anshul jindal-0512-ex ante cost of equity estimates of nifty stocks

8/7/2019 Anshul Jindal-0512-Ex Ante Cost of Equity Estimates of Nifty Stocks

1/49

Ex Ante Cost of Equity Estimates of S&P 50 Firms: The Choice Between Global and Domestic CAPM------------------------------------------------------------------------------------------------------------------------------------------------

____________________________________________________________________________________________ M P BIRLA INSTITUTE OF MANAGEMENT, BANGALORE

1

MPBIM / DISSERTATION / FINANCE / 2005-2007

Ex Ante Cost of Equity Estimates of Nifty Stocks:

The choice between Global and Domestic CAPM

By

ANSHUL JINDAL

05XQCM6012

This dissertation report under the guidance of Prof Nagesh S. Malavalli,

MPBIM submitted in the partial fulfillment of the requirement for MBA

degree of Bangalore University

M P BIRLA INSTITUTE OF MANAGEMENT

(Associate Bharatiya Vidya Bhavan)

Bangalore-560001

MAY 2007


2/49



2

DECLARATION

I hereby declare that this dissertation titled EX ANTE COST OF EQUITY

ESTIMATES OF NIFTY STOCKS: THE CHOICE BETWEEN GLOBAL AND

DOMESTIC CAPM is a record of independent work carried out by me, towards the

partial fulfillment of requirements for the M.B.A. degree course of Bangalore University

at M.P. Birla Institute of Management. This has not been submitted in part or full towards

the award of any other degree or diploma.

PLACE: Bangalore Anshul Jindal

DATE:


3/49



3

PRINCIPALS CERTIFICATE

This is to certify that this research report titled EX ANTE COST OF EQUITY


DOMESTIC CAPM has been prepared by Mr. Anshul Jindal bearing the registration no.

05XQCM6012 under the guidance and supervision of Dr. N. S. Malavalli, M.P. Birla

Institute of Management.

PLACE: Bangalore Dr. N. S. Malavalli

DATE: (Principal)


4/49



4

GUIDES CERTIFICATE

This is to certify that this research report entitled EX ANTE COST OF EQUITY


DOMESTIC CAPM, done by Mr. Anshul Jindal bearing the registration no.

05XQCM6012 is a bonafied work done carried under my guidance and supervision

during the academic year 2005-2007 in the partial fulfillment of the requirement for the

award of MBA degree by Bangalore University. To the best of my knowledge this report

has not formed the basis for the award of any other degree.

PLACE: Bangalore Dr. N. S. Malavalli

DATE: (Professor)


5/49


6/49


7/49



7

RESEARCH ABSTRACT

Cost of Equity estimation is an important area of research in financial markets and lot of

effort has been extended towards finding out new and advanced models for calculating

cost of equity and checking the relevance of models in the practical world.

In this direction, this paper attempts to estimate the ex ante cost of equity for a sample of

S&P CNX Nifty stocks over the period 2004-2006.

This research studied the choice between global and domestic CAPM by examining

which of the two models provides the better fit with a sample of ex ante expected equity

return estimates for large Indian firms. In contrast to many prior studies that use realized

returns, we estimate implied expected returns based on the theorys call for a forward

looking measure. This study period covers 2004 to 2006.

The ex ante estimates show a better overall fit with the domestic version of the single

factor CAPM than with the global version, but the difference is small. The finding has no

trend in time. The findings suggest that for estimating the cost of equity, the choicebetween domestic and global CAPM may not be a material issue.

The studys practical implications are based on the widespread use of the CAPM in cost

of capital estimation by the investors, where the traditional use of the S&P 500 index as

the market portfolio continues to be the standard.

The paper is organized as follows. Section 1 provides the introduction and theoretical

background which is followed by reviews related literature. This review includes the

domestic and global versions of the CAPM and why the two models are theoretically

likely to result in different expected rates of return for a given asset. Section 2 discusses

the methodology and data for the empirical analysis. Section 3 reports the results of the


8/49



8

empirical comparison of the ex ante expected return estimates with the estimates of the

two CAPM versions and with corresponding measures of risk. Section 4 provides

summary and conclusion.


9/49



9

CHAPTER 1

INTRODUCTION AND

THEORETICAL BACKGROUND


10/49



10

INTRODUCTION AND THEORETICAL BACKGROUND

The estimation of a firms cost of equity capital remains one of the most critical andchallenging issues faced by financial managers, analysts, and academicians. Although

theory provides several broad approaches, recent survey evidence reports that among

large firms and investors, the capital asset pricing model (CAPM) is by far the most

widely used model.

Among the variety of decisions to be made in implementing the CAPM is the choice

between a domestic or global index for the market portfolio. Although theory suggests

that using a domestic market index is appropriate only for an asset traded in a closed,

national market, empirical research has thus far failed to establish whether a global or

domestic pricing model performs better with Indian stocks.

Capital (money) used to fund a business should earn returns for the capital owner who

risked their saved money. For an investment to be worthwhile the estimated return on

capital must be greater than the cost of capital. Otherwise stated, the risk-adjusted return

on capital (incorporating not just the projected returns, but the probabilities of thoseprojections) must be higher than the cost of capital.

COST OF CAPITAL:

The cost of capital for a firm is a weighted sum of the cost of equity and the cost of debt.

Firms finance their operations by three mechanisms: issuing stock (equity), issuing debt

(borrowing from a bank is equivalent for this purpose) (those two are external financing),

and reinvesting prior earnings (internal financing).

