available) + Income Statement Deferred Taxes (If available)
BEt
: Book common equity at the end of fiscal year t
Book Equity (BEt) = Stockholder’s Equity (or common equity + preferred stock par
value or asset – liabilities) – Preferred Stock + Balance Sheet Deferred Taxes and
Investment Tax Credit (if available) – Post Retirement Asset (if available)
Investment Opportunities: The two ratios which are used as indicators of investment
opportunities include the following: the ratio of aggregate market value to the aggregate
book value of assets (Vt/At) and the ratio of the change in last year’s assets in baht to this
year’s total assets (dAt/At). The components of these ratios are discussed below:
For Investment opportunities, the indicator is Vt/At
Vt/At is the ratio of aggregate market value to the aggregate book value of assets (Fama
and French 2001).
Where:
Vt
: Total market value at the end of fiscal year t
Market Value of Firm (Vt ) = Assets – Book Equity + Market Equity
At
: Assets at the end of fiscal year t
Total Assets (At) = Short-term and Long-term assets
233
Chapter 5: Analysis of Data and Findings
For Investment opportunities, the additional indicator is dAt/At
dAt/At is the ratio of change in the asset value (in baht) of this year, compared to last year
divided by the total asset value (in baht) of this year. It is also the rate of growth of assets
of the firm (Fama and French 2001).
Where:
At
: Assets at the end of fiscal year t
Total Assets (At) = Short-term and Long-term assets
dAt
: Change in value of assets (in baht)
Change in Assets (dAt) = Time t assets in baht – Time t-1 assets in baht
Size is measured by a firm’s aggregate assets (At) and the ratio of aggregate liabilities to
aggregate assets (Lt/At). The components of this ratio are discussed below:
For Size, the indicator is At
At
: Assets at the end of fiscal year t (Fama and French 2001).
Where:
Total Assets (At) = Current assets and Long-term assets
For Size, the indicator is Lt/At
Lt/At is the ratio of aggregate liabilities to aggregate assets, or the value of liabilities
created from a single baht of assets (Fama and French 2001).
234
Chapter 5: Analysis of Data and Findings
Where:
Lt
: Liabilities at the end of fiscal year t
Total Liabilities (Lt) = Short-term and Long-term liabilities
At
: Assets at the end of fiscal year t
Total Assets (At) = Short-term and Long-term assets
The information required for classifying the characteristics of listed firms suggested by
Fama and French (2001) include:
Earnings Before Interest: are earnings on a firm’s assets in Thai baht at the end of every
fiscal year from 1990 to 2002. The earnings before interest is found on the income
statements of firms listed on SET’s database (SIMS).
Total Assets at year end for every listed firm is found on the published balance sheets
which are available on SIMS.
After-Tax Earnings to Common Stock is a ratio of available earnings to common
stocks and it is an important factor which influences the dividend policy of firms.
Book Value of Common Equity, data for every fiscal year is available on SIMS.
Total Market Value of listed firms is also available on SIMS from 1990 to 2002.
The Et, dAt, At, Yt, BEt, At,Vt, MEt and Lt items are also presented for each year and are
used for calculating the percentage average aggregate values of firms that pay dividends.
The discussion on descriptive statistics concludes with a summary report which identifies
the percentage of each dividend group (payers, non-payers, former payer, non-payer, new
listed) in Thailand’s capital market (1990 to 2002) and the average (mean) of each ratio
which has been used as an indicator of a firm’s characteristics (profitability, investment
235
Chapter 5: Analysis of Data and Findings
opportunities and size). This will include a three-stage discussion of the relative changes
in the size of the dividend groups before and after the Asian economic crisis (1997), and
during the total period of investigation (1991 to 2002).
The figure 4.7 indicates how the results for the three characteristics will be presented.
Figure 4.7: Average Firm Profitability, Investment Opportunities and Size
Listed Firms Characteristics
Profitability
Et/At,
Yt/BEt
Investment Opportunities
dAt/At
Vt/At
All Firms
Payers
Size
At
Lt/At
All New Listed
Non-Payers
Former Payers
Payers
Non-Payers
Never Paid
Source: Developed for this research
After calculating the mean values of each ratio and comparing the results for each group
according to the period scenarios, an ANOVA will be calculated for each year between
236
Chapter 5: Analysis of Data and Findings
1991 and 2002. ANOVA is the analysis of variance which is used for confirming that the
means of each group are significantly different (Mason, Lind and Marchal 1999;
Ticehurst and Veal 2000; Tabachnick and Fidell 2001). ANOVA is a statistical tool
which will confirm that the mean of the three characteristics for each group is statistically
different. An F-test is used to express the level of significance of the variation in the
means. The variation will be tested at the significance level of 0.05 (Mason, Lind and
Marchal 1999; Tabachnick and Fidell 2001).
Two ANOVA tests will be conducted for this study. The first ANOVA test will be for
payers and non-payers, and will confirm whether characteristics of payers differ from
non-payers. The second test will be conducted for payers and the two sub-groups of nonpayers (former payers and never paid firms). This test will confirm whether the
characteristics of payers differ from those of former payers and never paid firms.
The Stock Exchange of Thailand (SET) and the Bank of Thailand (BOT) will provide
information on fluctuations in the SET index, economic data, trends and forecasts,
benchmarks and other market indicators for this research.
4.4.3.2.1
THE EFFECTS OF PROFITABILITY, INVESTMENT
OPPORTUNITIES AND FIRM SIZE
To assess the effects of firm characteristics (profitability, investment opportunities, and
size) on the payment of dividends, Fama and French (2001) suggested that an estimation
of logistic regression is appropriate. Logistic regression (hereafter logit regression)
‘allows prediction of group membership, when predictors are continuous, discrete, or a
combination of the two’ (Tabachnick and Fidell 2001, p. 24). Thus, ‘it is an alternative to
both discriminant function analysis and logit analysis’ (Tabachnick and Fidell 2001, p.
24). Logit regression ‘allows one to evaluate the odds (or probability) of membership in
one of the groups based on the combination of values of the predictor variables’
(Tabachnick and Fidell 2001, p. 24).
237
Chapter 5: Analysis of Data and Findings
Therefore ‘Logit regression documents more formally the marginal effects of
profitability, investment opportunities and size on the likelihood that a firm pays
dividends’ (Fama and French 2001, p. 12).
The results of logit regression are expected to confirm the inferences on the role of
profitability, investment opportunities and firm size in influencing a firm’s dividend
policies (Fama and French 2001). The hypotheses developed in chapter 3 shall be tested
and answered using the t-statistics provided with the logit regression. Fama and
MacBeth (1973, p. 621) stated that ‘using t-statistics in the usual way does not lead to
serious errors’. This research applies t-statistics to test the significance of the
predetermined hypotheses.
As stated earlier the independent variables are profitability, investment opportunities and
size. Fama and French (2001) defined each characteristic as follows:
Probability is measured by the Et/At (Fama and French 2001) or earnings before interest
but after taxes standardized by total assets.
Investment opportunities are measured by dAt/At and Vt/At (Fama and French 2001) or
the growth rate in assets and the market-to-book ratio.
Firm size is measured by the percentage of SET firms that have the same or smaller
market capitalization as the specific firms (Fama and French 2001).
Following Fama and French (2001), the logit regression is divided and presented in two
sets. The first set is the logit regression which controls for two investment variables,
238
Chapter 5: Analysis of Data and Findings
namely, dAt/At and Vt/At. The second set is the logit regression control for the
investment variable dAt/At. In both sets, the remaining characteristics are unchanged.
Table 4.2: Comparison of Different Ratios Applied in Different Stages of this
Research
Characteristics
Approach (1)
Descriptive
Statistic
Profitability
Investment
Opportunities
Size
Et/At
Yt/BEt
dAt/At
Vt/At
At
Lt/At
Approach (2)
Logit Regression
Logit Regression
(controlling 2
(controlling 1
investment variables)
investment variable)
Et/At
Et/At
dAt/At
Vt/At
At
dAt/At
At
Source: Developed for this research
Later, logit regression is also used to estimate the probability that firms will pay
dividends. It will therefore allow the researcher to analyse and confirm the effects of a
firm’s characteristics on its dividend policy. In addition, logit regression tests whether the
findings of Fama and French (2001), Benito and Young (2001) and Aivazian, Booth and
Cleary (2003) could be applied for a developing capital market such as Thailand.
Figure 4.8 indicates how the characteristics of firms are confirmed by logit regression
analysis.
239
Chapter 5: Analysis of Data and Findings
Figure 4.8: How Logit Regressions is Used for Explaining the Dividend Policies of Firms
Average Coefficient
SETt
Size
Vt/At
t-statistic
dAt/At
Investment
Et/At
Profitability
Source: Developed for this research
Using t-statistics, each characteristic is confirmed as significant (<0.05) or nonsignificant (>0.05) (Tabachnick and Fidell 2001) variable in predicting dividend
behaviour of the listed firms. T-statistic values indicate the strength of relationship
between each characteristic and the likelihood that a firm will pay dividends (Tabachnick
and Fidell 2001).
4.4.3.3 CHANGING PROPENSITY TO PAY DIVIDENDS
Fama and French (2001) found that the characteristics of firms have changed and this has
affected the payment of dividends. As stated earlier, they also suggest that firms are less
likely to pay dividends if these characteristics are controlled over time. This phenomenon
was entitled ‘the lower propensity to pay dividends of the listed firms’.
The methodology adopted to test the propensity to pay dividends has several steps as
suggested by Fama and French (2001). The first step is to apply descriptive statistics by
categorizing listed firms into four groups, namely, (1) firms with positive earnings (Yt >
0), (2) firms with negative earnings (Yt < 0), (3) firms with earnings above investment
opportunities (Et > dAt) and (4) firms with earnings below investment opportunities (Et <
240
Chapter 5: Analysis of Data and Findings
dAt) (Fama and French 2001). This is to investigate the trend in the propensity to pay
dividends of listed firms from 1990 to 2002.
The next step is to estimate the probability that firms with constant characteristics
(profitability, investment opportunities and size) paid dividends between 1990 and 2002.
Later, the probabilities which were calculated for the base years 1991 to 1996 or before
the Asian Economic Crisis will be applied to the sample firm’s characteristics. This will
forecast the expected percentage of dividend payers from 1991 to 2002. The probabilities
of the characteristics are fixed at their base year (1991 to 1996). Therefore, any variation
in the number of expected payers will result from changes in the characteristics of the
sample firms (Fama and French 2001). The actual percentage of payers must be
calculated for each year from 1991 to 2002. The difference between the expected
percentage of payers and the actual percentage of payers will be an estimate of the
change in the propensity to pay dividends by listed firms between 1991 and 2002. A
decline in the propensity is indicated by the difference between the expected percentage
of payers and actual percentage of payers.
Figure 4.9 below indicates how the number of total firms and payers is counted and how
the proportion of payers in the sample is expressed as a percentage.
241
Chapter 5: Analysis of Data and Findings
Figure 4.9: Logit Regression Estimates of the Effect of Changing Firm Characteristics and
Decline in the Propensity to Pay Dividends
Firms
Actual
Percent
Payers
Expected
Percent
Expected Actual
Logit
Regression
forecasted
N
n
n/N x 100
Characteristics
changes
Propensity
to pay
Source: Developed for this research
Note: N is total number of firms, n is number of payers in time t
The Total number of firms (N) and payers (n) are counted. Then the actual percentage of
payers to total firms is found and compared to the expected percentage of payers
forecasted by logit regression. The results show the changes in characteristics over time.
Lastly, the differences between expected and actual percentage of payers in the market
determine the changes in propensity to pay dividends of listed firms.
Therefore the propensity to pay dividends would be discussed using descriptive statistics
and logit regression and the changes in the characteristics will also be formally identified
in this stage.
4.5
PLAN FOR DATA ANALYSIS
This section explains how the data will be analysed after the descriptive statistics and
logit regression have been conducted. In particular, this section discusses the data
processing, screening and analysis techniques which are applied in this research.
242
Chapter 5: Analysis of Data and Findings
4.5.1
DATA PROCESSING
Data processing has two main functions, namely, editing and coding of the raw data
(Zikmund 2000). In this study, the raw data is the published financial statements of firms
listed on the Thailand stock exchange. This data must be converted into an organised
form which can be used for answering the research problem. If the database is large, a
computerized data processing program is needed to arrange, analyse and interpret the
meaning of the data.
In this study, the raw data is presented on the financial statements as dividend payments
and firm characteristics. The information which is selected will be entered onto
spreadsheets with the aid of Microsoft EXCEL and SPSS (The Statistical Package for
Social Sciences) software. This will ensure the data can be analysed effectively.
4.5.2
VALIDITY
Validity refers to the extent that information collected in a research study, actually
reflects the phenomenon in the real world (Ticehurst and Veal 2000, Zikmund 2000).
There are two types of validity (1) internal validity which addresses whether the
independent variables are related to the dependent variables and (2) external validity
which addresses whether the scope of this research is applicable to the real world
(Ticehurst and Veal 2000; Zikmund 2000).
The ratios which estimate the importance of each firm characteristic have readily been
used by professionals and academics and are regarded as sound measures of profitability,
investment opportunities and firm size. The methodology is also widely accepted in the
field of finance throughout the world (Lintner 1963; Fama and Babiak 1968; Fama and
MacBeth 1973; James, Dodd and Kimpton 1985; Rappaport 1986; Fama and French
1988, 1995, 1997, 1999, 2001; Fisher and Jordan 1991; Hurley and Johnson 1994;
Haugen 1997; Yao 1997; Bringham, Gapenski and Ehrhardt 1999; Brealey and Myers
243
Chapter 5: Analysis of Data and Findings
2000; Melicher and Norton 2000; Petty et al. 2000; Reilly and Brown 2000; Gitman,
Juchau and Flanagan 2002; Wetherilt and Weeken 2002). The terms, categories, ratios
and methodologies are commonly used by researchers in the field of finance and
investment.
4.6
CONCLUSION
This chapter discussed the research design and methodology which will be used in this
study. A range of statistical techniques for investigating the disappearance of dividends
have been discussed. Chapter 5 provides the results of the data analysis which is used for
explaining the significance of the phenomenon of disappearing dividends in Thailand.
244
Chapter 5: Analysis of Data and Findings
C
HAPTER
5.1
5
ANALYSIS OF DATA AND FINDINGS
INTRODUCTION
Chapter 4 discussed the methodology and the plan for investigating the disappearing
dividends phenomenon in the capital market of Thailand. This chapter applies the
methodology and data together to answer the research problem, questions and
hypotheses developed in chapter 3.
Chapter 5 presents the results of the descriptive statistics and logistic regression. In
particular, significant variables are identified which demonstrate the characteristics of
listed firms which pay dividends, changes in these characteristics over time and the
propensity to pay dividends of these firms. This chapter also discusses the findings of
the hypothesis tests which were developed in chapter 3, and the implications of this
study for Thailand’s capital market.
The chapter is presented in eight sections as indicated by Figure 5.1. Section 5.2 defines
the descriptive statistics and provides the sample with these descriptions. The initial
findings of the descriptive statistics and a discussion of the phenomenon of
disappearing dividends in Thailand are provided in section 5.3. Section 5.4 outlines the
trends identified in the payment of cash dividends. Section 5.5 analyses the
characteristics of the dividend groups with the aid of ANOVA and logic regression.
Section 5.6 analyses the propensity to pay dividends of listed firms with descriptive
statistics. Section 5.7 uses logit regression to test the significance of the findings in
relation to the propensity to pay dividends of firms and change in the characteristics of
firms. Finally, section 5.8 is a conclusion to this chapter.
245
Chapter 5: Analysis of Data and Findings
Figure 5.1: The Structure of Chapter Five
5.1
5.2
5.2.1
Introduction
Descriptive Statistics
Descriptive Statistics
Definitions
5.2.2
A Description of the
sample
5.3 An Initial analysis of Descriptive Analysis
5.4 Time Trends in Cash dividend
5.4.1
Payers and their
dividends yield
5.5
5.5.1
5.2.3
5.4.2
Characteristics of dividend payers
Firm Characteristics
(Descriptive Statistics
Approach)
5.5.2 ANOVA test
5.5.4 Confirmation from Logit
Regressions
Review of Descriptive
Statistics Results
5.5.6
5.5.5
Dividend Pattern
Hypothesis Testing (1)
Average coefficient
5.6
5.7
Hypothesis Testing (2)
Group coefficient
The Propensity to pay dividends
(Descriptive approach)
Changes in characteristics
and propensity to pay
(Logit Regressions approach)
5.8
Conclusion
Source: Developed for this research
246
Chapter 5: Analysis of Data and Findings
5.2
DESCRIPTIVE STATISTICS
This section describes the descriptive statistics which are used in this tool in the study.
5.2.1
DESCRIPTIVE STATISTICS DEFINITIONS
Statistics is the science of collecting, organising, presenting, analysing, and
interpreting data (Mason, Lind and Marchal 1999, p. 3). In business, statistics is often
used for improving the quality of investment decisions (Mason, Lind and Marchal
1999). Statistics can be presented in two ways, namely: (1) numerically and (2)
graphically (Mason, Lind and Marchal 1999). The statistical findings of this study, in
both of these forms is discussed in the later part of this chapter.
There are two main types of statistics: (1) inferential statistics and (2) descriptive
statistics. The objective of inferential statistics is to observe a phenomenon from a
sample of a population (Mason, Lind and Marchal 1999). Descriptive statistics are used
for organising, summarising and presenting data in a simple, but informative way
(Mason, Lind and Marchal 1999; Ticehurst and Veal 2000; Zikmund 2000). Descriptive
statistics enable a researcher to better understand a set of data before undertaking more
advanced statistical analysis (Huck, Cormier and Bounds 1974). Frequencies and
means are two of the most common descriptive statistics. Frequencies show the
‘counting and percentages for each variable’ while the means is an average of the total
population or sample (Ticehurst and Veal 2000, p. 181).
5.2.2
A DESCRIPTION OF THE SAMPLE
As discussed in chapter 4, the sample used in this study consists of firms listed on the
Stock Exchange of Thailand between 1990 and 2002. Listed firms are classified into
thirty-one sectors, or industries. The names of these industries are as follows:
(1) Agribusiness (2) Banking (3) Commerce (4) Communication (5) Construction,
building materials and furniture (6) Electronic components (7) Electrical and computer
247
Chapter 5: Analysis of Data and Findings
parts (8) Energy (9) Entertainment (10) Finance (11) Food and Beverage (12) Health
Care Services (13) Hotels (14) Insurance (15) Jewelry (16) Mining (17) Packaging (18)
Pharmaceutical (19) Printing (20) Professional Services (21) Property (22) Pulp and
Paper (23) Textiles (24) Transportation (25) Vehicles (26) Warehouse (27) Household
(28) Machinery (29) Chemicals (30) Others (31) Rehabilitation.
An objective of this study is to identify and explain the pattern of cash dividends from
1990 to 2002. As discussed earlier, this study will explore the effect of the Asian
Economic Crisis on the payment of dividends. The study focuses on three periods of
time, (1) Step A, pre-crisis (2) Step B, post-crisis and (3) Step C, all times (1990-2002).
The ‘pre-crisis’ period contains data from 1990 to 1996 and represents the period
before the Asian Economic Crisis. The ‘post-crisis’ is represented by the period 1997
to 2002 and finally the ‘all times’ period focuses on the period 1990 to 2002. The data
will be classified according to the framework and will be analysed with descriptive
statistics and logistic regression.
A number of listed firms were deleted from the sample because their financial
statements were incomplete. SEC requires that the financial statement of listed firms
should be reviewed by an auditor of SET. However, as discussed in chapter 4, a number
of listed firms provided insufficient information on their financial statements and have
not been audited. If a firm’s financial statement has not been provided, it will be forced
to cease trading (a ST, or trading suspended sign will be placed on its securities).
If a firm fails to provide an annual financial statement for 2 years, or only provides a
quarterly financial statement, it will be delisted. The Minister of Finance (MOF) and
the Bank of Thailand (BOT) forced a number of financial institutions to cease
operating, or present them with plans for a merger or acquisition, during the Asian
Economic Crisis. Some of these financial institutions survived a while, some closed. In
short, the financial statements and trading information of firms which closed are
incomplete. To ensure the analysis is accurate these firms have been deleted from the
sample.
248
Chapter 5: Analysis of Data and Findings
5.3
AN INITIAL ANALYSIS OF DESCRIPTIVE STATISTICS
The percentage of listed firms paying cash dividends in Thailand (SET) fell from 84.2
percent in 1990, to 46.4 percent in 2002, a decline of about 40 percent (table 5.1). As
stated earlier in chapter 3, the payment of dividends has also declined in the US (Fama
and French 2001). These researchers suggested that this trend resulted from a change in
the characteristics and the propensity to pay dividend of firms and their results are
supported by a number of studies (La Porta et al. 2000; DeAngelo, DeAngelo and
Skinner 2002; Baker and Wurgler 2003a).
To explain the phenomenon disappearing dividends in Thailand, samples of listed firms
from 1990 to 2002 have been collected. This section compares the results of this study
with Fama and French (2001). These results include changes in the characteristics and
the propensity to pay dividends of listed firms in developed and emerging capital
markets.
5.4
TIME TRENDS IN CASH DIVIDENDS
Generally, dividends are taxed at a higher rate than capital gains and they are less
valuable for investors (Fama and French 2001). Similar taxation policies are applied in
Thailand because dividends are taxed at 10 percent, although a tax credit of 3/7 is
received at the end of the year. Consequently, the real tax rate of dividends is 5.71
percent in Thailand. Capital gains are currently exempt from taxation (Cholviroj 1989;
SEC 2002). As discussed in chapter 3, firms paying dividends appear to be
disadvantaged (they experience a higher tax rate and low investment opportunities) as
their equity costs are higher than those of firms that do not pay dividends. However,
some firms continue to pay dividends and investors and academics have found it
difficult to explain this puzzle (Fama and French 2001).
This section focuses on trends in the payment of dividends between 1990 and 2002.
The sample firms are classified as payers, or non-payers of dividends in the present
249
Chapter 5: Analysis of Data and Findings
time (t). As explained in chapter 4, payers and non-payers are classified according to
whether they paid dividends in the past period (t-1) and present period (t).
Table 5.1: Counts and Percent of SET firms in Different Dividend Groups
Counts of SET firms
All Firms
New Lists
1990
209
1991
228
45
1992
254
38
1993
288
33
1994
328
38
1995
344
23
1996
378
34
1997
393
4
1998
388
3
1999
385
4
2000
368
11
2001
376
7
2002
386
20
Percent of SET firms
Payers
Non-Payer
Former-Payer
Never-Paid
New Lists
1990
84.2
15.8
1991
84.6
15.4
10.1
5.3
19.7
1992
90.6
9.4
5.1
4.3
15.0
1993
85.1
14.9
8.0
6.9
11.5
1994
83.8
16.2
9.5
6.7
11.6
1995
84.3
15.7
12.8
2.9
6.7
1996
81.0
19.6
14.6
5.0
9.0
1997
71.5
28.5
22.9
5.6
1.0
1998
28.4
71.6
65.2
6.4
0.8
1999
29.9
70.1
62.9
7.3
1.0
2000
35.3
64.9
56.5
8.4
3.0
2001
40.7
59.3
48.7
10.6
1.9
2002
46.4
53.6
40.2
13.5
5.2
82.2
73.7
51.5
52.6
82.6
73.5
50.0
33.3
0.0
27.3
57.1
20.0
New Lists that Pay
Note: Payers paid dividends in time t while non-payers did not. Non-payers are dividend into two subgroups, namely, former payers (firms that do not pay in time t but paid in the previous year) and firms
that have never paid dividends. New lists are firms which were listed in time t and new lists that paid
dividends are classified as payers.
Source: SET (1990, 1993, 1996, 1999, 2002a)
Patterns in the payment of dividends are explored from 1990 to 2002 with statistics
from the SIMS database. In 1990, 84.2 percent of listed firms paid dividends (see table
5.1). The percentage of payers rose to its peak of 90.6 percent in 1992 and fell
gradually to 71.5 percent in 1997. The percentage of payers fell to 28.4 percent in 1998.
This figure was recorded a year after the Asian crisis. However, the percentage of firms
paying dividends recovered slightly in 1999 (29.9 percent) and rose to 46.4 percent by
2002. However, the percentage of firms paying dividends in 2002 (46.4 percent) was
well below the 1990 level (84.2 percent).
To better understand changes in the pattern of dividend payments, the sample firms
were classified by industry according to SET’s classification.
Table 5.2 shows the number of firms in each industry group from 1990 to 2002.
250
Chapter 5: Analysis of Data and Findings
Table 5.2: Number of Sample Firms in Each Industry
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
AGRIBUSINESS
Number of firms in each industry
11
15
17
24
28
27
28
26
24
21
21
20
20
BANKING
15
14
10
10
10
9
10
11
10
13
14
14
14
BUILDING AND FURNISHING MATERIALS
13
19
21
24
25
27
30
33
29
25
23
24
17
CHEMICALS AND PLASTICS
5
7
8
8
10
13
15
15
14
13
13
12
12
COMMERCE
9
9
12
12
14
13
14
15
14
14
12
13
14
1
2
3
5
9
9
10
10
10
10
10
11
12
13
2
10
10
11
11
10
7
22
42
43
51
49
ELECTRICAL PRODUCTS AND COMPUTER
6
7
8
9
10
11
12
13
11
9
11
8
10
ELECTRONIC COMPONENTS
2
2
2
3
4
6
8
8
8
8
7
8
8
ENERGY
3
3
3
5
7
8
9
9
9
9
10
10
11
COMMUNICATION
COMPANIES UNDER REHABILITATION
1
1
1
1
2
3
6
7
7
8
7
7
10
FINANCE AND SECURITIES
24
18
13
13
14
17
20
21
20
21
19
21
26
FOODS AND BEVERAGES
15
15
16
18
20
20
21
24
24
23
21
22
22
HEALTH CARE SERVICES
4
6
8
9
11
11
13
13
13
13
11
10
10
10
11
11
12
12
12
12
12
12
12
12
12
10
2
4
6
8
8
8
8
8
8
6
7
7
17
18
19
19
19
20
22
22
22
21
21
21
ENTERTAINMENT AND RECREATION
HOTELS AND TRAVEL SERVICES
HOUSEHOLD GOODS
INSURANCE
JEWELRY AND ORNAMENTS
13
2
MACHINARY AND EQUIPMENT
MINING
OTHERS
PACKAGING
PHARMACEUTICAL PRODUCTS AND
COSMETICS
1
4
4
4
4
4
4
4
3
2
2
2
2
1
2
2
3
4
6
6
5
4
4
3
3
1
1
1
1
1
1
1
1
1
1
1
1
1
2
4
4
5
5
5
6
5
4
5
5
5
11
13
12
13
17
17
17
17
15
14
14
12
13
2
2
2
2
2
2
2
2
2
2
2
2
2
PRINTING AND PUBLISHING
4
6
7
8
8
7
10
10
10
8
7
8
8
PROFESSIONAL SERVICES
1
2
2
2
2
2
2
2
2
2
2
2
2
PROPERTY DEVELOPMENT
6
11
15
21
25
28
32
37
35
26
21
22
29
3
2
2
2
2
3
5
5
5
5
5
4
4
24
22
25
27
28
28
27
28
25
25
24
24
24
TRANSPORTATION
2
3
3
4
5
6
7
7
8
8
8
8
8
VEHICLES AND PARTS
3
6
6
6
8
9
10
10
10
8
8
8
9
WAREHOUSE AND SILO
4
4
4
4
4
4
4
4
5
5
4
4
4
PULP AND PAPER
TEXTILES, CLOTHING AND FOOTWEAR
Source: SET (1990, 1993, 1996, 1999, 2002a)
Table 5.2 indicates that the number of listed firms has increased in most sectors since
1990. The highest growth has been in rehabilitation, electronic components, energy,
entertainment and recreation, transportation vehicles and parts, household goods and
property development. The number of firms in these industries more than tripled
between 1990 and 2002.
To comply with the research problem stated in chapter 1, trends in the number of
dividend payers in each industry group are analysed.
251
Chapter 5: Analysis of Data and Findings
Table 5.3: Number of Dividend Payers in Each Industry
Number of payer in each industry
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
9
12
17
20
25
23
22
20
9
11
13
18
16
BANKING
11
13
9
9
10
9
10
9
7
0
0
0
0
BUILDING AND FURNISHING MATERIALS
12
16
18
21
18
19
24
19
2
2
2
3
6
CHEMICALS AND PLASTICS
3
3
8
8
9
11
9
10
3
5
6
6
7
COMMERCE
9
8
12
12
13
13
13
13
4
6
7
7
8
COMMUNICATION
0
2
3
3
5
6
6
8
2
1
0
2
3
COMPANIES UNDER REHABILITATION
ELECTRICAL PRODUCTS AND
COMPUTER
9
2
5
4
5
5
4
1
0
0
0
0
0
6
7
7
8
9
9
10
9
3
3
4
4
7
ELECTRONIC COMPONENTS
2
2
1
2
4
5
6
7
2
3
2
3
4
ENERGY
3
3
3
2
5
6
7
7
4
5
6
5
8
ENTERTAINMENT AND RECREATION
1
1
1
1
2
3
5
7
4
3
3
3
3
FINANCE AND SECURITIES
21
17
13
12
14
17
19
18
2
4
6
4
7
FOODS AND BEVERAGES
13
11
15
15
15
18
18
19
11
12
13
13
13
HEALTH CARE SERVICES
4
6
8
8
10
10
11
10
2
1
4
5
6
HOTELS AND TRAVEL SERVICES
7
11
9
10
11
10
10
9
3
5
7
9
9
2
4
5
6
7
5
8
4
3
1
4
4
11
16
16
17
17
17
19
19
14
14
13
17
17
2
4
4
4
4
4
3
2
1
1
1
1
2
0
2
2
3
4
6
6
1
1
1
1
1
0
AGRIBUSINESS
HOUSEHOLD GOODS
INSURANCE
JEWELRY AND ORNAMENTS
MACHINARY AND EQUIPMENT
MINING
1
1
1
1
1
0
0
0
0
0
0
0
OTHERS
0
2
1
2
3
3
3
5
2
2
3
4
3
PACKAGING
PHARMACEUTICAL PRODUCTS AND
COSMETICS
7
10
11
11
12
14
13
11
4
4
5
4
7
1
2
1
2
2
2
2
2
0
1
1
1
1
PRINTING AND PUBLISHING
4
5
7
8
8
6
8
8
4
4
5
5
5
PROFESSIONAL SERVICES
1
2
2
2
2
2
2
2
2
2
2
2
2
PROPERTY DEVELOPMENT
5
9
12
15
20
25
25
16
1
2
1
3
6
2
2
2
2
2
2
4
2
0
0
0
2
2
23
13
25
24
24
23
23
19
11
15
15
16
17
TRANSPORTATION
2
3
3
3
4
5
6
4
2
0
1
1
4
VEHICLES AND PARTS
3
6
6
6
8
8
9
7
3
2
5
6
7
WAREHOUSE AND SILO
4
2
4
4
4
4
4
4
3
3
3
4
4
PULP AND PAPER
TEXTILES, CLOTHING AND FOOTWEAR
Source: SET (1990, 1993, 1996, 1999, 2002a)
Table 5.3 indicates that the number of firms which pay dividends expanded in most
industry sectors between 1990 and 2002. The number of payers declined in eight
industries. These industries include agribusiness, banking, building, commerce,
rehabilitation, finance, mining and textiles.
As discussed earlier, table 5.2 indicates that the number of listed firms increased
between 1990 and 2002. Therefore, it is not surprising that the number of payers in
each industry has also increased over this period. However, table 5.4 presents the
percentage of firms which pay dividends in each industry as a standard measurement of
252
Chapter 5: Analysis of Data and Findings
the increase, or decrease in the proportion of dividend payers in each industry over
time.
Table 5.4: Percentage of Payers to Total Number of Listed Firms in Each Industry
Percent of payers to total number of
firms in each industry
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
AGRIBUSINESS
81.8
80.0
100.0
83.3
89.3
85.2
78.6
76.9
37.5
52.4
61.9
90.0
80.0
BANKING
BUILDING AND FURNISHING
MATERIALS
73.3
92.9
90.0
90.0
100.0
100.0
100.0
81.8
70.0
0.0
0.0
0.0
0.0
92.3
84.2
85.7
87.5
72.0
70.4
80.0
57.6
6.9
8.0
8.7
12.5
35.3
CHEMICALS AND PLASTICS
60.0
42.9
100.0
100.0
90.0
84.6
60.0
66.7
21.4
38.5
46.2
50.0
58.3
100.0
88.9
100.0
100.0
92.9
100.0
92.9
86.7
28.6
42.9
58.3
53.8
57.1
0.0
100.0
100.0
60.0
55.6
66.7
60.0
80.0
20.0
10.0
0.0
18.2
25.0
45.5
45.5
40.0
14.3
0.0
0.0
0.0
0.0
0.0
COMMERCE
COMMUNICATION
COMPANIES UNDER
REHABILITATION
ELECTRICAL PRODUCTS AND
COMPUTER
69.2
100.0
50.0
40.0
100.0
100.0
87.5
88.9
90.0
81.8
83.3
69.2
27.3
33.3
36.4
50.0
70.0
ELECTRONIC COMPONENTS
100.0
100.0
50.0
66.7
100.0
83.3
75.0
87.5
25.0
37.5
28.6
37.5
50.0
ENERGY
ENTERTAINMENT AND
RECREATION
100.0
100.0
100.0
40.0
71.4
75.0
77.8
77.8
44.4
55.6
60.0
50.0
72.7
100.0
100.0
100.0
100.0
100.0
100.0
83.3
100.0
57.1
37.5
42.9
42.9
30.0
FINANCE AND SECURITIES
87.5
94.4
100.0
92.3
100.0
100.0
95.0
85.7
10.0
19.0
31.6
19.0
26.9
FOODS AND BEVERAGES
86.7
73.3
93.8
83.3
75.0
90.0
85.7
79.2
45.8
52.2
61.9
59.1
59.1
HEALTH CARE SERVICES
100.0
100.0
100.0
88.9
90.9
90.9
84.6
76.9
15.4
7.7
36.4
50.0
60.0
70.0
100.0
81.8
83.3
91.7
83.3
83.3
75.0
25.0
41.7
58.3
75.0
90.0
0.0
100.0
100.0
83.3
75.0
87.5
62.5
100.0
50.0
37.5
16.7
57.1
57.1
HOTELS AND TRAVEL SERVICES
HOUSEHOLD GOODS
INSURANCE
JEWELRY AND ORNAMENTS
84.6
94.1
88.9
89.5
89.5
89.5
95.0
86.4
63.6
63.6
61.9
81.0
81.0
100.0
100.0
100.0
100.0
100.0
100.0
75.0
50.0
33.3
50.0
50.0
50.0
100.0
33.3
0.0
0.0
100.0
100.0
100.0
100.0
100.0
100.0
20.0
25.0
25.0
33.3
MINING
100.0
100.0
100.0
100.0
100.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
OTHERS
0.0
100.0
25.0
50.0
60.0
60.0
60.0
83.3
40.0
50.0
60.0
80.0
60.0
63.6
76.9
91.7
84.6
70.6
82.4
76.5
64.7
26.7
28.6
35.7
33.3
53.8
50.0
MACHINARY AND EQUIPMENT
PACKAGING
PHARMACEUTICAL PRODUCTS
AND COSMETICS
50.0
100.0
50.0
100.0
100.0
100.0
100.0
100.0
0.0
50.0
50.0
50.0
PRINTING AND PUBLISHING
100.0
83.3
100.0
100.0
100.0
85.7
80.0
80.0
40.0
50.0
71.4
62.5
62.5
PROFESSIONAL SERVICES
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
PROPERTY DEVELOPMENT
83.3
81.8
80.0
71.4
80.0
89.3
78.1
43.2
2.9
7.7
4.8
13.6
20.7
PULP AND PAPER
TEXTILES, CLOTHING AND
FOOTWEAR
66.7
100.0
100.0
100.0
100.0
66.7
80.0
40.0
0.0
0.0
0.0
50.0
50.0
95.8
59.1
100.0
88.9
85.7
82.1
85.2
67.9
44.0
60.0
62.5
66.7
70.8
TRANSPORTATION
100.0
100.0
100.0
75.0
80.0
83.3
85.7
57.1
25.0
0.0
12.5
12.5
50.0
VEHICLES AND PARTS
100.0
100.0
100.0
100.0
100.0
88.9
90.0
70.0
30.0
25.0
62.5
75.0
77.8
WAREHOUSE AND SILO
100.0
50.0
100.0
100.0
100.0
100.0
100.0
100.0
60.0
60.0
75.0
100.0
100.0
Source: SET (1990, 1993, 1996, 1999, 2002a)
Table 5.4 indicates that the percentage of payers have declined in 22 out of 31
industries. The number of listed dividend payers has therefore declined in most
industries in Thailand.
253
Chapter 5: Analysis of Data and Findings
5.4.1 PAYERS AND THEIR DIVIDEND YIELD
SET publishes data on the annual dividend yield of listed firms that paid dividends in
time t. The dividend yield is the total common stock dividend of the last 12 months
divided by the market capitalisation of the stock and multiplied by 100 (SET 2002a).
During the study period (1990 to 2002), the average market dividend yield has
fluctuated due to changes (decreased and increased) in the number of payers and stock
prices. Figure 5.2 shows the pattern of the annual average yields for the sample firms
between 1990 and 2002.
However, there could be some bias when using the average dividend yield for
monitoring the pattern of dividend yields of common stocks in Thailand. The dividend
yield is the annual cash dividend in time t divided by the rate of market capitalisation in
Thai baht (SET 2003). This is not a standardised tool because firms could have a
different number of common stocks outstanding and stock prices will make the
dividend yield unstable. The Weighted Average Dividend Yield could be more
accurate because it is the dividend yield multiplied by the percentage of market
capitalisation rate of a firm to total market capitalisation and divided by 100. Figure 5.2
shows trends in the weighted average yield of dividends.
254
Chapter 5: Analysis of Data and Findings
Figure 5.2: Annual Average Dividend Yields and the Pattern in the Yield of the
Weighted Average Dividend
Average Dividend Yield and Weighted Average Dividend Yield
Average Dividend Yield
Weighted Dividend Yield
18.00%
16.00%
14.00%
Percent
12.00%
10.00%
8.00%
6.00%
4.00%
2.00%
0.00%
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002
Year
Source: SET (1990, 1993, 1996, 1999, 2002a)
Figure 5.2 indicates that the yield of the weighted average dividend fluctuated less than
the yield of the arithmetic average dividend, especially, during the Asian Economic
Crisis. It is also observed that the market dividend yield for the average dividend yield
and the weighted dividend yield rarely exceeded 4 percent after 1990. Considering this,
it is interesting to explore the range of the dividend yield paid to shareholders. SET
noted there are four ranges for the yield of dividends (1) lower than 3 percent (2) 3 to 7
percent (3) 7 to 12 percent and (4) greater than 12 percent (SET 2003). Table 5.5
presents the percentage of firms that paid dividends in the four ranges between 1990
and 2002.
255
Chapter 5: Analysis of Data and Findings
Table 5.5: Annual Dividend Yield by Range
Year
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
<3%
40.91%
38.86%
46.09%
58.89%
47.31%
30.34%
22.55%
12.81%
24.55%
20.00%
7.69%
13.55%
15.08%
Dividend Yield (Annual)
3%-7% 7%-12%
43.18%
14.77%
49.74%
8.29%
46.52%
6.52%
35.57%
4.74%
41.22%
10.39%
47.59%
18.97%
36.60%
29.41%
21.35%
22.78%
39.09%
18.18%
32.17%
34.78%
33.85%
36.92%
42.58%
32.90%
55.87%
23.46%
>12%
1.14%
3.11%
0.87%
0.79%
1.08%
3.10%
11.44%
43.06%
18.18%
13.04%
21.54%
10.97%
5.59%
Source: SET (1990, 1993, 1996, 1999, 2002a)
However, this method which separates the sample firms into four groups based on the
yield on dividends does not reveal a clear trend in dividend yields in the market.
Therefore, by taking the mid point of the four ranges, table 5.6 presents only two ranges
of dividend yields and the percentages of payers according to their new and wider
ranges (lower than 7 percent and higher than 7 percent). The results are illustrated in
figure 5.3.
