1

2

INTRODUCTION

subsequent years, Kato et al. (2002) investigating the Japanese stock market,
Choi et al. (2011) researching on the Korean stock market also found a positive
relationship between dividend changes and changes in profits one year after the
dividend change, Chinpiao Liu and An-Sing Chen (2015) studying the Chinese
stock market showed that dividend changes is a sign of prospects for return on
equity.
On the other hand, many studies have found little or no correlation
between dividend changes and profitability of enterprises. Typical examples
include: DeAngelo (1996), Benartzi et al. (1997), Grullon et al. (2005), Lie
(2005) did not find a significant relationship between dividend changes and
changes in earnings, Jensen et al. (2010) proved that the profits of enterprises
will bounce back after reducing dividends.
In summary, there have been many different conclusions when predicting
the enterprises’ profitability based on information about dividend changes.
Moreover, these studies focused on changing of dividends in cash without
considering the possibility of an enterprise buying back shares during the year
of paying dividends in cash, because repurchasing shares is a form of profit
distribution among shareholders. In addition, many researchers have
demonstrated that ownership structure affects the dividend payment of listed
companies. Therefore, the correlation between dividend changes and
profitability may vary according to ownership structure. However, so far there
have been no studies clarifying dividend signaling in the form of ownership and
in the case of enterprises repurchasing shares in the market.
In Vietnam, many studies on dividends, but so far there has been no
research on the effect of dividend changes on the forecast of profitability of the

has becoming a topic attracting attention of many subjects. Empirical results
provide different conclusions, divided into 2 basic groups: The first group
supports the viewpoint that dividend changes reveal the future profitability of
enterprises; the second group does not support this view and believes that the
increases or decreases of dividend are not linked with changes in future
earnings. Studies in the first group include: Healy and Palepu (1988), Nissim
and Ziv (2001) implemented in the US Stock Exchange showing that dividend
changes are positively correlated with changes in profitability in two


3

4

changes and profitability of the next 2 years of listed companies on Vietnam
stock market, thereby finding empirical evidence on dividend signaling on
Vietnam Stock Market. Specific objectives include:
- Evaluating current status of dividends, dividend changes and profitability
of listed companies on Vietnam stock market in the period 2008-2017.
- Examining one-way relationship between dividend changes and
profitability in 2 subsequent years of listed companies on Vietnam stock
market.
- Examining one-way relationship between dividend changes and
profitability in 2 subsequent years of state-owned companies and privately
owned companies.
- Examining one-way relationship between dividend changes and
profitability in 2 subsequent years of companies repurchasing stocks in the
market during the dividend change year
Based on the above objectives, the research develops the following research
questions:

Whereby, the thesis has established and tested the hypothesis of dividend
signaling on the frontier stock market in Vietnam. The empirical evidence
obtained from the study is a significant source of information in terms of
dividend signaling theory that other developing countries can refer to.
(i) Is there a relationship between dividend changes and future profitability
of firms?
(ii) Are there any differences in the relationship between dividend changes
and profitability of state-owned companies and privately owned companies?
(iii) Does repurchasing shares in the market in the dividend change year
reduce the content of information from dividend signaling?
Moreover, the thesis has added 3 control variables into the model to test
the above relationship which are all statistically significant. The 3 variables are:
Enterprise size, investment opportunities and growth rate. Finally, the thesis
also contributed to providing evidence to support the signaling theory of
dividend in emerging markets like Vietnam.
Practical contributions
The research sample includes: 657 observations of dividend increases, 893
observations of dividend decreases and 643 observations of no-change


