Journal of Indonesian Economy and Business

Volume 35, Number 3, 2020, 257 – 269

Day-of-the-Week Effect and Investors’ Psychological Mood

Testing in a Highly Mispriced Capital Market

Rizky Luxianto 1, Usman Arief2* and Muhammad Budi Prasetyo1

1
Department of Management, Faculty of Economics and Business, Universitas Indonesia, Depok,
16424, Indonesia
2
Department of Management, Faculty of Economics and Business, Universitas Gadjah Mada,
Yogyakarta, 55281, Indonesia

ABSTRACT ARTICLE INFO

Research Aims: This research examines investors’ psychological moods Article information:
which cause day-of-the-week anomalies in highly mispriced stock Received in 21 February
markets. Design/methodology/approach: We use a sample from the 2020.
Indonesian capital market as, in the Asian region, this country is Received in revised form
considered to have a highly mispriced capital market. We decompose the 26 April 2020.
Received in revised form
stock price index in Indonesia into speculative, less speculative, and non-
27 May 2020.
speculative indexes. We employ the mean and variance regressions to
Accepted 3 June 2020.
control the heteroscedasticity and serial correlation. Novelties: Our
novelties are two fold. We postulate a method to decompose stock price Keywords:
indexes in Indonesia (the JKSE, LQ 45, and Kompas 100) into Indonesia, day-of-the
speculative, less speculative, and non-speculative indexes. Secondly, we week, anomaly,
estimate the mean and variance levels simultaneously to get a robust decomposition,
estimation result of the anomaly. Research Findings: We empirically psychology, mood,
find that the behavior mood hypothesis is supported only during normal behavioral finance
periods, when investors tend to be irrational and use their good mood to
trade on speculative stocks on a Wednesday and sell them on Monday. In JEL Code:
other periods, rationality and psychological effects play a role with G12, G15, G40, F30
Indonesian investors, when their mood is good they are more active in
trading less speculative stocks, to avoid higher risks and earn higher
returns from those less speculative and non-speculative stocks.

* Corresponding Author at Department of Management, Faculty of Economics and Business, Universitas Gadjah Mada,
Yogyakarta, 55281, Indonesia.
E-mail address: [email protected]
ISSN 2085-8272 (print), ISSN 2338-5847 (online) http://journal.ugm.ac.id/jieb
258 Luxianto, et al

INTRODUCTION returns at that time. The last argument is related

The efficient market hypothesis (EMH) to the investors’ psychology and moods. This
proposed by Fama (1970) stated that all the hypothesis says that the mood of traders tends to
available information is incorporated into a increase from Thursday to Friday and then
stock’s price, and no one can beat the market decrease on Monday. Hence, traders tend to
from taking advantage of that kind of infor- evaluate prospects optimally when they are in a
mation. Nevertheless, in a capital market, there good mood, rather than in a bad mood.
are several empirical findings which reveal Some empirical findings (Wright & Bower,
several anomalies, namely the calendar effect, 1992; Bagozzi et al., 1999) supported the
size effect, low book value, and winner-loser. investor psychology hypothesis. They argued
These findings argue that capital markets are not that people who are in a good mood tend to
fully efficient since the theoretical frameworks evaluate good stimuli such as consumer goods,
cannot clearly explain the existence of life satisfaction, or past life experiences more
anomalies. positively than when they are in a bad mood.
One of the well-known and most research This happens with investors too,if they are in a
anomalies is the calendar effect. This anomaly good mood at the end of the week, they tend to
explains that the high and the low returns in a buy stocks since they see the prospective value
capital market have a seasonality pattern. As a of a stock increasing in the future.
consequence, investors can beat the market using A recent study byChiah and Zhong (2019)
a simple strategy of buying at a low price and examined the day-of-the-week effect
selling at a high price, or buying at a high price internationally. They useddata from 24
and selling at a higher price. Some calendar developed stock market countries around the
effects that are often examined by researchers world. Their finding is quite remarkable, since
are time-of-the-day effects, day-of-the-week positive global returns on Friday, and negative
effects, week-of-the-month effects, and month- returns on Monday are a manifestation of the
of-the-year effects. The first study of these mispricing effect, which is driven by investor
calendar effects is the day-of-the-week effect, by sentiment. They also reveal that the optimistic
Kelly (1930), which revealed that the United investors tend to buy speculative stocks on
States stock markets’ returns are relatively lower Fridays, and vice versa, the pessimist investors
on Mondays and then at their highest on Fridays. tend to sell (or avoid trading) on Mondays. They
In general, the day-of-the-week phenomenon concluded that psychological investors lead the
is a tendency for stock returns to be relatively day-of-the-week effect around the world.
higher at certain times (e.g., Friday) and lower at Another study that supports Chiah and Zhong
other times (e.g., Monday). With this anomaly, (2019) is by Ali et al. (2018) which revealed the
investors can easily get an abnormal return by short-term reversal seen on Mondays was due to
buying on Monday and then selling on Friday. investor sentiment.
Birru (2018) mentioned three hypotheses that Speculative stocks are stocks that are traded
cause day-of-the-week effects, including the in small volumes, they are volatile, unprofitable,
possibility that institutional traders trade at a non-dividend paying, lottery-like, potentially
particular time so that the return on that day is close to distress and have extreme growth;
higher. Second, there is a tendency for news stocks like these are difficult to judge properly
releases at certain times, causing abnormal (Birru, 2018; Asness et al., 2018; Chiang &
Journal of Indonesian Economy and Business, Vol. 35, No. 3, 2020 259