Table 5.6: Percentage of Payers in Two Ranges of Dividend Yield
Year
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
Dividend Yield (Annual)
<7%
>7%
84.09%
15.91%
88.60%
11.40%
92.61%
7.39%
94.47%
5.53%
88.53%
11.47%
77.93%
22.07%
59.15%
40.85%
34.16%
65.84%
63.64%
36.36%
52.17%
47.83%
41.54%
58.46%
56.13%
43.87%
70.95%
29.05%
Source: SET (1990, 1993, 1996, 1999, 2002a)
256
Chapter 5: Analysis of Data and Findings
Figure 5.3: Trends in the Dividend Yield 1990-2002
Percent of Dividend Payers with reported dividend yield
<7%
Percent of Payers
>7%
100.00%
90.00%
80.00%
70.00%
60.00%
50.00%
40.00%
30.00%
20.00%
10.00%
0.00%
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002
Year
Source: Developed for this research
Figure 5.3 shows that more than 80 percent of the dividends of listed firm’s yield less
than 7 percent before the crisis. However, during the crisis, the percentage of payers
with high yield dividends increased sharply to 65.85 percent and this figure fluctuated
strongly throughout the remainder of the study period. However, the percentage of
firms with low yield dividends has increased sharply since 2000. In 2002, over 70
percent of firms paid dividends with a yield of less than 7 percent.
Low yields may result from the following: (1) the payment of high dividends by large
and mature firms could appear small when compared to their stock price, therefore,
dividend yields could be small or (2) listed firms really make small dividend payments
which are unrelated to the size of the firm.
As discussed earlier, the dividend yield is related to market capitalisation, or change in
the price of stocks. The unstable pattern of dividends could be explained by observing
the percentage change in stock prices during the period of investigation.
257
Chapter 5: Analysis of Data and Findings
Table 5.7: Average Dividend Yield, Average Weighted Dividend Yield, Average
Percentage Changes in Security Prices and Percentage of Market
Capitalisation of Payers to Total Firms.
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
Average
Dividend
Yield
3.69%
3.72%
3.26%
2.73%
3.26%
4.40%
5.73%
16.64%
2.00%
2.22%
3.17%
2.97%
2.78%
Average
Weighted
Dividend
Yield
3.53%
3.32%
2.45%
1.32%
1.54%
1.78%
2.81%
4.89%
1.86%
0.53%
1.53%
1.92%
2.50%
Average
Percentage
Changes in
Security
Prices
-26.01%
-4.44%
8.15%
25.47%
-24.37%
-19.71%
-24.39%
-44.25%
6.94%
51.27%
-15.33%
33.72%
27.31%
Percentage of
Market Cap.
Of Payers to
Total Firms
89.78%
91.00%
92.89%
79.74%
81.39%
85.46%
84.37%
88.02%
53.37%
13.68%
28.35%
43.71%
62.33%
Source: SET (1990, 1993, 1996, 1999, 2002a)
With a similar number of shares outstanding, the lower the stock price, the higher the
dividend yield. Therefore, the high payment of dividends in 1997 could be a direct
result of low stock prices due to the Asian Economic Crisis. This view is consistent
with the trend, which is shown in table 5.7. The table indicates that the average
percentage change in security prices was largest in 1997 (-44.25 percent). Therefore, in
1997 the average dividend yield was 16.64 percent and the average weighted dividend
yield was 4.89% (figure 5.2) and these figures were the highest recorded during the
study period.
The Asian economic crisis influenced the dividend yields by making the figures
unstable over time but how the crisis influenced the characteristics, the propensity to
pay dividends of payers and the phenomenon of disappearing dividends is to be
investigated in the later sections.
258
Chapter 5: Analysis of Data and Findings
This section compares trends in the payment of dividends and dividend. This analysis
explores the extend at which dividends are disappearing from the capital market of
Thailand. The study aims to explain changes in the pattern of dividend payments
before, during and after the Asian Economic Crisis. The changes in the characteristics
of listed firms and the propensity to pay dividends are explored in the following
analysis.
5.4.2
DIVIDEND PATTERN
Figure 5.4 shows the total number of listed firms between 1990 and 2002 that (1) pay
dividends (2) do not pay dividends (3) formerly paid dividends and (4) have never paid
dividends. Figure 5.5 shows the percentage of the sample of listed firms which are
presented in each of the four dividend groups.
Figure 5.4: The Number of SET Firms in Each Group
Number of SET firms in different dividend groups
All firms
Payers
Non-Payer
Former Payers
Never-Paid
Number of SET firms
450
400
350
300
250
200
150
100
50
0
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002
Year
Source: Developed for this research
259
Chapter 5: Analysis of Data and Findings
Figure 5.5: Percentage of All SET Firms in Different Dividend Groups
Percent of all SET firms in different dividend groups
Percent
Payers
Non-Payer
Former-Payer
Never-Paid
100
90
80
70
60
50
40
30
20
10
0
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002
Year
Source: Developed for this research
Figure 5.4 indicates that the number of listed firms increased sharply from 200 firms in
1990 to 400 firms in 1996 (pre-crisis). Table 5.1 indicates that the number of listed
firms in the sample declined to below 400 immediately after the crisis. This figure has
recovered slightly and stood at 397 firms in 2002. However, current number of listed
firms is still below the figure for 1996 and 1997. The number of firms paying dividends
increased from almost 200 in 1990 to greater than 300 firms in 1996. However, the rate
of increase in the number of payers has declined. The number of firms which pay
dividends has partly recovered although it was still below 200 firms in 2002.
There are four possible explanations for the decline in the number of firms paying
dividends in the short-term and the long-term. Firstly, existing and new listed firms
exhibit a lower propensity to pay dividends, or are less willing to pay dividends.
Secondly, firms have recently listed with the characteristics of non-payers tend to avoid
paying dividends. Thirdly, the characteristics of mature firms appear to have changed
and many of these firms have ceased paying dividends. Fourthly, many firms have been
distressed since the crisis (both new lists and existing firms) and ceased paying
dividends. This will be discussed more in detail later in the study.
260
Chapter 5: Analysis of Data and Findings
The number of firms which did not pay dividends before the crisis (the figure was less
than 100) was consistently less than the number of payers. However, the number of
firms which did not pay dividends rose sharply from 112 in 1997, to a peak of 280 in
1998. This figure has declined slowly since 1998, although it remained slightly above
200 in 2002. Fama and French (2001) suggested that firms which do not pay dividends
can be classified into the two groups, former payers and firms which have never paid
dividends. Former payers are firms that do not pay dividends in the present year (time
t) but paid dividends in a previous year. Figure 5.4 indicates that the number of former
payers (before and after the crisis) is similar to the number of non-payers. In addition,
the number of firms which have never paid dividends was relatively small in 1990. The
number of non-payers and former payers increased sharply in 1997 and peaked in 1998
(see figure 5.4). This increase in the number of non-payers and former payers can be
attributed to the Asian Economic Crisis. Indeed, a strong decline in the number of firms
which pay dividends also appears to result from the crisis.
The number of non-payers declined from about 300 in 1998 to about 200 in 2002. In
addition, the number of former payers declined from 250 in 1998 to about 150 in 2002.
A trend that has emerged since the crisis is that the number of firms that have never
paid dividends accounts for an increasingly larger share of the non-payer group (see
figure 5.4). It appears that (1) firms that were listed between 1997 and 2002 are more
likely to have never paid dividends and (2) most of Thailand’s listed firms tend to pay
dividends if they are capable of doing so. Therefore, most of the non-payers (former
payers are the majority of this group) were forced to cease paying dividends due to
financial distress flowing from the crisis. This view is consistent with the findings of
Fama and French (2001). Former payers are likely to be distressed and have not paid
dividends in time t. However, it remains to be seen whether non-payers and former
payers will pay dividends in the future. This issue will be addressed later in this section.
As stated earlier, the percentage of firms which paid dividends decreased from 84.2
percent in 1990 to 46.4 percent in 2002. Before the Asian Economic Crisis (19901996), the percentage of payers declined slightly from 84.2 percent in 1990 to 81.0
261
Chapter 5: Analysis of Data and Findings
percent in 1996. However, the percentage of payers fell sharply from 71.5 percent to
28.4 percent during 1997 and 1998. In 1999, the percentage of payers increased slightly
to 29.9 percent. Indeed, the percentage of payers has continued to rise and stood at 46.4
percent in 2002. This figure is about 20 percent higher than the figure recorded during
the crisis, although it remains about 40 percent lower than 1990. The trends in the
percentage of payers indicate (1) in the long-term, the payment of dividends appears to
be disappearing because the percentage of listed payers has declined, and (2) in the
short-term (the pre-crisis period), the percentage of payers declined slightly. This
outcome may be an early indicator of the financial collapse which occurred in 1997,
and (3) in the short-term (post-crisis period), the percentage of firms paying dividends
has rebounded from the low point recorded during the crisis and may rise to pre-crisis
levels in the future.
The decline in the percentage of dividend payers may also result from growth in the
non-payer group due to new listings. Before the crisis, the number of firms in most
sectors expanded by more than 10 percent per annum due to new listings (see table 5.1).
However the percentage of new listings declined to less than one percent of the total
number of listed firms during the crisis and currently stands at about 5.2 percent. Table
5.1 indicates that some newly listed firms paid dividends, although the percentage of
these firms fluctuated between 1990 and 2002. The percentage of new firms that paid
dividends stood at 82.2 percent (1991), 73.3 percent (1992), 51.5 percent (1993), 52.6
percent (1994), 82.6 percent (1995), 73.5 percent (1996) and 50 percent in 1997. After
the crisis, less than 30 percent of new lists paid dividends. In addition, the percentage of
new firms that paid dividends was lower than the total percentage of payers in all years
with the exception of 1998 and 2001 (see table 5.1). Therefore, it appears that new
firms are less likely to pay dividends and this partly explains the decline in the
percentage of payers between 1990 and 2002.
Fama and French (2001) found that the percentage of listed firms which paid dividends
decreased sharply from almost 70 percent in 1978 (a period of recession) to 20.8
percent in 1999. As stated earlier, Fama and French (2001) averaged the figures for the
period preceding 1978 and explored the decline in the payment of dividends after 1978.
262
Chapter 5: Analysis of Data and Findings
However, Thailand’s capital market is relatively new and has only experienced one
major economic downturn (1997-1998). As stated earlier the percentage of listed payers
in Thailand’s market has recovered slightly since the crisis. The percentage of firms
paying dividends in the United States has continued to decline.
Fama and French (2001) did not indicate whether there was any change in the
percentage of payers in the US market after 1999. These researchers presented their
data over several time periods and did not investigate trends from one year to the next.
Therefore, more research is needed to determine whether the decline in the number of
payers in the US has been sustained. This study presents the pattern of dividends year
by year from 1990 to 2002 and in several time periods (range of years). This analysis is
similar to Fama and French (2001). This study found that the percentage of payers in
Thailand’s capital market increased from 20.8 percent in 1999 to 50.4 percent in 2002
(see figure 5.5).
Table 5.8 presents the dividend behaviour of payers in year t-1 and non-payers in year
t-1 undertake in time t. The table is divided into three parts. The first part presents
behaviour in year t of firms that paid dividends during the previous year. These firms
continue to pay dividends, ceased paying or were forced to delist. The second part
presents the behaviour of firms that were non-payers in time t-1. These firms
commenced paying dividends, remained a non-payer, or were delisted from the stock
exchange. The last part shows the percentage of non-payers that commenced paying
dividends in time t. The table also indicates if these firms were former payers, or if they
have never paid dividends.
263
Chapter 5: Analysis of Data and Findings
Table 5.8: What Happens in Year t to SET Firms that Do and Do not Pay
Dividends in Year t-1?
What Happens in Year t to Firms that Pay Dividends in Year t-1
1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002
78.4 88.1 93.0 93.5 91.3 91.4 81.7 38.4 75.5 89.6 92.3 95.4
Continue to Pay
13.1
6.7
5.7
6.1
6.2
6.6 17.0 59.4 23.6
5.2
6.9
3.9
Stop Paying
0.6
0.5
0.0
0.0
0.4
0.3
0.0
3.0
0.0
3.5
0.0
0.0
Delist
What Happens in Year t to Firms that Do Not Pay Dividends in Year t-1
54.5 91.4 45.8 55.8 35.8 16.7 24.3
Start Paying
12.1
2.9 54.2 44.2 62.3 83.3 64.9
Do not Pay
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Delist
0.9
94.6
0.9
11.2
84.2
1.4
8.9
83.3
4.1
11.7
84.9
1.3
12.1
82.5
2.2
Percent of Non-Payers in Year t-1 that Start Paying in Year t-1
54.5 91.4 45.8 55.8 35.8
All Non-Payers (t-1)
63.6 11.4 29.2
7.0 15.1
Former Payers
33.3 20.0 70.8 37.2
1.9
Never Paid
0.9
27.7
0.0
11.2
8.6
0.0
8.9
10.0
0.4
11.7
9.6
1.7
12.1
0.0
0.0
16.7
20.4
13.0
24.3
1.4
0.0
Source: SET (1994, 1997, 2000, 2002a)
Note: Firms that continue to pay were payers in year t-1 and continue to pay dividends in time t. Firms
that cease paying dividends are firms that were payers in time t-1, but non-payers in time t. Firms that
delist are firms that were payers in time t-1 but delisted in time t. Numbers shown in percentage.
Table 5.8 shows the likelihood that payers and non-payers of the last period (t-1) will
continue to pay, commence or cease paying dividends in this year. Firms which paid
dividends last year (t-1) tend to continue paying dividends in year t. Before the crisis,
almost 90 percent of the payers continued to pay dividends (table 5.8). In 1997,
however, the percentage of payers in the previous year (t-1) that continued to pay (in t)
fell to 81.7 percent and to 38.4 percent during 1998. After the crisis, the percentage of
payers in the previous period, that continued to pay in the next period, increased to 75.5
percent in 1999, 89.6 percent in 2000, 92.3 percent in 2001 and finally 95.4 percent in
2002. Although the number of firms that continue to pay dividends has increased to
more than 90 percent, it should be stated that the figures after the crisis, are compared
with the lowest percentage of payers that continued to pay dividends in 1998 (38.4
percent). Therefore, after 1998, the denominator (the total number of payers in the last
period) is less than the figure observed in the pre-crisis period. The question which
arises is what has occurred to the remaining 60 percent of firms which are firms that
ceased paying dividends during the crisis?
264
Chapter 5: Analysis of Data and Findings
Firms which paid dividends before the crisis, but did not pay during the crisis, ceased
paying dividends at the rate of 59.4 percent and delisted at the rate of 3 percent in 1998.
Before the crisis, the percentage of payers that ceased paying dividends was about 10
percent per annum. However, this figure rose to 17 percent in 1997 and peaked at 59.4
percent in 1998. In addition, the percentage of payers that delisted in the next period (t)
was less than one percent per annum before the crisis, but rose to 3 percent in 1998.
Finally, after the crisis, a greater proportion of firms which paid dividends continued
doing so while less ceased paying and delisted. This indicates that the market is
recovering.
Firms that did not pay dividends in the last period commenced paying at the rate of 91.4
percent in 1991 and 45.8 percent in 1992. This indicates that the vast majority of firms
possessed a high propensity to pay, or resumed paying dividends at this time. However,
this figure fell to 35.8 percent in 1994 and to less than one percent in 1998. Finally,
after the crisis, this figure increased to above 10 percent, although, it is still lower than
the percentage of non-payers that commenced paying dividends before the crisis. In
summary, less firms re-commenced paying dividends after the crisis.
Before the crisis, non-payers appeared less willing to commence paying dividends in
the next period (decreasing their propensity to pay dividends). The percentage of nonpayers that continued to avoid this payment in the next period also supports this finding
(see table 5.8). After 1993, a large proportion of non-payers continued to avoid paying
dividends, even though it was still several years before the crisis. In addition, only 0.9
percent of non-payers commenced the payment of dividends in 1998. This figure
increased to 12.1 percent in 2002, indicating that the market was recovering and a high
propensity to pay dividends by former non-payers.
The last part of table 5.8 indicates that, former payers resumed the payment of
dividends more frequently than firms that have never paid dividends. This view
supports the earlier discovery that former payers omit the payment of dividends but do
not intend to cease the payment on a permanent basis. The descriptive statistical
analysis of Fama and French (2001) suggests that former payers are distressed firms
265
Chapter 5: Analysis of Data and Findings
which ceased paying dividends to preserve their cash. The relationship between former
payers and distress will be analysed for Thailand with the aid of a variety of ratios. This
analysis is presented in section 5.5.
A preliminary analysis of the descriptive statistics reveals that the percentage of firms
which paid dividends started to decline before the crisis, and fell sharply during the
crisis. The percentage of payers rose slightly after the crisis, but remains much lower
than the level recorded before the crisis. It also appears that payers were less likely to
continue the payment of dividend during the crisis. The analysis suggests that the
propensity to pay dividends of existing payers is lower. However, this suggestion will
be confirmed in the later sections of the chapter.
As stated earlier, the percentage of non-payers has increased over time and increased
sharply immediately after the crisis. Most of the non-payers were former payers which
were forced to avoid the payment of dividends due to the crisis. In addition, it appears
that non-payers were less willing to re-commence paying dividends. This supports the
proposed explanation for the observed decline in the percentage of payers; non-payers
are less likely to re-commence paying dividends. Consequently, the total number of
payers is smaller and represent a smaller proportion of the total number of firms.
Many new firms were listed at SET before the crisis and large proportion of these firms
did not pay dividends. The percentage of new firms that paid dividends was less than
the percentage of listed firms that paid dividends, in most years (see table 5.1).
Therefore, new lists usually do not pay dividends and the denominator (total number of
listed firms) is increasing over time, while the numerator (the total number of payers) is
becoming smaller leading to a decline in the percentage of payers.
The findings of this section have raised several questions which require further
investigation. (1) What happened to listed firms in 1997 to 1998? Why do payers tend
to cease paying dividends? What are the characteristics of firms which pay dividends?
Has the propensity to pay dividends fallen? Have characteristics of these firms
changed? (2) What are new lists’ characteristics? Are the characteristics of new lists
266
Chapter 5: Analysis of Data and Findings
different from mature firms? Is the decline in percentage of dividend payers due to new
listing? Section 5.5 presents the descriptive statistics on the characteristics of the
dividend groups and provides answers to these questions.
5.5
CHARACTERISTICS OF THE DIVIDEND PAYERS
The preliminary discussion of the characteristics of dividend payers, non-payers and
other dividend found that payers and non-payers differ in terms of (1) profitability (2)
investment opportunities and (3) size. Section 5.5.2 and 5.5.4 will confirm the findings
of the descriptive statistics with the aid of ANOVA (An Analysis of Variance) and logit
regression. As stated earlier, Fama and French (2001) suggested there are three
characteristics of firms which pay dividends (1) profitability (2) investment
opportunities and (3) size. Each characteristic will be measured and compared with the
aid of a variety of ratios. The analysis will compare the changes in these characteristics
during the pre-crisis and post-crisis periods by using descriptive statistics. In addition,
each characteristic will be compared with the findings of the literature. Finally, this
analysis will test differences in the mean of these characteristics to determine whether
they are significantly different.
5.5.1 FIRM CHARACTERISTICS DESCRIPTIVE STATISTICS APPROACH
5.5.1.1 PROFITABILITY
Table 5.9 presents the characteristics of firms in a variety of dividend groups. The
profitability of the dividend groups is expressed by two ratios ‘(1) Et/At (the ratio of
aggregate earnings before interest to aggregate assets) (2) Yt/BEt (the ratio of aggregate
common stock earnings over aggregate book equity)’ (Fama and French 2001, p. 7).
The discussion is divided into two parts, in line with the two ratios for measuring
profitability.
267
Chapter 5: Analysis of Data and Findings
Table 5.9: Average Values of Six Characteristics Ratios
19912002
19911996
19972002
1991
1992
1993
1994
1995
1996
6.16
Et/At (percent): The Ratio of Aggregate Earnings Before Interest to Aggregate Assets
All Firms
3.64
8.71
-1.43
11.16
10.14
8.96
8.07
7.78
Payers
6.68
9.47
3.89
10.97
10.59
9.88
8.97
8.84
7.53
-0.45
4.35
-5.24
10.75
5.84
3.73
3.37
2.08
0.31
Non-Payers
Never Paid
Former Payers
0.48
6.22
-5.25
11.41
8.66
6.33
7.18
2.78
0.97
-0.96
3.25
-5.16
12.63
3.45
0.75
0.67
1.92
0.08
All New Lists
4.62
11.77
-2.53
11.83
10.86
12.07
20.12
10.06
5.67
Payers
9.95
11.19
8.71
11.62
11.3
13.9
12.41
11.6
6.33
0.4
7.95
-7.14
12.82
9.63
10.13
8.57
2.74
3.84
19912002
19911996
19972002
1991
1992
1993
1994
1995
1996
Non-Payers
Yt/BEt (percent): The Ratio of Aggregate Common Stock Earnings Over Aggregate Book Equity
-108.62
8.65
-225.89
16.64
14.44
13.01
-8.42
10.57
5.64
-9.21
13.21
-31.63
16.67
16.26
12.79
13.17
11.92
8.44
Non-Payers
-164.63
-15.91
-332.83
16.48
-3.01
14.28
-3.51
3.31
-6.1
Never Paid
-43.23
5.85
-92.31
14.64
7.88
6.04
5.58
-11.52
12.51
0.08
All Firms
Payers
Former Payers
-0.96
3.25
-5.16
12.63
3.45
0.75
0.67
1.92
All New Lists
-10.1
14.63
-34.83
18
17.75
13.1
25.23
11.15
2.56
10.5
14.07
6.92
17.96
20.63
12.54
16.11
15.07
2.14
-18.34
8
-44.67
18.19
9.72
13.69
10.13
-7.48
3.72
19912002
19911996
19972002
1991
1992
1993
1994
1995
1996
Payers
Non-Payers
dAt/At (percent): The Growth Rate in Assets
All Firms
-8.98
23.15
-41.1
29.09
30.58
23.52
23.78
19.25
12.65
Payers
11.95
22.38
1.51
31.75
28.66
19.22
21.07
20.45
13.16
Non-Payers
-18.49
28.72
-65.7
14.44
48.98
48.05
37.86
12.8
10.22
Never Paid
47.6
69.02
26.18
63.29
95.88
72.83
88.28
46.75
47.11
-39.78
3.74
-83.29
-11.05
9.29
19.55
2.08
5.09
-2.53
All New Lists
96.39
113.05
79.74
100
97.87
97.13
186.24
100
97.06
Payers
87.57
98.89
76.25
100
97.11
100
96.24
100
100
Non-Payers
87.57
97.16
77.97
100
100
94.07
100
100
88.89
19912002
19911996
19972002
1991
1992
1993
1994
1995
1996
1.18
Former Payers
Vt/At: The Ratio of the Aggregate Market Value to the Aggregate Book Value of Assets
All Firms
1.77
2.15
1.39
4.35
1.92
2.42
1.67
1.38
Payers
1.55
2.09
1
4.02
1.9
2.41
1.65
1.39
1.18
Non-Payers
2.05
2.5
1.6
6.16
2.08
2.46
1.81
1.36
1.11
Never Paid
1.85
2.52
1.18
4.02
2.36
2.9
2.27
2.05
1.51
Former Payers
2.06
2.46
1.66
7.28
1.85
1.97
1.48
1.2
0.98
All New Lists
1.27
2.09
0.46
3.94
1.89
1.94
2.27
1.6
0.89
Payers
1.95
2.75
1.15
4.79
2.56
3.77
2.27
1.93
1.21
Non-Payers
2.01
2.69
1.33
3.29
2.48
3.68
2.4
2.35
1.92
268
Chapter 5: Analysis of Data and Findings
Table 5.9: Average Value of Six Characteristic Ratios (Continued)
19912002
19911996
19972002
1991
1992
1993
1994
1995
1996
All Firms
18107.71
14816.16
21399.25
12497.04
12378.22
13372.01
14835.83
17179.76
18634.12
Payers
16160.28
16593.41
15727.15
14185.74
13086.91
14633.33
16802.84
19275.35
21576.29
Non-Payers
14819.71
5246.12
24393.3
3185.04
5586.69
6185.43
4629.64
5925.68
5964.26
Never Paid
13830.14
8507.63
19152.64
1250.8
2546.08
4359.1
7971.55
19891.7
15026.57
Former Payers
15355.32
4747.11
25963.53
4194.21
8159.52
8285.72
2257.97
2751.59
2833.64
6936.83
5440.35
8433.31
3041.72
2628.18
3226.89
8066.69
10383.35
5295.26
3196.8
3987.73
2405.87
3335.83
2588.8
1572.31
2273.34
8989.73
5166.41
7869.11
13987.61
1750.6
15428.2
7248.63
1670.57
2525.93
42701.22
14351.13
19912002
19911996
1991
1992
1993
1994
1995
1996
At: Total Assets
All New Lists
Payers
Non-Payers
19972002
Lt/At: The Ratio of Total Liabilities to Total Assets
All Firms
0.73
0.53
0.93
0.53
0.51
0.55
0.5
0.53
0.56
Payers
0.45
0.52
0.39
0.5
0.5
0.55
0.5
0.52
0.55
Non-Payers
0.92
0.58
1.25
0.67
0.57
0.54
0.53
0.6
0.58
Never Paid
0.63
0.45
0.81
0.29
0.54
0.39
0.41
0.57
0.49
1
0.67
1.33
0.86
0.59
0.72
0.61
0.61
0.61
All New Lists
0.62
0.53
0.72
0.44
0.53
0.57
0.67
0.47
0.49
Payers
0.41
0.51
0.31
0.46
0.53
0.72
0.33
0.45
0.54
Non-Payers
0.61
0.43
0.8
0.32
0.55
0.42
0.37
0.56
0.36
Former Payers
Source: Developed for this research
Note: At is total asset, BEt is the book value for common equity, MEt is the market value of common
equity, Lt is the book liabilities, Vt is the market value of the entire firm, Et is the earnings before interest
but after taxes, Yt is the after-tax earnings to common stock and dAt is the changes in the assets from
time t and t-1. Results for ranges of years are grouped and averaged according to the dividend groups.
Et/At, Yt/Bet, dAt/At are shown in percentages. Vt/At and Lt/At are shown in ratio values. At is shown
in Baht values.
269
Chapter 5: Analysis of Data and Findings
Table 5.9: Average Values of Six Characteristic Ratios (Continued)
1997
1998
1999
2000
2001
2002
Et/At (percent): The Ratio of Aggregate Earnings Before Interest to Aggregate Assets
All Firms
-9.43
4.11
-3.25
-3.92
-1.55
Payers
-5.36
9.56
8.84
8.5
-7.28
9.08
Non-Payers
-19.64
2.52
-8.4
-10.66
2.38
2.34
Never Paid
-22.57
1.8
-6.53
-5.75
-1.27
2.8
Former Payers
-18.92
2.6
-8.62
-11.39
3.17
2.18
-9.72
-1.11
-19.39
-7.39
12.51
9.9
-10.18
19.99
0
7.17
20.72
14.56
-9.26
-11.67
-19.39
-12.85
1.57
8.73
1997
1998
1999
2000
2001
2002
All New Lists
Payers
Non-Payers
5.47
Yt/BEt (percent): The Ratio of Aggregate Common Stock Earnings Over Aggregate Book Equity
All Firms
-289.65
-706.33
-63.82
-43.08
-259.24
6.76
Payers
-246.84
10.54
10.87
10.92
12
12.71
Non-Payers
-397.05
-988.27
-95.63
-72.28
-445.34
1.61
Never Paid
-61.75
-76.07
-354.55
-44.79
-10.46
-6.27
Former Payers
-18.92
2.6
-8.62
-11.39
3.17
2.18
All New Lists
-52.13
-9.7
-147.88
-30.99
18.11
13.65
Payers
-42.88
21.88
0
7.71
30.79
24.04
Non-Payers
-61.39
-25.5
-147.88
-45.51
1.2
11.05
1997
1998
1999
2000
2001
2002
-2.98
dAt/At (percent): The Growth Rate in Assets
-1.23
-88.14
-5.78
-19.64
-128.86
0.14
-2.36
5.43
3.14
-2.02
4.73
Non-Payers
-4.66
-121.49
-10.56
-31.96
-215.88
-9.64
Never Paid
65.87
10.85
11.58
13.64
25.89
29.27
Former Payers
-21.9
-134.57
-13.12
-38.75
-268.72
-22.7
All New Lists
100
54.16
49.82
90.38
88.6
95.49
Payers
100
100
0
100
80.05
77.47
Non-Payers
100
31.24
49.82
86.77
100
100
1997
1998
1999
2000
2001
2002
All Firms
Payers
Vt/At: The Ratio of the Aggregate Market Value to the Aggregate Book Value of Assets
All Firms
1.04
1.88
1.26
1.27
1.59
Payers
1.05
1.01
1
0.87
0.99
1.34
1.11
Non-Payers
1.02
2.22
1.36
1.48
1.99
1.54
Never Paid
0.99
0.93
1.27
1.16
1.27
1.48
Former Payers
1.02
2.34
1.37
1.52
2.15
1.57
All New Lists
0.76
0.45
0
0.23
1.08
0.27
Payers
1.52
1.34
0
0.83
1.89
1.33
Non-Payers
0.81
1.2
1.43
1.41
1.46
1.66
270
Chapter 5: Analysis of Data and Findings
Table 5.9: Average Values of Six Characteristic Ratios (Continued)
1997
1998
1999
2000
2001
2002
22974.21
At: Total Assets
All Firms
20729.08
20147.08
21303.15
22159.39
21082.6
Payers
25384.91
47848.21
3034.98
3893.16
5289.25
8912.41
Non-Payers
9047.92
9173.24
29084.04
32002.28
31918.4
35133.93
Never Paid
17754.73
17074.25
26090.25
23752.6
18282.7
11961.32
Former Payers
6919.59
8392.5
29430.43
33231.79
34898.88
42907.97
All New Lists
2272.65
7479.3
27672.49
6902
2670.88
3602.55
Payers
2772.81
1192.13
0
1436.5
4003.96
5029.8
Non-Payers
2772.81
596.07
0
538.69
5338.61
1257.45
1997
1998
1999
2000
2001
2002
Lt/At: The Ratio of Total Liabilities to Total Assets
All Firms
0.68
1.34
0.81
0.93
1.11
0.71
Payers
0.63
0.39
0.34
0.32
0.33
0.33
Non-Payers
0.82
1.72
1
1.26
1.65
1.05
Never Paid
0.82
0.75
0.84
0.92
0.86
0.71
Former Payers
0.82
1.81
1.02
1.31
1.82
1.16
All New Lists
0.68
0.8
1.29
0.8
0.36
0.37
Payers
0.65
0.09
0
0.29
0.4
0.44
0.7
1.15
1.29
0.99
0.31
0.35
Non-Payers
Source: Developed for this research
Note: At is total asset, BEt is the book value for common equity, MEt is the market value of common
equity, Lt is the book liabilities, Vt is the market value of the entire firm, Et is the earnings before interest
but after taxes, Yt is the after-tax earnings to common stock and dAt is the changes in the assets from
time t and t-1. Results for ranges of years are grouped and averaged according to the dividend groups.
Et/At, Yt/Bet, dAt/At are shown in percentages. Vt/At and Lt/At are shown in ratio values. At is shown
in Baht values.
271
Chapter 5: Analysis of Data and Findings
The ratio of aggregate earnings before interest to aggregate assets (Et/At)
Before the Asian Economic Crisis (1991-1996) payers, on average, have higher
profitability (9.47 percent) than non-payers (4.35 percent), firms which have never paid
dividends (6.22 percent) and former payers (3.25 percent). All samples before the crisis
indicate that firms which pay dividends are the most profitable (8.71 percent). This
result is consistent with the findings of Fama and French (2001). Therefore, when using
the Et/At ratio for measuring profitability, payers have higher profitability than nonpayers. When comparing the two groups of non-payers, firms that have never paid
dividends were more profitable (6.22 percent) than former payers (3.25 percent).
As explained in chapter 2, during the Asian Economic Crisis almost 70 percent of firms
listed in Thailand suffered large losses (SET 2002a). Therefore, payers were less
profitable during and after the crisis (1997-2002) although they remained the most
profitable (3.89 percent) of the four dividend groups. While payers experienced positive
profitability, non-payers experienced losses (-5.24 percent). Within the non-payer
group, firms that have never paid dividends suffered slightly higher losses (-5.25
percent) than former payers (-5.16 percent). The average profitability (Et/At) of all
samples after the crisis was -1.43 percent. This was lower than the profitability of
dividend payers, but higher than the figure recorded for non-payers and their subgroups. This finding is consistent with the literature (Fama and French 2001).
During the period 1991 to 2001, payers of dividends were the most profitable (6.68
percent) followed by firms that have never paid dividends (0.48 percent). Non-payers of
dividends were more profitable than former payers. Non-payers experienced losses of
-0.45 percent while former payers experienced losses of -0.96 percent. The all firm
average level of profitability was 3.64 percent.
Therefore, the Et/At ratio indicates that payers are more profitable than non-payers
during favourable and poor economic conditions. Former payers tend to be distressed
171
Chapter 5: Analysis of Data and Findings
because they are the least profitable in good times and experience large losses in bad
times. Hence, these firms often cease paying dividends. The descriptive statistics have
shown that when non-payers are divided into two sub-groups (never paid and former
payers), firms which have never paid dividends perform better than former payers. This
view is consistent with the findings of La Porta et al. (2000), Benito and Young (2001),
DeAngelo, DeAngelo and Skinner (2001), Fama and French (2001), and Aivazian,
Booth and Cleary (2003).
Generally, firms that were newly listed on the market before the crisis were highly
profitable (11.7 percent) and more profitable than the average of the existing listed
firms (see table 5.9). The profitability of the new lists that paid dividends was also high
(11.19 percent), while the profitability of new lists that did not pay dividends was lower
(7.95 percent). All figures for profitability measured by Et/At were higher for new lists
than for existing listed firms before the crisis period.
After the crisis, the losses of the new lists increased slightly (-2.03) percent, when
compared to the figure recorded for the existing listed firms (table 5.5). However, new
lists that paid dividends were highly profitable (8.71 percent) when compared to all
existing listed firms that pay dividends.
For the whole period of investigation, new lists were relatively profitable (4.62 percent)
and generally more profitable than existing firms during the same period. New lists that
paid dividends were highly profitable (9.95 percent) and this figure is higher than
existing firms which paid dividends.
172
Chapter 5: Analysis of Data and Findings
Fama and French (2001) found that new lists tended to be highly profitable, early on,
although their profitability declined after the crisis in 1978 (Fama and French 2001).
Therefore, Fama and French (2001) concluded that the decline in the percentage of
payers after 1978 is partly due to the new characteristics of the new lists (low
profitability, high investment opportunities and small size). However, in the case of
Thailand, a clear trend in the disappearing of dividends is not evident. It appears that
Thai firms have resumed paying dividends after a crisis. Therefore, new lists with
higher profitability than mature firms are more common in Thailand. There is no clear
explanation for this result although it appears that changes in the characteristics of new
lists and mature firms are very small and the Thai capital market has not reached its
turning point.
173
DBA Thesis: Malinee Ronapat
The ratio of aggregate common stock earnings over aggregate book equity
(Yt/BEt)
Fama and French (2001) suggested that Et or earnings before interest may not be the
relevant tool affecting firms’ dividend payment decisions. These researchers suggested
that Yt, or earnings available for common stocks could be a better tool. Table 5.9
indicates that the gap between the profitability of payers and non-payers tends to be
wider when measured by the Yt/BEt ratio. This view is consistent with the findings of
Fama and French (2001).
During the pre-crisis period dividend payers were more profitable (13.21 percent) than
the other dividend groups. When focusing on non-payers, firms that have never paid
dividends were more profitable (5.85 percent) than former payers (3.25 percent). The
average level of profitability for all firms was 8.65 percent, which is lower than that
recorded for payers, but higher than the other dividend groups. This finding is similar to
the findings of Fama and French (2001).
During and after the Asian Economic Crisis, all firms, on average, experienced losses
of -225.89 percent. The largest losses were recorded in 1997 and 1998 (table 5.9).
Furthermore, when analysing each dividend group, the data indicates that every
dividend group experienced losses.
In relation to the period 1991 to 2002, the average Yt/BEt ratio for payers was -9.21
percent, while firms which have never paid dividends experienced greater losses of
-43.23 percent. Non-payers and the sub-group of former payers, experienced very large
losses of -164.63 percent and -43.23 percent respectively. The average level of
profitability of all firms was -108.62 percent.
174
DBA Thesis: Malinee Ronapat
When using Yt/BEt as a measure of profitability, payers are more profitable than nonpayers during positive economic conditions. This finding is consistent with La Porta et
al. (2000), Benito and Young (2001), DeAngelo, DeAngelo and Skinner (2001), Fama
and French (2001), and Aivazian, Booth and Cleary (2003). During difficult times,
payers experienced smaller losses than non-payers.
The profitability of new lists was 14.63 percent before the crisis (using Yt/BEt). The
average profitability of new lists that paid dividends was 14.07 percent, which is much
higher than the level experienced by the existing firms which paid dividends. During
and after the crisis, new lists made losses of -34.83 percent. However, new lists that
paid dividends were highly profitable and this performance was superior to the results
achieved by listed firms that pay dividends.
For the whole period of investigation (1990-2002), new lists experienced losses of
-10.10 percent, although the payer groups were more profitable (10.5 percent) than the
returns achieved by existing listed firms that paid dividends (see table 5.9). These
findings are similar to Fama and French (2001). However, Fama and French (2001)
stated that the characteristics of newly listed firms changed from high to low
profitability, after a strong downturn in the US in 1978. It should be noted that Fama
and French (2001) concluded that this outcome is not consistent, over time, for new
lists that paid dividends and mature firms that paid dividends. Mature firms which pay
dividends are less profitable than new listed payers when using both ratios (Et/At and
Yt/BEt) in the US. This finding is similar to the results of this study on the Thai stock
market (see table 5.9). The study focuses on the characteristics of the dividend payers.
Therefore, it is important to highlight the issue that new lists which pay dividends are
more profitable than mature firms that do not pay dividends in Thailand. Furthermore,
the profitability of new lists has been consistently higher than mature firms over the
period of investigation.
175
DBA Thesis: Malinee Ronapat
5.5.1.2 INVESTMENT OPPORTUNITIES
The investment opportunities of firms are measured by the three ratios which were
recommended by Fama and French (2001) (1) dAt/At (growth rate in assets) (2) Vt/At
(the ratio of the aggregate market value to the aggregate book value of assets) (3)
RDt/At (research expenditure) in table 5.9. However, this study will apply the first two
ratios because data on research and development is not available in Thailand. This
information is not required by SEC and need not be presented on the financial
statements of listed firms (SET 2003b). The two ratios will be calculated and analysed
separately in this study.
The growth rate in assets (dAt/At)
Before the Asian Economic Crisis, the percentage rate of growth in assets for payers
was 22.38 percent, however, this was lower than that achieved by non-payers (28.72
percent). This finding is consistent with Fama and French (2001). These researchers
suggested that payers have poorer investment opportunities than non-payers, because
they pay a proportion of their earnings as dividends (Fama and French 2001). Former
payers and firms that have never paid dividends experienced growth in their assets of
3.74 percent and 69.02 percent respectively. The average rate of growth in assets of all
samples was 23.15 percent.
Firms that pay dividends out of their earnings have fewer opportunities for investing in
projects with good returns. Hence, firms which pay dividends have a low rate of growth
in assets (as a proxy of investment opportunities) than firms which retain their earnings
and invest in assets (Fama and French 2001).
When the Asian Economic Crisis commenced in 1997, every dividend group
experienced a negative rate of growth in their assets, with the exception of the payer
group (1.51 percent) and firms which have never paid dividends (26.18 percent). After
the crisis, these two groups continued to experience significant growth in their assets,
while the other groups continued to record negative growth in their assets. Negative
176
DBA Thesis: Malinee Ronapat
growth rates could be due to downsizing, or the liquidation of some of the assets owned
by non-payers and former payers.
During the period 1990 to 2002, the rate of growth of assets for payers was positive
(11.95 percent). The highest rate achieved among the four dividend groups was
achieved by firms that have never paid dividends (47.6 percent). The asset growth rate
of non-payers was -18.49 percent, while the former payers experienced losses of -39.78
percent. Firms that have never paid dividends recorded the highest growth in their
assets followed by the payers. This finding is consistent with the literature (Fama and
French 2001).