5

6

dividends of non-financial companies listed on Ho Chi Minh Stock Exchange
and Hanoi Stock Exchange in the period 2008-2017. The thesis has provided
convincing evidence of the causal relationship between dividend changes and
profitability in 2 subsequent years in Vietnam stock market – which has not
been proved by any prior research. In general, on Vietnam stock market,
dividend increases imply that the profits in one following year of the company

after taxes for existing shareholders, form and dividend payout ratio.
1.1.1.2. Forms of dividend payment of joint-stock companies
Dividends can be paid in cash, by shares of the company or by other
assets as stipulated in the company's Charter.
1.1.1.3. Dividend payment procedures
1.1.1.4. Models of paying dividends of joint-stock companies
1.1.1.5. Dividend ratios
Dividend per Share (DPS)
DPS = Total dividends paid/Number of shares
Dividend payout ratio
Dividend payout ratio = Dividends per share/ Earnings per share
Dividend Yield
Dividend yield = Dividends per share / Market value per share
1.1.2. Definition of profitability and profitability ratios
1.1.2.1. Definition of profitability
Profitability reflects the amount of profits that an enterprise receives per
unit of cost, per unit of input or per unit of revenue.
1.1.2.2. Profitability ratios
- Return on Sales (ROS)
- Return on Assets (ROA)
- Return on Equity (ROE)
1.1.3. Theories on the relationship between dividend changes and profitability
of enterprises
1.1.3.1. Information asymmetry and signaling theory
According to Kyle (1985), information asymmetry on the stock market
occurs when one or more investors have their own information. Kim and
Verrecchia (1994) provided that information asymmetry occurs when a group
of investors has more public information about a company. Signaling theory is



Watt (1973)
Weak correlation is found
Aharony and - Positive correlation is found with increases in
Swary (1980)
dividends
- Negative correlation is found with decreases in
dividends
Healy
and Earnings increase (decrease) at least one year after the
Palepu (1988)
first dividend payment (not paying dividend)
DeAngelo et - If profits and dividend decrease, profits in 2 subsequent

Researchers
al. (1992)

Benartzi et al.
(1997)

Fukada (2000)

Nissim and Ziv
(2001)
Grullon et al.
(2005)

Harada
and
Nguyen (2005)
Lie (2005)

(2011)
and changes in profits one year after
- No correlation is found with dividend increases
- Non-linear model does not support signaling theory
Reza et al. - Positive correlation is found between dividend


9

Researchers
(2014)

10

Results

increases and profitability in 2 subsequent years
- Positive correlation is found between dividend
decreases and profitability in one subsequent year and
negative correlation is found between dividend decreases
and profitability in 2 subsequent years
Fairchild et al. - Negative correlation is found between dividend
(2014)
decreases and changes in profits in one following year
- Negative correlation is found between dividend
decreases and changes in ROA in 2 subsequent years
Liu and Chen - No correlation is found between dividend changes and
(2015)
changes in ROE
- Negative correlation is found between dividend

2.2. Current situation of dividends of listed enterprises on Vietnam stock
market
2.2.1. Dividend payment
Forms of dividend payment
The statistics of companies paying dividends on Vietnam stock market in the
period from 2008 to 2016 showed that paying dividends in cash is the most
comment form of payment compared to stock dividends
The number of companies paying dividends in cash by ownership
structures
In 2008, the proportion of state-owned enterprises paying cash dividends
accounted for 81.58%, while privately owned enterprises paying cash dividends
only accounted for 18.42%. The proportion of cash dividends of state-owned
enterprises tends to decrease, whereas the proportion of cash dividends paid by
privately owned enterprises tends to increase rapidly.
The number of companies paying dividends in cash by sectors
The industrials has the largest number of companies paying cash
dividends, followed by consumer goods and raw materials, especially, oil and
gas and banking sectors have very small number of companies paying cash
dividends.
2.2.2. Dividend rates
The annual dividend rates of enterprises listed on Vietnam stock market
in the period of 2008-2016 are mainly from 1,000-2,000 VND/share. From


11

12

2013 to 2016 the proportion of companies paying dividends below 1,000
VND/share increased compared to previous years. The average dividend rates

