International Journal of Business and Social Science

Vol. 2 No. 6; April 2011

The Reliability of Dividend Discount Model in Valuation of Common Stock at the

Nairobi Stock Exchange
Tobias Olweny
Department of Commerce and Economics, JKUAT-Kenya
Email: [email protected]
Abstract
Valuation of common stock is very important yet a very complex process. The stock requires a deeper analysis
compared to preferred stock or debts. The major techniques of valuation of common stock are:
(i) Relative valuation models which is based on the earnings power of the firm, the book value and sales.
(ii) The discounted cash flow techniques, where the value of stock is estimated based upon the present value of
some measure of cash flow including dividends, operating cash flow among others.
The study was conducted to establish the reliability of the dividend discount model (which is based on the
discounted cash flow techniques) on the valuation of common stock at the Nairobi Stock Exchange. Data was
collected in form of share prices, market indices and dividend per share from the Nairobi Stock Exchange
secretariat, and were used to predict share prices for each of the eighteen companies studied. Market model was
used as a model of equilibrium to provide a link between the expected values which are non observable and real
values that were used in testing the model. Predicted share prices were compared with the actual prices by
computing the differences between them. The differences were then subjected to t-test. The test of significance
showed that out of the eighteen companies studied; only three showed that the differences were significant. I
therefore concluded that the dividend discount model is not reliable in the valuation of common stock at the
Nairobi Stock Exchange.

Keywords: Dividend discount model, stock valuation, stock exchange

1.0 Introduction
The investment process involves decisions by an investor on what marketable securities to invest in, the extent of
the investment and when the investment should be made. The investment environment includes the kinds of
marketable securities that exist, where and how they are bought or sold. Investment is a commitment of funds for
a certain period of time in order to derive a rate of return to compensate for the time funds are invested, the
expected rate of inflation during that time, the liquidity premium and the risk involved. When an investor commits
certain funds, he expects a stream of returns over the period of ownership. The investor could be an individual, a
government, a pension fund or a corporation. The investor therefore trades a known shilling amount today for
some expected future stream of payments that will be greater than the current outlay. Since an investment
involves sacrifice of a current shilling for a future shilling, time and risk must be taken into consideration. The
sacrifice made today is certain while the returns expected in future are uncertain. Discounted cash flow formulas
take into account the risk on the value of an investment; hence the value can be determined as follows:
Vo = C1
+
C2
+ .. + Ct
+ Cn
(1)
(1 + k1)1
(1 + k2 )2
(1 + kt )t
(1 + kn) n
Vo = the current or present value of an investment.
Ct = expected returns at time t.
kt = required rate of return for each period
n = the number of periods over which returns are expected to be generated.
PV (stock) = PV (Expected future dividends, interest payments, earnings or capital gains).
Valuation of common stocks is very important; however it is more complex than that of other stocks. The investor
will ensure that the expected rates of returns correspond with the risk involved. Equity shareholders are the
residual owners of a corporation. Their return is less certain than the return to lenders or preferred stockholders.
The book value of equity is the shareholders equity of a corporation less the par value of preferred stock divided
by the number of shares outstanding (Van Horne,2001).In valuation of ordinary shares a concept known as
intrinsic value is commonly used as means of estimating the anticipated returns. The intrinsic or true value of any
asset is based on cash flows that the investor expects to receive in the future from owning the asset.
127

 Centre for Promoting Ideas, USA

www.ijbssnet.com

The current market price can be compared with the intrinsic to find out whether a share is undervalued or for an
investor to be willing to invest in the stock he/she requires a market capitalization rate and hence the price of a
share of stock is the present value of all expected future dividends per share discounted at market capitalization
rate.
Vj =
D1
+
D2
+
D3
+ +
Dn (2)
(1 + k1)1
(1 + k2)2
(1 + k3) 3
(1 + kn)
Vj= value of common stock j
Dt = dividend during period t
k= required rate of return of stock j (market capitalization rate)
t= the holding period
As t approaches infinity:

Vj
=
Dn .. .. (3)
t=1
(1 + k n)n
The model was initially set forth by Williams (1938) and subsequently expanded by Gordon (1963) cited in
Brealey &Myers (2000: 64-66).For the above formula to apply, the capital markets must be well functioning i.e.
where all securities in an equivalent risk class are priced to offer the same expected returns. The focus of the
dividend discount model is on determining the true value of one share of a particular companys common stock,
even if larger purchases are being contemplated because it is assumed that larger purchases can be made at cost
that is a simple multiple of the cost of one share. To use the above equation (3), an investor must forecast all
future dividends. Certain assumptions have to be made, these assumptions concern dividend growth rates. That is,
the dividend per share at any time t can be viewed as being equal to the dividend per share at time ,t-1 times the
growth rate of gt (Sharpe et al 1999).
Dt =Dt-1(1+gt) (4)
Or Equivalently
Dt-Dt-1/Dt-1=gt(5)
Earnings per share model relates to the earnings per ordinary share at any given time multiplied by the price
earnings ratio at time (t):
Pit = EPSit x (P/E) it .. (6)
Pit
=
the estimated value of ordinary share
EPSit
=
the estimated earnings per share i at time t
(P/E) it =
The estimated price earning ratio of share i at time t
The application of EPS valuation model requires that:
i)
The analysts must select some time horizon for the analysis and once this is done, the growth in
earnings per share over this time horizon must be forecast. The EPS forecast facilitates a forecast of
the horizon period.
ii)
An appropriate price earnings ratio must be selected.
iii)
The firms performance must be considered as well as the market performance of the horizon period.
Earnings are important to investors because they provide cash flows necessary for paying dividends. Earnings per
share method is also simpler and easier to use and can apply to stocks that do not pay dividends. Reported
earnings are important determinants of stock prices. Empirical studies suggest that stock price movements are
associated with earnings changes and differences between actual and predicted change lead to price adjustments
(Elton and Grubber 1995). Despite the simplicity of the model, it is difficult to estimate price earnings ratio. The
major determinants of price earnings ratio are dividends payout, earnings growth, and earnings volatility cannot
be easily forecasted.Miller and Modigliani (1961) argue that dividends are irrelevant and that it does not matter
whether a firm capitalizes dividends or earnings, because price changes in shares will be reflected on both
earnings and dividends and those investors would select whether to receive income as dividends or by sale of
shares. In the real world it is generally accepted that dividends policy matters because of presence of transactions
cost, taxation effects, monopolistic effects in the markets for borrowing and investment and indivisible investment
opportunities (Wilkes, 1977). The dividends discount model therefore has a strong foundation for share valuation.
The dividend discount model is perceived as an appropriate model in this study because: first there is no sound
methodology for evaluating price earnings ratio which in essence is the reciprocal of the required rate of return.
128

