Beruflich Dokumente
Kultur Dokumente
Thus( the abnormal returns were averaged by dividing it by the number of days to
find out daily average abnormal returns The process was repeated for all the
dates and finally average cumulative abnormal returns were obtained This is the
second measure <8A)=( it measures the investors total return over a period
starting from before the announcement of dividend to after the dividend
announcement day The cumulative abnormal returns from day t
$
through t
%
(8A)
t
( are .C*7D
8AA)
t X
_ Avg <A)t= where t X t
$
to t
%
YYYYYYYYYYYYYYY
<$/=
8AA) may be positive or negative "f 8AA) is negative in periods after
dividend announcements( this suggests dividend announcements do not carry
information about future earnings and cash flows of the companies A positive
8A) indicates distribution of dividend adds to shareholders value by conveying
good news to the market Ge use a +$ day event window period
starting from L%9to Z%9 day relative to the dividend
announcement day <9 day= 0or the purpose of analysis both interim and final
dividend announcements has been taken
To compute the tBstatistic( first( all abnormal returns are standardi>ed as.
:A)
it
X A)
it
! :i <A)=YYYYYYYYYYYYYYYYYYYYYYY
<$2=
Dayananda sagar college of arts, science and commerce Page 54
IMPACT OF DIVIDEND POLICY ONSHAREHOLDERS VALUE: A STUDY OF INDIANFIRMS
where (:i <A)= is the standard deviation of the abnormal returns of stock Mi in the
estimation period The tBstatistic for the sample of N observations for each day Mt
in the event window is calculated as.
Dayananda sagar college of arts, science and commerce Page 55
IMPACT OF DIVIDEND POLICY ONSHAREHOLDERS VALUE: A STUDY OF INDIANFIRMS
t<:A)= X <_
iX$ to I
:A)
it
= $!bI
YYYYYYYYYYYYYYYYYYY<$6 : ANALYSES AND
FINDINGS
:-1- LINTNER MODEL IN ITS SECTOR
The regression results <refer to Anne'ure %= of one way 0i'ed effect
model shows that divided paid during previous year is significant at ,U level of
significance The Adjusted ) s#uare is 69U0 statistics is significant at ,U
level of significance showing overall validity of the model The results
highlight that there is -ow dividend smoothing in this sector as it is
characteri>ed by high target payout ratio and high speed of adjustment
coefficient
:-.- LINTNER MODEL IN FMCG SECTOR
The regression results <refer to Anne'ure %= show that 4AT and
dividend paid during previous year are significant at ,U level of significance The
value of Adjusted ) s#uare is 7,U The 0 statistics are also significant at 7,U
confidence interval showing the overall validity of the model in the 0M8N
sector Target payout is high but speed of adjustment factor is between the range
suggested by -intner<$7,/=Therefore( it can be said that in this sector dividend
signaling and smoothing effects are present
"t may be noted that -M test results show that 8lassical linear regression model
could also be used
Dayananda sagar college of arts, science and commerce Page 56
IMPACT OF DIVIDEND POLICY ONSHAREHOLDERS VALUE: A STUDY OF INDIANFIRMS
%*
Dayananda sagar college of arts, science and commerce Page 57
IMPACT OF DIVIDEND POLICY ONSHAREHOLDERS VALUE: A STUDY OF INDIANFIRMS
:-8 LINTNER MODEL IN SERVICE SECTOR
Dividend paid during the previous year and 4AT both <refer to Anne'ure
%= are found to be significant at ,U level of significance The value of adjusted
) :#uare is 2/$6U The overall validity of the model has been found out with the
help of 0 statistics
The findings show very high target payout ratio of *$,U combined with high
adjustment coefficient of ,%26%/, indicates absence of dividend signaling and
smoothing effects
6- FACTOR ANALYSIS AND REGRESSION RESULTS ON
EBTRACTED FACTORS
6-1- ANALYSIS OF ITS SECTOR
Table %$ <refer to Anne'ure *= shows RaiserBMeyerBPlkin Measure
of :ampling Ade#uacy values "t is measure that judges the sampling ade#uacy
The value obtained is ,/9 which ensures the sample si>e is ade#uate to apply
0actor Analysis
6-1-1 FACTOR EBTRACTION
4rincipal 8omponent Analysis method was used to e'tract the factors The
Table %% <refer to anne'ure *= shows the factor pattern matri'( which
highlights variance e'hibited by e'tracted factors Nenerally( the identification
of the factors is determined by the factor loadings( and the relationship of the
factor with the variable is based on the signs of factor loadings A factor loading
is simply the correlation of an original variable with factor As suggested by
Dillion and Noldstein( variables with factor loadings greater than absolute value
Dayananda sagar college of arts, science and commerce Page 58
IMPACT OF DIVIDEND POLICY ONSHAREHOLDERS VALUE: A STUDY OF INDIANFIRMS
of 9*9 or more are considered significant and( thus( used in labelling of factors
As shown in the factor pattern matri' a set of 6 factors have been been
e'tracted These factors have been labelled as 0actor of dividend signaling and
promoter holding( 0actor of li#uidity ratios( 0actor of longterm solvency(
0actor of financial and systematic risk( 0actor of firm si>e( 0actor of retained
earnings and dividend stability( 0actor of growth and e'pansion and 0actor of
valuation and capital market ratios
FIGURE 1: SCREE PLOT :cree 4lot
,
+
*
%
Eigenvalue
$
9
$ * , 2 7 $$$*$,$2$7%$
Dayananda sagar college of arts, science and commerce Page 59
IMPACT OF DIVIDEND POLICY ONSHAREHOLDERS VALUE: A STUDY OF INDIANFIRMS
8omponent Iumber
An elbow in the scree plot indicates the point at which the inclusion of
additional factors does not contribute significantly in e'plaining the variance in
data set 0actors above the elbow of the plot are retained The :cree plot
