Asymmetric Dividend Smoothing in Listed Industrial Goods Firms in Nigeria: Analysis Based on Growth Potentials

Journal of Environmental Treatment Techniques

2020, Volume 8, Issue 1, Pages: 73-84

J. Environ. Treat. Tech.

ISSN: 2309-1185

Journal weblink: http://www.jett.dormaj.com

Asymmetric Dividend Smoothing in Listed

Industrial Goods Firms in Nigeria: Analysis Based

on Growth Potentials

Nuhu Abubakar

Department of Accounting, A.B.U. Business School, Ahmadu Bello University, Zaria, Nigeria

Received: 27/03/2019

Accepted: 05/08/2019

Published: 13/09/2019

Abstract

Literature on asymmetric dividend smoothing mostly focused on establishing the existence of the behavior within firms without

accounting for firm characteristic variables that influence dividend smoothing. Given that firms that smooth dividend payment more

tend to have low growth potentials and likely to be affected by severe agency conflict; therefore, asymmetric smoothing behavior is

expected to vary according to growth potentials. In this regards, the study aims at examining the determinants of asymmetric dividend

smoothing behavior after accounting for whether the behavior exists in firms with high/low growth potentials. The paper used data

obtained from a sample of 9 out of 16 Industrial Goods firms listed on the Nigerian Stock Exchange for the period of 11 years - 2006-

2016. The model of the study was estimated using Fixed Effect Model. The findings reveal that Industrial goods firms in Nigeria

smooth dividend payment and have asymmetric dividend smoothing behavior. The adjustment rate is not only asymmetric below and

above their Target Pay Out Ratio (TPR) but also asymmetric below TPR for both High and Low growth potentials. However, the

behavior only exists above TPR when the firms have high growth potentials. The paper concludes that firms with high growth potentials

have slower adjustment rates than the ones with low growth potentials when they are below their TPR.

Keywords: Target dividend payout; asymmetric dividend smoothing; growth potentials; industrial goods firms in Nigeria

investors put premium on firms with stable dividend policy

and punish those that decrease dividends.

Introduction

Dividend smoothing can be seen as paying dividends

Subsequently, Lintner findings have been upheld by

studies in the both the developed and developing countries. In

developing countries like Nigeria, where firms hardly adopt

constant payout ratio over long period of time probably due the

prevalence of economic instability and other related issues.

Ozo opines that despite the fact that Nigeria has different

economic environment, it has similar dividend setting process

to US and other developed countries (10,19). Firms are

conservative in setting their dividend policy and focused on

stability of earnings, current earnings and availability of cash

in determining their dividend policy. However contrary, to the

developed counties, Nigerian firm do not have target payout

rather they set a target dividend per share when determining

the magnitude of dividend level.

based on long run target payout ratio instead of current

earnings. According to Lintner who conducted an interview

with 28 managers from different companies in US identified

that, the main concern of managers is the stability of dividends

(

19). He argued that firms instead of setting dividends

periodically based on current earnings they rather first decide

on whether there is need to change the dividends or not. If they

consider the change to be necessary, then decide on how large

the change should be. Hence, they decrease dividends when it

happens to be the last option and increase dividends when they

are confident that the increase can be sustained. Mangers

believe that investors are unhappy with dividend decrease

because they use it to cater for their own needs and some

managers consider it as bad signal to future retained earnings

of a firm. Managers achieve stability through maintaining long

run target which they adjust their dividend policy gradually

toward it. Thus, two important points can be identified;

Earlier, Adelegan reported that, the issue of dividend

behaviour of firms was not given serious attention until the

work of Uzoaga and Aloziewa (4). The study and others that

Corresponding author: Nuhu Abubakar, Department of Accounting, A.B.U. Business School, Ahmadu Bello University, Zaria,

Nigeria. E-mail: aljazairyne@gmail.com.

Journal of Environmental Treatment Techniques

2020, Volume 8, Issue 1, Pages: 73-84

followed it such as Adelegan have confirmed findings of

Lintner and others that assessed the smoothness of dividend

based on factors other than that of Lintner, indicate that,

dividend smoothing should be an important factor in dividend

policy and Lintner variables are strong in explaining the

behavior (4). However, in the era of unclaimed dividend,

Moreover, Kighir cautioned regulatory authorities in Nigeria

on the activities of earnings management through dividend

smoothing by firms and described the behaviour as potential

danger to investors and government policies, if it persists in

the period of increasing incidences of unclaimed dividend.

Therefore, dividend smoothing has been established under

Nigerian economic environment (15,19).

Furthermore, Industrial Goods firms fall among the top

three sub sectors that strive to maintain stable dividend in

Nigeria over the years. More often than not, this may be

unconnected with the preferences of majority of shareholders

in Nigeria that are described as conservative who need cash

dividend and consider cash payment from dividend as means

of reducing high risks associated with expected future income

in the event of adverse economic conditions (“5 Years

Dividend Payment Review for NSE Quoted Companies”,

small/large companies and high/low profitable companies

have an asymmetric smoothing behaviour, his study suffered

some limitations (2). One, he measured dividend level as

dividend payout instead of dividend per share. It is believe that

per share value of dividend is a better factor of interest when

compared with dividend payout. Two, though he examined

smoothing based on the two adjustment modes and accounted

for some firm characteristics, the approach is still considered

inadequate in the context of Nigeria because it failed to capture

the effect of the Economic Recession that Nigeria experienced

from 2015 to 2016. In Nigeria, availability of cash and growth

potentials are important factors that could motivate dividend

smoothing among firms. Despite the availability of pointers to

firms’ dividend smoothing, previous studies have been

completely silent about it. The implication is that the studies

were not able to present a complete picture of dividend

smoothing of listed firms in the country.

In view of the fact that dividend stability is of greatest

importance to the overall success of firms and the fact that new

emerging country specifics are pointing to the need for further

studies in the area, it is imperative to undertake a study that

will consider some of these factors as they relate to the

industrial goods firms sector. This study therefore, is

motivated by the need to provide new evidence on the possible

factors that determine dividend smoothing of listed industrial

goods firms in Nigeria? Specifically the study seeks to offer

answer to the following question; Do firms with lowly growth

potentials adjust their dividend slower than the highly growth

potential ones in listed non-financial firms in Nigeria?

2016). However, due to the operating cost problem, the

banning of 40 raw materials from souring foreign exchange,

problem of poor infrastructure, the competitive ability of

manufacturing firms in Nigeria has been eroded and large

number of them are utilizing only about 20% of their capacity.

The interesting issue is that how the firms in the sector survive

the unfavorable economic environment when growth

potentialities are likely to be very low; compete with the other

sectors of the economy in maintaining relative cash dividend

stability, despite the cash flow problems that faced them.

