Content uploaded by Zaiby Zahid
Author content
All content in this area was uploaded by Zaiby Zahid
Content may be subject to copyright.
270 Afro-Asian J. Finance and Accounting, Vol. 2, No. 3, 2011
Copyright © 2011 Inderscience Enterprises Ltd.
Impact of stock splits on stock price performance of
selected companies in Indian context
Kavita Chavali* and Zaiby Zahid
School of Business, Alliance University,
Chikkahagade Cross Chandapura – Anekal Main Road,
Anekal, Bangalore – 562106, Karnataka, India
E-mail: kavita.chavali@gmail.com
E-mail: Zaiby.zahid@gmail.com
*Corresponding author
Abstract: This paper aims to investigate the impact of stock splits on the stock
price performance of selected companies in the Indian stock market. A
purposive sampling method was employed, and a sample of 20 stock splits
announced by BSE-listed companies from the beginning of April 2006 to the
end of September 2008 was selected pertaining to different sectors. The study
employs the market model-event study methodology with an event window of
81 days (40 days prior to split and 40 days post-split) and split announcement
date (An date, t0) as the event date, to examine the market reaction. The
findings indicate that the market is found to react positively with significantly
positive average abnormal results on t0 and very near to the An date especially
evident during t–1 to t+1. The empirical results find the semi-strong form of
efficient market hypothesis to be true in the Indian context.
Keywords: stock split; market model; event study; stock returns; India.
Reference to this paper should be made as follows: Chavali, K. and Zahid, Z.
(2011) ‘Impact of stock splits on stock price performance of selected
companies in Indian context’, Afro-Asian J. Finance and Accounting, Vol. 2,
No. 3, pp.270–282.
Biographical notes: Kavita Chavali is currently an Associate Professor in the
School of Business, Alliance University, Bangalore. She received her Masters
in Business Administration and PhD from Andhra University. She has over 12
years of teaching experience in various B-schools in India and the West Indies.
She has many publications in peer-reviewed journals to her credit.
Zaiby Zahidis is a post-graduate programme student at School of Business,
Alliance University, Bangalore. She specialises in the area of finance. She has
prior experience working in the financial services industry in the Middle East.
1 Introduction
Stock splits are considered as financial puzzles by many researchers. The paradoxical
nature of stock splits has two contrasting views: one view holds that stock split is a costly
exercise that cannot affect the value of the firm, while the other view advocates that the
value of the firm immediately and significantly increases upon the announcement of
Impact of stock splits on stock price performance of selected companies 271
stock splits. There have been numerous research studies in India and abroad on the
market reaction to stock splits. However, these studies have been conducted mostly for
companies listed on NSE and BSE sensex and for periods up till 2005, which gave mixed
results. This provided the primary impetus to conduct this study for the BSE500 listed
companies in the more recent period from April 2006 to September 2008. The objective
of this research is to analyse the impact of stock splits on the stock price performance.
2 Theoretical background
The financial literature explains stock splits and positive abnormal returns accompanying
their announcement. Some of the noteworthy ones which this research study undertakes
to test are: signalling hypothesis, liquidity hypothesis, trading range hypothesis and the
semi-strong form of efficient market hypothesis (EMH).
2.1 Signalling hypothesis or information asymmetry
A signalling explanation of splits based on information asymmetry received considerable
attention in the academic literature. The basic notion is that managers use splits to signal
good information to investors. According to this view, the key role of splits is to convey
information, not to seek out an optimal price level. The value increase on split
announcements is often attributed to this signalling effect. One common interpretation of
this phenomenon is that by splitting the firm’s stock managers are attempting to signal to
outsiders that management believes that the firm’s stock price will increase (Ikenberry
and Ramanathan, 2002). According to the information asymmetry (signalling) hypothesis
(Brennan and Copeland, 1988; McNicholas and Dravid, 1990; Woolridge and Chambers,
1983), managers of undervalued firms use stock splits to signal positive information
about their future prospects to investors. The presence of positive abnormal returns
around the stock split announcement is observed in studies conducted in the past which
provide evidence for the signalling hypothesis. This is because when companies split
their shares the shares seem affordable to investors giving the investors the impression
that they are getting the shares for cheap. The present study conducted investigates
whether this signalling hypothesis holds true in the selected sample of companies in the
Indian scenario.
2.2 Semi-strong efficient market hypothesis
The hypothesis used to explain the stock price reaction of stock splits is the semi-strong
form of market efficiency (Fama, 1970). Semi-strong tests are called event studies (Fama,
1991). In an event study, it is measured how rapidly security prices respond to
announcements such as stock split or dividend announcement or news of a takeover. The
studies conducted on stock price reaction of stock splits or bonus issues are based on test
of semi-strong form of market efficiency. According to semi-strong EMH, current market
prices not only reflect all information content of historical prices/stocks but also reflect
all the information, which are publicly available about the companies being studied. The
present study attempts to gather evidence in support (if any) of the semi-strong form of
EMH in the Indian stock market.