shown above has an elbow at 0actor 6Therefore a set of 6 0actors were chosen
which accounts for about 22U of the variations in the data
Dayananda sagar college of arts, science and commerce Page 60
IMPACT OF DIVIDEND POLICY ONSHAREHOLDERS VALUE: A STUDY OF INDIANFIRMS
6-1-. REGRESSION RESULTS ON EBTRACTED FACTORS
The Table %*<refer anne'ure *= shows the regression results on e'tracted
factors 0actors *( +(,(/ and 6 have e'pected signs Put of these factors only two
factors ie 0actor / and 6 have regression coefficients( which are statistically
significant at ,U level of significance Both factor $ and % have e'actly
opposite signs of regression coefficients compared to what was e'pected based
on previous research studies The value of Adjusted )
%
is 92,2 The 0 values are
also significant at ,U level of significance
6-1-8- FINDINGS
A set of 6 factors has been e'tracted through the techni#ue of 4rincipal
8omponent analysis The regression results show that 0actor of dividend
signaling and ownership( li#uidity ratios are significantly negatively related
with D4 ratio Also a positive significant relationship e'ists between )1
earnings and D4 ratio This shows that in "T sector capital gains are preferred
are cash dividends The information environment is highly symmetrical
Therefore( cash dividends are not used to signal their profitability to
shareholders
6-. ANALYSIS OF FMCG SECTOR
The first step was to calculate RMP The value obtained is /79 which
ensures the sample si>e is ample to apply 0actor Analysis <refer to Table %+ in
anne'ure * for test values=
Dayananda sagar college of arts, science and commerce Page 61
IMPACT OF DIVIDEND POLICY ONSHAREHOLDERS VALUE: A STUDY OF INDIANFIRMS
6-.-1 FACTOR EBTRACTION
The Table %, in anne'ure * shows the factor pattern matri'( which
highlights variance e'hibited by e'tracted factors "t also depicts the loadings
of each variable on a given factor The e'tracted factors has been labeled as
0actor of Dividend
:ignaling and :moothing( 0actor of cash flow #uality and firm si>e( 0actor of
future e'pansion and growth( 0actor of ownership and li#uidity( 0actor of
earning variability and systematic risk( 0actor of long term solvency and financial
leverage
FIGURE .: SCREE PLOT
S,r&& P+'$
/
,
Eigenvalue
+
*
%
Dayananda sagar college of arts, science and commerce Page 62
IMPACT OF DIVIDEND POLICY ONSHAREHOLDERS VALUE: A STUDY OF INDIANFIRMS
$
9
As discussed( 0actors above the elbow of the plot are retained The :cree plot
shown above has an elbow at 0actor /Therefore a set of / 0actors were chosen
which accounts for about 2/U of the variations in the data
6-.-. REGRESSION RESULTS OF EBTRACTED FACTORS
Table %/ <refer to anne'ure *= shows 0actors $(%(* and / have
e'pected signs Put of these factors only one factor ie 0actor % has regression
coefficient( which is statistically insignificant at ,U level of significance
0actor + and , have e'actly opposite signs of regression coefficients compared
to what was e'pected based on previous research studies The value of Adjusted
)
%
is 9/*% 0 statistics are significant at ,U level of significance
6-.-8 FINDINGS
Put of si' e'tracted factors , were found to be significantly related to D4
ratioA 4ositive and significant relationship between factor of systematic risk(
Dividend signaling and smoothing( -ong term solvency and financial
leverage Iegative and significant relationship has been found between
0actor of li#uidity and ownership and 0actor of growth and e'pansion This
implies that if the systematic risk increases these firms increase their
dividend payout The agency conflicts are not so grave since 0M8N firms
operate with low levels of debt &igher the growth opportunities available to a
firm lower will be the dividend payout ratio in the 0M8N sector
Dayananda sagar college of arts, science and commerce Page 63
IMPACT OF DIVIDEND POLICY ONSHAREHOLDERS VALUE: A STUDY OF INDIANFIRMS
Dayananda sagar college of arts, science and commerce Page 64
IMPACT OF DIVIDEND POLICY ONSHAREHOLDERS VALUE: A STUDY OF INDIANFIRMS
6-8- ANALYSIS OF SERVICE SECTOR
As done in the other sectors the first step was to calculate RMP "t is a
measure that judges the sampling ade#uacy The value obtained is ,9/ which
ensures the sample si>e is sufficient to apply 0actor Analysis<refer to table %2
in anne'ure *= Bartlett test of sphere city is the statistical test for overall
significance of all correlations with in a correlation matri' "t also judges the
appropriateness of factor analysis
6-8-1 FACTOR EBTRACTION
The table %6<refer to anne'ure *= shows the variance e'hibited by
e'tracted factors "t shows that the first factor accounts for highest amount of
variance( the second factor accounts for second highest and so onThe principal
components analysis using Mvarima' rotation method of correlation matri' of
the %% variables have led to the e'traction of seven broad components of
dividend policy of the corporate "ndia These factors accounted for
%9U($%U($9U($9U(2U(2U and ,U of the total variance e'plained(
respectively Accordingly( these factors have been labeled as 0actor of dividend
signaling and profitability( 0actor of li#uidity ratios and systematic risk( 0actor of
firm si>e( 0actor of agency conflicts and ownership( 0actor of cash flow
#uality and dividend stability( 0actor of growth and e'pansion and 0actor of
longterm solvency
Dayananda sagar college of arts, science and commerce Page 65
IMPACT OF DIVIDEND POLICY ONSHAREHOLDERS VALUE: A STUDY OF INDIANFIRMS
FIGURE 8: SCREE PLOT
S,r&& P+'$
,
+
Eigenvalue
*
%
$
9
$ % * + , / 2 6 7 $9 $$ $% $* $+ $, $/ $2 $6 $7
%9
%$ C'3?'"&"$ N%3C&r
Dayananda sagar college of arts, science and commerce Page 66
