Decision (September 2014) 41(3):327–337

DOI 10.1007/s40622-014-0050-4

RESEARCH PAPER

Settlement cycle and day of the week anomaly: empirical

evidence from Indian stock market
Samveg A. Patel • M. Mallikarjun

Published online: 25 May 2014

Ó Indian Institute of Management Calcutta 2014

Abstract Indian stock market was functioning with both return and volatility of Indian stock market.
Accounting Period settlement cycle till December 31, Empirical result reveals that before introduction of
2001. But as per the recommendation of G-30 which T?2 settlement, returns of Monday and Thursday
was a group to determine the best international were negative, while after the introduction of T?2
practices for securities clearing and settlements, settlement, returns of all days are positive. By
Indian stock market has adopted T?2 rolling settle- applying GARCH model, this study concludes that
ment cycle on April 1, 2003. Difference of settlement after the introduction of T?2 settlement, day of the
and clearing cycles of different stock exchanges might week effect exists only in return of Indian stock
lead to the day of the week anomaly. Day anomaly in market, but it does not exist in volatility of Indian
return and volatility of stock market was found to be stock market.
affected by settlement cycle, and it changes over the
period of time. Therefore, present study tries to Keywords Day of the week effect  Settlement
investigate the effect of introduction of T?2 settle- cycle  GARCH model
ment cycle on day of the week anomaly in return and
volatility of Indian stock market. Study analyzes daily
closing price data of Sensex and Nifty over the period
January 2, 1991 to January 31, 2013 covering 5,763 Introduction
samples. Study computes descriptive statistics of each
week days for both subperiods as well as for complete Indian stock market was functioning with Accounting
sample period. It also applies GARCH (1, 1) model for Period settlement cycle till December 31, 2001. In
investigating the presence of day of the week effect in accounting period settlement method, trades of whole
week were cumulated and settled on one particular day
of next week. However, rolling settlement was
S. A. Patel (&) recommended by G-30 which was a group to deter-
Goa Institute of Management, Sanquelim Campus, mine the best international practices for securities
Poriem, Sattari 403505, Goa, India
e-mail: [email protected]; [email protected] clearing and settlements. A majority of developed
world stock markets are operating on rolling settle-
M. Mallikarjun ment basis. In rolling settlement, trades of a particular
Institute of Management, Nirma University, Sarkhej- day are settled independently from the trades of any
Gandhinagar Highway, Ahmedabad 382481, Gujarat,
India other day of week. Furthermore, in case of rolling
e-mail: [email protected] settlement, buyer and seller get their security and

123
328 Decision (September 2014) 41(3):327–337

money earlier than accounting period settlement. 2004; Holden et al. 2005). Similarly, change in
Hence, settlement risk is considerably reduced in security settlement cycle from accounting period to
rolling settlement. Initially, Indian stock market has rolling period settlement also affects the day of the
adopted T?5 days rolling settlement cycle. This was week anomaly pattern of stock market (Keef and
further reduced to T?3 and T?2 on April 1, 2002 and McGuinness 2001 for New Zealand; Draper and
April 1, 2003, respectively. Difference of settlement Paudyal 2002 for UK; Chander et al. 2008 for India).
and clearing cycles of different stock exchanges might Hence, the main objective of present study is to
lead to the day of the week anomaly. investigate the effect of introduction of T?2 settle-
Day of the week anomaly means regular pattern in ment cycle on day of the week anomaly in return and
stock market return across all week days (Islam and volatility of Indian stock market. With this back-
Watanapalachaikul 2005). Day of the week anomaly ground, remainder of the paper is organized in the
indicates that average returns and volatility of all days following sections. Second section provides Review
are different from each other. The presence of day of of the Literature. Third section discusses Data and
the week anomaly provides abnormal return on a Methodology. Empirical Analysis is presented in
particular day of the week. The term ‘‘day of the week fourth section. Findings and conclusion of the study
effect’’ in stock market was documented in early are pinpointed in fifth section.
nineties (Fields 1931; Kelly 1930). Testing stock
market seasonality had received considerable atten-
tion in post nineties (French 1980; Gibbons and Hess Review of literature
1981; Lakonishok and Levi 1982; Roll 1983; Lako-
nishok and Smidt 1990; Peiro 1994; Goswami and Previous studies related to day of the week anomaly
Angshuman 2000). Studies which had supported day are classified into two broad categories. First part
of the week effect had provided two contradicting majorly focuses on day of the week effect in return of
views. First, average return of Monday was higher stock market. While second more specifically spot-
than all other week days returns (French 1980). lighting on day of the week effect in both return and
Second, average return of Friday was higher than volatility stock market.
Monday return (Keim and Stambaugh 1984; Agrawal
and Tandon 1994; Mills and Coutts 1995). Further- Day of the week effect in return of stock markets
more, Monday effect was significant only in the last
two weeks of a month (Wang et al. 1997). Even returns US stock market had provided significant negative
on preholidays were significantly higher than week - return on Monday (Mehdian and Perry 2001; Gao and
days return (Arumugam 1998; Chong et al. 2005; Kling 2005; Chiaku 2006), while it had offered higher
McGuinness 2005). return on Friday (Cho et al. 2007). Similarly, Madure-
Later studies have included volatility of stock ira and Leal (2001) found negative Monday return in
market return to the day of the week anomaly. Higher Brazil stock market. Continuation of day of the week
return of a particular day might be a result of higher effect in recent times for US market was opposed by
variance on that particular day. Therefore, day of week Prokop (2010).
effect should be tested for both return and volatility of Day of the week effect was existed in European
the stock market (Ho and Cheung 1994; Choudhry stock markets (Apolinario et al. 2006); Swedish stock
2000). market (Thomas 2002); Istanbul stock exchange
Day of the week effect is a global phenomenon, and (Dicle and Hassan 2007). Weekend effect was existed
this might be the result of spillover effect of developed in UK stock market at the time of bad news, but at the
stock markets (Jaffe and Westerfield 1985; Wong et al. time of good news that effect was vanished (Steeley
1992; Agrawal and Tandon 1994). Seasonality of 2001). Ajayi et al. (2004) found mix evidence for
return and volatility of stock market were found to be Monday effect in European stock markets. Contradict-
change over the period of time (Bhattacharya et al. ing to this, day of the week effect was disappeared
2003; Doyle and Chen 2007). Stock market season- after 1990s from UK stock market (Steeley 2001);
ality was also affected by change in macroeconomic stock market of Germany (Alt et al. 2002; Prokop
variable like GDP and financial crisis (Rosenberg 2010); Turkish stock market (Demirer and Karan