272 K. Chavali and Z. Zahid
2.3 Optimal trading range hypothesis
According to the optimal trading range hypothesis, managers use stock splits as tools to
realign the share price and to broaden the ownership mix of the firm, i.e., to increase the
number of shareholders and decrease the institutional ownership of the firm. The
ownership mix becomes broader because the lower post-split share price makes it easier
for individual investors to purchase shares. Lamoureux and Poon (1987) document that
the average number of stockholders increases by 35.6% in the year the firm goes in for a
split. The management’s motivation to bring the share price to an optimal trading range
arises from the desire to improve liquidity which also forms the basis of the liquidity
hypothesis. Surveys of corporate managers (Baker and Gallagher, 1980) suggest that the
increase in round lots enabled by stock splits is used to increase a firm’s shareholder’s
base and thereby, improve liquidity. The present study attempts to test the existence of
the optimal trading range hypothesis in the Indian stock market.
2.4 Liquidity hypothesis
According to the liquidity hypothesis, firms split their stocks so as to increase the number
of trading shares in order to increase liquidity of the stock in the market. Liquidity
hypothesis suggests that with increasing volumes of trade in post-split the stock market
makes positive returns which get reflected in the decreasing expected returns on stocks in
the market. The evidence for the liquidity hypothesis from the previous studies conducted
is mixed. Desai et al. (1998) and Muscarella and Vetsuypens (1996) observe an increase
in trading volume during the post-split period provide support for the liquidity hypothesis
of stock splits. On the other side, Conroy et al. (1990) and Hwang (1995) present results
which indicate that corporate liquidity decreases after the split. This study investigates if
the sample of stock splits taken provides support for the liquidity hypothesis in the Indian
context. Goyenko et al. (2006) find that the worsened liquidity experienced by the firms
going in for a stock split is only a temporary phenomenon. The spread of the firms revert
back to the same level in 12–24 months for NYSE firms and in 12–39 months for
NASDAQ firms.
3 Review of research
Past studies have shown that the markets generally react positively to the announcement
of a stock split (Asquith et al., 1989; Bar-Joseph and Brown, 1977; Brennan and
Copeland, 1988; Grinblatt et al., 1984; Macey and Hara, 1997; McNicholas and Dravid,
1990; Lakonishok and Lev, 1987; Lijleblem, 1989; Woolridge and Chambers, 1983).
Grinblatt et al. (1984) document significant positive abnormal returns of about 3.3% in
the two days around a split announcement for a sample of 244 stocks having no other
confounding announcement effects. It was found in their study that the stocks listed on
the New York Stock Exchange (NYSE) and American Stock Exchange (AMEX) reacted
positively to stock split announcements and is not affected by other firm-specific news.
Lakonishok and Lev (1987) find that splits are aimed primarily at restoring stock prices
to a normal trading range; however, they also find some support for the signalling-based
hypothesis. Lijleblem (1989) confirms the presence of stock split announcement effects
for stocks traded on the Stockholm Stock Exchange. For the long-term observation, Fama
Impact of stock splits on stock price performance of selected companies 273
et al. (1969) find stock returns give 30% abnormal return two years after stock splits.
Ikenberry and Ramanathan (2002) note that an abnormal upward drift of 9% for one year
after the split in 1986 to 1997 time period which indicates positive abnormal returns
associated with stock split. Mishra (2007) found that splits have affected the share price
of the sample stocks negatively and these findings suggest that the splitting firms use
stock splits to reduce their share price to a preferred level. It induced brokers to revise
their optimistic valuation about the future prospects of the company.
Table 1 The sample list of companies which went for stock split and split ratio
Company Event (announcement)
date Split ratio Industry
Graphite India Ltd. 30 October 2006 5:1 Capital goods
Pantaloon Retail India
Ltd.
27 November 2006 5:1 Miscellaneous
ABB Ltd. 28 May 2007 5:1 Capital goods
Suzlon Energy Ltd. 23October 2007 5:1 Capital goods
GMR Infrastructure Ltd. 2 July 2007 5:1 Diversified
Electrosteel Castings 23 July 2007 10:1 Metal, metal products and
mining
BL Kashyap and Sons
Ltd.
30 July 2007 2:1 Housing related
Ahluwalia Contracts Ltd. 31 July 2007 5:1 Housing related
Peninsula Land Ltd. 21 September 2007 5:1 Housing related
GVK Power &
Infrastructure Ltd.