IMPACT OF DIVIDEND POLICY ONSHAREHOLDERS VALUE: A STUDY OF INDIANFIRMS
An elbow in the scree plot indicates the point at which the inclusion of
additional factors does not contribute significantly in e'plaining the variance in
data set 0actors above the elbow of the plot are taken The procedure involves
certain amount of subjectivity( if no clear elbow appears in the curve The
:cree plot shown below shows a clear elbow at 0actor 2 These seven factors
account for about 2%U of the variations in the data 8onse#uently these
seven 0actors are retained in the analysis
Dayananda sagar college of arts, science and commerce Page 67
IMPACT OF DIVIDEND POLICY ONSHAREHOLDERS VALUE: A STUDY OF INDIANFIRMS
6-8-. REGRESSION RESULTS OF EBTRACTED FACTORS
The regression results are highlighted in the Table %7<refer to the anne'ure
*= Put of 2 factors / factors have statistically significant regression
coefficients Pnly one 0actor ie0actor of dividend signaling and profitability
has statistically insignificant regression coefficient at ,U significance level
0actor * and , have e'actly opposite signs as established by previous
research studies The value of Adjusted )
%
is /,2 which indicates that these
factor combined together e'plain //U of the dividend payout pattern of "ndian
:ervice sector The 0 values are also found to be significant at ,U level of
significance
6-8-8 FINDINGS
A set of / factors out of 2 are found to be significantly related to D4 ratio This
shows that capital gains are preferred to cash dividends The regression
results have indicated a negative and significant relationship between D4
ratio and 0actor of li#uidity( firm si>e( growth 5 e'pansion &owever( the
0actor of longBterm solvency is significantly positively related Thus "t can be
said( :maller firms tend to pay more dividends in order to allure shareholders
and compensate them for risk involved %+ 0irms in :ervice sector prefer to
retain funds whenever any future investment opportunity is foreseen for further
growth and e'pansion
Dayananda sagar college of arts, science and commerce Page 68
IMPACT OF DIVIDEND POLICY ONSHAREHOLDERS VALUE: A STUDY OF INDIANFIRMS
D- QUADRATIC POLYNOMIAL REGRESSION ANALYSIS AND
FINDINGS
D-1 ANALYSIS OF IT SECTOR
Tables *$ and *%<refer to anne'ure += show the hypothesis testing results of
the different models Table*$ show that the null hypothesis &
9
$ K O
O
% X 9 and
&
9
% K O
O
%X O
O
% X9 are rejected The 0B test results show that both firm
and time effects are present in the data-angrange Multiplier test statistics
presented in the Table *% indicate that either the fi'ed effect firm and firm and
time models or the random effects firm and firm and time models are to be
preferred to 8lassical -inear regression model&ausman specification test results
presented in this Table*% conclude to prefer random effect model to fi'ed
effect model But we restrict our interpretation to fi'ed effect firm and time models
<two way=
Table **<refer to anne'ure += depicts the results from 0i'ed effect firm
Model estimation assuming nonBmonotonic relationship between regressors and
regressand Table *+<refer to anne'ure += shows regression results of 0i'ed effect
two way model Model represented in Table *+ assumes linear relationship
between D4 ratio and ownership variables and Table *,<refer to anne'ure +=
depicts the 0i'ed 1ffect firm and time effects results of #uadratic polynomial
model
Dayananda sagar college of arts, science and commerce Page 69
IMPACT OF DIVIDEND POLICY ONSHAREHOLDERS VALUE: A STUDY OF INDIANFIRMS
"n Model " none of the variable is found to be significant at ,U and $9U
respectively 0 statistics also show that this model does not fit well in "T sector in
"ndia "n Model "" only debt e#uity and time effects are found to be significant
The regression coefficient of debt e#uity ratio is positive at $9U level of
significance This implies that conflicts of interest do not e'ist between debt
holders and shareholders and they do not consider dividend payment a way to
e'propriate their value
A panel data #uadratic polynomial regression analysis increased the value of
Adjusted ) s#uare to *$U from $6U in Model """ &owever none of the
regression coefficient is significant at ,U level of significance The regression
coefficients of "nstitutional holding areL6$*9 in level and %+*+2+2 in s#uare
These coefficients are significant for confidence interval of 79UThis implies a
non monotonic <inverted U shaped= relationship between "nstitutional holding and
dividend payout of "T firms in "ndia 0or $9U level of significance the
regression coefficient of debt e#uity ratio is also found to be positive and
significant The 0 values are also significant at ,U level of significance <refer to
anne'ure +=
Thus the results show that relationship between institutional ownership and
dividends is nonBlinear Ghen institutional ownership is low( an increase in
ownership percentage tends to reduce agency costs At the margin( as agency
costs falls( cash dividends become less desirable as a tool for further reducing
agency costs( and thus dividends tend to decrease At( high levels of institutional
ownership agency costs tend to rise with further increases in ownership percentage
Dayananda sagar college of arts, science and commerce Page 70
IMPACT OF DIVIDEND POLICY ONSHAREHOLDERS VALUE: A STUDY OF INDIANFIRMS
and the increased scrutiny placed on the firm by higher dividends become
necessary Thus (dividends are e'pected to decrease over range of "nstitutional
ownership( and increase after the point of entrenchment indicating a parabolic
relation
Dayananda sagar college of arts, science and commerce Page 71