Firms that have never paid dividends and non-payers exhibit higher investment
opportunities than payers before and after the Asian Economic Crisis. The securities of
these firms are therefore likely to be real growth stocks. Payers maintained their asset
sizes at levels recorded prior to the crisis (table 5.9). After the crisis, payers continued
to have positive investment opportunities, while the other listed firms engaged in
downsizing. Former payers were distressed and experienced low investment
opportunities and growth. Overall, payers tend to have a lower rate of asset growth,
than firms that have never paid dividends (see table 5.9). However, payers experienced
a higher rate of growth of their assets than non-payers and former payers.
During 1991-1996, newly listed firms experienced high growth in investment
opportunities (113.05 percent) and the new lists which paid dividends experienced
lower asset growth than the new lists which did not pay dividends. Therefore, after
1991 new lists experienced high growth and tended to have similar characteristics to
non-payers. This result is consistent with Fama and French (2001).
The ratio of the aggregate market value to the aggregate book value of assets
(Vt/At)
During the pre-crisis period, investors perceived that the assets of firms which pay
dividends, should be valued highly (2.09 times), but lower than that of non-payers (2.05
177
DBA Thesis: Malinee Ronapat
times). This implies that (1) the market value of payers’ stocks could be undervalued or
(2) investors are willing to pay more for the stocks of firms that do not pay dividends.
An appreciation in stock prices is therefore more likely for firms that do not pay
dividends. The Vt/At ratio for never paid firms is 2.52 times and 2.46 times for former
payers. The average Vt/At ratio for all firms was 2.15 times. The payer group had the
lowest Vt/At ratio.
In the post-crisis period, the assets of payers valued more lowly (1.00 times) than other
dividend groups. The average Vt/At for all firms was 1.39 times. These results were
also consistent with Fama and French (2001).
Over the period 1990 to 2002, the Vt/At ratio of payers has averaged 1.77 times. At
present, payers have a lower Vt/At than non-payers and other dividend groups.
In summary, it is observed that payers experience lower growth than non-payers
because they tend to maintain the size of their assets during recessions, while nonpayers, never-paid and former-payers accounted for more than 50 percent of the firms
engaged in downsizing (BOT 1998). Over the period 1990 – 2002 payers had lower
investment opportunities than the other dividends groups. Former payers were
distressed firms with low profitability and low investment opportunities. This result is
also consistent with the findings of Fama and French (2001). Furthermore, Fama and
French (2001) suggested that in relation to investment opportunities (1) the
understatement of profitability (measured by Et/At) by growth firms, could take some
time to reach its maximum level and (2) profitability (measured by Et/At) of payers
could be overstated due to inflation (Fama and French 2001).
178
DBA Thesis: Malinee Ronapat
5.5.1.3 SIZE
The size of firms can be measured with two ratios (1) At (total assets) and (2) Lt/At (the
ratio of total liabilities to total assets) in (table 5.9). In line with Fama and French 2001,
each ratio will be analysed separately.
Total assets (At)
From 1991 to 1996, the total assets of payers was 16593.41 million Baht. This was
higher than the figure for non-payers’ (5246.12 Million Baht), never paid (8507.63
Million Baht) and former payers (4747.11 Million Baht). Therefore, the assets of payers
tend to be larger than the other dividend groups and the assets of former payers were
larger than non-payers and firms which have never paid dividends. This result is
consistent with the findings of Fama and French (2001).
After 1997, the total assets of payers were maintained at 15727.15 million Baht, while
the assets of non-payers increased to 24393.3 million Baht. The asset sizes of never
paid and former payers more than doubled, (19152.64 and 25963.53 Million Baht
respectively) and the assets of the payer group was smaller than the assets of nonpayers. This outcome is due to (1) a reduction in number of payers (2) an increase in
the total number of listed firms (3) the new lists were smaller in size than the mature
firms and (4) a reduction in the percentage of new lists that paid dividends (see table
5.1 and 5.9)
However over the period 1991 to 2002, the total assets of payers’ was maintained at
16160.28 mil. Baht. Indeed, the total assets of the payer group was less than all other
dividend groups. This is consistent with Fama and French’s (2001).
179
DBA Thesis: Malinee Ronapat
In terms of total assets, firms that paid dividends were larger firms before and after the
crisis. These firms also maintained their size. In addition, more firms became nonpayers during the period and the asset size of non-payers also increased.
The ratio of total liabilities to total asset (Lt/At)
This ratio represents the value of a firm’s liabilities to its total assets in baht. Therefore,
it represents the leverage of the firm. Fama and French (2001) used Lt/At to measure the
size of firms and assumed that the higher a firm’s liabilities, the higher is its ability to
access funds and its size.
Before the crisis, the Lt/At ratio for payers was 0.52 times. This was higher than the
ratio for never paid firms (0.45 times) and indicates that payers were well balanced in
their debt and equity capital. The ratios of non-payers and former payers were higher at
0.58 and 0.67 respectively.
After the crisis, the ratio of most dividend groups was large, indicating that the debt
levels of all firms had increased. Only the group that paid dividends maintained its
Lt/At ratio. This indicates that these firms were not experiencing high levels of debt or
financial difficulties. On average, between 1997 and 2001, the lowest Lt/At ratio (0.39
times) was achieved by payers, followed by never paid firms (0.81 times) and former
payers (1.33 times). The non-payer group averaged 1.25 times.
From 1991 to 2002, the payer group recorded the lowest Lt/At, or highest level of
leverage (0.45 times). The ratio of all firms’s was 0.73, non-payers 0.92, never paid
0.63 and former payers 1.00 times. After the economic crisis, a large number of firms
were counted as non-payers, never paid and former payers. Therefore, the total assets of
these groups increased after the crisis.
180
DBA Thesis: Malinee Ronapat
In term of leverage, firms which pay dividends have high leverage and they tend to be
well balanced between debt and equity capital, while non-payers are burdened with
debt. When measuring total assets, firms which paid dividends were larger than the
other dividend groups, before the crisis, although the assets of some of these firms were
maintained after the crisis, while the value fell for some payers. The number of former
payers and non-payers grew after the Asian Economic Crisis. However, the ratio of
total liabilities to total assets was small for payers over time and the total asset size of
the payer group remained higher than the other dividend groups.
The debt and equity of new lists was well balanced before the crisis as indicated by a
Lt/At ratio of 0.53 times. After the crisis, this ratio increased to 0.72 times. This figure
was greater than the figure recorded for the existing listed firms. This indicates that the
new lists were greater issuers of debt than existing firms.
Table 5.10 summarises the characteristics of the dividend groups after the analysis
using descriptive statistics.
Table 5.10: Summary of the Characteristics of the Listed Firms in Thailand from
Descriptive Statistics Approach
CHARACTERISTICS OF THE LISTED FIRMS
All times
Payer
Profit
(Highest)
Highly
Positive
Before Crisis
Investment
Size
Highly
Positive
Largest
After Crisis
Profit
(Highest)
Highly
Positive
Investment
Size
Positive
Largest
Profit
(Highest)
Slightly
Positive
Investment
Highly
Negative
Non-Payer
Slightly
Negative
Negative
Slightly
Positive
Highly
Positive
Highly
Negative
Never
Paid
Slightly
Positive
(Highest)
Highly
Positive
Slightly
Positive
(Highest)
Highly
Positive
Highly
Negative
Former
Payer
(Lowest)
Slightly
Negative
Highly
Negative
(Lowest)
Slightly
Positive
Slightly
Positive
Highly
Negative
Size
Positive
(Highest)
Highly
Positive
(Lowest)
Highly
Negative
Source: Developed for this research
181
Largest
DBA Thesis: Malinee Ronapat
Table 5.11 shows the percentage aggregate value accounted attributed to firms paying
dividends. Table 5.11 shows the real effect of the numerator and denominator for the
six ratios used to measure the characteristics of firms.
Table 5.11: Percent of Aggregate Values Accounted for by Firms Paying Dividends
Yt
Et
dAt
At
Vt
BEt
MEt
Lt
19912002
91.9
47.5
56.0
63.5
62.6
70.4
67.1
62.7
19911996
97.9
96.6
90.0
94.7
92.2
89.0
86.0
95.9
19972001
85.8
-1.6
22.0
32.3
33.0
51.9
48.3
29.5
1991
90.4
95.4
95.2
96.1
95.3
92.2
91.0
96.7
1992
97.0
96.3
92.4
95.7
95.0
94.2
92.9
96.0
1993
96.7
95.0
83.0
93.1
88.2
86.4
80.7
94.5
1994
98.0
97.0
86.8
95.0
90.6
86.3
81.3
96.9
1995
100.0
97.4
91.1
94.6
92.1
88.0
85.4
96.1
1996
105.2
98.6
91.7
93.7
92.4
86.7
84.4
95.3
1997
51.3
-66.5
86.6
87.6
88.1
82.4
88.0
88.1
1998
64.3
65.7
31.4
67.3
66.6
58.2
53.4
68.6
1999
-5.8
-104.5
10.1
4.3
4.4
23.3
13.7
1.8
2000
-20.0
38.9
29.0
6.2
6.1
39.3
28.4
2.6
2001
333.7
9.7
-52.8
10.2
11.5
48.7
44.1
5.1
2002
91.5
47.4
27.9
18.0
21.0
59.3
62.3
10.8
Source: Developed for this research
Note: At is asset, BEt is the book value for common equity, MEt is the market value of common equity, Lt is
the book liabilities, Vt is the market value, Et is the earnings before interest but after taxes, Yt is the after-tax
earnings to common stock and dAt is the change in assets in time t and t-1. Results for each time period are
grouped and averaged according to the dividend groups.
The aggregate values of both Yt and Et, are used for presenting the profitability of
firms. Firms which paid dividends accounted for almost 100 percent of the total Yt in
the market before the crisis. This confirms the findings of section 5.5 and payers are
more profitable than the other dividend groups. After the crisis, the figure for Yt and Et
decreased to 85.8 percent. This may result from the smaller number of payers in that
year. Indeed, most of the firms, including those that experienced high profitability,
suspended the payment of dividends in 1998. In 2001 and 2002, Yt increased to 91.5
percent because a large number of firms paid dividends (see table 5.1). The dAt which
represents the growth in assets, or investment opportunities displayed a very high
aggregate value ratio from 1991 to 1996 (table 5.11) although this declined to only 22
percent after the crisis. The total assets (At) accounted for by payers was a very high
ratio between 1991 and 1996. However, after the crisis At fell to 4 percent, before
182
DBA Thesis: Malinee Ronapat
rebounding to 18 percent in recent years due to the recovery. A similar pattern has
occurred with Vt, BEt, MEt and Lt (table 5.10). This outcome has resulted from (1) a
decline in the numbers of payers (2) downsizing of firms during the crisis (3) an
increase in the number of listed firms and (4) the fact that new lists are more likely to
be growth rather than value firms (table 5.1, 5.9).
5.5.2
ANOVA TEST
An analysis of variance (ANOVA) was undertaken to confirm that the groups are
sampled from different populations. ANOVA can also be used for comparing more than
two means at the same time (Moore and McLabe 1998; Ticehurst and Veal 2000;
Tabachnick and Fidell 2001). ANOVA is used for calculating the F statistic. The F
statistic is the ratio of the magnitude of the difference between the dividend groups to
the magnitude of the differences within these groups (Mason, Lind and Marchal 1999).
An ANOVA test was used to statistically confirm the results which are summarised in
table 5.9, 5.10 and 5.11. The first ANOVA test will (1) analyse the variance of the
payer and non-payer groups to statistically confirm the characteristics of the payers
from 1991 to 2002 and (2) analyse three groups: payers, former payers and firms that
have never paid dividends to statistically confirm the characteristics of these groups
from 1991 to 2002.
Table 5.12 identified the year and names of the ratios (presented earlier as
characteristics) which differed significantly between payers and non-payers for every
year over the period 1991 to 2002. The analysis was based on the F-statistic at the 0.05
level of significance. Table 5.12 summarises the characteristics of payers in a general
form.
183
DBA Thesis: Malinee Ronapat
Table 5.12: ANOVA F-statistics Test of Characteristics from 1991 to 2002
ANOVA
Year
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
NONE
-
Sum of Squares
-
df
-
Mean Square
-
F
-
Sig.
-
Et/At
Between Groups
0.049
1
0.049
7.941
0.005
Yt/BEt
Between Groups
0.807
1
0.807
14.172
0.000
dAt/At
Between Groups
0.897
1
0.897
5.570
0.019
Et/At
Between Groups
0.140
1
0.140
13.359
0.000
dAt/At
Between Groups
3.022
1
3.022
12.842
0.000
Et/At
Between Groups
0.118
1
0.118
29.695
0.000
Yt/BEt
Between Groups
76.584
1
76.584
6.444
0.012
Et/At
Between Groups
0.208
1
0.208
63.380
0.000
Yt/BEt
Between Groups
0.338
1
0.338
3.929
0.048
Lt/At
Between Groups
0.350
1
0.350
7.895
0.005
0.000
Et/At
Between Groups
0.303
1
0.303
51.805
Yt/BEt
Between Groups
1.261
1
1.261
13.242
0.000
Et/At
Between Groups
1.655
2
0.827
11.861
0.000
Lt/At
Between Groups
3.055
2
1.528
12.980
0.000
At
Between Groups
117807901049
1
117807901049
11.211
0.001
0.000
Et/At
Between Groups
2.397
1
2.397
17.776
dAt/At
Between Groups
2.062
1
2.062
8.638
0.003
Vt/At
Between Groups
10.566
1
10.566
6.267
0.013
AT
Between Groups
54724905688
1
54724905688
5.485
0.020
Lt/At
Between Groups
35.904
1
35.904
36.901
0.000
Et/At
Between Groups
3.099
1
3.099
6.728
0.010
Vt/At
Between Groups
32.035
1
32.035
6.657
0.010
0.014
At
Between Groups
67067466395
1
67067466395
6.127
Lt/At
Between Groups
75.784
1
75.784
15.778
0.000
Vt/At
Between Groups
91.620
1
91.620
5.912
0.016
At
Between Groups
64346225910
1
64346225910
6.093
0.014
Lt/At
Between Groups
159.137
1
159.137
10.592
0.001
Et/At
Between Groups
0.437
1
0.437
12.163
0.001
Vt/At
Between Groups
18.053
1
18.053
7.130
0.008
At
Between Groups
66001204222
1
66001204222
5.775
0.017
Lt/At
Between Groups
49.175
1
49.175
24.205
0.000
Source: Developed for this research
Note: Et/At is the ratio of aggregate earnings before interest to aggregate assets, Yt/BEt is the ratio of
aggregate common stock earnings over aggregate book equity, dAt/At is the growth rate in assets, Vt/At
is the aggregate market value to the aggregate book value of assets, At is total assets, Lt/At is the ratio of
total liabilities to total assets. Statistic is significant at 0.05 level. See appendix A for details.
184
DBA Thesis: Malinee Ronapat
Table 5.13: Summary of the Payers’ Characteristics Confirmation
Year
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
Characteristics confirmed
Investment
Profitability
Size
Opportunities
Et/At Yt/BEt dAt/At
Vt/At At Lt/At
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Source: Developed for this research
Note: Y is ‘Yes’ for confirming the characteristics in each year from 1991 to 2002. Et/At is the ratio of
aggregate earnings before interest to aggregate assets, Yt/BEt is the ratio of aggregate common stock
earnings over aggregate book equity, dAt/At is the growth rate in assets, Vt/At is the aggregate market
value to the aggregate book value of assets, At is the total asset, Lt/At is the ratio of total liabilities to
total asset. See appendix A for details.
After conducting a one-way analysis of variance, firms which paid dividends displayed
significantly higher profitability (as measured by Et/At) (table 5.12) than non-payers in
1992, 1993, 1994, 1995, 1996, 1997, 1999, 2000 and 2002. In 1998 and 2001, there is
no significant difference between the means of the two groups. However, most firms
made losses in these years. The Yt/BEt (the ratio of aggregate common stock earnings
over aggregate book equity) of payers indicated significantly higher profitability for
firms which paid dividends in 1992, 1994, 1995 and 1996. It appears that profitability is
an important characteristic of firms which paid dividends before and after the crisis.
The investment opportunities (this was measured by dAt/At which is an indication of
the growth rate of assets) of the payers was significantly lower than non-payers in 1992,
1993, 1997 and 1999.
Vt/At (market value of stocks to book value of assets) identified significant variation
between payers and non-payers in 1999, 2000, 2001 and 2002. Firms which pay
185
DBA Thesis: Malinee Ronapat
dividends have a lower Vt/At than firms which did not pay dividends. Significant
differences in firm size emerged after At (or total assets) (see table 5.12). Firms which
paid dividends have less assets than non-payers. This could be a result of the
downsizing which many firms undertook due to the Asian Economic Crisis. The
variable Lt/At (total liabilities to total assets) was used as an alternative measurement of
size. This variable was significantly smaller for payers than non-payers in 1995, 1997,
1999, 2000, 2001, 2002.
The arithmetic means are presented in tables 5.9 and 5.11 and indicate that the
characteristics of payers and non-payers differ. ANOVA suggested that profitability, as
measured by the Et/At ratio (earnings to assets) was a significant characteristic of firms
which paid dividends before and after the crisis. A firm’s Yt/BEt (earnings to book
equity) was only important before the crisis. The results for investment opportunities
were mixed. However, it was clear that Vt/At was low and a significant characteristic of
payers after the crisis. In addition to size, both At (total assets) and Lt/At (total liabilities
to total assets) tended to be characteristics of payers after the crisis (smaller At and
Lt/At than non-payers).
The second ANOVA test divided the total samples into three groups (1) payers (2)
former payers and (3) firms that have never paid dividends. It is appropriate to divide
the non-payers into two groups (group (2) and (3)) because this is a better classification
of the characteristics of firms which do not pay dividends (Fama and French 2001).
Table 5.14 presents the values for the ANOVA F-statistics over a range of firm
characteristics. The table presents the results of six ratios for each year between 1991
and 2002. Table 5.15 is a summary of the characteristics of payers, former payers and
firms which have never paid dividends.
186
DBA Thesis: Malinee Ronapat
Table 5.14: ANOVA Test for Payers, Former Payers and Never Paid Firms
Year
Sum of
Squares
1991
1992
1993
1994
1995
dAt/At
2
2.615
F
Sig.
6.849
0.001
(I)
Multiple Comparisons: Tukey HSD
Std.
(J)
Mean Difference (I-J)
Error
Sig.
1
2
0.426
0.136
0.006
3
2
0.743
0.22
0.002
Lt/At
0.494
2
0.247
4.402
0.013
1
3
0.209
0.07
0.009
Et/At
0.065
2
0.033
5.317
0.005
1
2
0.071
0.022
0.004
Yt/BEt
1.048
2
0.524
9.319
0
1
2
0.285
0.068
0
dAt/At
5.365
2
2.683
18.644
0
1
3
-0.672
0.117
0
2
3
-0.866
0.155
0
Et/At
0.187
2
0.094
9.026
0
1
2
0.093
0.022
0
dAt/At
7.789
2
3.894
17.753
0
1
3
-0.645
0.109
0
2
3
-0.668
0.143
0
Et/At
0.194
2
0.097
25.546
0
1
2
0.083
0.012
0
3
2
0.065
0.017
0.001
Yt/BEt
138.99
2
69.495
5.981
0.003
1
2
2.23
0.646
0.002
dAt/At
10.814
2
5.407
24.561
0
1
3
-0.672
0.104
0
2
3
-0.862
0.131
0
Vt/At
9.238
2
4.619
4.243
0.015
1
3
-0.625
0.231
0.02
2
3
-0.793
0.291
0.019
Lt/At
0.524
2
0.262
4.367
0.013
2
3
0.195
0.068
0.013
Et/At
0.209
2
0.104
31.705
0
Yt/BEt
0.608
2
0.304
3.556
0.03
dAt/At
1.68
2
0.84
6.297
0.002
Vt/At
1996
5.23
ANOVA
Mean
df
Square
5.915
2
2.958
5.384
0.005
1
2
0.069
0.009
0
3
1
-0.061
0.018
0.003
1
3
0.234
0.094
0.035
1
2
0.154
0.059
0.026
3
2
0.417
0.128
0.004
1
3
-0.665
0.238
0.015
2
3
-0.85
0.26
0.003
Lt/At
0.36
2
0.18
4.053
0.018
1
2
-0.094
0.034
0.017
Et/At
0.305
2
0.152
25.96
0
1
2
0.075
0.011
0
3
1
-0.064
0.019
0.002
Yt/BEt
Vt/At
2.174
6.69
2
2
1.087
3.345
11.683
6.239
0
0.002
1
2
0.21
0.045
0
3
2
0.265
0.085
0.005
1
3
-0.506
0.182
0.016
2
3
-0.713
0.203
0.001
187
DBA Thesis: Malinee Ronapat
Table 5.14: ANOVA Test for Payers, Former Payers and Never Paid Firms
(Continued)
Year
ANOVA
Sum of
Squares
1997
Et/At
dAt/At
Lt/At
1998
At
1999
Et/At
Yt/BEt
2000
Mean Square
2
0.827
11.861
F
Sig.
0
13.802
2
6.901
28.071
0
(I)
(J)
Mean Difference (I-J)
Std. Error
1
2
0.136
0.032
Sig.
3
1
-0.172
0.058
0.01
1
2
0.22
0.06
0.001
0
3
1
0.657
0.11
0
3
2
0.878
0.118
0
2
-0.196
0.042
0
3.055
2
1.528
12.98
0
1
3
1
0.191
0.076
0.033
1.17833E+11
2
58916744022
5.592
0.004
1
2
39057.12
11685.737
0.003
3
1
-29663.95
23106.115
0.405
2.408
2
1.204
8.908
0
1
2
0.175
0.042
0
300.909
2
150.455
4.548
0.011
3.594
2
1.797
7.635
0.001
Vt/At
10.846
2
5.423
3.21
0.041
Lt/At
36.749
2
18.374
18.878
0
1
3
3.654
1.212
0.008
2
3
2.889
1.148
0.033
1
2
0.186
0.055
0.002
3
2
0.247
0.097
0.03
1
2
-0.373
0.147
0.031
1
2
-0.686
0.112
0
3
1
0.503
0.208
0.042
Et/At
3.186
2
1.593
3.451
0.033
1
2
0.199
0.076
0.025
Vt/At
35.687
2
17.843
3.705
0.026
1
2
-0.665
0.246
0.019
69572911571
2
34786455758
3.171
0.043
1
2
-29499.17
11720.662
0.033
Lt/At
80.099
2
40.049
8.336
0
1
2
-1.002
0.245
0
Vt/At
117.491
2
58.745
3.797
0.023
1
2
-1.164
0.431
0.02
73409148645
2
36786455758
3.474
0.032
1
2
-29609.63
11260.019
0.024
189.731
2
94.865
6.332
0.002
1
2
-1.497
0.424
0.001
At
Lt/At
2002
df
1.655
dAt/At
At
2001
Multiple Comparisons: Tukey HSD
Et/At
dAt/At
Vt/At
At
Lt/At
0.438
2
0.219
6.086
0.002
1
2
0.069
0.021
0.003
12.497
2
6.249
4.499
0.012
2
3
-0.52
0.189
0.017
18.363
2
9.181
3.618
0.028
1
2
-0.456
0.175
0.025
1.03291E+11
2
51645588378
4.546
0.011
1
2
-38995.56
11694.419
0.011
57.183
2
28.592
14.182
0
1
2
-0.83
0.156
0
Source: Developed for this research
Note: Et/At is the ratio of aggregate earnings before interest to aggregate assets, Yt/BEt is the ratio of
aggregate common stock earnings over aggregate book equity, dAt/At is the growth rate in assets, Vt/At
is the aggregate market value to the aggregate book value of assets, At is total assets, Lt/At is the ratio of
total liabilities to total assets. Statistic is significant at 0.05 level. (1) is payers, (2) is former payers (3) is
never paid firms. See appendix B for details.
Table 5.15: Summary of the Characteristics of Payers, Former Payers and Never
Paid Firms
188
DBA Thesis: Malinee Ronapat
Year
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
Characteristics confirmed Payers, Former payers and Never paid
Profitability
Investment Opportunities
Size
Et/At
Yt/BEt
dAt/At
Vt/At
At
Lt/At
1(2), 2(3)
1(3)
1(2)
1(2)
1(2), 2(3)
1(2)
1(3), 2(3)
1(2), 2(3) 1(2)
1(3), 2(3)
1(3), 2(3)
2(3)
1(2), 1(3) 1(3)
1(2), 2(3)
1(3), 2(3)
1(2)
1(2), 1(3) 1(2), 2(3)
1(3), 2(3)
1(2), 1(3)
1(2), 1(3), 2(3)
1(2), 1(3)
1(2), 1(3)
1(2)
1(3), 2(3)
1(2), 2(3)
1(2)
1(2), 1(3)
1(2)
1(2)
1(2)
1(2)
1(2)
1(2)
1(2)
1(2)
2(3)
1(2)
1(2)
1(2)
Source: Developed for this research
Note: 1(2) is payers (1) to former payers (2). 1(3) is payers (1) to never paid firms (3). 2(3) is former
payers (2) and never paid firms (3). Et/At is the ratio of aggregate earnings before interest to aggregate
assets, Yt/BEt is the ratio of aggregate common stock earnings over aggregate book equity, dAt/At is the
growth rate in assets, Vt/At is the aggregate market value to the aggregate book value of assets, At is the
total asset, Lt/At is the ratio of total liabilities to total assets. See appendix B for details.
The second stage of the ANOVA testing repeated the results of the first stage. Table
5.15 indicates that payers have significantly higher profitability than former payers and
firms that have never paid dividends in most years (measured by both Et/At and
Yt/BEt). However, the results from the Vt/At ratio are mixed. The analysis of dAt/At
indicates that payers have lower investment opportunities than firms that have never
paid dividends, but payers have higher investment opportunities than former payers. In
relation to firm size there was minimal variation between the three groups before the
crisis. However, the value of payers’ assets was significantly smaller than the asset
value of former payers after the crisis. This could result from the unsuccessful
elimination of NPLs (Non-Performing Loans) and the restructuring which was
undertaken by former payers. The results for Lt/At were mixed in the early years,
although payers had less liabilities (lower Lt/At or ratio of total liabilities to total assets)
than former payers.
The ANOVA results are similar to the descriptive tests and confirm that firms which
pay dividends are more profitable (as presented by the Et/At ratio) and have low
investment opportunities. These firms are well balanced between equity and debt (as
189
DBA Thesis: Malinee Ronapat
presented by Lt/At ratio). Former payers are usually distressed firms with low
profitability, low investment opportunities and a large asset size. This may have
resulted from the NPLs and low quality assets which are owned by these firms. Firms
that never paid dividends have been relatively profitable (lower than payers but higher
than former payers) and enjoyed very high investment opportunities. It therefore
appears that firms are more successful if they have never paid dividends.
5.5.3 REVIEW OF DESCRIPTIVE STATISTIC RESULTS
The evidence from the descriptive statistic is largely consistent with the findings of
Fama and French (2001) and suggests the three characteristics: profitability, investment
opportunities and size are factors affecting the dividend policies of listed firms.
However, the relationships between characteristics and dividend payments were
confirmed for only some of the years which were investigated. The higher the level of
profitability, the more likely firms will pay dividends. This trend is consistent over time
and also relevant to the findings of Fama and French (2001). In relation to investment
opportunities, evidence from the pre-crisis period suggests that payers have lower
investment opportunities than the firms which do not pay dividends and are larger in
size than firms in the other groups. All of these findings are consistent with Fama and
French (2001). After the crisis, many listed firms tended to downsize and ceased paying
dividends. Most of the non-payers were heavily in debt and owned a high proportion of
low quality assets. It appears that the characteristics of the payers changed in the latter
part of 1997. The asset size of the payers has stabilized since 1997. Indeed, the asset
growth of the payer group has been slightly positive, while the other dividend groups
engaged in downsizing and liquidated some of their assets. These firms therefore
experienced negative growth in their assets. Furthermore, the percentage of firms which
pay dividends has fallen and more firms are now non-payers. Lastly, firms that have
never paid dividends have greater investment opportunities than payers and this is
consistent over time.
190
DBA Thesis: Malinee Ronapat
Firms which pay dividends were large in size and profitable before the crisis
(confirmed by descriptive statistic approach). However, they tended to have low
investment opportunities (confirmed by descriptive and ANOVA approaches). Former
payers are distressed firms with low profitability and investment opportunities, but
large in terms of assets sizes after the crisis. Firms that have never paid dividends were
more profitable and had better investment opportunities than former payers.
The newly listed firms were added to the total sample. The findings suggest that new
firms tend to be small, but quickly adopt the characteristics of non-payers, or growth
stock and contribute to the decline in the percentage of payers. After the crisis, new lists
continued to be small and grew rapidly, but their profitability declined. Thus, new lists
that pay a dividend become unrare. Therefore, new lists rarely paid dividends.
The decline in the percentage of payers from 1991 is partly due to (1) the characteristics
of new lists are increasingly becoming similar to firms that do not pay dividends and
(2) the Asian Economic Crisis in 1997 resulted in financial distress for many firms
(both new and mature) and they ceased paying dividends. The findings suggest there is
an unusual pattern in the payment of cash dividends and the percentage of dividend
payers during, and a year after the Asian Economic Crisis (1997-1998). However, the
descriptive statistics suggest that this trend is temporary and the market appears to be
recovering because: (1) the percentage of payers is increasing and (2) the profitability
of firms is increasing. A formal analysis of the characteristics of firms will be discussed
in the section 5.5.4.
191
DBA Thesis: Malinee Ronapat
5.5.4
CONFIRMATION FROM LOGIT REGRESSIONS
Table 5.16 summarises the results of the logit regression which was undertaken to
identify the relative importance of the three characteristics (profitability, investment
opportunities and size) which influence a firm’s dividend behaviour. Logit regression is
used for explaining why firms pay dividends and for developing a formula for
predicting the firms which are likely to pay dividends. Profitability is measured as the
ratio of a firm’s earnings before interest to its total assets (Et/At) (Fama and French
2001). The proxies for investment opportunities are a firm’s rate of growth of assets,
dAt/At and its market-to-book ratio, Vt/At (Fama and French 2001). The size, SET, is
measured by the percentage of listed firms with the same, or smaller market
capitalization. This will ‘eliminate the effects of the growth in size through time’ (Fama
and French 2001, p. 12). Logit regression is estimated yearly and applies the time-series
standard deviations of the annual coefficients. This permits a correlation of the
regression residuals across firms and ensures judgements about the average coefficient
can be made (Fama and MacBeth 1973; Fama and French 2001, p. 12). The t-statistics
table shows the mean divided by the standard error. This is the times-series standard
deviation of the regression coefficient divided by the square root of the number of years
in the period of the study (Fama and French 2001). From 1991 to 2001, the slopes from
the regressions match the roles of profitability, investment opportunities and size as
characteristics which affect the dividend decisions of firms between 1990 and 2002.
192
DBA Thesis: Malinee Ronapat
Table 5.16: Logit Regression to Explain Which Firms Pay Dividends
Average Coefficient
t-statistic
Int
SETt
Vt/At
dAt/At
Et/At
Int
SETt
Vt/At
dAt/At
Et/At
1991-1996
1997-2002
1991-2002
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
1.38
-0.77
0.30
1.52
2.61
1.89
1.11
0.66
0.48
0.29
-2.25
-0.91
-0.68
-0.66
-0.44
0.87
1.85
1.36
1.69
0.25
-0.02
1.09
0.83
1.36
2.01
2.69
1.24
1.74
1.73
1.70
-0.36
-0.59
-0.47
-0.05
-0.59
-0.07
-0.46
-0.64
-0.34
-0.08
-0.25
-0.73
-1.17
-0.54
-0.77
-0.92
0.18
-0.37
0.30
-1.34
-2.85
-1.75
0.30
-0.18
-0.19
0.26
0.53
0.75
0.00
-0.29
14.54
3.14
8.84
-3.04
14.64
15.06
20.82
24.16
15.62
1.79
1.64
4.31
5.58
-0.06
5.61
4.22
-2.28
0.76
3.25*
9.45*
6.27*
-1.77
1.05
-1.20
-3.46*
-3.69*
-4.87*
3.79*
3.27*
3.46*
1991-1996
1997-2002
1991-2002
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
1.12
-1.22
-0.05
1.62
1.89
1.85
0.83
0.25
0.28
0.24
-2.37
-1.37
-1.49
-1.27
-1.05
0.39
1.63
1.01
0.91
0.05
-0.20
0.42
0.11
1.01
1.96
2.49
0.75
1.32
1.76
1.52
-1.01
0.28
-0.36
0.34
-1.51
-2.93
-2.05
0.32
-0.21
-0.19
0.44
0.52
0.80
0.01
0.08
13.53
2.43
7.98
-2.90
11.43
15.01
19.90
22.39
15.37
1.83
1.57
3.97
5.06
-0.07
2.20
3.59
-3.53
-0.12
1.92
6.77*
4.19*
-1.81
1.83
-1.09
3.71*
3.25*
3.27*
Source: Developed for this research
Note: Logit regressions are estimated for each year from 1991-2002. The dependent variable is 1.0 in
year t if the firm is a payer and 0.0 otherwise. The explanatory variables are Et/At for profitability,
dAt/At for investment opportunities, Vt/At for investment opportunities and SETt for size. SETt
represents the percentage of SET firms with an equal, or rate of lower market capitalization in time t. For
the range of years, means of the regression intercepts and independent variables. The t-statistics table is
defined as the mean divided by the standard error which provides the coefficient means. (*) the figure is
significant at <0.05 level. See appendix C, D, E and F for details.
193
DBA Thesis: Malinee Ronapat
During the pre-crisis period (1991-1996), larger firms tended to pay dividends. This is
confirmed by the average slope of 3.25 of SETt. Firms with higher profitability are
more likely to pay dividends. The slope on Et/At is 3.79. In relation to investment
opportunities, the t-statistics suggest that payers tend to have lower investment
opportunities than all other firms, due to relatively low growth in their assets (-3.46)
and Vt/At (-1.77).
The coefficient for each year from 1991 to 1996 confirmed that Thai firms which paid
dividends possess the three characteristics which were suggested by Fama and French
(2001). This finding is valid especially for Et/At, SET. The ratio Et/At showed a high
positive coefficient every year from 1991 to 1996 in addition to SET. The coefficient
for Vt/At was negative in most years (table 5.15).
In relation to the post-crisis period (1997-2002), the logit regression suggests that large
firms with large capitalisation are more likely to pay dividends. This is confirmed by
the slope of 9.45 for SETt. The t-statistics suggest that profitable firms tend to pay
dividends. In this time period, the investment opportunities of firms is consistent with
Fama and French (2001) because the slope for the growth in assets is negative (-3.69)
and Vt/At (1.05) is positive. Therefore, firms which pay dividends tend to have low
investment opportunities.
In relation to the period 1991 to 2002, the slopes from the regressions confirm that
firms with high profitability (3.46) and of larger size (6.27) are more likely to pay
dividends. This finding is consistent with Fama and French (2001). Furthermore the
two measures of investment opportunities are also consistent with Fama and French
(2001). These ratios have negative slopes of -4.87 and -1.2 respectively. The result for
the growth in assets are -4.87. This indicates that firms with low investment
opportunities tend to pay dividends. The Vt/At ratio is also negative (-1.2) and infers
194
DBA Thesis: Malinee Ronapat
that firms with low investment opportunities tend to pay dividends. This conclusion is
also similar to the evidence which is presented in the literature.
The conclusion that firms which pay dividends tend to be more profitable, larger and
experience low investment opportunities is consistent with Fama and Babiak (1968),
Easterbrook (1984), Jensen (1986), Fama and French (1988 1995, 1997, 1998, 2001),
DeAngelo DeAngelo and Skinner (2002), Aivazian, Booth and Cleary (2003).
Furthermore, Fama and French (1999) suggested that dividend paying firms are
profitable or have higher payout ratios than firms with higher investment opportunities
(Fama and French 2001). Firms with high investment opportunities tend to achieve high
growth with small value. This is consistent with the characteristics of non-payers (Fama
and French 2001). Hence, high growth firms are reluctant to issue equities because the
trasaction costs are high (Myers 1984, Fama and French 1999, Fama and French 2001).
In short, the findings for Thailand are consistent with the literature (see table 5.9).
Figure 5.6 re-states the hypotheses which were developed in chapter 3.
Controlling for one investment variable, suggests similar results to the earlier parts.
Payers, before, during and after the crisis, have high profitability and size but low
investment opportunities (see table 5.16).
195
DBA Thesis: Malinee Ronapat
Figure 5.6: Hypotheses Revisiting
CHARACTERISTICS OF DIVIDEND PAYERS BEFORE THE ASIAN ECONOMIC CRISIS (Step A)
(1) PROFITABILITY
Ho
:
Ha
:
There is no significant relationship between the profitability of the firms and the probability
that the listed firms will pay dividend before the Asian Economic Crisis.
There is a significant relationship between the profitability of the firms and the probability
that the listed firms will pay dividend before the Asian Economic Crisis.
(2) SIZE
Ho
:
Ha
:
There is no significant relationship between the size of the firms and the probability that the
listed firms will pay dividend before the Asian Economic Crisis.
There is a significant relationship between the size of the firms and the probability that the
listed firms will pay dividend before the Asian Economic Crisis.
(3) INVESTMENT OPPORTUNITIES
Ho
:
Ha
:
There is no significant relationship between the investment opportunities of the firms and the
probability that the listed firms will pay dividend before the Asian Economic Crisis.
There is a significant relationship between the investment opportunities of the firms and the
probability that the listed firms will pay dividend before the Asian Economic Crisis.
CHARACTERISTICS OF THE DIVIDEND PAYERS DURING AND AFTER THE ASIAN ECONOMIC CRISIS (Step B)
(1) PROFITABILITY
Ho
:
Ha
:
There is no significant relationship between the profitability of the firms and the probability
that the listed firms will pay dividend after the Asian Economic Crisis.
There is a significant relationship between the profitability of the firms and the probability
that the listed firms will pay dividend after the Asian Economic Crisis.
(2) SIZE
Ho
:
Ha
:
There is no significant relationship between the size of the firms and the probability that the
listed firms will pay dividend after the Asian Economic Crisis.
There is a significant relationship between the size of the firms and the probability that the
listed firms will pay dividend after the Asian Economic Crisis.
(3) INVESTMENT OPPORTUNITIES
Ho
:
Ha
:
There is no significant relationship between the investment opportunities of the firms and the
probability that the listed firms will pay dividend after the Asian Economic Crisis.
There is a significant relationship between the investment opportunities of the firms and the
probability that the listed firms will pay dividend after the Asian Economic Crisis.
CHARACTERISTICS OF THE DIVIDEND PAYERS FOR THE WHOLE PERIOD (Step C)
(1) PROFITABILITY
Ho
:
Ha
:
There is no significant relationship between the profitability of the firms and the probability
that the listed firms will pay dividend for the whole period of investigation.
There is a significant relationship between the profitability of the firms and the probability
that the listed firms will pay dividend the whole period of investigation.
(2) SIZE
Ho
:
Ha
:
There is no significant relationship between the size of the firms and the probability that the
listed firms will pay dividend the whole period of investigation.
There is a significant relationship between the size of the firms and the probability that the
listed firms will pay dividend the whole period of investigation.
(3) INVESTMENT OPPORTUNITIES
Ho
:
Ha
:
There is no significant relationship between the investment opportunities of the firms and the
probability that the listed firms will pay dividend the whole period of investigation.
There is a significant relationship between the investment opportunities of the firms and the
probability that the listed firms will pay dividend the whole period of investigation.
Source: Developed for this research
196
DBA Thesis: Malinee Ronapat
5.5.5
HYPOTHESES TESTING USING LOGIT REGRESSION (AVERAGE
COEFFICIENT)
Each year during the investigation period (1991 to 2002) has its own coefficient. As
stated earlier, the twelve year period (1991 to 2002) is grouped into three steps (A, B,
and C). Step A represents the period 1991 to 1996 (before the crisis), step B the period
from 1997 to 2002 (during and after the crisis), and step C is while period from 1991 to
2002. The coefficient of each variable, in each year, is averaged to obtain an average
coefficient for each step (A, B and C) (Fama and MacBeth 1973; Fama and French
2001).