Zhang, 2019). Furthermore, Chiah and Zhong stocks, less speculative stocks, and non-
(2019), and Ali et al. (2018) have empirically speculative stocks. According to Birru (2018),
found that investors tend to be optimistic about Asness et al., (2018), and Chiang & Zhang
the futures of speculative stocks on a Friday, (2019) the day-of-the-week effect was more
rather than on other days. significant in a speculative index rather than a
This study is motivated by Chiang and less speculative index, or a non-speculative
Zhong (2019) who used a developed countries index. Hence, in this paper, we seek to answer
dataset. The primary purpose of this paper is to the following questions:
investigate the day-of-the-week phenomenon in 1. Is there a day-of-the-week effect anomaly in
developing countries such as Indonesia. three indexes in Indonesia?
Compared to previous studies, we found gaps in 2. Is the day-of-the-week effect more signi-
the research: First of all, and to the best of our ficant in a speculative index instead of less or
knowledge, this is the first study to investigate non-speculative indexes?
the hypothesis of investors’ moods in emerging
markets. Secondly, the previous studies which Our novelties from this study for the asset
examined the hypothesis only used one major pricing literature are two-fold. We postulate a
stock index for each country. We divided the method to decompose stock price indexes in
Jakarta Stock Exchange Composite index Indonesia (JKSE, LQ 45, and Kompas 100) into
(JKSE), as the stock market in Indonesia, into speculative, less speculative, and non-specula-
three indexes consisting of speculative, less tive indexes. Conventionally, to identify the
speculative, and non-speculative. Since the speculative stocks, we would need to list all of
effect of the investors’ moods will be different the speculative stocks manually and create a
for each index, we consider that our findings will value-weighted index, as suggested by Chiah
contribute to the existing literature. and Zhong (2019). This work requires a lot of
We chose to use data from the Indonesian effort to identify the speculative and non-
stock market when we became aware of the speculative stocks. Hence, to make the work
uniqueness of this index. This decision was more efficient, we use the characteristics of
based on research by Stambaugh, Yu, and Yuan indexes in Indonesia based on the performance
(2015)and Jacobs (2016), who all revealed that of the companies. We believe that our method is
the Indonesian stock market had a greater more efficient for creating the speculative index,
mispricing condition compared to other markets compared to the previous research papers.
in Southeast Asia. This mispricing condition Secondly, we test the anomaly of the day-of-the-
meant the existing information was not incor- week effect at the mean and variance levels
porated perfectly in the stock prices. Secondly, simultaneously, and our results will be robust for
the proportion of foreign investors in Indonesia time-series stylized facts (heteroscedasticity and
was greater, at around 50.64%, compared to serial correlation).
domestic investors. This causes the majority of The rest of this study proceeds as follows.
the stock trading to be dominated by foreign Section 2 describes the literature review. Section
investors who have some preferences and 3 explains the methodology, Section 4 presents
characteristics in trading. To re-examine the the empirical findings, and Section 5 offers the
findings from Chiang and Zhong (2019), we conclusion.
260 Luxianto, et al