Even though, studies all over the world have consistently

re-affirmed the robustness of Lintner model in explaining

dividend smoothing, the model was considered as being bias

due to the short comings of it assumptions such as; small

sample bias, constant response coefficient and in ability of the

model to incorporate cross-sectional characteristics such as

leverage, size, growth and liquidity which are considered as

important factors that affect dividend smoothing (18; 2). In

addition, due to countries specific factors such as tax policies,

economic and institutional characteristics, cross-sectional

differences and the fact that Lintner study was conducted 60

years back, some studies have gone beyond the boundaries of

testing validity of Lintner model in explaining dividend

smoothing behaviour of firms. In this respect, Leary &

Michaely and Abu-Khalaf, examined the determinant of

dividend smoothing by means of more sophisticated measures

that include firm median payout as Target Dividend per Share

Literature review

On the empirical studies in the field of dividend

smoothing, the researchers mostly started by highlighting the

theoretical foundation that underlies the debate employing

several approaches. Early researches to a great extent adopted

field investigation by studying the opinion of some corporate

managers in getting insight on what influenced the dividend

policy decision of their firms. Studies in the direction include

Lintner, Pruitt and Gitman (19, 21). Among the ones that

followed survey approach, Lintner further set up the

theoretical models and used statistical tests in order to provide

reliable estimates that could explain the pattern of corporate

dividend smoothing behaviour and policy. These studies found

that managers have divergent view on the factors that explain

dividend changes (19).

Moreover, Lintner variables which include, current

earnings and previous dividend have been found

predominantly important over past decades. Though, some

researchers have emphasized on the explanatory power of

factors such as cash flows (7, 16) and cross-sectional

characteristics using more robust smoothing measure (18 &

(

DPS) Policy and Relative Volatility for measuring divided

smoothing in order to address the Lintner’s model

shortcomings as earlier noted. Leary and Michaely empirically

tested firm characteristics as a function of dividend smoothing

2011 & 2), in explaining dividend smoothing, current

researches are still confirming the appropriateness of Lintner

variables either using static models given specific nature of

their economic policies, employing more robust techniques

and dynamic models that control both firm and temporal effect

in their estimations (3, 11). Other also employed the Lintner

model and used control variables (6). In assessment of the

determinants smoothness of dividend mostly three models

were employed; the Lintner partial adjustment model and

(

2, 18).

Furthermore, Lambrecht and Myers, explained that there

exists asymmetric adjustment smoothing behaviuor among

firms and this limits the desire of firms to adjust faster to their

target (17). Though, Abu-khalaf studied dividend payout,

propensity to pay dividend and dividend smoothing, and in

addition, examined whether high/low leveraged companies,

Journal of Environmental Treatment Techniques

2020, Volume 8, Issue 1, Pages: 73-84

Modified Lintner model by Brittain (1964) based on

symmetric adjustment and Dividend Deviation model based on

asymmetric adjustment (19).

∆DIVit = ἇit + ἇ

CEit + ἇ

DIVi(t-1) + ɛit

(3)

Where: ∆DIVit is the Expected change in dividend

payment , ἇit is the constant term and is expected to be zero for

some companies but will generally be positive to reflect the

2.1 Lintner (1956) Partial Adjustment Model

The mathematical first model to be considered in this study

greater reluctance to increase than to increase dividend, ἇ

equal to ƴ and ƴ is the speed of adjustment coefficient (That

is speed of movement of current dividend to target payout ratio

and is 0 < ƴ < 1 . therefore the closer the value to zero the

higher the smoothing behaviour and vice versa), p is the target

payout ratio and ἇ is (1- ƴ ). CE is the current earning for firm

i in period t, DIV(t-1) is the previous year’s dividend and ɛ is

the error term. Result of the empirical analysis revealed that

there is significant positive relationship earnings, previous

dividend and dividend changes. The study document Speed of

Adjustment (SOA) of 0.25 and Target Payout Ratio (TPR) of

0.6.

Lintner model performed well in explaining dividend

changes of the sample companies. Nonetheless, it suffers from

shortcomings such as; the study was conducted over sixty

years ago, covers only 28 firms and can only be applied on

cash dividend payments firms only (18). In addition, it is based

on the assumptions of symmetric movement and only previous

dividend and current earnings determine dividend changes,

constant response coefficient which suggests that investors’

reactions to the explanatory of all firms are identical, which

was criticized by (18; 2, 11) as being affected by the dynamic

firm-specific, industry specific and economic factors.

is Lintner partial-adjustment model of dividend. He conducted

interviews of corporate managers of US firms and enquired

about their dividend decisions (19). The series of the questions

asked were: firstly, whether their firms were primarily

concerned with dividend stability; secondly, whether earnings

were considered as the most critical factor that determines

dividend stability; thirdly, whether financial decisions were

taken in pursuance of dividend policy. The conduct of this

interviewrevealed the following findings; managers determine

their dividend policies based on a long-run target payout ratio,

the likelihood of paying and stabilizing dividend is highly

present with matured firms with stable earnings than growing

firms. This indicates that dividend changes are important to

managers than dividend level and firms smooth their dividend

to follow shift in the long-run sustainable earnings. He finally

concluded that, managers are reluctant to cut dividends

because they believe that shareholders prefer stable dividends

as such any cut in dividend may send bad signals about the

future prospect of the firm. Therefore, Lintner model assumes

that changes in dividend are gradually moving toward

achieving target payout not immediately to changes in

earnings.

Under this assumption of target payout, the partial-

adjustment model was developed to examine the smoothing

process in dividend policy. Thus, the target dividend Payments

are a proportion of the firms earning per share.

Mathematically,

2.2 Modified Lintner/Brittain (1964) Model

Ample studies have tested the modified version of Lintner

model or after extending it using US and other countries data

around the world. Brittain (1964) modified and tested the

Lintner model by deflating the variables using total number of

ordinary shares outstanding rather than using aggregate data in

US. The empirical result revealed SOA of 0.23 and TPR of

0.66 and confirms the findings of Lintner at lower level of

SOA. Observation has shown that most of the researchers that

tested Linter model employed Lintner/Brittain model, notably,

Al-najjar, Al-Yahyee, Pham & Walter, Jeong, Omar & Rizuan

and Sibanda (7, 9, 14, 20, 22)

The findings corroborated the study of Al-Najjar, and

Jeong in recent time. Among the variables introduced in the

various modifications of the model include; capital structure

variables, liquidity measures, firms growth variables and cash

flows (7, 14).

DIV*it = p

CEit

(1)

Where; DIV*it is the expected dividend payment for firm i

in period t, pi is the target payout ratio; CEit is the current

earnings after tax for firm i in period t

Knowing that DIV*it = piCEit, the model suggests that a

firm will only gradually adjust to the target dividend payment

in any given year, therefore change in dividend payment from

previous year (t-1) to current year (t) are explained by the

partial adjustment as thus:

DIVit -DIVi(t-1) = ἇ

+ ƴ

(DIV*it – DIVi (t-1) )+ ɛit

(2)

Rather than extending the model, several other researchers

used advanced methodology to capture the dynamic nature of

dividend behaviour in specific term. The result of the two-way

fixed and time effect model employed in the study revealed

that both the current earnings and previous dividend have

positive significant relationship with dividend changes. It also

revealed a SOA of 0.5 and above across sectors and low TPR

of zero and 0.17 in Industrial goods and Consumer goods

companies respectively. This signifies that there is a

significant difference in dividend policies across individual

firms over time. His findings are consistent with Abubakar,

who also used Least Square Dummy variable model in

examining the dividend smoothing behaviour among listed

manufacturing firms in Nigeria for the period of 2000 to 2009

Where: ἇ

is the constant term, ƴ

is the speed of adjustment

coefficient, DIV*it is the dividend which the company would

have paid in the current year if its dividend were based simply

on its fixed target pay-out ratio p applied to current earnings,

DIVit the actual dividend payment for firm i in year t and

DIVi(t-1) is the actual dividend payments for firm i in year t-1.