International Journal of Business and Social Science

Vol. 2 No. 6; April 2011

Secondly, dividends are the flow of returns received by the investors. Thirdly others have intensively used the
dividend discount model in valuation of securities. There is evidence that complex dividend discount models
improve the accuracy of the forecast and therefore are useful in selecting shares (Fuller and Chi Cheng 1984;
Sorensen and Williamson 1985).Fourthly, the dividend discount model is based on a simple, widely understood
concept. The fair value of any security should be equal to the discounted value of cash flows expected to be
produced by that security. Fifth, the basic inputs for the model are standard outputs for many large investment
management firms, that is these firms employ security analysts who are responsible for projecting corporate
earnings(Sharpe et al 1999).Finally it is argued that the dividend discount model provides a consistent and
plausible framework for imbedding analysts judgments of share value(Michaud and Davis,1982).As a
qualification of security value, the dividend discount model is often a first and critical step in a quantitative
investment management program.
The dividends and earnings valuation methods have not gained widespread or wholehearted acceptance by
investors because of the choice of required rate of return. It has been the most difficult variable to estimate.
According to Brigham and Gapenski (1996), the required rate of return of an investment is determined by:
1) The economys real risk-free rate of return plus
2) The expected inflation rate during the holding period plus
3) A liquidity premium plus
4) A risk premium.
The required rate of return therefore depends on both systematic and the unsystematic risk. The two elements are
separated clearly when the return for a single stock is related to the return on the market portfolio of all stocks. Of
the two, systematic risk is the most dominant determinant of the required rate of return. The market offers the
investor a risk premium in excess of his risk less rate of return for taking systematic risk (Copeland and Weston
1988).According to Elton and Grubber it is the systematic risk that is important to the investor:
systematic risk is the only important ingredient in determining expected returns and
that non systematic risk plays no role. Put in another way, the investor gets rewarded for
bearing systematic risk. Elton and Grubber (1995:301)
Systematic risk =

Cov (j, m)... (7)

2m
Where Cov (j, m) = Covariance between the securitys return and the market.
2m = Market Variance
Systematic risk is referred to as Beta
Therefore: Bj = Cov (j, m) (8)
2
The required rate of return can be calculated once beta is known using Capital Asset Pricing Model:
E (Rj) = Rf + (Rm Rf) Bj. (9)
Where E (Rj) = the required rate of a security
Rf = the risk-free rate,Rm = the expected market return and Bj = the systematic risk of security j
Capital Asset Pricing Model can be used to value assets like ordinary shares.Risk premium is the market risk
premium (Rm-Rf) weighted by the index of the unsystematic risk Bj of an individual sec urity. If the general
economy is static, industry characteristic are unchanged and management policies have continuity, the measure of
Bj of a security will be relatively stable when calculated for different time periods. If the condition of stability
does not exist the value of Bj will vary over different periods. As indicated above:
Rj = f (expected real rate, expected inflation and liquidity).
E (Rj) = Rf + Cov (Rm,Rj) (E (Rm) Rf) (10)
2m
E (Rm Rf) Can be replaced by
2m
E(Rj) = Rf + Cov (Rm,Rj).(11)
For the model to be useful in this study, Bj must remain constant over time. The beta values in CAPM can be
computed using the market model since forces within the market and the stock market have common significant
influence or changes in prices in many if not all stocks.
129

 Centre for Promoting Ideas, USA

www.ijbssnet.com

The stock prices are therefore sensitive to the above forces hence the required return of a share:
E(Rj) = + BjRm + E1..(12)
Where E (Rj) = average monthly rate of return of a given share j
Bj = beta, the market sensitivity of share j
Rm = monthly rate of return of NSE index
Ei = random variable representing variability in E (Rj) not associated with variations
in Rm.
Therefore CAPM allows us to determine the appropriate discount rate for discounting expected dividends and
terminal value to their present value. CAPM has a number of assumptions and some of them do not hold in the
real world; however it is still useful in evaluating financial decisions. The question of whether investors
emphasize on dividends or earnings per share observed cannot be easily resolved. However it has been observed
that the dividend discount model is useful for valuation of a stable mature entity where assumption of a relatively
constant growth for along term is appropriate (Reilly & Brown 2000). Earnings per share can be used when the
aggregate market is not either seriously overvalued or under valued, implying that markets are slow or inefficient
processors of information.
1.2 Approaches to Valuation
The major schools of thought in determining security value and behavior of prices are:
i)
Fundamentalists.
ii)
Technicians.
iii)
Efficient market hypothesis.
According to fundamentalists, the price of a security at any time is equal to the discounted value of the stream of
income from the security. They believe that the value of a security depends on the underlying economic factors
and hence the value of a stock is determined by analyzing variables such as current and future earnings, cash
flows, interest rates and risk variables (Reilly and Brown 2000).Fundamental analysis therefore involves market
analysis, company analysis and portfolio management. Technicians argue that the market value of a share is
determined by the interaction of supply and demand having very little to do with earnings and dividends. The
supply and demand are governed by several factors both national and international. They believe that the prices of
individual securities and overall value of the market move in trends, which persist for appreciable length of time,
and that prevailing trends change in reaction to shifts in supply and demand relationships. These shifts no matter
why they occur can be detected sooner or later in the action of the market itself. The analysis focuses upon the
study of the stock market itself and not upon external factors that influence the market. The external factors are
assumed to be fully reflected in the share prices and the volume of stock exchange. The market itself provides all
information for analyzing and predicting stock price behavior.
Efficient Market Hypothesis contends that a change in stock prices occurs randomly. It is not possible to predict
future prices. They argue that price movement whether up or down occurs as a result of new information and
since investors cannot predict the kind of new information it is not possible to predict future price movements.
Efficient Market Hypothesis clearly conflicts with the technical analysis. The theory states that previous prices
changes or changes in returns are useless in predicting future prices implying that the work of technical analysis is
useless. The vast majority of studies that have tested the weak form efficient market hypothesis have found that
prices adjust rapidly to stock market information, supporting the random walk theory (Fama: 1970, 1991).
Most security analysts support fundamental analysts, and even technical analysts admit that a fundamental analyst
with good analytical ability and a good sense of informations impact on the market should achieve above average
returns. Technicians argue that the fundamental analyst can achieve these returns only if they can obtain new
information before investors and process it correctly and quickly. It is difficult for an investor to obtain new
information frequently and processes it quickly. This study is conducted in line with the fundamentalists
perspective. In conclusion; superior analysts or successful investors must understand what variables are relevant
to the valuation process and have the ability to do a superior job of estimating these variables. Alternatively one
can be superior if he or she has the ability to interpret the impact or estimate the effect of some public information
better to others.
1.3 Effects of Dividends on Share Prices
The price of common stock is a function of the level of a companys earnings, dividend risk, the cost of money
and future growth rates (Elton &Grubber 1995).
130