123
Decision (September 2014) 41(3):327–337 329

(2002); and Spanish stock market (Gardeazabal and Tan and Wong (1998), Poshakwale and Murinde
Regulez 2004). A majority of European and Asian (2011), Fountas and Segredakis (2002), Alejandro
markets had demonstrated significant negative Mon- (2006), Wickremasinghe (2007), and Tsangarakis
day effect (Tong 2000). The argument that speculators (2007).
close their position on Friday and reopen it on Monday Day of week effect was existed in return of Indian
was denied in the study of Istanbul stock exchange stock market (Kaur 2004; Sarma 2004; Chia and Liew
(Aksoy and Dastan 2011). 2010), while it was rejected by Nageswari et al. (2011).
Day of the week effect was found in South Ascan Before the introduction of rolling settlement, Friday
(Alagidede 2008) and West African stock markets return was lowest, and Monday return was highest
(Tachiwou 2010). Ghana stock market has provided (Kumari and Mahendra 2006), while after introduction
negative return on Monday (Paul and Theodore 2006). of rolling settlement, Friday return was highest, and
Australian companies had provided significant Monday return of Indian stock market was lowest
positive return on Monday (Liu and Li 2010). Day of (Nath and Dalvi 2004; Gupta 2006; Chander et al.
the week effect was found in Australian stock market, 2008; Singhal and Bahure 2009). Similarly, Karmakar
but this effect was becoming insignificant in recent and Chakraborty (2000) found significant positive
years (Worthington 2010). Opposite to this, Marrett return on Friday. On the other hand, Amanulla and
and Worthington (2008) did not find day of the week Thiripalraju (2001) reported that Wednesday return
effect in Australian stock market. was significant. Within Indian stock market indices,
Day of the week anomaly was identified for Asian day of the week effect was found in Sensex and Nifty
stock markets (Foo and Kok 2000; Hui 2005; Islam only, while this effect was not detected in BSE 100,
and Watanapalachaikul 2005; Basher and Sadorsky BSE 500, and S&P CNX 500 (Padhi 2010). Day of the
2006; Lean et al. 2007; Anwar and Mulyadi 2009); week effect was also reported for IT companies of
Japanese stock market (Dhaoui et al. 2011); Malaysian Indian stock market (Debasish 2012).
stock market (Kok and Wong 2004; Lim and Dollery
2010; Muhammad and Rahman 2010); Taiwan stock
market (Huang and Liao 2010); Chinese stock market Day of week effect in both return and volatility
(Gao and Kling 2005; Cai et al. (2006); Pakistani stock of stock market
market (Zafar et al. 2010); Nepalese stock market
(Bahadur and Joshi 2005); and Dhaka Stock Exchange According to Capital Asset Pricing Model, risk and
(Chowdhury and Sharmin 2010) return are positively related with each other, and
Opposite to this, day of the week effect in five therefore for a particular day on which volatility is
ASEAN equity markets like Malaysia, Singapore, higher than other day volatility, return of that day
Thailand, Indonesia, and Philippines after Asian should also be higher than return of other week days
financial crisis was opposed by Chen and Liang (Mookerjee and Yu 1999). This risk and return
(2004); for Korea (Kamath and Chusanachoti 2002) relationship in stock market seasonality was chal-
and for Pakistan (Hussain 2000; Ullah and Usman lenged by Clare et al. (1998), Beller and Nofsinger
2010). (1998), and Lucey (2000).
Yakob et al. (2005) investigated stock market Day of the week effect in return and volatility was
seasonality in ten Asian countries and found that day recognized for US stock market (Berument and Dogan
of the week effect was existed in only five countries 2012); European stock markets (Högholm and Knif
including India, while month of the year effect was 2009; Savva et al. 2006); French stock market
found for eight countries. Similarly, Hui (2005) (Dimitrios and Katerina 2003); Hungarian and Polish
investigated day of the week effect in Singapore, stock markets (Poshakwale and Murinde 2011);
Korea, Hong Kong, and Taiwan. He found this effect Chilean stock market (Kamath and Liu 2011); Saudi
only in Singapore stock market. stock exchange (Ulussever et al. 2011); and Karachi
The presence of day of the week effect was also stock Exchange (Shaheen 2006). However, day of the
supported by Tong (2000) and Lim and Dollery week effect in volatility was refused for European
(2010). On the other hand, the presence of day of the stock markets (Apolinario et al. 2006) and Italian
week hypothesis was rejected by some studies such as stock market (Guidi 2010).

