FOREIGN TRADE UNIVERSITY
HO CHI MINH CITY CAMPUS
------***------

MID-TERM ASSIGNMENT
SUBJECT: FINANCIAL MARKETS AND INSTITUTIONS
Topic: Stock return seasonalities in the Vietnamese stock market
Lecturer: Ms. Trương Thị Thuỳ Trang
Supervisor: Ms. Nguyễn Thu Hằng
Class: K57CLC2
Group 5:

TABLE OF CONTENTS


LIST OF TABLES
The list below shows the tables included in the paper and the page on which each one is
located.
Name of table

Page

TABLE OF ABBREVIATIONS
The following table describes the significant of various abbreviations and acronyms used
throughout the paper. The the page on which each one is defined or first used is also
given.
Abbreviation
HNX
HNX-Index
HOSE
TOM

UPCOM
UPCoM
VN-Index

Meaning
Hanoi Stock Exchange
Hanoi Stock Exchange Index
Ho Chi Minh Stock Exchange
Turn-of-the-month
Composite index of the Unlisted Public Company Market
Unlisted Public Company Market
Ho Chi Minh Stock Exchange Composite Index

Page
19
6
19
8
6
14
6


Stock return seasonalities in the Vietnamese stock market

ABSTRACT
In this paper, we focus on explaining the existence of seasonality in monthly and weekly
rates of return on the Vietnamese Stock Market over 12 years, from January 2010 to
January 2021. Vietnamese Stock Market, as its nature, is a significant emerging market,
which is attentively attended by both domestic investors and foreign investors. The

investors’ purpose is mainly seeking opportunities to earn abnormal returns through the
market and such market efficiency is named anomaly of which seasonality is one pattern.
We test the seasonality in the Vietnamese Stock Market using monthly and weekly rates
of return of three Vietnamese Stock Exchanges. This research is carried out by
Quantitative research method and OLS estimation. The results indicated that the
Vietnamese Stock Market is not fully efficient yet. The research also found that there
certainly is seasonality in monthly rates of return on the Vietnamese Stock Market, of
which January generates the most significant positive returns. Moreover, the research
proved the existence of seasonality in weekly rates of return on the Vietnamese Stock
Market, of which Friday generates the most significant positive returns. This result shows
that the Vietnamese Stock Market is influenced by the so-called “January effect” and
“Friday effect”.
Keywords: Vietnamese Stock Market, seasonality, January effect, Friday effect.

1 INTRODUCTION
Ever since the recognition of seasonality in stock return, it has remained a hard yet
interesting challenge for researchers to confront. The reason behind this attention to
seasonality in stock return is that seasonality in stock return could be an important factor
in building and explaining capital market theory, capital market efficiency, and the nature
of the distribution of stock returns. Much research has been carried out whose purpose is
to find out whether the monthly and weekly pattern in stock returns exist and to test
hypotheses explaining such pattern. However, the results are diverse and some are even
contradictory to others. Research carried out by Bonin and Moses (1974) and Officer
3


Stock return seasonalities in the Vietnamese stock market

(1975) is typical research that found out the seasonality in the stock returns. Bonin and
Moses (1974) concluded that seasonality had existed in individual stock returns for

reasonably long periods and expressed their suggestion for further research concerning
variables that served as causal factors for this phenomenon. Officer (1975) found out the
existence of seasonality using Australian share market data and suggested research for
underlying causes of this seasonality. The research by Granger and Morgenstern (1963,
1970), on the other hand, indicated that the seasonal variations which also means
seasonality herein were small, insignificant and that such seasonal variations could not
play as a factor for predictability of market prices. According to Predictability of Stock
Market Prices (1971), the variations of stock market prices were random on long-term and
undetectable trends. The lack of consensus in older research is due to limited evidence
and the application of different techniques to measure seasonality. The similarities
between the research are the omission of possible causal factors of seasonality and the use
of seasonality for finance models and estimations. Also, there are only a little research and
papers relating to seasonality that could be applied to emerging markets.
Realizing the shortcomings and the diversity in results of precedent research, we decided
to carry out this research concerning stock return seasonalities in the Vietnamese stock
market. We herein put our attention on finding the existence of seasonality in the monthly
cycle and daily cycle and giving our possible explanations as well as further discussions.
With the monthly cycle, we focus and also realize the “January effect” in the Vietnamese
stock market. “January effect” means that the stock market has a tendency to increase
between the last day of December and the end of the first week of January. The main
reason for this effect is that the investors decide to sell their stocks right before the end of
the year in order to claim a capital loss for tax purposes and that the investors want to play
it safe with their investments (In the Vietnamese stock market, the stock exchanges close
on the Tet holiday at the beginning of the year, which makes it riskier for investors to hold
their stocks). Therefore, the stocks’ prices go down at the end of the year and go up again
at the beginning of the year when the investors start buying back the stocks. With the
4