28 September 2007 10:1 Diversified
Jaiprakash Associates 07 November 2007 5:1 Housing related
City Union Bank 02 January 2008 10:1 Banking/finance
EMCO 17 January 2008 5:1 Capital goods
Shree Renuka Sugars 23 January 2008 10:1 Agriculture
NMDC Ltd 10 March 2008 10:1 Metal, metal products and
mining
Tanla Solutions Ltd 17 March 2008 2:1 Telecom/IT
Thomas Cook India Ltd. 30 April 2007 10:1 Tourism
MIC Electronics 09 May 2008 5:1 Telecom
Areva T&D India Ltd. 29 July 2008 5:1 Capital goods
Jagran Prakashan Ltd 24 December2008 5:1 Media and publishing
4 Methodology
The data pertaining to the information regarding stock splits announcements is based
mainly on press reports in leading newspapers such as Economic Times, The Hindu and
Business Standard. The time period selected for the study is April 2006 to September
274 K. Chavali and Z. Zahid
2008 during which the global financial crisis has occurred. The impact of this global
financial crisis on the price performance of the sample companies is not considered
which can be a limitation of the study. The An dates have been obtained from BSE
website and the ex dates have been obtained from Reuters. The BSE 500 index open and
close prices for the respective months have been taken from bseindia.com. The stock
price data is collected for a total of 81 trading days, 40 days prior to split, the share price
on the An date t0 and also share prices of 40 trading days post the An date. The event date
is defined as the announcement date of the board meeting recommending or approving
the stock splits. This approach assumes that the information was first known to the
market on the event date itself.
The companies in the research sample (Tables 1 and 2) are listed in BSE and have not
had any other stock split for at least one year before. The companies that had also
announced bonus shares along with the stock split announcement on the same date or
very near to it have been excluded since the researcher felt there would be an impact on
stock performance also due to bonus shares announcements. Those stock splits have been
excluded from the sample whose split ex date coincided with the ex date of another
corporate event of the stock such as stock dividend (bonus issue) or rights issue. Sample
selected represents nine industries and the scope of cross correlation among industries
would be negligible if any.
In this research study conducted, the market model event study methodology has been
employed to measure the effect of stock split announcements and its impact on the stock
price. The market model used by Fama et al. (1969) assumes a linear relationship
between the return of the security to the return of the market portfolio. This model has
been chosen because it is powerful in detecting abnormal returns when compared to other
elaborate methodologies like index model and also free from criticisms of capital asset
pricing model. Brown and Warner (1985) have specified that a simple methodology
based on market model is well specified and relatively powerful under wide variety of
conditions.
Estimation period and event window: An estimation period and event window are
selected to develop the market model. The historical stock price data of 40 days prior to
the stock split announcement is collected. The daily prices of 30 trading days are
included in the estimation period for computation of the regression coefficients for the
sample companies that went for stock splits. The estimation period spans 30 trading days
but ends ten days prior to the An date. This estimation period has been selected so that it
does not overlap with the announcement date and it is free from any post-split effect. The
ten days prior to An date have not been included in the estimation period so as to avoid
any effect on stock prices resulting from leakage of information (if any) regarding the
company’s decision to make split announcements. The study conducted uses a 81-day
event window, t-40 to t+40, where (event date) t0 is the announcement date or An date for
computing the abnormal returns (AR), for each of the stocks, the average abnormal
returns (AAR) of the sample of 20 companies and the cumulative average abnormal
returns (CAAR) around the An date of stocks in the sample. The study undertakes to
determine any significant positive AR associated with the stock splits around the event
dates (announcement dates) and the speed with which the information relating to split
announcements is absorbed into the security prices in the market.
The study is based on the following hypothesis:
Impact of stock splits on stock price performance of selected companies 275
Null hypothesis (H0) There is a significant positive AAR around the event date
(announcement date), or the AAR around the event date
(announcement date) is significantly greater than zero i.e.,
1/n.ΣAR ≥ 0, where n is the number of companies in the
sample selected.
Alternative hypothesis (Ha) There is a significant negative AAR around the event
date (announcement date), or the AAR around the event date
(announcement date) is significantly less than zero, i.e.,
1/n.ΣAR< 0.
For testing the hypothesis, ‘t score’ is computed on MS Excel and critical value approach
is used to reject or accept the null hypothesis.