The hypotheses tested by logit regression were also divided into the three steps (A, B,
and C). Step A’s null hypothesis is to test each characteristic of the firms and their
relation to the probability that firms will pay dividends before the crisis (1991-1996),
after the crisis in step B (1997-2002) and for the whole period of investigation in step C
(1991-2002). Table 5.16 presents the logit regression for two sets of results (1)
controlling two variables for investment opportunities (Et/At, dAt/At, Vt/At and SETt)
and (2) controlling one variable for investment opportunities (Et/At, dAt/At and SETt).
As stated in chapter 3, the hypothesis from step A (1) states there is no significant
relationship between the profitability of the firms and the probability that listed firms
will pay dividends before the Asian Economic Crisis (Ho). With the ratio of Et/At (the
ratio of aggregate earnings before interest to aggregate assets) as a measure of
profitability, the logit regression results indicate that the null hypothesis is rejected
(0.012). This infers there is a relationship between a firm’s profitability and its
probability to pay dividends before the crisis (table 5.16). Table 5.16 indicates that the
value of the t-statistic is 3.79 for Et/At. This was calculated from the average coefficient
during 1991 to 1996.
The null hypothesis (1) in steps B (1997-2002) and C (1991-2002) states that there is no
significant relationship between the profitability of the firms and the probability that the
listed firms will pay dividends in step B (0.022) and C (0.005) (Ho) is also rejected.
197
DBA Thesis: Malinee Ronapat
Table 5.16 presents the t-statistics for Et/At (earnings to total assets) for the period 1997
to 2002 (3.27) and the period 1991 to 2002 (3.46). Hence, there is a relationship
between the firm’s profitability and the probability that it will pay dividends before the
crisis, after the crisis and during the whole period of investigation. Firms with higher
profitability are more likely to pay dividends. Therefore, Et/At (earnings to total assets)
has significant power for predicting the dividend behaviour of firms.
When controlling only one investment variable, the null hypothesis (1) of steps A
(0.013), B (0.022) and C (0.007) is also rejected (Ho). The lower part of table 5.16
shows that the value for the t-statistics of Et/At (earnings to total assets) are highly
positive for steps A (3.71), B (3.25) and C (3.27). This infers that Et/At (earnings to
total assets), as a measure of a firm’s profitability, is a significant variable for
predicting the dividend behaviour of the firm in every time period. Therefore, the
higher the level of profitability, the more likely a firm will pay dividends.
Null hypothesis number 2 from step A (before the crisis), B (during and after the crisis)
and C (the whole period of investigation) stated that there is no significant relationship
between the size of the firms and the probability that these firms will pay dividends
(Ho). SETt (the percentage of SET firms with the same or a lower level of market
capitalization) is a variable which represents the size of the firm. The null hypotheses
(Ho) in step A (3.25), B (9.45) and C (6.27) are rejected. Table 5.16 presents the value
of SET in all three steps. This infers that the firm size is significant for predicting
dividend behaviour in all time periods (1991-1996 (0.02), 1997-2002 (0.000) and 19912002 (0.000)). Consequently, larger firms are more likely to pay dividends.
By controlling one investment variable, the variable SETt represents the size of the
firm. However, null hypothesis number 2 in step A (0.11) is accepted. This infers there
is no relationship between the size of the firm and its dividend behaviour after the crisis
(Ho). Table 5.16 indicates that the t-statistic for SET is 1.92 for the period 1991 to
1996. For groups B (0.001) and C (0.001), null hypothesis number 2 is rejected. This
infers there is a relationship between a firm’s size and its dividend behaviour. The
results indicate that the variable SETt in steps B and C has significant power for
198
DBA Thesis: Malinee Ronapat
predicting the dividend behaviour of a firm. In addition, firms with high investment
opportunities are less likely to pay dividends.
Null hypothesis number 3 in steps A (1991-1996) (0.018), B (1997-2002) (0.014) and C
(1991-2002) (0.000) states that there is no significant relationship between the
investment opportunities of the firms and the probability that they will pay dividends
(Ho). The investment opportunities are represented by two variables, Vt/At (the marketto-book ratio) and dAt/At (the growth rate of assets). Table 5.16 indicates that the tstatistics (for dAt/At) are strongly negative for group A (1991-1996), B (1997-2002)
and C (1991-2002). Therefore, the null hypotheses are rejected in all groups. This infers
that there is a negative relationship between investment opportunities (measured by
dAt/At) and the probability that firms will pay dividends in step A (-3.46), B (-3.69)
and C (-4.87). Investment opportunities (dAt/At) is a significant variable for predicting
the dividend behaviour of the firm. If a firm’s investment opportunities rise, it will be
less likely to pay dividends.
Table 5.16 indicates that the t-statistics for Vt/At (book value to total assets) in every
time period (A (0.137), B (0.340), C (0.253)), are small and the signs of t are randomly
positive and negative. This infers that the null hypotheses can not be rejected and there
is no significant relationship between investment opportunities (measured by Vt/At) and
the dividend behaviour of firms in every period. Therefore, when controlling two
investment variables only dAt/At (growth rate of assets) is significant for predicting the
dividend behaviour of the firm in steps A, B and C. The variable Vt/At was found to be
insignificant.
After controlling one investment variable, null hypothesis number 3 is accepted for
steps A (0.129), B (0.127) and C (0.300). The values for the t-statistic in table 5.16 are
randomly positive and negative and are not sufficiently significant to reject the null
hypothesis. This infers that dAt/At, (when used as the only measure of investment
opportunities) has no relationship and is not a significant predictor of a firm’s dividends
behaviour.
199
DBA Thesis: Malinee Ronapat
Table 5.17 summarises the results of hypotheses testing the applying average
coefficient method.
Table 5.17: Summary of Hypothesis Testing Results (Average Coefficient)
Periods
Controlled Variables
2
Variables
Before the Crisis
(Group A) 1991-1996
Characteristics
Profitability
Size
Investment (dAt/At)
Accepted
Investment (Vt/At)
Profitability
Size
Investment
(dAt/At)
1
Variable
Profitability
Size
Investment (dAt/At)
Investment (Vt/At)
2
Variables
During and After the
Crisis
(Group B) 1996-2002
Profitability
Size
Investment
(dAt/At)
1
Variable
The Whole Period of
Investigation
(Group C) 1991-2002
Accepted
Accepted
Rejected
Rejected
Rejected
Accepted
Rejected
Rejected
Accepted
Accepted
Investment (Vt/At)
1
Variable
Rejected
Rejected
Rejected
Rejected
Profitability
Size
Investment (dAt/At)
2
Variables
Null Hypothesis
Rejected
Rejected
Rejected
Profitability
Size
Investment
(dAt/At)
Rejected
Rejected
Accepted
Source: Developed for this research
Note: See appendix C, D, E and F for details.
5.5.6 HYPOTHESES TESTING USING LOGIT REGRESSION (GROUP
COEFFICIENT)
Different from average coefficient, calculating coefficient using firm period (hereafter
group coefficient) allows data to be entered simultaneously into the logit regression
equation to identify the coefficient for each step. Each year during the investigation
period (1991 to 2002) has its own coefficient. As stated earlier, the twelve years (1991
200
DBA Thesis: Malinee Ronapat
to 2002) are grouped into three steps (A, B, and C). The coefficient of each variable in
each step is observed by running a logit regression.
Tables 5.18 and 5.19 present the results of the logit regression. These results were
obtained by using the group coefficient controlling two investment variables (Et/At,
dAt/At, Vt/At and SETt) and one investment variable (Et/At, Vt/At and SETt)
respectively. Where: Et/At is the ratio of aggregate earnings before interest to aggregate
assets, dAt/At is the growth rate in assets, Vt/At is the aggregate market value to the
aggregate book value of assets, SETt represents the percentage of SET firms with an
equal or lower rate of market capitalisation in time t.
Table 5.18: Logit Regression Results (Controlling Two Investment Variables)
Variables in the Equation
Group A (1991-1996)
Variables in the Equation
Group B (1997-2002)
Variables in the Equation
Group C (1991-2002)
SETt
Vt/At
dAt/At
Et/At
Constant
SETt
Vt/At
dAt/At
Et/At
Constant
SETt
Vt/At
dAt/At
Et/At
Constant
B
0.630
-0.054
-0.577
12.705
0.809
B
1.699
-0.464
0.156
0.059
-0.676
B
0.741
0.005
0.465
1.491
-0.009
S.E.
0.255
0.018
0.195
1.229
0.142
S.E.
0.167
0.082
0.093
0.096
0.115
S.E.
0.119
0.014
0.074
0.212
0.066
Wald
6.079
8.559
8.788
106.957
32.597
Wald
103.767
32.339
2.808
0.381
34.352
Wald
38.657
0.124
39.226
49.547
0.019
df
1
1
1
1
1
df
1
1
1
1
1
df
1
1
1
1
1
Sig.
0.014
0.003
0.003
0.000
0.000
Sig.
0.000
0.000
0.094
0.537
0.000
Sig.
0.000
0.725
0.000
0.000
0.890
Exp(B)
1.877
0.947
0.562
1.000
2.246
Exp(B)
5.467
0.629
1.169
1.061
0.509
Exp(B)
2.098
1.005
1.593
4.443
0.991
Source: Developed for this research
Note: The explanatory variables are Et/At for profitability, dAt/At for investment opportunities, Vt/At for
investment opportunities and SET for size. SETt represents the percentage of SET firms with equal or
lower market capitalization rate in time t. Statistic is significant at 0.05 level. The constant is the
autonomous and B is the beta coefficient (Ticehurst and Veal 2000).
Table 5.19: Logit Regression Results (Controlling One Investment Variable)
201
DBA Thesis: Malinee Ronapat
Variables in the Equation
Group A (1991-1996)
Variables in the Equation
Group B (1997-2002)
Variables in the Equation
Group C (1991-2002)
SETt
dAt/At
Et/At
Constant
SETt
dAt/At
Et/At
Constant
SETt
dAt/At
Et/At
Constant
B
0.473
-0.610
12.454
0.802
B
1.523
0.249
0.139
-1.085
B
0.747
0.459
1.502
-0.004
S.E.
0.248
0.196
1.215
0.142
S.E.
0.159
0.096
0.132
0.094
S.E.
0.118
0.074
0.212
0.065
Wald
3.643
9.685
105.001
32.063
Wald
92.053
6.708
1.114
133.582
Wald
39.983
38.399
50.056
0.005
df
1
1
1
1
df
1
1
1
1
df
1
1
1
1
Sig.
0.056
0.002
0.000
0.000
Sig.
0.000
0.010
0.291
0.000
Sig.
0.000
0.000
0.000
0.945
Exp(B)
1.605
0.544
1.000
2.231
Exp(B)
4.587
1.283
1.150
0.338
Exp(B)
2.110
1.583
4.492
0.996
Source: Developed for this research
Note: The explanatory variables are Et/At for profitability, dAt/At for investment opportunities and SET
for size. SETt represents the percentage of SET firms with equal or lower market capitalization rate in
time t. Statistic is significant at 0.05 level.
The following results are presented in table 5.18: Et/At (0.000), dAt/At (0.003), Vt/At
(0.003) and SETt (0.014). In step A, the null hypothesis is rejected for each component
of the firms’ characteristics (profitability, investment opportunities and size). In other
words, variable Et/At (earnings to total assets), dAt/At (growth rate of assets), Vt/At
(market value to total assets) and SETt (SETt represents the percent of SET firms with
equal or lower market capitalization rate) are significant for predicting the dividend
payment behaviour of firms. In step A (1991-1996), large firms with high profitability,
and low investment opportunities are more likely to pay dividends.
However, the null hypothesis is rejected in step B for Et/At (0.537) and dAt/At (0.094)
and it is accepted in step B for Vt/At (0.000) and SETt (0.000). Therefore, during and
after the Asian Economic Crisis, only Vt/At and SETt are significant for predicting the
dividend behaviour of firms. During and after the Asian Economic crisis (step B), large
firms with low investment opportunities (measured by Vt/At or the ratio of market value
to total assets) are more likely to pay dividends.
In step C (1991-2002), the logit regression suggests that the hypotheses number 1 and 2
should be rejected. In relation to null hypothesis 3, dAt/At (growth rate of assets) should
202
DBA Thesis: Malinee Ronapat
be rejected and Vt/At, is accepted (0.725). This infers that, dAt/At, (a measure of
investment opportunities) is a significant variable for predicting dividend behaviour in
step C (1991-2002) while Vt/At (market value to total assets) is insignificant. Therefore,
over the period 1990 to 2002, large firms with high profitability and positive growth in
their assets are more likely to pay dividends.
Table 5.19 presents the results of the logit regression after controlling one investment
variable. The variables include: Et/At (earnings to total assets), dAt/At (growth rate of
assets), and SETt (SETt represents the percent of SET firms with equal or lower market
capitalization rate). Table 5.17 indicates in step A, null hypothesis number 2 is accepted
(0.056) while number 1(0.000) and 3 (0.002) are rejected. This infers that investment
opportunities (dAt/At) and profitability (Et/At) are significant variables for predicting
the dividend behaviour of a firm and size (SETt) is insignificant between 1991 and
1996. In short, before the Asian Economic Crisis, highly profitable firms with low
investment opportunities were more likely to pay dividends.
In step B, null hypothesis number 1 is accepted (0.291) because there is no relationship
between profitability and the dividend behaviour of firms. Null hypotheses numbers 2
(0.000) and 3 (0.10) are also rejected. This infers that firm size and investment
opportunities are significant for predicting the dividend behaviour of a firm.
Consequently, in step B (1997-2002), firms which are large and experiencing high
growth in their assets are more likely to pay dividends.
Lastly, step 3 considers the period 1991 to 2002 and null hypotheses numbers 1 (0.000),
2 (0.000) and 3 (0.000) were rejected. This implies that profitability, size and
investment opportunities are significant variables when predicting the dividend
behaviour of a firm. This suggests that large firms which experience high profitability
and a positive growth rate in their assets are more likely to pay dividends.
Table 5.20 shows the summary of the hypotheses testing results applying group
coefficient method.
203
DBA Thesis: Malinee Ronapat
Table 5.20: Summary of Hypothesis Testing Summary (Group Coefficient)
Periods
Controlled variables
2
Variables
Before the Crisis
(Group A) 1991-1996
Characteristics
Profitability
Size
Investment (dAt/At)
Rejected
Investment (Vt/At)
Profitability
Size
Investment
(dAt/At)
1
Variable
2
Variables
During and After the
Crisis
(Group B) 1996-2002
The Whole Period of
Investigation
(Group C) 1991-2002
Rejected
Accepted
Profitability
Size
Investment (dAt/At)
Rejected
Rejected
Rejected
Accepted
Investment (Vt/At)
Rejected
Profitability
Size
Investment
(dAt/At)
1
Variable
2
Variables
Null Hypothesis
Rejected
Rejected
Rejected
Accepted
Rejected
Profitability
Size
Investment (dAt/At)
Rejected
Rejected
Rejected
Rejected
Investment (Vt/At)
Accepted
1
Variable
Profitability
Size
Investment
(dAt/At)
Rejected
Rejected
Rejected
Source: Developed for this research
Similar to the average coefficient method, the group coefficient method has identified
firms characteristics which influence the dividend behaviour of firms. It is observed
that testing null hypotheses applying the group coefficient method results in less null
hypotheses being accepted (see table 5.17 and 5.20). However, the rest of this study, in
the spirit of Fama and MacBeth (1973) and Fama and French (2001), will apply the
average coefficient method to predict the dividend behaviour and present the propensity
to pay dividends of listed firms.
Hosmer and Lemeshow Test for goodness of fit was undertaken to ensure that this
study is consistent with the methodology (average coefficient) of Fama and French. The
results of this analysis are presented in table 5.21.
204
DBA Thesis: Malinee Ronapat
Table 5.21: Hosmer and Lemeshow Test (Goodness of Fit)
Step A
Step B
Step C
Hosmer and Lemeshow Test
Chi-square df
46.587 8
C1
49.184 8
C2
92.958 8
C1
99.265 8
C2
130.946 8
C1
130.742 8
C2
Sig.
0.000
0.000
0.000
0.000
0.000
0.000
Source: Developed for this research
Note: Step A covers 1991 to 1996. Step B covers 1997 to 2002. Step C covers 1991-2002. C 1 and C 2
controlled two and one investment variable(s) respectively. The statistics are significant at the 0.05 level.
The table presents the chi-square of the Hosmer and Lemeshow test and indicates that
the data was well fit in every group and step (0.000). This infers that the data which
was collected before, during and after the crisis was sufficient for predicting or
explaining the dependent variable (payers/non-payers) (Tabachnick and Fidell 2001).
Therefore, it is suggested that the logit regression model could be a valuable tool for
predicting and identifying firms that are likely to pay dividends.
5.6
THE PROPENSITY TO PAY DIVIDENDS (DESCRIPTIVE STATISTICS
EVIDENCE)
Fama and French (2001) suggested that changes in firm characteristics may not be the
only factor affecting the payment of dividends. The researchers suggested that given the
characteristics of firms, they will be less likely, rather than unlikely, to pay dividends
(Fama and French 2001). Therefore, it is important to investigate changes in propensity
to pay dividends of listed firms since 1991. Three factors are considered to identify the
propensity to pay dividends. Yt is a proxy for common stock earnings (representing the
profitability), Et represents earnings before interest that exceed investment outlays and
dAt is the change in assets (representing investment opportunities and asset growth)
(Fama and French 2001).
205
DBA Thesis: Malinee Ronapat
A firm’s propensity to pay dividends is expressed by a time series plot in figure 5.7.
This figure shows the percentage of dividend payers among (1) firms with positive
common stock earnings (Y>0) (2) firms with negative common stock earnings (Y<0)
(3) firms with earnings before interest that exceed investment outlays (Et>dAt) and (4)
firms with earnings before interest which are lower than investment outlays (Et<dAt)
(Fama and French 2001). Firms with earnings before investment, which are higher than
the growth in their assets have low investment opportunities and are low growth firms.
Likewise, firms with earnings before investment which are lower than changes in their
assets are considered growth firms (Fama and French 2001).
Figure 5.7 indicates that most firms with positive common stock earnings (83.93
percent) paid dividends in 1991 and 100 percent of firms with negative earnings also
paid dividends. In addition, 91.7 percent of firms with Y>0 paid dividends in 1992
although this figure declined after 1992. This figure fell to its minimum level in 1998,
when only 38.19 percent of firms with positive earnings paid dividends. Furthermore,
in 1999, 13.10 percent of firms with negative earnings paid dividends. Importantly, the
percentage of firms with positive earnings that paid dividends almost doubled from
39.19 percent in 1998 to 56.7 percent in 2002. This analysis suggests that firms with
positive earnings exhibited a small decline in the propensity to pay dividends before the
Asian Economic Crisis (96.95 percent to 91.26 percent). During the crisis, (1997 to
1998), firms with Yt>0 exhibited a substantial decline in the propensity to pay a
dividends (89.92 percent in 1997 to 39.19 percent in 1998). It appears that firms with
positive earnings were not willing to pay dividends during the crisis. Firms with
negative earnings were also less willing to pay dividends during the crisis. However
after the crisis, the propensity to pay dividends these groups has increased. The
propensity to pay of firms with Yt>0 increased from 39.19 percent in 1998 to 56.7
percent in 2002, while the propensity to pay of firms with Yt<0 increased from 5.76
percent in 1999 to 7.78 percent in 2002. In short, the firms in the sample are more
206
DBA Thesis: Malinee Ronapat
willing to pay dividends if they have positive earnings, or negative earnings after the
crisis. This finding was also supported by the descriptive statistics.
Similarly, before the crisis, firms with earnings before interest which are higher, or
lower than changes in their assets (investment) exhibit a slight decline in the propensity
to pay dividends from 76.25 percent to 75.43 percent (when Et>dAt) and 88.59 percent
to 83.25 percent (when Et<dAt). During the crisis, firms with Et>dAt and Et<dAt
exhibited a significant decline in the propensity to pay a dividends from 69.01 percent
to 28.53 percent and 69.66 percent to 24.36 percent respectively. After the crisis, both
types of firms exhibit an increase in the propensity to pay dividends. The propensity to
pay of firms with earnings higher than investment increased from 30.2 percent in 1999
to 47.52 percent in 2002. Firms with lower earnings than investment, payers of this
group (Et<dAt) accounted for 26.62 percent in 1998 and this increased to 40.72 percent
in 2002 (figure 5.7). These findings suggest that the Asian Economic Crisis also
influenced the propensity to pay dividends of listed firms.
207
DBA Thesis: Malinee Ronapat
Figure 5.7: Percentage of Payers Among Firms with (1) Positive Earnings (2)
Negative Earnings (3) Earnings above Investment, and (4) Earnings
Below Investment
Percentage of Payers Among Firms with (i) positive earnings, (ii) negative
earnings, (iii) earnings above investment, and (iv) earnings below investment
Y>0
Y<0
E>dA
E<dA
120.00
100.00
Percent
80.00
60.00
40.00
20.00
0.00
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
Year
Source: Developed for this research
In summary, the results suggest that irrespective of whether firms have high or low
profitability and high or low investment opportunities, they demonstrated a lower
propensity to pay dividends before the crisis. The propensity to pay dividends of these
types of firms, declined significantly during the crisis, although this has gradually
recovered since the downturn. However, the percentage of payers with positive
earnings in recent years (2002) has been lower than the figure which was observed
before the crisis. This indicates that firms with high profitability and investment
opportunities now have a lower propensity to pay dividends.
208
DBA Thesis: Malinee Ronapat
5.7
CHANGING
CHARACTERISTICS
AND
PROPENSITY
TO
PAY
DIVIDENDS (LOGIT REGRESSIONS)
This section focuses on the quantitative analysis which is used to confirm changes in
the characteristics of firms and the propensity to pay dividends. The analysis will
estimate the probability that firms paid dividends between 1991 and 1996 (pre-crisis),
with given characteristics (profitability, investment opportunities and size). The next
step will be to apply these probabilities (those observed during 1991-1996) to the
sample firms with their observed characteristics from 1997 to 2002 (post-crisis). This
will estimate the expected percentage of dividend payers. Any variation in the expected
percentage of payers after 1996 will result from changes in characteristics of the sample
firms (Fama and French 2001). The actual percentage of payers is the observed number
of payers, to the total number of firms in the sample from 1991 to 2002. Any
differences between the expected percentage of payers (calculated from the base
probabilities in 1991-1996) and actual percentage of payers will represent a change in
the propensity to pay dividends. Positive variation between the expected and actual
percentage of payers will imply a decline in the propensity to pay dividends, while
negative variation will imply an increase in the propensity to pay dividends.
Expected is the expected percentage of payers. Actual is the actual percentage of firms
that pay dividends in time t. Expected – Actual is the change in propensity to pay
dividends over time. Decline in the expected percentage of payers represents changes in
the characteristics of the payers since the base period.
5.7.1
REGRESSION ESTIMATES
Table 5.22 presents the expected percentage of dividend payers which was obtained by
applying the average coefficient to the proxies of the characteristics of the sample firms
209
DBA Thesis: Malinee Ronapat
from 1991 to 2002. Two sets of results are presented so that the methodology is
consistent with Fama and French (2001).
Firstly,
proxy for profitability is Et/At while the two proxies for investment
opportunities dAt/At and Vt/At.
Second, one proxy biased for profitability but only one controlled proxy is used for
investment opportunities.
While the other proxies for profitability, investment opportunities and size are constant
over time, Vt/At is variable (Fama and French 2001, p. 15). Fama and French (2001)
stated that the elimination of Vt/At is due to the unstable and unclear tendency of the
Vt/At trend (see table 5.8) which results from three factors:
(1) increasing profitability of assets in place
(2) more profitable or more abundant expected investments, or
(3) lower discount rates for expected cash flows (Fama and French 2001, p. 15)
Therefore, Vt/At results in an over-statement of the level of investment opportunities
and the level of understating of the propensity to pay dividends.
210
DBA Thesis: Malinee Ronapat
Table 5.22: Logit Regression Estimates of the Effect of Changing Firm
Characteristics and Decline in the Propensity to Pay Dividends
19911996
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
Firms
Payers
302
225
255
289
322
345
378
394
389
385
369
377
388
256
192
230
245
275
290
306
281
110
115
130
153
179
Vt/At and dAt/At
Expected
Actual Expected - Actual
84.7
85.3
90.2
84.8
82.8
84.1
81.0
71.3
28.3
29.9
35.2
40.6
46.1
88.8
98.8
90.7
97.6
98.3
97.9
62.3
86.3
81.0
85.3
88.8
91.7
3.5
8.6
5.9
14.8
14.2
16.9
-9.0
58.0
51.1
50.1
48.2
45.6
dAt/At
Actual
Expected
Expected
- Actual
85.3
90.2
84.8
82.8
84.1
81.0
71.3
28.3
29.9
35.2
40.6
46.1
99.6
99.2
93.0
98.5
98.5
97.6
61.8
86.9
80.8
86.4
88.8
93.0
14.3
9.0
8.2
15.7
14.4
16.6
-9.5
58.6
50.9
51.2
48.2
46.9
Source: Developed for this research
Note: To explain why firms pay dividends, all firms for 1991-1996 are used to estimate the logit
regression. The explanatory variables are Et/At (for profitability), dAt/At (for investment opportunities),
Vt/At (for investment opportunities) and SET (for size). SETt is the percentage of SET firms with the
same, or a lower rate of market capitalization. The firm column presents the number (or average number)
of firms in the sample in time t. Payers is the number (or average number) of sample firms that paid
dividends in time t. The actual percentage is the percentage of payers to the number of sample listed
firms. Expected is obtained from applying the average logit regression coefficient for 1991-1996 to the
values of the explanatory variables mentioned above for each firm in every year from 1991 to 2002. This
summing over firms and divided by number of firms then multiply by 100. The changes in the percentage
of the expected column represents the effect of changing characteristics of the listed firms over time.
Expected – Actual measures the effect of the propensity to pay dividends of the listed firms. Two sets of
results are presented, one with Vt/At and dAt/At as a measurement of investment opportunities. Another
only uses dAt/At as a measurement of investment. See appendix G and H for details.
This study conducts regression by using Et/At, dAt/At, Vt/At and SETt first as the factors
for explaining the expected percentage of payers in each year from 1991-2002. The
base years (1991-1997) are applied to each of the twelve years (1991-2002) to estimate
the expected percentage of payers. Any change in the expected percentage of payers in
any year is due to changes in characteristics of the sample. Before the Asian Economic
Crisis (the base year period 1991-1996), the expected percentage of payers is above 90
percent for most years. This implies that firm characteristics have changed little since
211
DBA Thesis: Malinee Ronapat
the base years. When the expected percentage of payers is subtracted from the actual
percentage of payers, the propensity to pay dividends of the samples in the pre-crisis
period is high. However, the figure tends to increase slightly over time from a variation
of 3.5 percent in 1991 and 16.9 percent in 1996. During the Asian Economic Crisis
(1997), most of the samples exhibit a decline largely in the characteristics from the base
years (62.3 percent expected as payers). The number of actual payers rose to 71.3
percent in the market. Therefore, in 1997, the propensity to pay dividends of the
samples increased (differences -9.0 percent). This may be due to (1) an expectation by
the firms that the crisis will not last long or (2) the result of policies by firms aimed at
stopping investors from panicking and raising the level of confidence. However, the
Asian Economic Crisis was long lasting and severally affected the Thai stock market.
In 1998, the characteristics of most of the sample firms tended to be similar to the
characteristics of payers in the base years (86.3 percent). However, firms were not
willing to pay dividends out of their earnings. This resulted in a strong divergence
between the expected percentage of payers and the actual percentage of payers (59
percent). Therefore, the sample of listed firms had the lowest propensity to pay in 1998.
After the crisis (1999-200), the number of expected papers increased from 81 percent in
1999 to 91.7 percent in 2002. This implies that the firm characteristics which were
present in the base years returned. However, the propensity to pay dividends was large
in 1998 (59 percent). Firms were not willing to pay dividends even though they had the
ability to pay (given these characteristics). After 1998, the difference between the
expected and the actual percentage of payers fell from 59 percent in 1998 to 45.6
percent in 2002. This indicates that a recovery is under way and the propensity to pay
dividends of the sample firms is increasing. This raises the expectation that (1) the base
year characteristics may return and (2) the propensity to pay dividends of listed firms
may rise.
Fama and French (2001) omitted the Vt/At ratio because they anticipated that it would
reduce the percentage of expected payers. This would reduce the contribution to the
decline in the percentage of dividend payers (Fama and French 2001). After discussing
the Vt/At ratio, this study found that the expected percentage of payers increased. Fama
212
DBA Thesis: Malinee Ronapat
and French (2001) concluded that ‘with or without Vt/At, the regression approach bares
the tracks of a potentially elusive phenomenon’ (Fama and French 2001, p. 17). This
was supported by the second set of findings of Fama and French (2001) and by this
study. Table 5.22 indicates there is no significant difference in the findings of this
analysis if two variables are controlled, or if one variable is controlled. Before the
crisis, there was minimal change in the characteristics of firms. Therefore, only a
slightly lower propensity to pay dividends was recorded. During the crisis, more firms
were willing to pay dividends because firms characteristics differed from those
observed during the base years. Therefore, during the crisis sample firms tended to have
a higher propensity to pay dividends. After the crisis, the characteristics of firms were
similar to the base years and the propensity to pay dividends rose slightly from a
variation of 50.9 percent in 1999 to 46.9 percent in 2002.
The evolution of Expected Percentage column (see table 5.22) indicates that the impact
of characteristics of firms on the percentage of payers was below 90 percent in only
1991. From 1992 to 1996, the Expected Percent column increased to over 90 percent
and has been above this level ever since. In 1997, the characteristics of payers which
are similar to the base period fell strongly to 62.3 percent. However, this figure has
increased and the characteristics of firms were similar to the base years after 1998.
Recently (2002), the Expected Percentage stood at 91.7 percent. This figure suggests
that the characteristics of the payers which were observed during the base period have
returned.
The logit regression of Fama and French (2001) indicated that changes in firm
characteristics were maintained over time. The results of this analysis are not consistent
with this result. This study indicates the characteristics of payers which existed
established during the base years have largely been maintained throughout the period of
investigation. These characteristics changed only a short period during the Asian
Economic Crisis.
Logit regression suggests that changes in firm characteristics and the propensity to pay
dividends has resulted in the partial disappearance of dividends from Thailand’s capital
213
DBA Thesis: Malinee Ronapat
market, before, during and after the crisis. However, the evidence suggests that before
the crisis, there was slight change in the characteristics of firms and the propensity to
pay dividends. During the crisis, large changes in firm characteristics were observed
and the propensity to pay dividends remained relatively high. After the crisis,
propensity to pay the dividends has slowly risen, although it is below the figures
observed before the crisis. This outcome appears to result from changes in the
characteristics of firms.
5.8
CONCLUSION
From a peak of 90.6 in 1992, the percentage of dividend payers in the Stock Exchange
of Thailand currently stands at 46.4 percent. The decline is partly due to the fact that
many newly listed firms and mature firms are adopting the characteristics of firms
which do not pay dividends. Changes in the characteristics of firms are also an
important factor leading to the observed decline in the payment of dividends. The
methodology which has been adopted for this study suggests that the lower propensity,
or reduced willingness to pay dividends among listed firms is common. The next
chapter will summarise findings of this study and discuss its implications for future
research.
214
DBA Thesis: Malinee Ronapat
C
HAPTER
6.1
6
CONCLUSIONS AND IMPLICATIONS
INTRODUCTION
The previous chapter presented the results of the analysis of the characteristics of listed
firms, the propensity to pay dividends and their changes over time in Thailand. This
chapter is a conclusion to this study. The objectives of this chapter are to summarise (1)
the characteristics of firms and propensity to pay dividends of firms which pay
dividends (2) and indicate how the findings of this study can be implemented by
investors, analysts, regulators and (3) to suggest areas where further research is needed.
The chapter is presented in six sections. Section 6.1 discusses the objectives and
structure of the chapter 6. Section 6.2 addresses the research hypotheses, questions and
problem. The implications of this study and its contribution to the body of knowledge
are presented in sections 6.3 and 6.4. Section 6.5 discusses the limitations of the study
and finally, section 6.6 provides several recommendations for further research.
215
DBA Thesis: Malinee Ronapat
Figure 6.1: The Structure of Chapter Six
6.1 Introduction
6.2 Conclusion
6.2.1 Research hypotheses and
6.2.2
Research problem
questions
6.3 Implications
6.3.1
Implications for theory
6.3.2
Implications for policy
6.3.3 Implications for investment
practices
6.4 Contributions of the study
6.5 Limitations of the study
6.6 Implications for further research
Source: Developed for this research
216
DBA Thesis: Malinee Ronapat
6.2
CONCLUSIONS
This section summarises the findings of the analysis of data which was conducted in
chapter 5 and addresses the research hypotheses, questions and the research problem.
6.2.1 CONCLUSIONS – RESEARCH HYPOTHESES AND RESEARCH
QUESTIONS
Research Questions and Hypotheses
Several research questions and hypotheses were developed to answer the research
problem which was developed in chapter 3. Two sets of research questions were also
developed. The first set of research questions related to the literature review, while the
second set related to the variables which were used for developing the formula which
predicts the dividend behaviour of firm. Both sets of research questions and hypotheses
are presented below.
6.2.1.1 CONCLUSIONS RELATING TO THE CHAPTERS
Research question # 1 is restated below:
What is the background and performance of the Stock Exchange of Thailand and
its listed firms?
This research question was answered in chapter 2. The history of SET has been short
although it provides a vital role in supporting the economic growth of the nation.
However, the performance of SET has varied through time and swings according to
several internal and external factors. The period which the study focuses upon (19902002) has been associated with major fluctuations in SET’s index. Indeed the index
rose to the highest point at 1682.9 in 1993 and fell to 269.19, its lowest level in its
history, in 2000. The background and performance of SET is a key factor which
217
DBA Thesis: Malinee Ronapat
explains why the payment of dividends disappeared and reappeared in Thailand during
the period of investigation.
Research question # 2 is asked:
How have researchers investigated dividends and their patterns throughout the
world?
This research question was answered in chapter 3. The nature of dividend policies and
theories were discussed in chapter 3 together with the relationships between dividends
and investment, firm characteristics, and the propensity to pay dividends of firms. Fama
and French (2001) raised awareness of the issues in the world’s capital markets with
their study on disappearing dividends and its relation to the characteristics of firms and
the propensity to pay dividends. A number of researchers and analysts around the world
have shown interest in this issue and have conducted studies aiming to support, or reject
the suggestions of Fama and French.
Research question # 3 stated that:
What is the appropriate methodology for collecting and analysing data?
Research question # 3 is answered in chapter 4. The data was obtained from the Stock
Exchange of Thailand and was divided into three periods of time (1) before the crisis
(1991-1996), (2) after the crisis (1997-2002) and (3) the whole period of investigation
(1991-2002). Descriptive statistics have been used by researchers and analysts to
initially understand the phenomenon. ANOVA was also used to confirm whether the
means of a variety of dividend groups were statistically different to each another. Logit
regression analysis was also used to obtain a formula to predict the percentage of
payers in the capital market for each year between 1990 and 2002. This method of
regression is commonly used in the area of finance when the dependent variable has
two possible outcomes, occurrence and non-occurrence (payers and non-payers).
218
DBA Thesis: Malinee Ronapat
Research question # 4 asked:
Does the data from Thailand support a model?
Research questions # 4.1-4.3 asked:
What are the characteristics of the dividend payers?
Do these characteristics change over time?
How has the propensity to pay dividends of firms changed over time?
The above research questions were answered in chapter 5. The data collected from the
Stock Exchange of Thailand was used to calculate the values for six ratios which were
used to represent three characteristics of firms. Et/At is the ratio of aggregate earnings
before interest to aggregate assets (profitability), Yt/BEt is the ratio of aggregate
common stock earnings over aggregate book equity (profitability), dAt/At is the growth
rate in assets (investment opportunities), Vt/At is the aggregate market value to the
aggregate book value of assets (investment opportunities), At is the total asset (size),
Lt/At is the ratio of total liabilities to total asset (size).
When applying logit regression to the data, the formula predicted the percentage of
dividend payers more accurately for the period before the crisis, while it was less
accurate after the crisis. Logit regression suggested there were changes in the
characteristics of the payers and the propensity to pay dividends of these payers given
the characteristics which were calculated from the predicted value. The changes in
characteristics were minimal before and after the crisis but very large during the crisis
in 1997. The characteristics of firms which were observed during the base years (19911996) appeared to re-occur after the crisis. An important factor was that the propensity
219
DBA Thesis: Malinee Ronapat
to pay dividends of firms is lower after the crisis than before the crisis. This finding is
consistent with the literature.
Research question # 5 stated the following:
How could a model to predict payers be implemented?
The answer to this research question is discussed in section 6.3. The model to predict
the percentage of payers given their characteristics, could be implemented in a capital
market such as SET by regulators, analysts and investors.
6.2.1.2 CONCLUSIONS RELATED TO TESTING THE MODEL
This section is a conclusion on the relationship between firm characteristics and the
probability that firms will pay dividends to predict the dividend behaviour of listed
firms.
Research question # 4, as stated earlier, has several sub-questions (research questions
#4.1-4.3) that lead to the hypotheses which were developed. To answer research
question 4, and its subquestions, these hypotheses were tested and analysed. These
hypotheses were answered in chapter 5 by using logit regression in step, A (1991 to
1996), B (1997 to 2002) and C (1991-2002). Every dividend group was analysed twice,
firstly with one investment variable and secondly with two investment variables. The
same methodology was applied by Fama and French (2001). These researchers used
logit regression with average coefficient and group coefficient techniques.
Three firm characteristics were analysed (1) profitability, (2) investment opportunities
and (3) size. Variable Et/At (the ratio of aggregate earnings before interest to aggregate
assets for profitability) as a proxy for profitability, dAt/At (the growth rate of assets for
investment opportunities) as a proxy for investment opportunities, Vt/At (the ratio of
aggregate market value to the aggregate book value of assets for investment
opportunities) as a proxy for investment opportunities and SET (the percent of sample
220
DBA Thesis: Malinee Ronapat
firms with lower or equal market capitalisation rate for size) as a proxy for size. All of
the variables were statistically significant for predicting payers, when applying logit
regression in every period (controlling two then one investment variables). However,
the results were mixed when different steps and methods are applied (see Chapter 5).
Research question # 4.1.1 stated the following:
Does the dividend payment behaviour depend on the profitability of the
listed firms?
Research question # 4.1.1 was tested with the research hypothesis # 1 which is:
Ho :
There is no significant relationship between the profitability of the firms
and the probability that the listed firms will pay dividends.
Ha :
There is a significant relationship between the profitability of the firms
and the probability that the listed firms will pay dividends.
Table 6.1 and 6.2 present the answers to research problem #4.1.1 in three different
period of times.
Table 6.1: Results of Hypotheses Testing Following Fama and French (2001)
Method of Averaging Coefficient (Profitability)
Periods
Before the Crisis
(Step A) 1991-1996
Controlled
Variables
Characteristics
2 Variables
Profitability
1 Variable
During and After the Crisis
(Group B) 1997-2002
2 Variables
2 Variables
Rejected
Yes
Rejected
Yes
Rejected
Yes
Profitability
Rejected
Yes
Rejected
Yes
Profitability
Rejected
Yea
Profitability
1 Variable
Answer to
Question #
4.1.1
Profitability
Profitability
1 Variable
The Whole Period of
Investigation
(Group C) 1991-2002
Null
Hypothesis
Source: Developed for this research
Note: ‘Yes’ means that dividend payment behaviour depends on the profitability of the listed firms
221
DBA Thesis: Malinee Ronapat
Table 6.2: Results of Hypotheses Testing Using Group Coefficient (Profitability)
Periods
Before the Crisis
(Step A) 1991-1996
Controlled
Variables
Characteristics
2 Variables
Profitability
1 Variable
During and After the
Crisis
(Stpe B) 1997-2002
2 Variables
2 Variables
Rejected
Yes
Rejected
Yes
Rejected
Yes
Profitability
Accepted
No
Rejected
Yes
Profitability
Rejected
Yes
Profitability
1 Variable
Answer to
Question
#4.1.1
Profitability
Profitability
1 Variable
The Whole Period of
Investigation
(Step C) 1991-2002
Null
Hypothesis
Source: Developed for this research
Note: ‘Yes’ means that dividend payment behaviour depends on the profitability of the listed firms
The analysis found that profitability (Et/At: the ratio of aggregate earnings before
interest to aggregate assets) is a significant factor for predicting the dividend behaviour
in all time periods, with two, or one controlled investment variable. The finding was
consistent with Fama and French (2001). When applying the group coefficient
technique, Et/At was not a significant factor for predicting the dividend behaviour from
1997 to 2002 controlling one investment variable.