The cost of debt is relatively simple to calculate, as it is composed of the interest paid

(interest rate), including the cost of risk (the risk of default on the debt). In practice, the

interest paid by the company will include the risk-free rate plus a risk component, which


11/49



11

itself incorporates a probable rate of default (and amount of recovery given default). For

companies with similar risk or credit ratings, the interest rate is largely exogenous.

Cost of equity is more challenging to calculate as equity does not pay a set return to itsinvestors. Similar to the cost of debt, the cost of equity is broadly defined as the risk-

weighted projected return required by investors, where the return is largely unknown. The

cost of equity is therefore inferred by comparing the investment to other investments with

similar risk profiles to determine the "market" cost of equity.

The cost of equity is also known as the discount rate, the rate at which projected earnings

will be discounted to give a present value.

Cost of debt

The cost of debt is computed by taking the rate on a non-defaulting bond whose duration

matches the term structure of the corporate debt, then adding a default premium. This

default premium will rise as the amount of debt increases (since the risk rises as the

amount of debt rises). Since in most cases debt expenses is a deductible expense, the cost

of debt is computed as an after tax cost to make it comparable with the cost of equity

(earnings are after-tax as well). Thus, for profitable firms, debt is discounted by the tax

rate. This is used for large corporations only.

Cost of equity

In finance, the cost of equity is the minimum rate of return a firm must offer shareholders

to compensate for waiting for their returns, and for bearing some risk.

The cost of equity capital for a particular company is the rate of return on investmentthat is required by the company's ordinary shareholders. The return consists both of

dividend and capital gains, e.g. increases in the share price. The returns are expected

future returns, not historical returns, and so the returns on equity can be expressed as the


12/49



12

anticipated dividends on the shares every year in perpetuity. The cost of equity is then the

cost of capital which will equate the current market price of the share with the discounted

value of all future dividends in perpetuity.

The cost of equity reflects the opportunity cost of investment for individual shareholders.

It will vary from company to company because of the differences in the business risk and

financial or gearing risk of different companies.

The cost of equity is calculated by the following formula:

Ke=D1/Po + g

Po=D1/ (Ke-g)

Where D1=Do (1+g)

The formula above calculates the cost of equity based on a firm's current rate of return. If one assumes a perfect market, industry-specific costs of equity reflect the riskiness of

particular industries. A high cost of equity would then indicate a higher-risk industry that

should command a higher return to compensate for the higher risk.

The various models which are used for calculating the cost of equity are as follows:

Capital asset pricing model:

The Capital Asset Pricing Model (CAPM) is used in finance to determine a theoretically

appropriate required rate of return (and thus the price if expected cash flows can be

estimated) of an asset, if that asset is to be added to an already well-diversified portfolio,


13/49



13

given that asset's non-diversifiable risk. The CAPM formula takes into account the asset's

sensitivity to non-diversifiable risk (also known as systematic risk or market risk), in a

number often referred to as beta ( ) in the financial industry, as well as the expected

return of the market and the expected return of a theoretical risk-free asset.

The model was introduced by Jack Treynor, William Sharpe, John Lintner and Jan

Mossin independently, building on the earlier work of Harry Markowitz on

diversification and modern portfolio theory. Sharpe received the Nobel Memorial Prize in

Economics (jointly with Harry Markowitz and Merton Miller) for this contribution to the

field of financial economics.

The formula:

The CAPM is a model for pricing an individual security (asset) or a portfolio. For

individual security perspective, we made use of the security market line (SML) and its

relation to expected return and systematic risk (beta) to show how the market must price

individual securities in relation to their security risk class. The SML enables us to

calculate the reward-to-risk ratio for any security in relation to that of the overall market.

Therefore, when the expected rate of return for any security is deflated by its beta

coefficient, the reward-to-risk ratio for any individual security in the market is equal to

the market reward-to-risk ratio, thus:

Individual securitys = Markets securities (portfolio)

Reward-to-risk ratio Reward-to-risk ratio


14/49



14

The market reward-to-risk ratio is effectively the market risk premium and by rearranging

the above equation and solving for E (Ri), we obtain the Capital Asset Pricing Model

Where:

is the expected return on the capital asset

is the risk-free rate of interest

(the beta coefficient) the sensitivity of the asset returns to market returns,

is the expected return of the market

is sometimes known as the market premium or risk premium (thedifference between the expected market rate of return and the risk-free rate of return).

Asset pricing:

Once the expected return, E(Ri), is calculated using CAPM, the future cash flows of the

asset can be discounted to their present value using this rate (E(Ri)), to establish the

correct price for the asset.

In theory, therefore, an asset is correctly priced when its observed price is the same as its

value calculated using the CAPM derived discount rate. If the observed price is higher

than the valuation, then the asset is overvalued (and undervalued when the observed price

is below the CAPM valuation).

Alternatively, one can "solve for the discount rate" for the observed price given a

particular valuation model and compare that discount rate with the CAPM rate. If the

discount rate in the model is lower than the CAPM rate then the asset is overvalued (and

undervalued for a too high discount rate).


15/49



15

Asset-specific required return:

The CAPM returns the asset-appropriate required return or discount rate - i.e. the rate at

which future cash flows produced by the asset should be discounted given that asset'srelative riskiness. Betas exceeding one signify more than average "riskiness"; betas below

one indicate lower than average. Thus a more risky stock will have a higher beta and will

be discounted at a higher rate; less sensitive stocks will have lower betas and be

discounted at a lower rate. The CAPM is consistent with intuition - investors (should)

require a higher return for holding a more risky asset.

Since beta reflects asset-specific sensitivity to non-diversifiable, i.e. market risk, the

market as a whole, by definition, has a beta of one. Stock market indices are frequently

used as local proxies for the market - and in that case (by definition) have a beta of one.