LITERATURE REVIEW mean level instead of the variance level using

1. Day-of-The-Week-Effect the S&P 500’s dataset. French’s study revealed
that the returns on Mondays were less than on
One theory dealing with financial markets that
any other trading day. Furthermore, the study by
has been the subject of much research is the Gibbons and Hess confirmed the findings of
efficient market hypothesis (EMH). Fama (1965) French (1980) that Monday’s returns were the
stated that a market would move based on every lowest compared to those on the other trading
new additional piece of information. In addition, days. In contrast, the returns on Wednesdays and
the price of an asset or stock on a market already Fridays were the highest.
reflected the historical information related to it.
Rogalski (1984) continued the work using
The further implication is that no investor will
the Dow Jones Industrial Average Index (DJIA)
be able to gain greater returns than the market,
and added a more detailed examination. Instead
using the information that is already available.
of studying the close-to-close price (Friday close
Given this proposition, many researchers had to Monday close price) as done by previous
challenged the EMH by identifying the source of researchers, he differentiated between the non-
abnormal returns that can beat the market. One trading days, which refers to the Friday close to
of the popular topics related to this is the Monday open and the trading day, which refers
calendar effect, where the researcher tries to find to the Monday open to Monday close. Rogalski
out whether the information in the calendar can (1984) found that what contributed to the low
generate abnormal returns, compared to the return on a Monday was the non-trading day
market. The scope of the calendar effect is the portion, while the trading portion on Monday is
weekday effect, January effect, and other special as normal as on any other trading day. This split
events related to the calendar effect. of trading and non-trading days inspired
The weekday effect was first studied by Smirlock and Stark (1986) to do a more detailed
Fama (1965), who found abnormalities during analysis. They examined the day-of-the-week-
trading days in the Dow-Jones Industrial effect using an hourly analysis, and also split the
Average from January, 1956 to April, 1958.His time frame into three sub-periods (1963 to 1968,
research disseminated abnormalities in variance, 1968 to 1974, and 1974 to 1983). They found
and the hypothesis was that the variance on a that the hourly Monday return was indeed lower
Monday would be three times higher than that than that on any other day, but only before 1974.
on any other day, since it would sum up the After 1974 there has been no difference between
variability of information during non-trading the hourly return of Monday compared to the
days – Saturday and Sunday. This finding was other trading days.
interesting because the variance’s return on The research into the Monday effect later
Mondays was 20% higher than on any other day became very popular in other countries. Jaffe
– later on, the abnormality always happened on and Westerfield (1985) examined a similar topic
Monday, and it became known as the Monday using datasets from the US, Canada, the UK,
effect. Japan, and Australia, and they empirically found
Many research papers have followed Fama’s that the results from the US, Canada, and the UK
(1965) ideas to find any abnormality in a trading were the same as those from the previous
day. One of the initial followers was French studies: the lowest returns were on Mondays, but
(1980), who examined the abnormality in the in Japan and Australia the lowest returns were
Journal of Indonesian Economy and Business, Vol. 35, No. 3, 2020 261

on Tuesdays. Condoyani (1987) extended the stocks have been included in the top 60
international day-of-the-week-effect to more companies with the highest transaction value on
countries by adding France and Singapore to the a regular market over the last 12 months.
sample. Consistent with Jaffe and Westerfield The second index is the Kompas 100,
(1985), in Canada, the effect was found on consisting of 100 companies with good funda-
Mondays, while in Australia, France, Singapore, mental performance, high liquidity, and large
Japan, and the UK, the effect was seen on market capitalization in Indonesia. But, in
Tuesdays. general, the performance of companies listed on
From previous research papers, the interna- the Kompas 100 index is below the LQ 45 index.
tional results of the Monday effect have As a consequence, if a company is expected to
diverged, since some countries have different have high performance but is not among the top
weekday effect anomalies. Then, the question is: companies for 12 consecutive months, then the
Is the weekday anomaly just a statistical artifact company will be listed on the Kompas 100
and not consistent with the asset pricing theory? index. The third index is a composite index in
Studies by Smirlock and Stack (1986), Mehdian Indonesia known as the JKSE. All shares listed
and Perry (2001), and Kohers et al. (2004) on the Indonesian capital market,which consists
answer this question; they found that the result of speculative and non-speculative shares are in
of the Monday effect reduces over time. Even this index. Therefore, based on the characte-
Kohers et al. (2004) stated that the effect ristics of the indexes in Indonesia, in the form of
disappears over time, especially in developed classifying stock performance, we can make an
markets. After 1980 the effect fades away, index for speculative stocks with this
meaning that the markets were getting more assumption.
efficient.
HYPOTHESES AND RESEARCH
2. Indexes in Indonesia METHOD
The capital market in Indonesia has unique In this section, we explain the data source used,
characteristics compared to the other developing the method to decompose the return of the index
countries in the Southeast Asia region. These into speculative, less speculative, and non
characteristics include the stock index being speculative stock index returns, and also the
based on the performance of individual shares. regression model using a dummy variable to get
There are two main indexes, namely LQ 45, and the evidence of the presence of the day-of-the-
Kompas 100, and one composite share price week effect.
index, JKSE. Each index has certain provisions
that are regulated by the regulator. 1. The Model of the Day-of-Week Effect
The first index, LQ 45, is an index consisting We used the daily returns data from January
of 45 individual stocks that have the following 2000 to July 2019 from the three main indexes in
characteristics: Firstly, the individual stock must Indonesia, namely; Jakarta Stock Exchange
have good financial conditions, the prospect of Composite index (JKSE), the LQ45 index, and
growth, a high transaction value and frequency. the Compass 100 index. JKSE represents the
Secondly, the stocks have been included in the performance of all types of stock traded on the
top 60 companies, with the largest market Indonesian stock exchange. The LQ45 index
capitalization in the last 12 months. Lastly, the represents the 45 most liquid stocks on the
262 Luxianto, et al