Stock return seasonalities in the Vietnamese stock market


daily cycle, we found that there is a “Day of the week effect” which briefly means that the
stock market’s average return tends to be especially higher or lower in a specific day in
the week. That day in the Vietnamese stock market that has been found by us is Friday,
and such special pattern on Friday is called the “Friday effect”.
We extracted the data from three Vietnamese Stock Exchanges between January 2010 and
January 2021: VN-Index (Ho Chi Minh Stock Exchange Composite Index), HNX-Index
(Hanoi Stock Exchange Index), and UPCOM (a composite index of the Unlisted Public
Company Market). The Vietnamese market is emerging dramatically, which builds a firm
foundation for the development and dynamic of the Vietnamese stock market. With that
promising characteristics, the Vietnamese stock market may become a prospect of
investment for both domestic and foreign investors, and this research is made with a view
to supply information for the knowledgeable as well as enthusiastic investors and
researchers if that day would come.
This paper is developed and organized in the following order. The second section would
be Literature Review, which brings out older research and some more detailed definitions
used throughout our research. The third section is Methodology where we build our
models and test the models. This is followed by Result and Discussion, the main objective
of which is to present the results of the models, explain the results and give our possible
further opinions upon the models and their results. The final section should be Conclusion
which contains outstanding and remarkable points of our research.

2 LITERATURE REVIEW
The Efficient Market Hypothesis, also known as the Efficient Market Theory, states that
at any given time, all new information is captured by market participants and instantly
reflected in market prices (Fama, 1970). Reilly (1989) implied that an efficient market is a
market in which stocks' prices change rapidly according to the spread of new information.
Hence the current stocks' price reflects all the available information. According to Mishra
(2009), stocks are always traded at their fair market value on exchanges, making it
5



Stock return seasonalities in the Vietnamese stock market

impossible for investors to buy or sell under or over-valued stocks. Therefore, investors
cannot increase their profit through selecting stocks from experts' advice or choosing
some point of time to trade; and the only way they can improve their capital gain is by
selecting riskier investments. If all market participants accept the efficient market
hypothesis, they may not be interested in the new information, and then prices will not
fully reflect this information. However, efficient market theorists argue that not everyone
responds so, and some make surplus profits based on old information. Many people will
act the same way, and as a result, the market returns to a state of efficiency. In short, an
efficient market is a steady-state and a self-monitoring mechanism.
Weak form efficiency is one of three forms of the efficient market hypothesis besides
strong form and semi-strong form efficiency. It states that stock's prices reflect all the
available public information but may not reflect new information that is not yet publicly
available. It additionally assumes that the previous information relating to prices,
volumes, and returns is independent of future prices. Besides, it also implies that technical
trading strategies cannot provide consistent abnormal returns. With weak form efficiency,
technical analysis is not considered accurate, and even fundamental analysis, at times, can
be false. Therefore, it is incredibly challenging to outperform the market with weak-form
efficiency, especially in the short term. For example, suppose an investor believes that the
market is weak-efficient. In that case, he will pick an investment or a portfolio randomly
instead of having a financial advisor or an active portfolio manager because all of them
provide similar returns. Malhotra et al. (2015) realized weak form efficiency for monthly
returns in ten stock exchanges in Asia-Pacific markets. Said and Harper (2015) used
autocorrelation and Box-Ljung test statistics to prove weak-form efficiency in daily index
return in the Russian stock market from 2003 to 2012. Besides, a clear demonstration of
the weak efficient market is the calendar effect.
The calendar effect is also known seasonal effect, refers to changes in market prices or

market indices due to date, month, or any specific point of time in the year relate to
6


Stock return seasonalities in the Vietnamese stock market

stocks' price. The calendar effect includes the “January effect”, the “Friday effect”, the
“Halloween effect”, the “Day of the Week effect”, the “Month of the year effect”, the
“Turn-of-the-year effect”, the “Holiday effect”, etc. People try to determine a certain
period to test the remarkable phenomenon about the stock returns, then find out if there
are any rules they can follow or any opportunities they can catch to generate capital gains.
Hereafter we will give some initial research as well as theories about the "Day of the
week effect", the "Friday effect", the "Turn-of-the-month (TOM) effect", and the "January
effect". However, we would like to stress that our focus is "Friday effect" and "January
effect", and we would also give our explanations considering such focus.
First of all, we would like to look into the "Day of the Week effect". The "Day of the
Week effect" is when one specific day of the week has different returns than the other
days' average returns. Friday's returns are usually significantly higher in the US market
than the other days' average returns, called the Friday effect. One possible reason for this
situation is that the equity market's trading date is not always the same as when payment
is made (settlement date). Investors have to wait for more days when their transactions are
recorded on Friday (because of the weekend holiday). Therefore, they will use money in
some alternative markets for a few days (if their transactions are realized on Friday) until
payment is made. Gibbons and Hess (1981) found that in the US Stock Market from 1962
to 1978, the Mondays' returns were usually lower than the other days' average returns and
the Friday returns were much higher than the other days' average returns. Keim and
Stambaugh (1984) also reported a similar result while using the data from 1928 to 1982.
Gay and Kim (1987) found high returns on Fridays and Wednesdays in the Commodity
Research Bureau future index. Muradoglu and Oktay (1993) found that returns on
Tuesdays were usually negative while on Fridays were positive in the Istanbul Stock

