See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/228440905

The application of the Capital Asset Pricing Model (CAPM): A South African
perspective

Article in African journal of business management · August 2011

CITATIONS READS

11 422

1 author:

Soon Nel
Stellenbosch University
14 PUBLICATIONS 53 CITATIONS

SEE PROFILE

All content following this page was uploaded by Soon Nel on 10 March 2016.

The user has requested enhancement of the downloaded file.

African Journal of Business Management Vol. 5(13), pp. 5336-5347, 4 July, 2011
Available online at http://www.academicjournals.org/AJBM
DOI: 10.5897/AJBM10.1054
ISSN 1993-8233 ©2011 Academic Journals

Full Length Research Paper

The application of the Capital Asset Pricing Model

(CAPM): A South African perspective
W. S. Nel
Department of Accounting, Stellenbosch University, South Africa. E-mail: snel@sun.ac.za
Tel: 021-808 3430. Fax: 021-886 4176.
Accepted 26 January, 2011

The focus of this paper is the capital asset pricing model (CAPM), with a specific emphasis on two of its
main components, namely the risk-free rate and beta. The CAPM is used extensively in practice to
calculate the cost of equity, which, in turn, is used to calculate the weighted average cost of capital for
equity valuation and investment appraisal purposes. The aim of this paper is to investigate how well
valuation theory regarding the CAPM in particular, as advocated by academia, is aligned with the CAPM
and alternative models that leading financial analysts and corporate financiers apply in practice. No
study has yet compared the methods of choice of investment practitioners when calculating the cost of
equity to that of academia. The research results revealed that, although both academia and investment
practitioners favor the CAPM, they disagree significantly with regard to the components of the CAPM
and the use of alternative models.

Key words: Academia, investment practitioners, capital asset pricing model, discount rate, arbitrage pricing
theory, risk-free rate, beta.

INTRODUCTION

Several researchers have indicated that the discounted asset pricing model (CAPM) and the arbitrage pricing
cash flow (DCF) approach is the most accurate and theory (APT) model (PwC, 2008). Leading financial
flexible equity valuation method (Goedhart et al., 2005; analysts and corporate financiers, who will be referred to
Courteau, Kao et al., 2003; Berkman et al., 2000). as “investment practitioners” in this paper, tend to focus
Central to the application of a DCF approach is the on the CAPM in practice (PwC, 2008). Academically, the
calculation of an appropriate required rate of return or South African Institute of Chartered Accountants (SAICA,
discount rate (Damodaran, 2007), which is typically the 2008) highlights the CAPM as a key approach in the
weighted average cost of capital (WACC). The WACC is calculation of the cost of equity, which, in turn, is used to
calculated by allocating weights to the costs of the calculate the WACC for equity valuation and investment
various components (interest-bearing debt, preference appraisal purposes.
shares and ordinary share equity) in a company’s target Much research has been devoted to determining which
capital structure (PricewaterhouseCoopers, (PwC) 2008). models are superior in calculating the cost of equity
Internationally, the two primary models that are applied (Ammann and Verhofen, 2007; Burton, 1998; Magni,
in practice to calculate the cost of equity are the capital 2005; Magni, 2007; Bello 2008). Although researchers
generally agree that the use of the CAPM is a key appli-
cation area of finance for investment decisions and equity
Abbreviations: APT, Arbitrage pricing theory; CAPM, capital valuations (Magni, 2005; Damodaran, 2007; Fernández,
asset pricing model; DCF, discounted cash flow; R(e), expected 2002), they seem to be torn between the relative merits
return; Rf, risk-free rate; β, beta; COV, covariance; NA, not of the CAPM and the APT model. Those in favor of the
applicable; NI, not included; PwC, PricewaterhouseCoopers;
APT model refer to the stringent assumptions of the
E(Ri), expected return on asset i; E(Rm), expected return on
the market; SAICA, South African Institute of Chartered
CAPM and the fact that it cannot be tested empirically
Accountants; WACC, weighted average cost of capital; σ², (Nawalkha, 2007; Fama and French, 2004; Roll, 1977).
variance. Others argue that the CAPM has undoubtedly made a
 Nel 5337

