Professional Documents
Culture Documents
com
vol.2, No. 3(s), pp. 1034-1042 ISSN 1805-3602
Abstract Introduction
The purpose of this study was to investigate the One of the basic products of accounting system is
effect of dividend earnings on the quality of earnings the profit and loss statements. This financial statement
with emphasis on the type of industry. In the study includes a financial key element namely net profit and it
earnings quality have been analyzed in four dimen- is expected that the net profit shows the performance re-
sions including net value of discretionary accruals, sult of a economic unit for a financial period and is one
accruals quality (quality of financial reporting), in- of the items that are considered by a wide range of users
come smoothing and association of earnings to the of financial statements.The accrual system is used in the
stock value. Also, two models of Eckel and TZ were profit measurement process. Thus, the accruals adjust
used for separation of smoothers from non-smoothers the recognition of cash flows during the period until the
companies. The present study has applied aim and in adjusted figures measure the performance of economic
terms of methodology refers to scientific-correlation unit better. For example, registration of a receivable ac-
research type. The statistical population of this study count accelerates the future cash flow in profit and co-
were the accepted companies in Tehran Stock Ex- incide the identification of accounting profit with the
change in 5 industries including food, automotive, economic benefits obtained from the sale. Paying sta-
chemical, pharmaceutical and cement that 100 com- tusof dividend earnings is associated with: (1) Net value
panies have been located in statistical sample of study of discretionary accruals, (2) Accruals quality (quality of
by systematic elimination sampling method. Thetime financial reporting), (3) Income smoothing and (4) As-
period of study was among 2005 to 2011 but the data sociation of earnings to stock value
of year 2004 were also used to calculate some research
variables. In this study, the ordinary least squares re- Theoretical Principles and background of study
gression and also specific regression of each company
were used in time series and annual cross-sectional Looking for Dechow and Schrand (2004), earn-
to calculate the some of the research variables. Mul- ings is focused as a criterion of the companys per-
tiple regression and logistic regression were used to formance and argues that earnings with high qual-
test the hypotheses of this study. The research results ity must have three characteristics: 1.The operational
indicated that dividend earnings had positive impact performance of the company be reflected accurately,
on the net value of discretionary accruals and income 2. Providing a good indicator of the future operational
smoothing in both models (TZ and Eckel) and a sig- performance and 3. Serve is as a useful summary for
nificant inverse impact on accruals quality (quality of assessing value of the company. Earning quality crite-
financial reporting). Also, no significant relation was rions is commonly designed in these three characteris-
observed between dividend earnings and association tics and when considers together, pays attention to the
of earnings to the stock value. support of the association between dividend earnings
and quality of earnings. Dividend earning contains in-
Keywords: dividend earning, earnings quality, formation about the future prospects of the companys
accruals, income smoothing profit. Companys managers are reluctant to increase
Corresponding author: Mahdiye Ebrahimpour, Science and Research Branch, Islamic Azad University,
East Azarbaijan, Tabriz, Iran. Email: mahdieh.ebrahimpour @ yahoo.com
Copyright Mahdiye Ebrahimpour et al., 2013
European Online Journal of Natural and Social Sciences; vol.2, No. 3(s), pp. 1034-1042
1034
Social science section
the profit of cash stock. Unless, they believe that they the criterion of existence of exceptional items, com-
can maintaina new level of the cash stocksprofit in panies with higher earning quality pay higher divi-
the future. Therefore, the regular distribution of cash dend earnings. Also, results of this research showed
stocksprofit reports from the sustainability of future that the applied criteria to measure earning quality
profits and it is assumed that the dividend earning con- have no significant relationship with dividend earn-
tains information about earnings quality. The profit of ing changes.