Exchange (ISE). Kohers and Patel (1996) analyzed the day of the week effect for junk
bonds and found that returns on Friday are all positive and higher than average returns of
any day of the week. Georgantopoulos, Kenourgios, and Tsamis (2011) reported that the
day of the week effect exists in the Greek and Turkish equity market returns. Rodriguez
7


Stock return seasonalities in the Vietnamese stock market

(2012) used GARCH(1,1) models and documented that Friday's returns were higher than
other days of the week for all the stock markets in Latin America over 1993 – 2007.
The initial research has inspired us that the "Day of the Week Effect", or more specifically
the "Friday effect", may occur in the Vietnamese Stock Market. The Vietnamese Stock
Exchanges all close on the weekends (Saturday and Sunday). The stock market is
susceptible and quickly reacts to the economy's changes and the country's political and
social situation. Therefore, a two-day-closure is a risky period for the investors as they
cannot adjust their portfolio under the economy's flow. For example, if there is news
saying that a company's factory caught fire on Saturday, many investors would become
terrified and consider such bad news a sale signal. The investors then try to sell that
company's stocks as soon as possible and receive lower than expected profit. For that
reason, investors in general and risk-averse investors may choose to sell their stocks on
Friday, which increases the mean return of Friday compared with other days of the week.
Keeping this argument in mind, we inquire into the existence of the "Friday effect".
Secondly, we come into the studies regarding the "Turn-of-the-month effect". The "Turnof-the-month (TOM) effect" is defined as the tendency of stock returns sudden change in
the period between the end of this month and early next month. Ariel (1987) first
documented the effect's pattern in an analysis of advice that sales should be postponed to
the latter half of the month and purchases should be made before month-ends to earn
unusually high returns accrued in the early day of the month. Lakonishok and Smidt
(1988) referred to the four days from the last trading day of the previous month to the first
three days of the current month as the TOM period usually creates positive returns for the

Dow Jones Industrial Average through 1897 – 1986. The turn-of-the-month effect, which
is considered a sign of an ineffective market, is usually noticed in seasonal reports as
anomalies to help investors predict the market's tendency and avoid risk.

8


Stock return seasonalities in the Vietnamese stock market

There are three plausible explanations for the TOM effect. The first one, which is called
the liquid fund hypothesis, relates to the payment day customs. Ogden (1990) found that
regularity in payment date of wages and interest or dividend income will generate the
supply of "liquid fund" at the end of the month, and the flow of these funds into the
market will push the stock price up, leading to higher mean returns at the turn of the
month. The second one, called the window dressing hypothesis, states that fund managers
adjust their portfolios to "window dress" their reports. Haugen and Lakonishok (1987)
found that the funds created when the fund managers return to their prior portfolios may
result in higher mean returns around reporting dates. The final explanation, which is
called the news clustering hypothesis, relates to good news and bad news spreading time.
McNichols (1988) showed that firms voluntarily spread the good news in the early days
of the month and prevent bad news until reporting deadlines.
After considering those possible causal factors of the TOM effect, we think some
characteristics of the Vietnamese Stock Market lowers the likelihood of the TOM effect.
Firstly, the wage is not always paid at the end of the month. Many companies decide to
pay wage semi-monthly or weekly instead of paying monthly from employees'
convenience, especially those living in big cities. This helps the employees to have money
to cover daily costs without putting much effort into estimating the budget. As a result, the
employees reduce daily costs, become less stressed, and perform better at work. Because
of such changes in paying wages, the liquid fund hypothesis is not fully applied in the
Vietnamese Stock Market. Secondly, considering the window dressing hypothesis and the

news clustering hypothesis, the absence and delay of news can cause the TOM effect.
Obviously, asymmetric information negatively affects the firm and leads to a decrease in
the firm's value (T.L.D. Huynh et al., 2020). Many efforts have been made to limit the
asymmetry of information in favour of the efficiency of the economy, and the
development of media is one of the solutions. The investors can have information about
companies from securities company's service registration, learning of companies' public
financial statements, periodic stock analysis from reputable analytics departments, etc.
9