undoubtedly made a fundamental contribution to the investment practitioners. Although the PwC survey
understanding of asset pricing (Perold, 2009; Fernández, afforded academia an insight into the nature and
2002; Harrington and Korajczyk, 1993). frequency of the application of equity valuation methods
However, the literature review revealed no evidence of in practice, no such research has yet been conducted
a study that compared academia’s preferences with among academia.
regard to the use of specific approaches to the calcula- The contribution of this study is that it facilitates the
tion of the cost of equity, with that of investment practitio- convergence of, firstly, academic thinking regarding the
ners. This study investigates academic consensus use of the CAPM, and, secondly, valuation practices
among chartered accountants regarding the use of the between academia and investment practitioners. To this
CAPM, and its constituents in particular, and whether the end, the research results will present academic consen-
general concern regarding a gap between theory and sus regarding the use of the CAPM and highlight
practice (Triantis, 2005; Ralston 2003; Bernstein 2008) is differences between academia and investment practi-
warranted. Although this exploratory study focuses on a tioners in this regard. Should the results reveal that there
specific target audience within the wider academic are approaches to the calculation of the cost of equity,
community, the research results indicate that the topic and the use of the CAPM in particular, that are used
warrants further investigation, which the author intends frequently in practice which are not advocated by
pursuing with future research. However, for the purpose academia, this may highlight the need for academia to
of this paper, the reference to academia will specifically reconsider their syllabus. Similarly, it could mean that
refer to chartered accountants who are in academia. The there are approaches to the calculation of the cost of
emphasis is on academia’s perception regarding the equity and the use of the CAPM that are advocated by
calculation of the cost of equity with a specific focus on academia that are not applied in practice. In this case it
the CAPM, and how these preferences compare to those may be necessary for investment practitioners to
of investment practitioners in South Africa. The next two reconsider their approach to the CAPM and other
sections set out the objective and the value of the methods that they use in practice.
research, followed by a literature review. Sections five The research also contributes to the preparation of
and six describe the research methodology and the students for the marketplace. If there is a gap between
survey results regarding the CAPM, as preferred by theory and practice, the nature of the gap should be
academia, followed by a gap analysis between theory investigated and resolved in order to better align aca-
and practice in section seven. Final remarks are offered demia with the real world. It is therefore, imperative that
in the last section of this paper. academic consensus regarding the calculation of the cost
of equity, and the application of the CAPM in particular, is
compared to the current application of the CAPM in
Objective of the research practice, to establish whether such a gap exists.

The chief objective of this research is to ascertain

whether there is a gap between what is lectured in aca- LITERATURE REVIEW
demia and what is applied in practice. If such a gap
exists, it may have implications for academia and invest- Modern finance theory revolves around maximizing an
ment practitioners. The paper will focus on the calculation investor’s return at a given level of risk (Mirza, 2005).
of the cost of equity with a specific emphasis on the Based on Fama’s (1970) work regarding the efficiency
CAPM. The research forms part of a wider research pro- market hypothesis and Markowitz and Tobin’s portfolio
ject, which is aimed at establishing the nature and size of theory, the CAPM was developed in the mid 1960s, in
the gap between theory and practice with regard to equity order to express the relationship between an asset’s risk
valuations, that is primary methods of equity valuation, and return (Sharpe, 1964; Lintner, 1965; Mossin, 1966).
secondary methods of equity valuation, discount rates, The underlying principle in the CAPM is that company
etc. or industry-specific events have very little impact on an
asset’s required return. The relevant risk is the market
risk, which refers to the sensitivity of the asset’s returns
Value of the research to the returns of the market as a whole, which is reflected
in beta (Mirza, 2005). The classical Sharpe-Lintner
Although the convergence of academia and investment CAPM equates an asset’s expected return to three
practitioners on mainstream valuation practices is a variables, namely beta (β), the risk-free rate (Rf) and the
common phenomenon in developed markets (Bruner, expected market return (Rm) (Fama and French, 2004),
Conroy, Estrada et al., 2002), no research has yet been so that:
conducted on this topic in South Africa. A PwC valuation
methodology field survey, which was conducted among Ε(Ri ) = Rf + βi[Ε(Rm ) − Rf ] ,
leading financial analysts and corporate financiers in
2008, provides a valuable benchmark for best practice for where
5338 Afr. J. Bus. Manage.

Capital asset pricing

model
3.00
2.50
2.00
1.50
1.00
0.50
-

Other Arbitrage pricing theory

Figure 1. Methods used most frequently to calculate the cost of equity in practice in South Africa.
Source: Adapted from PwC (2008)

E(Ri) = expected return on asset i, Rf = risk-free rate, Bi = The APT model equated the expected return of an asset
beta of asset i, E(Rm) = expected return on the market. to various macro-economic factors, each with a specific
Mathematically, beta is the covariance of asset returns beta factor, so that
and market returns divided by the variance of market
returns, so that Ε(Re ) = Rf + β 1Ρ1 + β 1Ρ1 + ... ,