An investor in a large, diversified portfolio (such as a mutual fund) therefore expects

performance in line with the market.

The market portfolio:

An investor might choose to invest a proportion of his or her wealth in a portfolio of risky

assets with the remainder in cash - earning interest at the risk free rate (or indeed may

borrow money to fund his or her purchase of risky assets in which case there is negative

cash weighting). Here, the ratio of risky assets to risk free asset does not determine

overall return - this relationship is clearly linear. It is thus possible to achieve a particular

return in one of two ways:

1. By investing all of one's wealth in a risky portfolio,2. Or by investing a proportion in a risky portfolio and the remainder in cash (either

borrowed or invested).


16/49



16

For a given level of return, however, only one of these portfolios will be optimal (in the

sense of lowest risk). Since the risk free asset is, by definition, uncorrelated with any

other asset, option 2 will generally have the lower variance and hence be the more

efficient of the two.

This relationship also holds for portfolios along the efficient frontier: a higher return

portfolio plus cash is more efficient than a lower return portfolio alone for that lower

level of return. For a given risk free rate, there is only one optimal portfolio which can be

combined with cash to achieve the lowest level of risk for any possible return. This is the

market portfolio.

Assumptions of CAPM:

All investors have rational expectations.

There are no arbitrage opportunities.

Returns are distributed normally.

Fixed quantity of assets.

Perfectly efficient capital markets.

Investors are solely concerned with level and uncertainty of future wealth

Separation of financial and production sectors.

Thus, production plans are fixed.

Risk-free rates exist with limitless borrowing capacity and universal access.

The Risk-free borrowing and lending rates are equal.

No inflation and no change in the level of interest rate exists.

Perfect information, hence all investors have the same expectations about security

returns for any given time period.


17/49



17

Shortcomings of CAPM:

The model assumes that asset returns are (jointly) normally distributed random variables.

It is however frequently observed that returns in equity and other markets are notnormally distributed. As a result, large swings (3 to 6 standard deviations from the mean)

occur in the market more frequently than the normal distribution assumption would

expect.

The model assumes that the variance of returns is an adequate measurement of risk. This

might be justified under the assumption of normally distributed returns, but for general

return distributions other risk measures (like coherent risk measures) will likely reflect

the investors' preferences more adequately.

The model does not appear to adequately explain the variation in stock returns. Empirical

studies show that low beta stocks may offer higher returns than the model would predict.

Some data to this effect was presented as early as a 1969 conference in Buffalo, New

York in a paper by Fischer Black, Michael Jensen, and Myron Scholes. Either that fact is

itself rational (which saves the efficient markets hypothesis but makes CAPM wrong), or

it is irrational (which saves CAPM, but makes EMH wrong indeed, this possibility

makes volatility arbitrage a strategy for reliably beating the market).

The model assumes those given a certain expected return investors will prefer lower risk

(lower variance) to higher risk and conversely given a certain level of risk will prefer

higher returns to lower ones. It does not allow for investors who will accept lower returns

for higher risk. Casino gamblers clearly pay for risk, and it is possible that some stock

traders will pay for risk as well.

The model assumes that all investors have access to the same information and agree

about the risk and expected return of all assets. (Homogeneous expectations assumption)


18/49



18

The model assumes that there are no taxes or transaction costs, although this assumption

may be relaxed with more complicated versions of the model.

The market portfolio consists of all assets in all markets, where each asset is weighted byits market capitalization. This assumes no preference between markets and assets for

individual investors, and that investors choose assets solely as a function of their risk-

return profile. It also assumes that all assets are infinitely divisible as to the amount

which may be held or transacted.

Arbitrage pricing theory:

Arbitrage pricing theory (APT), in Finance, is a general theory of asset pricing, that has

become influential in the pricing of shares.

APT holds that the expected return of a financial asset can be modeled as a linear

function of various macro-economic factors or theoretical market indices, where

sensitivity to changes in each factor is represented by a factor specific beta coefficient.

The model derived rate of return will then be used to price the asset correctly - the asset

price should equal the expected end of period price discounted at the rate implied by

model. If the price diverges, arbitrage should bring it back into line.

The theory was initiated by the economist Stephen Ross in 1976.

The APT model:

If APT holds, then a risky asset can be described as satisfying the following relation:


19/49



19

Where,

E(rj) is the risky asset's expected return,

RPk is the risk premium of the factor,

rf is the risk-free rate, Fk is the macroeconomic factor,

bjk is the sensitivity of the asset to factor k, also called factor loading,

and j is the risky asset's idiosyncratic random shock with mean zero.

Relationship with the capital asset pricing model :

The APT along with the capital asset pricing model (CAPM) is one of two influential

theories on asset pricing. The APT differs from the CAPM in that it is less restrictive in

its assumptions. It allows for an explanatory (as opposed to statistical) model of asset

returns. It assumes that each investor will hold a unique portfolio with its own particular

array of betas, as opposed to the identical "market portfolio". In some ways, the CAPM

can be considered a "special case" of the APT in that the securities market line represents

a single-factor model of the asset price, where Beta is exposure to changes in value of the

Market.

Additionally, the APT can be seen as a "supply side" model, since its beta coefficients

reflect the sensitivity of the underlying asset to economic factors. Thus, factor shocks

would cause structural changes in the asset's expected return, or in the case of stocks, in

the firm's profitability.

On the other side, the capital asset pricing model is considered a "demand side" model.

Its results, although similar to those in the APT, arise from a maximization problem of each investor's utility function, and from the resulting market equilibrium (investors are

considered to be the "consumers" of the assets).