Firstly, profits and profitability of Vietnamese enterprises in the period
2008-2017 fluctuated considerably.
Secondly, the average profit of state-owned enterprises was lower than
that of privately owned enterprises.
Thirdly, the profitability of Vietnamese enterprises in the period 20082017 reached the highest value in 2009 and began to decline since 2010, and
reached the lowest values in 2012 and 2013.
2.4. Actual situation of dividend changes and changes in profitability of
listed joint-stock companies on Vietnam stock market
2.4.1. Actual situation of dividend changes of listed joint-stock companies on
Vietnam stock market
In the period 2008 - 2016, out of 2700 observations: 844 cases
experiencing dividend increases (accounted for 31.26%), 1133 cases
experiencing dividend decreases (accounted for 41.96%) and 723 cases
experiencing no-change dividends (accounted for 26.78%). The results show
that the proportion of enterprises listed on the stock market in Vietnam
experiencing annual dividend changes (increases or decreases) is relatively
high, accounting for 73.22%.
2.4.2. Actual situation of changes in profits and profitability of listed jointstock companies on Vietnam stock market
Enterprises with dividend increases or no-change dividends often have
increases in profits in the dividend change year and 2 subsequent years, but
with enterprises reducing dividends, profits decrease in the year of decreasing
dividends and one year after. However, ROA and ROE of companies that
increase dividends usually only increase in year 0, but tend to decrease in 2
subsequent years. In contrast, with firms reducing dividends, ROA and ROE
tends to reduce in years 0 and 1, but increase again in 2 subsequent years.
CHAPTER 3:
METHODOLOGY
3.1. Research gaps



shares in the year of changing dividend is positively correlated (+) with the future
profitability of listed companies on the stock market.
Hypothesis H4: In Vietnam, dividend changes along with repurchasing
shares in the year of changing dividend will decrease information on profitability
of listed companies on the stock market.

3.3.2. Research model
In order to test the research hypotheses above, two models of Nissim and
Ziv (2001) are applied:
MH1: (Et - Et-1)/B-1 = α0 + α1 DIV0 + α2ROEt-1 + εt
MH2: (Et - Et-1)/B-1 = α0 + α1DPC. DIV0 + α2DNC. DIV0 + α3ROEt-1
+ α4(E0 – E-1)/B-1 + εt
Then, the model proposed by Choi et al. (2011) is applied:
MH3: (Et - Et-1)/B-1 = α0 + α1 DIV0 + α2ROEt-1+ α3 DIV-1 + α4ROEt-2
+ α5(E0 – E-1)/B-1 + εt
Next, nonlinear model of Grullon et al. (2005) is applied:
MH4: (Et - Et-1)/B-1 = α0 + α1DPC. DIV0 + α2DNC. DIV0 +
+ (β1DFE0+β2NDFED0.DFE0+β3NDFED0.DFE02+β4PDFED0.DFE02)
+ (λ1CE0+ λ2NCED0.CE0+ λ3NCED0.CE02 + λ4PCED0.CE02) + εt
Finally, the regression model of Nissim and Ziv (2001) is extended by
adding control variables in the model including: Enterprise size (SIZE), investment
opportunities (MTB) and growth rate (GROW).
MH5: (Et - Et-1)/B-1 = α0 + α1DPC. DIV0 + α2DNC. DIV0 + α3ROEt-1
+ α4(E0 – E-1)/B-1 + α5SIZE0+ α6MTB0 + α7GROW0 + εt
In which:
t = 0 is the dividend change year, the research tests with t =1 and t =2
DIV0: the dividend change in year 0
Et: Earnings in year t after dividend change of enterprise i
Et-1: Earnings one year before year t of enterprise i
B-1: Book value of equity one year before the dividend change year

City Stock Exchange and Hanoi Stock Exchange in the period 2008-2017.
Data on dividends and financial ratios of non-financial companies listed on the
Vietnam stock market are taken from StoxPlus data.
After collecting data on dividends and financial indicators of enterprises
in the research sample, the author calculated the value of dependent variables
(Et - Et-1)/B-1 and (ROAt – ROAt-1) with t=1 and t=2, independent variable is the
annual dividend change DIV0: DIV0 = (DIVi,0– DIVi,-1)/DIVi,-1
Finally, research data includes 657 observations of dividend increases,
893 observations of dividend decreases and 643 no-change observations.
3.5. Methodology
In order to study the relationship between dividend changes and future
profitability of enterprises on Vietnam stock market, both qualitative research
and quantitative research methods are used. Qualitative research is conducted
by evaluating the correlation between dividend changes and profitability
changes in the dividend change year (year 0) and the two subsequent years
(year 1 and year 2) of companies listed on Vietnam stock market. Quantitative