Table 2 Firm characteristics (T stat, regression coefficients) using market model
Company α β T statistics
Graphite India Ltd. 0.0042 0.5386 1.3052
Pantaloon Retail India Ltd. 0.0045 0.3636 0.8482
ABB Ltd. 0.0039 0.6681 1.7676
Suzlon Energy Ltd. 0.0073 0.3321 1.6472
GMR Infrastructure Ltd. 0.0102 –0.1256 2.4162
Electrosteel Castings 0.0020 –0.3391 0.7019
BL Kashyap and Sons Ltd. 0.0052 1.0084 1.3089
Ahluwalia Contracts 0.0043 0.3150 0.7808
Peninsula Land 0.0022 1.4897 0.5040
GVK Power & Infrastructure Ltd. 0.0007 1.3249 0.1484
Jaiprakash Associates 0.0002 1.5802 0.0351
City Union Bank 0.0207 0.2552 2.3023
EMCO 0.0023 0.3447 0.5424
Shree Renuka Sugars 0.0145 0.6149 1.6609
NMDC Ltd –0.0038 0.5215 –0.4999
Tanla Solutions Ltd 0.0011 0.6958 0.2334
Thomas Cook India Ltd. –0.0033 0.5531 –1.5846
MIC Electronics –0.0012 0.4122 –0.2199
Areva T&D India Ltd. 0.0061 0.9439 0.9131
Jagran Prakashan Ltd 0.0017 0.7076 0.4068
Notes: *For computation of regression coefficients in the market model, (stock) split
adjusted prices are used and share prices are adjusted for any cash dividends paid
out.
**Close prices are adjusted for split so as to avoid the extreme deviation resulting
from stock sub division on the ex date.
276 K. Chavali and Z. Zahid
5 Empirical results
Table 3 presents results for each of the 81 days in the sample period for the entire sample
of 20 stock splits. It reports the average daily AR, AARs, CARs and CAAR for days t−81
to t+81 along with the test statistics obtained using the MS Excel descriptive analysis tool
for testing the null hypothesis. The table indicates that for the 40 days before the
announcement date there is no consistent pattern of AARs of companies engaging in
stock splits. The AARs are negative and significantly far from An date which is on days
t−35, t−30 and t−19 at 5% (0.05) significance level. Thus, null hypothesis is rejected on these
respective days, t−35, t−30 and t−18. However, it is observed that there are significantly
positive AARs on all other pre-announcement days (nearer to An date ‘t0’) particularly on
the three days t−24, t−13, t−6 prior and near to An date. Hence, null hypothesis cannot be
rejected on these days i.e., that there is a significantly positive AAR around An date. This
might be due to leakage of information regarding the company’s decision to recommend
the stock split or insider trading activities on these respective days very close to the split
announcements. CAAR is also seen to be significantly positive very near to the An date,
before and after the split. The build-up of AR prior to the announcement is consistent
with the semi-strong EMH as stock prices respond to the information leaking out prior
and very near to the announcements. The AR, AAR and CAAR increase when the
information becomes available in those days for the particular stocks. In the study
conducted, it is seen that majority of the stocks experienced AARs significantly greater
than zero on the An date t0 (significantly positive at 0.05, stronger evidence in support of
semi-strong EMH) and also particularly near (t−24, t−19, t−13, t−11,t−6, t+8, t+18, t+21, t+24) to the
An date.
On An date and one day after on t+1 day there were positive AAR while the AAR
was positive and significant on t0 at 5% significance level. Also, CAAR of 2.98%
(+0.0298) on t0 and 3.13% (+0.0313) on the first trading day following the
announcement (t+1) is observed. This corroborates with the semi-strong EMH since the
stock prices respond rapidly as the published information becomes available to the
public or investors especially the retail investors to whom the prior information had
not reached. However, AARs observed post An date were significantly greater than
zero for majority of the days post announcement. The only exceptions were on
days t+5 and t+34 with 5% significance level (strong evidence to reject H0) and day t+21
with 10% (weaker evidence to reject H0) where there were significant and
negative AARs. From the An date to post announcement days, the AARs were positive
for a majority of 19 days – t+0, t+1, t+4, t+8, t+10, t+16, t+17, t+18, t+19, t+23, t+24, t+26, t+31, t+33,
t+35, t+37, t+38, t+39, and t+40, most significant positive AARs being on days t0 (An date),
t+18 and t+24. This leads to acceptance of the null hypothesis (H0). It indicates that
post announcements, the returns of the stocks did boost for majority of the stocks in the
sample. This positive market reaction post announcement can be attributed to the rapid
absorption of the information which is perceived positively and reflected in the stock
prices. This could also be explained on account of other announcements regarding the
record date fixed etc., it can be partly explained by the fact that most companies
announced their respective record date fixed, between the period from as less as 15 days
to even 35 days after t0. Another explanation lies in the positive market reaction post
execution day. It is seen that the split ex date of most companies was on days t+8, t+11, t+15,
t+16, t+19, t+23, t+29, t+30, to even as late as t+34. The results obtained also indicate that there
Impact of stock splits on stock price performance of selected companies 277
is a signalling impact on the share prices gradually resulting in a price increase post
announcements.