cash stock may also contain information about future Nourvash et al (2009) in a study entitled Ex-
earnings changes and when acash stocksprofit chang- amining the effects of accruals onearning quality
es, this change inform from the futureearnings changes. of accepted companies in Tehran Stock Exchange
Since, companies undertake to pay cash stocksprofit, concluded that, in Tehran Stock Exchange,earning
particularly, since this profit has economic content quality reduced by reducing the accruals quality and
help investors to convince about thequality of reported consequently increasing the estimating error of ac-
profits. So, even if cash profit changes dont inform cruals, earning stability coefficient and so on.
about the future profits changes in reliable form assist
investors to convince that the reported earnings will be External research
maintained in future periods. This study is in line with Dechow and Dichev (2002) investigated the re-
study of Mojtahedzadeh et al (2009) and Tang & Bin lationship betweenthe accruals quality andearning
Miao (2010) intends to answer this question that how persistence. Their empirical criterion from the ac-
the impact of dividend earning on earnings quality is. crual quality was the regressions of changes in work-
ing capital of past and present, and the future cash
Literature review flows. They finally came to the conclusion that there
is a direct correlation between the quality of accruals
Internal research and earning persistence.
Khajouiand Nazemi (2005) in a study entitled Chan et al (2006) in a research entitled The
The investigation of the relation between earnings quality of earnings and stock returns concluded
quality and stock returns, with emphasis on the role that stock returns of companies with high accrual
of accrual accounting in Tehran Stock Exchange figures reduce in the period after the financial in-
reached to this conclusion that the average stock re- formation reporting. One interpretation of these
turns of companies are not influenced by the accrual results is that the returns of companies with low
rate and its related components. In other words, it earnings quality (i.e. companies with high accrual
can notbe accepted that there is a significant dif- figures) decline in period after the reporting because
ference between the average returns of companies investors realize the companies lower earnings qual-
that their accrual are reported to the minimum and ity andadjust stock prices accordingly. This subject
maximum rate. was done by separating the components of accrual
Tehraniand Zakeri (2009) in a study entitled and also classifying based on the discretionary and
Surveying the relationship between the earnings non-discretionary accruals and similar results were
quality and dividend earning of accepted companies obtained.
in Tehran Stock Exchange concluded that com- Markarion (2008) conducted a research on the
panies which pay profit have more profit stability in relationship between stock returns fluctuations and
future periods and also this impact is more specified income smoothing and concluded that the fluctua-
for companies with higher distribution ratio. Thus, tions risk has strong and inverse correlation with in-
dividend earning contains information about the come smoothing. Also, this negative relationship is
quality of reported earnings. stronger in high operational risk companies, small
Mojtahedzadeh et al. (2009), in another study, companies and bankrupt companies.
examined the relationship between the earning Tong and Bin Miao (2010) in a study entitled
quality and dividend earning of companies. The Is dividend earning associated with earnings
statistical population and time period consisted of quality?concluded that when the size of the divi-
85 companies during the 2002 to 2006. The results dend earning payment is larger, the relationship
showed that,with regard to the criterion of the re- between the paying status of dividend earning and
vised providing of profit and loss statement, com- earning quality is very strong. Generally, the results
panies with higher earnings quality do not pay the show that the paying status of dividend earningindi-
dividend earning on time and by taking into account cates companies reporting quality.
Research hypotheses dividend earning and those that do not pay dividend
earningsand service companies because their divi-
The main hypothesis dend earnings are not comparable to other compa-
The dividend earning has an impact on earning nies and hence, they were removed from sample.
quality. Thus, the statistical sample includes all companies of
the 5 industries of food, automotive, chemical, phar-
Sub-hypotheses maceutical and cement accepted in Tehran Stock
1. The dividend earning has an impact on the Exchange, which have the following conditions:
net value of discretionary accruals. 1. Until the end of March 2004 are accepted in
2. The dividend earning has an impact on accru- Tehran Stock Exchange.
als quality (quality of financial reporting). 2. For 8 consecutive years (2004 to 2011) the
3. The dividend earning has an impact on in- basic data required for this research have been pro-
come smoothing. vided to the Stock Exchange.
4. The dividend earning has an impact on the 3. Their fiscal period terminates to the end of
relevance of earning value. March to remove the effects of seasonal fluctuations.