research is performed with 5 regression models on the relationship between
dividend changes and future profitability, Fixed Effects Model (FEM) and
Random Effects Model (REM) based on panel data, with statistical software
STATA14. However, Hausman test results show that FEM is more suitable
than REM. After that, the author detects autocorrelation, heteroscedastic errors
and multicollinearity in the models. The results show that both autocorrelation,
heteroscedastic errors present in the models. The cluster command in stata is
used to overcome the autocorrelation and heteroscedasticity in the FEM.
CHAPTER 4
TEST RESULTS OF THE RELATIONSHIP BETWEEN
DIVIDEND CHANGES AND PROFITABILITY OF THE LISTED
COMPANIES ON VIETNAM STOCK MARKET
4.1. Descriptive Statistics

again.
Thus, the results of empirical research support hypothesis H1: In
Vietnam, dividend changes have a positive correlation (+) with the
profitability in the first year subsequent to the dividend change of listed
companies on the stock market; in addition, it supports the hypothesis H2: In
Vietnam, dividend decreases have a negative correlation (-) with profitability
of 2 years subsequent to dividend decreases of listed companies on the stock
market. These result are different from the research results of Nissim and Ziv
(2001) which supported the signaling theory with t = 1 and t = 2 on the US
stock market, Harada and Nguyen (2005) also supported the signaling
hypothesis and provided that dividend changes are positively correlated with
changes in profits in the following years on the Japanese stock market;
Grullon et al. (2005) investigating the US stock market and Reza et al.
(2014) studying the Chinese stock market supported the signaling theory of
dividends in case of dividend increases but found no evidence when dividend
decreases. In contrast, Choi et al. (2011) examined the Korean stock market
only found evidence of signaling theory in case of dividend reduction with t
= 1, but found no relationship between dividend increases and changes in
profits in the following years.
Besides, the correlation coefficients of ROEt-1 in 3 models show that ROE of
the previous year has an adverse effect on changes in profitability of the following
year with T = 1 and T = 2
4.2.1.2. Regression results of nonlinear model (Model 4)
Regression results of nonlinear model of Grullon et al. (2005) (MH4) are
similar to results of the last linear model of Nissim and Ziv (2001) (MH2).
4.2.1.3. . Regression results of extended linear model (MH5)

The results of the extended linear model also support the hypothesis H1
and hypothesis H2 in Vietnam stock market. The Beta coefficient of enterprise
size (SIZE) is negative and significance level at 1% with T = 1 and T = 2,

19

20

enterprises will increase again after 2 years subsequent to dividend
decreases.
4.2.3. Results of the relationship between dividend changes and future
profitability: in case companies repurchasing shares and in case companies not
repurchasing shares during the year of the dividend change.
When comparing the regression results of MH5 with 2 separate data sets: (1)
The enterprises change dividends and repurchase shares in the market during the
year of the dividend change; (2) The enterprises change their dividends and did
not repurchase shares during the dividend change year, hypothesis H3 is
supported: In Vietnam, dividend changes and not repurchasing shares are
positively correlated (+) with future profitability of listed companies on the stock
market. At the same time, the hypothesis H4 is supported: In Vietnam, dividend
changes and repurchasing stocks will reduce the content of information about
future profitability of listed companies on the stock market. These results
contribute to finding specific cases in favor of signaling dividends.
4.3. Robustness analysis with dependent variable of changes of return on
assets
The model with dependent variable ROAt – ROAt-1 is rewritten as follows:
ROAt - ROAt-1 = α0 + α1DPC. DIV0 + α2DNC. DIV0 + α3ROAt-1
+ α4(ROA0 – ROA-1) + α5SIZE0+ α6MTB0 + α7GROW0 + εt
The results of Robustness analysis with entire data
The study also found evidence to support the signaling theory of dividend
when changing dependent variable by changes in ROA. These results are
different from the research results of Grullon et al. (2005) on the US stock
market that there is a negative relationship between dividend changes and
changes in ROA in 2 years subsequent to the dividend change year. Choi et al.