Thus, the model predicts that a dividend change is a function

of the target payout less the previous period dividend payout

multiplied by the speed of adjustment factor. From equation

1) and (2) by substituting p CEit for target dividend payment

DIV*it in Equation (2), Lintner drives dividend smoothing

model as follows:

(

Journal of Environmental Treatment Techniques

2020, Volume 8, Issue 1, Pages: 73-84

(

3). He found that manufacturing firm in Nigeria adopt

agency problems (13, 14). On this premise, they proposed the

use of debt as the substitute mechanisms for controlling the

problem.

unstable dividend policy.

The Lintner model was developed in the US where

dividends are taxed higher than capital gains, a system that

motivates smoothing. This may cast doubt on its valid

application on economies where their tax policies are different

from that of US. For instance like Oman where there is no tax

on divided, firms are highly levered, high concentration of

stock ownership and variability in cash dividend payment. Al-

Yahyee, Pham & Walter examined whether Oman financial

firms smooth their dividend and have target payout ratio using

Lintner model (9). Panel Tobit regression model was

estimated using 377 firm-year observations obtained from both

dividend-paying and non-dividend- paying firms covering the

period of 16 years 1989 to 2004. They found a positive

significant relationship between dividend per share and

previous year’s dividend per share and earnings per share. The

analysis returns a value of 0.94 and 0.56 for SOA and TPR

respectively. Though, study attempted to contribute to the

existing literature by accounting for censoring problem

associated with zero dividends and covered larger period

which is good to the study of dividend behaviour. However,

the effort has failed because dividend smoothing is assessed on

firms with dividend record not on both paying and non-paying

firms as in the case of other dividend policy studies.

Some other researchers have confirmed the findings of

Lintner partial adjustment model in recent time (19).

According to Omar & Rizuan Malaysian firms follow the same

determinant of dividend smoothing and stability as suggested

by Lintner (20). The empirical model was estimated using

ordinary least square. The study covered the period of 15 years

in examining 319 firms listed on Bursa Malaysia. Only

companies with history of nine years cash dividend payment

were used. The results revealed that firms in Malaysia were

involved in smoothing activities. The study also provided

evidence that the firms had target payout and they adjusted to

their target ratios with SOA of 0.447 and TPR of 0.64. Despite

the fact that the study confirmed the validity of Lintner model

in recent time, it suffers the following shortcoming; failure to

measure dividend smoothing using dividend changes as

suggested by the large number of previous and current studies

such as Lintner and Sibanda (19, 22)

Furthermore, given emphasis on the current development

on cash flows, researchers have questioned that earnings

provide superior information in explaining dividend

smoothing over cash flows. On this grounds, the trends of

researches such as Al-Najjar & Belghitar and Kighir, Omar

and Mohamed in recent times has shifted from emphasizing on

strong link between current earnings and dividend smoothing

to the superiority of cash flows over earnings (7,16). In attempt

to capture this effect, studies mostly follow two approaches;

by modifying Lintner model to include cash flow variables

instead of earnings measure or segregating the earnings into

two: cash flow and accrual components. In this regard, Al-

Najjar & Belghitar examined the superiority of cash flows over

earnings in explaining dividend smoothing by exchanging

earning measures with cash flows variables- free and operating

cash flows (7). The proposed Al-Najjar and Belghitar modified

partial adjustment model based on cash flows was developed

using Generalized Least squared model (GLS) and

Generalized Method of Moments (GMM) in order to control

un-observed firm-specific effect for the potential relationship

of both the previous year’s dividend and cash flows or earnings

in case of Lintner model. In consistence with most previous

literature, the study excludes financial firms in the analysis but

extended by including zero dividend. 432 firm-year

observations were used covering the period of seventeen (17)

years, 1991 to 2007. The study found that, United Kindom

(UK) firms smooth both the operation and free cash flows as

source for smoothing dividend and that original version of

Lintner partial adjustment is not working in UK, given the

lower result of SOA in relation to new version of Lintner’s

Model. The study has performed remarkably by confirming the

superiority of cash flows over earnings and also mitigates the

shortcoming of other researches by controlling firm-effect and

capturing both paying and non-paying firms. However, it fails

to account for asymmetric nature of dividend smoothing and

the effect of firm characteristic on the smoothness of dividend.

Contrary, Kighir et al. also contributed to the debate using data

from non-financial firms quoted on Bursa Stock Exchange in

Malaysia and found that non-financial firms in Malaysia

consider current earnings and proceeding years cash flow more

important than current cash flow and previous year’s profit is

establishing their dividend payout decisions (16).

Similarly, Sibanda also tested the model and found that

firms in South Africa smooth their dividend in line with Linter

argument (22). The result of the analysis revealed that firms

smooth their dividend by means of speed of adjustment

coefficient. The study returns a value of 0.73 and 0.41 for SOA

and TPR respectively. The study also suffers limitation similar

to Omar & Rizuan, in that it has avoided the problem of short

period and captured dividend smoothing as change in dividend

per share (20).

In Nigeria, earliest studies mostly focus on dividend

behaviour without establishing whether paying firms smooth

dividend and used Liner variables in the studies. Even the

dividend behaviour studies, Adelegan affirm that the earliest

attempt was conducted during the Indigenization period by

Uzoaga & Aloizeuwa who analyzed the dividend payment

pattern of sample of 13 companies for the period of 1969 to

1972. They found that there is not enough evidence to prove

the validity of traditional variables using Nigerian data and

concluded that the best predicators for dividend behaviour are

fear and resentment. The findings was later challenged by large

number of studies such as Inanga who asserted that both the

Lintner variables and non-conventional factor such as excess

cash obtained from issue of new capital and share pricing

policy of the Capital Issue Commission are the major drivers

.3 Modified Lintner/Brittain (1964) Model with Cash flow

Variables

Arguing from the angle of the free cash theory introduced

by Jensen who posits that the presence of excess cash in the

possession of managers provide them an opportunity to use the

fund to feather their own nest (13). It also argued that the free

cash flow provide managers with the incentive to make

investment in less profitable ventures, thereby increasing

Journal of Environmental Treatment Techniques

2020, Volume 8, Issue 1, Pages: 73-84

of dividend payment pattern (4).

investigated the smoothing behaviour of Jordanian firms using

data obtained from sample of 38 listed non-financial firms in

Amman Stock Exchange covering the period of 1998 to 2009

(23). The study employed Fixed effect panel regression and

used Lintner model after segregating the data into positive and

negative earnings firms in order to ascertain whether dividend

adjustment below and above target is present in the two

different conditions and firms with less than six years dividend

payments were excluded from the study. The empirical results

indicate that Jordanian firms have target dividend payout with

low rate of adjustment below 0.5 and it is asymmetrical

adjustment process depending on whether they are above or

below target with both positive and negative earnings. Though

the study attempted to assess dividend smoothing in both

positive and negative earning and captured firm’s specific

effect that has not been considered by most researches, it fails

to account for the inherent, asymmetric adjustment when firms

is at different levels of firm characteristics and small period

against the tradition of dividend payment studies without

justification.