Stock return seasonalities in the Vietnamese stock market

Besides, due to the close linkage between economy, politics and society in the Vietnamese
Stock Market, any news that can create significant effects on a company as specific and
the economy as general shall be public not to cause more significant repercussions on
many industries. Also, although the insiders are illegal, the insider-outsider theory (Assar
Lindbeck and Dennis J. Snower, 1989), which explains how the "insiders" (incumbent
employees of the company) get market power in comparison with the "outsiders"
(investors or people are not in such company), is still applicable in the Vietnamese Stock
Market. Therefore, it is tough for companies to keep all information private and disclose it
whenever they want. The window dressing hypothesis and the news clustering hypothesis
are not applicable for the reasons mentioned above. Overall, the TOM effect is negligible
in the Vietnamese Stock Market, and we do not analyze the TOM effect in this paper.
Finally, the "January effect" is a hypothesis that there is a seasonal anomaly in the
financial market where securities' prices increase in January more than in any other
month, according to Mills and Coutts (1995). This calendar effect would create an
opportunity for investors to buy stocks for lower prices before January and sell them after
their value increases. Wachetel (1942) was the first to report the January Effect in the
Dow Jones Industrial Average from 1927 to 1942. The phenomenon attracted little
attention until Keim (1983) again documented the abnormal return in January for the

common stock in the New York Stock Exchange (NYSE) and the American Stock
Exchange (AMEX) throughout 1963 -1979.
A large number of literature investigations about the January effect were documented.
Rozeff and Kinney (1976) found an abnormally high mean return in the stock market
during January. Keim (1983) documented that these high January returns accrue
disproportionately to small firms during the early days of January. Roll (1983) found that
the small firms' stocks tend to earn excess returns in January and with much of the effect
concentrated in the first few days of the month. Chan et al. (1985) stated that holding
stocks during January is significantly riskier than in other months of the year. Some of the
10


Stock return seasonalities in the Vietnamese stock market

information is ignored at the beginning of the year will bring significantly high returns to
the investors in January.
The January effect in the corporate bond market is well-documented. Chang and Pinegar
(1986) examined the monthly holding returns of treasuries and bonds rated Aaa, Aa, A,
Baa, Ba, and B; and found that positive excess return in January for bonds rated Ba and B.
Chang and Huang (1990) used a different methodology. However, they had the same
result as Chang and Pinegar (1986) proving, that the lower the bonds' quality, the more
apparent the January effect.
Two proposed explanations for the January effect include the tax-loss selling hypothesis
of Roll (1983) and the window dressing hypothesis of Haugen and Lakonishok (1987).
The first one, the tax-loss selling hypothesis, proposes that individual investors tend to
sell stocks that fall in price at the end of the year to generate capital losses to offset
taxable income and avoid tax on capital gains. After that, the selling pressure disappears,
and prices rebound to the equilibrium level. D'Mello, Ferris and Hwang (2003) found an
abnormal selling pressure on the stocks that experience large capital losses before yearends and found that investors tend to postpone stock sales that experience capital gains
until after the New Year. These findings show that individual investors, rather than

institutional investors, are the major sellers at the end of the year; individual tax-loss
selling is the fundamental explanation for abnormal January returns. However, Fortune
(1991) noted that the tax-loss selling hypothesis is insufficient with the efficient market
hypothesis. According to the efficient market hypothesis, investors who do not have
taxable income will identify any trend toward the abnormally low price in December and
become oversold buyers at the end of December. This means that taxable sales affect
ownership of the shares, not their price. The other explanation, the window dressing
hypothesis, argues that institutional investors eliminate losers and buy winners and wellknown stocks at the end of the year to "window dress" their annual reports. Ritter and
Chopra (1989) stated that institutional investors' window dressing puts pressure on small
11


Stock return seasonalities in the Vietnamese stock market

stocks at the turn of the year. Haugen and Lakonishok (1987) found that the funds created
when the fund managers return to their prior portfolios may result in higher mean returns
around reporting dates.
January Effect affects stock return not only in the US market but also in the international
market. Ayadi et al. (1998) provided evidence of abnormal January returns in Ghana.
Reyes (2001) reported the January Effect's observation in small-firm stock in the Japanese
Stock Market. Yakob et al. (2005) reported the January Effect in Taiwan and Malaysia for
2001-2005. According to the efficient market hypothesis, these abnormally high returns in
January due to higher risk, but Keim (1983) found that these abnormally high returns had
a close relationship with the January effect.
We realize that January is also a particular month for the Vietnamese Stock Market. Most
of the time, the Tet Holiday - the most extended and most important holiday of Vietnam in
January (Tet Holiday depends on the Lunar calendar, which deviates from the Gregorian
calendar). The Vietnamese Stock Exchanges close down on this holiday, which raises
concerns to the investors because they could not react to the economy's changes and
adjust their portfolio. A typical example of this is January of 2019. The COVID-19

pandemic occurred right on the Tet holiday of 2019, which negatively affected the world
economy in general and Vietnam. Vietnamese stock exchanges' stock prices fluctuated
dramatically and unpredictably, which raised awareness and fear upon investors and
brought enormous losses for investors. To prevent such cases, the investors sell their
stocks at the end of the year and start trading actively to buy back the stocks at the
beginning of the following year. Besides, risk-averse investors try to avoid this holiday's
risks by selling to take profit at the end of the year and trading back at the beginning of
next year. Regarding the tax-loss selling hypothesis of Roll (1983), we argue that such a
hypothesis is more applicable in developed countries than in developing countries. The
reason is that taxation laws in Vietnam are preferable in comparison with other countries.
Due to the Government's preferable taxation laws, Vietnamese citizens and corporations'
12