20/49



20

CHAPTER 2

REVIEW OF LITERATURE


21/49



21

REVIEW OF LITERATURE

Review of literature .means examining and analyzing the various literatures available inany field either for references purposes or for further research.

Further research can be done by identifying the areas which have not been studied and in

turn undertaking research to add value to the existing literature.

For the purpose of literature review various sources of information have been used.

Sources include books, journals as well as some literature papers.

Recent survey evidence (Bruner, Eases, Harris, and Higgins, 1998) and Graham and

Harvey, 2001) reports that the capital asset pricing model (CAPM) is widely used by

large US firms and investors. The CAPM also continues to have the wide popularity in

academic textbooks and applied articles (e.g. Kaplan and Peterson, 1998 and Ruback,

2002).

These applications use the traditional domestic CAPM, Ki = Rf + Bid [Kmd Rf]; where

Ki is the equilibrium expected rate of return for asset i; Rf is the risk free rate; Bid is the

beta of asset i against the domestic market portfolio returns; Kmd is the equilibrium

required rate of return on the domestic market portfolio; and Kmd- Rf is the risk premium

on the domestic market portfolio.

A.

Global CAPM and domestic CAPM:

Stehle (1997) and Stulz (1995a, 1995b, 1999) argue that using a domestic market indexis only appropriate for an asset traded in a closed, national financial market. Although

equilibrium international asset pricing models are multifactor in general, if the purchasing

power parity (PPP) condition holds, then the single factor CAPM equation can be adapted


22/49



22

to a international context for asset in the global market portfolio, as discussed in Stulz

(1995c). We emphasize the difference between the domestic and global CAPMs by

equation (1).

Ki = Rf + Bi G [KmG - Rf] (1)

Where, Ki is the equilibrium expected rate of return for asset I in a specific pricing

currency, Rf is the nominal rate of return on an asset that is risk free and denominated in

the pricing Currency, Bi G is the beta of asset is returns against the unhedged global

market index returns, with returns computed in pricing currency, Km G is the equilibrium

required rate of return in the pricing currency on the unhedged global market portfolio,

and Km G - Rf is the risk premium on the unhedged global market portfolio.

Karolyi and Stulz (2003) point out that only in special case in which Bi G equals Bid Bd G

does the global CAPM result in the same expected return as the domestic CAPM, i.e.,

when an assets global beta is equal to its domestic beta times the global beta of the

domestic market portfolio. Generally, this condition does not hold. Instead, when Bi G is

greater than Bid Bd G, the domestic CAPM is likely to underestimate the assets expected

return relative to the global CAPM, because there is more global systematic risk in theassets returns than is accounted for by the domestic market index. Similarly, when Bi G

less than Bid BdG, the domestic CAPM is is likely to overestimate the assets expected

return relative to the global CAPM, because the asset has less global systematic risk in its

returns than is accounted for by the domestic market index.

Stehle (1977) reports empirical support for the global CAPM over the domestic version

in realized returns for the US stocks from 1956 to 1975. Harveys (1991) study provides

further empirical support of the global pricing of US equities. Black (1993) asserts that

the issue of whether a global or domestic index should be used in CAPM applications is

not yet settled. However, given the significant globalization of the world financial

markets, Stulz (1995a, 1995b, 1999) advocates the use of global version. In contrast to


23/49



23

Stehles (1977) findings, Griffin (2002) reports that for the period between 1981 and

1995, a three factor (Fama- French) domestic model had lower pricing errors for firms

than did an analogous Three factor world version. His results indicate that a domestic

pricing model is a better fit with realized return data than a global pricing model.Campbells (1996) empirical analysis of a multifactor domestic pricing model finds that

the single facto domestic CAPM is a good approximate model for stock and bond prices,

since the additional factors (returns to human capital and changes in expected market

return) are highly correlated with the market index returns. Ng (2003) reaches a similar

conclusion in the context of the global CAPM, with the additional factors of FX risk and

shifts in both expected market returns and expected FX changes. Therefore, we only

examine the two Single- factor CAPMs. Griffin (2002) does not report results on

domestic compared to world Single- Factor (Market index) models. Griffin reported that

the domestic version of the single factor model had lower pricing errors than did the

world model.

A domestic index could be the preferred benchmark for investors with a significant

home bias, as in the Cooper and Kaplanis (2000) model of partially integrated world

markets. However, we do not know whether the popularity of the domestic CAPM

among firms is for this reason.

Finally, Robert S. Harris, Dev R Mishra (2003) advocates that Single factor domestic

CAPM has a better fit than the global CAPM but at the end they gave a lead in their

article that after extending our study to smaller companies, we might shed more light on

DCAPM and GCAPM.

B.

Ex Ante Expected Return Estimates:

Empirical tests comparing global to domestic pricing models usually rely on realized

returns. However, Elton (1999) points out that ex ante estimates of expected returns are

more desirable. We obtain ex ante expected return estimates through discounted cash


24/49



24

flow (DCF) models, as in a number of prior studies, including Claus and Thomas

(2001), Fama and French (2002) , and others discussed below.

In contrast to research that uses realized returns, almost all of the studies using ex anteexpected return estimates find an empirical relation between expected return and beta

risk, despite differences in approaches and time periods. Harris and Marston (1992) and

Harris (1993) report a significant relation between ex ante expected return estimates and

(domestic0 betas for a sample of US stocks in the 1982-1987 periods. At the same time

they confirm the findings of previous empirical studies of no significant relation between

realized returns and betas.

