Theories and Evidence on Dividend Smoothing in Corporate Finance

This report provides a comprehensive analysis of dividend smoothing, a well-documented phenomenon in corporate financial policy. It examines the evolution of dividend smoothing behavior, cross-sectional changes, and time trends, utilizing unique datasets. The study delves into the theories surrounding dividend smoothing, including agency considerations, asymmetric information, and tax planning, along with their empirical implications. It addresses the impact of market frictions, such as agency conflicts, taxes, and information asymmetry, on the extent of dividend smoothing. The research highlights the significance of share repurchases and their relation to dividend payout and smoothing policies, documenting a stable rise in dividend smoothing over time. The findings reveal that traditional smoothing procedures may be biased, and alternative measures are proposed. The report also explores the relationship between earnings dynamics and smoothing, and the unevenness in dividend speed changes from the target. The study's conclusions provide insights into the factors influencing dividend smoothing and the implications for corporate financial decision-making. The article is divided into sections, including theories of dividend smoothing, measurement methods, data analysis, and conclusions.

The most documented well known phenomenon in corporate financial policy is called
dividend Smoothing. It was observed in many studies that the concern of firms is primarily
about the stability of dividends (Linther, 1956). The firms need to change the dividend policy
as they need to set the policy to pay the dividends quarterly. They need to decide when the
change is necessary and how do they implement that change and whether the change needs to
a big change or small. The firms with stable dividend policy will get premiums from the
market theses are manager’s contentions. The Linther’s study was done 50 years ago and the
sample selected for the study was almost 28 firms, his study was based on a wide range of
firms and with the most recent ones. For example: Fama and Babiak, 1968; Bray et al. 2005.
Latest evidence says that this change is proved to be costly for firms. The study findings says
that the managers are willing to forgo the investments with positive NPV and willing to raise
capital with external sources in order to evade cutting dividends (Brave et al., 2005).
In spite of the importance and prevalence of dividend smoothing, there is a little
propensity as if why firms need to do dividend Smoothing. The research has shown an
improved interest in elucidating dividend smoothing; the evaluation of alternative
explanations has some limited empirical evidence. The article is to address this gap and give
an understanding as to why firms need to smooth out their dividends by reviewing the cross-
sectional changes in their dividend Smoothing behaviour (say total pay-out). We also assess
the smoothing of dividend that has evolved over time by using unique dataset.
Dividend Smoothing is done with our realistic analysis according to the market
frictions by classifying the existing hypothetical markets. The smoothness of dividend is
affected by many market frictions which are associated to the extent of dividend levels that is
by agency conflicts, taxes and facts asymmetry. The evidences are provided which are
empirical and a broad list of proxies are used to prove which type of the firms are more likely
or less likely to dividend smoothing. The given firms prove a significant shift near to
repurchases, the cross sectional properties of dividend smoothing are also analysed. For
exploring the time trends in dividend smoothing behaviour we need to use an exceptional
dataset, starting from 1920s and over a longer time duration than has been previously
documented. The theories of dividend smoothing which are in existence as well as the
directions for future work both are severed in our findings.
The evidences which are existing and empirical on the smoothing behaviour suggest
that the smoothing of dividend is widespread (For example, empirical evidence by Fama and
Babiak, 1968; Lintner, 1956 and Choe, 1990). However there were limited studies which
examine cross sectional variances. In United Kingdome, smoothing of dividend is more
clearly visible in public firms as compared to private firms, as we did analyse from the data
(Michaely and Roberts, 2011). The members of Keiretsu group which are Japanese firms are
smooth less and likely to face rarer conflict of interest and lower information asymmetry.
(Dewenter and Warther, 1998). The firms with only private debts are less smooth then the
firms with the public bond rating as they are smoother (Aivazian, Booth and Cleary, 2006).
The studies which were conducted earlier suggest that there is logical difference in the
smoothing behaviour, this article made several contributions for that. First of all there are
some cross sectional evidences for a large number of firms which are selected for sample in