The few identified that assessed dividend smoothing use

robust technique that could capture firm effect to investigate

whether smoothing behaviour exists in Nigeria, Abubakar

studied the sample of listed non-financial firms in Nigeria

using fixed and temporal effect model which captures what

was neglected by the previous researches (3). It is worthy of

note that, Abubakar only examined smoothing based on

symmetric assumption, which assumes firms movement from

current dividend payment to the target symmetrical (3). That

is there is no difference between speed of adjustment when

firm is above or below its target. Thus, the study is deficient

by its inability to account for asymmetric adjustment as

established in the literature (17). Similary, Kighir assessed the

smoothness of non-financial firms in Nigeria during the era of

unclaimed dividend and found that the firms use dividend

smoothing to manage their reported earnings (15).

.4 Dividend Deviation Model

In another development, other researchers are of the view

that the argument that firms set a long-run target payout and

move toward it does not hold water nowadays based on Lintner

SOA assumptions, since the advancement in the nature of

business activities, conflicted economics situations and

policies are not as the way they were in the sixties. In this

regard, they proposed factors such as cross-sectional

characteristics, and asymmetric adjustments as the most

important variables that derive dividend smoothing. Under this

argument smoothing is also assessed based on the SOA and

TPR but the method differs. Here, TPR is either firm /industry

average or firm/industry median dividend payout. And the

speed of adjustment is the coefficient of dividend deviation

variables when it is regressed against dividend changes. The

dividend deviation variable is calculated by subtracting the

respective year’s dividend from TPR.

Similarly, Abu-Khalaf examined the smoothness of

dividend of listed non-financial firms in Jordan covering the

period of 1997 to 2006 using random effect model (2). He

extended the prior researches by examining both symmetric

and asymmetric adjustments toward the target payout in

context, three cross-sectional characteristics; profitability,

leverage and size were used in order to ascertain whether firms

in Jordan differ in movement toward the target if they are

below or above their target payout and the differences in the

movements in large/small size, high/low leverage and

high/low profit companies. The study revealed that firms in

Jordan adjust toward target moderately and the process is

asymmetrical instead of symmetrical; they move at different

rate when below or above target and the asymmetric

movement is influenced by the firms characteristics; size,

leverage and profitability. In line with signaling theory in the

case of profitability but contradict the theory in the context of

size and supports agency cost theory in respect of leverage.

Based on this background, therefore, an analysis of the

dividend smoothing based on asymmetric adjustment and

growth potentials was conducted in order to ascertain where

dividend smoothing behaviour in listed industrial goods is

asymmetric whether firm with high growth potential smooth

dividend slower than those with low growth potentials. Thus

the following hypotheses are proposed:

According to Leary & Michaely traditional variables for

testing smoothing are biased and are not the best measures for

explaining firm cross-sectional differences in dividend policy

(

18). They estimated dividend smoothing using more

sophisticated SOA measures that include Relative Volatility

and Target Dividend Per share (TDPS) Policy as against the

traditional Lintner’s SOA factor. Firm cross-sectional

characteristic were employed in predicting the behaviour of the

dependent variables; dividend changes. These measures were

suggested based on the argument that dividend target today is

different from what Lintner understands and in recent time

managers are more concerned about Target DPS than Target

POR. The result revealed that firms with more tangible assets,

lower price volatility, lower earnings volatility, institutional

investors, higher payout ratio and large size tend to smooth

more. This result is consistent with Agency consideration that

lower growth firm, firm with free cash flow and better

corporate governance tend to smooth dividend more.

Furthermore, other researchers extended by testing the

smoothing behaviour in both negative and positive earnings on

the assumption that the adjustment towards target differs in the

two conditions differs or is asymmetric. Hence, firms with

positive earnings and having dividend lower than its target are

expected to move faster to the target compared with those with

negative earnings. In this regard, Zurigat & Gharaibeh

H0 Dividend smoothing in listed industrial goods firms in

Nigeria is symmetric

H0 Firms with high growth potentials with below target

dividend smooth their dividend payment faster than the ones

with low growth potentials in listed industrial goods firms in

Nigeria.

H0 Firms with high growth potentials with above target

dividend smooth their dividend payment slower than the ones

with low growth potentials in listed industrial goods firms in

Nigeria.

2.5 Theoretical framework

This study was anchored on signaling theory. The theory

was chosen predicated form the view that dividend is a

signaling device for investors regarding firm’s prospect (19).

Journal of Environmental Treatment Techniques

2020, Volume 8, Issue 1, Pages: 73-84

The theory is based on the idea of information asymmetry

between owners and managers, which assumes that managers

possess higher knowledge than investors about the firm’s

prospect. Managers use dividend as a signaling device to

convey their performance to investors thereby aligning their

knowledge with that of investors. The fact that higher

dividends translate into higher stock price and consequently

increase in firm value is the idea behind the use of dividend as

a signaling mechanism. Since smoothing is a movement to a

target payout, it can be either upward when firms are below

their target or downward when they are above. The fact that

increases in dividend signals good news and increases

represent bad news to investors the firms are expected to move

slower above than when they are below target (2). Therefore,

the level of information asymmetry determines the smoothness

of dividend which suggests that the firms with more severe

information asymmetries are likely to smooth more (20-23).

Furthermore, dividend smoothing behaviour is more prevalent

in a firm with high growth potentials because they have severe

information asymmetry. This indicates that firms with high

investment opportunities and very little tangible assets smooth

more and it reduces over time in accordance with information

revealed to the market increases (18).

smoothing behaviour can only be examined in dividend paying

firms and to minimize the likelihood of spurious result (2, 8).

Therefore, the adjusted population is nine (9) firms and these

firms includes, Ashaka, Avon Crown Caps &Containers Plc,

Berger Paints Plc, Beta Glass Co. Plc, Chemical and Allied

Product Plc, Cement company of Nigeria Plc, Cutix Plc, Greif

Nigeria Plc, and Lafarge Africa Plc. The study used

secondary data because the variables under investigation can

be best measured using data in the form of financial

information available in the selected firms’ financial reports.

The sources of the data therefore are the annual reports and

accounts of the selected firms for the period covered by the

study - 2007 to 2016.

Multiple regression technique was employed in modeling

the relationship that exists between the dependent and

independent variables of the study. In checking the smoothness

of dividend payment, Panel data analysis technique was

employed through the use of Fixed and Random effect models

in estimating the coefficients of the model employed for the

study. The technique was chosen because it produces more

informative data, less collinearity among variables, more

variability, more efficiency and more degree of freedom (11,

12). Due to nature of the measure chosen for measuring

dependent variable of dividend smoothing models; dividend

changes, which is the change between current year’s and the

previous year’s dividend, some negative figures were obtained

whenever the previous dividend is higher than the current one,

as such the use of Tobit will be biased since the technique is

meant for analyzing restricted variables model. Having any

value below zero the restriction cannot be obtained and the

values will now be continuous to both left and right of zero.