IMPACT OF DIVIDEND POLICY ONSHAREHOLDERS VALUE: A STUDY OF INDIANFIRMS
D-. ANALYSIS OF FMCG SECTOR
The techni#ue of panel fi'ed effect firm and time model has been applied
for analysis 0 Test results indicate the presence of firm and time effects Table
*7<refer to anne'ure += depicts the results from 0i'ed effect firm Model
estimation assuming an inverted U shaped relationship between regressors and
regressand Table *7 shows regression results of 01 firm and time estimations
assuming linear relationship between D4 ratio and ownership variables and
Table *$9 depicts the 0i'ed 1ffect firm and time effects results of #uadratic
polynomial model&ausman Test shows random effect model should be preferred
to fi'ed effect firm and time model Therefore Table *$$ reports PneBway
random group effect model using Neneralised -east s#uare Model
The 0N-: estimates show that corporate holding is only the
significant ownership determinant of D4 ratio A significant negative
relationship has been observed between corporate holding and dividend payout
ratio The results of Model " shows promoter holding have a positive
relationship with dividend payout in level and negative in s#uare This implies at
lower level of promoter holding( Dividend payments are high but gradually as
their holding increases they prefer lesser dividend distribution 0inally according
to the results of Model """ corporate holding has negative significant relationship
with D4 ratio in level and promoter holding has negative statistically significant
relationship with Dividend
payout in s#uare &owever we do not consider the results of
Model """ in data interpretation as &ausman test suggests use of )andom
effect Model over this model Thus( it can be stated there is no significant
influence of ownership pattern on dividend payout ratio in the 0M8N sector
Dayananda sagar college of arts, science and commerce Page 72
IMPACT OF DIVIDEND POLICY ONSHAREHOLDERS VALUE: A STUDY OF INDIANFIRMS
Dayananda sagar college of arts, science and commerce Page 73
IMPACT OF DIVIDEND POLICY ONSHAREHOLDERS VALUE: A STUDY OF INDIANFIRMS
CAARs/AARs
D-8 ANALYSIS OF SERVICE SECTOR
Table*$+ <refer to anne'ure += depicts the results from 0i'ed effect firm
Model estimation assuming nonBmonotonic relationship between repressor
and regress and Table*$, portrays regression results of fi'ed effect twoBway
model assuming linear relationship between dependent and independent
variables Table *$/ depicts the 01 firm and time affects results of #uadratic
polynomial regression model
&ousman test results show that fi'ed effect firm and time model is preferred to
random effect model Model " <Table *$+= and "" <Table *$,= do not fit well as
0 values are not significant Also the results presented in these two models are
corroborated by the results of fi'ed effect firm and time model estimations The
results of Model """ <Table *$= show 0"" negative relationship in level Ione of the
other ownership variable is significant at ,U and $9U level of significance
Thus( it can be said that in :ervice sector -inear relationship holds well 0""
holding (
corporate holding and promoter holding are inversely related to D4 ratio
Dayananda sagar college of arts, science and commerce Page 74
IMPACT OF DIVIDEND POLICY ONSHAREHOLDERS VALUE: A STUDY OF INDIANFIRMS
E- EVENT STUDY: ANALYSIS AND FINDINGS
E-1 "T SECTOR
FIGURE 9: Average abnormal and 8AA)s of %9 "T companies over a window
period starting from day L%9 to day Z%9 relative to dividend announcement day <9B
day=
,
B$9 AA)s
8AA
B%9
B%,
Dayananda sagar college of arts, science and commerce Page 75
IMPACT OF DIVIDEND POLICY ONSHAREHOLDERS VALUE: A STUDY OF INDIANFIRMS
The results show that abnormal returns are negative during the entire
window period e'cept for the dividend announcement day Though small
positive abnormal returns are generated on dividend announcement but they are
not statistically significant at $9 U This shows that dividend announcements
do not contain signaling effect in this sector Pther reason may be that
investors prefer other modes of dividend distribution to cash dividends
Dayananda sagar college of arts, science and commerce Page 76
IMPACT OF DIVIDEND POLICY ONSHAREHOLDERS VALUE: A STUDY OF INDIANFIRMS
7
9
11
13
15
17
19
1
CAARs/AARs
E-.- EVENT STUDY OF FMCG SECTOR
FIGURE : . AA)s and 8AA)s of $, 0M8N companies over a window period
starting from day L%9 to day Z%9 relative to dividend announcement day <9B day=
*9
%9
$9
9
-19
-17
-15
-13
-11
-9
-7
-5
-3
-1
1
3
5
B$9
B%9
Dayananda sagar college of arts, science and commerce Page 77
IMPACT OF DIVIDEND POLICY ONSHAREHOLDERS VALUE: A STUDY OF INDIANFIRMS
B*9
B+9
8AA)s
AA)s
Dayananda sagar college of arts, science and commerce Page 78
EVENT WINDOW
&uge abnormal returns are created * days post dividend announcement and
sustained till $6
th
day<refer anne'ure ,= in the event window with minor
fluctuations 4ositive Abnormal returns are also witnessed * days before dividend
announcements
E-8 EVENT STUDY OF SERVICE SECTOR
FIGURE 6 .AA)s and 8AA)s of :ervice companies over a window period
starting from day L%9 to day Z%9 relative to dividend announcement day <9B day=
CAAR1@AAR1 (F*
AA)s
8A
A)s
$
CAARs AND AARs
9,
EVENT WINDOW
Abnormal returns are generated prior to dividend announcements but are not
sustained as 8AA)s start falling 4ositive abnormal returns occur from 2
th
till
%
nd
day pre dividend announcement as shown in Table+*<results appended to
Anne'ure,
7- CONCLUSION
This study has tested empirically the agency cost theory( -intner model(
dividend signaling and smoothing effects using a framework of various
econometric models