International Journal of Business and Social Science

Vol. 2 No. 6; April 2011

A valuation model converts a set of forecasts of a series of company and economic variables into a forecast of
market value for the companys stock. Inputs to a valuation model include future earnings, dividends and
variability of earnings. Valuation model therefore is a formal relationship that is expected to exist between a set
of corporate and economic factors and the markets valuation of these factors. The dividend discount model
explains the relationship between the share price and dividends paid in a particular period. In a world of no taxes,
Miller and Modigliani (1961) proved that payout has no effect on shareholders wealth (share prices).Dividend
policy is therefore irrelevant. They argue that the value of the firm depends on the firms earnings which results
from its investment policy. When corporate and personal taxes are introduced into the model, shareholders wealth
decreases when dividends are paid out. Empirical research on the relationship between dividend yields and
common stock prices has, in most cases not looked at the effect of departures from an optimal dividend pay out
(Weston and Copeland 1992).
Although managers behave as though dividend policy is a critical variable, their behavior does not imply that
market actually values that attention. Given the conflicting impacts of market imperfections, the relevance of
dividend policy becomes an empirical question. A critical question may be asked what does real world stock
price suggest about how dividend policy affects equity valuation? In a real world there are market imperfections
which include taxation effects, transactions costs, monopolist effects in the markets for borrowings, asymmetric
information and agency costs. Therefore a firms dividend policy might impact on the value of its shares.Brennan
(1970) added a dividend yield variable to the capital asset pricing model, and reasoned that firms with higher
dividend yields should have higher pre tax returns than equity in firms with lower payouts. This higher yield
would compensate investors for higher taxes and, therefore equates after tax returns holding constant for
systematic risk. Empirical tests of Brennans model however, have not yielded definitive results with respect to
dividend yield coefficient as noted by Black and Scholes (1974).
Long(1978) conducted a unique study on the relationship between dividend yield and market returns. He
examined prices of two classes of common stock in a firm (Citizens Utilities Company of Atlanta, Georgia) with
two classes of common stock. One pays cash dividend while the other class provides an equivalent dollar value in
extra shares via stock split. Tax models of dividend policy predict the stock split shares will sell at a premium
relation to the cash dividend shares. Surprisingly, Long found the opposite. The cash dividend shares sold at a
significant premium to the other class of shares. Although this result represents only one firm, it suggests the
market value cash dividend over capital gains. If taxes play a large role in the composition of investors
portfolios, high yield stocks to escape taxes, while low tax bracket investors should be more indifferent to the
dividend policies of firms. In other words tax induced dividend clienteles should exist. Lewellen et al (1978)
examined the dividend yields on portfolios held by individual investors in a cross section of tax brackets and
found weak support, suggesting that high tax bracket investors chose stocks that paid lower dividend yields.
Miller and Modigliani state that the tax differential in favor of capital gains is undoubtedly the major systematic
imperfection in the market. Implying that existence of differential taxes on income and capital gains should make
the shares of corporations that pay low more desirable, and thus a corporation can increase the value of its shares
by reducing its payout ratio. Nevertheless, Miller and Modigliani still conclude that dividend policy has no effect
on the share values.Finally, a popular avenue of research of tax effect and tax-induced clientele effect has been the
stock price behavior across the dividend day. Elton and Grubber (1970) authored an influential study of stock
price behavior around the ex-dividend day, they found less than full dividend price drop on the dividend day
during periods of differential taxation. Their study concludes that ex dividend price behavior of stocks is
evidence of investors preference for capital gains over cash dividends. Empirical studies that clearly model how
dividend policy impacts firms value due to corporate flotation costs and investors translates are, unfortunately not
available. The Agency theory models that suggest dividend policy can help reduce agency conflicts between
bond shares and stockholders have, to date, not been tested.
With respect to whether managers use dividend policy to convey news about changes in firms value based on
their inside or asymmetric information, empirical studies are more definitive. Studies have shown that stock
prices significantly rise when dividends are increased by more than the expected amount, and vice versa. The
stocks splits study by Fama et al (1969) as cited in Fama (1976), found that when splits were accompanied by
dividend announcements there was an increase in adjusted share prices for the group that announced dividend
increase and a decline in share prices for the dividend decrease group. Other studies of the effect of unexpected
dividend changes on share prices were made by Pettit (1972), Watts (1973) Kwan (1981) and Aharony and Swary
(1980).
131

 Centre for Promoting Ideas, USA

www.ijbssnet.com

Healy and Palepu(1988) found that investors interpret announcements of dividend initiations and omissions as
managers forecast of future earnings changes. Further, Brickley (1983) has shown that specially designated
dividends which bear such labels as special or extra when announced by the board, convey less favorable
information than do increases in regular dividend. These findings suggest that market regards specially
designated dividends as more temporary versus the permanent increase implied by an increase in regular dividend.
Empirical evidence also shows that stocks prices do respond positively when firms announce repurchase
programs. However, the economic factors that lead managers to choose cash dividends versus stock repurchases
are not well understood.
To develop a theory that explains choice between payout mechanisms, the differential costs and benefits between
the alternatives must be specified. Based on asymmetric information arguments, Barclay and Smith (1988), say
that if managers time their repurchases in periods when they think, based on outside information, that their stock
is undervalued, selling shareholders lose while remaining shareholders, including non selling managers, win. Such
gaining activity cannot be conducted to the disadvantage of selling shareholders since the market is aware of
managers ability to exploit inside information. A higher market price will be attached to firms with a regular cash
dividend policy versus a more sporadic share repurchase policy. This observation might explain the reason why
cash dividends are much more commonly used as a method of cash disbursement than stock repurchase.
In conclusion, it is difficult to summarize the dividend puzzle. As Black (1976) noted, The harder we look at the
dividend picture, the more it seems like a puzzle with pieces that just do not fit together. In a perfect capital
market world with both certainty and uncertainty cases; dividend policy is irrelevant, a trivial detail that managers
could as well ignore. In a world of imperfections dividends policy is favored. However, certain market
imperfections seem to favor a managed dividend policy, others favor residual dividend policy, yet other
imperfections are ambiguous as to their impact. The empirical evidence on whether dividend policy affects stock
value or required returns is mixed and generally inconclusive. What is unknown dominates what is known about
dividends policy. Little evidence suggests an appropriate dividends payout level. However, compelling evidence
suggests that stock price changes accompany changes in cash dividends and stock repurchase announcements.
1.4 Valuation of New Issues
In the past decade, the Kenya government has embarked on privatization of state corporations. It is therefore
critical to discuss how share valuation using the fundamental analysis can be used to determine shares to be
offered to the general public for subscription. The price of a firms shares is influenced by all factors that affect
the expectations of the firm and its share.Reilly and Brown (2000) recommend a three step valuation process:
i)
Analysis of alternative economies and security markets.
ii)
Analysis of the alternative industries.
iii)
Analysis of individual companies and stocks.
Economic factors exert force on all industries in the economy. They include monetary and fiscal policies, political
forces and international environment. A number of models have been developed which have found an important
linkage between the money supply and the level of share prices. These include Hamburger and Kochin(1972),
Homa and Jafee (1971) and Kraft and Kraft (1977). Chen et al (1986) found that inflation, industrial production,
risk premium and the slope of the term structure of interest rates are the main factors that affect expected returns.
These factors change the business environment and add to the uncertainty of sales and carrying expectation and
therefore the risk premium required by investors (Kerandi 1993).Industry analysis is critical to valuation, since it
is a prospect within the global basis environment, and determines how well or poorly an individual firm will
perform. The firms do well in poor industries and vice versa. Finally an enumerator can analyze and compare the
entire industry using financial data (Page and Paul, 1979). This is difficult especially for the firms offering
ordinary shares to the public for the first time, since financial data provided in the prospects are likely to be
limited to a short time.
2.0 Research Methodology
2.1 Population
All the companies quoted in the Nairobi Stock Exchange as at 31st December 1999.
2.2 Sampling Plan
The sample consists of only the companies trading on ordinary shares. The assumption made here is that investors
require five years to assess the risk of the stock. The study covered five years from 1st January 1995 to 31st
December 1999. The five-year period, and especially December 1999 was chosen to fall within the period used in
a previous study (Sawaya, 2000).
132