market, and the Kompas 100 index represents used the LQ45 as a proxy for the non-
the 100 most liquid index in the market. Daily speculative index’s return because the index lists
returns are calculated using the continuously the 45 most liquid stocks in Indonesia. Below is
compounded method with the natural logarith- the formula to calculate the return of the NS
mic first difference of the closing price: index returns:
rt = lnPt – ln Pt-1 𝐿𝑄45𝑡
(1) 𝑁𝑆𝑡 = ( ) −1 (4)
𝐿𝑄45𝑡−1
Where Pt is the closing stock price of each
In preparing the stocks included in the
index at time t. Financial data have a stylized
speculative index, first, we defined the
fact that there is the possibility of serial
speculative stocks. Speculative stocks are stocks
correlation and heteroscedasticity. To overcome
that are not included in the LQ45 index. We
this, the study used the standard Ordinary Least
divided all of the stocks that were not included
Square methodology by assuming the data was
into two groups, so we have two speculative
non-normal and volatility clustering occurred.
index groups. To form the first group’s
Next, we added one lag value from the daily
speculative stock index, we used the Kompas
return into the regression model along with five
100 index, which contains the 100 best stocks on
dummy variables to measure the parameters of
the Indonesian stock market. The first
each week-day. Here is the regression model:
speculative stock index had stocks that were not
rt = α + γ rt-1 +∑5𝑗=2 β𝑗𝐷𝑎𝑦 + εt (2) included in the LQ45 index but were included in
the Kompas 100 index. As for the second
Where rt is the daily return of each index, α is
group’s speculative stock index, we included all
the constant term, and Monday is the base
the stocks that were not included in the Kompas
condition of the dummy variable. Meanwhile, to
100 index. Therefore, the speculative stocks in
control volatility clustering, this study used the
the first group had a lower level of speculation
GARCH (p,q) model by incorporating weekday
than the second group. We called the first
dummies into the conditional variance equation
group’s speculative stock index S1, and for the
with the following models in Equation 3. The
second group it was S2. We used the following
two models above were joint estimation
formula to estimate S1’s returns:
processes, which mean they were estimated
simultaneously. The advantage of the two [(RetKompas100t * Kompas100Vol .) 
models above is that they can see the weekdays ( RetLQ 45t * LQ 45Vol .)]
S1t  (5)
effect that occurs at the mean level and at the ( Kompas100Vol .  LQ 45Vol .)
variance level.
Where RetKompas100t is the return of the
σ2t = η + λ (L) ω2t + θ(L) σ 2t-1 + Kompas 100 index, RetLQ45t is the return of the
∑5𝑗=2 β𝑗𝐷𝑎𝑦 (3) LQ45 index, LQ45Volt is the volume of LQ45
index, and the Kompas100Volt is the volume of
2. Speculative vs. Non-Speculative Kompas100 index. We used volume data from
Decomposition the Kompas 100 index and LQ45 index as
proxies to estimate the weight component in
We identified the non-speculative index return
Equation (5).
(NS) as the return of the stocks on the most
liquid index on the Indonesian stock market. We For S2’s returns, we used the returns of the
Jakarta Composite Index (JKSE) as a benchmark
Journal of Indonesian Economy and Business, Vol. 35, No. 3, 2020 263