Research question # 4.1.2 stated the following:
Does the dividend payment behaviour depend on the investment
opportunities of the listed firms?
Research hypothesis # 2 was generated from research question 4.1.2:
Ho :
There is no significant relationship between the investment opportunities
of the firms and the probability that the listed firms will pay dividends.
Ha :
There is a significant relationship between the investment opportunities of
the firms and the probability that the listed firms will pay dividends.
Two measurements were used for controlling investment opportunities: dAt/At (the
growth rate of assets for investment opportunities) and Vt/At (the ratio of aggregate
222
DBA Thesis: Malinee Ronapat
market value to the aggregate book value of assets). Tables 6.3 and 6.4 present the
answers to the null hypothesis and research question 4.1.2.
Table 6.3: Results of Hypotheses Testing Following Fama and French (2001)
Method of Averaging Coefficient (Investment Opportunities)
Periods
Before the Crisis (Step A)
1991-1996
Controlled
variables
2
Variables
Null
Hypothesis
Answers to
question
#4.1.2
Investment (dAt/At)
Rejected
Yes
Investment (Vt/At)
Accepted
No
Accepted
No
Investment (dAt/At)
Rejected
Yes
Investment (Vt/At)
Accepted
No
Accepted
No
Investment (dAt/At)
Rejected
Yes
Investment (Vt/At)
Accepted
No
Accepted
No
Characteristics
Investment
(dAt/At)
1 Variable
During and After the
Crisis (Step B) 1997-2002
2
Variables
Investment
(dAt/At)
1 Variable
The Whole Period of
Investigation (Step C)
1991-2002
2
Variables
Investment
(dAt/At)
1 Variable
Source: Developed for this research
Note: ‘Yes’ means that dividend payment behaviour depends on the investment opportunities of the
listed firms.
Table 6.4: Results of Hypotheses Testing Using Group Coefficient Technique
(Investment Opportunities)
Periods
Before the Crisis (Step A)
1991-1996
Controlled
Variables
2
Variables
Characteristics
Null
Hypothesis
Answers to
Question
#4.1.2
Investment (dAt/At)
Rejected
Yes
Investment (Vt/At)
Rejected
Yes
Rejected
Yes
1
Variable
During and After the
Crisis (Step B) 1997-2002
2
Variables
Investment
Investment (dAt/At)
Accepted
No
Investment (Vt/At)
Rejected
Yes
Rejected
Yes
Investment (dAt/At)
Rejected
Yes
Investment (Vt/At)
Accepted
No
Rejected
Yes
1
Variable
The Whole Period of
Investigation (Step C)
1991-2002
2
Variables
1
Variable
Investment
Investment
Source: Developed for this research
Note: ‘Yes’ means that dividend payment behaviour depends on the investment opportunities of the
listed firms.
223
DBA Thesis: Malinee Ronapat
The results suggest that by controlling two investment variables, dAt/At is a significant
variable for predicting the dividend behaviour of firms in every period. However, Vt/At.
was found to be insignificant. When controlling one investment variable, the results
were mixed for Vt/At, especially when using the average coefficient technique. It
appears that Vt/At was not related to the dividend behaviour of firms in any time period
which was analysed. However, the analysis with logit regression which used a group
coefficient suggested that there is a relationship between the dividend behaviour of
firms and dAt/At (growth rate in assets) in every time period.
Table 6.5 presents the answer to research question # 4.1.3.
Research question # 4.1.3 stated that:
Does the dividend payment behaviour depend on the size of the listed firms?
Research hypothesis # 3 was developed to answer research question #4.1.3:
Ho :
There is no significant relationship between the size of the firms and the
probability that the listed firms will pay dividends.
Ha :
There is a significant relationship between the size of the firms and the
probability that the listed firms will pay dividends.
Size is measured in term of SET or the percentage of firms with identical or less market
capitalisation. Size was found to be a significant variable for predicting the dividend
behaviour of firms in most time periods. Table 6.5 and 6.6 summarise the answers to
research questions number 4.1.3.
224
DBA Thesis: Malinee Ronapat
Table 6.5: Results of hypotheses testing following Fama and French (2001) method
of averaging coefficient (size)
Periods
Before the Crisis (Step A)
1991-1996
Controlled
Variables
2
Variables
Characteristics
Null
Hypothesis
Answers to
Question
#4.1.3
Size
Rejected
Yes
Accepted
No
Rejected
Yes
Rejected
Yes
Rejected
Yes
Rejected
Yes
1
Variable
During and After the
Crisis (Step B) 1997-2002
2
Variables
Size
Size
1
Variable
The Whole Period of
Investigation (Step C)
1991-2002
2
Variables
Size
Size
1
Variable
Size
Source: Developed for this research
Note: ‘Yes’ means that dividend payment behaviour depends on the size of the listed firms.
Table 6.6: Results of hypotheses testing using group coefficient method (size)
Periods
Before the Crisis (Step A)
1991-1996
Controlled
Variables
2
Variables
Characteristics
Null
Hypothesis
Answer to
Question
#4.1.3
Size
Rejected
Yes
Accepted
No
Rejected
Yes
Rejected
Yes
Rejected
Yes
Rejected
Yes
1
Variable
During and After the
Crisis (Step B) 1997-2002
2
Variables
Size
Size
1
Variable
The Whole Period of
Investigation (Step C)
1991-2002
2
Variables
Size
Size
1
Variable
Size
Source: Developed for this research
Note: ‘Yes’ means that dividend payment behaviour depends on the size of the listed firms.
The logit regression, which applied an average coefficient and the regression with a
group coefficient both suggested that the dividend behaviour of firms depends on firm
size (as measured by the percent of firms with same or lesser market capitalisation) in
every time period, with the exception of step A (before the crisis) controlling one
investment variable and the null hypothesis is accepted.
Although the results are mixed, every variable in the step that controlled two
investment variables were significant for predicting the dividend behaviour of listed
225
DBA Thesis: Malinee Ronapat
firms in every time period. However, the tests for goodness of fit suggested that every
model which was developed from the samples is significant for predicting the
percentage of payers in the market (chapter 5).
The logit regression model which was developed identified the following
x
There was slight change (or decline) in the characteristics of the dividend payers
even before the Asian Economic Crisis in 1997.
x
The characteristics which were present during the base years changed
significantly (during the crisis) in 1997 but resumed during the period 19982002.
x
The characteristics of the dividend payers during the base years (1991 to 1996)
and whole period of investigation (1991 to 2002), were slightly different.
x
The predicted percentage of payers is, in almost every year, higher than the
actual percentage of payers in the market.
x
The decline in the propensity to pay dividends of the payers is evidenced even
before the crisis.
x
The divergence between the actual percentage of payers are the predicted
percentage of payers increased for every year after 1997 (the actual percentage
was always lower than the expected percentage).
x
Declining in the propensity to pay dividend of payers was large after the crisis.
x
The divergence between the actual and the expected percentage of payers has
tended to decline in recent years and indicates that the market is recovering.
x
The empirical findings of this study are consistent with the literature.
x
The model could be used for predicting the percentage of dividend payers in the
market.
The reason why listed firms in Thailand continue to pay dividends or resume paying
dividends after a major economic downturn is still a puzzle. Several researchers have
studied this issue although there is a consensus as to why some firms pay dividends and
some do not and why some investors prefer dividends and some do not. Trends such as
the disappearance of dividends in the short-term (before crisis), the long-term (19902002) and the reappearance of dividend payments of firms (1998 till current) could be
226
DBA Thesis: Malinee Ronapat
explained by the general advantages and disadvantages of dividends as discussed
below.
Lintner (1956) observed that managers are reluctant to omit dividends. If firms are
reluctant to omit dividends, especially, during bad times, the benefit of dividends
increase, because investors do not need to sell stocks to receive returns from their
investment. Therefore, reliability in the payment of dividends will make investors feel
that their assets are secure, while capital gains are likely to fluctuate. Dividends could
therefore be viewed as a cushion during bad times, or when stock prices fall (Linter
1956, Gwilym, Morgan and Thomas 2000).
Dividends are useful when the market has many high-tech stocks which do not pay
dividends but experience high growth rates. Dividends also provide investors with
some control over the management of firms as they can demand that dividends should
be paid. As stated earlier, cash dividends are real money and can provide investors with
an indication of a firm’s performance. Fama and French (2001) suggested that firms
with positive earnings are more likely to pay dividends, while young firms are not.
Therefore, the payment of dividends does not disclose a firm’s poor performance in the
last accounting period, although it appears sometimes that earnings do.
Researchers have also suggested that dividend payments provide information to
investors on the current and future performance of the firm, external financing, as well
as the true value of firms. However, forgoing dividends should actually provide
investors with higher returns in the long-run when compared with other types of
investment (Bhattacharya 1979, 1980; Asquith and Mullins 1983; John and Williams
1985; Miller and Rock 1985; Richardson, Sefcik and Thompson 1986; Healey and
Palepu 1988).
In Thailand, the SEC (2002b) has surveyed investors understanding of investment, their
behaviours and the information they need to undertake informed decisions. SEC found
that 26 percent of the sample investors prefer and are interested in firms which pay
dividends, while only 16 percent prefer firms that provide large capital gains. The
227
DBA Thesis: Malinee Ronapat
research also stated that, when making investment decisions, Thai investors perceive
that dividend history and financial statements are more important than the investment
opportunities of the firms (SEC 2002b). Knowing that dividend payment received is
taxed at the rate of 5.71 percent in Thailand but capital gain is tax-free (see section 5.4
in Chapter 5), Thai investors seem to be fond of dividend payments. This investment
culture of Thai investors could be a result of reappearing dividend pattern after the
Asian Economic Crisis in 1997. Whether this culture is a country factor and how it
affects the Thailand’s capital market, is an issue for future research.
However, deciding whether to invest in dividend-paying stocks or mutual funds, many
factors such as taxation, risk and the need for income will be considered. Therefore, the
phenomenon of disappearing dividends could be related to the above issues. Several
explanations for the disappearance of dividends could include the following.
In many countries, including Thailand, capital gains are usually taxed less than
dividends. Investors can decide to pay tax on dividends or sell their shares and pay tax
on capital gains.
From the perspective of the firm, managers can be more flexible when they report their
earnings if they do not pay dividends. As dividend payment is a reduction of retained
earnings, managers of non-payer firms are more likely to have higher retained earnings
in their balance sheets. Furthermore, issuing stocks are costly, while borrowing is less
expensive as the interest from borrowing is tax deductible (Banker and Wulgler 2003b).
The declining propensity to pay dividends of the listed firms could result from six
factors according to Baker and Wugler (2003b).
1. Agency: Dividends are viewed as a tool for good corporate governance
(Easterbrook 1984) to keep managers close to capital markets. Dividends could
be disappearing due to a new corporate governance mechanism which help
reduce the agency problems between manager and investors. Therefore the need
to control agency costs through dividends is diminish (Fama and French 2001,
Banker and Wulgler 2002, 2003a, 2003b).
228
DBA Thesis: Malinee Ronapat
2. Information Asymmetry: Dividends are a signal of the profitability of the
firm, and could be used as a tool for predicting the performance of the firm and
its stocks (Ross 1977; Battacharya 1979; John and Williams 1985; Miller and
Rock 1985). However, as information transmission rates are relatively fast and
accurate nowadays, the gap between managers and investors is narrowed and
prediction models (such as Dividend Discount Model) may be used less often
(Baker and Wulgler 2002, 2003a, 2003b). The role of dividends in transmitting
information is reduced.
3. Stock Options: Managers could hold options which are not dividend protected
and this could create personal incentives for not paying dividends. Further
research has suggested that options correlate with lower dividend payments
(Lambert, Lanen and Larcker 1989; Jolls 1998; Weisbenner 1999; Fama and
French 2001; Baker and Wulgler 2002, 2003a, 2003b).
4. Clientele Equilibrium: Baker and Wurgler (2003b) explained that the decline
in the payment of dividends could result from changes in the demand for
dividends from the firms by investors.
5. Catering: Baker and Wurgler (2003b) suggested a new theory on the
disappearance of dividends. They found that dividend payments depend on the
demand from investors. Therefore, managers will pay dividends if their share
prices are perceived to be at a premium and omit dividends if their share prices
are discounted (Baker and Wurgler 2002).
6. Slow Learning about Taxes: Baker and Wurgler (2003b) stated that managers
may have been slow to understand the tax disadvantages of paying dividends.
After fully understanding these disadvantages, their propensity to pay dividend
declined (Grinblatt and Titman 1998).
Several of these causes and effects could be used to explain the pattern of disappearing
and reappearing dividends on the Thai stock market of Thailand. However, this study
concentrates on the characteristics of firms, changes in these characteristics and the
propensity to pay dividends of listed firms over time.
229
DBA Thesis: Malinee Ronapat
CONCLUSIONS – THE RESEARCH PROBLEM
Research Problem
The research problem is restated as follows:
‘How do the characteristics of publicly traded firms and their propensity to pay
dividends influence the payment of dividends and reason for their disappearance
from the stock market of Thailand?
The payment of dividends disappeared during the crisis and re-appeared after the Asian
Economic Crisis. In addition, changes in the characteristics of payers is evident before
and after the crisis. It is more important, however, that listed firms have shown a lower
propensity to pay dividends before and after the crisis. The percentage of dividend
payers in the market declined irrespective of changes in the characteristic of firms.
Figure 6.2 shows the relationships between the findings of the logit regression,
ANOVA and the descriptive statistical techniques.
The figure below shows that firms started to have lower propensity to pay dividends
and changed characteristics before the crisis and the Asian Economic Crisis triggered
firms to have even lower propensity to pay dividends. These results in the decline in
percentage of payers continue paying dividends, non-payers to start paying dividends.
The percentage of never paid firms increased, as well as the new lists that tend not
paying dividends. All of these result in the decline in the percentage of dividend payers
in the market.
230
DBA Thesis: Malinee Ronapat
Figure 6.2: Relationships of Logit Regression, ANOVA and Descriptive Statistics
Findings
FIRMS HAVE LOWER
PROPENSITY TO PAY
DIVIDENDS
% PAYERS CONTINUE
TO PAY DECLINED
RESULTS IN
FIRMS CHANGED
CHARACTERISTICS
ASIAN ECONOMIC
CRISIS TRIGGERED
% NON-PAYER START
TO PAY DECLINED
RESULTS IN
%
PAYER
DECLINED
% NEVER PAID FIRMS
INCREASED
LOWER PROPENSITY
CHANGES TO
CHARACTERISTICS
% NEW LISTS THAT DO
NOT PAY HAS
INCREASED
Source: Developed for this research
6.3
IMPLICATIONS
This section discusses the implications of the findings of this study for theory, policy
and investment practice. The contribution of this study to the body of knowledge is also
discussed.
6.3.1
IMPLICATIONS FOR THEORY
As this research is concerned with the disappearance of dividends from the stock
market of Thailand, it will be compared with the existing literature and will provide
additional support for several of the findings of the Fama and French (2001) study. It
231
DBA Thesis: Malinee Ronapat
will also be relevant to scholars who are interested in the pattern of dividend payment
in developed and developing capital markets. In particular, the findings of this study
provide insight for emerging stock markets such as Thailand.
The findings of this research have implications for countries with similar market
conditions, especially in developing economies such as those of Asian countries.
Therefore, this study will help direct further research.
It could also be used as a teaching tool and add to the knowledge on finance and
investment in the business schools at both undergraduate and postgraduate levels. The
methodology, sample size, analysis and implications could be used for discussion in
classes and students could use this thesis as a guideline for developing their own
research papers.
6.3.2
IMPLICATIONS FOR POLICY
Since this research concentrates on the dividend payment of listed firms in Thailand, its
findings have implications for the Securities Exchange Commission of Thailand (SEC)
and the Stock Exchange of Thailand (SET). Several of these possible implications are
discussed below.
The research also provides a tool to supervise and monitor the characteristics of listed
firms and their dividend behaviour. It will help maintain the safety and soundness of the
financial system and capital market in Thailand. The findings of this research could be
used as a preliminary investigation of the changes in the characteristics of listed firms.
It could help predict the expected number of payers in the market and monitor the
performance of new and existing listed firms.
232
DBA Thesis: Malinee Ronapat
Listed firms could be grouped into three dividend groups as specified by Fama and
French (2001) (payers/non-payers, payer/former payers/never paid) according to their
dividend behaviour. SEC and SET could provide this information to the public to
enhance the quality of information for investors. This will help increase the level of
understanding of investments in different types of securities with various risk-return
trade-offs. The SEC and SET should issue CDs and annual reports to enhance the flow
of information to investors. Therefore, investors will be able to make decisions using
accurate information. The SEC and SET should impose a strict regulation, or a code of
conduct, stating that listed firms must provide at least 5 to 10 years of historical
financial statements and a copy of the company’s annual report to investors free of
charge.
This study could provide a tool for the early detection of an increase, or decrease in the
number of payers and non-payers in the market, an upward or downward trend in the
payment of dividends and changes in the propensity to pay dividend in the next period.
This research could be an ideal report for investors on the pattern of dividend payments
and for the SEC or SET. Furthermore, the analysis of listed firms characteristics is
based on information from the financial statements of the listed firms. It may therefore
increase public awareness of the information which is currently provided by SET and
SEC. This will help enhance the protection and education of investors.
6.3.3
IMPLICATION FOR INVESTMENT PRACTICE
The information provided in the research could also be used as a guide on investment
practice for the different types of stock being traded in the market (e.g. value and
growth stocks, payers and non-payers). It helps provide information on the dividend
behaviour of listed firms with the given characteristics and provides a warning system
for the firms that may or may change their dividends behaviour. This study also
highlights changes in firm characteristics and the propensity to pay dividends. It can be
233
DBA Thesis: Malinee Ronapat
used as a tool for investors in the stock selection process by providing insight on the
expected returns. Furthermore, the research could also aid investors decisions by
suggesting stocks for a hedging strategy or helping to diversify their portfolios.
Investors could shift their funds to different dividend groups (from payers to nonpayers or to former-payers or firms that have never paid dividend) to enhance the
performance of securities in their diversified portfolio and magnify investors’ tax
preferences. This would aid technical and fundamental analysis in the stock market,
thus, increasing efficiency. Importantly, this research could increase confidence and
encourage investors to invest in the capital market in Thailand.
6.4
CONTRIBUTIONS OF THE STUDY
This research has contributed to the level of knowledge on the phenomenon of
disappearing dividends in the following ways.
The logit regression model which was developed to predict the percentage of payers in
the market in a particular year, is important and a significant indicator of changes in
firm characteristics and the propensity to pay dividends of the listed firms. The
descriptive statistical findings of this research also contribute to the literature on
investment and financial management.
The year-to-year analysis of the data provides a clearer picture of the characteristics,
the propensity to pay dividends and the pattern of dividend payments than a
methodology which only focuses on a range of years. A range of years may contain
mixed and misleading signals of the phenomenon.
The findings of this study contribute to the level of knowledge of the phenomenon of
disappearing dividends and the dividend payments pattern in an emerging capital
market such as Thailand. Minimal research has focused on the disappearance of
dividends because the issue is relatively new. Furthermore, most of the relevant studies
234
DBA Thesis: Malinee Ronapat
have obtained data from developed stock markets such as the US and UK. This research
expands and contributes to the literature by concentrating on Thailand’s stock market.
The findings of this research support the view that emerging markets (Thailand) often
follow the same pattern as developed markets (US and UK).
The results of this research indicate that the phenomenon of disappearing dividends is
present in Thailand and it is related to firm characteristics, changes in these
characteristics and the propensity to pay dividends. Furthermore, it is interesting to note
that the propensity to pay has also declined in Thailand. The review of literature noted
that it has also occurred in the US although it has not been empirically explained.
The research also helps confirm the market efficiency and the implications for technical
and fundamental analysis in an emerging capital market. Market participants will be
able to use the findings of this study to more accurately predict expected cash inflows
and the value of stocks. With this information, market participants will be able to make
more informed investment decisions.
6.5
LIMITATIONS OF THE STUDY
Doctoral research is constrained by resources, time, funding, scope and focus.
Consequently, this research and its findings maybe influenced by the following
limitations.
The period of investigation covers the period 1990 to 2002. However, the data was
obtained from the SET database was only complete from 1997 onwards. The data
search was conducted to the highest possible level, however, the data from the 19901996 period is incomplete. Consequently, there could be some sample bias.
235
DBA Thesis: Malinee Ronapat
The characteristics of the listed firms which were suggested by Fama and French
(2001) may be limited and may overlook other characteristics which may be important.
Each ratio which represents a proxy for a firm’s characteristics has its own limitations
in terms of expressing the relationship between the characteristics of the firm and its
dividend behaviour. This problem occurs even when firm characteristics are being
analysed in developed markets. This study has analysed an emerging market, and some
of the homogenous characteristics may not be presented in the study.
Lastly, listed firms may have their own accounting policies, and these accounts may
represent different characteristics of firms.
6.6
SUGGESTIONS FOR FURTHER RESEARCH
It is suggested that more research is needed on the characteristics of the different
dividend groups. In particular, the issue of whether the characteristics which were
suggested by Fama and French (2001) are adequate needs further consideration.
Furthermore, expanding the size of the sample of listed firms and the length of time
frame should be considered because the findings of this study indicate an upward trend
in the percentage of payers in recent years in the Thailand’s capital market. In addition,
more than three dividend groups could be assessed in further research. Different
patterns of dividends payments may also be identified if the time frame and size of
samples are expands.
Factors such as inflation rates, interest rates, business cycles, GDP, GNP could be
incorporated in the analysis to enhance the predictive power of the logit regression.
The fact that many firms resumed paying dividends after the crisis in Thailand indicates
that investors appreciate value stocks and firms still want to pay dividends. This implies
further research is needed to investigate the performance of value and growth stocks in
Thailand. This is to provide additional information for investors and expand the
236
DBA Thesis: Malinee Ronapat
literature of the dividend behaviour of firms. Behaviourial finance could also be used to
explain the behaviour of the Thai investors, why they prefer value stocks or growth
stocks, and is it culturally different.
Finally, though descriptive statistics and logit regression are recommended as an
appropriate methodology in the literature (Fama and French 2001), researchers could
expand the research effort by applying different statistical tools to verify the
characteristics which were suggested by Fama and French (2001), or analyse the
propensity to pay dividends of listed firms over time.
237
DBA Thesis: Malinee Ronapat
References
ADB 2000a, Asian Recovery Report 2000a, Asia Recovery Information Center 2000a,
Asian Development Bank, Manila.
ADB 2000b, Asian Recovery Report 2000b, Asia Recovery Information Center 2000b,
Asian Development Bank, Manila.
ADB 2000c, Asian Recovery Report 2000c, Asia Recovery Information Center 2000c,
Asian Development Bank, Manila.
Aivazian, V., Booth, L. & Cleary, S. 2003, ‘Do Emerging Market Firms Follow Different
Dividend Policies From U.S. Firms?’, Journal of Financial Research, vol. 26, no. 3, Fall,
pp. 371-387.
Ambarish, R., John, K. & Williams, J. 1987, ‘Efficient Signaling with Dividends and
Investments’, Journal of Finance, vol. 42, no. 2, June, pp. 321-343.
Asquith, P. & Mullins, D.W. Jr. 1983, ‘The Impact of Initiating Dividend Payments on
Shareholders' Wealth’, Journal of Business, vol. 56, no. 1, February, pp. 77-96.
Auerbach, A.J., 1979a, ‘Share Valuation and Corporate Equity Policy’, Journal of Public
Economics, vol. 2, no. 2, June, pp. 291-305.
Auerbach, A.J., 1979b, ‘Wealth Maximization and the Cost of Capital’, Quarterly Journal
of Economics, vol. 93, no. 3, August, pp. 433-446.
Baker, H.K. & Farrelly, G.E. 1988, ‘Dividend Achievers: A Behavioral Perspective’,
Akron Business and Economic Review, vol. 19, no. 1, Spring, pp. 79-92.
Baker, H.K., Farrelly, G.E. & Edelman, R.B. 1985, ‘A Survey of Management Views on
Dividend Policy’, Financial Management, vol. 14, no. 3, Autumn, pp. 78-84.
Baker, M. & Wurgler, J. 2002, ‘Why are Dividends Disappearing’, Unpublished Working
Paper, Stern School of Business, New York University, December.
Baker, M. & Wurgler, J. 2003a, ‘A Catering Theory of Dividends’, National Bureau of
Economic Research Ind., NBER Working Papers, no. 9542, March,
http://pages.stern.nyu.edu/~jwurgler/papers/catering.pdf (accessed 14th, October 2003).
Baker, M. & Wurgler, J. 2003b, ‘Appearing and Disappearing Dividends: The Link to
Catering Incentives’, National Bureau of Economic Research Ind., NBER Working
Paper, no. 9995, September, http://pages.stern.nyu.edu/~jwurgler/papers/appdisdiv.pdf
(accessed 14th October 2003).
238
DBA Thesis: Malinee Ronapat
Bar-Yosef, S. & Huffman, L. 1986, ‘The Information Content of Dividends: A Signalling
Approach’, Journal of Financial and Quantitative Analysis, vol. 21, no. 1, March, pp. 4758.
Bar-Yosef, S. & Kolodny, R. 1976, ‘Dividend Policy and Capital Market Theory’, Review
of Economics and Statistics, vol. 58, no. 2, May, pp. 181-190.
Bar-Yosef, S. & Kolodny, R. 1978, ‘Dividend Policy and Capital Market Theory: A
Reply’, Review of Economics and Statistics, vol. 60, no. 3, August, pp. 477-478.
Benito, A. & Young, G. 2001, ‘Hard Times or Great Expectations?: Dividend Omissions
and Dividend Cuts by UK Firms’, Working Papers no. 147, Bank of England, London.
Berle, A.A. & Means, G.C. 1932, The Modern Corporation and Private Property,
MacMillan Company, New York.
Bhattacharya, S. 1979, ‘Imperfect Information, Dividend Policy, and "The Bird in the
Hand" Fallacy’, Bell Journal of Economics, vol. 10, no. 1, Spring, pp. 259-270.
Bhattacharya, S. 1980, ‘Nondissipative Signaling Structures and Dividend Policy’,
Quarterly Journal of Economics, vol. 95, no. 1, August, pp. 1-24.
Black, F. 1976, ‘The Dividend Puzzle’, Journal of Portfolio Management, vol. 2, no. 2,
Winter, pp. 5-8.
Black, R.T. 1999, Doing Quantitative Research in the Social Sciences, An Integrated
Approach to Research Design, Measurement and Statistics, SAGE Publication, London.
BOI 2004, ‘Country Review’, http://www.boi.go.th/english/thailand/index.html, Accessed
16th February 2004.
BOT 1990, Economic Review (1990), Publication Department, Bank of Thailand,
Bangkok.
BOT 1991, Economic Review (1991), Publication Department, Bank of Thailand,
Bangkok.
BOT 1992, Economic Review (1992), Publication Department, Bank of Thailand,
Bangkok.
BOT 1993, Economic Review (1993), Publication Department, Bank of Thailand,
Bangkok.
BOT 1994, Economic Review (1994), Publication Department, Bank of Thailand,
Bangkok.
BOT 1995, Economic Review (1995), Publication Department, Bank of Thailand,
Bangkok.
239
DBA Thesis: Malinee Ronapat
BOT 1996, Economic Review (1996), Publication Department, Bank of Thailand,
Bangkok.
BOT 1997, Economic Review (1997), Publication Department, Bank of Thailand,
Bangkok.
BOT 1998, Economic Review (1998), Publication Department, Bank of Thailand,
Bangkok.
BOT 1999, Economic Review (1999), Publication Department, Bank of Thailand,
Bangkok.
BOT 2000, Economic Review (2000), Publication Department, Bank of Thailand,
Bangkok.
BOT 2001, Economic Review (2001), Publication Department, Bank of Thailand,
Bangkok.
BOT 2002, Economic Review (2002), Publication Department, Bank of Thailand,
Bangkok.
Brealey, R.A. & Myers, S.C. 1991, Principles of Corporate Finance, 4th edition, McGraw
Hill, New York.
Brealey, R.A. & Myers, S.C. 2000, Principles of Corporate Finance, 6th edition, Irwin
McGraw-Hill, London.
Brennan, M.J. 1970, ‘Taxes, Market Valuation, and Corporation Financial Policy’,
National Tax Journal, vol. 23, no. 3, December, pp. 417-427.
Bringham, E.F., Gapenski, L.C. & Ehrhardt, M.C. 1999, Financial Management, Theory
and Practice, 9th edition, Dryden Press, Fort Worth, Texas.
Bryman, A. & Cramer, D. 1999, Quantitative Data Analysis with SPSS Release 8 for
Windows: A Guide for Social Scientists, Routledge, London.
Bryman, A. & Cramer, D. 2001, Quantitative Data Analysis with SPSS Release 10 for
Windows: A Guide for Social Scientists, Routledge, London.
Case, K.E. & Fair, R.C. 1996, Principles of Economics, 4th edition, Prentice-Hall
International.
Cavana, R.Y., Delahaye, B.L. & Sekeran, U. 2001, Applied Business Research,
Qualitative and Quantitative Methods, John Wiley and Sons, Sydney.
Cholviroj, B. 1989, Taxation, Tana Publishing, Bangkok.
240
DBA Thesis: Malinee Ronapat
Christensen, T. & Jordan, W. 2001, Fundamentals of Corporate Finance, 2nd edition,
McGraw-Hill, Roseville, New South Wales.
Chua, S.J., Sim, W.Y. & Tham, S.S. 1999, ‘The Asian Financial Crisis – A Critical
Study’, Applied Research Project, Nanyang Technological University, pp. 36-47.
Clover, V.T. & Balsley, H.L. 1984, Business Research Methods, 3rd edition, Wiley &
Sons, New York.
Darling, P.G. 1957, ‘The Influence of Expectations and Liquidity on Dividend Policy’,
Journal of Political Economy, vol. 65, no. 2, June, pp. 209-224.
DeAngelo, H. & DeAngelo, L. 1990, ‘Dividend Policy and Financial Distress: An
Empirical Examination of Troubled NYSE Firms’, Journal of Finance, vol. 45, no. 5,
December, pp. 1415-1431.
DeAngelo, H., DeAngelo, L. & Skinner, D.J. 1992, ‘Dividends and Losses’, Journal of
Finance, vol. 57, no. 5, December, pp. 1837-1863.
DeAngelo, H., DeAngelo, L. & Skinner, D.J. 2002, ‘Are Dividends Disappearing?
Dividend Concentration and the Consolidation of Earnings’, Centre in Law, Economics
and Organization Research Papers, Working Paper no.02-09, University of Southern
California.
DeAngelo, H. & Masulis, R.W. 1980. ‘Leverage and Dividend Irrelevancy Under
Corporate and Personal Taxation’, Journal of Finance, vol. 8, no. 1, March, pp. 3-29.
Easterbrook, F.H. 1984, ‘Two Agency-Cost Explanations of Dividends, The American
Economic Review, vol. 74, no. 4, September, pp. 650-659.
Emory, C.W. 1995, Business Research Methods, 5th edition, Irwin, Homewood, Illinois.
Emory, C.W. & Cooper, D.R. 1991, Business Research Methods, 4th edition, Irwin,
Homewood, Illinois.
Fama, E.F. 1970, ‘Efficient Capital Markets: A Review of Theory and Empirical Work’,
Journal of Finance, vol. 25, no. 2, March, pp. 383-423.
Fama, E.F. 1976, Foundations of Finance: Portfolio Decisions and Securities Prices,
Basic Books, New York.
Fama, E.F. 1998, ‘Market Efficiency, Long-Term Returns, and Behavioral Finance’,
Journal of Financial Economics, vol. 49, no. 3, September, pp. 283-306.
Fama, E.F. & Babiak, H. 1968, ‘Dividend Policy: An Empirical Analysis,’ Journal of the
American Statistical Association, vol. 63, no. 324, December, pp. 1132-1161.
Fama, E.F. & French, K.R. 1988, ‘Dividend Yields and Expected Stock Returns’, Journal
of Financial Economics, vol. 22, no. 1, October, pp. 3-25.
241
DBA Thesis: Malinee Ronapat
Fama, E.F. & French, K.R. 1995, ‘Size and Book-to-Market Factors in Earnings and
Returns’, Journal of Finance, vol. 50, no. 1, March, pp. 131-155.
Fama, E.F. & French, K.R. 1997, ‘Dividends, Debt, Investment and Earnings’, Working
Paper, Graduate School of Management, University of Chicago, Chicago.
Fama, E.F. & French, K.R. 1998, ‘Value versus Growth: The International Evidence’,
Journal of Finance, vol. 53, no. 6, December, pp. 1975–1999.
Fama, E.F. & French, K.R. 1999, ‘Testing Tradeoff and Pecking Order Predictions about
Dividends and Debt’, Working Paper, Sloan School of Management, MIT.
Fama, E.F. & French, K.R. 2000, ‘Forecasting Profitability and Earnings’, Journal of
Business, vol. 72, no. 2, April, pp. 161-175.
Fama, E.F. & French, K.R. 2001, ‘Disappearing Dividends: Changing Firms
Characteristics or Lower Propensity to Pay?’, Journal of Financial Economics, vol. 60,
no. 1, April, pp. 3-44.
Fama, E.F & French, K.R. 2002, ‘The Equity Premium’, Journal of Finance, vol. 57, no.
2, April, pp. 637-659.
Farrar, D.E. & Selwyn, L.L. 1967, ‘Taxes, Corporate Financial Policy and Return to
Investors, National Tax Journal, vol. 20, no. 3, May, pp. 444-462.
Fazzari, S., Hubbard, R.G. & Petersen, B.C. 1988, ‘Financing Constraints and Corporate
Investment’, Brookings Papers on Economic Activity, pp. 1141-1195.
Fama, E.F. & MacBeth, J.D. 1973, ‘Risk, Return and Equilibrium: Empirical Tests’,
Journal of Political Economic, vol. 81, no. 3, May, pp. 607-636.
Fama, E.F & Miller, M.H. 1972, The Theory of Finance, Dryden Press, Hinsdale, Illinois.
Feldstein, M.S. & Green, J. 1983, ‘Why Do Companies Pay Dividends?’, American
Economic Review, vol. 73, no. 1, March, pp. 17-30.
Ferris, P.S., Sen, N. & Yui, H.P., 2002, ‘International Dividend Policy: Legal Influences,
Firm Attributes and the Desire to Pay’, Research Proposal, Nanyang Technological
University, Singapore.
Firth, M. 1977, The Valuation Of Shares and The Efficient-Markets Theory, 1st edition,
Macmillan Press, London.
Fisher E.D. & Jordan, J.R. 1991, Security Analysis and Portfolio Management, 5th edition,
Englewood Cliffs, New Jersey.
Frankfurter, G.M. & Gong, J. 1992, ‘Empirical Tests of a Dividend Signaling Equilibrium
Model’, Unpublished Working Paper, Bucknell University.
242
DBA Thesis: Malinee Ronapat
Frankfurter, G.M. & Gong, J. 1993, ‘Time-Series Cross-Sectional Tests of Dividend
Policy Determinants’, Unpublished Working Paper, Bucknell University.
Frankfurter, G.M. & Lane, W.R. 1992, ‘The Rationality of Dividends’, International
Review of Financial Analysis, vol. 1, pp. 115-129.
Frankfurter, G.M. & Wood, B.G. 2001, ‘Dividend Policy Theories and Their Empirical
Tests’, Unpublished Working Paper, Bucknell University.
Fung, W.K.H. & Theobald, M.F. 1984, ‘Dividends and Debt Under Alternative Tax
Systems’, Journal of Financial and Quantitative Analysis, vol. 19, no. 1, February, pp. 5972.
Gitman, L.J. 2000, Principles of Managerial Finance, 9th edition, Addison Wesley
Publisher, Melbourne.
Gitman L.J., Juchau, R. & Flanagan, J. 2002, Principles of Managerial Finance, 3rd
edition, Addison Wesley Publisher, Melbourne.
Goldstein, M. 1998, ‘The Asian Financial Crisis: Causes, Crues, and Systematic
Implication’, Institute for International Economics, Washington, D.C., June, pp. 1-22.
Gordon, M.J. 1962, The Investment, Financing and Valuation of the Corporation, Richard
D. Irwin, Homewood, Illinois.
Gordon, M.J. 1963, ‘Optimal Investment and Financing Policy’, Journal of Finance,
vol.18, no. 2, May, pp. 264-72.
Grinblatt, M. & Titman, S. 1998, Financial Markets and Corporate Strategy,
Irwin/McGraw-Hill, Boston, Massachusetts.
Gwilym, O.A., Morgan, G. & Thomas, S. 2000, ‘Dividend Stability, Dividend Yield and
Stock Returns: UK Evidence, Journal of Business Finance and Accounting, vol. 27, no. 3
and 4, April/May, pp. 261-281.
Hakansson, N.H. 1982, ‘To Pay or Not to Pay Dividend’, Journal of Finance, vol. 37,
May, pp. 415-428; reprinted in summary form in the C.F.A. Digest, 1983; vol. 13,
Summer, pp. 16-17.
Harkins, E.P. & Walsh, F.J. Jr. 1971, Dividend Policies and Practices, The Conference
Board, Inc., New York.
Haugen, R. 1995, The New Finance: The Case against Efficient Markets, Prentice Hall,
Englewood Cliffs, New Jersey.
243
DBA Thesis: Malinee Ronapat
Healey, P.M. & Palepu, K.G. 1988, ‘Earnings Information Conveyed by Dividend
Initiations and Omissions’, Journal of Financial Economics, vol. 21, no. 2, September, pp.
149-275.
Higgins, R.C. 1972, ‘The Corporate Dividend-Saving Decision’, Journal of Financial and
Quantitative Analysis, vol. 7, no. 6, December, pp. 1527-1541.
Huck, S.W., Cormier, W.H., & Bounds, W.G. 1974, Reading Statistics and Research,
Harper & Row publishers, New Jersey.
Hurley, W.J. & Johnson, D.L. 1994, ‘A Realistic Dividend Valuation Model’, Financial
Analysts Journals, vol. 50, no. 1, July-August, pp. 50-54.
Hurley, W.J. & Johnson, D.L. 1998, ‘Generalized Markov Dividend Discount Models’,
Journal of Portfolio Management, vol. 25, no. 1, February, pp. 27-31.
James H. L., Dodd, P. & Kimpton, M.H. 1985, The Stock Market: Theories and Evidence,
2nd edition, Irwin Professional Publication, February, Homewood, Illinois.
James, L.D., Fama, E.F. & French, K.R. 2000, ‘Characteristics, Covariances and Average
Returns 1929 to 1997’, Journal of Finance, vol. 55, no. 1, February, pp. 389-406.
Jensen, M.C. 1978, ‘Some Anomalous Evidence Regarding Market Efficiency’, Journal
of Financial Economics, vol. 6, no. 2, June/September, pp. 95-101.
Jensen, M.C. 1986, ‘Agency Costs of Free Cashflow, Corporate Finance and Takeovers’,
American Economic Review, vol. 76, no. 2, May, pp. 323-329.
John, K. & Williams, J. 1985, ‘Dividends, Dilution, and Taxes: A Signalling Equilibrium’,
Journal of Finance, vol. 40, no. 4, September, pp. 1053-1070.
Jolls, C. 1998, The Role of Incentive Compensation in Explaining the Stock-Repurchase
Puzzle, Harvard University Working Paper, Harvard University.
Kahneman, D. & Tversky, A. 1982, ‘The Psychology of Preferences’, Scientific American,
vol. 246, no. 14, December, pp. 167-173.
Kale, J.R. & Noe, T.H. 1990, ‘Dividends, Uncertainty, and Underwriting Costs Under
Asymmetric Information’, Journal of Finance, vol. 13, no. 2, Winter, pp. 265-277.
Kindleberger, C.P. 1984, A Financial History of Western Europe, George Allen & Unwin,
London.
Kumar, P. 1988, ‘Shareholder-Manager Conflict and the Information Content of
Dividends’, Review of Financial Studies, vol. 1, no. 2, June, pp. 111-136.