enterprises in one following year will increase, dividend decreases indicate that
the profits of enterprises in one following year will decrease. However, the
decreases in profits are higher than the increases in profits
In addition, in companies with dividend decreases, the profitability will
rebound after 2 years. This result is contrary to the signaling theory.
5.1.1.2. With dependent variable of changes in ROA (ROAt – ROAt-1)
The research results also found empirical evidence supporting the
signaling theory of dividends when changing dependent variable to changes in


21

22

ROA with T = 1. Moreover, in case of dividend increases, ROA in 2 years
subsequent to the dividend change year may still increase.
5.1.2. Summary of research results on the relationship between dividend
changes and future profitability of enterprises by sector and ownership.
5.1.2.1 Summary of research results on the relationship between dividend
changes and future profitability of enterprises by sector
Dividend increases in all sectors is a positive signal of profitability for
one subsequent year. However, when enterprises reduce dividends, it can only
include that industrial enterprises will reduce profits in one following year, but
there is not enough evidence to confirm the profitability in one following year
of other non-financial companies. In addition, with other non-financial
enterprises, profitability may increase in 2 years subsequent to the dividend
change year.
5.1.2.2. Summary of research results on the relationship between dividend
changes and future profitability of enterprises by ownership.
The results of the relationship between dividend changes and profitability

changes to predict the profitability of enterprises in the short term. In the long
term, investors can rely on the ratio of return on equity of previous year (ROEt1), size of enterprises (SIZE) to predict the profitability of the following year
due to their negative effects on future profitability. Investors should also learn
about investment opportunities and the growth rate of the business in the
dividend change year to make a forecast of profitability in the following year.
Thirdly, when predicting the future profitability based on the information
of dividend changes of enterprises, investors need to understand the sectors of
enterprises, because the signaling dividends vary by sector.
Fourthly, when realizing that a company changes its dividends, investors
should consider whether the enterprise is state-owned or privately owned, and
whether the enterprise repurchases shares in the dividend change year.
Fifthly, in fact there are cases in which enterprises use dividends as a tool
to “polish” themseves. Therefore, when studying the dividend policy, investors
need to carefully analyze to identify the situation more accurately.
5.2.2. Recommendations for enterprises when developing dividend policy
5.2.2.1. Basic perspectives when establishing dividend policy.
5.2.2.2. Recommendations for enterprises when developing dividend policy
Consider the information content of dividend payment notice


23

Develop a stable dividend policy, suitable with investor psychology
Enterprises need to try to maintain a stable dividend policy, in order to convey
positive signals to shareholders about the company's prospects.
Avoid abrupt dividend cuts due to investor psychology
Develop dividend policy in line with the cycle of enterprises
An enterprise in a period of high growth rates with many attractive
projects should pay low dividends to reinvest, and improve profitability. By the
time the growth rate decreases and the investment opportunities are less, the

considering investing in a stock. Based on information asymmetry between
investors and business managers, the signaling theory of dividends, the thesis
has studied the relationship between dividend changes and future profitability of
non-financial companies listed on stock market in Vietnam. The results of
empirical research on dividend signaling in the world give many different
conclusions, but the thesis has provided convincing evidence of the causal
relationship between dividend changes and profitability in 1 subsequent years in
Vietnam stock market. In general, on Vietnam stock market, dividend increases
imply that the profits in one following year of the company will increase,
dividend decreases indicate that the profits in one following year of the
company will decrease, but the decreases in profits will be higher than the
increases in profits. However, the signaling theory is not completely confirmed
with enterprises of different sectors, enterprises categorized by state-owned and
privately owned factors, or enterprises that buy back/not buy back shares in the
dividend change year. Specifically: State-owned enterprises do not support
signaling theory of dividends; with companies changing dividends at the same
time repurchasing stocks in the market, research does not find any evidence of
the relationship between dividend changes and future profitability. Besides, the
thesis also proves that factors: the return on equity of the previous year,
enterprise size, investment opportunities and growth rate are also affects on
future profitability of firms.
From the research results of the thesis, the author has provided some
recommendations for investors when predicting the profitability of enterprises
in 2 subsequent years based on signs of dividend changes; recommendations for
enterprises to provide more reasonable and effective dividend policies develop
dividend policy; as well as proposing some suggestions for state management
agencies in issuing regulations on information disclosures of dividends.