The results of Gebhardt , Lee, and Swaminathan (2001) provide the only exception that

we know of to a positive empirical relation between ex ante expected return and beta risk

estimates. Their study, which uses IBES forecasts and a clean surplus residual income

valuation model, reports no significant association between their ex ante expected returns

estimates and domestic betas for a sample of US stocks from the period 1979-1995.

There is some controversy about IBES forecasts. La Porta (1996) asserts that analysts

growth forecasts tend to be too extreme, but Lee, Myers, and Swaminathan (1999) findthat IBES forecasts improve their intrinsic value estimates over forecasts based on a time

series model.


25/49



25

CHAPTER 3

RESEARCH METHODOLOGY


26/49



26

3.1 STATEMENT OF PROBLEM

The estimation of a firms cost of equity capital remains one of the most critical and

challenging issues faced by financial managers, analysts, and academicians. Although

theory provides several broad approaches, recent survey evidence reports that among

large firms and investors, the capital asset pricing model (CAPM) is by far the most

widely used model.

In this research project the question we ask is -

Whether the domestic or the global version of the single factor CAPM provides the

better fit with the dispersion of the ex ante expected return estimates for a sample of Nifty

stocks.


27/49



27

3.2 OBJECTIVES

To test which model provides the better fit with the sample of ex ante expected

equity return estimates for Nifty stocks

To analyze the risk premium differences for DCAPM and GCAPM.


28/49



28

3.3 METHODOLOGY AND DATA

A.

Ex Ante Expected return Estimation:

For each year from 2004 through 2006, we calculate an ex ante expected return estimate

for S&P CNX Nifty stock for which data is available. The firm is eliminated if the

standard deviation around the mean forecast exceeds 20%, or if there are not sufficient

historical returns for the prior 36 months to perform the beta estimations. The dividend

and other firm specific information is obtained by the various sources like Capitaline and

Prowees databases.

The ex ante expected rates of returns are estimated by using the constant dividend growth

model.Ki = D 1i /P0i + Gi

Where Ki is the ex ante expected rate of return (cost of Equity) estimate for the company

I, D 1i is the dividend per share expected to be received at time 1, P 0i is the current price

per share, and Gi the expected growth rate in dividends per share, which are calculated by

taking the EBIT growth rate for the 10 years.

This study is a joint test of the underlying model and the empirical constructs used.

Therefore this cannot be concluded whether rejection is due to failure of the model or of

the empirical proxies.

Finally, using the widespread use of the CAPM, the conflicting empirical results on the

impact of using a domestic or global index warrants additional study using a variety of

approaches. Furthermore, additional empirical results on the constant growth model,

given its longstanding history and continued use, could be useful.

B.

Global CAPM compared to Domestic CAPM:

To use either the global or the domestic CAPM to estimate a firms cost of equity, a time

varying approach is applied to estimate betas and market risk premia. The equity betas


29/49



29

are estimated for a particular year with yearly excess returns (the stock return minus 91

days Treasury bill (T bill) return) for three years prior to the year for which the cost of

equity is estimated. The equity betas for all companies are estimated by using an ordinary

least squares (OLS) of the excess returns on excess market index returns. The monthlystock returns are obtained from January 2001 through December 2006 from

CAPITALINE database. T Bill returns are obtained from the website of Reserve Bank of

India. S&P CNX 500 index is used as the domestic Indian market index. The Dow Jones

Wilshire Global Total Market Index is used as global portfolio or index. The data for the

global index is obtained from the website of Dow Jones: www.djindexes.com/global.

The question in this research report is which of the two CA versions, if we assume that

version is the correct model, has less variation in its fit with the ex ante expected return

estimates for the individual firms. To implement this investigation, we back out the

estimated market risk premia (domestic and global) for each month from the ex ante

expected returns of the individual stocks. To do so, for a given month, we first turn each

stocks ex ante expected return estimate into an ex ante risk premium estimate by

subtracting the yield on the 91 days T-bond. Then we aggregate the stocks ex ante risk

premia estimates with value weighting, producing an ex ante portfolio risk premium

estimate. For the domestic CAPM, we value-weight the firms domestic beta estimatesinto a portfolio domestic beta estimate for the month. Since the portfolio risk premium

should be equal to the portfolio beta times the market risk premium, the domestic market

risk premium estimate for the month is found implicitly by dividing the portfolio risk

premium estimate by the portfolio domestic beta estimate. To ensure a fair comparison

between the domestic CAPM (DCAPM) and the global CAPM (GCAPM), we use an

analogous procedure (each year) to estimate the implicit global market risk premium

from the ex ante portfolio risk premium estimate and the portfolios global beta estimate.

In other words, we estimate the domestic market risk premium by assuming that the

domestic CAPM is valid for the average stock, and estimate the global market risk

premium by assuming that the global CAPM is valid for the average stock By design, this


30/49



30

approach implies that the average difference between the model estimates and the ex ante

estimates is zero for both CAPM versions.

We then investigate how much variation exists for individual firms between the ex ante

risk premium estimates and the corresponding estimates of each of the two CAPMversions. For each year from January 2004 until December 2006, we analyze each stock

as follows. We begin by using the stocks domestic beta and the domestic market risk

premium estimates to find the firms risk premium estimate under the DCAPM, We also

estimate the stocks risk premium under the GCAPM with the stocks global beta and the

global market risk premium estimates. We then compare the ex ante risk premium

estimate for the stock with the risk premium estimates of both CAPM versions.