Therefore, this suggests the use of FE and RE models.

Hausman specification test is used for testing the FE against

RE model estimates under the null hypothesis that the

coefficient estimated by the consistent FE estimator are same

as the ones estimated by the consistent RE. The significant

Hausman test leads to the rejection of the null hypothesis in all

our three estimated models, suggesting that FE estimates are

the most appropriate estimator in the relation to FE estimates.

3.1 Models specification and variables measurement

The study employed both symmetric and asymmetric

Partial adjustment models for testing smoothing behaviour. In

addition, Target DPS was used to estimate firm median DPS

and it was used to investigate the impact of the high/low

growth potentials using both symmetric and asymmetric

partial adjustment toward the long run target.

Figure 1: Interaction of below and above target dividend

between low and high growth potentials and dividend

changes.

3.2 Symmetric Adjustment Model

Symmetric adjustments model assumes a symmetric speed

of adjustment for dividends above and below the long run

target ratio. This implies that the costs and benefits associated

with adjustment below and above target are equal (2). In line

with him, the model was estimated in order to determine

whether actual lag dividend deviates from the long run target

Methodology / Materials

The study employed correlational research design. The

design was adopted predicated to the objectives of the study

which is to examine the factors that determine dividend

smoothing patterns of firms. The population of this study

consists of all sixteen (15, 16) listed Industrial Goods firms in

Nigeria as at 31 December, 2016. Seven (7, 8) firms were

excluded from the population as a result of not having five (4,

payout. In achieving this dividend deviation variable (₯evit

)

was calculated by taking the difference between target payout

ratio (TPR) of the current year and actual lag dividend

payments. The target payout ratio was calculated by

multiplying the target ratio (TR) with the current earnings (CE)

) years dividend payment history. This is because dividend

Journal of Environmental Treatment Techniques

2020, Volume 8, Issue 1, Pages: 73-84

that is TPR is equal to TPRE. The median of the firm payout

ratio was used as against the industry or firm mean because its

use is associated with the problems of outliers.

For estimating dividend smoothing using firm’s median

payout ratio as Target Dividend Per share, the model is

presented in equation 1:

above

below

represents dividends

where ₯evit

and₯evit

payment that are above and below long run target ratio

respectively. ƴ and ƴ are the adjustment coefficients to be

estimated. We hypothesize that the two adjustment

coefficients must be significant, greater than zero (ƴ > 0, ƴ

) and not equal (ƴ ≠ ƴ ). Furthermore, in line with Abu-

∆DPSit = ἇ

+ ἇ

₯evit + ɛit ------------ I

Khalaf, the work argues that if the cost of decreasing dividend

is higher than that of increasing it, then the coefficient of

above

below

Where ∆DPSit change in dividend payment (DPSit

- DPSi(t-

₯evit

is greater than that of ₯evit

< ƴ2. Hence, the

)

) , ₯evit = dividend deviation , ἇ = the adjustment coefficient

speed of adjustment for increasing dividend below the target

will be faster than that for decreasing when firms are above

their target (2).

which captures the adjustment in dividend changes to the

target dividend payout ratio, ɛit error term. The adjustment

coefficient is assumed to lie between 0< α≤ 1, indicating that

firms have target ratio and they do not instantly adjust their

dividend payment ratio to the TPR. If α = 1, indicates no

absence of adjustment cost and then instant shift towards the

target occurred, implying no dividend smoothing exist. On the

other hand if α = 0, implies that no movement towards the

target exist, since the actual adjustment at a time t is equal to

the observed in the previous time period and then dividend

In line with signaling hypothesis, it also expects the rate of

adjustment to differ when firms are above and below the target

and the smoothing behaviour should also to be affected by

high/low growth potentials. This is because firms with high

growth potentials tend to have high investment opportunities

and low free cash and the reverse is the case with those with

low growth potentials. Lintner argues that earnings influences

dividend smoothing and firms mostly smooth their dividend in

relation to changes in their earnings (cash flows) (19).

Therefore, adjustment below or above the target dividends

differs in firms with high or low growth potentials. Hence, two

new variables Hgwth for high growth potentials and Lgwth for

growth potentials were introduced as interaction variable in the

model. Based on the above variables introduced in the model,

the new model is presented as thus:

payment is completely stable. α > 1 indicates abnormal

adjustment take place and the target has not been attained.

Therefore if the coefficient is found to be significant, it

indicates that the selected firms have TPR and gradually move

to the target overtime when paying dividend.

.3 Asymmetric Adjustment Model

Given the controversies in the literature that there is

difference in cost of adjustment between firms that are below

and above the target or otherwise, this study further adopted

asymmetric model from the studies of Leary and Michaely and

Abu-Khalf, in order to prosper a solution to the controversies

∆DPSit = Ԓ

+ Ԓ

Lgwthit *₯evit

₯evit ^a^b^o^v^e+ Ԓ

₯evit ^b^e^l^o^w+Ԓ

Lgwthpit

above above

+ Ԓ Hgwthit *₯evit

above

Hgwthit +Ԓ

below

Lgwthit * ₯evit

+ Ԓ

Hgwthit * ₯evit

+ Ҙit -------

(III)

(

2, 18). This test is important especially when dividend

payment differs significantly in a situation where firms are

below or above the target, that is, if firms are above the target

they tend to adjust downwards and accordingly, if they are

below the target, they adopt upward adjustment. In such a case,

Mangers of a value a maximizing firms adopt a dividend policy

that always maximizes their share value (19). Hence, the

managers of these firms achieve this by smoothing the actual

dividend payments toward a target level.

where Lgwthit, Hgwthit, are low growth potentials and high

above

below

and ₯evit

growth potentials respectively. ₯evit

represents dividends payments that are above and below long

above

run target ratio respectively. ₯evit

(Lgwthit + Hgwthit) is

the interaction between dividend deviation above the target

blow

and low/high free growth potentials. ₯evit

(Lgwthit +

Lgwthit) is the interaction between dividend deviation below

the target and the low/high growth potentials for firm i in year

When dividend deviation is less than zero (₯evit < 0) it

implies that the company is above it target payout ratio, while

it’s below the target if ₯evit > 0. Therefore, to examine

whether adjustment rate differs across firm that are below or

above their target level, the following models are used by

splitting the values of ₯evit into two new variables:

t. If the coefficients; Ԓ5, Ԓ Ԓ7,and Ԓ8, are significant then

dividend smoothing exists and they should not be jointly equal

to zero for asymmetric adjustment to exist. Ҙit Error is the term.