International Journal of Business and Social Science

Vol. 2 No. 6; April 2011

Sawayas study dealt with estimation of systematic risk for the Nairobi Stock Exchange and his findings,
especially market portfolio beta and the percentage of diversification of the total unsystematic risk, are important
to this study. The assumption that the market portfolio beta is approximately one, can only be assured by using the
period that he used to estimate the beta. Stratified sampling was used in the sample selection for the study. The
quoted companies were divided into two groups; actively traded and non-actively traded companies. Stratifying
was done by observing changes in the shares prices and the rate of buying and selling using daily price lists
supplied by NSE secretariat. The sample is made up of eighteen companies classified as actively traded. The first
six months of the year 2000 was used to test the model. The period was chosen because it is expected that the
parameters involved were almost constant.
2.3 Data Collection
Data required was collected in form of secondary data, and in particular the bid prices of the stock. Annual
dividends per share were used, as monthly dividend per share; since the investors reaction to these figures are the
same irrespective of whether they are looked at from a monthly or annual point of view. Secondary data was used
in the study. The following data were collected:
1.
Bid prices of the stock.
2.
Annual dividend per share.
2.4 Data Analysis
As indicated in the introduction of this paper price of a share:
Po
=
D1 +
D2 +... +
Dt
+Dn+Pn
(14)
1 + k1 )1 (1 + k2)2
(1 + k) t (1 + kn)n
Where Dt= expected dividend at some time horizon t.
n
=
time horizon n
k
=
required rate of returns
Pn
=
expected terminal price
Po =
the present price
The dividend discount model represents a formal notation for the statement that share prices depend on expected
returns, but this is not sufficient to make the statement testable. To provide a level between expected values and
real values a model of equilibrium is required. The market model therefore can be used as a model of
equilibrium. It is a single factor model, which shows the relationship between the security return and the market
return. Following Fama (1976) the model can be used in an efficient capital market.
Assume that all events of interest take place at discrete points in time t-1,t,t+1,e.t.c.
Then define t-1 = the set of information available at time t-1, which is relevant for determining security
prices at t-1.mt-1 = the set of information that the market uses to determine security prices at t-1. Thus
mt-1 is a subset of t-1;mt-1 contains at most the information in t-1, but it could be less.Pj, t-1 = the price
of security j at time t = 1, j = 12,n) where n is the number of securities in the market.fm(P1t+T.Pn,
t+T/ mt-1) = The joint probability density function for security prices at time t+T( T>= 0) assessed by
market at time t-1 on the basis of the information mt-1.f(P1 , t+T,.,Pn, t+T/ t-1) = the true joint
probability density function for the security prices at time t+T(T>= 0) that is implied by the
information t-1.
(Fama, 1976:134)
The market model assumes joint distribution of security prices is multivariate normal. The market assesses a
joint distribution of security at time t. The market equilibrium is obtained at time t-1 at price sets P1, t-1,.,Pn,t1 when the investors demand for individual securities equals to the outstanding supply of the security. Since the
true joint distribution of the prices of different securities at time t is multivariate, the joint of security returns f
(R1 ,., Rm/ t-1),is also multivariate normal (Fama,1976).If a bivariate normal distribution is obtained
from the multivariate function, a linear regression equation results:
E (Rjt/Rmt) = a + Bj Rmt . (15)
t = 1, 2 ..t
Rjt
=
the returns on securitys from time t= 1 to time t.Rmt = average of the returns of these stocks
from time t=1 to time t.Where Bj
= cov (Rjt/Rmt) / 2Rmt and
a = E(Rjt t-1) - BjE (Rmt)/ t-1 .. (16)
T=1,2.t
133

 Centre for Promoting Ideas, USA

www.ijbssnet.com

t = 1 the set of information available at time t=1, which is relevant for determining security prices at time t=1.P j
t=1 The price of security at time t=1, j = 1,2.n where n is the number of securities in the market.
Rit =
(Pjt Pj,t-1)(17)
(Bjt1)
a=E (Rit)/t-1, Rmt) = a+Bi Rmt).(18)
j at time t which is reduced to
Rjt = a + Bj Rmt + Ejt(19)
Ejt = the deviation of Rjt from its conditional expected value.
Therefore E(Ejt/t-1, Rmt) = 0.0 . (20)
In deriving our expected values using market model, we will assume that during each period the market sets
prices, so that fm (Rjt, Rmt/mt-1) is perceived as a bivariate normal distribution of Rjt, and Rmt, and is constant
through time, implying that aj, Bj, and the time distribution of Ejt are the same period after period. The expected
terminal price will be computed from the market model to obtain the monthly returns for each company. The
market portfolio m will contain all ordinary shares on the Nairobi Stock Exchange. To derive Rmt, we will
average the returns of these shares for the period 1995 1999. The estimators of the market model cov Bj and j
involves substituting unbiased estimators of E(Rj) , E(Rmt) and Cov (Rj, Rmt) .Where P o = present value of
ordinary share
ke
=
required rate of return on share j
T
= holding period
Po
=
n
Dt
+ Dn
(21)
(1+ke)t
(1 + kn)n
t=1
T
Rj
=