4. Companies should notinclude the banks and
Methodology financial institutions (investment companies, finan-
cial intermediation, holdings and leasing), because
The selection of method depends on the research they have different structures of the strategic prin-
objectives and the nature of research subject and its ciples and financial disclosures.
executive facilities. So, when we can decide aboutthe By applying the above conditions, 100 companies
investigating and conducting a study that the nature of were included in the statistical sample of this research.
the study subject, its objectives and scope and extent The dependent variable: in this study,earnings
are determined. Therefore, it is necessary to define quality will be examinedin four dimensions as fol-
the research method clearlyand by describing and ex- lowing:
pressing, wecan specify and limit the implementation 1.In this study, the net value of discretionary
scope of the study. This study has an applied purpose accruals is measured by the modified Jones model,
and correlational-scientific nature. discretionary accruals is annually calculated by fol-
The statistical sample of this research consists lowing cross-sectional regression coefficientsesti-
of two groups includingthose companies that pay mating:
1
TACCi,t= 0 + 1 + 2 ( SALE i,t AR i,t ) + 3PPEi,t + i,t
ASSETi,t
TACC: total accruals divided by average of the cruals (financial reporting quality), provided
total assets model is also consistent with the idea that the
ASSET: average of the total assets accrual accounting improves the detection of
SALE: sale changes cash flows over time. So, the adjusted numbers
AR:receivableaccounts changes (e.g. profit) better reflect the current operating
PPE:isthe gross asset of the property and equipment performance and offers good indicator of future
that all be the same scaled by average of total assets. performance.
2. Also, the criterion of Dechow and Di- Reporting quality is estimated by the following
chev(2002) is usedto assess the quality of ac- model:
CACCi,t = 0 + 1CFOi,t1 + 2 CFOi,t + 3CFOi,t+1 + 4 SALEi,t + 5PPEi,t + i,t
CACC: Current accruals model and when it is lower, it expresses the higher
CFO: Cash Flows earning quality. For coordinationof calculations
SALE: changes in sale with the title of variable, the calculated numberis
PPE: is the gross asset of the property and equipment multiplied in - 1.
All variables become same scaled by average to- 3. Income smoothing: in this study, income
tal assets at the end of period. smoothing has been measured by two models: (1)
The criterion of accruals quality was the stan- Eckelindex (2) TZ criterion and the smoothing and
dard deviation of the residual error of this regression non-smoothing companies are separated.
According to the results of first hypothesis test- suitablefor predicting. The results of the above table
ing which is presented in Table 1, the significance show the eigenvalueand status indicator are in situa-
level of statistical F (000/0) was less than the accept- tion that confirms the utilization of regression. Due
able error level (5%) and the total regression model to the low (P-Value) of the t-statistic of the accept-
is significant. DurbinWatson statistic (1.909) is able error level for the coefficient 1, the test re-
located in the interval between 1/5 and 2/5.There- sults show that the dividend earning has positive and
fore, there is no correlation between the error com- significant impact on the net value of discretionary
ponents model.Investigating the correlation be- accruals in reliability level of 95%. The results of
tween the independent variables indicates that the study also show that among the control variables,
eigenvalueandstatus indicator is less than 15.What- sales growth and rate of assets return has positive
ever the status indicator be lower, and whatever the and significant effect on the amount of discretion-
eigenvaluebe larger and closer to 1, regression is ary accruals. The coefficient of determination and
the adjusted coefficient of determination indicate The second hypothesis test results
that the independent and controlvariables entered The second hypothesis states thatthe dividend
into the regression can explain6.9% of the depen- earning has an impact on accruals quality (quality of
dent variablevariability. financial reporting).