because the index contained all the stocks caused by systematic risk’s composition, with a
available on the Indonesian stock market. beta value of 1.19.
Therefore, S2’s stocks are all the stocks that
were included on the Jakarta Composite Index, Table 1. Index return before decomposition
but excluded from the LQ45 and Kompas 100 LQ45 K100 IHSG
indexes. We used the following equation to Mean 0.0373% 0.0377% 0.0422%
estimate S2’s returns: Standard Deviation 1.51% 1.44% 1.26%
Min. 10.30% 9.13% 7.92%
[(RetIHSGt * IHSG Vol .) 
Max. -11.86% -11.16% -10.38%
( RetKompas100t * Kompas100Vol .)] Beta 1.19 1.13 1.00
S1t  (6)
( IHSG Vol .  Kompas100Vol .)
Figure 1 reflects an upward and downward
Where RetIHSGt is the return of the Jakarta trend in the three indexes in Indonesia. Overall,
Composite Index, and IHSGVolt is the volume stock price trends in all three indices rose with
of the Jakarta Composite Index. the highest share prices in 2018, and the lowest
We defined the traditional Monday effect share prices were seen from 2008 to 2009. In
accusing the returns on Mondays to be lower 2008 and 2009, there was a global crisis, better
than on any other trading day of the week. There known as the subprime mortgage crisis in
was also the possibility of other weekday effects America, resulting in a decline in share prices in
(such as the Wednesday effect). Therefore we Indonesia. Two indexes, the Kompas 100 and
used 2-tailed testing using a notation Bj(1 = LQ 45 experienced relatively similar price
Monday, 2 = Tuesday, etc.) following Doyle and movements for the entire observation period.
Chen (2009). The null hypothesis was: Besides, these indexes have a pattern of
H0= B1 +B2 +B3 + B4 +B5= 0 (null for Monday movement that is almost the same as the IHSG,
effect) where there was a sharp decline in 2008 and
2016, and an increasing pattern in 2013 and
H1 = B1 +B2 +B3 + B4 +B5 ≠ 0 (for Monday or
2017.
weekday effect)
Once we had done the decomposition
process for the three major indexes in the capital
RESULTS AND DISCUSSIONS
market, we got three indexes consisting of
1. Descriptive Statistics speculative, less speculative, and non-specula-
Table 1 outlines the descriptive statistics of the tive indexes. Recalling that a speculative index
index returns before their decomposition into the is a stock price index consisting of stocks in
new speculative index. Indonesia that have the characteristics of a small
Table 1 illustrates the highest average return trading volume, volatility, unprofitable, non-
from January 2000 to July 2019 was from the dividend paying, lottery-like, potentially close to
IHSG with a value of 0.0422% while the average distress, and having extreme growth, so that
return for other indexes was similar, at around shares like this make it difficult to judge them
0.037%. The total risk measured by the standard properly. Table 2 illustrates the descriptive
deviation showed the LQ 45 index was the statistics of the three indexes. On average, the
riskiest, with a value of 1.51% compared to the highest return was from the speculative stock
Kompas 100 index, which was 0.10% lower. price index, while the lowest return was from the
The risk in the LQ 45 index may have been non-speculative index. Interestingly, the lowest
264 Luxianto, et al

standard deviation was in the speculative index, figure above, it can be seen that the price
with a value of 0.84% compared to the other movements (trends) in the non-speculative
three indexes. This seems to contrast with the stocks were relatively stable compared to the
risk-return trade-off in asset pricing, but this less speculative and speculative stocks. The
might be because the speculative index in speculative index experienced an increase in
Indonesia consists of less attractive stocks; stock pricesdue to price movements from 2009
hence investors rarely deal in these stocks. In to 2012, and reached a peak in 2018. On the
Table 3, we segregate the sample period into other hand, the non speculative index expe-
during the crisis and after the crisis, and we find rienced a decline in 2018.
the consistent pattern of the speculative index
has the highest return and the lowest risk during Table 3. Descriptive Statistics Decomposition
the two periods. Index During and After the Crisis Period
Crisis 2007-2010
Table 2. Descriptive Statistics Index Non- NS S1 S2
Speculative (NS), Less Speculative Mean 0.05% 0.07% 0.07%
(S1), and Speculative (S2) Median 0.00% 0.04% 0.04%
Maximum 10.30% 8.10% 4.70%
NS S1 S2
Minimum -11.86% -7.79% -8.72%
Mean 0.0373% 0.0374% 0.0545%
Std. Dev. 2.16% 1.66% 1.28%
Standard Deviation 1.51% 1.23% 0.84%
After Crisis (2010-2019)
Min. 10.30% 8.10% 4.70%
NS S1 S2
Max. -11.86% -7.79% -8.72%
Mean 0.03% 0.03% 0.05%
Beta 1.19 0.80 0.57
Median 0.01% 0.00% 0.04%
Figure 2 depicts the movement of the three Maximum 6.19% 4.90% 3.40%
Minimum -10.13% -7.12% -6.15%
indexes resulting from the decomposition
Std. Dev. 1.22% 1.05% 0.62%
process into speculative stocks. Based on the