3.4 Variables measurement

Having specified the model of the study, the variables

employed for the study, their measurements are presented in

the Table 1:

₯

evit ^a^b^o^v^e= ₯evit if TPREit – DPSi(t-1)

< 0 and zero

otherwise

₯

otherwise

evit ^b^e^l^o^w= ₯evit if TPREit – DPSi(t-1) ≥ 0 and zero

Analysis of results

The results obtained from the descriptive and inferential

statistics are presented in this section. It started from the

empirical distribution of the variables and then determines the

existence and direction of relationship between the variables

of the study.

above

below

When the variables ₯evit

and₯evit

are

substituted in place of ₯evitin model (I) above, model (II) will

be presented as thus:

∆Divit = ƴ

+ ƴ

₯evit ^a^b^o^v^eandƴ ₯evit^b^e^l^o^w+ ɛit -------------

Journal of Environmental Treatment Techniques

2020, Volume 8, Issue 1, Pages: 73-84

.1 Descriptive statistics

From table 2, it can be seen that the rate of changes in

clear indication that the selected firms may likely smooth

dividend in line with Linter’s findings. Dividend deviation

below target returns an average value of 0.55; indicating that

average changes from actual lag dividend to target pershare

when the firms are below the target is an increase by average

by about 55kobo per share. A mean value of -0.13 for deviation

above implies that dividend deviation above target per share

has recorded a decrease on average by 13 kobo per share

within the period of the study.

dividend payment over number of ordinary shares has

increased by an average rate of 38 kobo per share while

dividend deviation below target and deviation above target

have average of about 0.42, 0.55 and ,-0.14, respectively.

These imply that dividend deviation with about 42 kobo per

share average is far above average changes in dividend per

share. This suggests that magnitude of dividend changes does

not change at the same rate with dividend deviation and also a

Table 1: specified the model of the study

Variables

Dividend smoothing measured by

dividend changes

Acronym

Div

Measurement

Current Dividend per share

minus lagged Dividend per share

Target Dividend payout less lagged Actual

Dividend payout Ratio

Source

Driver et al, Al-Najjar and Klinkarslan

(8,11)

Abu-khalaf and Zurigatand Gharaibeh

(2,23)

∆

Dividend Deviation

₯ev

A firm is above the target if actual payout

is higher than target payout. That is ₯ev

less than zero

A firm is below the target if actual payout

is lower than target payout. That is ₯ev is

greater than zero.

Dividend Deviation above the

Target

above

Abu-khalaf , Zurigat and Gharaibeh,

Leary and Michaely (2,18,23)

₯

evit

Dividend Deviation below the

Target

below

Abukhalaf, Zurigat and Gharaibeh, Leary

and Michaely (2,18,23)

Median of firm’s dividend paid over

number of ordinary shares.

Abu-khalaf and Zurigat and Gharaibeh,

Leary and Michaely (2,18,23)

Abu-Khalaf Xian, (2)

Target Dividend per share

Growth potentials

TDPS

Gwth

Market to book value of equity

A dummy variable; 1 for firms that have

growth potentials lower than the median

and zero otherwise multiply by growth

potential value

A dummy variable; 1 for firms that have

growth potentials higher than the median

and zero otherwise multiply by growth

potential value

Low growth potentials

High growth potentials

Lgwth

Hgwth

Abu-Khalaf (2)

Abu-Khalaf (2).

Table 2: Descriptive statistics

Variables

DPS

Gwth

Obs.

Mean

Std. Dev

0.6214

1.1442

5.0082

1.0181

0.3441

1.0856

0.1178

Min

Max

2.3277

4.8761

24.1346

4.8761

∆

₯

0.0379

0.4157

3.7480

0.5537

-0.1379

0.9723

2.7756

-2.6062

-1.9293

0.1528

-1.9293

below

evit

above

₯

evit

Lgwthf

Hgwthf

10.0235

24.1346

Source: STATA output (2019)

Table 3: Correlation Matrix

above

below

∆DPS

₯ev

GWTH

₯evit

Lgwth

Hgwth

∆

₯

DPS

GWTH

1.0000

0.3407

-0.0857

0.1975

0.5483

0.2115

-0.1704

1.0000

0.5249

0.9559

0.4982

0.4190

0.3430

1.0000

0.5849

0.0159

0.1582

0.9150

below

₯

evit

1.0000

above

₯

evit

0.2217

0.4471

0.3903

1.0000

0.0711

-0.0135

Lgwth

Hgwth

1.0000

-0.2538

1.0000

Source: STATA output (2019)

It also shows that the selected firms on average are slightly

decreasing their dividend from actual lag dividend payment to

target whenever they are above target per share. The presence

of average increase above and average decrease below target

is an indication that the selected firms are likely to have

asymmetric smoothing behavior. Furthermore, the result

shows that the average value of growth potentials is 3.75. This

indicates that the total value of market to book value of equity

is about ₦3.75 in the sub-sector over the period of the study.

The result also shows that the average value of low growth

potentials and high growth potentials are 0.97 and 2.78

respectively. This indicates that on the average the selected

firms have 97 kobo and 2.78 kobo market to book value of

equity for low and high market to book value of equity

Journal of Environmental Treatment Techniques

2020, Volume 8, Issue 1, Pages: 73-84

respectively. Deviation above has the lowest standard

deviation of 0.34 while high growth potentials have the highest

value of 5.11. This implies that observations for high growth

potentials have widely dispersed away from their mean value

in relation to other variables and a variable with observations

not far away from the average value is dividend deviation

above target payout.

is FE model and the summary of the result is presented in table

4. The result in the table shows that the Wald Chi2 is statistics

is 35.42 with p-value of 0.0000 implying that the model is

fitted. The adjustment coefficient of common intercept

exhibits a value -0.2104 and a p-value of 0.002. This indicates

that constant term is negative significant at 1% level. In

consistent with Al-Malkawi, Bhatti and Magableh the finding

implies low reluctant to cut dividend and that if a given firm

decides to reduce dividend, the action will negatively affect it

reputation in the market (6). On the other hand, this finding

contradicts Lintner and Abu-Khalaf who posit that the constant

term will usually be positive to show greater reluctance to

decrease than to increase (2,19). The results further shows that

dividend deviation variable ₯evi,t returns a coefficient of

0.5035 with a p-value of 0.000 suggesting that the adjustment

coefficient is statistically significant at 1% level and that listed

Industrial Goods firms in Nigeria have target dividend payout

and adjust gradually toward the target rate. Also their speed of

adjustment rate is 50% which is equal to 50% (that is the rate

that signifies average rate of adjustment) higher than the ones

obtained by studies such as 25% in Oman by Al-Malkawi et al

and 45% in Malaysia by Omar and Rizwan (6, 20). In Nigeria,

Abubakar found SOA of 235% after covering the period of

2000-2009. This shows that the SOA has drastically reduced

from 235% to 50% indicating that now Nigerian firms have

transformed form unstable dividend behaviour to excessive

dividend smoothing as compared to about 8 years ago (3).

.2 Correlation matrix result

From Table 3 it can be seen that there is positive

relationship between dividend changes, dividend deviation,

deviation below, deviation above and low growth potentials.

While the relationship between the dividend changes, growth

potentials and high growth potentials are negative. On the

relationship among the independent variables themselves, the

correlations that calls for concern is 0.96 which is between

deviation and deviation above and 0.95 between growth

potentials and high growth potentials Though, this implies that

there is tendency of harmful multi-collinearity among the

variables but since the variables are extractions from each

other the condition is inevitable.