Rjt (22)
t=1
T
T
Rm
=

Rmt.. (23)
t=1
T
T
S 2 (Rm) =
Rmt Rm (24)
t=1 T 1
T
Sjm
=
(Rjt Rj) (Rmt Rm)..(25)
t=1
T1
Therefore Bj =
Sjm ...(26)
S2 (Rmj)
and aj = Rj + Bjm .(27)
The basic CAPM was used to derive the beta for each of the companies to be studied. Bjs computed for each
company will be our beta values. One year government of Kenya Treasury bills rate plus market returns Rm
computed when deriving the market model will give us full market returns.Therefore average market returns is
computed using:Rm = (mt mt-1) (28)
Mt - 1
Where Rm = monthly market returns at period t,Mt = market index at period t and Mt-1= market index at period t1.
The results were summarized using descriptive statistics such as mean and standard deviation. Each price
obtained was compared to the actual price for that period.This was done by finding the difference between the
actual and predicted prices then testing whether the difference between the two are significant.The following
hypothesis was tested:
Ho: There is no significant difference between the actual share prices
and the predicted share prices
using dividend discount model.
Ha:
There is a significant difference between the actual and the predicted share price using dividend
discount model.The t- test was used as the appropriate test statistic.
134

International Journal of Business and Social Science

Vol. 2 No. 6; April 2011

T = (d-)
sn
(29)
d=
the means of the differences between the two samples.
s=
the standard deviations of the differences.
n=
number of observations
The dividend discount model qualified as a reliable model depending on the number of companies for which it
predicts share prices that are not significantly different from the actual one

3.0 Data Analysis and Findings

3.1The Market Model
Monthly returns computed from the share prices and market indices were used to derive the market model for
each company, as indicated in appendix C. I, obtained beta values using CAPM. CAPM was assumed to estimate
the required rate of return for each company (table 1). The market model was then used to forecast expected share
prices for the first six months of the year 2000, and the results summarized in table 2 for each of the companies
studied. To determine the significance of relationship between the two prices (the predicted and actual prices), the
differences computed were used to carry out hypothesis testing for each company. The market model was not a
good predictor for fourteen companies (about 78 percent) and was a good predictor for only four companies
(about twenty-two percent).This further suggest the possibility of market inefficiency (NSE).
TABLE 1 THE MARKET MODEL DERIVED FOR EACH COMPANY

COMPANY
BROOKE BOND
GEORGEWILLIAMSON
KAKUZI
SASINI TEA AND COFFEE
DIAMOND TRUST
NATION MEDIA GROUP
STANDARD N.PAPER
BARCLAYS BANK
C.F.C LTD
B.A.T
BAMBURI PORTLAND LTD
E.A.B.L
K.P.L.C LTD
TOTAL KENYA LTD
STANDARD CHARTERED
K.C.B LTD
CAR&GENERAL LTD
I.C.D.C LTD

MARKET MODEL
R=-0.0017-0.2104RM
R=0.0029+0.2100RM
R=0.0015+0.0740RM
R=-0.0094+0.2043RM
R=-0.0150+0.1599RM
R=0.0145+0.0783RM
R=0.0625+0.4430RM
R=-0.0048+0.1933RM
R=0.0015-0.3909RM
R=0.0114+0.1385RM
R=0.0303+0.1088RM
R=-0.0043+0.1881RM
R=0.0132+0.0727RM
R=-0.0149+0.1499RM
R=-0.0068+0.1871RM
R=0.0145+0.0783RM
R=0.0476+1.0678RM
R=0.0079+0.2059RM

BETA
-0.2104
0.21
0.074
0.2043
0.1599
0.0783
0.443
0.1933
0.3909
0.1385
0.1088
0.1881
0.0727
0.1499
0.1871
0.0783
1.0678
0.2059

TABLE 2 PREDICTED SHARE PRICES USING THE MARKET MODEL

MONTHS
COMPANY
BROOKE BOND

GEORGE WILLIAMSON

KAKUZI

SASINI

Actual
Predicted
Difference
Actual
Predicted
Difference
Actual
Predicted
Difference
Actual

Jan-00
104
103.69
0.31
93
93.39
-0.39
97.5
87.14
10.36
45

Feb-00
104
104.32
-0.32
87
92.94
-5.94
77.5
87.04
-9.54
36

Mar-00
88
104.26
-16.26
87
93.1
-6.1
70
87.14
-17.14
35.5

Apr-00
78
104.82
-26.82
90
92.71
-2.71
67
87.05
-20.05
31.75

May-00
76
104.39
-28.39
77
93.2
-16.2
67
87.25
-20.25
36.5

Jun-00
74
100.38
-26.36
75
96.87
-21.87
66.5
88.51
-22.01
35

135

 Centre for Promoting Ideas, USA

DIAMOND TRUST

NATION MEDIA

STANDARD
NEWSPAPER

BARCLAYS

C.F.C

B.A.T

BAMBURI

E.A.B.L

www.ijbssnet.com

Predicted
Difference
Actual
Predicted
Difference
Actual
Predicted
Difference

44.64
0.36
25
25.64
-0.64
93
101.5
-8.5

43.89
-7.89
28
25.1
2.9
90.5
102.68
-12.18

43.42
-7.92
26.75
24.7
2.05
87.5
104.18
-16.68

42.71
-10.96
24.75
24.2
0.55
75
105.41
-30.41

42.41
-5.91
21.25
23.88
-2.63
74
107.03
-33.03

43.53
-8.53
20
24.19
-4.19
75
110.05
-35.05

Actual
Predicted
Difference
Actual
Predicted
Difference
Actual
Predicted
Difference
Actual
Predicted
Difference
Actual
Predicted
Difference
Actual
Predicted
Difference

10.75
10.49
0.26
101
102.63
-1.63
14.05
14.23
-0.18
73
78.45
-5.45
26.25
27.06
-0.81
66.5
69.78
-3.28

10.5
10.97
-0.47
115
101.41
13.59
14
14.46
-0.46
94
78.94
15.06
26
27.77
-1.77
70
69
1

10.05
11.63
-1.58
90
100.81
-10.81
15.15
14.51
0.64
64
79.78
-15.78
27.5
28.59
-1.09
70
68.63
1.37

8.05
12.18
-4.13
90
99.66
-9.66
16
14.72
1.38
62
80.31
-18.31
26.5
29.35
-2.85
69
67.9
1.1

8.75
13
-4.25
87
99.4
-12.4
13.65
14.68
-1.03
61
81.35
-20.35
28.5
30.27
-1.77
66.5
67.75
-1.25

6.1
14.82
-8.72
86
102.34
-16.34
9.8
13.7
-3.9
57
84.25
-27.25
29.25
31.76
-2.51
65.5
69.69
-4.19

Actual
Predicted
Difference
Actual
Predicted
Difference

Jan-00
93.5
96.81
-3.31
49
47.58
1.42

Feb-00
91.5
97.83
-5.8
65
46.61
18.39

Mar-00
88
99.08
-11.08
48.75
45.87
2.88

Apr-00
78
100.14
-22.14
49
44.95
4.05

May-00
50
101.54
-51.54
49
44.36
4.64

Jun-00
51.5
104.17
-52.67
51
44.86
6.14

Actual
Predicted
Difference
Actual
Predicted
Difference
Actual
Predicted
Difference
Actual
Predicted
Difference