Put of the chosen sectors -intner model fits well in the 0M8N sector signifying
dividend signaling and smoothing effects are present in this sector Thus these
firms follow stable dividend payments year on year basis( even though earnings
might change dramatically The findings in the 0M8N sector are in alignment
with Brave etal that mangers are very reluctant to cut dividends once they are
initiated This reluctance leads to dividends that are sticky( smoothed from year
to year and tied to long run profitability of the firm &owever "T sector and
service sector demonstrate a pattern( which is seen in emerging economies like
Tunisia( Simbabwe and Turkey These sectors are characteri>ed by high target
payouts coupled with high speed of adjustment coefficient
Through the analysis of second objective it was found that there are sectoral
differences in corporate dividend policy determinants The results are
consistent with conclusion of Baker( 0arrelly( and 1delman <$76,= and &o
&orace <%99%= that firms industry type influence dividend policy A factor
which may be relevant for one industry becomes irrelevant for another
depending upon the "ndustry characteristics like growth phase( ownership
pattern( si>e( systematic risk and earnings variability
0M8N companies score high on dividend stability and consistency as -agged
dividend and 4AT are important factors governing dividend distribution The
#uality of cash flows( which is measure of li#uidity of the firm and firm si>e
are found be inconse#uential in determining the dividend payout The
opportunities for future growth and e'pansion are found to be negatively
related to dividend payout ratio -arger is the growth and investment
opportunities available to the firm( lesser is the incentive to pay dividends by
retaining larger proportion of profits The regression results also disclose
negative and significant relationship with )etained earnings and 8apital
1'penditure during the current year which is in conformity with the e'isting
literature A company which prefers retention of profits for financing the
capital e'penditure from internal resources distributes fewer dividends compared
to a firm which finances the investment e'penditure from e'ternal sources
Also larger the retention of profits by a company lesser is the dividend
distributed <4ecking order hypothesis= Thus( the e'tent to which the company
decides to finance 8A41@ from retained earningsK both retained earnings and
8A41@ would be negatively related to dividend payments The results
establish a negative relationship between li#uidity and Dividend payout ratio and
promoter holding Though systematic risk and earning variability obstruct the
stable dividend payout but the results report that Dividend 4ayout ratio is
positively related to risk Dividend 4ayout ratio is found to be significantly
positively related to longterm solvency of the firm The firms in 0M8N
sector operate with very low level of debt These firms are highly li#uid
firms( any increase in debt proportion in capital structure do not put pressure on
firms capacity to pay dividend A positive and significant relation has been
obtained between Debt 1#uity ratio and Dividend 4ayout ratio through the
results( which is consistent with 1asterbrook
analysis This positive relation can also be attributed to the fact that 0irms in
0M8N sector use dividends as a source to allay conflicts that may arise
between bondholders and shareholders with increase in Debt e#uity ratio
"t has been noted that "T sector score high on the dividend stability The
dividend paid during previous year is an important governing factor 0irms in
"T sector do not use dividends as a medium to signal their prosperity to the
shareholders This also reflects that there is lesser information asymmetry in this
sector "T sector is a human intensive sector and do not re#uire huge capital
asset base like manufacturing companies for their operations The major asset
of this sector is manpower The funds re#uired for recruitment and retention of
manpower is comparatively less than funds re#uired for purchasing capital assets
:o these firms can easily release funds for payment of dividends Also a
negative relationship between profitability can be attributed to the fact that agency
problems are not very relevant and thus Dividend payout as a monitoring
mechanism may be less needed The results demonstrate that 4romoters holding
in this sector also negatively influence the Dividend 4ayout ratio
A negative regression coefficient of 0actor of li#uidity ratio and Dividend
4ayout ratio can be attributed to the fact that in "T sector capital gains are
preferred to cash dividends &igher debts e#uity ratio and changeability in the
earnings per share may negatively influence the dividend payout of company
But in case of "T firms which are very low debt or >ero debt companies eg
"nfosys is a >ero debt company( these variables may not be an important
determinant of dividend payout Therefore( 0actor of financial and systematic
risk has not emerged as an imperative factor affecting the dividend payout
ratios of firms in "T sector
The results signify that :ervice companies do not score high on
dividend stability 4rofitability is not a primary determinant of dividend
payout though it is positively associated with Dividend payout ratio 0irms in
service sector do not use cash dividends to signal their prosperity to the
shareholders The results are in contrast to
the previous studies on banking industry( which state that banks use
their dividend history to set their dividend These results were established by
Dickens I)oss and IewmanA?oseph in their study EBank Dividend policy.