123
330 Decision (September 2014) 41(3):327–337

Day of the week effect in Turkey market was by 100, Rt = ln (Pt/Pt - 1) 9 100, where Pt and
present during 2003–2005, but that effect was disap- Pt - 1 are daily closing value of stock market indices
peared during 2005–2007 (Kamath and Liu 2011). at day t and t - 1, respectively.
Similarly, significant negative return on Monday was The presence of day of the week effect depends on
found for small and mid cap companies of New data period and methodology used for the study (Chen
Zealand, but Monday effect was absent in large cap et al. 2001). For the first approach, this study that
companies (Bin and Benjamin 2011). exercises to detect day of the week effect is to compute
Day of week effect was existed in both return and descriptive statistics of all individual weekdays for
volatility of stock market (Aggarwal and Rivoli 1989; both subperiods as well as for complete period. These
Yadav and Pope 1992; Hakan and Halil 2001; descriptive statistics help in comparing mean return of
Bhattacharya et al. 2003). But this effect was not all week days. If mean return varies across all days, it
contributing toward improvement in volatility fore- indicates that day of the week effect may exist.
casting (Charles 2010). Contradicting to this, some For the second approach, this study has three
studies had opposed day of the week effect in return available alternatives. As a first option, some studies
and volatility of stock markets (Balaban et al. 2001; had used nonparametric Kruskal–Wallis test for inves-
Kiymaz and Berument 2003). tigating day of the week anomaly (Hui 2005; Chander
Monday had a significant negative effect on stock et al. 2008; Lim et al. 2010). They argued that the error
market return of Indonesia, Malaysia, and Thailand, term resulting from application of OLS method might
but Monday effect was absent in both return and not be stationary because of heteroscedasticity problem.
volatility of Indian stock market (Choudhry 2000). Second, some studies had applied parametric test like t
Application of GARCH model to Indian stock market test and regression analysis (Agrawal and Tandon 1994;
revealed that Thursday and Friday were providing Choudhry 2000; Coutts et al. 2000; Al-Loughani and
significantly positive return, while OLS method found Chappel 2001). They had utilized only dummy vari-
that Monday was providing significantly higher return ables for week days as independent variables in
than all other week days (Bhattacharya et al. 2003). regression analysis. However, results of these studies
were challenged by Bhattacharya et al. (2003); Muham-
mad and Rahman (2010) and Guidi (2010). They have
Data and methodology added lagged values of stock return to mean equation of
stock return as shown in Eq. 1.
The aim of present study is to investigate the effect of
introduction of T?2 settlement cycle on day of the Rt ¼ a1 þ a2 D2t þ a3 D3t þ a4 D4t þ a5 D5t
week anomaly in return and volatility of Indian stock Xm

market. This study analyzes daily closing price data of þ ai Rti þ lt ; ð1Þ
i¼1
two leading Indian stock market indices namely
Sensex and Nifty over the period January 2, 1991 to where Rt is return of stock index, and D2t, D3t, D4t, and D5t
January 31, 2013 covering 5763 samples. Five obser- are dummy variables for Tuesday, Wednesday, Thurs-
vations per week are considered. Wherever, data day, and Friday, respectively. Here, D2t = 1, if t is
points that are missing are calculated by using linear Tuesday and 0 otherwise, similarly, D3t = 1, if t is
interpolation method in EViews. Rolling settlement of Wednesday and 0 otherwise, and so on. Dummy variable
T?2 types was introduced on April 1, 2003. There- for Monday is not included to avoid dummy variable trap.
fore, this study divides total sample period in two Constant a1 represents average Monday return.
subperiods. First, Period before the introduction of Furthermore, Stock market return has time varying
T?2 settlements is from January 2, 1991 to March 31, volatility, and therefore, application of normal OLS
2003. And second, Period after the introduction of method to stock return might provide misleading
T?2 settlements is from April 1, 2003 to January 31, information for day of week effect (Connolly 1989).
2013. Time series data for Sensex and Nifty are And therefore, third option is to apply Generalized
collected from the websites of BSE and NSE, respec- Autoregressive Conditional Heteroscedasticity
tively. Daily return is calculated by taking natural (GARCH) model (Bollerslev 1986) for accessing the
logarithmic first difference of the series and multiply it day of the week effect in both return and volatility of