350

300

250

200
LQ45
150
K100
100 IHSG
50

Figure 1. Trend index before decomposition

Journal of Indonesian Economy and Business, Vol. 35, No. 3, 2020 265

600
500
400
300 NON

200 SPEC1

100 SPEC2

Figure 2. Trend index after decomposition

2. Empirical Analysis and Findings Wednesdays, the highest and most significant
In this section, we investigate the day-of-the- average returns with the highest average returns
week effect on the three indexes (speculative, are from the less speculative stocks with a value
less speculative, and non-speculative) using a at 0.19%, as opposed to the lowest return from
dummy regression for the mean equation, and to the speculative indexes at about 0.14%. The
control the heteroscedasticity we use the average return value then decreased gradually to
GARCH model. We only show the regression reach a negative value on Mondays, while the
result from the mean level to save space. less speculative index was the only index to have
However, the full estimation can be shown by a significant negative return on that day. On a
request. specific day such as Friday, the speculativeS2
index experienced the highest positive and
Table 4 provides the estimation results for
significant returns of 0.09% compared to less
the three indexes from January to July 2019.
speculative S1index and the non-speculative
Overall, we found that the weekday effect
(NS) at 0.08% and 0.05% consecutively.
anomaly does exist in Indonesia. The three
indexes experienced a similar pattern namely, on

Table 4. Estimation Result Mean Level Regression Full Sample Period

Day NS. S1 S2
Monday -0.0888% -0.1733% *** -0.0492%
(0.203) (2.874) (1.042)
Tuesday 0.0254% -0.0016% 0.0193%
(0.357) (0.479) (0.564)
Wednesday 0.1802% *** 0.1941% *** 0.1460% ***
(3.127) (4.501) (3.571)
Thursday 0.0172% 0.0856% * 0.0659% **
(1.009) (1.982) (2.493)
Friday 0.0526% 0.0825% * 0.0907% ***
(0.305) (1.194) (3.007)
R-Square 0.004 0.011 0.010
F-Stat 2.402 6.746 6.545
Sig. 0.035 0.000 0.000
Notes t-statistics is reported in parentheses.
*** Significant at 1% level; ** Significant at 5% level; * Significant at 1% level
266 Luxianto, et al

To answer the second question, as well as to significant, compared to the other indexes,
check the consistency of our results, we where the lowest returns were on Mondays at -
conducted a robustness check by splitting the 0.01% and the highest return was on
estimation period into during and after the global Wednesdays at 0.01%.
financial crisis (GFC) in Table 5. During the Our findings support the mood hypothesis
GFC period, we found a significant Wednesday from Birru (2018) and Chiah and Zhong (2019)
effect for all the indexes (NS, S1, and S2)and the that investors' moods tend to rise on Fridays and
highest return, of 0.25% for the non-speculative fall on Mondays. Investors who are in a good
and speculative indexes. For the speculative mood tend to buy speculative stocks on a Friday,
index, we found the return on Tuesdays was the causing the returns to be positive and significant.
lowest at 0.16% and significant at 95%. After Investors make these transactions only due to
the GFC period, we found consistent results for their psychological impulses, without consider-
the Wednesday effect in all the indexes. For the ing other relevant information such as the
S2 index, we found that all the days were company’s fundamentals.

Table 5. Estimation result for the mean level regression during and after crisis period
During Crisis (2007 - 2010)
Day NS. S1 S2
Monday 0.09% -0.01% 0.01%
0.6344 0.3877 0.9546
Tuesday 0.12% 0.00% 0.16% **
0.3544 0.2048 0.0275
Wednesday 0.25% * 0.00% * 0.25% ***
0.0733 0.0545 0.0012
Thursday 0.11% 0.00% 0.12%
0.4242 0.1941 0.1051
Friday 0.15% 0.00% 0.19% ***
0.2101 0.0814 0.0036
After Crisis (2010 - 2019)
Day NS. S1 S2
Monday -0.00% -0.01% *** -0.01% **
0.1785 0.0002 0.0607
Tuesday 0.00% 0.04% 0.00% *
0.1169 0.4616 0.0858
Wednesday 0.00% *** 0.01% *** 0.01% ***
0.0000 0.0000 0.0000
Thursday 0.00% 0.01% ** 0.00% **
0.8591 0.015 0.0057
Friday 0.00% 0.00% ** 0.00% ***
0.4478 0.0613 0.0002
Notes t-statistics is reported in parentheses.
*** Significant at 1% level
** Significant at 5% level
* Significant at 1% level
Journal of Indonesian Economy and Business, Vol. 35, No. 3, 2020 267