e'planatory factorsF and 4al Raram and Noyal 4uja E-eading determinants of
Dividend policy. A case study of the "ndian Banking "ndustryF Their study
displayed that stable dividend policy is followed by "ndian banking industry as
lagged dividend emerged as the most significant determinant of dividend payout
A negative relationship between systematic risk and earnings variability
potray that higher the earnings variability lower will be dividend paid by the
companies in :ervice sector The results also highlight that promoter holding is
positively related to dividend payout The results also show that there is a
negative relationship between growth and investment opportunities and dividend
payout ratio This result is in alignment with pecking order hypothesis
0irms with high leverage are those whose value shifting is potentially
costly :uch firms are e'pected to pay large dividends Also low leverage
firms are high growth firms Therefore they pay low dividends This positive
relationship between debt e#uity and dividend payout in :ervice sector can be
considered consistent with 1asterbrook Analysis
A finding in the :ervice sector that refutes the e'isting literature is a
negative relationship between firms si>e and the dividend payout ratio This
finding is not in agreement with 4ecking order hypothesis and stands in sharp
contrast with results of :mith and Gatts <$77%= -arger companies despite
having the opportunity to tap easily the financial markets by issuing stocks
or bonds prefer to retain dividends so as to avoid the costly e'ternal financing
Moreover( small firms( which are more risky( need to have a high payout
ratio( in order to attract investors to but their stocks
The analysis of third objective demonstrates that the influence of
ownership pattern on the dividend payout is heterogeneous "t has been observed
that there are sectoral differences in impact and influence of ownership groups on
dividend payout "ndia is a common -aw country characteri>ed by strong
investor protection and dispersed ownership <the role of the insider is played by
the manager=( hence the agency conflicts are not so severe and cash dividends
may not be essential to mitigate the agency conflicts According to -aporta etal
greater the investor protection in a country dividend payouts tend to be higher "n
the "T sector results are consistent with Manager 1ntrenchment hypothesis
depicting that institutional holding regression coefficient is positive in level and
negative in s#uare This implies that upto a certain threshold
$%
( dividends
act as substitute for corporate governance After the threshold the direct
monitoring efforts of institutional holders are insufficient or become too costly
Therefore( dividend payments are increased so that managers are forced to
raise finance from e'ternal capital markets and acts as an e'ternal monitoring
device These results are in agreement with the findings of :hort( Shang and
Reasey %99%( and 0arinha( %99*= A nonBmonotonic and parabolic relationship
has been established by the research in "T sector for the period under study
&owever in the 0M8N sector none of the ownership groups were found to
have considerable influence on dividend payout As regards :ervice sector( the
relationship between dividend payout and various ownership groups ie 0""(
8orporate holding and
"nstitutional holding has been found to be linear These ownership
groups negatively impact dividend payout of the companies supporting the
hypothesis that dividend payments are the means to alleviate the agency conflicts
These ownership groups act as monitoring device reducing the need of high
dividend payments
Through the analysis of the fourth objective it has been found that cash
dividends may not always create abnormal returns for the shareholders "n the
modern scenario a gradual drift to other modes of payment of dividends has
been observed :mall abnormal returns on dividend announcement can also be
attributed to the fact that the dividend announced is below the investors
e'pectations More so( dividend income( being a marginal constituent in
investment return( may not inspire much to the over enthused investors in rising
capital markets The findings of the research highlight that in 0M8N sector
investors respond positively to cash dividends announcements whether increasing
or decreasing Thus( 8ash dividends are welcomed by the investors in this
sector This implies that signaling
?ayesh Rumar in his study on association between corporate Novernance and
dividend payout identified this threshold level to be %, hypothesis holds &owever
abnormal returns are created in service sector but they are not sustained over the
event window and gradually 8AA)s <8umulative average abnormal returns=
become negative 0inally it can be stated that dividend announcements create
shareholders wealth in the 0M8N and :ervice sector Thus( the investors tend to
applaud the dividend announcements
&owever( inspite of the fact that managers view dividend decisions as
important it cannot be concluded that market rewards a carefully managed
dividend policy with higher share price"n "ndia financial managers typically view
dividend decisions as an important part of their job The typical firm does not
follow a residual policy nor leave its dividend payout to chance )ather( firms