Stock return seasonalities in the Vietnamese stock market

tax amount is much lower than in other countries. This tax rate and tax set are suitable for
citizens' income and are supported by the Government in some particular cases.
Moreover, the time to submit tax is scattered throughout the year, not only at the end of
the year, like in developed countries. Therefore, we argue that the "January effect" is
mainly due to the Tet holiday and investors' risk aversion. We proceed to test such
existence in this paper.
This paper examines the "Friday effect" and the "January effect" to illumine the seasonal
effects on Vietnam's stock returns to prove our previous arguments. Research papers on
this kind of topic are abundant in the world but extremely rare in Vietnam. Moreover, they
are also not entirely suitable and applicable for Vietnam's circumstances, an emerging
country whose stock market still seems to be not as efficient as other countries (Chung
Tien Luu, Cuong Hung Pham & Long Pham, 2016). Therefore, we decide to study this
topic specifically on the Vietnamese stock market.


3 METHODOLOGY
3.1 DATA
Daily data on the VN-Index, HNX-Index and UPCOM were collected from Ho Chi Minh
Stock Exchange, Hanoi Stock Exchange and UPCoM Stock Exchange. Data used in the
research is gathered over the period of time from 4th January 2010 until 22nd January
2021.
3.2 JANUARY EFFECT
3.2.1 Hypothesis
As been mentioned in Literature Review, due to the effect of the Tet holiday and risk
aversion of investors, we argued that the “January effect” exists in the Vietnamese Stock
Market. We decide to proceed to check the validity of arguments with the following
hypotheses:
H0: µ1 = µ2 (The average daily log return of January is equal the daily log return of rest of
the months);
HA: µ1 ≠ µ2.
13


Stock return seasonalities in the Vietnamese stock market

Where:
-

µ1 = the average daily log return of January;
µ2 = the average daily log return of rest of the months.

3.2.2 Model
To test whether there is January in returns of 3 stock market, we use OLS regression
model:
Rt= α + β2 * FEB + β3 MAR+ β4*APR+ β5*MAY+ β6*JUN+ β7*JUL+ β8*AUG+

β9*SEP+ β10*OCT+ β11*NOV+ β12*DEC
Where:
-

Rt: The daily return of each index in t times. Rt = ln(Rt/Rt-1)*100;
FEB is dummy variables. For FEB = 1, if month t is February and 0 otherwise;
MAR is dummy variables. For MAR = 1, if month t is March and 0 otherwise;
APR is dummy variables. For APR = 1, if month t is April and 0 otherwise;
MAY is dummy variables. For MAY = 1, if month t is May and 0 otherwise;
JUN is dummy variables. For JUN = 1, if month t is June and 0 otherwise;
JUL is dummy variables. For JUL = 1, if month t is July and 0 otherwise;
AUG is dummy variables. For AUG = 1, if month t is August and 0 otherwise;
SEP is dummy variables. For SEP = 1, if month t is September and 0 otherwise;
OCT is dummy variables. For OCT = 1, if month t is October and 0 otherwise;
NOV is dummy variables. For NOV = 1, if month t is November and 0 otherwise;
DEC is dummy variables. For DEC = 1, if month t is December and 0 otherwise;
Β2 to β12: the difference in return between month t and January and the i month
with i runs from 2 to 12.

3.3 DAY OF THE WEEK EFFECT
3.3.1 Hypothesis
We argued that the existence of the “Friday effect” is because of the closure of the
Vietnamese Stock Exchanges on the weekends and the risk aversion of the investors. In
order to check that argument, we develop the following hypotheses:
H0: µ1= µ2 (The average daily log return of investigated day is equal the daily log return
of other weekdays);
HA: µ1 ≠ µ2.
Where:
14


µ1 = the average daily log return of the investigated day;


Stock return seasonalities in the Vietnamese stock market

-

µ2 = the average daily log return of the other weekdays.

3.3.2 Model
Daily data on the VN-Index, HNX-Index and UPCOM were collected from Ho Chi Minh
Stock Exchange (HOSE) and Hanoi Stock Exchange(HNX). Data used in the research is
gathered over the period of time from 4th January 2010 until 22nd January 2021.
The following formula is used to calculate daily returns:
Rt = ln(Pt/Pt-1)*100
Where:
-

Rt: the return over the period t;
Pt: the daily closed share price index of day t;
Pt-1: the daily closed share price index of day t-1.

To examine the impact of each day of the week on the stock returns in Vietnamese stock
exchange, we establish the following OLS regression model:
Ri = α + β1 * MON + β2 * TUE + β3 * WED + β4 * THU
Where:
-

Ri: the daily return of each index;
MON: dummy variable on Monday (MON = 1 for the observation on Monday;


-

otherwise MON = 0);
TUE: dummy variable on Tuesday (TUE = 1 for the observation on Tuesday;

-

otherwise TUE = 0);
WED: dummy variable on Wednesday (WED = 1 for the observation on

-

Wednesday; otherwise WED = 0);
THU: dummy variable on Thursday (THU = 1 for the observation on Thursday;

-

otherwise THU = 0);
Β1 to β4: the difference between the return on Friday and the returns on other days
of the week.