244
DBA Thesis: Malinee Ronapat
Lambert, R.A., Lanen, W.N. & Larcker, D.F. 1989, ‘Executive Stock Option Plans and
Corporate Dividend Policy, Journal of Financial and Quantitative Analysis, vol. 24, no. 3,
August, pp. 409-425.
Laub, P.M. 1976, ‘On the Informational Content of Dividends, Journal of Business, vol.
49, no. 1, February, pp. 73-80.
La Porta, R., Lopez-de-Silanes, F., Shleifer, A., Vishny, R., 2000, ‘Agency problems and
Dividend Policies Around the World’, Journal of Finance, vol. 55, no. 2, June, pp. 859874.
Lakonishok, J., Shleifer, A. & Vishny, R.W. 1994, ‘Contrarian Investment,
Extrapolation, and Risk’, Journal of Finance, vol. 49, no. 5, December, pp. 1541–1578.
Lim, C.Y. 2000, ‘From Recession to Recovery in East Asian: a Non-IMF and Non-world
Bank Explanation’, Accounting and Business Review, vol. 7, no. 2, July, pp. 1-10.
Lintner, J. 1956, ‘Distribution of Incomes of Corporations Among Dividends, Retained
Earnings, and Taxes’, American Economic Review, vol. 46, no. 2, May, pp. 97-113.
Lintner, J. 1962, ‘Dividends, Earnings, Leverage, Stock Prices, and the Supply of Capital
to Corporations’, Review of Economics and Statistics, vol. 44, no. 2, August, pp. 243-69.
Madura, J. 2000, International Financial Management, 6th edition, South-Western College
Publishing, Cincinnati, Ohio.
Makhija, A.K. & Thompson, H.E. 1986, ‘Some Aspects of Equilibrium for a CrossSection of Firms Signaling Profitability with Dividends: A Note’, Journal of Finance, vol.
41, no. 2, May, pp. 249-253.
March, T.A. & Merton, R.C. 1987, ‘Dividend Behavior for the Aggregate Stock Market’,
Journal of Business, vol. 60, no. 1, January, pp. 1-40.
Mason, R.D., Lind, D.A. & Marchal W.G. 1999, Statistical Techniques in Business and
Economics, 9th edition, Irwin McGraw-Hill, Chicago.
Masulis, R.W. & Trueman, B. 1988, ‘Corporate Investment and Dividend Decisions
Under Differential Personal Taxation’, Journal of Financial and Quantitative Analysis,
vol. 23, no. 4, September, pp. 369-386.
Melicher, R.W. & Norton, E.A. 2000, Finance: Introduction to Institutions, Investments,
and Management, 10th edition, South-Western College Publishing, Thomson Learning,
Cincinnati, Ohio.
Miller, M.H. 1977, ‘Debt and Taxes’, Journal of Finance, vol. 32, no. 2, May, pp. 261275.
Miller, M.H. 1986, ‘Behavior Rationality in Finance: The Case of Dividends’, Journal of
Business, vol. 59, no. 4, October, pp. 451-468.
245
DBA Thesis: Malinee Ronapat
Miller, M.H. & Modigliani, F. 1961, ‘Dividend Policy, Growth, and the Valuation of
Shares’, Journal of Business, vol. 34, no. 4, October, pp. 411-413.
Miller, M.H. & Rock K, 1985, ‘Dividend Policy Under Asymmetric Information’, Journal
of Finance, vol. 40, no. 4, September, pp. 1,031-1,051.
Miller, M.H. & Scholes, M.S. 1978, ‘Dividends and Taxes’, Journal of Financial
Economics, vol. 6, no. 4, December, pp. 333-364.
Miller, M.H. & Scholes, M.S. 1982, ‘Dividends and Taxes: Some Empirical Evidence’,
Journal of Political Economy, vol. 90, no. 6, December, pp. 1,118-1,141.
Modigliani, F. 1982, ‘Debt, Dividend Policy, Inflation and Market Valuation, Journal of
Finance, vol. 37, no. 2, May, pp. 255-273.
Moore, D.S. & McLabe, G.P. 1998, Introduction to the Practice of Statistics, 2nd edition,
W.H. Freeman, New York.
Morgan, I.G. 1980, ‘Dividend and Stock Price Behavior in Canada’, Journal of Business
Administration, vol. 12, no. 1, February, pp. 91-107.
Morgan, I.G. 1982, ‘Dividends and Capital Asset Prices’, Journal of Finance, vol. 37, no.
5, December, pp. 1071-1086.
Myers, S.C. 1984, ‘The Capital Structure Puzzle’, Journal of Finance, vol. 39, no. 3, July,
pp. 575-592.
Ofer, A. R. & Thakor, A.V. 1987, ‘A Theory of Stock Price Responses to Alternative
Corporate Cash Disbursement Methods: Stock Repurchases and Dividends’, Journal of
Finance, vol. 42, no. 2, June, pp. 365-394
Page, C & Meyer, D. 2000, Applied Research Design for Business and Management,
Irwin McGraw-Hill, London.
Penman, S.H. & Sougiannis T. 1998, ‘A Comparison of Dividend, Cash Flow, and
Earnings Approaches to Equity Valuation’, Contemporary Accounting Research, vol. 15,
no. 3, Fall, pp. 343-383.
Perry, C. 1995, ‘A Structured Approach to Presenting PhD Theses: Notes for Candidates
and Their Supervisors’, Research Workshop – Handout, pp. 45-85, Southern Cross
University.
Perry, C. 1998, ‘A Structured Approached for Presenting Theses’, Australasian Marketing
Journal, vol. 6, no. 1, pp. 63-86.
Petty, J.W. et al. 2000, Financial Management, 2nd edition, Prentice Hall, Sydney, New
South Wales.
246
DBA Thesis: Malinee Ronapat
Pringvanich, N. 2002, Thailand Overview, December, Bangkok.
Radelet, S. & Schs, J. 1998, ‘The Onset of the East Asian Financial Crisis’, Paper
Presented to the National Bureau of Economic Research Currency Crises Conference,
February 6-7 1998.
Rahman S., Yui H.P. & Foon, H.S. 2003, ‘Comparative Speed of Recovery of the Asian
Economic Crisis Driven Countries’, International Trade Counselling Centre, The Seventh
International Conference on Global Business and Economic Development, pp. 495-511.
Rappaport, A. 1986 ‘The Affordable Dividend Approach to Equity Valuation’, Financial
Analysts Journal, vol. 42, no. 4, July/August, pp. 52-59.
Reilly F.K. & Brown K.C. 2000, ‘Investment Analysis and Portfolio Management’, 6th
edition, Harcourt College Publishers, Fort Worth, Texas.
Reilly F.K. & Norton, E.A. 1999, Investment, 5th edition, Dryden Press, Fort Worth,
Texas.
Richardson, G., Sefcik, S.E. & Thompson, R. 1986, ‘A Test of Dividend Irrelevance
Using Volume Reaction to a Change in Dividend Policy, Journal of Financial Economics,
vol. 17, no. 2, April, pp. 313-333.
Rodriguez, R.J. 1992, ‘Quality Dispersion and the Feasibility of Dividends as Signals’,
Journal of Financial Research, vol. 15, no. 4, September, pp. 307-315.
Roll, R., 1977, ‘A Critique of the Asset Price Theory's Test; Part I: On Past and Potential
Testability of Theory’, Journal of Financial Economics, vol. 4, no. 2, March, pp. 129-176.
Roll, R. 1983, ‘Vas ist das?’, Journal of Portfolio Management, vol. 9, no.2, Winter, pp.
18-28.
Saengkaew, V. 1996, Statistic for Business Management, 6th edition, Assumption
University, Sarm Dee Press, Bangkok.
SCU, 1995, DBA Manual, Southern Cross University.
SCU, 2000, DBA Manual, Southern Cross University.
SCU, 2002, DBA Manual, Southern Cross University.
Sekeran, M. 2000, Research Methods for Business.
SEC 2001, Annual Report 2001, Investment Management Supervision Department,
Investment Advice Supervision Department and Broker-Dealer Supervision Department,
The Stock Exchange Commission of Thailand, Bangkok.
247
DBA Thesis: Malinee Ronapat
SEC 2002a, Annual Report 2002, Investment Management Supervision Department,
Investment Advice Supervision Department and Broker-Dealer Supervision Department,
The Stock Exchange Commission of Thailand, Bangkok.
SEC 2002b, ‘Investors Behaviour and Understanding about Investment’, Strategic
Management and Development, The Stock Exchange Commission of Thailand, Bangkok.
SEC 2003, Annual Report 2003, Investment Management Supervision Department,
Investment Advice Supervision Department and Broker-Dealer Supervision Department,
The Stock Exchange Commission of Thailand, Bangkok.
SET 1990, Fact Book 1990, Publication Department, The Stock Exchange of Thailand,
Bangkok.
SET 1991, Fact Book 1991, Publication Department, The Stock Exchange of Thailand,
Bangkok.
SET 1992, Fact Book 1992, Publication Department, The Stock Exchange of Thailand,
Bangkok.
SET 1993, Fact Book 1993, Publication Department, The Stock Exchange of Thailand,
Bangkok.
SET 1994, Fact Book 1994, Publication Department, The Stock Exchange of Thailand,
Bangkok.
SET 1995, Fact Book 1995, Publication Department, The Stock Exchange of Thailand,
Bangkok.
SET 1996, Fact Book 1996, Publication Department, The Stock Exchange of Thailand,
Bangkok.
SET 1997, Fact Book 1997, Publication Department, The Stock Exchange of Thailand,
Bangkok.
SET 1998, Fact Book 1998, Publication Department, The Stock Exchange of Thailand,
Bangkok.
SET 1999, Fact Book 1999, Publication Department, The Stock Exchange of Thailand,
Bangkok.
SET 2000a, Annual Report, Investor Relations and Communications Department, The
Stock Exchange of Thailand, Bangkok.
SET 2000b, Fact Book 2000, Publication Department, The Stock Exchange of Thailand,
Bangkok.
248
DBA Thesis: Malinee Ronapat
SET 2001a, Annual Report, Investor Relations and Communications Department, The
Stock Exchange of Thailand, Bangkok.
SET 2001b, Fact Book 2001, Publication Department, The Stock Exchange of Thailand,
Bangkok.
SET 2002a, Annual Report, Investor Relations and Communications Department, The
Stock Exchange of Thailand, Bangkok.
SET 2002b, Fact Book 2002, Publication Department, The Stock Exchange of Thailand,
Bangkok.
SET 2002c, The Stock Exchange in Thailand 2002, Investor Relations and
Communications Department, The Stock Exchange of Thailand, Bangkok.
SET 2003, Fact Book 2003, Publication Department, The Stock Exchange of Thailand,
Bangkok.
Shefrin, H.M. & Statman, M. 1984, ‘Explaining Investor Preference for Cash Dividends’,
Journal of Financial Economics, vol. 13, no. 2, May, pp. 253-282.
Shiller, R.J. 1984, ‘Stock Prices and Social Dynamics’, Brookings Papers on Economic
Activity, no. 2, October, pp. 457-510.
Smith, A. 1937, The Wealth of Nations, Random House, New York.
Sutcliffe, C. 1993, Stock Index Futures: Theories and International Evidence, Chapman
and Hall Series in Accounting and Finance, Chapman and Hall, London.
Sutcliffe, C. 1997, Stock Index Futures: Theories and International Evidence, Second
edition, International Thomson Business Press, London.
Tabachnick, B.G. & Fidell, L. 2001, Using Mutivariate Analysis, 4th edition, McGrawHill, New York.
Talmor, E., 1981, ‘Asymmetric Information, Signaling, and Optimal Corporate Financial
Decisions’, Journal of Financial and Quantitative Analysis, vol. 16, no. 4, August, pp.
413-435.
Thaler, R.H. & Shefrin, H.M. 1981, ‘An Economic Theory of Self Control’, Journal of
Political Economy, vol. 89, no. 1, April, pp. 392-410.
Ticehurst, G.W. & Veal, A.J. 2000, Business Research Methods: A Managerial Approach,
Longman, French Forest, New South Wales.
Tobin, J. 1958, ‘Liquidity Preference as Behaviour Towards Risk’, Review of Economic
Studies, vol. 25, no. 67, February, pp. 65-86.
249
DBA Thesis: Malinee Ronapat
TRIS 2001, ‘Moody’s and S&P’s Credit Rating’, TRISVIEW, no. 2, January.
TRIS 2002, ‘Sovereign Credit Rating of Thailand’, TRISVIEW, no. 2, December.
Turnovsky, S.J. 1967, ‘The Allocation of Corporate Profits Between Dividends and
Retained Earnings’, Review of Economics and Statistics, vol. 49, no. 1, January, pp. 583589.
Watts, R. 1973, ‘The Information Content of Dividends’, Journal of Business, vol. 46, no.
2, Spring, pp. 191-211.
Watts, R. 1976a, ‘Comments on The Informational Content of Dividends’, Journal of
Business, vol. 49, no. 1, January, pp. 81-85.
Watts, R. 1976b, ‘Comments on "The Impact of Dividend and Earnings Announcements":
A Reconciliation’, Journal of Business, vol. 49, no. 1, January, pp. 97-106.
Weisbenner, S.J. 1999, Corporate Share Repurchase in the Mid-1990s: What Role Do
Stock Option Play?, Federal Reserve Board Working Paper.
Wetherilt, V. & Weeken, O. 2002, ‘Equity Valuation Measures: What Can They Tell
Us?’, Quarterly Bulletin, Bank of England Financial Stability Review, Monetary
Instruments and Markets Division, Winter, pp. 391-403.
Yao, Y. 1997, ‘A Trinomial Dividend Valuation Model’, Journal of Portfolio
Management, pp. 99-103.
Yardeni, E. 2003, ‘Stock Valuation Models 4.1’, Prudential Financial Research, January.
Zikmund, W.G. 2000, Business Research Methods, 6th edition, Dryden Press, Fort Worth,
Texas.
250
DBA Thesis: Malinee Ronapat
Appendices
Appendix A:
Descriptive statistics and ANOVA comparing payers and non-payers 251
Appendix B:
ANOVA comparing payers, former payers, never paid firms
264
Appendix C:
Logit Regression results controlling two investment variables
277
Appendix D:
Logit Regression results controlling one investment variable
280
Appendix E:
t-statistic results (two investment variables)
283
Appendix F:
t-statistic results for (one investment variable)
285
Appendix G:
Expected percentage of payers controlling two investment variables
287
(crosstabs)
Appendix H:
Expected percentage of payers controlling one investment variables
(crosstabs)
292
251
DBA Thesis: Malinee Ronapat
APPENDIX A:
DESCRIPTIVE STATISTICS AND ANOVA: COMPARING
PAYERS AND NON-PAYERS
252
DBA Thesis: Malinee Ronapat
Table A1: Descriptive statistics: comparing payers and non-payers (1991)
N
Mean
Std. Deviation
Std. Error
95% Confidence Interval for Mean
Lower Bound
Et/At
Yt/BEt
dAt/At
Vt/At
At
Lt/At
Minimum
Maximum
Upper Bound
.00
32
.1220
.06134
.01084
.0999
.1442
.02
.30
1.00
192
.1096
.06955
.00502
.0997
.1195
-.42
.43
Total
224
.1113
.06845
.00457
.1023
.1203
-.42
.43
.00
32
.1661
.10292
.01819
.1290
.2032
.00
.48
1.00
192
.1663
.18261
.01318
.1403
.1923
-.49
2.00
Total
224
.1662
.17330
.01158
.1434
.1891
-.49
2.00
.00
32
.1247
1.22796
.21707
-.3181
.5674
-6.18
1.00
1.00
192
.3185
.47411
.03422
.2511
.3860
-2.76
1.00
Total
224
.2908
.63778
.04261
.2069
.3748
-6.18
1.00
.00
32
6.4437
12.92814
2.28539
1.7826
11.1048
.46
63.74
1.00
192
4.0363
7.22362
.52132
3.0080
5.0646
.28
69.34
Total
224
4.3802
8.28492
.55356
3.2894
5.4711
.28
69.34
.00
32
3435.8506
6880.07394
1216.23674
955.3194
5916.3818
58.67
32611.85
1.00
192
14257.5330
58657.01194
4233.20520
5907.6966
22607.3693
137.67
595801.1
Total
224
12711.5783
54478.47848
3639.99645
5538.3867
19884.7700
58.67
595801.1
.00
32
.4366
.24846
.04392
.3470
.5262
.04
.95
1.00
192
.5007
.24053
.01736
.4665
.5349
.01
1.15
Total
224
.4915
.24215
.01618
.4597
.5234
.01
1.15
Source: Developed for this research
Table A2: ANOVA: comparing payers and non-payers (1991)
Sum of Squares
Et/At
Yt/BEt
dAt/At
Vt/At
At
Lt/At
Between Groups
df
Mean Square
.004
1
.004
Within Groups
1.040
222
.005
Total
1.045
223
Between Groups
.000
1
.000
Within Groups
6.698
222
.030
Total
6.698
223
Between Groups
1.031
1
1.031
Within Groups
89.678
222
.404
Total
90.709
223
Between Groups
158.965
1
158.965
Within Groups
15147.748
222
68.233
Total
15306.712
223
Between Groups
3212127325.688
1
3212127325.688
Within Groups
658630602430.805
222
2966804515.454
Total
661842729756.492
223
Between Groups
.113
1
.113
Within Groups
12.964
222
.058
Total
13.076
223
F
Sig.
.913
.340
.000
.995
2.552
.112
2.330
.128
1.083
.299
1.928
.166
Source: Developed for this research
253
DBA Thesis: Malinee Ronapat
Table A3: Descriptive statistics: comparing payers and non-payers (1992)
N
Mean
Std. Deviation
Std. Error
95% Confidence Interval for Mean
Lower Bound
Et/At
Yt/BEt
dAt/At
Vt/At
At
Lt/At
Minimum
Maximum
Upper Bound
.00
24
.0584
.11604
.02369
.0094
.1074
-.21
.33
1.00
230
.1059
.07370
.00486
.0963
.1154
-.01
.71
Total
254
.1014
.07958
.00499
.0916
.1112
-.21
.71
.00
24
-.0301
.49116
.10026
-.2375
.1773
-1.81
1.01
1.00
230
.1626
.19602
.01293
.1371
.1880
-.29
2.52
Total
-1.81
2.52
254
.1444
.24475
.01536
.1141
.1746
.00
24
.4898
.47189
.09632
.2905
.6890
-.25
1.00
1.00
230
.2866
.39351
.02595
.2355
.3377
-2.58
1.00
Total
254
.3058
.40491
.02541
.2558
.3558
-2.58
1.00
.00
24
2.0842
.80515
.16435
1.7442
2.4241
.99
4.05
1.00
230
1.8991
1.26107
.08315
1.7352
2.0629
.52
17.03
Total
254
1.9165
1.22528
.07688
1.7651
2.0680
.52
17.03
.00
24
5586.6921
17914.72551
3656.82803
-1978.0331
13151.4172
285.50
89174.37
1.00
230
13086.9055
60483.38340
3988.15620
5228.7331
20945.0779
59.83
666008.7
Total
254
12378.2239
57837.88692
3629.07179
5231.1851
19525.2628
59.83
666008.7
.00
24
.5655
.21255
.04339
.4758
.6553
.04
.91
1.00
230
.5031
.21815
.01438
.4747
.5314
.04
1.09
Total
254
.5090
.21798
.01368
.4821
.5359
.04
1.09
Source: Developed for this research
Table A4: ANOVA comparing: payers and non-payers (1992)
Sum of Squares
Et/At
Yt/BEt
dAt/At
Vt/At
Between Groups
.049
Within Groups
1.553
252
.006
Total
1.602
253
.807
1
.807
Within Groups
14.348
252
.057
Total
15.155
253
Between Groups
Between Groups
.897
1
.897
Within Groups
40.583
252
.161
Total
41.480
253
.745
1
.745
379.087
252
1.504
Between Groups
Total
Lt/At
Mean Square
1
Within Groups
At
df
.049
379.831
253
1222510516.344
1
1222510516.344
Within Groups
845118443932.337
252
3353644618.779
Total
846340954448.681
253
Between Groups
Between Groups
.085
1
.085
Within Groups
11.937
252
.047
Total
12.022
253
F
Sig.
7.941
.005
14.172
.000
5.570
.019
.495
.482
.365
.547
1.789
.182
Source: Developed for this research
254
DBA Thesis: Malinee Ronapat
Table A5: Descriptive statistics: comparing payers and non-payers (1993)
N
Mean
Std. Deviation
Std. Error
95% Confidence Interval for Mean
Lower Bound
Et/At
Yt/BEt
dAt/At
Vt/At
At
Lt/At
Minimum
Maximum
Upper Bound
.00
43
.0373
.10815
.01649
.0040
.0706
-.30
.35
1.00
244
.0992
.10145
.00649
.0864
.1120
-.13
1.12
Total
287
.0900
.10465
.00618
.0778
.1021
-.30
1.12
.00
43
.1428
.54845
.08364
-.0260
.3116
-.62
2.71
1.00
244
.1285
.20362
.01304
.1028
.1541
-1.77
1.27
Total
-1.77
2.71
287
.1306
.28183
.01664
.0979
.1633
.00
43
.4805
.44028
.06714
.3450
.6160
-.11
1.00
1.00
244
.1930
.49242
.03152
.1309
.2551
-6.10
1.00
Total
287
.2360
.49503
.02922
.1785
.2935
-6.10
1.00
.00
43
2.4644
2.01057
.30661
1.8456
3.0831
.77
10.74
1.00
244
2.4169
2.73211
.17491
2.0723
2.7614
.74
32.56
Total
287
2.4240
2.63364
.15546
2.1180
2.7300
.74
32.56
.00
43
6185.4342
17854.48186
2722.78296
690.6357
11680.2327
359.78
112231.5
1.00
244
14693.3015
69420.14214
4444.16920
5939.2908
23447.3122
61.42
782870.4
Total
287
13418.6036
64425.64905
3802.92582
5933.3304
20903.8769
61.42
782870.4
.00
43
.5441
.26420
.04029
.4628
.6254
.09
1.15
1.00
244
.5515
.49332
.03158
.4893
.6137
.00
5.65
Total
287
.5504
.46587
.02750
.4963
.6045
.00
5.65
Source: Developed for this research
Table A6: ANOVA: comparing payers and non-payers (1993)
Sum of Squares
Et/At
Yt/BEt
dAt/At
Vt/At
Between Groups
.140
Within Groups
2.992
285
.010
Total
3.132
286
.008
1
.008
Within Groups
22.709
285
.080
Total
22.716
286
Between Groups
Between Groups
3.022
1
3.022
Within Groups
67.063
285
.235
Total
70.084
286
.082
1
.082
1983.630
285
6.960
Between Groups
Total
Lt/At
Mean Square
1
Within Groups
At
df
.140
1983.713
286
2646170367.499
1
2646170367.499
Within Groups
1184443806742.514
285
4155943181.553
Total
1187089977110.013
286
Between Groups
Between Groups
.002
1
.002
Within Groups
62.070
285
.218
Total
62.072
286
F
Sig.
13.359
.000
.094
.759
12.842
.000
.012
.913
.637
.426
.009
.924
Source: Developed for this research
255
DBA Thesis: Malinee Ronapat
Table A7: Descriptive statistics: comparing payers and non-payers (1994)
N
Mean
Std. Deviation
95% Confidence Interval for Mean
Std. Error
Lower Bound
Et/At
Yt/BEt
dAt/At
Vt/At
At
Lt/At
Minimum
Maximum
Upper Bound
.00
52
.0374
.08448
.01172
.0138
.0609
-.27
.22
1.00
273
.0894
.05828
.00353
.0825
.0964
-.14
.48
Total
325
.0811
.06588
.00365
.0739
.0883
-.27
.48
.00
52
-1.1973
8.67080
1.20242
-3.6112
1.2167
-62.37
2.31
1.00
273
.1269
.12339
.00747
.1122
.1416
-.61
1.15
Total
325
-.0850
3.47613
.19282
-.4643
.2943
-62.37
2.31
.00
52
.2693
.48244
.06690
.1350
.4036
-1.57
1.00
1.00
273
.2321
.50844
.03077
.1715
.2927
-5.84
1.00
Total
325
.2381
.50383
.02795
.1831
.2931
-5.84
1.00
.00
52
1.6024
.87804
.12176
1.3580
1.8469
.73
6.28
1.00
273
1.6913
1.08964
.06595
1.5615
1.8211
.35
10.04
Total
325
1.6771
1.05791
.05868
1.5616
1.7925
.35
10.04
.00
52
11566.3871
41476.29233
5751.72687
19.3136
23113.4606
336.90
286428.5
1.00
273
15581.7186
76425.69599
4625.49354
6475.3991
24688.0380
64.43
898373.5
Total
325
14939.2655
71947.33881
3990.92030
7087.8771
22790.6540
64.43
898373.5
.00
52
.5625
.21827
.03027
.5017
.6232
.07
1.10
1.00
273
.4930
.25252
.01528
.4629
.5231
.00
2.96
Total
325
.5041
.24836
.01378
.4770
.5312
.00
2.96
Source: Developed for this research
Table A8: ANOVA: comparing payers and non-payers (1994)
Sum of Squares
Et/At
Between Groups
Within Groups
Total
Yt/BEt
dAt/At
Vt/At
.118
1.288
323
.004
1.406
324
1
76.584
Within Groups
3838.464
323
11.884
Total
3915.048
324
Between Groups
.060
1
.060
Within Groups
82.186
323
.254
Total
82.247
324
.345
1
.345
362.266
323
1.122
Between Groups
Total
Lt/At
Mean Square
1
76.584
Between Groups
Within Groups
At
df
.118
362.611
324
704247691.139
1
704247691.139
Within Groups
1676455690198.802
323
5190265294.733
Total
1677159937889.941
324
Between Groups
Between Groups
.211
1
.211
Within Groups
19.775
323
.061
Total
19.986
324
F
Sig.
29.695
.000
6.444
.012
.238
.626
.308
.580
.136
.713
3.442
.064
Source: Developed for this research
256
DBA Thesis: Malinee Ronapat
Table A9: Descriptive statistics: comparing payers and non-payers (1995)
N
Mean
Std. Deviation
Std. Error
95% Confidence Interval for Mean
Lower Bound
Et/At
Yt/BEt
dAt/At
Vt/At
At
Lt/At
Minimum
Maximum
Upper Bound
.00
54
.0208
.08060
.01097
-.0012
.0428
-.22
.21
1.00
290
.0884
.05193
.00305
.0824
.0944
-.07
.37
Total
344
.0778
.06232
.00336
.0712
.0844
-.22
.37
.00
54
.0331
.68460
.09316
-.1538
.2200
-2.27
3.90
1.00
290
.1192
.12572
.00738
.1047
.1338
-.72
1.03
Total
-2.27
3.90
344
.1057
.29448
.01588
.0745
.1369
.00
54
.1280
.32943
.04483
.0381
.2179
-.60
1.00
1.00
290
.2045
.37735
.02216
.1609
.2481
-4.04
1.00
Total
344
.1925
.37084
.01999
.1531
.2318
-4.04
1.00
.00
54
1.3616
.63364
.08623
1.1886
1.5345
.60
3.77
1.00
290
1.3891
.77130
.04529
1.3000
1.4783
.29
6.20
Total
344
1.3848
.75059
.04047
1.3052
1.4644
.29
6.20
.00
54
5925.6809
11092.62927
1509.51564
2897.9760
8953.3859
495.38
69338.75
1.00
290
19275.3505
88614.49312
5203.62221
9033.5478
29517.1532
301.84
1035447
Total
344
17179.7628
81602.30361
4399.70135
8525.9714
25833.5543
301.84
1035447
.00
54
.6028
.24187
.03291
.5368
.6688
.00
1.45
1.00
290
.5151
.20429
.01200
.4915
.5387
.01
.94
Total
344
.5289
.21266
.01147
.5063
.5514
.00
1.45
Source: Developed for this research
Table A10: ANOVA: comparing payers and non-payers (1995)
Sum of Squares
Et/At
Yt/BEt
dAt/At
Vt/At
At
Lt/At
Between Groups
df
Mean Square
.208
1
.208
Within Groups
1.124
342
.003
Total
1.332
343
Between Groups
.338
1
.338
Within Groups
29.407
342
.086
Total
29.745
343
Between Groups
.266
1
.266
Within Groups
46.903
342
.137
Total
47.169
343
.035
1
.035
Within Groups
193.205
342
.565
Total
193.240
343
Between Groups
Between Groups
8112866862.121
1
8112866862.121
Within Groups
2275902165295.430
342
6654684693.846
Total
2284015032157.551
343
Between Groups
.350
1
.350
Within Groups
15.161
342
.044
Total
15.511
343
F
Sig.
63.380
.000
3.929
.048
1.940
.165
.061
.805
1.219
.270
7.895
.005
Source: Developed for this research
257
DBA Thesis: Malinee Ronapat
Table A11: Descriptive statistics: comparing payers and non-payers (1996)
N
Mean
Std. Deviation
Std. Error
95% Confidence Interval for Mean
Lower Bound
Et/At
Yt/BEt
dAt/At
Vt/At
At
Lt/At
Minimum
Maximum
Upper Bound
.00
72
.0032
.12794
.01508
-.0269
.0333
-.73
.25
1.00
306
.0753
.05837
.00334
.0688
.0819
-.11
.52
Total
378
.0616
.08151
.00419
.0534
.0698
-.73
.52
.00
72
-.0627
.58979
.06951
-.2013
.0759
-2.80
2.63
1.00
306
.0844
.19082
.01091
.0629
.1059
-2.05
1.34
Total
-2.80
2.63
378
.0564
.31355
.01613
.0247
.0881
.00
72
.1050
.39639
.04671
.0119
.1981
-1.36
1.00
1.00
306
.1316
.93697
.05356
.0262
.2370
-15.12
1.00
Total
378
.1265
.86020
.04424
.0395
.2135
-15.12
1.00
.00
72
1.1447
.83235
.09809
.9491
1.3402
.42
6.82
1.00
306
1.1837
.72083
.04121
1.1027
1.2648
.25
6.51
Total
378
1.1763
.74234
.03818
1.1012
1.2514
.25
6.82
.00
72
6129.9290
12463.49563
1468.83705
3201.1511
9058.7069
248.14
85913.50
1.00
306
21576.2876
97755.05651
5588.28463
10579.8156
32572.7596
294.15
1155109
Total
378
18634.1240
88301.55432
4541.74188
9703.8042
27564.4439
248.14
1155109
.00
72
.5944
.22530
.02655
.5415
.6473
.11
1.44
1.00
306
.5534
.38156
.02181
.5105
.5963
.00
5.75
Total
378
.5612
.35721
.01837
.5251
.5973
.00
5.75
Source: Developed for this research
Table A12: ANOVA: comparing payers and non-payers (1996)
Sum of Squares
Et/At
Yt/BEt
dAt/At
Vt/At
Between Groups
.303
Within Groups
2.201
376
.006
Total
2.505
377
Between Groups
1.261
1
1.261
Within Groups
35.803
376
.095
Total
37.064
377
Between Groups
.041
1
.041
Within Groups
278.918
376
.742
Total
278.959
377
.089
1
.089
207.665
376
.552
Between Groups
Total
Lt/At
Mean Square
1
Within Groups
At
df
.303
207.754
377
13906388138.733
1
13906388138.733
Within Groups
2925624626874.058
376
7780916560.835
Total
2939531015012.791
377
Between Groups
Between Groups
.098
1
.098
Within Groups
48.008
376
.128
Total
48.106
377
F
Sig.
51.805
.000
13.242
.000
.056
.814
.161
.688
1.787
.182
.767
.382
Source: Developed for this research
258
DBA Thesis: Malinee Ronapat
Table A13: Descriptive statistics: comparing payers and non-payers (1997)
N
Mean
Std. Deviation
95% Confidence Interval for Mean
Std. Error
Lower Bound
Et/At
Yt/BEt
dAt/At
Vt/At
At
Lt/At
Minimum
Maximum
Upper Bound
.00
112
-.1964
.34283
.03239
-.2605
-.1322
-2.08
.38
1.00
281
-.0536
.22506
.01343
-.0801
-.0272
-1.60
.51
Total
393
-.0943
.27133
.01369
-.1212
-.0674
-2.08
.51
.00
112
-3.9705
31.00026
2.92925
-9.7750
1.8340
-325.37
13.21
1.00
281
-2.4684
33.22680
1.98214
-6.3702
1.4334
-550.44
10.03
Total
393
-2.8965
32.57562
1.64322
-6.1271
.3342
-550.44
13.21
.00
112
-.0466
.71001
.06709
-.1796
.0863
-4.33
1.00
1.00
281
.0014
.43729
.02609
-.0500
.0527
-2.98
1.00
Total
393
-.0123
.52897
.02668
-.0647
.0402
-4.33
1.00
.00
112
1.0166
.44518
.04207
.9333
1.1000
.40
3.49
1.00
281
1.0482
.57407
.03425
.9808
1.1156
.08
5.09
Total
393
1.0392
.54011
.02724
.9856
1.0928
.08
5.09
.00
112
9047.9238
21683.09435
2048.85983
4987.9712
13107.8763
377.12
190277.4
1.00
281
25384.9138
122294.36702
7295.47030
11023.9811
39745.8464
248.85
1408619
Total
393
20729.0795
104261.44910
5259.29439
10389.1273
31069.0316
248.85
1408619
.00
112
.8211
.39637
.03745
.7469
.8953
.18
3.46
1.00
281
.6258
.31882
.01902
.5883
.6632
.00
2.03
Total
393
.6814
.35339
.01783
.6464
.7165
.00
3.46
Source: Developed for this research
Table A14: ANOVA: comparing payers and non-payers (1997)
Sum of Squares
Et/At
Yt/BEt
dAt/At
Vt/At
At
Lt/At
Between Groups
df
Mean Square
1.631
1
1.631
Within Groups
27.228
391
.070
Total
28.860
392
Between Groups
180.685
1
180.685
Within Groups
415798.385
391
1063.423
Total
415979.070
392
.185
1
.185
Within Groups
109.499
391
.280
Total
109.684
392
Between Groups
Between Groups
.080
1
.080
Within Groups
114.273
391
.292
Total
114.353
392
Between Groups
21373511506.616
1
21373511506.616
Within Groups
4239842797717.739
391
10843587717.948
Total
4261216309224.355
392
Between Groups
3.055
1
3.055
Within Groups
45.901
391
.117
Total
48.956
392
F
Sig.
23.423
.000
.170
.680
.659
.417
.273
.602
1.971
.161
26.023
.000
Source: Developed for this research
259
DBA Thesis: Malinee Ronapat
Table A15: Descriptive statistics: comparing payers and non-payers (1998)
N
Mean
Std. Deviation
Std. Error
95% Confidence Interval for Mean
Lower Bound
Et/At
Yt/BEt
dAt/At
Vt/At
At
Lt/At
.00
278
.0195
.70414
.04223
1.00
110
.0956
.09681
Total
388
.0411
.59892
.00
278
-9.8999
1.00
110
Total
.00
1.00
Minimum
-.0636
.1026
-2.75
.00923
.0774
.1139
-.15
.44
.03041
-.0187
.1009
-2.75
10.23
155.80777
9.34473
-28.2956
8.4958
-2597.13
25.05
.1054
.39044
.03723
.0316
.1792
-.77
3.43
388
-7.0633
131.89511
6.69596
-20.2283
6.1017
-2597.13
25.05
278
-1.2208
15.44686
.92644
-3.0445
.6030
-253.88
1.00
110
-.0236
.20114
.01918
-.0616
.0144
-1.12
1.00
Total
388
-.8814
13.08006
.66404
-2.1869
.4242
-253.88
1.00
.00
278
2.2227
15.10458
.90591
.4393
4.0060
.22
239.23
1.00
110
1.0109
.59571
.05680
.8984
1.1235
.20
3.87
Total
388
1.8791
12.79449
.64954
.6021
3.1562
.20
239.23
.00
278
9186.2069
23507.93125
1409.91202
6410.7034
11961.7105
20.38
208724.5
1.00
110
47848.2083
189230.91295
18042.45962
12088.6399
83607.7767
281.56
1266949
Total
388
20147.0836
103853.32174
5272.35360
9781.0418
30513.1254
20.38
1266949
.00
278
1.7203
12.65716
.75913
.2259
3.2146
.01
207.51
1.00
110
.3895
.24863
.02371
.3426
.4365
.01
.95
Total
388
1.3430
10.72594
.54453
.2724
2.4136
.01
207.51
Source: Developed for this research
Table A16: ANOVA: comparing payers and non-payers (1998)
Sum of Squares
Et/At
Between Groups
Within Groups
Total
Yt/BEt
dAt/At
Vt/At
Mean Square
1
.457
138.362
386
.358
138.819
387
1
7889.733
Within Groups
6724485.843
386
17420.948
Total
6732375.575
387
Between Groups
112.966
1
112.966
Within Groups
66098.111
386
171.239
Total
66211.077
387
115.725
1
115.725
63235.767
386
163.823
Between Groups
Total
Lt/At
df
.457
7889.733
Between Groups
Within Groups
At
Maximum
Upper Bound
63351.492
387
117807901049.125
1
117807901049.125
Within Groups
4056185411677.855
386
10508252361.860
Total
4173993312726.979
387
Between Groups
Between Groups
139.564
1
139.564
Within Groups
44383.159
386
114.982
Total
44522.723
387
F
Sig.
1.275
.260
.453
.501
.660
.417
.706
.401
11.211
.001
1.214
.271
Source: Developed for this research
260
10.23
DBA Thesis: Malinee Ronapat
Table A17: Descriptive statistics: comparing payers and non-payers (1999)
N
Mean
Std. Deviation
Std. Error
95% Confidence Interval for Mean
Lower Bound
Et/At
Yt/BEt
dAt/At
Vt/At
At
Lt/At
.00
270
1.00
Total
-.0840
.43517
.02648
115
.0884
.07908
385
-.0325
.37518
.00
270
-.9563
1.00
115
Total
.00
1.00
Minimum
Maximum
-3.55
1.60
Upper Bound
-.1362
-.0319
.00737
.0738
.1030
-.16
.36
.01912
-.0701
.0051
-3.55
1.60
6.90977
.42052
-1.7842
-.1284
-88.22
4.87
.1087
.12965
.01209
.0848
.1327
-.40
.43
385
-.6382
5.80427
.29581
-1.2198
-.0565
-88.22
4.87
270
-.1056
.56979
.03468
-.1739
-.0373
-4.27
1.00
115
.0543
.18969
.01769
.0193
.0894
-.33
1.00
Total
385
-.0578
.49345
.02515
-.1073
-.0084
-4.27
1.00
.00
270
1.3639
1.47699
.08989
1.1869
1.5409
.07
17.50
1.00
115
1.0019
.71886
.06703
.8691
1.1347
.27
6.98
Total
385
1.2558
1.30733
.06663
1.1248
1.3868
.07
17.50
.00
270
29084.0368
119161.81966
7251.95740
14806.2237
43361.8499
219.57
1182878
1.00
115
3034.9757
3855.20365
359.49959
2322.8098
3747.1417
288.50
21983.67
Total
385
21303.1484
100468.89164
5120.36885
11235.6791
31370.6177
219.57
1182878
.00
270
1.0047
1.17016
.07121
.8645
1.1449
.05
10.87
1.00
115
.3374
.19475
.01816
.3015
.3734
.01
.79
Total
385
.8054
1.03149
.05257
.7020
.9087
.01
10.87
Source: Developed for this research
Table A18: ANOVA: comparing payers and non-payers (1999)
Sum of Squares
Et/At
Yt/BEt
dAt/At
Vt/At
Between Groups
2.397
Within Groups
51.655
383
.135
Total
54.052
384
Between Groups
91.477
1
91.477
Within Groups
12845.297
383
33.539
Total
12936.774
384
Between Groups
2.062
1
2.062
Within Groups
91.437
383
.239
Total
93.499
384
Between Groups
10.566
1
10.566
645.736
383
1.686
Total
Lt/At
Mean Square
1
Within Groups
At
df
2.397
656.302
384
54724905688.052
1
54724905688.052
Within Groups
3821370398100.955
383
9977468402.352
Total
3876095303789.008
384
Between Groups
Between Groups
35.904
1
35.904
Within Groups
372.657
383
.973
Total
408.562
384
F
Sig.