According to the results of second hypothesis test- variables entered to the regression can explain 24.1%
ing which is presented in Table 2, the significance level of the dependent variable variability.
of statistical F (000/0) was less than the acceptable er-
ror level (5%) and the total regression model is signifi- The third hypothesis test results
cant. DurbinWatson statistic (1.505) is located in the The third hypothesis states:The dividend earn-
interval between 1/5 and 2/5. Therefore, there is no ing has impact on income smoothing.
correlation between the error components model. So, This hypothesiswas tested in two approaches. In
correlation survey between the independent variables the first case, the Eckel income smoothing modeland
indicates the eigenvalue and status index is less than in the second case, TZincome smoothing model was
15. Whatever the status index be lower, and whatever entered into the logit modelas the dependent variable.
the eigenvaluebegreater and closer to 1, regression is Following the provided model of company that is iden-
suitable for forecasting.The results of the above table tified smoothing accordingto the definition,number
indicate the eigenvalueand status index are in situation one and otherwise number zerohas been considered.
that confirms the utilization of regression. Due to the According to the results of the third hypothesis
low (P-Value) of the t-statistic of the acceptable error testing,in first case,which is presented in Table 3, the
level for the coefficient 1, the test results show that significancelevel of statistical chi square (0.000) was
the dividend earning has negative and significant im- less than the acceptable error level (5%) and the total
pact on the accruals quality in reliability level of 95%. regression model was significant and is indicator of
Also, results of study show that among the control vari- a good fit of the model. Due to the low (P-Value) of
ables, sales growth has positive and significant effect on the Wald statistic of the acceptable error level for the
the quality of accruals andin contrast, the ratio of book coefficient 1, the test results showed that the divi-
value to market value, rate ofassets return and indus- dend earning has a positive and significant impact on
try class has negative and significant impact on the qual income smoothing of Eckel model in reliability level
ity of accounting information. Also, the coefficient of of 95%.Also, the results of study showed that, among
determination and the adjusted coefficient of deter- the control variables, company size has a positive
mination indicate that the independent and control and significant effect on income smoothing of Eckel
model and in contrast, the ratio of assets return and coefficient of determination shows that the indepen-
industry class has a negative and significant impact on dent variables entered into the regression can explain
income smoothing of Eckel model. Also, the Pseudo 17.8% to 24% of the dependent variable variability.
Table 3. Results of the third hypothesis testing in the first case (Eckel model)
Table 4.Results of the third hypothesis testing in the second case (model TZ)
Variable name Coefficient () Wald statistic Sig
Fixed value 1.445 3.822 0.051
Dividend earning DIV (1) 0.583 10.083 0.001
Company Size Size (2) - 0.078 1.968 0.161
Book valueto market value B/M (3) - 0.309 3.590 0.058
Sales Growth Growth (b4) 0.153 0.839 0.260
Rate of assets returns ROA (b5) 0.294 0.161 0.688
Classification of Industry
IND (b6) - 0.335 22.276 0.000
(virtual variable)
statistic Log likelihood statistic Coefficients of determination
2 P-Value
All model of (Log-Likelihood) of Pseudo (Pseudo r-square)
Logistic Regression Cox&Snell= 0.050
35. 680 0.000 921.033
NagelKerke= 0.067
According to the results of the third hypothesis impact on income smoothing of TZ model in reli-
testing, in second case, which is presented in Ta- ability level of 95%.Also, results of study show that
ble 4, the significance level of statistical chi square among the control variables, industry classsales has
(0.000) was less than the acceptable error level (5%) negative and significant effect on income smooth-
and the total regression model was significant and is ing of Eckle model.Further, results indicatedthat,
indicator of a good fit of the model.Due to the low by increasing the acceptable error level to 10%, the
(P-Value) of the Wald statistic of the acceptable er- ratio of book value to market value has also a nega-
ror level for the coefficient 1, the test results show tive andsignificant impact on income smoothing of
that the dividend earning has positive and significant TZ model.Also, thePseudo coefficient of determi-
nation shows that the independent variables entered Results of the fourth hypothesis testing
into the regression can explain 5% to 6.7% of the The fourth hypothesis stated the dividend earn-
dependent variablevariability. ing has an impact on the relevance of earning value.