15


Stock return seasonalities in the Vietnamese stock market

4 RESULT AND DISCUSSION
4.1 JANUARY EFFECT
4.1.1 Descriptive data


Table 1. Descriptive statistic of the 3 indexes for 12 months
VN-Index
Standard
deviation

Mean

HNX-Index
Standard
deviation

Mean

UPCOM
Standard
deviation

Mean

JAN

1.2782500

0.2335

1.324965

0.1245


0.8461916

- 0.0098

FEB

1.3442507

0.0770

1.496107

0.1718

1.1253879

0.0126

MAR

1.3427028

- 0.0784

1.473182

- 0.0352

1.2295055


0.0720

APR

1.1866174

0.1181

1.320743

0.0264

1.1740924

- 0.0488

MAY

1.4827529

- 0.0628

1.4827529

- 0.1322

1.2800481

- 0.0407


JUN

1.0799501

0.0038

1.195789

0.0067

0.9122743

- 0.0437

JUL

1.0653150

0.0144

1.219449

- 0.0547

1.0230223

0.0083

AUG


1.2052826

0.0032

1.459101

- 0.0141

0.9023847

0.0292

SEP

0.8422387

0.0375

1.119112

0.0243

0.9127943

0.0886

OCT

0.9209891


0.0010

1.090351

- 0.1063

1.2951478

0.2166

16


Stock return seasonalities in the Vietnamese stock market

NOV

0.8352404

- 0.0199

1.042561

- 0.0171

1.2959758

- 0.2388

DEC


0.9877240

0.0546

1.342858

0.1824

1.1099055

0.0661

In descriptive statistic table, it can be seen that VN-Index has highest average return in
January. While HNX-Index has and UPCOM do not have highest highest average returns
in January (highest average returns month of HNX-Index and UPCOM is May and
November respectively). Even average return in January of UPCOM is negative which
means investor, on average, may face losses when invest in January. Therefore, the
January effect may only occur in VN-Index.
4.1.2 Results of OLS estimation
We use Augmented Dickey-Fuller Test to test the whether the time series is stationary or
not. The result of Augmented Dickey-Fuller Test is that p-value is smaller than 0.01.
Therefore, the regression is stationary.
We use OLS estimation to test whether there is January effect in Vietnam stock market.
The result of the estimation is shown in the table below.

Table 2. Estimated results of the January effect model
VN-Index

HNX-Index


UPCOM

Estimate*

t-value

Estimate*

t-value

Estimate*

t-value

(Intercept)

0.233469**

2.7338

0.132533

1.5107

0.0035012

0.0619

FEB


- 0.156440

- 1.1925

0.039279

0.2779

0.0090929

0.0903

MAR

- 0.311855*

- 2.5709

- 0.167697

- 1.3004

0.0684987

0.7058

APR

- 0.115381


- 0.9786

- 0.106099

- 0.8419

- 0.0522764

- 0.5317

MAY

- 0.296289*

- 2.2795

- 0.264742

- 1.8633

- 0.0442464

- 0.4349

17


Stock return seasonalities in the Vietnamese stock market


JUN

- 0.229670*

- 2.0745

- 0.125833

- 1.0719

- 0.0472160

- 0.5752

JUL

- 0.219105*

- 2.0047

- 0.187227

- 1.5942

0.0047613

0.0550

AUG


- 0.230275*

- 2.0006

- 0.146640

- 1.1442

0.0257391

0.3184

SEP

- 0.196005

- 1.9139

- 0.108236

- 0.9372

0.0851334

1.0212

OCT

- 0.232434*


- 2.2396

- 0.238856*

- 2.1304

0.2131038*

2.1233

NOV

- 0.255123*

- 2.5229

- 0.164037

- 1.4790

- 0.242350*

- 2.3722

DEC

- 0.178908

- 1.6799


0.049865

0.4047

0.0625959

0.6867

Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1
Regarding VN-Index, the result show that the January effect occurs in Ho Chi Minh Stock
Exchange. In more details, all variable (except FEB) are statistically significant which
means the coefficients are valid. The coefficient are all negative showing that the mean
returns in other months are lower than that in January, in other words, return in January is
the highest in a year in HOSE. Therefore, there is January effect in HOSE. Especially,
stock return in HOSE is lowest in March, which 0.31% lower than return in January.
Regarding to HNX-Index and UPCOM, there are only 2 statistically significant variables
among explanatory variables and not all coefficients are negative. It means return in
January statistically is not the highest in a year in Hanoi Stock Exchange and UPCOM.
Therefore, there is no clear evidence for January effect in UPCOM and HNX-Index.
4.2 FRIDAY EFFECT
4.2.1 Descriptive data