17.776
.000
2.728
.099
8.638
.003
6.267
.013
5.485
.020
36.901
.000
Source: Developed for this research
261
DBA Thesis: Malinee Ronapat
Table A19: Descriptive statistics comparing payers and non-payers (2000)
N
Mean
Std. Deviation
95% Confidence Interval for Mean
Std. Error
Lower Bound
Et/At
Yt/BEt
dAt/At
Vt/At
At
Lt/At
.00
238
1.00
Total
-.1070
.84208
.05458
130
.0850
.06391
368
-.0392
.68396
.00
238
-.7258
1.00
130
Total
.00
1.00
Minimum
Maximum
Upper Bound
-.2145
.0005
-10.38
1.09
.00560
.0739
.0961
-.16
.30
.03565
-.1093
.0309
-10.38
1.09
9.11072
.59056
-1.8892
.4376
-115.08
23.63
.1092
.09381
.00823
.0929
.1255
-.20
.36
368
-.4308
7.33251
.38223
-1.1825
.3208
-115.08
23.63
238
-.3209
2.06700
.13398
-.5849
-.0570
-28.92
1.00
130
.0314
.29794
.02613
-.0203
.0831
-1.86
1.00
Total
368
-.1964
1.67890
.08752
-.3685
-.0243
-28.92
1.00
.00
238
1.4839
2.69624
.17477
1.1396
1.8282
.22
35.69
1.00
130
.8666
.54561
.04785
.7719
.9613
.13
5.55
Total
368
1.2658
2.21055
.11523
1.0392
1.4924
.13
35.69
.00
238
32136.7395
129928.04388
8421.98418
15545.2284
48728.2505
187.39
1236328
1.00
130
3893.1621
6539.65238
573.56546
2758.3488
5027.9754
302.36
51960.20
Total
368
22159.3888
105353.32673
5491.92180
11359.8049
32958.9726
187.39
1236328
.00
238
1.2658
2.71992
.17631
.9184
1.6131
.00
35.51
1.00
130
.3164
.18941
.01661
.2835
.3492
.00
.73
Total
368
.9304
2.23530
.11652
.7012
1.1595
.00
35.51
Source: Developed for this research
Table A20: ANOVA comparing payers and non-payers (2000)
Sum of Squares
Et/At
Yt/BEt
dAt/At
Vt/At
Between Groups
3.099
Within Groups
168.584
366
.461
Total
171.683
367
58.614
1
58.614
Within Groups
19673.390
366
53.752
Total
19732.005
367
Between Groups
Between Groups
10.436
1
10.436
Within Groups
1024.034
366
2.798
Total
1034.470
367
32.035
1
32.035
1761.329
366
4.812
Between Groups
Total
Lt/At
Mean Square
1
Within Groups
At
df
3.099
1793.364
367
67067466394.794
1
67067466394.794
Within Groups
4006384241002.584
366
10946405030.062
Total
4073451707397.378
367
Between Groups
Between Groups
75.784
1
75.784
Within Groups
1757.952
366
4.803
Total
1833.736
367
F
Sig.
6.728
.010
1.090
.297
3.730
.054
6.657
.010
6.127
.014
15.778
.000
Source: Developed for this research
262
DBA Thesis: Malinee Ronapat
Table A21: Descriptive statistics comparing payers and non-payers (2001)
N
Mean
Std. Deviation
Std. Error
95% Confidence Interval for Mean
Lower Bound
Et/At
Yt/BEt
dAt/At
Vt/At
At
Lt/At
Minimum
Maximum
7.28
Upper Bound
.00
223
.0238
.54671
.03661
-.0484
.0959
-1.41
1.00
153
-.0728
2.03734
.16471
-.3982
.2527
-25.09
.41
Total
376
-.0155
1.36442
.07036
-.1539
.1228
-25.09
7.28
.00
223
-4.4534
56.32197
3.77160
-11.8861
2.9793
-823.12
3.88
1.00
153
.1200
.10318
.00834
.1035
.1365
-.13
.52
Total
376
-2.5924
43.39339
2.23784
-6.9927
1.8079
-823.12
3.88
.00
223
-2.1588
31.32864
2.09792
-6.2932
1.9756
-467.83
1.00
1.00
153
-.0202
1.12787
.09118
-.2004
.1599
-13.61
1.00
Total
376
-1.2886
24.13837
1.24484
-3.7363
1.1592
-467.83
1.00
.00
223
1.9945
5.08704
.34065
1.3232
2.6659
.26
55.65
1.00
153
.9897
.58141
.04700
.8968
1.0826
.18
5.31
Total
376
1.5856
3.96246
.20435
1.1838
1.9875
.18
55.65
.00
223
31918.3958
133049.88853
8909.67953
14360.0244
49476.7671
72.52
1248748
1.00
153
5289.2451
11474.18865
927.63317
3456.5259
7121.9643
265.60
100742.0
Total
376
21082.5978
103463.52221
5335.72375
10590.9099
31574.2856
72.52
1248748
.00
223
1.6500
5.02831
.33672
.9865
2.3136
.01
55.19
1.00
153
.3258
.20244
.01637
.2934
.3581
.00
.88
Total
376
1.1112
3.92544
.20244
.7131
1.5092
.00
55.19
Source: Developed for this research
Table A22: ANOVA comparing payers and non-payers (2001)
Sum of Squares
Et/At
Yt/BEt
dAt/At
Vt/At
At
Lt/At
Between Groups
df
Mean Square
.846
1
.846
Within Groups
697.266
374
1.864
Total
698.112
375
Between Groups
1897.999
1
1897.999
Within Groups
704222.008
374
1882.947
Total
706120.007
375
Between Groups
415.004
1
415.004
Within Groups
218082.793
374
583.109
Total
218497.797
375
91.620
1
91.620
Within Groups
5796.295
374
15.498
Total
5887.915
375
Between Groups
Between Groups
64346225910.489
1
64346225910.489
Within Groups
3949916434699.211
374
10561273889.570
Total
4014262660609.700
375
Between Groups
159.137
1
159.137
Within Groups
5619.254
374
15.025
Total
5778.392
375
F
Sig.
.454
.501
1.008
.316
.712
.399
5.912
.016
6.093
.014
10.592
.001
Source: Developed for this research
263
DBA Thesis: Malinee Ronapat
Table A23: Descriptive statistics comparing payers and non-payers (2002)
N
Mean
Std. Deviation
Std. Error
95% Confidence Interval for Mean
Lower Bound
Et/At
Yt/BEt
dAt/At
Vt/At
At
Lt/At
Minimum
Maximum
-1.52
2.09
Upper Bound
.00
207
.0234
.25154
.01748
-.0111
.0579
1.00
179
.0908
.06516
.00487
.0812
.1004
-.09
.39
Total
386
.0547
.19223
.00978
.0354
.0739
-1.52
2.09
.00
207
.0161
1.03165
.07170
-.1253
.1574
-6.14
8.87
1.00
179
.1271
.10380
.00776
.1118
.1424
-.14
.58
Total
386
.0676
.75995
.03868
-.0085
.1436
-6.14
8.87
.00
207
-.0964
1.56272
.10862
-.3106
.1177
-19.32
1.00
1.00
179
.0473
.47023
.03515
-.0221
.1167
-4.81
1.00
Total
386
-.0298
1.18914
.06053
-.1488
.0892
-19.32
1.00
.00
207
1.5445
2.10332
.14619
1.2562
1.8327
.30
23.35
1.00
179
1.1108
.58533
.04375
1.0245
1.1971
.28
5.13
Total
386
1.3434
1.60387
.08163
1.1829
1.5039
.28
23.35
.00
207
35133.9306
143651.74933
9984.48678
15449.0490
54818.8123
93.47
1245098
1.00
179
8912.4088
27779.52726
2076.33935
4815.0004
13009.8172
261.19
234465.8
Total
386
22974.2094
107562.65648
5474.79507
12209.9694
33738.4493
93.47
1245098
.00
207
1.0454
1.93700
.13463
.7799
1.3108
.00
23.26
1.00
179
.3296
.20163
.01507
.2999
.3594
.00
.84
Total
386
.7134
1.46767
.07470
.5666
.8603
.00
23.26
Source: Developed for this research
Table A24: ANOVA comparing payers and non-payers (2002)
Sum of Squares
Et/At
Yt/BEt
dAt/At
Vt/At
Between Groups
.437
Within Groups
13.790
384
.036
Total
14.227
385
1.183
1
1.183
Within Groups
221.165
384
.576
Total
222.348
385
Between Groups
Between Groups
1.983
1
1.983
Within Groups
542.430
384
1.413
Total
544.412
385
18.053
1
18.053
972.315
384
2.532
Between Groups
Total
Lt/At
Mean Square
1
Within Groups
At
df
.437
990.368
385
66001204221.708
1
66001204221.708
Within Groups
4388342947562.820
384
11427976425.945
Total
4454344151784.528
385
Between Groups
Between Groups
49.175
1
49.175
Within Groups
780.140
384
2.032
Total
829.315
385
F
Sig.
12.163
.001
2.055
.153
1.404
.237
7.130
.008
5.775
.017
24.205
.000
Source: Developed for this research
264
DBA Thesis: Malinee Ronapat
APPENDIX B:
ANOVA COMPARING PAYERS, FORMER PAYERS, NEVER
PAID FIRMS
265
DBA Thesis: Malinee Ronapat
Table B1: ANOVA comparing payers, former payers and never paid firms (1991)
Dependent
Variable
(I) PAYER3
(J) PAYER3
Et/At
1.00
2.00
3.00
-.0165
.01502
.517
Lower Bound
-.0519
Upper Bound
.0190
-.0042
.02026
.976
-.0520
.0436
2.00
1.00
.0165
.01502
.517
-.0190
.0519
3.00
.0122
.02425
.869
-.0450
.0695
1.00
2.00
.0042
.02026
.976
-.0436
.0520
-.0122
.02425
.869
-.0695
.0450
2.00
-.0097
.03809
.965
-.0996
.0801
3.00
.0203
.05137
.918
-.1010
.1415
1.00
3.00
1.00
.0097
.0300
.03809
.06150
.965
.877
-.0801
-.1151
.0996
.1751
3.00
Yt/BEt
1.00
2.00
3.00
dAt/At
Vt/At
95% Confidence Interval
Sig.
-.0203
.05137
.918
-.1415
.1010
-.0300
.06150
.877
-.1751
.1151
1.00
2.00
.4264(*)
.13627
.006
.1049
.7479
-.3170
.18382
.198
-.7506
.1167
2.00
3.00
1.00
-.4264(*)
.13627
.006
-.7479
-.1049
3.00
-.2242
-.7434(*)
.22005
.002
-1.2625
3.00
1.00
.3170
.18382
.198
-.1167
.7506
1.00
2.00
2.00
3.00
.7434(*)
-2.9025
.22005
1.80915
.002
.246
.2242
-7.1707
1.2625
1.3658
-.0196
2.44035
1.000
-5.7771
5.7378
2.00
1.00
2.9025
1.80915
.246
-1.3658
7.1707
3.00
2.8828
2.92139
.586
-4.0095
9.7751
1.00
2.00
.0196
2.44035
1.000
-5.7378
5.7771
-2.8828
2.92139
.586
-9.7751
4.0095
2.00
9922.2383
11908.01231
.683
-18171.8597
38016.3363
3.00
12865.6516
16062.61734
.703
-25030.2402
50761.5434
1.00
3.00
1.00
-9922.2383
2943.4133
11908.01231
19228.86596
.683
.987
-38016.3363
-42422.4825
18171.8597
48309.3090
-12865.6516
16062.61734
.703
-50761.5434
25030.2402
2.00
-2943.4133
19228.86596
.987
-48309.3090
42422.4825
1.00
2.00
.0110
.05226
.976
-.1123
.1343
3.00
1.00
.2091(*)
.07049
.009
.0428
.3755
2.00
-.0110
.05226
.976
-.1343
.1123
3.00
.1981
.08439
.051
-.0010
.3972
1.00
-.2091(*)
.07049
.009
-.3755
-.0428
-.1981
.08439
.051
-.3972
.0010
1.00
2.00
3.00
Lt/At
Std. Error
2.00
3.00
At
Mean Difference (I-J)
3.00
2.00
* The mean difference is significant at the .05 level.
Source: Developed for this research
Table B2: ANOVA comparing payers, former payers and never paid firms (1992)
266
DBA Thesis: Malinee Ronapat
Dependent
Variable
(I) PAYER3
(J) PAYER3
Et/At
1.00
2.00
3.00
.0713(*)
.02231
.004
Lower Bound
.0187
Upper Bound
.1239
.0192
.02415
.705
-.0377
.0762
2.00
1.00
-.0713(*)
.02231
.004
-.1239
-.0187
3.00
-.0521
.03206
.237
-.1277
.0235
1.00
2.00
-.0192
.02415
.705
-.0762
.0377
.0521
.03206
.237
-.0235
.1277
2.00
.4442
3.00
Yt/BEt
1.00
2.00
3.00
dAt/At
Vt/At
95% Confidence Interval
Sig.
.2848(*)
.06759
.000
.1255
3.00
.0838
.07317
.487
-.0887
.2563
1.00
3.00
1.00
-.2848(*)
-.2010
.06759
.09712
.000
.098
-.4442
-.4300
-.1255
.0280
-.0838
.07317
.487
-.2563
.0887
2.00
.4300
.2010
.09712
.098
-.0280
2.00
.1937
.10814
.174
-.0612
.4487
-.6722(*)
.11707
.000
-.9482
-.3962
2.00
3.00
1.00
-.1937
.10814
.174
-.4487
.0612
3.00
-.8660(*)
.15540
.000
-1.2323
-.4996
3.00
1.00
.6722(*)
.11707
.000
.3962
.9482
1.00
2.00
2.00
3.00
.8660(*)
.0520
.15540
.34961
.000
.988
.4996
-.7722
1.2323
.8763
-.4654
.37850
.437
-1.3577
.4270
2.00
1.00
-.0520
.34961
.988
-.8763
.7722
3.00
-.5174
.50241
.559
-1.7019
.6671
1.00
2.00
.4654
.37850
.437
-.4270
1.3577
.5174
.50241
.559
-.6671
1.7019
2.00
4927.3893
16540.21785
.952
-34069.5252
43924.3039
3.00
10540.8237
17906.96329
.826
-31678.4702
52760.1175
1.00
3.00
1.00
-4927.3893
5613.4343
16540.21785
23769.02047
.952
.970
-43924.3039
-50426.8469
34069.5252
61653.7156
-10540.8237
17906.96329
.826
-52760.1175
31678.4702
1.00
2.00
3.00
Lt/At
Std. Error
1.00
3.00
At
Mean Difference (I-J)
2.00
-5613.4343
23769.02047
.970
-61653.7156
50426.8469
1.00
2.00
-.0822
.06214
.384
-.2287
.0643
-.0391
.06728
.830
-.1977
.1195
2.00
3.00
1.00
.0822
.06214
.384
-.0643
.2287
3.00
.0431
.08930
.880
-.1674
.2536
.0391
.06728
.830
-.1195
.1977
-.0431
.08930
.880
-.2536
.1674
3.00
1.00
2.00
* The mean difference is significant at the .05 level.
Source: Developed for this research
Table B3: ANOVA comparing payers, former payers and never paid firms (1993)
267
DBA Thesis: Malinee Ronapat
Dependent
Variable
(I) PAYER3
(J) PAYER3
Et/At
1.00
2.00
3.00
.0928(*)
.02221
.000
Lower Bound
.0404
Upper Bound
.1451
.0265
.02369
.503
-.0293
.0823
2.00
1.00
-.0928(*)
.02221
.000
-.1451
-.0404
3.00
-.0663
.03113
.086
-.1396
.0071
1.00
2.00
-.0265
.02369
.503
-.0823
.0293
.0663
.03113
.086
-.0071
.1396
2.00
-.0781
.06140
.412
-.2228
.0666
3.00
.0590
.06548
.640
-.0953
.2133
1.00
3.00
1.00
.0781
.1371
.06140
.08607
.412
.250
-.0666
-.0657
.2228
.3399
-.0590
.06548
.640
-.2133
.0953
2.00
.0657
3.00
Yt/BEt
1.00
2.00
3.00
dAt/At
Vt/At
95% Confidence Interval
Sig.
-.1371
.08607
.250
-.3399
2.00
.0230
.10216
.973
-.2177
.2637
-.6446(*)
.10893
.000
-.9012
-.3879
2.00
3.00
1.00
-.0230
.10216
.973
-.2637
.2177
3.00
-.6675(*)
.14319
.000
-1.0049
-.3301
3.00
1.00
.6446(*)
.10893
.000
.3879
.9012
1.00
2.00
2.00
3.00
.6675(*)
.7060
.14319
.57236
.000
.434
.3301
-.6425
1.0049
2.0545
-.9140
.61033
.294
-2.3520
.5240
2.00
1.00
-.7060
.57236
.434
-2.0545
.6425
3.00
-1.6201
.80229
.109
-3.5103
.2702
1.00
2.00
.9140
.61033
.294
-.5240
2.3520
1.6201
.80229
.109
-.2702
3.5103
2.00
7488.3241
14085.92710
.856
-25699.0841
40675.7323
3.00
9680.3420
15020.37202
.796
-25708.6824
45069.3664
1.00
3.00
1.00
-7488.3241
2192.0179
14085.92710
19744.40037
.856
.993
-40675.7323
-44327.1405
25699.0841
48711.1763
-9680.3420
15020.37202
.796
-45069.3664
25708.6824
1.00
2.00
3.00
Lt/At
Std. Error
1.00
3.00
At
Mean Difference (I-J)
2.00
-2192.0179
19744.40037
.993
-48711.1763
44327.1405
1.00
2.00
-.0732
.10171
.752
-.3129
.1664
.1001
.10845
.626
-.1554
.3557
2.00
3.00
1.00
.0732
.10171
.752
-.1664
.3129
3.00
.1733
.14256
.445
-.1625
.5092
1.00
-.1001
.10845
.626
-.3557
.1554
-.1733
.14256
.445
-.5092
.1625
3.00
2.00
* The mean difference is significant at the .05 level.
Source: Developed for this research
Table B4: ANOVA comparing payers, former payers and never paid firms (1994)
Dependent
(I) PAYER3
(J) PAYER3
Mean Difference (I-J)
Std. Error
Sig.
95% Confidence Interval
268
DBA Thesis: Malinee Ronapat
Variable
Et/At
1.00
2.00
3.00
.0831(*)
.01168
.000
Lower Bound
.0556
.0179
.01366
.391
-.0143
.0500
2.00
1.00
-.0831(*)
.01168
.000
-.1106
-.0556
3.00
-.0652(*)
.01718
.001
-.1056
-.0247
1.00
2.00
-.0179
.01366
.391
-.0500
.0143
.0652(*)
.01718
.001
.0247
.1056
2.00
3.00
Yt/BEt
1.00
2.00
3.00
dAt/At
Vt/At
.64580
.002
.7093
3.7504
.0759
.75524
.994
-1.7023
1.8541
1.00
3.00
1.00
-2.2299(*)
-2.1540
.64580
.95023
.002
.062
-3.7504
-4.3913
-.7093
.0834
-.0759
.75524
.994
-1.8541
1.7023
2.00
4.3913
2.1540
.95023
.062
-.0834
2.00
.1899
.08889
.084
-.0194
.3992
-.6721(*)
.10396
.000
-.9168
-.4273
2.00
3.00
1.00
-.1899
.08889
.084
-.3992
.0194
3.00
-.8620(*)
.13080
.000
-1.1700
-.5540
3.00
1.00
.6721(*)
.10396
.000
.4273
.9168
1.00
2.00
2.00
3.00
.8620(*)
.1677
.13080
.19769
.000
.673
.5540
-.2978
1.1700
.6332
-.6248(*)
.23119
.020
-1.1692
-.0805
2.00
1.00
-.1677
.19769
.673
-.6332
.2978
3.00
-.7925(*)
.29088
.019
-1.4774
-.1077
1.00
2.00
.6248(*)
.23119
.020
.0805
1.1692
.7925(*)
.29088
.019
.1077
1.4774
2.00
1.00
14544.8699
13583.15096
.533
-17436.6886
46526.4284
3.00
8831.2903
15885.01192
.843
-28570.0048
46232.5855
1.00
3.00
1.00
-14544.8699
-5713.5796
13583.15096
19986.31626
.533
.956
-46526.4284
-52771.4047
17436.6886
41344.2455
-8831.2903
15885.01192
.843
-46232.5855
28570.0048
2.00
5713.5796
19986.31626
.956
-41344.2455
52771.4047
1.00
2.00
-.1092
.04641
.050
-.2185
.0001
.0856
.05428
.257
-.0422
.2135
2.00
3.00
1.00
.1092
.04641
.050
-.0001
.2185
3.00
.1949(*)
.06829
.013
.0341
.3557
-.0856
.05428
.257
-.2135
.0422
-.1949(*)
.06829
.013
-.3557
-.0341
2.00
3.00
Lt/At
2.2299(*)
3.00
1.00
3.00
At
Upper Bound
.1106
3.00
1.00
2.00
* The mean difference is significant at the .05 level.
Source: Developed for this research
Table B5: ANOVA comparing payers, former payers and never paid firms (1995)
Dependent
(I) PAYER3
(J) PAYER3
Mean Difference (I-J)
Std. Error
Sig.
95% Confidence Interval
269
DBA Thesis: Malinee Ronapat
Variable
Et/At
Yt/BEt
.0692(*)
.00929
.000
.0607(*)
.01846
.003
.0172
.1041
2.00
1.00
-.0692(*)
.00929
.000
-.0911
-.0474
3.00
-.0086
.02011
.905
-.0559
.0388
3.00
1.00
2.00
-.0607(*)
.01846
.003
-.1041
-.0172
.0086
.02011
.905
-.0388
.0559
1.00
2.00
.0524
.04729
.509
-.0589
.1638
3.00
.2345(*)
.09402
.035
.0131
.4558
1.00
3.00
1.00
-.0524
.1820
.04729
.10240
.509
.179
-.1638
-.0590
.0589
.4231
-.2345(*)
.09402
.035
-.4558
-.0131
2.00
-.1820
.10240
.179
-.4231
.0590
1.00
2.00
.1536(*)
.05909
.026
.0145
.2927
-.2630
.11747
.066
-.5395
.0135
2.00
3.00
1.00
-.1536(*)
.05909
.026
-.2927
-.0145
3.00
-.4166(*)
.12795
.004
-.7178
-.1154
3.00
1.00
.2630
.11747
.066
-.0135
.5395
1.00
2.00
2.00
3.00
.4166(*)
.1849
.12795
.11991
.004
.273
.1154
-.0974
.7178
.4672
-.6646(*)
.23839
.015
-1.2258
-.1035
2.00
1.00
-.1849
.11991
.273
-.4672
.0974
3.00
-.8495(*)
.25965
.003
-1.4607
-.2383
1.00
2.00
.6646(*)
.23839
.015
.1035
1.2258
.8495(*)
.25965
.003
.2383
1.4607
2.00
3.00
Vt/At
3.00
At
1.00
16523.7653
13210.49180
.424
-14573.6465
47621.1771
3.00
-616.3515
26262.29757
1.000
-62437.6292
61204.9262
1.00
3.00
1.00
-16523.7653
-17140.1168
13210.49180
28604.97886
.424
.821
-47621.1771
-84476.0510
14573.6465
50195.8174
616.3515
26262.29757
1.000
-61204.9262
62437.6292
2.00
17140.1168
28604.97886
.821
-50195.8174
84476.0510
1.00
2.00
-.0942(*)
.03410
.017
-.1745
-.0140
-.0588
.06780
.661
-.2184
.1008
2.00
3.00
1.00
.0942(*)
.03410
.017
.0140
.1745
3.00
.0354
.07384
.881
-.1384
.2092
.0588
.06780
.661
-.1008
.2184
-.0354
.07384
.881
-.2092
.1384
2.00
3.00
Lt/At
Upper Bound
.0911
2.00
3.00
2.00
dAt/At
Lower Bound
.0474
1.00
3.00
1.00
2.00
* The mean difference is significant at the .05 level.
Source: Developed for this research
Table B6: ANOVA comparing payers, former payers and never paid firms (1996)
Dependent
(I) PAYER3
(J) PAYER3
Mean Difference (I-J)
Std. Error
Sig.
95% Confidence Interval
270
DBA Thesis: Malinee Ronapat
Variable
Et/At
Yt/BEt
.0745(*)
.01122
.000
.0645(*)
.01909
.002
.0196
.1094
2.00
1.00
-.0745(*)
.01122
.000
-.1009
-.0481
3.00
-.0100
.02126
.885
-.0600
.0400
3.00
1.00
2.00
-.0645(*)
.01909
.002
-.1094
-.0196
.0100
.02126
.885
-.0400
.0600
1.00
2.00
.3148
3.00
Vt/At
.04467
.000
.1046
-.0555
.07601
.746
-.2343
.1234
1.00
3.00
1.00
-.2097(*)
-.2652(*)
.04467
.08464
.000
.005
-.3148
-.4643
-.1046
-.0660
.0555
.07601
.746
-.1234
.2343
2.00
.2652(*)
.08464
.005
.0660
.4643
2.00
.1569
.12541
.424
-.1382
.4520
-.3949
.21337
.155
-.8970
.1072
2.00
3.00
1.00
-.1569
.12541
.424
-.4520
.1382
3.00
-.5518
.23762
.054
-1.1109
.0073
3.00
1.00
.3949
.21337
.155
-.1072
.8970
1.00
2.00
2.00
3.00
.5518
.2074
.23762
.10724
.054
.131
-.0073
-.0449
1.1109
.4598
-.5056(*)
.18246
.016
-.9349
-.0762
2.00
1.00
-.2074
.10724
.131
-.4598
.0449
3.00
-.7130(*)
.20319
.001
-1.1911
-.2349
1.00
2.00
.5056(*)
.18246
.016
.0762
.9349
.7130(*)
.20319
.001
.2349
1.1911
2.00
1.00
18742.6503
12930.55809
.317
-11683.6920
49168.9926
3.00
4781.8852
21999.95609
.974
-46985.2690
56549.0394
1.00
3.00
1.00
-18742.6503
-13960.7651
12930.55809
24500.00481
.317
.836
-49168.9926
-71610.6768
11683.6920
43689.1466
-4781.8852
21999.95609
.974
-56549.0394
46985.2690
2.00
13960.7651
24500.00481
.836
-43689.1466
71610.6768
1.00
2.00
-.0557
.05238
.538
-.1789
.0676
.0065
.08911
.997
-.2032
.2162
2.00
3.00
1.00
.0557
.05238
.538
-.0676
.1789
3.00
.0622
.09924
.806
-.1714
.2957
1.00
-.0065
.08911
.997
-.2162
.2032
-.0622
.09924
.806
-.2957
.1714
2.00
3.00
Lt/At
.2097(*)
3.00
1.00
3.00
At
Upper Bound
.1009
2.00
3.00
2.00
dAt/At
Lower Bound
.0481
1.00
3.00
2.00
* The mean difference is significant at the .05 level.
Source: Developed for this research
Table B7: ANOVA comparing payers, former payers and never paid firms (1997)
Dependent
(I) PAYER3
(J) PAYER3
Mean Difference (I-J)
Std. Error
Sig.
95% Confidence Interval
271
DBA Thesis: Malinee Ronapat
Variable
Et/At
Yt/BEt
.1355(*)
.03199
.000
.1721(*)
.05847
.010
.0345
.3096
2.00
1.00
-.1355(*)
.03199
.000
-.2108
-.0603
3.00
.0365
.06282
.830
-.1113
.1843
3.00
1.00
2.00
-.1721(*)
.05847
.010
-.3096
-.0345
-.0365
.06282
.830
-.1843
.1113
1.00
2.00
2.3217
3.95331
.827
-6.9792
11.6226
3.00
-1.8509
7.22612
.964
-18.8518
15.1499
1.00
3.00
1.00
-2.3217
-4.1726
3.95331
7.76291
.827
.853
-11.6226
-22.4364
6.9792
14.0911
1.8509
7.22612
.964
-15.1499
18.8518
2.00
-14.0911
22.4364
3.00
Vt/At
4.1726
7.76291
.853
1.00
2.00
.2204(*)
.06005
.001
.0791
.3617
-.6573(*)
.10977
.000
-.9155
-.3990
2.00
3.00
1.00
-.2204(*)
.06005
.001
-.3617
-.0791
3.00
-.8777(*)
.11793
.000
-1.1551
-.6002
3.00
1.00
.6573(*)
.10977
.000
.3990
.9155
1.00
2.00
2.00
3.00
.8777(*)
.0246
.11793
.06556
.000
.925
.6002
-.1296
1.1551
.1788
.0601
.11983
.871
-.2218
.3420
2.00
1.00
-.0246
.06556
.925
-.1788
.1296
3.00
.0355
.12873
.959
-.2673
.3384
1.00
2.00
-.0601
.11983
.871
-.3420
.2218
-.0355
.12873
.959
-.3384
.2673
2.00
3.00
At
1.00
18465.3199
12625.50595
.310
-11238.6592
48169.2990
3.00
7630.1860
23077.72876
.942
-46664.6973
61925.0694
1.00
3.00
1.00
-18465.3199
-10835.1338
12625.50595
24792.06301
.310
.900
-48169.2990
-69163.3247
11238.6592
47493.0571
-7630.1860
23077.72876
.942
-61925.0694
46664.6973
2.00
10835.1338
24792.06301
.900
-47493.0571
69163.3247
1.00
2.00
-.1963(*)
.04155
.000
-.2940
-.0985
-.1913(*)
.07595
.033
-.3700
-.0126
2.00
3.00
1.00
.1963(*)
.04155
.000
.0985
.2940
3.00
.0050
.08159
.998
-.1870
.1969
1.00
.1913(*)
.07595
.033
.0126
.3700
-.0050
.08159
.998
-.1969
.1870
2.00
3.00
Lt/At
Upper Bound
.2108
2.00
3.00
2.00
dAt/At
Lower Bound
.0603
1.00
3.00
2.00
* The mean difference is significant at the .05 level.
Source: Developed for this research
Table B8: ANOVA comparing payers, former payers and never paid firms (1998)
Dependent
(I) PAYER3
(J) PAYER3
Mean Difference (I-J)
Std. Error
Sig.
95% Confidence Interval
272
DBA Thesis: Malinee Ronapat
Variable
Et/At
1.00
2.00
3.00
.0544
.06830
.705
Lower Bound
-.1063
Upper Bound
.2151
.0617
.13506
.891
-.2561
.3795
2.00
1.00
-.0544
.06830
.705
-.2151
.1063
3.00
.0073
.12814
.998
-.2942
.3087
1.00
2.00
-.0617
.13506
.891
-.3795
.2561
-.0073
.12814
.998
-.3087
.2942
2.00
3.00
Yt/BEt
1.00
10.8456
15.04379
.751
-24.5496
46.2408
3.00
.8539
29.74597
1.000
-69.1328
70.8406
1.00
3.00
1.00
-10.8456
-9.9917
15.04379
28.22295
.751
.933
-46.2408
-76.3950
24.5496
56.4116
-.8539
29.74597
1.000
-70.8406
69.1328
2.00
9.9917
28.22295
.933
-56.4116
76.3950
1.00
2.00
1.3075
1.49124
.655
-2.2011
4.8161
-.1511
2.94862
.999
-7.0887
6.7864
2.00
3.00
1.00
-1.3075
1.49124
.655
-4.8161
2.2011
3.00
-1.4587
2.79765
.861
-8.0410
5.1237
3.00
1.00
.1511
2.94862
.999
-6.7864
7.0887
1.00
2.00
2.00
3.00
1.4587
-1.3278
2.79765
1.45860
.861
.634
-5.1237
-4.7596
8.0410
2.1040
.0474
2.88409
1.000
-6.7384
6.8331
2.00
1.00
1.3278
1.45860
.634
-2.1040
4.7596
3.00
1.3751
2.73642
.870
-5.0631
7.8134
1.00
2.00
-.0474
2.88409
1.000
-6.8331
6.7384
-1.3751
2.73642
.870
-7.8134
5.0631
2.00
66551.4820
2.00
3.00
dAt/At
Vt/At
3.00
At
1.00
39057.1241(*)
11685.73697
.003
11562.7662
3.00
29663.9558
23106.11544
.405
-24700.4209
84028.3325
1.00
3.00
1.00
-39057.1241(*)
-9393.1683
11685.73697
21923.05977
.003
.904
-66551.4820
-60974.0358
-11562.7662
42187.6992
-29663.9558
23106.11544
.405
-84028.3325
24700.4209
2.00
9393.1683
21923.05977
.904
-42187.6992
60974.0358
1.00
2.00
-1.4221
1.22204
.476
-4.2973
1.4532
-.3850
2.41634
.986
-6.0702
5.3002
2.00
3.00
1.00
1.4221
1.22204
.476
-1.4532
4.2973
3.00
1.0371
2.29262
.893
-4.3570
6.4312
.3850
2.41634
.986
-5.3002
6.0702
-1.0371
2.29262
.893
-6.4312
4.3570
2.00
3.00
Lt/At
3.00
1.00
2.00
* The mean difference is significant at the .05 level.
Source: Developed for this research
Table B9: ANOVA comparing payers, former payers and never paid firms (1999)
Dependent
Variable
(I) PAYER3
(J) PAYER3
Mean Difference (I-J)
Std. Error
Sig.
95% Confidence Interval
273
DBA Thesis: Malinee Ronapat
Et/At
1.00
2.00
3.00
.1746(*)
.04164
.000
Lower Bound
.0766
.1537
.07748
.118
-.0287
.3360
2.00
1.00
-.1746(*)
.04164
.000
-.2726
-.0766
3.00
-.0209
.07340
.956
-.1936
.1518
1.00
2.00
-.1537
.07748
.118
-.3360
.0287
.0209
.07340
.956
-.1518
.1936
2.00
.7654
.65140
.469
-.7672
2.2981
3.00
3.6542(*)
1.21202
.008
.8025
6.5060
1.00
3.00
1.00
-.7654
2.8888(*)
.65140
1.14806
.469
.033
-2.2981
.1875
.7672
5.5901
-3.6542(*)
1.21202
.008
-6.5060
-.8025
2.00
-2.8888(*)
1.14806
.033
-5.5901
-.1875
1.00
2.00
.1855(*)
.05495
.002
.0562
.3148
-.0615
.10224
.819
-.3021
.1790
2.00
3.00
1.00
-.1855(*)
.05495
.002
-.3148
-.0562
3.00
-.2470(*)
.09684
.030
-.4749
-.0192
3.00
1.00
.0615
.10224
.819
-.1790
.3021
1.00
2.00
2.00
3.00
.2470(*)
-.3729(*)
.09684
.14722
.030
.031
.0192
-.7193
.4749
-.0265
-.2673
.27393
.593
-.9118
.3773
2.00
1.00
.3729(*)
.14722
.031
.0265
.7193
3.00
.1056
.25948
.913
-.5049
.7162
1.00
2.00
.2673
.27393
.593
-.3773
.9118
-.1056
.25948
.913
-.7162
.5049
2.00
-26395.4496
11327.64108
.053
-53048.0880
257.1888
3.00
-23055.2743
21076.67301
.518
-72646.2651
26535.7165
1.00
3.00
1.00
26395.4496
3340.1753
11327.64108
19964.44118
.053
.985
-257.1888
-43633.8617
53048.0880
50314.2124
23055.2743
21076.67301
.518
-26535.7165
72646.2651
2.00
-3340.1753
19964.44118
.985
-50314.2124
43633.8617
1.00
2.00
-.6862(*)
.11174
.000
-.9492
-.4233
-.5028(*)
.20791
.042
-.9920
-.0136
2.00
3.00
1.00
.6862(*)
.11174
.000
.4233
.9492
3.00
.1835
.19694
.621
-.2799
.6468
1.00
.5028(*)
.20791
.042
.0136
.9920
-.1835
.19694
.621
-.6468
.2799
3.00
Yt/BEt
1.00
2.00
3.00
dAt/At
Vt/At
3.00
At
1.00
2.00
3.00
Lt/At
3.00
2.00
* The mean difference is significant at the .05 level.
Upper Bound
.2726
Source: Developed for this research
Table B10: ANOVA comparing payers, former payers and never paid firms (2000)
Dependent
(I) PAYER3
(J) PAYER3
Mean Difference (I-J)
Std. Error
Sig.
95% Confidence Interval
274
DBA Thesis: Malinee Ronapat
Variable
Et/At
1.00
2.00
3.00
.1994(*)
.07603
.025
Lower Bound
.0205
.1425
.13580
.546
-.1771
.4621
2.00
1.00
-.1994(*)
.07603
.025
-.3783
-.0205
3.00
-.0569
.13085
.901
-.3649
.2510
1.00
2.00
-.1425
.13580
.546
-.4621
.1771
.0569
.13085
.901
-.2510
.3649
2.00
3.00
Yt/BEt
1.00
2.00
3.00
dAt/At
Vt/At
.82153
.535
-1.0567
2.8099
.5571
1.46731
.924
-2.8960
4.0101
1.00
3.00
1.00
-.8766
-.3195
.82153
1.41379
.535
.972
-2.8099
-3.6466
1.0567
3.0076
-.5571
1.46731
.924
-4.0101
2.8960
2.00
.3195
1.41379
.972
-3.0076
3.6466
2.00
.4208
.18676
.064
-.0187
.8603
-.1050
.33357
.947
-.8900
.6800
2.00
3.00
1.00
-.4208
.18676
.064
-.8603
.0187
3.00
-.5258
.32140
.232
-1.2822
.2305
3.00
1.00
.1050
.33357
.947
-.6800
.8900
1.00
2.00
2.00
3.00
.5258
-.6652(*)
.32140
.24557
.232
.019
-.2305
-1.2431
1.2822
-.0873
-.2972
.43862
.777
-1.3294
.7350
2.00
1.00
.6652(*)
.24557
.019
.0873
1.2431
3.00
.3680
.42262
.659
-.6265
1.3626
1.00
2.00
.2972
.43862
.777
-.7350
1.3294
-.3680
.42262
.659
-1.3626
.6265
2.00
1.00
-29499.1726(*)
11720.66169
.033
-57081.5743
-1916.7709
3.00
-19859.4418
20934.10720
.610
-69123.9774
29405.0938
1.00
3.00
1.00
29499.1726(*)
9639.7308
11720.66169
20170.49182
.033
.882
1916.7709
-37827.7777
57081.5743
57107.2393
19859.4418
20934.10720
.610
-29405.0938
69123.9774
2.00
-9639.7308
20170.49182
.882
-57107.2393
37827.7777
1.00
2.00
-1.0015(*)
.24529
.000
-1.5788
-.4243
-.6015
.43811
.356
-1.6325
.4295
2.00
3.00
1.00
1.0015(*)
.24529
.000
.4243
1.5788
3.00
.4000
.42213
.610
-.5934
1.3934
.6015
.43811
.356
-.4295
1.6325
-.4000
.42213
.610
-1.3934
.5934
2.00
3.00
Lt/At
.8766
3.00
1.00
3.00
At
Upper Bound
.3783
3.00
1.00
2.00
* The mean difference is significant at the .05 level.
Source: Developed for this research
Table B11: ANOVA comparing payers, former payers and never paid firms (2001)
Dependent
(I) PAYER3
(J) PAYER3
Mean Difference (I-J)
Std. Error
Sig.