manage their dividends as proposed by -intners model and partially follow
stable dividend policy
10- CHAPTER PLAN
The study is organi>ed in following nine chapters.
C>#?$&r 1: I"$r')%,$'"
This chapter introduces the dividend pu>>le "t throws light on the theoretical
background( genesis( concept and meaning of dividends The primacy and
importance of dividend decision has also been discussed in this chapter
C>#?$&r .: L$&r#$%r& R&5&G
8hapter two reviews the literature in detail and discusses the various research
studies on the topic under study
C>#?$&r 8: R&1&#r,> M&$>')'+'!0
8hapter three traces the research methodology and discusses in detail the various
models developed( tools and techni#ues used for analy>ing the research objectives
C>#?$&r 9: O5&r5&G '( $>& ")%1$r0
This chapter gives brief overview of the financial performance( growth
prospects( characteristics of the various sectors under study
C>#?$&r :: D#$# #"#+011: L"$"&r )5)&") 3')&+
This chapter covers the empirical analysis of the -intner model proposed by ?ohn
-intner <$7,/= in the three sectors under study The chapter highlights the target
payout ratios and speed of adjustment coefficients of each sector respectively using
panel data analysis
C>#?$&r 6: D#$# #"#+011: C'r?'r#$& )5)&) ?'+,0 )&$&r3"#"$1
8hapter si' contains the analysis and findings of factor analysis( which is used to
develop model of corporate dividend policy determinants in each of the sector
respectively
C>#?$&r D: D#$# #"#+011: I3?#,$ '( 5#r'%1 'G"&r1>? !r'%?1 '" )5)&")
?#0'%$ r#$'1
8hapter seven discusses in detail the data analysis and findings of #uadratic
polynomial regression analysis This model has been developed to find the
impact of various
Pwnership groups on the dividend payout ratios in all the three sectors
undertaken for study
C>#?$&r E: D#$# #"#+011: E5&"$ 1$%)0
This 8hapter unfolds the impact of dividend announcement on shareholders
wealth as reflected by the shareprices through the use of most sophisticated
techni#ue in 8orporate 0inance ie 1vent study
C>#?$&r 7: C'",+%1'"
This chapter summari>es and concludes the research Areas for future research
are also discussed in this chapter
SELECTED REFERENCES
$ Agarwal ) I <$77$=( EDividends and :tock 4rices. A 8ase :tudy of
8ommercial Jehicle :ector in "ndia $7//B/2 to $76/B62F( Indian
ournal of !inance and "esearch(Jol $( pp /$B/2
% Aiva>ian Jarouj( Booth -aurence( 8leary :ean< %99*=( EDividend policy
and the organi>ation of capital marketF( ?ournal Pf Multinational 0inancial
Management (pp$9$B$%$
* Aiva>ian( J( and Booth( -<%99*=( EDo emerging firms follow different
dividend policies from U: firmsHF( ?ournal of 0inancial research( Jol%/
Io*(pp*2$B62
+ Allen 0ranklin and Michael )oni<April %99%= (4ayout 4olicy( IorthB&olland
handbook of 1conomics and 0inance( Gharton 0inancial "nstitutions center
, Amidu( Mohammed( and( Abor ?oshua <%99/=( EDeterminants of Dividend
payout ratios in NhanaF ( The ?ournal Pf )isk 0inance( Jol 2 Io% ( pp$*/B$+,
/ As#uith( 4 and Mullins( D <$76*=( MMThe impact of initiating dividend
payments on shareholders Mwealth( ?ournal of Business( Jol ,/( pp 22B7/
2 Baker( &R <$777=( EDividend 4olicy issues in regulated and unregulated
firms. a managerial perspectiveF( Managerial 0inance( Jol%, Io/( pp$B$7
6 Baker(&Rent( Narry 14owell( Theodore 1 Jeit <%99$=( E0actors
influencing dividend policy decisions of Iasda# firmsF( The 0inancial )eview( pp
$7B*6
7 Baker(&Rent(and Narry 14owell(<%999= (EDeterminants of corporate
dividend policy. a survey of IT:1 firmsF ( 0inancial 4ractice and education 7
(pp%7B+9
$9 Black( 0ischer( EThe Dividend 4u>>leF( The ?ournal of 4ortfolio Management(
winter $72/( pp/*+B/*7
$$ Brown : ? and Garner ? B <$769=( EMeasuring :ecurity 4rice 4erformanceF(
ournal of !inancial #conomics( Jol 6( pp %9,B%,6
$% D:ou>a( ?<$777=( EAgency cost ( market risk( investment opportunities and
dividend policyBan international perspectiveF( Managerial 0inance( Jol%,
Io/(pp*,B+*
$* Dillion and Nolstein ($76+= Alli - Rasin( Rhan 3ayyum()amire> N
Nabriel ( EDeterminants of 8orporate Dividend 4olicy . A factorial
AnalysisF( The 0inancial )eview ( Jol%6Io+ ( Iovember $77*
$+ 0arrelly( Nail 1 and & Rent Baker( )ichard B 1delman<$76/= (
E8orporate Dividends .Jiews of 4olicy makersF ( Akron Business and 1conomic
review,$2.+ <Ginter= ( pp /%B2+
$, 0rankfurter( M( Neorge and Good Bob (N < %99*=( EDividend 4olicy
Theory and 4racticeF( Academic 4ress
$/ &iggins()8<$72%=( EThe corporate dividend Lsaving decisionsF( ?ournal of
0inancial and 3uantitative Analysis( Jol2 Io%( pp$,%2B+$
$2 ?ensen( M <$76/=( EAgency costs of free cash flow( corporate finance and
takeoversF( American 1conomic)eview( vol 2/( pp *%*L7
$6 ?ensen( M and Meckling( G <$72/= ETheory of the firm. managerial
behaviour( agency costs and ownership structureF( ?ournal of 0inancial 1conomics(
vol *( pp *9,L/9
$7 Revin( : <$77%=( Dividend 4olicy. An analysis of some determinants( 0inance
"ndia( vol/( Io%
%9 -a 4orta( )( -ope>Bde :ilanes( 0( :hleifer A and Jishny( ) <%999=(
EAgency problems and dividend policies around the worldF( ?ournal of 0inance(
vol ,,( pp $L**
%$ -ease( 8()onald( ?ohn Rose( Ralay Avner(-oewenstein Uri(:ari# &Pded
<August %99$=(FDividend 4olicy. "ts "mpact on 0irm JalueF( 0inancial
management Association :urvey and synthesis series( &arvard Business :chool