United States. Most importantly, while Keiretsu/ non-keiretsu; public/ private and rated/
unrated divisions are associated with the numerous frictions of market, these measurements
are only not sufficient to distinguish in between the competing theories. The empirical
evidence which is in existence not only provides importance of dividend smoothing but also
provides guidance to growing literature which is limited.
This research provides new results in bulk for the firm’s smoothing policies. The
primary findings of the firms are highlighted here. The first thing is that the traditional
procedures of smoothing of dividend are biased and are not ideal for analysing the cross-
sectional changes in the policy. Our concern is regarding the connection among the speed of
adjustment and the small- sample from the Lintner’s model in estimating the autoregressive
models (Hurwicz, 1950). The another concern is regarding the dividend targets of different
firms as the targets of dividend today are somewhat changed from the Lintner’s model (Brav
et. al., 2005). To overcome these concerns we did performed two alternative measures
regarding smoothing and used a simulation exercise for that.
The variation in cross- sectional dividend smoothing is then documented and even
better cross- sectional dissimilarity in total dividend smoothing. The same policy of dividend
smoothing is not followed by all the firms. The pay-out policy is linked to market frictions
and the smoothing behaviour varies in accordance with the previously linked pay-out policy.
The proxies for information asymmetry are found to be correlated negatively with smoothing.
Low dividend yield funds, younger firms, firms with very high return and earning volatility,
smaller firms and the firms with accurate analyst forecast and more isolated and with less
analysts are all forecasted to be smooth less. By that time the results indicate that the firms
that smooth the most are usually substance to agency conflict. The firms are considered to be
cash cows, with weaker governance, with low growth prospects and which are observed by
institutional investors are considered to be smoother. Results that we figured out are robust
across various measures of empirical and smoothing methods.
Over the past few centuries we documented a stable and more substantial rise in the
amount of dividend smoothing. As we encountered that most of the hike in smoothing have
ensued before the use of repurchasing system this shows that the contribution of repurchases
is less in this process (Grullon & Michaely, 2002). We came to know that dividend
smoothing is not attributable to the nature of the firms as we did the analysis using both panel
regression and subsample from constant firm.
Our practical findings make available evidences of between which firms and when
dividend smoothing is extreme prevalent. The underlying theories can be related with
dividend smoothing to elaborate the motive of this study. Theories which present the
dividend smoothing of the firms are primarily based on agency consideration or on
asymmetric information (Kumar, 1988; Brennan and Thakor, 1990; Fudenberg and Tirole,
1995; DeMarzo and Sannikov, 2008; Guttman, Kadan and Kandel, 2010). Soothing can be
related to tax planning and external finance cost as we do analyse other studies (Miller and
Scholes, 1978)

Asymmetric information model is challenged by our present results. For example: this
model implies that the important determinant of dividend stickiness is adverse section.
Contrary to that our results shows that the firms which are more uncertain and where
information asymmetry is more they smooth the tiniest. The similar critique is applied to the
signalling model that is the firms which are facing high level of asymmetric information
smooth the least is also applied by Allen and Michaely, 2003 and DeAngelo, DeAngelo, and
Skinner, 2009.
Dividend smoothing arises from the effort to evade external finances which are costly
(Almeida, Campello & Weisbach, 2004). The firms with least access to highest dividend
levels and external capital are found to be most prevalent. The firms that are held mostly by
the individual investors will tend to smooth more as what the tax based model imply.
Apart from this, the evidence is all the more reliable with the agency conflicts
that motivate smoothing. For example: A favourable connection between smoothing and the
level of dividend and among the severity of free cash flows and smoothing issue is predicted
in it (Easterbrook, 1984; Allen, Bernardo, and Welch, 2000 and DeAngelo and DeAngelo,
2007). The results of this class suggest that these agency models present the future
development.
We additionally document that dividend pay-out is regulated by the time-series things
of the firm’s revenues. In line with this survey firms with more cyclical incomes smooth
more and firms with constant earnings chains smooth less (Lintner, 1956). We also establish
that firms with smooth incomes tend to smooth dividend less. By that time, our outcome
reflect that the differences in the policy of dividend in excess of the differences in behaviour
of the smooth incomes. We also marked asymmetry in the smoothing activities: Firms tend to
adjust dividends faster when they are not able to meet their target as compared to when they
accomplish their target.
We did explored two suggestions of the significance of the importance of share
repurchase as it is growing and is a component of payout policy. First of all we examined that
whether the cross sectional variances are carried by the differences in the repurchase in the
dividend payout. We found that the firms that repurchase most will smooth dividend more as
compared to the firms that do not repurchase. All the more, cross-sectional variances in the
smoothing of dividend not clarified by variations in the repurchase policies. Secondly, we
discover the range to which smoothing approaches (as opposed to dividend) are steady with
firms’ total payout or smoothing policies. The firms are not able to cope with the time-series
things of total smoothing to the amount for dividends. As the payout is further unstable than
dividends, issues that elucidate smoothing of dividend are not that effective in elucidating the
cross-sectional difference in payout smoothing. The biggest factor to determine the total
payout and dividend smoothing is the amount of repurchase done by the firm.
The articles section 1 is about the theories of smoothing of dividend and their
consequences in regard to the cross-sectional variations in the smoothing policies. The
Section 2 of the article throws light on a vital issue which is somewhat technical and is about