123
Decision (September 2014) 41(3):327–337 331

stock market. The GARCH (p,q) model is shown in subperiods viz first before the introduction of T?2
Eq. 2. settlement and second after the introduction of T?2
settlement. Further, descriptive statistics for each
Yt ¼ a þ b0 Xt þ lt
individual week days are also calculated. Results of
lt j Xt  iid N ð0; ht Þ these descriptive statistics for Sensex and Nifty are
X p X
q ð2Þ
presented in Tables 1 and 2, respectively. Descriptive
h2t ¼ c0 þ di h2ti þ cj l2tj statistics reveals that mean return for both Sensex and
i¼1 j¼1
Nifty has been increased, while standard deviation has
To investigate the presence of day of the week been decreased after the introduction of T?2 settle-
effect in Indian stock market, the present study applies ment. Moreover, both Monday and Thursday have
GARCH (1,1) model and is demonstrated in Eq. 3. provided negative return before the introduction of
Most of the recent studies have used this methodology T?2 settlements, while returns of all week days are
(Choudhry 2000; Balaban et al. 2001; Yakob et al. positive after the introduction of T?2 settlement.
2005; Kamath and Liu 2011; Ulussever and Kar 2011; Tuesday and Thursday returns of both Sensex and
and Berument and Dogan 2012). Nifty are negatively skewed before the introduction of
T?2, while after introduction of T?2, Wednesday,
Rt ¼ a1 þ a2 D2t þ a3 D3t þ a4 D4t þ a5 D5t
Thursday, and Friday returns of both Sensex and Nifty
Xm
þ ai Rti þ lt are negatively skewed. Consequently, return of all
i¼1 weekdays after the introduction of T?2 becomes
lt j Xt  iid N ð0; ht Þ negatively skewed. Kurtosis values of both stock
market indices for both the time periods disclose that
h2t ¼ c0 þ d1 h2t1 þ c1 l2t1 ð3Þ
stock market return of whole week and return of each
Before applying GARCH model to stock market individual week days follow Leptokurtic distribution.
indices, it is essential to test stationary of these stock Similarly, Jarque–Bera statistics concludes that all
market indices. For the same, Augmented Dickey– week days returns of Sensex and Nifty for both the
Fuller (ADF) (Dickey and Fuller 1979, 1981) test has time period do not follow normal distribution.
been applied. It is based on simple logic that stationary Result of ADF unit root test is reported in
process possesses mean reversion characteristic, and Table 3. Here, ADF test statistic values are lower
variance of such series remains constant. This test is than the 1 % critical values of both Constant and
conducted by adding the lagged value of dependent Trend models. Hence, The null Hypotheses of
variable. Therefore, it behaves like AR (1) process Sensex and Nifty have unit root, can be rejected at
with q = 1. Dickey Fuller test is designed to examine level for both subperiods and for whole sample
if q = 1. The complete model with deterministic period. Thus, returns of both stock market indices
terms such as intercepts and trends is shown in Eq. (4). are stationary series, and they can safely be used for
X
m econometric modeling.
DRt ¼ a þ p þ dRt1 þ bi DRt1 þ et : ð4Þ For investigating day of the week anomaly, this
i¼1 study has applied GARCH (1, 1) model. Result of
The null hypothesis of series has a unit root is tested GARCH (1, 1) model for Sensex is presented in
for ADF unit root test against alternative hypothesis of Table 4. It indicates that for the whole sample period
no unit root. If ADF calculated value is smaller than and for the period of before the introduction of T?2
the test critical value, then null hypothesis of unit root settlement, Tuesday return is significantly different
that exists will be rejected. from all other days. While after the introduction of
T?2 settlement, Monday return is significantly dif-
ferent from all other days. Additionally, volatilities of
Empirical analysis Monday and Thursday are significantly different from
all other days’ volatility. However, before the intro-
First, this study presents descriptive statistics like duction of T?2 settlement, Thursday volatility is
Mean, Standard Deviation, Skewness, Kurtosis, Jar- significant, but after the introduction of T?2 settle-
que–Bera Statistic, and Probability Value for both ment, volatilities of all days are insignificant.

123
332 Decision (September 2014) 41(3):327–337

Table 1 Descriptive statistics of sensex

Before T?2 all days Mon Tue Wed Thu Fri

Mean 0.0349 -0.0291 0.1025 0.0986 -0.00515 0.00805

Standard deviation 1.6609 1.6088 1.5864 1.949 1.4793 1.6444
Skewness 0.49189 0.050124 -0.078155 1.452747 -0.459273 0.422297
Kurtosis 11.09518 4.74876 5.191029 18.14873 7.898817 7.672322
JB Statistics 8847.247 81.69111 128.2657 6324.866 661.422 600.2322
Prob. 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000
After T?2 all days Mon Tue Wed Thu Fri

Mean 0.073 0.0345 0.0881 0.108 0.00851 0.1261

Standard deviation 1.5711 1.8704 1.4261 1.4466 1.4506 1.6201
Skewness -0.126506 0.301259 0.118247 -0.28230 -0.235376 -0.73318
Kurtosis 11.68607 17.09885 7.386054 6.410624 5.795649 9.373479
JB statistics 8079.77 4256.63 413.2004 255.9535 172.1313 914.2398
Prob. 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

Table 2 Descriptive Statistics of Nifty

Before T?2 Mon Tue Wed Thu Fri

Mean 0.0351 -0.0554 0.0997 0.0158 -0.0641 0.1795

Standard Deviation 1.6126 1.5198 1.4989 1.9337 1.4455 1.6095
Skewness 0.657701 0.125377 -0.08411 1.82856 -0.738706 0.620129
Kurtosis 12.95776 4.960955 5.780681 19.76328 10.93399 8.632446
JB Statistics 13422.21 104.0564 206.2996 7825.653 1734.112 885.6189
Prob. 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000
After T?2 Mon Tue Wed Thu Fri

Mean 0.0709 0.0356 0.09 0.0919 0.011 0.1258

Standard Deviation 1.5942 1.9089 1.4411 1.4494 1.4815 1.6448
Skewness -0.294546 0.060274 0.130419 -0.17809 -0.352031 -1.12952
Kurtosis 12.89596 18.38778 7.483239 5.680194 5.346522 12.52433
JB Statistics 10515.65 5061.567 431.92 156.5625 128.5401 2048.07
Prob. 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

Table 3 Results of Augmented Dickey-Fuller Unit Root Test

Name of variable Before T?2 After T?2 Whole period
Constant Constant and trend Constant Constant and trend Constant Constant and trend

Sensex -48.75241* -48.817* -47.022* -47.061* -67.552* -67.55182*

Nifty -48.43334* -48.487* -47.235* -47.266* -67.548* -67.54644*
For constant model and constant and trend model, critical values at 1 % level of significance are -3.432377 and -3.961154
respectively
* Indicates ADF test value is significant at 1 % level of significance