Empirically, we found that the highest conditions such as the subprime mortgage crisis,
investor mood in Indonesia is on Wednesdays or the up and downturn in macroeconomic
and gradually decreases to the lowest level on conditions, both of which affected the funda-
Mondays. The high investor mood can be seen mental value of stocks. Therefore we divided the
from the highest average return estimation estimation period according to the method used
results in Table 3. Interestingly, we empirically by Zhang, Lai, and Lin (2017) with three
found that the psychological mood of the periods, namely the high upside period, the
investors’ moods was more significant in the less normal period, and the low upside period based
speculative index, which tended to be more on the rise and fall of the stock index.
stable in performance than the speculative index. Table 6 shows the results of the estimated
We may argue that investors in Indonesia are not weekday effects with three time periods. The
completely irrational by using their mood when highest average return consistently occurred on
trading. Rationality and psychology can play a Wednesdays for the whole index. Meanwhile,
role for Indonesian investors since they are more the speculative index was very sensitive to the
active in trading less speculative stocks when division of this period, where during the upside
their mood is good, to avoid higher risk. Our period, the return value became positive and
findings are the same as the findings from significant for the whole of every day (except
Boubaker et al. (2017) and Anwar and Mulyadi Mondays), with the highest returns on
(2012), who examined the weekday effects of Wednesdays at 0.28% and the lowest returns on
the global major capital markets. Their results Mondays at 0.14%. In contrast, during the
show that there is a positive significant normal period, the speculative index experienced
Wednesday return. the highest returns on Wednesdays and the
The estimation results from our previous poorest returns on Mondays.
findings might have been influenced by global

Table 6. Estimation result mean level in three periods

High Upside Period (2009 2012) Normal period (2012 2015) Low Upside Period (2016 2019)
Day - - - -
NON SPEC. SPEC. 1 SPEC. 2 NON SPEC. SPEC. 1 SPEC. 2 NON SPEC. SPEC. 1 SPEC. 2
Monday -0.0462% -0.1391% 0.0149% -0.2252% ** -0.2807% *** -0.1546% *** 0.0438% -0.0863% 0.0269%
1.644 1.129 0.914 2.039 3.001 5.002 1.557 1.187 1.391
Tuesday 0.1156% 0.1927% ** 0.1634% ** 0.0698% -0.1234% -0.0005% 0.0193% -0.0719% -0.0111%
1.249 4.036 2.010 1.521 1.155 0.992 1.795 1.271 1.724
Wednesday 0.3469% *** 0.3648% *** 0.2808% *** 0.1851% * 0.1107% 0.1185% ** 0.0773% 0.2241% *** 0.0947% ***
3.001 4.000 5.000 1.989 1.202 2.016 1.301 3.648 3.003
Thursday -0.0138% 0.1282% 0.1454% ** 0.0564% 0.0845% 0.0194% 0.0293% 0.1125% * 0.0630% **
0.890 1.161 2.022 1.604 1.330 0.916 1.695 1.986 2.045
Friday 0.1313% 0.1486% 0.1362% ** 0.0066% 0.0708% 0.0686% -0.0037% 0.0296% 0.0770% **
1.190 1.105 2.032 0.952 1.416 1.163 0.961 1.650 2.014
R-Square 0.014 0.025 0.034 0.010 0.020 0.022 0.002 0.019 0.021
F-Stat 3.049 *** 5.432 *** 7.214 *** 1.567 3.134 *** 3.570 *** 0.329 3.567 *** 4.015 ***
Sig. 0.010 0.000 0.000 0.167 0.008 0.003 0.896 0.003 0.001

Notes t-statistics is reported in parentheses.

*** Significant at 1% level
** Significant at 5% level
* Significant at 1% level
268 Luxianto, et al

Using these three periods, we show that the (Indonesia). Instead of manually listing indi-
mood hypothesis from Birru (2018) and Chiah vidual speculative stocks, we postulated a new
and Zhong (2019) is supported in Indonesia. We method to decompose indexes into speculative,
may argue that during normal conditions, less speculative, and non-speculative indexes,
investors tend to only use the pure behavioral We empirically find that the mood of the
mood to trade on Wednesdays when the mood is investors is highest on Wednesdays, while the
good for speculative stocks, because they are mood is lowest on Mondays. Using the full
considered to have good prospects. This happens sample period, we reveal that less speculative
since the investors experience indifference stocks are supported by the behavior of the
between gaining the return and avoid risk. mood investor hypothesis. To get clearer results,
Conversely, on Mondays, investors tend to sell we divided our sample period into three testing
speculative stocks when the mood is bad. When periods (upside, normal, and low side). We
the index conditions are good, and the stock found that the behavior mood hypothesis is
price rises significantly (high-upside period), supported only in normal periods when investors
and when the stock price only rises steadily (low tend to be irrational and use their good mood to
upside period), investors tend to be more trade in speculative stocks on a Wednesday and
rational. sell them on a Monday.
Besides, we find that investors in Indonesia
CONCLUSION
are not completely irrational by using their mood
Research related to the anomaly of the day-of- when trading in the high-upside and low-upside
the-week effect, which began in 1930 with a periods. Rationality and psychology play a role
study by Kelly and is still ongoing, as shown by for Indonesian investors since they are more
the recent studies from Richards and Willows active in trading less speculative stocks to avoid
(2019), found that the returns on Mondays are greater risk and earn a positive return in those
always negative compared to other days, so this periods when the mood is good.
phenomenon is called the Monday effect The research’s implication for investors is
anomaly. This anomaly can generally only be that they should buy stocks on Mondays when
explained by the existence of persistent seasonal the prices are very low and sell on Wednesdays
ties in the capital market. when the average investors are in a good mood
Birru (2018) and Chiah and Zhong (2019) to buy shares, which causes the stock prices to
suggested the hypothesis that the Monday effect rise significantly. In this study, we have a
could be influenced by the psychological bias limitation, in that the weekday effect is static
from investors. They argued that the investor (does not vary overtime). However, in the real
mood hypothesis is a condition where an case, investors have mood swings that result in
investor’s mood will rise on a Friday and fall on dynamic weekday effects (wandering effects).
a Monday. The consequence of this hypothesis is For future research, we suggest using the
that investors will buy stocks with good dynamic weekday effect model to study the
prospects when the mood is good and sell them psychological mood of the investors.
when the mood is bad.
The motivation for our research was to REFERENCES
examine the mood behavior of investors using a Ali, F., & Ülkü, N. (2018). Monday Effect in the
highly mispriced capital market as a sample RMW and the Short‐Term Reversal Factors.
Journal of Indonesian Economy and Business, Vol. 35, No. 3, 2020 269