We use three metrics to assess which of the two CAPM versions has the better overall fit

with the ex ante estimates. First, we examine the average of the absolute differences

between the model estimates and the ex ante estimates. We decide that the model with the

lower overall average of absolute differences across all observations for the individual

firms is the better-fitting model for this metric. Second, we determine the percentage of

the ex ante estimates for which the DCAPM provides a closer fit than the GCAPM. In the

third metric, we compare the results of cross-sectional OLS of ex ante risk premiumestimates for the individual stocks against both the estimated domestic betas and the

estimated global betas. Whichever regression has the higher r-squared indicates the

better-fitting CAPM version with this approach. We also examine the regression results

for relative consistency with the theory: an intercept of zero and a positive slope.


31/49



31

3.4 LIMITATIONS OF THE RESEARCH

1. Sample is restricted to S&P CNX Nifty index.

2. Data considered for six years only.3. The models are tested on the basis of 3 years forecasted expected returns values

only.


32/49



32

CHAPTER 4

DATA ANALYSIS


33/49



33

4.1 RISK PREMIUM DIFFERENCES FOR DCAPM AND GCAPM

Table I summarizes the average absolute differences between the ex ante risk premium

estimates and the DCAPM and GCAPM estimates, and the percentage of instances inwhich the ex ante estimates are closure to DCAPM estimate than to the GCAPM

estimate. For all the observations in the sample, over all years from 2004 through 2006,

the DCAPMs estimated expected return differs in absolute terms from the corresponding

ex ante estimate by an average of 0.093. The GCAPMs estimated expected return differs

in absolute terms from the corresponding ex ante estimate by an average of 0.176.

For every year, the average absolute difference between the DCAPM estimates and the ex

ante estimates is less than or equal to the average absolute difference between the

GCAPM estimates and the ex ante estimates. Based on the average absolute difference

criterion, it is found that the DCAPM has the better overall fit with the ex ante risk

premium estimates.

Table 1.Summary of Risk Premium Differences for DCAPM and GCAPM

Year Ex Ante Bi D RPD Ex- D Bi G RPG Ex- G%DCAPM

Closure

2004 0.310 0.956 0.109 0.201 0.832 0.108 0.202 0.433

2005 0.308 0.923 0.295 0.014 0.628 0.103 0.205 0.733

2006 0.287 0.917 0.222 0.065 1.076 0.166 0.122 0.900

The column s show, respectively the value weighted averages of the estimated ex ante

risk premia (Ex Ante), average domestic beta estimates ( BiD), the average domestic

market risk premium estimates (RP D), the average absolute differences between the ex

ante estimates and those of the DCAPM (Ex- D), the average global beta estimates ( BiG),

the average global market risk premium estimates (RP G), the average absolute differences


34/49



34

between the ex ante estimates and those of the GCAPM (Ex- D), and the percentage of

cases in which the ex ante estimates is closure to the DCAPM estimate than to GCAPM

estimate (% DCAPM Closure).


35/49



35

4.2 CROSS SECTION REGRESSIONS ON SYSTEMATIC RISK

Table 2 reports the results of the cross section regression of the firms ex ante risk

premium estimate on the beta estimates. Overall, the cross section regressions provide

further evidence that consistently throughout the time period 2004 -2006, the ex ante

estimates have the better fit with those of DCAPM than with the GCAPM. Table 2 shows

that the R- squares of all the regressions are higher when we use the domestic beta as the

independent variable than with the global beta. More over, the DCAPM regression results

are consistently better aligned with the theory. The regression intercepts are closer to zero

for DCAPM than for the GCAPM, and their T-statistics on the slope coefficients aremore significant for the DCAPM than for the GCAPM. These observations apply to the

entire period, to all three years covered in the study.

The findings of significant, positive slope coefficients in each of the 3 years cross

section regressions appear to strongly confirm the basic asset pricing theory prediction

that expected returns are positively related to beta risk. However, the positive regression

intercepts suggest the possible omission of risk factor(s) or systematic optimism in the

growth forecasts. Further exploration of this issue is beyond the scope of this study.

Table 2.Cross Section Regressions

Versus Domestic Beta Versus Global Beta

Year Intercept Slope R - Sq Intercept Slope R - Sq

20040.201 0.114 0.053 0.383 0.088 0.038

2005 0.406 0.106 0.065 0.417 0.174 0.054

2006 0.355 0.074 0.088 0.428 0.131 0.061


36/49



36

The table presents the results of cross section regression of ex ante risk premium

estimates and systematic risk estimates for individual firms. This research used ordinary

least squares, with ex ante risk premium estimates as the dependent variable and firmbeta against indicated market portfolio as independent variable.

INTERPRETATION:

Together, Tables 1 and 2 lead us to conclude that using all three metrics (average

absolute differences, percentage of cases with the better fit, and cross- section regression

results), the domestic CAPM fits the dispersion of ex ante risk premium estimates better

than does the global CAPM. This finding is consistent with the Cooper and Kaplanis

(2000) model of partially segmented global capital markets and home bias.


37/49



37

CHAPTER 5

CONCLUSION


38/49



38

CONCLUSION

We compared ex ante expected return estimates, which are implicit in share prices,growth rates, and the dividend growth model, with expected return estimates from the

global CAPM and the domestic CAPM. We use the Dow Jones Wilshire World global

index as the market benchmark for computing betas for the Global CAPM, and S&P

CNX 500 Index for computing betas for the domestic CAPM. Our sample comprises

S&P CNX NIFTY companies over the period 2004 2006.

We find that the domestic CAPM has a better fit with the dispersion of ex ante expected

return estimates, overall and for all samples the previous researches also emphasis on this

fact and in this study we also found that the DCAPM has the less risk premium

differences than compared to GCAPM for all the three years. And only in the case of

year 2004 the %DCAPM Closure value is less than 0.5 which shows that the DCAPM is

very much desirable while calculating the cost of capital.