123
Decision (September 2014) 41(3):327–337 333

Table 4 Result of GARCH (1, 1) model for sensex

Before T?2 After T?2 Whole period
Coefficient SE Coefficient SE Coefficient SE

Constant -0.093837 0.060616 0.126759* 0.055674 0.00852 0.042585

Tue 0.284099* 0.084021 -0.004512 0.073732 0.154617* 0.057499
Wed 0.132818 0.088087 0.002199 0.071737 0.075256 0.057921
Thu 0.091804 0.077146 -0.040466 0.074793 0.030034 0.05431
Fri 0.119076 0.079738 0.034221 0.075128 0.07603 0.056022
Sensex(-1) 0.159583* 0.021193 0.075008* 0.024122 0.125181* 0.016274
C 0.316694 0.17431 0.191038 0.143877 0.238091* 0.112646
RESID(-1)^2 0.142237* 0.020108 0.133371* 0.02256 0.133969* 0.015121
GARCH (-1) 0.809965* 0.021391 0.850596* 0.019938 0.845565* 0.014274
Tue -0.05384 0.334448 -0.462899 0.246295 -0.287454 0.207441
Wed 0.286276 0.366596 -0.124292 0.176307 -0.025596 0.187498
Thu -1.170645* 0.295067 -0.087399 0.186138 -0.540725* 0.165574
Fri 0.096011 0.242779 -0.042353 0.22579 0.018062 0.166551
* Indicates test value is significant at 5 % level of significance

Table 5 Result of GARCH (1, 1) model for nifty

Before T?2 After T?2 Whole period
Coefficient SE Coefficient SE Coefficient SE

Constant -0.125281* 0.057354 0.129302* 0.057587 -0.012346 0.042269

Tue 0.327817* 0.07888 -0.004027 0.078243 0.177672* 0.056509
Wed 0.091809 0.084584 -0.010355 0.073407 0.051868 0.057827
Thu 0.083444 0.073505 -0.039995 0.078572 0.028436 0.054348
Fri 0.283979* 0.077238 0.020779 0.078172 0.151241* 0.055676
Nifty(-1) 0.182444* 0.020634 0.074851* 0.025531 0.135294* 0.016736
C 0.092912 0.147467 0.212442 0.154751 0.156023 0.106475
RESID(-1)^2 0.115872* 0.016363 0.138853* 0.02491 0.128445* 0.014354
GARCH (-1) 0.851583* 0.018248 0.844971* 0.02146 0.853434* 0.013162
Tue 0.058337 0.273714 -0.452724 0.26569 -0.197592 0.192335
Wed 0.476942 0.326721 -0.177228 0.189597 0.047167 0.176614
Thu -0.795555* 0.282302 -0.039103 0.203597 -0.363387* 0.161933
Fri 0.277093 0.231096 -0.125993 0.245115 0.055489 0.166213
* Indicates test value is significant at 5 % level of significance

Similarly, result of GARCH (1, 1) model for return is significantly different from all other days.
Nifty is presented in Table 5. It indicates that for the Additionally, volatility of Thursday is significantly
whole sample period, Tuesday and Friday returns different from all other days’ volatility for whole
are significantly different from all other days, and sample period and for the period of before the
for the period of before the introduction of T?2 introduction of T?2 settlement. Although after the
settlement, Monday return is also significant. While introduction of T?2 settlement, volatilities of all
after the introduction of T?2 settlement, Monday days are insignificant.

123
334 Decision (September 2014) 41(3):327–337

Conclusion new position on first day of the week (Padhi 2010).

This study has tried to contribute to the existing
This study has tested day of the week anomaly in both literature of stock market efficiency by using day of
return and volatility of Indian stock market indices. the week anomaly approach, and findings of present
This study has analyzed daily closing price of Sensex study will be help to investors and portfolio managers
and Nifty for the period January 2, 1991 to January 31, in designing trading strategies. Security Exchange
2013. It has divided total sample period into two parts Board of India (SEBI) should undertake amendment in
viz before the introduction of T?2 settlement and after the rules to eradicate day of the week anomaly from
introduction of T?2 settlement. This study has used Indian stock market, and this may help in development
descriptive statistics and GARCH model for investi- of Indian economy.
gating day of the week effect of Indian stock market. Results obtained of this study are based on sample
The empirical analyses find that after the introduction size used and are reported particularly for stock market
of T?2 settlement, returns of Indian stock market indices used in this study only. Application of some
indices have increased, while volatility has been other stock market indices or use of high frequency
decreased, which represents safe and healthy situation data may produce different results. Furthermore,
for Indian stock market investors. Results obtained by conclusion of this study is based on application of
this study also indicate that before the introduction of GARCH model; the use of some other volatility
T?2 settlement, Monday and Thursday were provid- models may produce varying results. Therefore, one
ing negative return (Mehdian and Perry 2001; Gao and needs to be careful while generalizing the results of this
Kling 2005; Chiaku 2006; Cho et al. 2007), while after study. In future, someone may focus on investigating
the introduction of T?2 settlement, returns of all day of the week anomaly for some other stock market
week days are positive. indices or sectoral indices. Results of present study can
Moreover, this study finds that after the introduction also be validated with the help of high frequency data.
of T?2 settlement, Monday return is significant for
both Sensex and Nifty, which provides some signal of
information inefficiency in return of Indian stock References
market (Choudhry 2000; Kaur 2004; Sarma 2004; Chia
and Liew 2010). While in case of volatility, before the Aggarwal R, Rivoli P (1989) Seasonal and day-of-the-week
introduction of T?2 settlement, only Thursday vola- effects in four emerging stock markets. Financ Rev
tility is significant, while after the introduction of T?2 24(4):541–550
Agrawal A, Tandon K (1994) Anomalies or illusions? Evidence
settlement, volatilities of all days are insignificant, from Stock Markets in Eighteen Countries. J Int Money
which specify that day of the week anomaly does not Financ 13(10):83–106
exist in volatility of Indian stock market (Balaban et al. Ajayi RA, Mehdian S, Perry MJ (2004) the day-of-the-week
2001; Kiymaz and Berument 2003; Apolinario et al. effect in stock returns: further evidence from eastern
European markets. Emerg Mark Financ Trade 40:53–62
2006). However, the presence of day of the week effect Aksoy M, Dastan I (2011) Short selling and the day of the week
does not reject Efficient Market Hypothesis for Indian effect for Istanbul stock exchange. Int Res J Financ Econ
stock market, because the presence of significant return 70:205–216
on Monday does not imply abnormal return (Yakob Alagidede P (2008) Day of the week seasonality in African stock
markets. Appl Financ Econ Lett 4:115–120
et al. 2005). Hence, Indian government and regulatory Alejandro RC (2006) Day of the week effect on European stock
bodies should monitor movements of Indian stock markets. Int Res J Financ Econ 2:53–70
market indices very closely to circumvent abnormal Al-Loughani N, Chappel D (2001) Modelling the day-of-the-
gain of some individual. week effect in Kuwait stock exchange: a nonlinear
GARCH representation. Appl Financ Econ 11:353–359
Surprisingly, Monday effect has been vanished Alt, Fortin, Weinberger (2002) The day-of-the-week effect
after the introduction of T?2 settlement (Choudhry revisited: an alternative testing approach. Working Paper,
2000; Aksoy and Dastan 2011; Bin and Benjamin Economics Series, Institute for Advanced Studies Vienna
2011). This result is contradicting with views that Amanulla S, Thiripalraju M (2001) Week end effect: new evi-
dence from the Indian stock market. Vikalpa 26(2):33–50
Friday return should be highest, and Monday return Anwar Y, Mulyadi MS (2009) The day of the week effects in
should be lowest because of speculators action of Indonesia, Singapore, and Malaysian Stock Market. MPRA
closing their position on last day of week and building Paper No. 16873