International Review of Finance, 19(3), Jacobs, H. (2016). Market maturity and

681-691. mispricing. Journal of Financial Economics,
Asness, C. S., Frazzini, A., & Pedersen, L. H. 122(2), 270-287.
(2019). Quality minus junk. Review of Jaffe, J., & Westerfield R., (1985). The week-
Accounting Studies, 24(1), 34-112. end effect in common stock returns: the
Baker, M., & Wurgler, J. (2006). Investor international evidence. The Journal of
sentiment and the cross‐section of stock Finance. Vol. 40, No.2, pp. 433-454.
returns. The journal of Finance, 61(4), Kelly, F. (1930). Why you win or lose: the
1645-1680. psychology of speculation. Boston:
Bagozzi, R. P., Gopinath, M., & Nyer, P. U. Houghton Mifflin.
(1999). The role of emotions in marketing. Mehdian, S. & Perry M.J. (2001). The reversal
Journal of the academy of marketing of the Monday effect: new evidence from
science, 27(2), 184-206. US equity markets.Journal of Business
Birru, J. (2018). Day of the week and the cross- Finance and Accounting, 28(7).
section of returns. Journal of financial Wright, F. W., & Bower, G.H (1992). Mood
economics, 130(1), 182-214. effects on subjective probability assessment.
Chiah, M., & Zhong, A. (2019). Day-of-the- Organizational Behavior and Human
week effect in anomaly returns: Decision Processes, 52, 276-291.
International evidence. Economics Letters, Richards, D. W., & Willows, G. D. (2019).
182, 90-92. Monday mornings: Individual investor
Condoyani, L., and J. O’Hanlon (1987). Day of trading on days of the week and times
the week effect on stock returns: the within a day. Journal of Behavioral and
international evidence. Journal of Business Experimental Finance, 22, 105-115.
Finance and Accounting, 14, 159 – 174. Rogalski, R. J. (1984). New findings regarding
Doyle, J. R., & Chen, C. H. (2009). The day-of-the-week returns over trading and
wandering weekday effect in major stock non-trading periods: a note. The Journal of
markets. Journal of Banking & Finance, Finance, Volume 39, Issue 5.
33(8), 1388-1399. Smirlock, M., & Starks, L. (1986). Day-of-the-
Fama, E. F. (1970). Efficient capital markets: A week and intraday effects in stock returns.
review of theory and empirical work. The Journal of Finance Economics, Volume 17,
journal of Finance, 25(2), 383-417. Issue 1, 197-210.

French, K. (1980). Stock returns and the Stambaugh, R. F., Yu, J., & Yuan, Y. (2015).
weekend effect.Journal of Financial Arbitrage asymmetry and the idiosyncratic
Economics, volume 8, issue 1, 55-69. volatility puzzle. The Journal of Finance,
70(5), 1903-1948.
G. Kohers, N. Kohers, V. Pandey & T. Kohers
(2004). The disappearing day-of-the-week Zhang, J., Lai, Y., & Lin, J. (2017). The day-of-
effect in the world’s largest equity markets. the-week effects of stock markets in
Applied Economics Letters. 11:3, 167-171, different countries. Finance Research
DOI:10.1080/1350485042000203797. Letters, 20, 47-62.