Table 3 Daily average AR and t statistics as per market model
Day t N AAR% T stat Significance Null hypothesis
test CAR CAAR
–40 20 –0.27% –0.5624 - Can’t reject H0 –1.0089 –0.0027
–39 20 –0.51% –0.7322 - Can’t reject H0 –1.0183 –0.0079
–38 20 0.01% 0.0236 - Can’t reject H0 –1.1570 –0.0077
–37 20 –0.66% –0.9582 - Can’t reject H0 –1.0973 –0.0144
–36 20 –0.50% –1.1898 - Can’t reject H0 –0.9143 –0.0193
–35 20 –0.81% –2.2481 ** Reject H0 –0.9557 –0.0274
–34 20 –0.35% –0.7131 - Can’t reject H0 –0.9319 –0.0309
–33 20 –0.23% –0.4657 - Can’t reject H0 –0.7885 –0.0332
–32 20 –0.10% –0.2736 - Can’t reject H0 –0.5822 –0.0341
–31 20 0.73% 0.7373 - Can’t reject H0 –0.5418 –0.0268
–30 20 –0.83% –2.4633 ** Reject H0 –1.7910 –0.0351
–29 20 –0.29% –0.4405 - Can’t reject H0 –1.3981 –0.038
–28 20 –0.60% –1.1237 - Can’t reject H0 –1.1233 –0.0439
–27 20 –0.12% –0.1784 - Can’t reject H0 –0.9974 –0.0451
–26 20 –0.02% –0.0418 - Can’t reject H0 –1.0180 –0.0454
–25 20 0.11% 0.1967 - Can’t reject H0 –0.8756 –0.0443
–24 20 1.11% 2.1302 ** Can’t reject H0 –0.8557 –0.0332
–23 20 0.10% 0.1683 - Can’t reject H0 –1.0390 –0.0322
–22 20 –0.16% –0.2983 - Can’t reject H0 –0.9235 –0.0338
–21 20 –0.68% –1.2962 - Can’t reject H0 –1.0245 –0.0406
–20 20 0.40% 0.5660 - Can’t reject H0 –0.8357 –0.0365
–19 20 1.05% 1.9198 ** Can’t reject H0 –0.5478 –0.0261
–18 20 –1.09% –2.4585 ** Reject H0 –1.1357 –0.0369
–17 20 –0.37% –0.6355 - Can’t reject H0 –1.1386 –0.0407
–16 20 1.08% 1.7211 * Can’t reject H0 –1.1405 –0.0298
–15 20 0.66% 1.2520 - Can’t reject H0 –0.9955 –0.0232
–14 20 –0.20% –0.2453 - Can’t reject H0 –1.0764 –0.0252
–13 20 1.38% 2.5660 ** Can’t reject H0 –0.8770 –0.0114
–12 20 0.72% 0.7487 - Can’t reject H0 –0.7785 –0.0042
–11 20 0.91% 1.6245 * Can’t reject H0 –0.7899 0.0049
–10 20 0.37% 0.4316 - Can’t reject H0 –0.8465 0.0086
–9 20 –0.46% –0.6659 - Can’t reject H0 –1.1949 0.004
–8 20 1.13% 1.1684 - Can’t reject H0 –0.8131 0.0153
–7 20 0.46% 0.5823 - Can’t reject H0 –0.9038 0.0198
–6 20 2.03% 2.0399 * Can’t reject H0 –0.5405 0.0401
278 K. Chavali and Z. Zahid
Table 3 Daily average AR and t statistics as per market model (continued)
Day t N AAR% T stat Significance Null hypothesis
test CAR CAAR
–5 20 –0.97% –0.7910 - Can’t reject H0 –1.2078 0.0305
–4 20 –0.52% –0.7137 - Can’t reject H0 –0.7241 0.0252
–3 20 –0.58% –0.8516 - Can’t reject H0 –0.8500 0.0194
–2 20 –0.23% –0.2889 - Can’t reject H0 –0.8381 0.0171
–1 20 –0.14% –0.1421 - Can’t reject H0 –0.6438 0.0158
0 20 1.41% 1.8670 ** Can’t reject H0 –0.8377 0.0298
1 20 0.15% 0.1829 - Can’t reject H0 –1.0940 0.0313
2 20 –0.25% –0.2672 - Can’t reject H0 –1.0854 0.0288
3 20 –0.87% –1.0271 - Can’t reject H0 –0.8919 0.0201
4 20 0.45% 0.5404 - Can’t reject H0 –1.0581 0.0246
5 20 –1.23% –2.2379 ** Reject H0 –0.9242 0.0123
6 20 –0.21% –0.2259 - Can’t reject H0 –0.8406 0.0102
7 20 –0.95% –1.1883 - Can’t reject H0 –1.1617 0.0007
8 20 1.46% 1.3755 * Can’t reject H0 –0.6929 0.0153
9 20 –0.68% –1.0994 - Can’t reject H0 –1.5422 0.0085
10 20 0.22% 0.2964 - Can’t reject H0 –0.9475 0.0107
11 20 –0.45% –0.7504 - Can’t reject H0 –1.0490 0.0062