57
56.18
0.82
35
31.97
3.03
10
10.54
-0.54
50
50.46
-0.46

75.5
55.41
20.09
31.5
35.54
-4.04
10
10.63
-0.63
45
50.48
-5.8

52.5
54.97
-2.47
25
36.03
-11.03
10
11.07
-1.07
40.5
50.82
-10.32

47.75
54.25
-6.5
26.5
36.46
-9.96
10
11.2
-1.2
46.75
50.86
-4.11

47
54
-7
27.5
37.02
-9.52
10.25
11.87
-1.62
47
51.38
-4.38

48
55.4
-7.4
28
38.06
-10.06
10.05
14.65
-4.6
49.5
53.64
-4.14

MONTHS
COMPANY
K.P.L.C

TOTAL KENYA

STANDARD
CHARTERED

K.C.B

CAR&GENERAL

I.C.D.C

136

International Journal of Business and Social Science

Vol. 2 No. 6; April 2011

TABLE 3 AN ANALYSIS OF THE DIFFERENCES BETWEEN ACTUAL AND PREDICTED PRICES BY THE MARKET
MODEL

COMPANY
BROOKE BOND
G.WILLIAMSON
KAKUZI
SASINI
DIAMOND TRUST
NATION MEDIA
STANDARD
N.PAPER
BARCLAYS BANK
C.F.C
B.A.T
BAMBURI
E.A.B.L
K.P.L.C
TOTAL KENYA
STANDARD BANK
K.C.B
CAR
AND
GENERAL
I.C.D.C

MEAN
-16.31
-8.868
-13.105
-6.808
-0.327
-22.64

VARIANCE
177.94
69.768
151.636
14.96
7.431
133.41

STD
DEV
13.339
8.353
12.314
3.868
2.726
11.55

T COMP.
(CONFID.)
13.999
8.766
12.923
4.059
2.861
12.121

NULL HYPOTHESIS
Reject Ho
Reject Ho
Reject Ho
Reject Ho
Reject Ho
Reject Ho

-3.148
-6.208
-0.675
-11.967
-1.798
-0.875
-24.423
6.253
-0.41
-6.93

10.894
117.402
3.361
225.975
0.622
5.874
501.81
37.902
111.077
30.01

3.3
10.835
1.833
15.032
0.789
2.424
22.401
6.156
10.539
5.478

3.464
11.371
1.924
15.776
0.828
2.543
23.508
6.46
11.06
5.749

Reject Ho
Reject Ho
Do not Reject Ho
Reject Ho
Do not Reject Ho
Do not Reject Ho
Reject Ho
Reject Ho
Reject Ho
Reject Ho

-1.61
-4.868

2.301
10.273

1.517
3.205

1.592
3.364

Do not Reject Ho
Reject Ho

Lev.of
significance=
Degrees of freed.
t critical

0.05
5
2.571

3.3 The Dividend Discount Model

In order to test the dividend discount model, we first estimated the required rate of return of each company as
shown in table 4. The rates of returns were then used to discount the forecasted dividend per share and the
terminal prices to their present values, for each company for the first six months of the year 2000. Table 5 shows
the actual, predicted and differences of prices for each of the eighteen companies. The results were tested for
significance by hypothesis testing on the difference for each company. Table 6 shows a summary including mean,
t-statistic and decision rule. All the eighteen companies had their shares predicted but only three had positive
results (about seventeen percent), while the rest were negative (Eighty three percent). We therefore reject our null
hypothesis and conclude that dividend discount model is not a good predictor of share prices at the Nairobi Stock
Exchange. The model cannot be relied on by companies listed in the Nairobi Stock Exchange to predict their
share prices: The results may be attributed to:
i) Inefficient market (NSE).
ii) Inappropriate discounting factors.
iii) Information differentials.
iv) Measurement and evaluation problems, among others.
As suggested earlier in this report the NSE could be inefficient, but the model can be used where all securities in
an equivalent class are priced to offer the same expected returns (where the market is efficient). Some managers
believe that the market is highly inefficient and that any valuation method (including the dividend discount
model) that is based on rationality of market participants will prove ineffective (Sharpe et al 1999). The study
assumed that the Nairobi Stock Exchange is an efficient market. Although there is active trading in the NSE,
improved liquidity, and investor protection regulations; its state of efficiency is still inconclusive. Inappropriate
discounting factors used may have contributed to the results above, since the discounting factors (rates of return)
for each company was obtained through CAPM. The assumptions of CAPM may not have existed for the period
of study, as explained above.
137

 Centre for Promoting Ideas, USA

www.ijbssnet.com

The market (NSE) may not have been efficient as such and hence the use of CAPM may not have been
appropriate. The assumption that the rate of return was constant for the six months of the period of testing the
model may also have affected the results. The rate of return might have been volatile since, even the government
of Kenya Treasury Bonds has been unstable. In July 1999 the Bonds rated at 14.5% (July 1999 issue).The
government of Kenya Treasury bonds rates has been falling as from December 2000. Although CAPM
assumption do not hold in the real world, CAPM still serves as a useful framework for evaluating financial
decisions. To reflect the real world, the assumptions may be relaxed by using extended versions of CAPM
(Sharpe et al 1999).Information differentials may have contributed to the results obtained in the study. The
presence of noise may cause markets to be inefficient, but prevents an investor from taking advantage from
inefficiencies. Noise makes it difficult to test either practical or academic theories about how the market works.
The estimated and /or the actual prices obtained above may be made up of both noise and information. This
may have led to imperfect observations and hence the knowledge of expectations on the stocks was
limited.Brennan (1973) noted that the possibility of inaccurate data should be obvious in any valuation model.
The estimates of the beta coefficients, expected market return among others may be debatable as preserved by
Sayawa (2000).
The study assumed that prices are determined by the expected dividend per share. However, since the results are
contrary, it therefore implies that the prices of shares do not only depend on dividends. This supports the widely
accepted view within the academic community that it is not the firms dividend policy that determines the value
of the shares, but also other critical variables like earnings power of the company. Most managers prefer that the
dividend discount model be incorporated into a broader framework of multiple valuation models. The basic idea
behind this approach is that different valuation models contain information about security mispricing, some of
these valuation models are based on market anomalities, such as over-reaction to the expected news about the
company. Due to the limitations of individual models, a combination of the models forecasts can produce
estimates of mispricing superior to any single model, an example of such is Franklin Portfolio Associates (FPA)
Model, used in Boston.
TABLE 4 THE REQUIRED RATES OF RETURN COMPUTED FOR EACH COMPANY
COMPANY
BROOKE BOND
GEORGE WILLIAMSON
KAKUZI
SASINI TEA AND COFFEE
DIAMOND TRUST
NATION MEDIA GROUP
STANDARD NEWSPAPER GROUP
BARCLAYS BANK
C.F.C BANK
B.A.T
BAMBURI PORTLAND
EAST AFRICAN BREWERIES
K.P.L.C
TOTAL KENYA
STANDARD CHARTERED BANK
K.C.B
CAR AND GENERAL
I.C.D.C