When we did the cross section regression of the firms ex ante risk premium estimate on

the beta estimates, we found that the ex ante estimates have the better fit with those of DCAPM than with the GCAPM with high R square values in the case of DCAPM

compared to the GCAPM. Regression intercepts are also closure to zero in the case of

DCAPM. This type of results are also found in the previous research works done by

several researchers in this area and most of them strongly emphasizing on the use of

DCAPM for calculating the firms cost of equity instead of GCAPM.

In this research we observe no trend in this fit over time. While the domestic model

provides a better fit of our data, the relatively small empirical difference between the

models suggests that for estimating the cost of equity, the choice between the domestic

and global CAPM may not be a material issues for many large firms.


39/49



39

CHAPTER 6

BIBLIOGRAPHY


40/49



40

6.1 BIBLIOGRAPHY

BOOKS1. Basic Econometrics: By Damodar N. Gujrati

2. Investments: By Bodie, Kane, Marcus, Mohanty

3. Financial Management: By I M Pandey

WEBSITES

1.

www.nseindia.com

2. www.yahoofinance.com

ECONOMETRICS SOFTWARE PACKAGES

1. Eviews

2. SPSS


41/49



41

6.2 REFERENCES

Robert S. Harris; Felicia C. Marston; Dev R. Mishra; Thomas J. OBrien, ExAnte Cost of Equity Estimates of S&P 500 firms: The choice between global and

domestic CAPM, Financial Management , Vol. 32, No. 3. (Autumn, 2003), pp.

51-66.

Black, F., 1993, Estimating Expected Return, Financial Analysts Journal 49,

September October, 36-38.

Campbell, J.Y., 1996, Understanding Risk and Return, Journal of Political

Economy 104, 298- 344.

Griffin, J.M., 2002, Are the Fama French Factors Global or Country Specific?

Review of Financial Studies 15, 783-803.


42/49



42

ANNEXURE

S&P CNX NIFTY STOCK BETAS FOR 2004 TO 2006

Company Year Domestic Beta Global BetaABB 2004 1.241 0.579

2005 0.863 0.2532006 0.836 0.896

ACC 2004 1.195 0.8942005 0.723 0.4892006 0.689 0.534

BAJAJ 2004 0.317 0.7262005 0.836 0.752

2006 0.708 0.608BHEL 2004 1.043 0.928

2005 0.904 0.4032006 0.795 0.929

BPCL 2004 0.722 0.8672005 1.225 0.8962006 1.042 1.741

CIPLA 2004 0.651 0.5462005 0.737 0.5732006 0.857 0.802

DABUR 2004 0.455 0.73

2005 0.874 0.7142006 0.886 0.886

Dr. REDDY 2004 0.734 0.3372005 0.793 0.5962006 0.724 1.026

GLAXO 2004 1.497 0.5932005 0.444 0.4972006 0.44 1.073

GRASIM 2004 0.817 0.8462005 0.825 0.3972006 0.962 1.062

AMBUJA2004 0.624 0.792005 0.72 0.3992006 0.608 0.576

HDFC 2004 1.185 0.4642005 0.787 0.6662006 0.867 1.209

HERO HONDA 2004 1.128 0.739


43/49



43

2005 0.865 1.1492006 0.682 1.332

HINDLEVER 2004 0.996 0.2392005 0.883 0.6522006 0.889 1.453

ICICI BANK 2004 1.661 0.5432005 0.93 0.2822006 0.919 1.018

INFOSYS 2004 1.2498 1.2782005 0.6014 0.5392006 0.6648 0.284

IPCL 2004 1.2183 1.4652005 1.1931 0.7362006 1.0419 0.528

ITC 2004 0.3208 0.5312005 0.6539 0.438

2006 0.673 0.933LT 2004 1.0534 1.129

2005 0.7627 0.9022006 0.7468 1.458

NATIONALUM 2004 0.7168 0.452005 1.2422 0.3122006 1.3818 1.726

ONGC 2004 0.647 0.7642005 0.9263 0.4352006 1.0485 1.333

RANBAXY 2004 0.4493 0.047

2005 0.5517 0.2432006 0.687 1.073

RELIANCE 2004 0.7292 0.9142005 0.739 0.5222006 0.8257 1.166

SAIL 2004 1.4955 1.0682005 1.8489 0.482006 1.8547 1.384

SBI 2004 0.8168 0.7742005 1.047 0.5632006 1.0734 1.384

SUNPHARMA 2004 0.6335 0.7352005 0.5456 0.6632006 0.5065 0.809

TATA POWER 2004 1.1766 1.3442005 1.4406 0.8692006 1.4226 1.39

TATA STEEL 2004 1.224 1.043


44/49



44

2005 1.2529 0.5032006 1.3906 1.286

VSNL 2004 1.0129 1.6942005 1.1836 1.5222006 1.1944 2.075

ZEEL 2004 1.6689 1.9022005 1.2869 1.3822006 1.0907 0.317


45/49



45

S&P CNX 5OO AND DOW JONES WILSHIRE GLOBAL MARKET

INDEX DATA FROM 2001 TO 2006

YEAR S&P 500 INDEX DJW global index200612 3190.63 3772.78200611 3212.33 3665.87200610 3044.23 3558.01200609 2879.70 3440.57200608 2689.50 3390.41200607 2479.00 3282.05200606 2409.00 3265.50200605 2831.38 3444.25

200604 2949.13 3434.54200603 2787.80 3322.00200602 2609.48 3274.19200601 2521.30 3219.04200512 2383.93 3125.61200511 2214.55 3016.07200510 2174.55 2948.61200509 2218.38 2998.41200508 2080.85 2942.79200507 1970.05 2863.13200506 1869.83 2819.15