12 20 –0.63% –0.7735 - Can’t reject H0 –1.4479 –0.0001
13 20 –0.78% –1.0393 - Can’t reject H0 –0.4923 –0.0079
14 20 –0.58% –0.8759 - Can’t reject H0 –1.1899 –0.0137
15 20 –0.55% –0.6136 - Can’t reject H0 –1.1613 –0.0191
16 20 0.26% 0.3398 - Can’t reject H0 –0.9749 –0.0166
17 20 0.52% 0.9163 - Can’t reject H0 –0.7790 –0.0114
18 20 1.24% 1.3896 * Can’t reject H0 –0.9419 0.001
19 20 0.03% 0.0378 - Can’t reject H0 –0.8840 0.0014
20 20 –0.84% –1.0563 - Can’t reject H0 –0.9242 –0.007
21 20 –1.82% –1.4413 * Reject H0 –1.2297 –0.0252
22 20 –0.81% –0.8785 - Can’t reject H0 –1.1552 –0.0334
23 20 0.57% 0.6642 - Can’t reject H0 –1.1566 –0.0277
24 20 1.41% 1.6432 * Can’t reject H0 –0.8237 –0.0136
25 20 –0.08% –0.1043 - Can’t reject H0 –1.0454 –0.0144
26 20 0.29% 0.3468 - Can’t reject H0 –1.2060 –0.0115
27 20 –0.45% –0.5357 - Can’t reject H0 –1.1906 –0.016
28 20 –4.68% –1.2474 - Can’t reject H0 –1.8758 –0.0628
29 20 –0.07% –0.0691 - Can’t reject H0 –0.9079 –0.0635
30 20 –0.92% –0.7364 - Can’t reject H0 –1.3974 –0.0728
31 20 0.37% 0.4634 - Can’t reject H0 –0.6518 –0.0691
Impact of stock splits on stock price performance of selected companies 279
Table 3 Daily average AR and t statistics as per market model (continued)
Day t N AAR% T stat Significance Null hypothesis
test CAR CAAR
32 20 –0.88% –0.9261 - Can’t reject H0 –1.0682 –0.0779
33 20 0.96% 0.7917 - Can’t reject H0 –0.7668 –0.0683
34 20 –1.26% –2.0213 ** Reject H0 –1.3404 –0.0809
35 20 0.09% 0.0930 - Can’t reject H0 –0.8296 –0.08
36 20 –0.81% –0.7885 - Can’t reject H0 –1.1490 –0.0881
37 20 1.17% 0.9182 - Can’t reject H0 –0.7748 –0.0763
38 20 0.41% 0.5464 - Can’t reject H0 –0.8631 –0.0723
39 20 0.18% 0.2027 - Can’t reject H0 –0.9016 –0.0705
40 20 0.57% 0.4544 - Can’t reject H0 –0.9751 –0.0647
Notes: Significance at 0.05 level ** and t0.05 = 1.729
Significance at 0.10 level * and t0.10 = 1.328
*Split adjusted share prices are used in the event study and share prices are
adjusted for any cash dividends paid out.
**Close prices are adjusted for split so as to avoid the extreme deviation resulting
from stock sub division on the ex date.
Table 4 Average value of AAR across different event windows
Event window AAR T stat Null hypothesis test Significance
t–40 to t–21 –0.20% –1.8716 Reject H0 **
t–20 to t–1 0.28% 1.4955 Cant reject H0 *
t–1 to t0 0.63% 0.8223 Cant reject H0 NS
t0 to t+1 0.78% 1.2388 Cant reject H0 NS
t–1 to t+1 0.47% 0.9982 Cant reject H0 NS
t+2 to t+20 –0.20% –1.1939 Cant reject H0 NS
t+21 to t+40 –0.29% –0.9698 Cant reject H0 NS
t–40 to t+40 –0.08% –0.7819596 Cant reject H0 NS
Notes: *Statistical significance at 0.10 and t0.10 = 1.328
**Statistical significance at 0.05 and t0.05 = 1.729
NS = not significant
It is observed from Table 4 that no significant positive AARs are present far from the An
date (30–40 days prior to An date) as shown across the event window t−40 to t−21 as
significant negative AARs are seen. Hence, null hypothesis of there being significant
positive returns associated with stock splits is rejected across this window.