the measurement of smoothing of dividend. We identified two methods of dividend
smoothing in this study one is the Lintern’s model of speed adjustment and another is the
relative model between dividend and earning. With the help of simulation, we analyzed these
measures and found out that these measures make a distinction among the degree of
smoothing in the firms and stick over to various policies of the firm for dividend. In the third
Section, we did present that how we extracted the sample firms and their dividend policies
from Compustat, CRSP and 13F and their data of the years 1985- 2005. Section 4 of the
article presents the main results regarding the cross sectional variances between smoothing of
dividend and the variations in the smoothing behavior of firms over the time span. Section
4.1 presents a report on the cross sectional findings and section 4.2 presents the smoothing
behavior Section 4.3 presents a relation between our outcomes and methods of payout policy
as well as smoothing of dividend. Section 4.4 of this article presents the link between
earnings dynamics and smoothing and Section 4.5 of this article presents the unevenness in
dividend speed that change from the target. The articles Section 5 elaborated the
consequences of the repurchase policy for total payout smoothing and dividend. Section 6
presents a number of tests of robustness, and the 7th Section is conclusion.
Theories of Dividend Smoothing:
In spite of the point that the smoothing of dividend is just about a piece of confidence
and was first acknowledged before 50 years. Here is shockingly petite harmony on economic
forces that lead the firms to act in this effective way. We render an outline of the prevailing
theories of smoothing of dividend as well as their pragmatic implications. The models which
we analyzed can be distributed into those which are mainly centered on uneven material and
those which are encouraged by agency concerns. Given is the nature of almost all the theories
we explained, we focused on the firms which are creating smoothing conduct rather than
proportional statics. It will focus on the firms of different types that show smoothing actions
and later describe which approaches of models are more reliable with empirical evidence.
Information asymmetry model:
Among uneven information models, Kumar, 1988; Kumar and Lee, 2001 and
Guttman, Kadan, and Kandel, 2010 offer the models in which dividend is used as a source of
private material about future and current cash flows. Conversely, comparable models which
are used to clarify the presence of dividends (For Example: Bhattacharya, 1979; John and
Williams, 1985 and Miller and Rock, 1985), these show the presence of somewhat revealing
equilibrium. Firm that are within an assured range group each other, however they are distinct
from the firms which are outer that range. Forceful extrapolations of all these simulations can
then create smoothing of dividend. Comparative statics propose that smoothing should be
effected with an increase in equity risk factors as volatility of cash flows increases. As the
investment opportunities expand and the investment prospect abbreviates (Guttman, Kadan,
and Kandel, 2010)
The results of this smoothing comes from the waving efforts, this action should be
predominant amongst the firms for whom the advantage of signaling is more than those

experiencing considerable amount of information irregularity. Let’s talk about an example of
more opaque firms, with more growth opportunities and fewer tangible assets will exhibit
somewhat more smoothing of dividend. We can expect that the amount of smoothing will
reduce over time as the evidence released by corporations and generated by analysts has
improved significantly with enhancements in market sophistication and information
technology.