CAPM RATE %
18.74
21.26
20.44
21.23
20.96
20.47
22.66
21.16
17.65
20.83
20.65
21.13
20.44
20.9
21.12
20.47
26.41
21.24

TABLE 5 PREDICTED PRICES USING THE DIVIDEND DISCOUNT MODEL

MONTHS
COMPANY
BROOKE BOND

138

Actual
Predicted
Difference

Jan-00
104
90.69
13.31

Feb-00
104
80.2
23.8

Mar-00
88
70.87
17.13

Apr-00
78
63.34
14.66

May-00
76
56.53
19.47

Jun-00
74
49.54
24.46

International Journal of Business and Social Science

GEORGE WILLIAMSON

KAKUZI

SASINI

DIAMOND TRUST

NATION MEDIA

STANDARD NEWSPAPER

BARCLAYS

C.F.C BANK

B.A.T

BAMBURI

E.A.B.L

Actual
Predicted
Difference
Actual
Predicted
Difference
Actual
Predicted
Difference
Actual
Predicted
Difference
Actual
Predicted
Difference
Actual
Predicted
Difference
Actual
Predicted
Difference
Actual
Predicted
Difference
Actual
Predicted
Difference
Actual
Predicted
Difference
Actual
Predicted
Difference

COMPANY

K.P.L.C

TOTAL KENYA

STANDARD CHARTERED

K.C.B

CAR AND GENERAL

I.C.D.C

Actual
Predicted
Difference
Actual
Predicted
Difference
Actual
Predicted
Difference
Actual
Predicted
Difference
Actual
Predicted
Difference
Actual
Predicted
Difference

93
77.7
15.3
97.5
73.56
23.94
45
36.89
8.11
25
21.86
3.14
93
85.71
7.29
10.75
8.55
2.2
101
92.89
8.11
14.05
12.66
1.39
73
77.09
-4.09
26.25
23.53
2.72
66.5
64.35
2.15
Jan-00
93.5
87.02
6.48
49
42.54
6.46
57
55.42
1.58
35
26.54
8.46
10
8.34
1.66
50
43.34
6.66

87
64.46
22.54
77.5
62.21
15.29
36
29.98
6.02
28
18.36
9.64
90.5
73.06
17.44
10.5
7.29
3.21
115
84.01
30.99
14
11.48
2.52
94
75.62
18.38
26
21.09
4.91
70
59.34
10.66
Feb-00
91.5
89.26
2.24
65
37.71
27.29
75.5
48.21
27.29
31.5
24.49
7.01
10
6.65
3.35
45
37.47
7.53

Vol. 2 No. 6; April 2011

87
53.93
33.07
70
52.91
17.09
35.5
24.54
10.96
26.75
15.62
11.13
87.5
63.25
24.25
10.05
6.3
3.75
90
77.18
12.82
15.15
10.5
4.65
64
75.22
-11.22
27.5
19.05
8.45
70
55.53
14.47
Mar-00
88
67.11
20.89
48.75
33.95
14.8
52.5
52.96
-0.46
25
20.61
4.39
10
5.48
4.42
40.5
32.81
7.69

90
44.98
45.2
67
45.09
21.91
31.75
19.98
11.77
24.75
13.34
11.41
75
54.06
20.94
8.05
5.38
2.67
90
71.35
18.65
16
9.5
6.5
62
75.14
-13.14
26.5
17.25
9.25
69
52.25
16.75
Apr-00
78
57.61
20.39
49
30.72
18.28
47.75
52.96
-5.21
26.5
17.31
9.19
10
4.39
5.61
46.75
28.8
17.95

77
37.96
39.04
67
38.72
28.8
36.5
16.43
20.07
21.25
11.56
9.69
74
47.36
26.64
8.75
4.68
4.07
87
66.97
20.03
13.65
8.62
5.03
61
74.76
-13.76
28.5
15.76
12.74
66.5
49.82
16.68
May-00
50
43.42
6.58
49
28.46
20.54
47
52.67
-5.67
27.5
14.59
12.91
10.25
3.68
6.57
47
25.67
21.33

75
33.15
41.85
66.5
33.77
32.73
35
13.97
21.07
20
10.32
9.68
75
44.04
30.96
6.1
4.35
1.75
86
64.38
21.62
9.8
7.15
2.65
57
74.97
-17.97
29.25
14.65
14.6
65.5
48.49
17.01
Jun-00
51.5
43.19
8.31
51
26.51
24.43
48
52.97
-4.97
28.5
12.45
16.05
10.05
3.59
6.46
49.5
23.6
24.24

139

 Centre for Promoting Ideas, USA

TABLE 6

www.ijbssnet.com

AN ANALYSIS OF THE DIFFERENCES BETWEEN THE ACTUAL AND PREDICTED PRICES USING THE
DIVIDEND DISCOUNT MODEL

COMPANY
BROOKE BOND
G.WILLIAMSON
KAKUZI
SASINI
DIAMOND TRUST
NATION MEDIA
STANDARD
N.PAPER
BARCLAYS BANK
C.F.C
B.A..T
BAMBURI
E.A.B.L
K.P.L.C
TOTAL KENYA
STANDARD BANK
K.C.B
CAR AND GENERAL
I.C.D.C

MEAN
18.805
32.833
23.293
13
9.115
21.253

VARIANCE
21.511
137.237
44.848
38.683
9.191
68.366

STD DEV.
4.638
11.715
6.697
6.22
3.032
8.268

T.COMP.
(CONF.)
4.867
12.294
7.028
6.527
3.181
8.677

NULL HYPOTHESIS
Reject Ho
Reject Ho
Reject Ho
Reject Ho
Reject Ho
Reject Ho

2.942
18.703
3.79
-6.967
8.695
12.953
10.815
18.633
2.093
9.668
4.678
14.51

0.808
61.613
3.659
174.831
19.566
33.806
61.944
55.037
161.031
17.57
3.712
68.987

0.899
7.849
1.913
13.222
4.423
5.814
7.87
7.419
12.69
4.192
1.927
8.306

0.944
8.237
2.007
13.876
4.642
6.102
8.26
7.785
13.317
4.399
2.022
8.716

Do not Reject Ho
Reject Ho
Do not Reject Ho
Reject Ho
Reject Ho
Reject Ho
Reject Ho
Reject Ho
Reject Ho
Reject Ho
Do not Reject Ho
Reject Ho