However, across the event window t−20 to t−1, on an average positive AAR is observed
which is significant. This can be attributed to the leakage of information (if any)
regarding the company’s decision to split the stock. On the An date and one day prior a
significant and increased AAR is observed for the event window t−1 to t0. The increasing
trend is followed even on one day post An date i.e., across event window t+1 to t0 where a
280 K. Chavali and Z. Zahid
higher positive AAR is observed. This indicates that a semi-strong form of EMH does
exist in the Indian stock markets as the information as and when available is more or less
quickly absorbed into the stock prices which is reflected by higher positive AR on An
date and very near the An date (across t−1 to t0, t0 to t+1).
A market correction is seen from trading days t+2 to t+20, i.e., the AAR is seen to
shrink back i.e., stock prices and daily returns on stock are observed to decrease. It took
two days for the market to correct from the speculations that had been built up prior to
announcement. The trading days across t+20 to t+40 the AR are seen to shrink further, daily
returns on stock once again declined between 20 to 40 days after An date and a reversal
in AAR i.e., in stock prices is seen post announcements.
When considering the AARs across the entire event window t−40 to t+40, a net negative
AAR is seen which is not significant and the null hypothesis cannot be rejected. This
implies that there is a positive impact on the stock prices around the split announcements.
This study works mainly around the signalling hypothesis and testifies the hypothesis.
The signalling hypothesis provides an explanation for the positive abnormal return
observed. A split may be interpreted as a signal that the firm’s managers are optimistic
regarding the company’s future prospects.
It is seen that there is a dip in the negative AR, i.e., an increase in AR is seen post
announcements. This is in conformity with the signalling hypothesis. The impact of stock
splits on stock price performance in the Indian stock market is shown by the empirical
results of AR, AAR and CAAR. Since, the AAR is found to be significantly positive
around the An date and on the An date, which confirms the existence of semi-strong form
of EMH in Indian stock market which states that the information is reflected in the stock
prices, implying the important role of information content regarding the corporate
announcement-stock splits. In most of the stocks in the sample selected a significant
positive price effect associated with stock split was observed on the An date itself when
the published information became available to the investors. There were also significant
positive AARs observed few days prior to the An date probably due to leakage of
information before it was made public. This again conforms to the semi-strong form of
EMH. The alternative hypothesis (Ha) does not hold true for most of the trading days as
the ARs very near and prior to An day and also post-announcement are found not to be
less than zero(negative) indicating positive market reactions very close before the An
date as well as post-split announcement. The market is found to react positively just after
the An date with a CAAR of 3.13% (0.0313) on the first trading day (an increase over
days prior to An date as well as An date) following the announcement. From the stock
price data collected for the companies in the sample, it is observed that some of the
stocks that were highly priced, when they were split, made the share price more
affordable and the stocks saw a general trend of higher positive AR post-split
announcement (short selling)and post execution (intraday trading) like Electrosteel
Castings, NMDC Ltd., Pantaloon Retail India Ltd., Jaiprakash Associates and Ahluwalia
Contracts (both from the outperforming housing related sector), Thomas Cook India Ltd.
This provides some evidence to support the optimal range trading hypothesis in the
Indian stock market. From the historical data, it is observed that there is a general trend of
decrease in trading volume (decrease in turnover) during the post-split period for most of
the 20 sample stocks (exception being Ahluwalia Contracts and Jaiprakash Associates).
Thus, the finding does not support the liquidity hypothesis of stock splits in the Indian
stock market. Hence, the findings confirm that the positive excess returns the stocks
Impact of stock splits on stock price performance of selected companies 281
made in the market are strongly attributed to the signalling effect of the stock split
announcements.
6 Discussion and conclusions
Stock splits are like optical illusions but in reality they are only adjustments in the
number of shares outstanding. The company’s equity and the value of shareholders’
holdings remain unchanged. The study conducted reveals that there exists a positive
market reaction post-split announcements in case of the Indian stock market. The study
done for a sample of 20 stock splits tracks the AR generated by BSE listed companies.
Computation of AR associated with the stock split is done for trading days around the
event date – split announcement date for the sample stocks, each making a stock split
announcement during April 2006 to September 2008. The study reports empirical results
conforming to the signalling hypothesis and are consistent with the results obtained by
Asquith et al. (1989), Brennan and Copeland (1988), Grinblatt et al. (1984) and
Lakonishok and Lev (1987). The study finds evidence of positive AR around the
announcement date and confirms that the stock splits have a favourable impact on the
stock price performance of the stocks as shown by positive market reaction post-split
announcements, in the Indian context. The findings of the study are consistent with the
findings of the past research done which give evidence to support the semi-strong form of
EMH in the context of Indian stock market. The study reveals that the information
regarding the corporate announcements of stock splits is absorbed quickly into the share
prices and has a measurable impact on the stock price performance of the stocks that
announce and execute splits. However, the result holds true for the selected sample of
BSE listed companies and during the period considered for the study. It cannot be
generalised for other exchange traded stocks, nor for the BSE500 listed companies in
other periods since even same companies making corporate announcements in the future
do so in a different market environment.