Table 3. Statistic description of VN-Index daily return
N

Range

Minimum Maximum

MON 539 11.34199 -6.481975 4.860016

18

Mean
-0.07374635

Std. Deviation Variance
1.394568

1.944821


Stock return seasonalities in the Vietnamese stock market

N
TUE

552

Range

Minimum Maximum

Mean

Std. Deviation Variance

8.77535 -5.247995 3.527356

-0.02805448


1.127841

1.272025

WED 554

8.71608 -4.114910

4.601170

0.09186578

1.022165

1.044821

THU

559

9.41388 -6.051190

3.362685

0.02146697

1.127787

1.271904


FRI

554

7.43327 -4.023881 3.409384

0.13308309

1.004790

1.009603

Table 4. Statistic description of HNX-Index daily return
N

Range

Minimum Maximum

Mean

Std. Deviation Variance

MON 539

12.16348 -6.657003 5.506476

-0.15716103

1.606730


2.581581

TUE

10.46084 -5.379204 5.081638

-0.06947156

1.285794

1.653267

WED 554

8.57737

-4.597660

3.979711

0.12009790

1.248582

1.558956

THU

559


11.40035

-6.615097 4.785252

0.00998943

1.298576

1.686298

FRI

554

11.12944

-6.754663 4.374778

0.14406921

1.123859

1.263059

552

Table 5. Statistic description of UPCOM daily return
N


Range

Minimum Maximum

MON 539 12.95666 -5.527088 7.429576
19

Mean

Std. Deviation

Variance

-0.07167765

1.180082

1.392594


Stock return seasonalities in the Vietnamese stock market

N

Mean

Std. Deviation

Variance


552 12.59453 -6.165094 6.429435

-0.00695428

1.153630

1.330862

WED 554 15.71628 -8.801088 6.915197

0.03001973

1.124458

1.264406

THU

559 13.28994 -7.204043 6.085895

0.03328590

1.098497

1.206697

FRI

554 11.80690 -5.101486 6.705410


0.07581961

0.972672

0.946090

TUE

Range

Minimum Maximum

The value shown in the descriptive statistic tables provides us some information related to
the distribution of the daily return during the days of the week. It can be seen that all of
the three indices have the highest returns on Monday. Besides, Monday also experienced
the lowest and negative average value of returns. Monday and Tuesday are the only days
of the week have the negative mean value over the period. Friday has the highest mean
value of daily returns compared with those values on other days of the week. It is
interested that Friday showed the most stable daily returns with the lowest Standard
deviation value, while Monday has the highest Standard deviation value.
4.2.2 Results of OLS estimation
We use OLS estimation to investigate the impact of each day of the week on the stock
returns in Vietnamese stock market. The result of the estimation is shown in the tables
below.

Table 6. Estimated results of the Friday effect model
VN-Index

(Intercept)
20


HNX-Index

Estimated*

t-value

Estimated*

t-value

0.13308**

3.118

0.14407**

3.017

UPCOM
Estimated
*
0.07582 .

t-value
1.833


Stock return seasonalities in the Vietnamese stock market


MON

- 0.20683**

- 2.807

- 0.30123***

- 3.583

- 0.14750*

- 2.250

TUE

- 0.16114*

- 2.508

- 0.21354**

- 2.942

- 0.08277

- 1.290

WED


- 0.04122

- 0.677

- 0.02397

- 0.336

- 0.04580

- 0.725

THU

- 0.11162 .

- 1.744

- 0.13408 .

- 1.842

- 0.04253

- 0.684

Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1

We analyze the Vietnam stock market to detect whether stock market exhibits Friday
effect, which can be seen as evidence of market inefficiency. The results show that

coefficients of all variables are negative which means the returns of other days are lower
than that of Friday. Regarding to the column of HNX-Index and VN-Index, all variables
(except Wednesday) are statistically significant. With regards to the column of UPCOM,
only one variable named MON is significant. Therefore, MON is the most statistically
significant variable amongst explanatory ones. Finally, the results of the studying
evidence that the mean returns on Friday is the highest in a week and is more statistically
significant regarding to VN-Index and HNX-index compared to UPCOM. Our findings
are similar to the previous ones conducted in the past, e.g. Gibbons and Hess (1981),
Keim and Stambaugh (1984).

5 CONCLUSION
As it has been mentioned hereinbefore, there are diverse results concerning seasonality
due to different sets of data, different techniques applied, and various applications of
methods and approaches. This is not only the inspiration for us to carry out this research
concerning the Vietnamese stock market but also the notice for any enthusiastic people
wanting to learn about seasonality that the results of each research are exclusive and
should be qualified as well as referred rationally. This research is not an exception, it is
21


Stock return seasonalities in the Vietnamese stock market

obvious that the different data sets used in the research (from three different stock
exchanges) generate different results in which the result from VN-Index is most
significant. However, the results definitely show the existence of seasonality in the
Vietnamese stock market, including both “January effect” and “Friday effect”.