.3 Result for symmetric partial adjustment model (model I)

In order to determine the best estimates of the symmetric

partial adjustment model for the listed Industrial Goods firms

in Nigeria both FE and FE models were estimated. The

diagnostic test between FE and RE model estimates shows that

Hausman specification has a chi2 of 24.40 and a probability

value of 0.0000. This provides evidence for rejecting the

hypothesis that the coefficients estimated by the efficient RE

estimator are the same as the ones estimated by the efficient

FE model estimator. This suggests that the best specification

4.4 Result for Asymmetric Partial adjustment Model

Given that dividend adjustment assumes firms are

reluctant to cut dividend indicates that the adjustment can take

the form of asymmetric not symmetric.

Table 4: Result of the Symmetric Partial Adjustments Model (model I)

Coefficients

t-value

(

0.2104***

0.002)

.5035***

0.000)

Common intercept

-3.20

₯

5.95

31.23%

35.42

Wald Chi2.

P-value

Hausman Test (chi2 value)

P-value

0.0000

24.40

0.0000

Note ***, **, * represents values are significant at 1%, 5%, and 10% level

Source: STATA output (2019)

Table 5: Result of the Asymmetric Partial Adjustments Model (model II)

Coefficients

t-value

(

0.0231

0.789)

.2823**

0.010)

.9736***

0.000)

Common intercept

-0.27

below

₯

evit

2.64

5.72

above

39.01%

24.62

Wald Chi2

P-value

Hausman Test (chi2 value)

P-value

0.0000

11.82

0.0027

Note ***, **, * represents values are significant at 1%, 5%, and 10% level

Source: STATA output (2018)

Journal of Environmental Treatment Techniques

2020, Volume 8, Issue 1, Pages: 73-84

In this regard, the asymmetric adjustment is analyzed in

this section in order to determine if adjustment rate above is

different from that of below target payout. Also, the best

specification of the model is FE model due the significant of

the Hausman specification test suggesting the rejection of the

hypothesis that says RE is the most appropriate model.

Therefore, FE robust model was estimated in order to take care

of the problem. The summary of result of the robust FE model

is presented in Table 5. It can be seen from Table 5 that model

is fitted due the value of Wald Chi2 of 24.62 and p-value of

than that of below the target 28%. The rate of adjustment for

dividend below the target than that of above is an indication

that the selected firms are more reluctant to increase rather than

reducing dividend. The result further contradicts the position

of agency cost theory which says that the cost of increasing in

lower than that of decreasing dividend. In addition, the finding

show that the behaviour of the firms is in conflicts with

signalling theory because they are more interested to decrease

than to increase dividend and asserts that the firms use

dividend as signalling device. The result further corroborates

with Lintner corroborate with Zurigat and Gharaibeh and Abu-

khalaf (2,19,23). A possible explanation to this is that listed

Industrial Goods in Nigeria do not increase dividend until

making sure that the increase can be sustained. Also with an

above target-dividend a slow reduction is expected but given

the nature of the firms and the period covered by the study, the

firms are likely to face liquidity problem due to the adverse

economic condition. As such it would not possible for them to

be fully reluctant to decrease than to increase.

0.000. It can also be seen that the adjustment coefficient is

asymmetric for movement below and above target payout.

This is determined after estimating adjustments coefficients

ƴ andƴ for deviation below and deviation above target

1 2

respectively are estimated and the hypothesis which proposes

that the two coefficients are jointly equal to zero and are equal

(

1 2 1 2

that is ƴ =0,ƴ =0, and ƴ =ƴ ) was tested. The result of the test

shows F-stat value of 24.62 and P-value of 0.000 and F-stat

value of 9.81 and P-value of 0.0025 for the ƴ =ƴ =0, and ƴ =ƴ

test respectively. Implying that adjustments coefficients

andƴ for deviation below and deviation above target have

4.5 Result of asymmetric partial adjustment model including

interaction with high/low Growth potentials (Model III)

The asymmetric partial adjustment model has been

examined not only to determine whether adjustment is

asymmetric below and above target payout ratio, but also

asymmetric if the adjustment rate varies for below/above-

targets dividend adjustment when firm is having low/growth

potentials as well as below/above-target dividend adjustment

when firm is experiencing high growth potentials. The best

fitted estimation of the result is based on FE model due to

significant of the Hausman specification test which exhibits a

chi2 value of 26.99 with p-value 0.0007. On the bases of this,

the hypothesis that says RE is the most appropriate estimate

was rejected. The FE model was estimated and summary result

is presented in Table 6.

satisfied the conditions for testing dividend smoothing. Thus

the two hypotheses are rejected at 1% level. Meaning that

dividend smoothing behaviour is symmetric in listed Industrial

Goods firms in Nigeria. The result reveals value of 0.2823 and

.9736 which are statistically significant at 1% and 5% for ƴ

and ƴ respectively. Evidence from the result the two

adjustment coefficients are positively and statistically

significant justify the rejection of the first hypothsis (Ho

)

which says dividend smoothing behaviour in listed industrial

goods in Nigeria is symmetric. Therefore, the smoothing

behaviour is asymmetric.

This result suggests that the dividend smoothing in listed

Industrial Goods firms is asymmetric instead of symmetric.

Also, the adjustment rate for above the target 97% is higher

Table 6: Result of the Asymmetric Partial Adjustments Model including interaction with growth potentials (Model III)

Coefficients

-0.0935

t-value

-0.58

Common intercept

(

0.566)

-0.0306

0.877)

1.5582***

0.000)

0.0957

below

₯

evit

-0.16

4.55

1.31

0.22

2.00

1.8

(

above

(

Lgwth

Hgwth

(0.195)

0.0052

(

0.829)

below

above

Hgwthx₯evit

Lgwthx₯evit

Hgwthx₯evit

Lgwthx₯evit

0.0213*

(

0.049)

0.0512*

0.076)

-0.0721

0.117)

-0.3934*

0.064)

(

-1.36

-1.88

(

52.59%

9.71

0.0000

26.99

Wald Chi2

P-value

Hausman Test (chi2 value)

P-value

0.0007

Note ***, **, * represents values are significant at 1%, 5%, and 10% level

Source: STATA output (2018)

Journal of Environmental Treatment Techniques

2020, Volume 8, Issue 1, Pages: 73-84

Result presented in Table 6 shows the model has a Wald

Chi2 of 9.71 with a p-value of 0.000 indicating that the model

is fitted at 1% level. It can also be seen that dividend smoothing

behaviour in listed Industrial Goods firms in Nigeria is not

only asymmetric for below and above target but also

asymmetric when firms experience high/ low growth

Malkawi et al who posits that firms with high profit (cash flow)

pay high dividend than low profitable ones (6). Therefore, high

growth potentilas enable listed Industrial Goods firms in

Nigeria to adjust to their target payout more quickly. The

findings is also in line with signalling theory because firms

with high free cash flow are likely to have severe information

asymmetry and tend to pay more and smooth less.

potentials when below.This is evident from the `coefficient of

below

.0213 and

Lgwthx₯evit

be statistically and positively significant all at 10% level

respectively. Also, the coefficeint of -0.3934 for Lgwthx₯evit

and

0.0512,

for

Hgwthx₯evit

below,

respectively. The coefficients are found to

Conclusion

The paper concludes that listed Industrial Goods firms in

Nigeria have target payout ratio and move to their target at a

moderate rate of adjustment. Also the process of the movement

is asymmetrical instead of symmetrical because the firm have

different down-ward/up-ward movement when they are

below/above their target payout ratio. Furthermore on

examining the effect of high/low growth potentials on the

smoothness of dividend, the paper concludes that firms smooth

their dividend when their dividend payment is below-target

payout with high and low growth potentials. The speed

adjustment is faster when below-target with low growth

potentials.

above

is significant at 10%, while the coefficient for

above

Hgwthx₯evit

is not significant at all levels of significant.