Stock split should not be the only deciding factor that entices an investor into
investing in a stock. There are some strong psychological reasons why companies split
their stock; the investors should understand that split does not change any of the business
fundamentals. An investor who can successfully predict which stocks are going to split
by looking at their outstanding performance and soaring share prices can profit from
investing in such stocks in the pre-announcement period because such stocks are likely to
appreciate in share value. The key to making profits for an investor from this stage is
being able to determine which stocks are the most likely to split during the next three to
six months. In conclusion, it can be stated that stock split announcements do attract
investors, thus generating a positive market reaction.
References
Asquith, P., Healy, P. and Palepu, K. (1989) ‘Earnings and stock splits’, The Accounting Review,
No. 64, pp.387–403.
Baker, H.K. and Gallagher, P.L. (1980) ‘Management view of stock splits’, Financial
Management, Vol. 9, pp.73–77.
Bar-Joseph, S. and Brown, L. (1977) ‘A re-examination of stock splits using betas’, Journal of
Finance, Vol. 32, pp.1069–1080.
282 K. Chavali and Z. Zahid
Brennan, M.J. and Copeland, T.E. (1988) ‘Stock splits, stock prices and transaction costs’, Journal
of Financial Economics, Vol. 22, pp.83–102.
Brown, S.J. and Warner, J.B. (1985) ‘Using daily stock returns: the case of event studies’, Journal
of Financial Economics, Vol. 14, pp.3–31.
Conroy, R.M., Harris, R.S. and Benet, B.A. (1990) ‘The effects of stock splits on bid-ask spreads’,
Journal of Finance, Vol. 45 No. 4, pp.1285–1295.
Desai, A.S., Nimalendran, M. and Venkataraman, S. (1998) ‘Changes in trading activity following
stock splits and their affect on volatility and the adverse information component of the bid-ask
spread’, Journal of Financial Research, Vol. 21, pp.159–183.
Fama, E., Fisher, L. Jensen, M.C. and Roll, R. (1969) ‘The adjustment of stock prices to new
information’, International Economic Review, Vol. 10, pp.1–21.
Fama, E. (1970) ‘Efficient capital markets: a review of theory and empirical work’, The Journal of
Finance, No. 2, pp.383–417.
Fama, E. (1991) ‘Efficient capital markets II’, The Journal of Finance, Vol. 46, pp.1575–1618.
Goyenko, R.Y., Holden, C.W. and Ukhov, A.D. (2006) ‘Do stock splits improve liquidity?’,
Working paper, Indiana University.
Grinblatt, M.S., Masulis, R.W. and Titman, S. (1984) ‘The valuation effects of stock splits and
stock dividends’, Journal of Financial Economics, Vol. 13, No. 4, pp.461–490.
Hwang, K.C. (1995) ‘The effects of reverse splits on the liquidity of the stock’, Journal of
Financial and Quantitative Analysis, March, pp.159–169.
Ikenberry, D. and Ramanathan, S. (2002) ‘Under reaction to self selected news items – a case of
stock split’, Review of Financial Studies, Vol. 15, pp.489–526.
Lakonishok, J. and Lev, B. (1987) ‘Stock splits and stock dividends: why, who and when’, Journal
of Finance, Vol. 42, pp.913–932.
Lamoureux, C.G. and Poon, P. (1987) ‘Market reaction to stock split’, Journal of Finance, Vol. 42,
pp.1347–1370.
Lijleblem, E. (1989) ‘The informational impact of announcements of stock dividends and stock
splits’, Journal of Business Finance and Accounting, Winter, Vol. 16, No. 5, pp.681–698.
Macey, R.M. and Hara, M.O. (1997) ‘The law and economics of best execution’, Journal of
Financial Intermediation, pp.188–223.
Mishra, A.K. (2007) ‘The market reaction to stock splits – evidence from India’, International
Journal of Theoretical and Applied Finance, Vol. 10, No. 2, pp.251–271.
McNicholas, M. and Dravid, A. (1990) ‘Stock dividend, stock split and signaling’, Journal of
Finance, Vol. 45, pp.857–879.
Muscarella, C.J. and Vetsuypens, M.R. (1996) ‘Stock splits: signaling or liquidity? The case of
ADR ‘solo-splits’’, Journal of Financial Economics, Vol. 42, No. 1, pp.3–26.
Woolridge, J.R. and Chambers, D.R. (1983) ‘Reverse splits and shareholders’ wealth’, Financial
Management.