200505 1762.48 2747.77200504 1751.83 2746.20200503 1810.15 2812.75200502 1798.43 2800.13200501 1742.55 2747.47200412 1723.65 2751.51200411 1578.43 2633.63200410 1508.28 2521.41200409 1429.20 2477.75200408 1346.88 2399.69200407 1285.68 2443.98

200406 1238.95 2476.08200405 1317.38 2404.65200404 1532.23 2491.00200403 1453.43 2476.29200402 1474.65 2487.37200401 1542.75 2449.09200312 1426.75 2337.55


46/49



46

200311 1249.55 2233.49200310 1189.30 2183.04200309 1089.78 2124.24200308 1020.00 2021.05200307 913.05 1999.87

200306 853.25 1983.35200305 755.30 1862.55200304 716.13 1738.42200303 734.40 1650.84200302 754.28 1676.97200301 763.08 1754.89200212 756.90 1773.71200211 717.20 1768.99200210 692.88 1637.14200209 717.75 1709.21200208 719.93 1784.85

200207 746.68 1821.24200206 772.58 1995.36200205 762.45 2101.16200204 782.95 2113.54200203 788.60 2121.42200202 766.53 2015.94200201 718.23 2089.97200112 712.35 2098.02200111 674.33 2047.56200110 599.13 1963.01200109 615.30 1938.30

200108 695.75 2180.99200107 706.35 2216.36200106 743.83 2302.31200105 775.75 2372.48200104 712.60 2225.37200103 850.60 2235.55200102 966.73 2458.19200101 950.95 2509.05


47/49



47

CALCULATED COST OF EQUITIES

COMPANY YEAR Ke D Ke G KeABB 2004 22.386 11.944 9.622

2005 22.263 27.991 7.731

2006 21.886 20.887 14.935

ACC 2004 40.006 11.781 11.1342005 40.358 24.570 9.362

2006 40.033 18.505 11.671

BAJAJ 2004 22.292 8.660 10.3282005 22.201 27.332 11.180

2006 21.801 18.813 12.338

BHEL 2004 25.019 11.241 11.2982005 25.036 28.994 8.768

2006 24.878 20.223 15.233

BPCL 2004 9.041 10.100 11.0052005 8.363 36.839 12.176

2006 6.104 24.225 22.555

CIPLA 2004 46.381 9.847 9.4632005 35.268 24.912 9.943

2006 35.076 21.227 14.088

DABUR 2004 20.512 9.150 10.3472005 22.150 28.260 10.918

2006 16.302 21.697 14.845

Dr. REDDY 2004 62.458 10.142 8.4602005 63.055 26.281 10.102

2006 62.702 19.072 16.108

GLAXO 2004 42.637 12.854 9.6892005 42.700 17.751 9.418

2006 41.182 14.471 16.531

GRASIM 2004 11.859 10.437 10.9042005 11.869 27.063 8.726

2006 11.992 22.929 16.432AMBUJA 2004 57.470 9.751 10.635

2005 40.497 24.496 8.740

2006 40.324 17.193 12.050

HDFC 2004 18.474 11.745 9.0702005 18.650 26.134 10.586


48/49



48

2006 18.002 21.389 17.758HERO HONDA 2004 46.356 11.543 10.390

2005 45.651 28.040 13.925

2006 44.378 18.392 18.867

HINDLEVER 2004 17.237 11.074 7.9892005 18.245 28.480 10.489

2006 17.941 21.746 19.958

ICICI BANK 2004 82.666 13.437 9.4492005 82.160 29.629 7.931

2006 81.701 22.232 16.035

INFOSYS 2004 63.937 11.976 12.9782005 66.780 21.598 9.708

2006 63.077 18.113 9.417

IPCL 2004 10.645 11.864 13.876

2005 12.021 36.059 11.0702006 11.612 24.223 11.617

ITC 2004 47.178 8.673 9.3912005 55.016 22.881 9.010

2006 20.756 18.246 15.269

LT 2004 17.596 11.278 12.2632005 17.508 25.540 12.217

2006 14.020 19.442 20.003

NATIONALUM 2004 23.965 10.081 9.0032005 23.975 37.259 8.139

2006 23.772 29.730 22.419

ONGC 2004 27.380 9.833 10.5102005 30.669 29.539 8.989

2006 28.605 24.330 18.876

RANBAXY 2004 16.307 9.130 7.0682005 16.236 20.383 7.662

2006 15.342 18.473 16.531

RELIANCE 2004 21.736 10.125 11.2302005 21.892 24.961 9.591

2006 21.793 20.720 17.370

SAIL 2004 47.484 12.849 11.9702005 53.107 52.088 9.300

2006 51.201 37.392 19.336

SBI 2004 13.681 10.437 10.5582005 14.173 32.489 9.874


49/49


2006 13.628 24.733 19.336SUNPHARMA 2004 31.131 9.785 10.371

2005 29.762 20.234 10.565

2006 29.801 15.549 14.151

TATA POWER 2004 13.317 11.716 13.2952005 13.528 42.109 11.989

2006 13.358 30.391 19.390

TATA STEEL 2004 33.257 11.884 11.8502005 33.421 37.521 9.459

2006 33.311 29.873 18.452

VSNL 2004 14.534 11.134 14.9752005 14.579 35.827 16.503

2006 13.055 26.694 25.566

ZEEL 2004 24.142 13.465 15.974

2005 23.967 38.352 15.5362006 24.159 25.014 9.714

anshul jindal-0512-ex ante cost of equity estimates of nifty stocks

Documents