This suggests that firms smooth their dividend when the

dividend payment only above their target payout ratio and

when they experience high growth potentials in listed

industrial goods firms in Nigeria. Thus test of hypotheses that

7 8 7 8

propose Ԓ = 0, Ԓ =0 and Ԓ =Ԓ with Fstat. Value of 2.98 and

.29 for the two test respectively were all found to be

statistically significant at 10% and 5% respectively. This

suggest that the variables are not equal and also not equal to

zero. However the test cannot be perfomed between

5 6

coefficents Ԓ & Ԓ beacuse they are not positive significant as

specified by the study requirement.

Based on these findings the hypothesisH0

with low growth potentilas and below -target dividend move

to their target payout slower than the firms with high growth

potentials in listed Industrial Goods firms was rejected. This is

References

Abdulkadir RI. Determinants of propensity to pay dividends

among Nigerian listed companies. Unpublished Doctor of

Philosophy Thesis. Universiti Utara, Malaysia, 2015.

that says firms

2. Abu Khalaf BK. Dividend smoothness, determinants and impact

of dividend announcements on share prices: empirical evidence

from Jordan (Doctoral dissertation, Heriot-Watt University),

2012.

because

Hgwth*₯evit

the

speed

adjustments

coefficient

below

(76%) are

below

(49%) and Lgwth*₯evit

Abubakar N. Do firms smooth their dividends over time?.

Evidence from listed manufacturing Firms In Nigeria. Nigerian

Journal of Accounting Research. 2014;10(1): 23-48, june

Adelegan O. The Impact of growth prospect, leverage and firm

size on dividend behaviour of corporate firms in Nigeria.

Manuscript, University Of Ibadan, 2003.

significant and statistically. This indicates that firms with low

growth potentilas adjust their dividend toward target payout

ratio in listed industrial goods firms faster than the ones with

high growth potentials. This is expected because dividend

adjustment below-target is usually up-ward movement and in

order to reduce agency cost, firms that are cash cows have high

free cash flow are supposed to adjust quickly than those

otherwise.

The finding also confirms dividend signalling, which says

managers of high profitable firms tend to signal their good

prospect in the future cash flow their smooth less and pay more

and Zurigat and Gharaibeh, who posit that Jordanian firms are

highly sensitive to smoothing when they experience negative

earnings as such with positive earning. But contradicts Abu-

Khalaf who found that high profitable firms smooth their

dividend firm faster than low profitable (2,23).

5. Al-attar AM, Al-Shattarat HK, Yusuf AN. The effects of

contextual factors on disaggregated earnings in explaining

dividend changes in Jordanian firms". International Journal of

Commerce and Management. 2015;25(2): 218 – 23.

Al-Malkawi HAN, Bhatti MI, & Magableh SI. On the dividend

smoothing, signalling and the global financial crises. Economics

Modeling. 2014;42:159–165.

Al-Najjar B, Belghitar Y. The information content of cash flows

in the context of dividend smoothing. Economics Issues.

2012;17(2): 57-70.

8. Al-Najjar B, Kilincarslan E. Corporate dividend decisions and

smoothing: New Evidence from an empircal study of Turkish

firms. International Journal of Management Science. 2017;13(3):

On the hypothesis H0 which states that firm with high

-32

growth potentials and above-target dividend smooth faster

than the firms with low growth potentials in listed Industrial

Goods firms in Nigeria. The results reveal that coefficient for

Al-yahyaee KH, Pham TM, Walter TS. Dividend smoothing when

firms distributes most of their earnings as dividend. Applied

Financial Economics. 2011;21: 1175–1183.

below

Hgwth*₯evit

(177%) and Lgwth*₯evit

(64%) are not

10. Ozo FK, Arun TG, Kostov P, Uzonwanne GC. Corporate dividend

policy in practice: the views of Nigerian financial managers.

Managerial Finance. 2015;41(11): 1159-1175

significant positive significant. in addition a test whether the

two coefficients are jointly significant for asymmetric

adjustment to exist was not satisfied. Thus, hypothesis H0 was

not rejected. A possible explanation to this is that with a above

target-dividend a continuous decrease is expected and the

firms with high growth potentilas are highly constrained as

such it would not rational for them to engage in an increase

that cannot be maintained. This result is consistent with Al-

1. Driver C, Grosman A, Scaramozzino P. Dividend Decision of UK

Firms: What do we really know? Seminar paper presented at

Aston Business School, 2016.

2. Gujarati ND, Porter CD. Essentials of econometrics (5th ed.).

New York, USA: Mc Graw-Hill/Irwin, 2009.

3. Jensen MC. Agency costs of free cash flow, corporate finance and

takeovers. American Economic Review. 1986;76(2): 323–329.

Journal of Environmental Treatment Techniques

2020, Volume 8, Issue 1, Pages: 73-84

4. Jeong J. Determinants of dividend smoothing in emerging market:

The case of Korea. Emerging Market Review. 2013;17: 76–88.

5. Kighir AE. Dividend Smoothing in an era of unclaimed dividends:

Review. 1956;46(2): 97–113.

20. Omar N, Rizuan S. Finacial deceiful trick through dividend

smoothing. Procedia Social and Behavioural Sciences.

2014;14(5): 300–307.

panel data analysis in Nigria. International Journal of

Economics and Management Engineering. 2016: 10(1).

6. Kighir AE, Omar NH, Muhammed N. Corporate cashflow and

dividends smoothing: A panel data analysis at Bursa Malaysia.

Journal of finance Reporting and Accounting. 2015;13(1): 2-19

7. Lambrecht BM, Myers SC. A Lintner model of payout and

managerial Rents. Journal of Finance. 2012;67(5): 1761–1810.

8. Leary MT, Michaely R. Determinats of dividend smoothing:

Empirical evidence. Working Paper, Cornell University of

Toronto, United State of America, 2011.

21. Pruitt SW, Gitman LJ. The interactions between the investment,

financing, and dividend decisions of major U.S.firms. Financial

Review. 1991;26(3): 409–430.

22. Sibanda M. Do firms smooth their dividend over time? Evidence

from the Johannesburg Stock Exchange. Journal of Economics.

2014;5(3): 333–339.

23. Zurigat Z, Gharaibeh M. Do Jordanin firms smooth their

dividends?Empirical test of symmetric and asymetric partial

adjustment models. International Research Journal of Finance

and Economics. 2011;8(1): 160-171.

9. Lintner J. Distribution of incomes of corporations among

dividends,retained earnings and taxes. American Economics