Lev.of Sig.
Degrees of
fred.
t critical

0.05
5
2.571

4.0 Conclusion
The main objective of the study was to establish the reliability of the dividend discount model on the valuation of
common stocks at the Nairobi Stock Exchange. In order to achieve this, share prices were predicted using the
dividend discount model and then compared with the actual prices. The differences between the two were
obtained .T-tests were carried out on the differences to establish whether the two prices were significantly
different from each other. Of the eighteen companies studied, only three companies showed that the differences
were significant. We can therefore; conclude that the dividend discount model cannot be relied on by companies
in the valuation of their common stocks at the NSE. The results are attributed to among other factors, the
inefficient market (NSE), inappropriate discounting factors, information differentials and measurement and
evaluation problems.
4.1 Further Research
The CAPM model assumptions are not practical in the real world situation; however it is possible to extend the
model by relaxing the assumptions without drastically changing it. For instance the study assumed a risk-free rate;
a better result could be obtained without the risk-free rate using the zero beta portfolios. This implies that the
Security Market Line (SML) will be more flat than the original version (with the risk-free rate).Many
organizations that estimate the SML generally find that it conforms to the zero betas CAPM than the original
CAPM (Sharpe et al 1999).It would therefore be interesting and more practical for one to conduct a study based
on the same and many other extensions of CAPM.A further study may also be conducted using a different model
of equilibrium rather than the market model .More dynamic models like Arbitrage Pricing Model may produce a
result with better significance. Since a firms share price is not only influenced by its dividends as indicated
elsewhere in the report, use of multiple models may result to more robust analysis than a single model like the
dividend discount model. The multiple models have produced accurate prices of stocks in Boston; they have been
successfully used by FPA as indicated in chapter four. Further studies may be conducted using the models in
various markets.
140

International Journal of Business and Social Science

Vol. 2 No. 6; April 2011

References
Aharony, J and Swary,I.,Quarterly Dividend and Earnings Announcements and Stockholders Returns: An Empirical Analysis
Journal of Finance, March 1980 pp.1-12.
Asquith, P. and Mullins, D., The Impact of Initiating Dividend Payments on Shareholders Wealth Journal of Business, January
1983 pp.77-96.
Barclay,M. and Smith,C., Corporate Payout Policy: Cash Dividends Versus Open-Market Repurchases Journal of Financial
Economics, October 1988 pp.61-82.
Black, F., Capital Market Equilibrium with Restricted Borrowing Journal of Business, July 1972, pp. 444 455.
., The Dividend Puzzle Journal of Portfolio Management,Winter 1976 pp. 5-8.
Black, F and Scholes,M.,The Effects of Dividend Yield and Dividend Policy on Common Stock Prices and Returns.Journal of
Financial Economics, May 1974 pp. 1-22.
Brennan,M.J., Taxes, Market Valuation and Corporation Financial Policy National Tax Journal, December 1970 pp.417-427.
..,An Approach to The Valuation of Uncertain Streams Journal of Finance June 1973 pp.61 74.
Brickley, J Shareholder Wealth, Information Signaling and the Specially Designated Dividend : An. Empirical study Journal of
Financial Economics, June 1983 pp.187-210
Brigham,E.F. and Gapenski L.C. ,Intermediate Financial management 5th Edition, Dryden Press, 1996.
Chen,N.,Roll,R.,&Ross,A.S, Economic Forces and the Stock Market Journal of Business July 1986 pp.383-403.
Craig, D. Johnson G. and Joy M. Accounting Methods and P/E Ratios Financial Analysts Journal, March April 1987 pp. 41-45.
Elton, B.J. & Grubber, M.J., Marginal Stockholder Tax Rates and the Clientele Effect Review of Economics and Statistics June
1970 pp.68-74.
., Modern Portfolio Theory and Investment Analysis 5th edition John Wiley & Sons,1995.
Fama, E.F., Efficient Capital Markets: A Review of Theory and Empirical Work Journal Of Finance May 1970 pp. 383 417.
., . Efficient Capital Markets: II Journal of Finance December 1991 pp. 1575-1617.
, Foundations of Finance, Basic Books Inc. 1976.
Fuller, R.J. and Chi-Cheng H., A Simplified Common Stock Valuation Model Financial Analysts Journal September October
1984. pp. 49-56.
Gordon, M Optimal Investment and Financing Policy Journal of Finance May 1963 pp. 264 272.
Harmbuger,M.J.& Kochin L.A., Money and Stock Prices: The Channels of Influence. Journal of Finance May 1972 pp.231-249.
Healy.P.M, and K.G, Palepu Earnings Information Conveyed by Dividend Initiations and Omissions Journal of Financial
Economics September 1988 pp.149-175.
Homa,K.E & Jaffee,D.M.,The Supply of Money and Common Stock Prices Journal of Finance ,December 1971 pp.1045-1066.
Jensen, M & Bennington G.A., Random Walks and Technical Theories: Some Additional Evidence. Journal of Finance May 1970
pp. 469 481.
Kwan, C.,Efficient Market Tests of the Information Content of Dividend Announcements:Critique and Extensions Journal of
Financial and Quantitative Analysis June 1981 pp.193-206.
Kerandi, A.M., Testing the Predictive Ability of the Dividend Valuation Model on Ordinary Shares Unpublished MBA
Project,University of Nairobi, 1993
Lewellen, W., Stanley.K, Lease.R. & Schlarbaum,G., Some Direct Evidence on Dividend Clientele Phenomenon Journal of
Finance December 1978pp.1385-99.
Lintner, J. Optimal Dividends and Corporate Growth Under Uncertainty Quarterly Journal of Economics, Feb. 1964 pp .49-95.
Long,J., The Market Valuation of Cash Dividends: A Case to Consider Journal of Financial Economics June 1978 pp.235-264
Michaud, R.O., &Davis, P.L., Valuation Model Bias and the Scale Structure of Dividend Discount Returns Journal of Finance,
May 1982 pp.563-573.
Miller,M.H& Modigiliani, F.,Dividend Policy,Growth and the Valuation of Shares Journal of Business October 1961 pp.411-33.
Pettit,R.R., Dividend Announcements, Security Performance, and Capital Market Efficiency Journal of Finance December 1972
pp.993-100
Reilly,F.K&Brown,K.C, Investment Analysis and Portfolio Management 6th Edition 2000 Dryden Press Philadelphia.
Sawaya,A.N,Beta Coefficients as a Measure of Risk of the Common Shares Listed at the Nairobi Stock Exchange., Unpublished
MBA Project, University of Nairobi,2000.
Sharpe,W.F,Gordon,J.A&Jeffery,V.B ,Investments 6th Edition 1999 Prentice Hall
Sorensen, E.H & Williamson P.A. ,Some Evidence on The Values of Dividends Discount Models Financial Analysts
JournalNov.-Dec.1985 pp. 60-69
Van Horne ,J.C., Financial Management and Policy 11th Edition. Prentice Hall 2001.
Watts,R.,The Information Content of Dividends Journal of Business, April 1973pp.191-211.
Weston, J.F., &Copeland, T.E, Managerial Finance 9th Edition. Dryden Press 1992.
Wilkes, F.M.,Dividend Policy and Investment Appraisal in Imperfect Markets
Journal of Business Finance and Accounting, Summer 1977 pp.187-199.

141