123
Decision (September 2014) 41(3):327–337 335

Apolinario RMC et al (2006) Day of the week effect on Euro- Coutts JA, Kaplanidis C, Roberts J (2000) Security price
pean stock markets. Int Res J Financ Econ 2:53–70 anomalies in an emerging market: the case of the Athens
Arumugam S (1998) Day of the week effects in stock returns: an stock exchange. Appl Financ Econ 10:561–571
empirical evidence from Indian equity market. Prajnan Debasish SS (2012) Stock price seasonality effect and trading
27(2):171–191 strategy: an empirical study of selected it companies in
Bahadur KCF, Joshi NK (2005) The Nepalese stock market: India. Bus Manag Edu 10(2):264–288
efficiency and calendar anomalies. Econ Rev 17(17):40 Demirer T, Karan M (2002) An investigation of the day-of-the-
Balaban E, Bayar A, Kan OB (2001) Stock returns, seasonality week effect on stock returns in Turkey. Emerg Mark Financ
and asymmetric conditional volatility in world equity Trade 38(6):47–77
markets. Appl Econ Lett 8:263–268 Dhaoui, Farhani, Garfatta (2011) Stock return and trading
Basher SA, Sadorsky P (2006) Day-of-the-week effects in volume distribution across the day-of-the-week: evi-
emerging stock markets. Appl Econ Lett 13:621–628 dence from the Japanese stock market. J Econ Social
Beller K, Nofsinger JR (1998) On stock return seasonality and Stud 2(1)
conditional heteroscedasticity. J Financ Res 21:229–246 Dickey DA, Fuller WA (1979) Distribution of the estimators for
Berument MH, Dogan N (2012) Stock market return and vola- autoregressive time series with a unit root. J Am Stat Assoc
tility: day-of-the-week effect. J Econ Finan 36:282–302 74(366):427–431. doi:10.2307/2286348
Bhattacharya K, Sarkar N, Mukhopadhyay D (2003) Stability of Dickey DA, Fuller WA (1981) Likelihood ratio statistics for
the day-of-the-week effect in return and in volatility at the autoregressive time series with a unit root. Econ J Econ Soc
Indian capital market: a GARCH approach with proper 49(4):1057–1072. doi:10.2307/1912517
mean specification. Appl Financ Econ 13:553–563 Dicle M, Hassan K (2007) Day of the week effect in Istanbul
Bin, Benjamin (2011) Day-of-the-week effects: another evi- stock exchange. Sci J Adm Dev 5:53–83
dence from top 50 Australian stocks. Eur J Econ Financ Dimitrios, Katerina (2003) Market anomalies. In: European
Adm Sci 2:24 Financial Management Association Conference Proceed-
Bollerslev T (1986) A generalized autoregressive conditional ings, pp 87–92
heteroscedasticity. J Econ 31(3):307–327 Doyle JR, Chen CH (2007) The wandering weekday effect in
Cai J, Li Y, Qi Y (2006) The day-of-the-week effect: new evi- major stock markets. J Bank Financ 33:1388–1399
dence from the Chinese stock market. Chin Econ 39:71–88 Draper P, Paudyal K (2002) Explaining monday return. J Financ
Chander R, Mehta K, Sharma R (2008) A reexamination of the Res 25(4):507–520
day-of-the-week effect on the Indian stock markets. Icfai J Fields MJ (1931) Stock prices: a problem in verification. J Bus
Appl Financ 14(4):5–20 7:415–418
Charles AL (2010) Does the day-of-the-week effect on volatility Foo MW, Kok KL (2000) Seasonal anomalies of stocks on the
improve the volatility forecasts? Appl Econ Lett 17:257–262 Kuala Lumpur stock exchange second board. Cap Mark
Chen Y, Liang B (2004) Timing ability in the focus market of Rev 8:123–145
Asian country. J Financ 59(6):2871–2901 Fountas S, Segredakis KN (2002) Emerging stock markets
Chen G, Kwok C, Rui O (2001) The day-of-the-week regularity return seasonalities: the January effect and the tax-loss
in the stock markets of China. J Multinatl Financ Manag selling hypothesis. Appl Financ Econ 12:291–299
11:139–163 French K (1980) Stock returns and the weekend effect. J Financ
Chia RCJ, Liew VKS (2010) Evidence on the day-of-the-week Econ 8(1):55–70
effect and asymmetric behavior in the Bombay stock Gao L, Kling G (2005) Calendar effects in Chinese stock mar-
exchange. IUP J Appl Financ 16(6):17 ket. Ann Econ Financ 6:75–88
Chiaku CN (2006) Stock market returns analysis, day of the week Gardeazabal J, Regulez M (2004) A factor model of seasonality
effect, volatility of returns: evidence from European Finan- in stock returns. Q Rev Econ Financ 44:224–236
cial Markets 1997–2004. Int Res J Financ Econ 1:112–124 Gibbons M, Hess P (1981) Day of the week effects and asset
Cho YH, Linton O, Whang YJ (2007) Are there Monday effects returns. J Bus 54:579–596
in stock returns: a stochastic dominance approach. J Empir Goswami R, Angshuman R (2000) Day of the week effect on the
Financ 14:736–755 bombay stock exchange. ICFAI J Appl Financ 6:31–46
Chong R et al (2005) Pre-holiday effects: international evidence Guidi F (2010) Day-of-the-week effect and market efficiency in
on the declinend reversal of a stock market anomaly. J Int the Italian stock market: an empirical analysis. IUP J Appl
Money Financ 24:1226–1236 Financ 16(2):5–51
Choudhry T (2000) Day-of-the-week effect in emerging Asian Gupta A (2006) Day-of-the-week effect on the Indian stock
stock markets: evidence from the GARCH model. Appl market: new evidence. ICFAI J Appl Financ 12(8):5–14
Financ Lett 10:235–242 Hakan B, Halil K (2001) The day of the week effect on stock
Chowdhury, Sharmin (2010) Does cross-sectional risk explain market volatility. J Econ Financ 25(2):181–190
day-of-the-week effects In Bangladesh stock market? Int Ho R, Cheung Y (1994) Seasonal pattern in volatility in Asian
Res J Financ Econ 93:84 stock markets. Appl Financ Econ 4:61–67
Clare AD, Ibrahim MSB, Thomas SH (1998) The impact of Högholm K, Knif J (2009) The impact of portfolio aggregation
settlement procedures on day-of-the- week effects: evi- on day-of-the-week effect: evidence from Finland. Global
dence from The Kuala Lumpur stock exchange. J Bus Financ J 20:67–79
Financ Account 25:401–418 Holden KH, Thompson J, Ruangrit Y (2005) The Asian crisis
Connolly R (1989) An examination of the robustness of the and calendar effects on stock returns in Thailand. Eur J
weekend effect. J Financ Quant Anal 24:133–169 Oper Res 163:242–252