Mentioning the “January effect”, many researchers have considered the institutional
factors (specifically the tax system) as the causal factors. Others possible factors are given
in the Literature Review section of this paper. However, with the taxation laws in

Vietnam, taxes cannot be the underlying cause for the “January effect”. Due to the
preferable taxation laws imposed by the Government, the amount of tax paid by
Vietnamese citizens and corporations is much lower than in other countries. This tax rate
and tax set are suitable with the income of citizens and are supported by the Government
in some particular cases. Therefore, rather than tax, the long holiday period and risk
preference of investors are supposed to play as underlying factors of the “January effect”.
As mentioned hereinbefore in the Introduction, the Vietnamese stock exchanges close
during the Tet holiday, which is quite a long and unpredictable period. COVID-19
pandemic spread widely and negatively affected the world economy in general and
Vietnam in particular during the Tet Holiday of 2019. The stock prices of Vietnamese
stock exchanges fluctuated dramatically and unpredictably, which raised awareness and
fear upon investors. Also, during the closed time of Vietnamese stock exchanges, it will
be riskier to hold the stocks as investors cannot sell or buy or do any actions in
accordance with the changes of marker to save their investments. As a result, they sell
their stocks at the end of the year and start trading actively to buy back the stocks at the
beginning of the following year.

Speaking of the “Friday effect”, the results show that Friday has the highest mean returns
in the week. The existence of the “Friday effect” maybe due to the risk aversion of
investors. As mentioned before, the Vietnamese Stock Exchanges close on the weekends,
of which many investors afraid. As a result, investors execute a massive sale of stocks on
22


Stock return seasonalities in the Vietnamese stock market

Friday, which explains the higher return of stocks on Friday in comparison with other
days of the week. By saying that, there might be chances for investors to make a profit in
the Vietnamese market from trading on Friday in accordance with the prevailed results.
The real question here is “Do investors really can earn abnormal returns by applying

seasonality?”, which will be answered hereinafter. Back to the research, it also indicates
that Monday has the most significant fluctuation in comparison with the rest. This, as
prevailed in other research, maybe due to the closed time during weekends of Vietnamese
stock exchanges. The stock market as a whole is very sensitive and two days off are big
problems and concerns for investors. If there are severe events during that time, Monday
tends to bear the most due to the reaction of investors to the weekend’s events and Friday
effect. For example, if investors follow the “Friday effect” and sell their stocks on Friday,
they may buy them back on the next Monday for reinvestment. If there is news
concerning problems of a certain company on the weekends, the investors may sell it on
the next Monday. These explain the most significance of Monday among other days of the
week.

The possibility for investors to earn a high profit or abnormal returns on this seasonality
effect, for us, is slight for these reasons. Firstly, we believe that the Vietnamese market is
opened and the investors are rational. Our market may seem less efficient than others, but
the investors can make a decision themselves by researching or read the available
information themselves concerning companies’ stocks and financial statements. Anyway,
the investors not only choose their investments individually but also base on the opinions
of the majority to decide. Therefore, if the seasonality is recognized, it is hard for
investors to earn significant abnormal returns in comparison with others. Secondly, the
data used to estimate the model is historical data. The stocks’ prices are affected by many
other factors and are unpredictable in the future. Knowing the seasonality of the stocks’
returns, but not the exact fluctuation or unpredictable events, as we believe, may not
allow any investors to have abnormal returns compared with others. However, we have
faith in the necessity of seasonality, of which investors should be aware in the assessment
23


Stock return seasonalities in the Vietnamese stock market


of their investment, development of portfolio strategies, and seizure of investment
opportunities.

The results of this paper are exclusive as we mentioned and explained above. Due to our
high belief in the promising future of the Vietnamese stock market, we do this research
with the hope that it would be a useful reference paper for any researchers in the future
who are interested in this topic and this research. We also think that this topic is suitable
with the current global and domestic situation and is worth being considered for further
research.

24


Stock return seasonalities in the Vietnamese stock market

REFERENCES
Al-Saad *, K., & Moosa, I. A. (2005). Seasonality in stock returns: evidence from an
emerging

market.

Applied

Financial

Economics,

15(1),

63–71.


doi:10.1080/0960310042000281185

Ariel, R. A. (1987). A monthly effect in stock returns. Journal of Financial Economics,
17, 161-174.
Assar Lindbeck and Dennis J. Snower (1990). The insider-outsider theory of employment
and unemployment. Long Range Planning, 23(2), 120. doi:10.1016/0024-6301(90)902302.
Ayadi, O. F., Dufrene, U. B., & Chatterjee, A. (1998). Stock return seasonalities in lowincome African emerging markets. Managerial Finance, 24(3), 22-33.

Bonin, J. M., & Moses, E. A. (1974). Seasonal Variations in Prices of Individual Dow
Jones Industrial Stocks. The Journal of Financial and Quantitative Analysis, 9(6), 963.
doi:10.2307/2329730

Chan, K. (1985). An exploratory investigation of the firm size effect. Journal of Financial
Economics, 14(3), 451–471. doi:10.1016/0304-405x(85)90008-x

Chang, E. C., & Huang, R. D. (1990). Time-Varying Return and Risk in the Corporate
Bond Market. The Journal of Financial and Quantitative Analysis, 25(3), 323.
doi:10.2307/2330699

25