4ress(
%% -inter( ? <$7,/=( EDistributions of incomes of corporations among dividends(
retained earnings and ta'esF( American economic )eview( +/ <$=( pp 72B$$*
%* Miller( M& and Modigliani( 0 <$7/$=( MMDividend policy( growth( and the
valuation of shares(?ournal of Business( Jol *+( pp +$$B**
%+ Iaceur Ben :amy( Noaied Mohamed( Belanes Amel<%99/=( EPn the
Determinants and Dynamics of dividend policyF( "nternational )eview of
0inance(Jol/(pp$B%*
%, Pffer( A and Thakor( D <$762=( MMA theory of stock price responses to
alternative corporate cash disbursement methods. stock repurchases and
dividends( ?ournal of 0inance( Jol +%(pp */,B7+
%/ 4al Raram( Noyal 4uja <?uly L December %992=( E-eading determinants of
Dividend policy. A case study of "ndian Banking "ndustryF( Decision( Jolume
*+ (Io%
%2 4etit( ) <$72%=( MMDividend announcements( security performance and capital
market efficiency ?ournal of 0inance( Jol %2( pp 77*B$992
%6 4ruitt( :G and Nitman( -G <$77$=( The interactions between the
investment( financing( and dividend decisions of major U: firms( 0inancial
review( Jol%/ Io**( pp+97B*9
%7 )odrigue>()?<$77%=( E3uality dispersion and the feasibility of dividends as
signalsF( ?ournal of 0inancial research( Jol$,(pp+$$B**
*9 )o>eff( M <$76%= (FNrowth( beta and agency costs as determinants of
dividends payout ratiosF( The ?ournal of 0inancial )esearch( vol ,6<*=( pp %+7B,7
*$ Golmaran( &4 <%99*=( EDoes -intner Model e'plain :outh African
Dividend paymentsF( Meditari Accountancy )esearch( Jol $$( pp %,,B,6
*% Iarasimhan( M: and 8 Asha <$772=( E"mplications of Dividend Ta' on
8orporate 0inancial 4oliciesF( The "80A" ?ournal of Applied 0inance( Jol *(
Io%( ?uly( pp $$B%6
** Iarasimhan( M: and : Jijayalakshmi <%99%=( E"mpact of Agency 8ost on
-everage and Dividend 4oliciesF( The "80A" ?ournal of Applied 0inance( Jol 6(
Io %( March( pp $/B%,
*+ Brealey( )A <$77+=( EDoes Dividend 4olicy MatterHF in :tern( ?M and D&
8hew <eds=()evolution in 8orporate 0inance( %nd edition( Blackwell
4ublishers "nc( 8ambridge( Massachusettes
*, Mahapatra( )4 and 4R :ahu <$77*=( EA Iote on Determinants of
8orporate Dividend Behaviour in "ndia L An 1conometric AnalysisF(
Decision( Jol %9( Io $( ?anuaryBMarch( pp $B%%
*/ Mohanty( 4 <$777=( EDividend and Bonus 4olicies of the "ndian
8ompaniesF( Jikalpa( Jol%+( Io+( PctoberBDecember( pp *,B+% Mahakud
?itendra <%99,=( E:hareholding 4atterns and Dividend policy . 1vidence from
"ndian 8orporate :ectorF("80A" ?ournal of Applied 0inance
*2 Baltagi (B& and ?M Nriffin <$766=( EA generalised 1rror 8omponent
Model with &eteroscedastic DisturbancesF( "nternational 1conomic )eview (Jol
%7 (pp2+, L2,*
*6 :chellenger & Michael( Good D David( Tashakori Ahmad <$767=(F
Board of Director 8omposition( shareholder Gealth( and Dividend 4olicy (?ournal
of Management ( Jol$,( Io*( +,2B+/2
*7 &sin Dar 8hen (-u &siangB&is ( &uang cheng LTing<%997= ( EThe
Announcement 1ffect of 8ash Dividned 8hanges on :hare pricesB An 1mpirical
Analysis of 8hinaF ( The 8hinese 1conomy ( Jol+% (no$ (?an L0eb %997 ( pp /%B
6,
+9 T:ubba )eddy and )ath :ubhrendu<%99,= ( E Disappearing dividends in
1merging MarketsH 1vidence from "ndia( 1merging Markets 0inance and trade(
Jol +$( no /( IovBDec %99,(pp ,6B6%
+$ Anand Manoj<%99$=( E0actors influencing dividend policy decisions of
8orporate "ndiaF( "80A" ?ournal Pf Applied 0inance(%99+ (Jol@ (Io%
+% :harma Dhiraj <%992=( E Are dividends in vogue in "ndiaH An 1mpirical
study of :ense' 8ompaniesF("80A" ?ournal of Management )esearch
((JolJ" (Io*( March %992pp %%B*9
PUBLICATIONS AND PAPER PRESENTATIONS
$ Rapoor :ujata (E)elevance of information asymmetry models in "ndian
conte'tF "ndian ?ournal Pf 0inance( "ssue $<?uneB ?uly 92issue=
% Rapoor :ujata (EDividend policy L A reviewF Management review( a
#uarterly journal by Delhi Management association( :eptember (%992
* Rapoor :ujata( Anil Ranwal E)elevance of "nformation asymmetry
ModelsBa study of "ndian "nformation technology :ectorF( "nternational ?ournal Pf
Business research and presented same paper at "AB1B%992 international
conference held at -as Jegas "n Pctober92
+ Anil( Ranwal( and Rapoor(:ujata( Determinants of Dividend 4ayout
ratiosB A study of "ndian "nformation Technology sectorK "nternational )esearch
?ournal Pf 0inance and 1conomics( "ssue $,( May %996( pp /*B2$( %996
, Ranwal( Anil( and Rapoor(:ujata( )elevance of :ignaling and
:moothing approaches to dividendB A study of "ndian 0M8N sector(
Business 4erspectives(Jolume$9(Io$ <?une %996 issue=
/ Ranwal( Anil( and Rapoor(:ujata( )elevance of "nformation Asymmetry
Dividend modelsB A study of "ndian 0M8N sectorK 1ffulgence (?uly %996
2 Rapoor(:ujata(Ranwal( Anil and Abidi( Iaseem(Determinants of dividend
payout ratiosB A study of "ndian 0M8N sector( The "nternational 8onference on
Nlobal "ssues in Business and Technology( Iational "nstitute of 0inancial
Management( "ndia( August %996
6 Ranwal( Anil( and Rapoor(:ujata( )elevance of :ignaling and
:moothing Approaches to Dividend. A study of "ndian "T :ector( Asia
4acific Business )eview( Jolume "J Io+(PctoberBDecember%996
7 Rapoor(:ujata(Ranwal( Anil( Determinants of 8orporate Dividend 4olicy in
"ndian "nformation Technology :ector. A factorial Analysis( "nternational
0inance 8onference (""M 8alcutta (December %997