123
336 Decision (September 2014) 41(3):327–337

Huang Hu, Liao (2010) Short sale, stock liquidity, and the day- McGuinness PB (2005) A re-examination of the holiday effect
of-the-week effect: evidence from the Taiwan stock mar- in stock returns: the case of Hong Kong. Appl Financ Econ
ket. Rev Pac Basin Financ Mark Polic 13(1):71–90 15:1107–1123
Hui T (2005) Day-of-the-week effects in U.S. And Asia-Pacific Mehdian S, Perry MJ (2001) The reversal of the monday effect:
stock markets during the Asian financial crisis: a non- new evidence from US equity markets. J Bus Financ
parametric approach. Omega 33:277–282 Account 28(7–8):1043–1064
Hussain F (2000) The day of the week effect in the pakistani Mills T, Coutts J (1995) Calendar effects in the London stock
equity market: an investigation. Lahore J Econ 5(1):93–98 exchange FT-SE indices. Eur J Financ 1:79–93
Islam S, Watanapalachaikul S (2005) Models for anomalies Mookerjee R, Yu Q (1999) An empirical analysis of the equity
studies. In: Muller WA, Bihn M (eds) Empirical finance: markets in China. Rev Financ Econ 8:41–60
modelling and analysis of emerging financial and stock Muhammad, Rahman (2010) Efficient market hypothesis and
markets. Physica-Verlag, Heidelberg, pp 107–122 market anomaly: evidence from day-of-the week effect of
Jaffe J, Westerfield R (1985) The week-end effect in common Malaysian exchange. Int J Econ Financ 2(2):35–42
stock returns: the international evidence. J Financ 40:433–454 Nageswari P, Selvam M, Gayathri J (2011) Analysis of monthly
Kamath R, Chusanachoti J (2002) An investigation of the day- effect in Indian stock market. Res J Bus Manag
of-the-week effect in Korea: has the anomalous effect 5(4):170–177
vanished in the 1990s? Int J Bus 7(1):47–62 Nath GC, Dalvi M (2004) Day of the week effect and market
Kamath & Liu (2011) An investigation of the day-of-the-week efficiency evidence from indian equity market using high
effect on the Istanbul stock exchange of Turkey. J Int Bus frequency data of national stock exchange. In: Paper Pre-
Res 9(1):15–27 sented at The Center for Analytical Finance, Indian School
Karmakar M, Chakraborty M (2000) A curious finding of day- of Business, Hyderabad, December, pp 19–21
of-the-week effect in the Indian stock market. Indian Padhi P (2010). Days-of-the-week-effect and stock return vol-
Capital Markets: Trends and Dimensions. Tata McGraw- atility: theory and empirical evidence. Adv Manag, 3(4)
Hill, New Delhi Paul A, Theodore P (2006) Modeling and forecasting volatility
Kaur H (2004) Time varying volatility in the Indian stock of returns on the Ghana stock exchange using GARCH
market. Vikalpa: J Decis Makers 29:25–42 models. J Emerg Mark Financ 4(2):115–132
Keef SP, McGuinness PB (2001) Changes in settlement regime Peiro A (1994) Daily seasonality in stock returns: further
and the modulation of day-of-the-week effect in stock international evidence. Econ Lett 45:227–232
returns. Appl Financ Econ 11:361–372 Poshakwale S, Murinde V (2011) Modeling the volatility in East
Keim DB, Stambaugh RF (1984) A further investigation of the European emerging stock markets: evidence on Hungary
weekend effect in stock returns. J Financ 39:819–835 and Poland. Appl Financ Econ 11:445–456
Kelly F (1930) Why you win or lose: the psychology of spec- Prokop J (2010) On the persistence of a calendar anomaly: the
ulation. Houghton Mifflin, Boston day-of-the-week effect in German and US stock market
Kiymaz H, Berument H (2003) The day of the week effect on returns. Int Res J Financ Econ 54:176–190
stock market volatility and volume: international evidence. Roll R (1983) The turn-of-the-year effect and the return premia
Rev Financ Econ 12:363–380 of small firms. J Q Rev Econ Financ 44:224–236