95% Confidence Interval
275
DBA Thesis: Malinee Ronapat
Variable
Et/At
1.00
2.00
3.00
-.1045
.14977
.765
Lower Bound
-.4569
Upper Bound
.2479
-.0600
.24279
.967
-.6313
.5113
2.00
1.00
.1045
.14977
.765
-.2479
.4569
3.00
.0445
.23863
.981
-.5170
.6060
1.00
2.00
.0600
.24279
.967
-.5113
.6313
-.0445
.23863
.981
-.6060
.5170
2.00
3.00
Yt/BEt
1.00
5.5240
4.75679
.477
-5.6692
16.7173
3.00
.2246
7.71115
1.000
-17.9206
18.3698
1.00
3.00
1.00
-5.5240
-5.2994
4.75679
7.57902
.477
.764
-16.7173
-23.1337
5.6692
12.5348
-.2246
7.71115
1.000
-18.3698
17.9206
2.00
5.2994
7.57902
.764
-12.5348
23.1337
1.00
2.00
2.6670
2.64710
.573
-3.5619
8.8959
-.2791
4.29117
.998
-10.3767
9.8185
2.00
3.00
1.00
-2.6670
2.64710
.573
-8.8959
3.5619
3.00
-2.9461
4.21764
.764
-12.8707
6.9784
3.00
1.00
.2791
4.29117
.998
-9.8185
10.3767
1.00
2.00
2.00
3.00
2.9461
-1.1641(*)
4.21764
.43087
.764
.020
-6.9784
-2.1780
12.8707
-.1502
-.2763
.69848
.917
-1.9199
1.3673
2.00
1.00
1.1641(*)
.43087
.020
.1502
2.1780
3.00
.8878
.68651
.400
-.7277
2.5032
1.00
2.00
.2763
.69848
.917
-1.3673
1.9199
-.8878
.68651
.400
-2.5032
.7277
2.00
2.00
3.00
dAt/At
Vt/At
3.00
At
1.00
-29609.6303(*)
11260.01861
.024
-56105.6645
-3113.5961
3.00
-12993.4564
18253.39590
.757
-55945.6591
29958.7463
1.00
3.00
1.00
29609.6303(*)
16616.1739
11260.01861
17940.62716
.024
.624
3113.5961
-25600.0504
56105.6645
58832.3982
12993.4564
18253.39590
.757
-29958.7463
55945.6591
2.00
-16616.1739
17940.62716
.624
-58832.3982
25600.0504
1.00
2.00
-1.4975(*)
.42403
.001
-2.4952
-.4997
-.5320
.68739
.719
-2.1495
1.0855
2.00
3.00
1.00
1.4975(*)
.42403
.001
.4997
2.4952
3.00
.9654
.67561
.327
-.6244
2.5552
.5320
.68739
.719
-1.0855
2.1495
-.9654
.67561
.327
-2.5552
.6244
2.00
3.00
Lt/At
3.00
1.00
2.00
* The mean difference is significant at the .05 level.
Source: Developed for this research
Table B12: ANOVA comparing payers, former payers and never paid firms (2002)
Dependent
(I) PAYER3
(J) PAYER3
Mean Difference (I-J)
Std. Error
Sig.
95% Confidence Interval
276
DBA Thesis: Malinee Ronapat
Variable
Et/At
1.00
2.00
3.00
.0690(*)
.02082
.003
Lower Bound
.0200
.0629
.02989
.090
-.0074
.1332
2.00
1.00
-.0690(*)
.02082
.003
-.1180
-.0200
3.00
-.0061
.03041
.978
-.0776
.0654
1.00
2.00
-.0629
.02989
.090
-.1332
.0074
.0061
.03041
.978
-.0654
.0776
2.00
3.00
Yt/BEt
1.00
2.00
3.00
dAt/At
Vt/At
.08329
.567
-.1114
.2806
.1898
.11960
.253
-.0916
.4711
1.00
3.00
1.00
-.0846
.1051
.08329
.12166
.567
.663
-.2806
-.1811
.1114
.3914
-.1898
.11960
.253
-.4711
.0916
2.00
-.1051
.12166
.663
-.3914
.1811
2.00
.2743
.12930
.087
-.0300
.5785
-.2454
.18565
.384
-.6822
.1914
2.00
3.00
1.00
-.2743
.12930
.087
-.5785
.0300
3.00
-.5197(*)
.18886
.017
-.9640
-.0753
3.00
1.00
.2454
.18565
.384
-.1914
.6822
1.00
2.00
2.00
3.00
.5197(*)
-.4561(*)
.18886
.17479
.017
.025
.0753
-.8673
.9640
-.0448
-.3668
.25096
.311
-.9573
.2237
2.00
1.00
.4561(*)
.17479
.025
.0448
.8673
3.00
.0893
.25530
.935
-.5114
.6899
1.00
2.00
.3668
.25096
.311
-.2237
.9573
-.0893
.25530
.935
-.6899
.5114
2.00
1.00
-33995.5603(*)
11694.41946
.011
-61510.9047
-6480.2158
3.00
-3048.9074
16790.96286
.982
-42555.7105
36457.8957
1.00
3.00
1.00
33995.5603(*)
30946.6529
11694.41946
17081.08486
.011
.167
6480.2158
-9242.7670
61510.9047
71136.0727
3048.9074
16790.96286
.982
-36457.8957
42555.7105
2.00
-30946.6529
17081.08486
.167
-71136.0727
9242.7670
1.00
2.00
-.8297(*)
.15579
.000
-1.1962
-.4631
-.3762
.22368
.214
-.9025
.1501
2.00
3.00
1.00
.8297(*)
.15579
.000
.4631
1.1962
3.00
.4535
.22754
.115
-.0819
.9889
.3762
.22368
.214
-.1501
.9025
-.4535
.22754
.115
-.9889
.0819
2.00
3.00
Lt/At
.0846
3.00
1.00
3.00
At
Upper Bound
.1180
3.00
1.00
2.00
* The mean difference is significant at the .05 level.
Source: Developed for this research
277
DBA Thesis: Malinee Ronapat
APPENDIX C: LOGIT REGRESSION: CONTROLLING TWO INVESTMENT
VARIABLES
Table C1: Logit Regression: controlling two investment variables from 1991 to 2002
Logit Regression 1991
278
DBA Thesis: Malinee Ronapat
Variables in the Equation
B
S.E.
Wald
df
Sig.
SET
1.69166191
0.7746301 4.76912164
1
0.0289746
PROFIT1
-3.0421817 3.00551323 1.02454964
1 0.31144225
INVEST1
0.30065078 0.25692757 1.36931467
1 0.24192967
INVEST2
-0.04984
0.0217351 5.25815479
1 0.02184417
Constant
1.52435137 0.50680503 9.04666144
1 0.00263175
a
Variable(s) entered on step 1: SET, PROFIT1, INVEST1, INVEST2.
Logit Regression 1992
Variables in the Equation
B
S.E.
Wald
df
Sig.
Step 1
SET
0.25025485 0.83675582 0.08944736
1
0.7648809
PROFIT
14.63854 4.07803407 12.8852791
1 0.00033118
INVEST1
-1.3425887
0.6029451 4.95827302
1 0.02596613
INVEST2
-0.5921966 0.22108188 7.17506076
1 0.00739239
Constant
2.61120472 0.54400168 23.0399406
1 1.5867E-06
a
Variable(s) entered on step 1: SET, PROFIT, INVEST1, INVEST2.
Logit Regression 1993
Variables in the Equation
B
S.E.
Wald
df
Sig.
Step 1
SET
-0.0227352 0.76644939
0.0008799
1 0.97633578
PROFIT
15.0560417 3.09832066 23.6139637
1 1.1773E-06
INVEST1
-2.8461384 0.55088734 26.6923268
1 2.3857E-07
INVEST2
-0.0725844 0.12960446 0.31365077
1 0.57544852
Constant
1.8909287 0.40828848 21.4494444
1 3.6328E-06
a
Variable(s) entered on step 1: SET, PROFIT, INVEST1, INVEST2.
Logit Regression 1994
Variables in the Equation
B
S.E.
Wald
df
Sig.
Step 1
SET
1.09242823 0.66000241 2.73964815
1 0.09788632
PROFIT
20.8225252 3.64845666 32.5722989
1 1.1484E-08
INVEST1
-1.7541047
0.5009255 12.2620971
1 0.00046225
INVEST2
-0.4561003 0.19003079 5.76066644
1 0.01638885
Constant
1.11172585 0.37305171 8.88090724
1 0.00288168
a
Variable(s) entered on step 1: SET, PROFIT, INVEST1, INVEST2.
Logit Regression 1995
Variables in the Equation
B
S.E.
Wald
df
Sig.
Step 1
SET
0.83479723 0.69185978 1.45588087
1 0.22758666
PROFIT
24.1569544 3.92324996 37.9133711
1 7.3957E-10
INVEST1
0.30063641 0.39517304 0.57877338
1 0.44679352
INVEST2
-0.6422795 0.26976171 5.66875274
1 0.01726973
Constant
0.66260573 0.38439776 2.97131488
1 0.08475294
a
Variable(s) entered on step 1: SET, PROFIT, INVEST1, INVEST2.
Logit Regression 1996
Variables in the Equation
B
S.E.
Wald
df
Sig.
Step 1
SET
1.36203933 0.58650508 5.39306611
1 0.02021692
PROFIT
15.6212727 2.87734407 29.4746892
1 5.6652E-08
INVEST1
-0.1821251 0.38740366 0.22101011
1 0.63827134
INVEST2
-0.3380858 0.21839834 2.39637727
1
0.1216166
Constant
0.48208197 0.30412863 2.51262204
1 0.11293788
a
Variable(s) entered on step 1: SET, PROFIT, INVEST1, INVEST2.
Logit Regression 1997
Variables in the Equation
Step 1
Exp(B)
5.428494873
0.047730643
1.350737555
0.951381603
4.592164004
Exp(B)
1.284352689
2277388.687
0.261168705
0.553110961
13.6154438
Exp(B)
0.977521272
3457449.26
0.058068126
0.92998728
6.62551893
Exp(B)
2.981505072
1104352741
0.173062112
0.633750283
3.039599764
Exp(B)
2.304346759
30990743365
1.350718145
0.526091831
1.939840452
Exp(B)
3.904147041
6084607.207
0.833497066
0.7131341
1.619442522
279
DBA Thesis: Malinee Ronapat
B
S.E.
Wald
df
Sig.
SET
2.01057411 0.48575242 17.1320849
1 3.4868E-05
PROFIT
1.78983085 0.49581226 13.0313512
1 0.00030632
INVEST1
-0.1917793 0.23938561 0.64181115
1 0.42305571
INVEST2
-0.0762184 0.28028488 0.07394693
1 0.78567442
Constant
0.29174874 0.28714911 1.03229308
1
0.3096207
a
Variable(s) entered on step 1: SET, PROFIT, INVEST1, INVEST2.
Logit Regression 1998
Variables in the Equation
B
S.E.
Wald
df
Sig.
Step 1
SET
2.68757859 0.52590526 26.1160395
1
3.215E-07
PROFIT
1.63713102 0.86163954 3.61007008
1
0.0574307
INVEST1
0.26203625 0.42856851 0.37383696
1 0.54092019
INVEST2
-0.2521342 0.26247227 0.92277658
1 0.33674699
Constant
-2.2458309 0.35074476 40.9988539
1 1.5232E-10
a
Variable(s) entered on step 1: SET, PROFIT, INVEST1, INVEST2.
Logit Regression 1999
Variables in the Equation
B
S.E.
Wald
df
Sig.
Step 1
SET
1.24191534 0.50070847 6.15196859
1 0.01312659
PROFIT
4.30982748 1.11488618 14.9437195
1 0.00011077
INVEST1
0.52716789
0.447829 1.38571356
1 0.23913031
INVEST2
-0.7253766 0.26091623 7.72902863
1
0.005434
Constant
-0.9079683 0.32811589 7.65749816
1 0.00565369
a
Variable(s) entered on step 1: SET, PROFIT, INVEST1, INVEST2.
Logit Regression 2000
Variables in the Equation
B
S.E.
Wald
df
Sig.
Step 1
SET
1.73732713 0.48178466 13.0034102
1 0.00031092
PROFIT
5.5765423 1.32574359
17.693385
1 2.5953E-05
INVEST1
0.75236727 0.34248438 4.82589923
1 0.02803524
INVEST2
-1.1656548
0.3536621 10.8633274
1 0.00098087
Constant
-0.6780377 0.35922275 3.56270578
1 0.05909134
a
Variable(s) entered on step 1: SET, PROFIT, INVEST1, INVEST2.
Logit Regression 2001
Variables in the Equation
B
S.E.
Wald
df
Sig.
Step 1
SET
1.72590489
0.4093318 17.7779934
1 2.4824E-05
PROFIT
-0.0555633 0.09701492
0.3280191
1 0.56682785
INVEST1
-0.0046421 0.02280901 0.04142055
1 0.83872831
INVEST2
-0.5444767 0.21879821 6.19257247
1 0.01282876
Constant
-0.656846 0.30524045 4.63066009
1 0.03140548
a
Variable(s) entered on step 1: SET, PROFIT, INVEST1, INVEST2.
Logit Regression 2002
Variables in the Equation
B
S.E.
Wald
df
Sig.
Step 1
SET
1.70082859 0.42250429 16.2053664
1 5.6833E-05
PROFIT
5.61125629 1.33189148 17.7493532
1
2.52E-05
INVEST1
-0.2860076 0.17600596
2.6405864
1 0.10416509
INVEST2
-0.7700316
0.2105792 13.3716877
1 0.00025545
Constant
-0.4395467 0.28131783 2.44126852
1 0.11818062
a
Variable(s) entered on step 1: SET, PROFIT, INVEST1, INVEST2.
Step 1
Exp(B)
7.467603327
5.988439457
0.825489055
0.926613824
1.338766591
Exp(B)
14.69604772
5.140400618
1.299573648
0.777140461
0.105839563
Exp(B)
3.462238484
74.42764751
1.694127546
0.484142192
0.403342844
Exp(B)
5.682135472
264.156651
2.122017457
0.311718487
0.507612127
Exp(B)
5.617602064
0.945952137
0.99536866
0.580145289
0.518484077
Exp(B)
5.47848493
273.4876026
0.751256907
0.462998415
0.644328417
Source: Developed for this research
Note: PROFIT represents Et/At ratio. INVEST1 represents dAt/At ratio and INVEST2 represents Vt/At.
280
DBA Thesis: Malinee Ronapat
APPENDIX D:
LOGIT REGRESSION: CONTROLLING ONE
INVESTMENT VARIABLE
Table D1: Logit Regression: controlling one investment variable
Logit regression 1991
281
DBA Thesis: Malinee Ronapat
Variables in the Equation
B
S.E.
Wald
SET
0.909227133
0.67056316 1.838507451
PROFIT
-2.903282463 2.927846101 0.983291063
INVEST1
0.341714711
0.24907159 1.882257189
Constant
1.617157642 0.501465958 10.39972373
a
Variable(s) entered on step 1: SET, PROFIT, INVEST1.
Logit regression 1992
Variables in the Equation
B
S.E.
Wald
Step 1
SET
0.051479911 0.808487973 0.004054417
PROFIT
11.43072322 3.710932953 9.488142579
INVEST1
-1.510840948 0.584779098 6.675039064
Constant
1.893909231 0.507278446 13.93880371
a
Variable(s) entered on step 1: SET, PROFIT, INVEST1.
Logit regression 1993
Variables in the Equation
B
S.E.
Wald
Step 1
SET
-0.195398328
0.69583429 0.078855145
PROFIT
15.01121885 3.101116518 23.43126549
INVEST1
-2.926270038 0.534767723
29.9432149
Constant
1.846103657
0.40176739 21.11362474
a
Variable(s) entered on step 1: SET, PROFIT, INVEST1.
Logit regression 1994
Variables in the Equation
B
S.E.
Wald
Step 1
SET
0.421364264 0.580402523 0.527056068
PROFIT
19.90443372 3.583042448 30.85998753
INVEST1
-2.04817657
0.49123395
17.3843333
Constant
0.828505261 0.357005393 5.385685487
a
Variable(s) entered on step 1: SET, PROFIT, INVEST1.
Logit regression 1995
Variables in the Equation
B
S.E.
Wald
Step 1
SET
0.113335512 0.592852892 0.036545856
PROFIT
22.39143171 3.755078097 35.55705463
INVEST1
0.315013545 0.402183652 0.613493023
Constant
0.252276658 0.355153006
0.504572
a
Variable(s) entered on step 1: SET, PROFIT, INVEST1.
Logit regression 1996
Variables in the Equation
B
S.E.
Wald
Step 1
SET
1.012963284 0.528232751 3.677365182
PROFIT
15.37261702 2.856105429 28.96991001
INVEST1
-0.207842807
0.39772938 0.273082991
Constant
0.283130654 0.278999929 1.029830142
a
Variable(s) entered on step 1: SET, PROFIT, INVEST1.
Logit regression 1997
Variables in the Equation
B
S.E.
Wald
Step 1
SET
1.956362569 0.441452572 19.63950727
PROFIT
1.834539726 0.471427319 15.14344723
INVEST1
-0.189181745 0.238378865 0.629829439
Constant
0.244061097 0.228288642 1.142953358
a
Variable(s) entered on step 1: SET, PROFIT, INVEST1.
Step 1
df
1
1
1
1
Sig.
0.17512598
0.32138764
0.17007783
0.00126034
Exp(B)
2.482403223
0.054842904
1.407358736
5.038748017
df
1
1
1
1
Sig.
0.94922956
0.00206804
0.00977724
0.00018886
Exp(B)
1.052828036
92108.56532
0.220724282
6.645295967
df
1
1
1
1
Sig.
0.77885492
1.2946E-06
4.4489E-08
4.3284E-06
Exp(B)
0.822506966
3305898.486
0.053596579
6.335087699
df
1
1
1
1
Sig.
0.46784719
2.7733E-08
3.0533E-05
0.02030261
Exp(B)
1.52403933
440946326.1
0.128969857
2.289893388
df
1
1
1
1
Sig.
0.84839264
2.4769E-09
0.43347565
0.47749808
Exp(B)
1.120007646
5302433516
1.370277871
1.286952033
df
1
1
1
1
Sig.
0.05515595
7.3511E-08
0.60127113
0.31019857
Exp(B)
2.753749077
4745071.331
0.81233472
1.327278565
df
1
1
1
1
Sig.
9.3515E-06
9.9644E-05
0.42741788
0.28502913
Exp(B)
7.073550661
6.262251128
0.827636075
1.276422314
282
DBA Thesis: Malinee Ronapat
Logit regression 1998
Variables in the Equation
B
S.E.
Wald
SET
2.490787506 0.481848774 26.72094666
PROFIT
1.567042419 0.717224534 4.773657229
INVEST1
0.438409329 0.211117835 4.312307558
Constant
-2.365617422 0.308193704 58.91711793
a
Variable(s) entered on step 1: SET, PROFIT, INVEST1.
Logit regression 1999
Variables in the Equation
B
S.E.
Wald
Step 1
SET
0.750371887 0.453141114
2.74211821
PROFIT
3.974769365 1.008424878 15.53591247
INVEST1
0.518463167 0.429289108 1.458599559
Constant
-1.367565114
0.2777623 24.24093846
a
Variable(s) entered on step 1: SET, PROFIT, INVEST1.
Logit regression 2000
Variables in the Equation
B
S.E.
Wald
Step 1
SET
1.324993571 0.439054176 9.107335127
PROFIT
5.060612144 1.187881476 18.14930031
INVEST1
0.80011661 0.334073122 5.736189623
Constant
-1.487770427 0.270983376 30.14300498
a
Variable(s) entered on step 1: SET, PROFIT, INVEST1.
Logit regression 2001
Variables in the Equation
B
S.E.
Wald
Step 1
SET
1.761378872 0.389166374 20.48495192
PROFIT
-0.069944305 0.089979679 0.604248864
INVEST1
0.009309897 0.016001202 0.338520153
Constant
-1.271783961 0.230614679 30.41254689
a
Variable(s) entered on step 1: SET, PROFIT, INVEST1.
Logit regression 2002
Variables in the Equation
B
S.E.
Wald
Step 1
SET
1.51588267 0.384204861 15.56701939
PROFIT
2.202351906
0.85389976
6.65211343
INVEST1
0.082779556 0.102829378 0.648054844
Constant
-1.045317985 0.224369269 21.70551518
a
Variable(s) entered on step 1: SET, PROFIT, INVEST1.
Step 1
df
1
1
1
1
Sig.
2.3506E-07
0.02889837
0.0378376
1.6446E-14
Exp(B)
12.07077819
4.792453142
1.550239335
0.093891312
df
1
1
1
1
Sig.
0.09773514
8.0953E-05
0.22715308
8.5006E-07
Exp(B)
2.117787449
53.23783699
1.67944464
0.254726435
df
1
1
1
1
Sig.
0.00254586
2.0424E-05
0.01661881
4.0133E-08
Exp(B)
3.762161169
157.6870139
2.225800463
0.225875701
df
1
1
1
1
Sig.
6.0102E-06
0.43696147
0.56068467
3.4926E-08
Exp(B)
5.820457532
0.932445751
1.009353369
0.280331076
df
1
1
1
1
Sig.
7.9632E-05
0.00990383
0.42080898
3.1788E-06
Exp(B)
4.553438537
9.046264467
1.086302313
0.351580004
Source: Developed for this research
Note: PROFIT represents Et/At ratio. INVEST1 represents dAt/At ratio.
283
DBA Thesis: Malinee Ronapat
APPENDIX E:
T-STATISTICS (TWO INVESTMENT VARIABLES)
Table E1: t-statistics for 1991-1996, 1997-2002, 1991-2002 (two investment
variables)
284
DBA Thesis: Malinee Ronapat
t-stat results 1991-1996
One-Sample Test
Test Value = 0
CONSTANT
SET
PROFIT
INVEST1
INVEST2
t
df
4.2243059
3.2457428
3.7868216
-3.4593826
-1.7689473
5
5
5
5
5
Sig. (2-tailed)
0.008293448
0.022803365
0.01280004
0.018054303
0.137133446
Mean Difference
1.378333333
0.866666667
14.54333333
-0.358333333
-0.92
95% Confidence Interval of the
Difference
Lower
Upper
0.539587586
2.21707908
0.180279063
1.55305427
4.670983385
24.41568328
-0.624601993
-0.092064674
-2.256916753
0.416916753
t-stat results 1997-2002
One-Sample Test
Test Value = 0
CONSTANT
SET
PROFIT
INVEST1
INVEST2
t
df
-2.2812049
9.4480851
3.2708552
-3.6908656
1.053322
5
5
5
5
5
Sig. (2-tailed)
0.071430364
0.000224296
0.022177572
0.014132908
0.340410564
Mean Difference
-0.775
1.851666667
3.145
-0.59
0.176666667
95% Confidence Interval of the
Difference
Lower
Upper
-1.648310806
0.098310806
1.347875617
2.355457716
0.673328419
5.616671581
-1.000918046
-0.179081954
-0.254479846
0.607813179
t-stat results 1991-2002
One-Sample Test
Test Value = 0
CONSTANT
SET
PROFIT
INVEST1
INVEST2
t
df
0.7642389
6.2704339
3.4649563
-4.873616
-1.2045706
11
11
11
11
11
Sig. (2-tailed)
0.46080455
6.08364E-05
0.00528676
0.000491864
0.253646707
Mean Difference
0.301666667
1.359166667
8.844166667
-0.474166667
-0.371666667
95% Confidence Interval of the
Difference
Lower
Upper
-0.567124248
1.170457582
0.882085539
1.836247794
3.226237404
14.46209593
-0.68830619
-0.260027144
-1.050774076
0.307440742
Source: Developed for this research
Note: PROFIT represents Et/At ratio. INVEST1 represents dAt/At ratio and INVEST2 represents Vt/At.
285
DBA Thesis: Malinee Ronapat
APPENDIX F:
T-STATISTICS (ONE INVESTMENT VARIABLE)
Table F1: t-statistics for 1991-1996, 1997-2002, 1991-2002 (one investment variable)
t-statistic results 1991-1996
One-Sample Test
286
DBA Thesis: Malinee Ronapat
Test Value = 0
CONSTANT
SET
INVEST2
PROFIT
t
df
Sig. (2-tailed)
3.588465874
1.918639616
-1.812972014
3.710863396
5
5
5
5
0.015733521
0.113125827
0.129575803
0.01384251
t
df
Sig. (2-tailed)
-3.526400773
6.766639201
1.827821418
3.2533474
5
5
5
5
0.016804068
0.001071098
0.127123447
0.022611779
t
df
Sig. (2-tailed)
-0.118025689
4.193188702
-1.087213535
3.27070573
11
11
11
11
0.908174919
0.001502741
0.30019141
0.007455828
Mean Difference
1.12
0.383333333
-1.006666667
13.53333333
95% Confidence Interval of the
Difference
Lower
Upper
0.317692898
1.922307102
-0.130254365
0.896921031
-2.434002019
0.420668686
4.158549858
22.90811681
t-statistic results 1997-2002
One-Sample Test
Test Value = 0
CONSTANT
SET
INVEST2
PROFIT
Mean Difference
-1.218333333
1.633333333
0.276666667
2.426666667
95% Confidence Interval of the
Difference
Lower
Upper
-2.10644156
-0.330225107
1.012845544
2.253821123
-0.112427312
0.665760645
0.509273752
4.344059582
t-statistic results 1991-2002
One-Sample Test
Test Value = 0
CONSTANT
SET
INVEST2
PROFIT
Mean Difference
-0.049166667
1.008333333
-0.365
7.98
95% Confidence Interval of the
Difference
Lower
Upper
-0.966044206
0.867710872
0.479063877
1.537602789
-1.103916098
0.373916098
2.609947471
13.35005253
Source: Developed for this research
Note: PROFIT represents Et/At ratio. INVEST2 represents Vt/At.
287
DBA Thesis: Malinee Ronapat
APPENDIX G:
EXPECTED PERCENTAGE OF PAYERS: CONTROLLING
TWO INVESTMENT VARIABLES (CROSSTABS)
Table G1: Expected percentage of payers: controlling two investment variables
(crosstabs) 1991
CODE2
Total
288
DBA Thesis: Malinee Ronapat
.00
PAYER
.00
1.00
Total
1.00
Expected Count
3.6
28.4
32.0
% within PAYER
15.6%
84.4%
100.0%
% within CODE2
20.0%
13.6%
14.3%
Expected Count
21.4
170.6
192.0
% within PAYER
10.4%
89.6%
100.0%
% within CODE2
80.0%
86.4%
85.7%
Expected Count
25.0
199.0
224.0
% within PAYER
11.2%
88.8%
100.0%
% within CODE2
100.0%
100.0%
100.0%
Source: Developed for this research
Table G2: Expected percentage of payers: controlling two investment variables
(crosstabs) 1992
CODE2
.00
PAYER
.00
1.00
Total
Total
1.00
Expected Count
.3
23.7
24.0
% within PAYER
8.3%
91.7%
100.0%
% within CODE2
66.7%
8.8%
9.4%
Expected Count
2.7
227.3
230.0
% within PAYER
.4%
99.6%
100.0%
% within CODE2
33.3%
91.2%
90.6%
Expected Count
3.0
251.0
254.0
% within PAYER
1.2%
98.8%
100.0%
% within CODE2
100.0%
100.0%
100.0%
Source: Developed for this research
Table G3: Expected percentage of payers: controlling two investment variables
(crosstabs) 1993
CODE2
.00
PAYER
.00
1.00
Total
Expected Count
Total
1.00
4.1
39.9
44.0
% within PAYER
20.5%
79.5%
100.0%
% within CODE2
32.1%
12.8%
14.6%
Expected Count
23.9
233.1
257.0
% within PAYER
7.4%
92.6%
100.0%
% within CODE2
67.9%
87.2%
85.4%
Expected Count
28.0
273.0
301.0
% within PAYER
9.3%
90.7%
100.0%
% within CODE2
100.0%
100.0%
100.0%
Source: Developed for this research
Table G4: Expected percentage of payers: controlling two investment variables
(crosstabs) 1994
CODE2
Total
289
DBA Thesis: Malinee Ronapat
.00
PAYER
.00
1.00
Total
Expected Count
1.00
1.3
51.7
53.0
% within PAYER
11.3%
88.7%
100.0%
% within CODE2
75.0%
14.6%
16.1%
Expected Count
6.7
269.3
276.0
% within PAYER
.7%
99.3%
100.0%
% within CODE2
25.0%
85.4%
83.9%
Expected Count
8.0
321.0
329.0
% within PAYER
2.4%
97.6%
100.0%
% within CODE2
100.0%
100.0%
100.0%
Source: Developed for this research
Table G5: Expected percentage of payers: controlling two investment variables
(crosstabs) 1995
CODE2
.00
PAYER
.00
1.00
Total
Expected Count
Total
1.00
.9
53.1
54.0
% within PAYER
11.1%
88.9%
100.0%
% within CODE2
100.0%
14.2%
15.7%
Expected Count
5.1
284.9
290.0
% within PAYER
.0%
100.0%
100.0%
% within CODE2
.0%
85.8%
84.3%
Expected Count
6.0
338.0
344.0
% within PAYER
1.7%
98.3%
100.0%
% within CODE2
100.0%
100.0%
100.0%
Source: Developed for this research
Table G6: Expected percentage of payers: controlling two investment variables
(crosstabs) 1996
CODE2
.00
PAYER
.00
1.00
Total
Expected Count
Total
1.00
1.5
70.5
72.0
% within PAYER
9.7%
90.3%
100.0%
% within CODE2
87.5%
17.6%
19.0%
Expected Count
6.5
299.5
306.0
% within PAYER
.3%
99.7%
100.0%
% within CODE2
12.5%
82.4%
81.0%
Expected Count
8.0
370.0
378.0
% within PAYER
2.1%
97.9%
100.0%
% within CODE2
100.0%
100.0%
100.0%
Source: Developed for this research
Table G7: Expected percentage of payers: controlling two investment variables
(crosstabs) 1997
CODE2
Total
290
DBA Thesis: Malinee Ronapat
.00
PAYER
.00
1.00
Total
Expected Count
1.00
42.2
69.8
112.0
% within PAYER
58.0%
42.0%
100.0%
% within CODE2
43.9%
19.2%
28.5%
Expected Count
105.8
175.2
281.0
% within PAYER
29.5%
70.5%
100.0%
% within CODE2
56.1%
80.8%
71.5%
Expected Count
148.0
245.0
393.0
% within PAYER
37.7%
62.3%
100.0%
% within CODE2
100.0%
100.0%
100.0%
Source: Developed for this research
Table G8: Expected percentage of payers: controlling two investment variables
(crosstabs) 1998
CODE2
.00
PAYER
.00
1.00
Total
Expected Count
Total
1.00
38.0
240.0
278.0
% within PAYER
18.3%
81.7%
100.0%
% within CODE2
96.2%
67.8%
71.6%
Expected Count
15.0
95.0
110.0
% within PAYER
1.8%
98.2%
100.0%
% within CODE2
3.8%
32.2%
28.4%
Expected Count
53.0
335.0
388.0
% within PAYER
13.7%
86.3%
100.0%
% within CODE2
100.0%
100.0%
100.0%
Source: Developed for this research
Table G9: Expected percentage of payers: controlling two investment variables
(crosstabs) 1999
CODE2
.00
PAYER
.00
1.00
Total
Expected Count
Total
1.00
51.2
218.8
270.0
% within PAYER
26.3%
73.7%
100.0%
% within CODE2
97.3%
63.8%
70.1%
Expected Count
21.8
93.2
115.0
% within PAYER
1.7%
98.3%
100.0%
% within CODE2
2.7%
36.2%
29.9%
Expected Count
73.0
312.0
385.0
% within PAYER
19.0%
81.0%
100.0%
% within CODE2
100.0%
100.0%
100.0%
Source: Developed for this research
Table G10: Expected percentage of payers: controlling two investment variables
(crosstabs) 2000
CODE2
Total
291
DBA Thesis: Malinee Ronapat
.00
PAYER
.00
1.00
Total
Expected Count
1.00
34.9
203.1
238.0
% within PAYER
22.3%
77.7%
100.0%
% within CODE2
98.1%
58.9%
64.7%
Expected Count
19.1
110.9
130.0
% within PAYER
.8%
99.2%
100.0%
% within CODE2
1.9%
41.1%
35.3%
Expected Count
54.0
314.0
368.0
% within PAYER
14.7%
85.3%
100.0%
% within CODE2
100.0%
100.0%
100.0%
Source: Developed for this research
Table G11: Expected percentage of payers: controlling two investment variables
(crosstabs) 2001
CODE2
.00
PAYER
.00
1.00
Total
Expected Count
Total
1.00
24.9
198.1
223.0
% within PAYER
18.4%
81.6%
100.0%
% within CODE2
97.6%
54.5%
59.3%
Expected Count
17.1
135.9
153.0
% within PAYER
.7%
99.3%
100.0%
% within CODE2
2.4%
45.5%
40.7%
Expected Count
42.0
334.0
376.0
% within PAYER
11.2%
88.8%
100.0%
% within CODE2
100.0%
100.0%
100.0%
Source: Developed for this research
Table G12: Expected percentage of payers: controlling two investment variables
(crosstabs) 2002
CODE2
.00
PAYER
.00
1.00
Total
Expected Count
Total
1.00
17.2
190.8
208.0
% within PAYER
15.4%
84.6%
100.0%
% within CODE2
100.0%
49.6%
53.7%
Expected Count
14.8
164.2
179.0
% within PAYER
.0%
100.0%
100.0%
% within CODE2
.0%
50.4%
46.3%
Expected Count
32.0
355.0
387.0
% within PAYER
8.3%
91.7%
100.0%
% within CODE2
100.0%
100.0%
100.0%
Source: Developed for this research
292
DBA Thesis: Malinee Ronapat
APPENDIX H:
EXPECTED PERCENTAGE OF PAYERS: CONTROLLING
ONE INVESTMENT VARIABLE (CROSSTABS)
Table H1: Expected percentage of payers: controlling one investment variable
(crosstab) 1991
CODE4
Total
293
DBA Thesis: Malinee Ronapat
.00
PAYER
.00
1.00
Total
Expected Count
1.00
.1
31.9
32.0
% within PAYER
.0%
100.0%
100.0%
% within CODE4
.0%
14.3%
14.3%
Expected Count
.9
191.1
192.0
% within PAYER
.5%
99.5%
100.0%
% within CODE4
100.0%
85.7%
85.7%
Expected Count
1.0
223.0
224.0
% within PAYER
.4%
99.6%
100.0%
% within CODE4
100.0%
100.0%
100.0%
Source: Developed for this research
Table H2: Expected percentage of payers: controlling one investment variable
(crosstab) 1992
CODE4
.00
PAYER
.00
1.00
Total
Expected Count
Total
1.00
.2
23.8
24.0
% within PAYER
8.3%
91.7%
100.0%
% within CODE4
100.0%
8.7%
9.4%
Expected Count
1.8
228.2
230.0
% within PAYER
.0%
100.0%
100.0%
% within CODE4
.0%
91.3%
90.6%
Expected Count
2.0
252.0
254.0
% within PAYER
.8%
99.2%
100.0%
% within CODE4
100.0%
100.0%
100.0%
Source: Developed for this research
Table H3: Expected percentage of payers: controlling one investment variable
(crosstab) 1993
CODE4
.00
PAYER
.00
1.00
Total
Expected Count
Total
1.00
3.1
40.9
44.0
% within PAYER
11.4%
88.6%
100.0%
% within CODE4
23.8%
13.9%
14.6%
Expected Count
17.9
239.1
257.0
% within PAYER
6.2%
93.8%
100.0%
% within CODE4
76.2%
86.1%
85.4%
Expected Count
21.0
280.0
301.0
% within PAYER
7.0%
93.0%
100.0%
% within CODE4
100.0%
100.0%
100.0%
Source: Developed for this research
Table H4: Expected percentage of payers: controlling one investment variable
(crosstab) 1994
CODE4
Total
294
DBA Thesis: Malinee Ronapat
.00
PAYER
.00
1.00
Total
Expected Count
1.00
.8
52.2
53.0
% within PAYER
5.7%
94.3%
100.0%
% within CODE4
60.0%
15.4%
16.1%
Expected Count
4.2
271.8
276.0
% within PAYER
.7%
99.3%
100.0%
% within CODE4
40.0%
84.6%
83.9%
Expected Count
5.0
324.0
329.0
% within PAYER
1.5%
98.5%
100.0%
% within CODE4
100.0%
100.0%
100.0%
Source: Developed for this research
Table H5: Expected percentage of payers: controlling one investment variable
(crosstab) 1995
CODE4
.00
PAYER
.00
1.00
Total
Expected Count
Total
1.00
.8
53.2
54.0
% within PAYER
9.3%
90.7%
100.0%
% within CODE4
100.0%
14.5%
15.7%
Expected Count
4.2
285.8
290.0
% within PAYER
.0%
100.0%
100.0%
% within CODE4
.0%
85.5%
84.3%
Expected Count
5.0
339.0
344.0
% within PAYER
1.5%
98.5%
100.0%
% within CODE4
100.0%
100.0%
100.0%
Source: Developed for this research
Table H6: Expected percentage of payers: controlling one investment variable
(crosstab) 1996
CODE4
.00
PAYER
.00
1.00
Total
Expected Count
Total
1.00
1.7
70.3
72.0
% within PAYER
9.7%
90.3%
100.0%
% within CODE4
77.8%
17.6%
19.0%
Expected Count
7.3
298.7
306.0
% within PAYER
.7%
99.3%
100.0%
% within CODE4
22.2%
82.4%
81.0%
Expected Count
9.0
369.0
378.0
% within PAYER
2.4%
97.6%
100.0%
% within CODE4
100.0%
100.0%
100.0%
Source: Developed for this research
Table H7: Expected percentage of payers: controlling one investment variable
(crosstab) 1997
CODE4
Total
295
DBA Thesis: Malinee Ronapat
.00
PAYER
.00
1.00
Total
Expected Count
1.00
42.7
69.3
112.0
% within PAYER
58.9%
41.1%
100.0%
% within CODE4
44.0%
18.9%
28.5%
Expected Count
107.3
173.7
281.0
% within PAYER
29.9%
70.1%
100.0%
% within CODE4
56.0%
81.1%
71.5%
Expected Count
150.0
243.0
393.0
% within PAYER
38.2%
61.8%
100.0%
% within CODE4
100.0%
100.0%
100.0%
Source: Developed for this research
Table H8: Expected percentage of payers: controlling one investment variable
(crosstab) 1998
CODE4
.00
PAYER
.00
1.00
Total
Expected Count
Total
1.00
36.5
241.5
278.0
% within PAYER
16.9%
83.1%
100.0%
% within CODE4
92.2%
68.5%
71.6%
Expected Count
14.5
95.5
110.0
% within PAYER
3.6%
96.4%
100.0%
% within CODE4
7.8%
31.5%
28.4%
Expected Count
51.0
337.0
388.0
% within PAYER
13.1%
86.9%
100.0%
% within CODE4
100.0%
100.0%
100.0%
Source: Developed for this research
Table H9: Expected percentage of payers: controlling one investment variable
(crosstab) 1999
CODE4
.00
PAYER
.00
1.00
Total
Expected Count
Total
1.00
51.9
218.1
270.0
% within PAYER
26.7%
73.3%
100.0%
% within CODE4
97.3%
63.7%
70.1%
Expected Count
22.1
92.9
115.0
% within PAYER
1.7%
98.3%
100.0%
% within CODE4
2.7%
36.3%
29.9%
Expected Count
74.0
311.0
385.0
% within PAYER
19.2%
80.8%
100.0%
% within CODE4
100.0%
100.0%
100.0%
Source: Developed for this research
Table H10: Expected percentage of payers: controlling one investment variable
(crosstab) 2000
CODE4
Total
296
DBA Thesis: Malinee Ronapat
.00
PAYER
.00
1.00
Total
1.00
% within PAYER
20.6%
79.4%
100.0%
% within CODE4
98.0%
59.4%
64.7%
% within PAYER
100.0%
.8%
99.2%
% within CODE4
2.0%
40.6%
35.3%
% within PAYER
13.6%
86.4%
100.0%
% within CODE4
100.0%
100.0%
100.0%
Source: Developed for this research
Table H11: Expected percentage of payers: controlling one investment variable
(crosstab) 2001
CODE4
.00
PAYER
.00
1.00
Total
Expected Count
Total
1.00
24.9
198.1
223.0
% within PAYER
18.4%
81.6%
100.0%
% within CODE4
97.6%
54.5%
59.3%
Expected Count
17.1
135.9
153.0
% within PAYER
.7%
99.3%
100.0%
% within CODE4
2.4%
45.5%
40.7%
Expected Count
42.0
334.0
376.0
% within PAYER
11.2%
88.8%
100.0%
% within CODE4
100.0%
100.0%
100.0%
Source: Developed for this research
Table H12: Expected percentage of payers: controlling one investment variable
(crosstab) 2002
CODE4
.00
PAYER
.00
1.00
Total
Expected Count
Total
1.00
14.5
193.5
208.0
% within PAYER
13.0%
87.0%
100.0%
% within CODE4
100.0%
50.3%
53.7%
Expected Count
12.5
166.5
179.0
% within PAYER
.0%
100.0%
100.0%
% within CODE4
.0%
49.7%
46.3%
Expected Count
27.0
360.0
387.0
% within PAYER
7.0%
93.0%
100.0%
% within CODE4
100.0%
100.0%
100.0%
Source: Developed for this research
297
DBA Thesis: Malinee Ronapat
298