Kok KL, Wong YC (2004) Seasonal anomalies of stocks in Rosenberg M (2004) The monthly effect in stock returns and
ASEAN equity markets. Sunway Coll J 1:1–11 conditional heteroscedasticity. Am Econ 48:67–73
Kumari D, Mahendra R (2006) Day-of-the-week and other Sarma SN (2004) Stock market seasonality in an emerging
market anomalies in the Indian stock market. Int J Emerg market. Vikalpa: J Decis Makers 29:35–41
Mark 1(3):235–246 Savva CS, Osborn DR, Gill L (2006) The day-of-the-week
Lakonishok J, Levi M (1982) Weekend effects in stock returns: a effect: evidence from six european stocks markets and the
note. J Financ 37(3):883–889 US, Working Paper, University of Manchester, England
Lakonishok J, Smidt S (1990) Are seasonal anomalies real? a Shaheen R (2006) Stock returns and the day of the week effect:
ninety-year perspective. Rev Financ Stud 1:403–425 Karachi stock exchange. Paper Presented At Campus for
Lean HH, Smyth R, Wong WK (2007) Revisiting calendar Finance Research Conference, Germany, January 16–17,
anomalies in Asian stock markets using a stochastic dom- 2008
inance approach. J Multinatl Financ Manag 17:125–141 Singhal A, Bahure V (2009) Weekend effect of stock rectums in
Lim Ho, Dollery (2010) An empirical analysis of calendar the Indian market. Great Lakes Herald 3:12–22
anomalies in the Malaysian stock market. Appl Financ Steeley JM (2001) A note on information seasonality and the
Econ 20:255–264 disappearance of the weekend effect in the UK stock
Liu, Li (2010) Day-of-the-week effects: another evidence from market. J Bank Financ 25:1941–1956
top 50 Australian stocks. Eur J Econ Financ Adm Sci Tachiwou AM (2010) Day-of-the-week-effects in West African
24:79–87 regional stock market. Int J Econ Financ 2(4):167–173
Lucey BM (2000) Anomalous daily seasonality in Ireland? Appl Tan RSK, Wong NT (1998) The diminishing calendar anoma-
Econ Lett 7:637–640 lies in the stock exchange of Singapore. Appl Financ Econ
Madureira LL, Leal RPC (2001) Elusive anomalies in the Bra- 8:119–125
zilian stock market. Int Rev Financ Anal 10:123–134 Thomas H (2002) Trends and calendar effects in stock returns.
Marrett GE, Worthington AC (2008) The day-of-the-week effect Aust Tech Anal Assoc J 40:222–234
in the Australian stock market: an empirical note on the Tong W (2000) International evidence on weekend anomalies.
market, industry and small cap effects. Int J Bus Manag 3:3–8 J Financ Res 23(4):495–522

123
Decision (September 2014) 41(3):327–337 337

Tsangarakis NV (2007) The day-of-the-week effect in the Athens Wong KA, Hui TK, Chang CY (1992) Day-of-the week effect:
stock exchange (ASE). Appl Financ Econ 17:1447–1454 evidence from the developing stock markets. Appl Financ
Ullah, Usman (2010) Market efficiency anomalies: a study of Econ 2:49–56
day of the week effect in Pakistani stock market. Inter- Worthington AC (2010) The decline of calendar seasonality in
discip J Contemp Res Bus 2(6):272–282 the Australian stock exchange, 1958–2005. Ann Financ
Ulussever Yumusak, Kar (2011) The day-of-the-week effect in 6:421–433
the Saudi stock exchange: a non-linear Garch analysis. Yadav PK, Pope PF (1992) Intraweek and intraday seasonalities
J Econ Social Stud 1(1):9–23 in stock market risk premia: cash and futures. J Bank
Wang K, Li Y, Erickson J (1997) A new look at the Monday Financ 16:233–270
effect. J Financ 52(5):2171–2186 Yakob Beal, Delpachitra (2005) Seasonality in the Asia Pacific
Wickremasinghe GB (2007) Seasonality of emerging stock mar- stock markets. J Asset Manag 6(4):298–318
kets: evidence from the colombo stock exchange. ICFAI J Zafar Urooj, Amjed (2010) Karachi stock exchange: testing day of
Appl Financ 13(6):43–65 the week effect. Interdiscip J Contemp Res Bus 2(5):78–89

123
Reproduced with permission of the copyright owner. Further reproduction prohibited without
permission.