βi =
cov(Ri , Rm ) where, E (Re) = expected return on equity capital, Rf =
σ ² (Rm ) risk-free rate, βP = risk premium reflecting sensitivity to
changes in a specific risk (PwC, 2008).
where; Bi = beta (systematic risk) of asset i, cov (Ri,Rm)
= the covariance between asset i and the market, σ²(Rm)
= the variance of the market returns (Fama and French, Methods that are used to estimate the cost of equity
2004). in practice in South Africa
Internationally, academia and investment practitioners
have been debating the merits of the CAPM since its A valuation methodology survey conducted by PwC in
inception in the mid-1960s. The CAPM is prone to 2008 among 25 leading financial analysts and corporate
criticism since it operates in a ceteris paribus financiers, confirmed that investment practitioners have a
environment and is based on various assumptions, preference for the CAPM when calculating the cost of
including the existence of a risk-free asset which enables equity. The 2008 PwC survey used a frequency table
investors to borrow unlimited amounts at a constant rate. between 0-3, where 0 indicates that the method is
Similarly, the beta factor, particularly its appropriateness seldom or never used, 1 indicates that the method is
as a measure of risk, has been the subject of various often used, 2 indicates the method is frequently used and
empirical studies. Research results by Blume (1971), 3 indicates that the method is always used. The most
Baesel (1974) and Roenfeldt (1978), for example, popular methods that are currently used to calculate the
indicated that share betas are unstable over time, while cost of equity in practice, according to the PwC survey,
research by Fama and French (1996), Jegadeesh (1992) are presented in Figure 1.
indicated that beta is not statistically related to returns. Figure 1 indicates that the most popular method to
Since research indicates that the third component of the calculate the cost of equity that is currently used in
CAPM, namely Rm, cannot be tested empirically (Fama practice is the CAPM, which scored a 3.00. Other
and French, 2004; Harrington and Korajczyk, 1993; methods (scored a 0.2) and the APT model (scored a
Mirza, 2005) the focus of this paper is on β and Rf. 0.10), were second and third best alternatives; and as
Ross (1976) introduced an alternative approach to their frequency scores indicate, negligible, according to
asset-pricing theory in the form of a multifactor model the PwC survey. Methods that were mentioned under the
known as arbitrage pricing theory (APT) in the late 1970s. “Other” category were peer composites and the build-up
 Nel 5339

R157 (6.7)
2.00

207 (11) 1.50 R153 (1.7*)

1.00

0.50

204 (10) - R203 (8.7)

R196 (0.2) R201 (6)

206 (5)
Figure 2. Proxies that are used for the Rf in practice in South Africa.
*The numbers in parenthesis indicates the bonds’ term to maturity.
Source: Adapted from PwC (2008)

method. The PwC survey also requested respondents to indicate

whether they use historical or forecast betas. The
majority (96%) of the respondents indicated that they use
Proxies for the risk-free rate that are used in practice historical betas, while only 4% indicated that they use
in South Africa forecast betas. The question is: To what extent do the
preferences of investment practitioners agree with that of
One of the key components of the CAPM is the risk-free academia? Although no recent literature was found that
rate (Rf). Although various proxies can be used for Rf, focused on such a gap, concern regarding the existence
the results of the PwC survey (2008) indicated that the of such a gap is not a new phenomenon.
majority of the respondents favored the R157
government bond. The most popular benchmarks that
were used at the time of the survey for the Rf in practice, Gap analysis
according to the PwC survey, are presented in Figure 2.
From Figure 2 it is evident that the majority of The first formally reported study that investigated a gap
investment practitioners use the R157, which scored a that presumably exists between theory and practice with
2.00. The use of the other bonds, for example, the now regard to investment management, was conducted by the
defunct R153 bond (scored a 0.5) and the R203 bond American Finance Association (Upton, 1949) in 1948.
(scored a 0.4), seemed negligible. The preference for the The American Finance Association held a convention at
R157 bond vis-à-vis the R153 bond is understandable, Cleveland in the United States, where 27 lecturers
given the former’s longer term to maturity. Respondents representing 20 schools, met to discuss methods of
were also asked whether they adjust the Rf for tax teaching Business Finance. The main conclusion
purposes. The majority (96%) indicated that they do not reached by the members present at the convention was
adjust the Rf for tax purposes. that business school education failed to meet the needs
of businesses in practice.
Approximately 15 years later, Wendt (1966) conducted
The use of beta in practice in South Africa a survey at 205 business schools and found that
investment faculties lagged investment practitioners by
Another key component of the CAPM is beta. The more than a decade and concluded that there was a
investment practitioners indicated that they consider a need for academia to catch up.
variety of sources for determining beta. The most popular In a later study Smith and Goudzwaard (1970) found
service providers that are used by investment that the gap between theory and practice does not only
practitioners are Bloomberg (scored a 1.3), the UCT exist, but is also widening. More recently, in a study
Financial Risk Service (scored a 1.1) and McGregor BFA regarding equity valuations, Nel (2010, 2009) concluded
(scored a 1.0). that, although similarities exist, academia and investment
5340 Afr. J. Bus. Manage.

Bloomberg
2.00

1.50 UCT Financial

Other
Risk Service
1.00

0.50

-
Reuters/Factiva McGregor BFA

In-house
MSCI Barra
calculation

Figure 3. Source of information for beta in practice in South Africa

Source: Adapted from PwC (2008)

practitioners differ significantly in their respective Response rate

approaches to equity valuation methods and multiples, in
The questionnaire was sent to 446 chartered accountants who are
particular. However, the literature review revealed no in academia. Of these emails, 36 were returned to the sender as a
research on how well the theory of the CAPM, as result of invalid addresses. A potential target audience of 81
advocated by academia, is aligned with the CAPM as it is lecturers opened the email. This percentage may seem small at first
applied in practice. To the best of the author’s know- glance, but it is important to bear in mind that not all lecturers in
ledge, no gap analysis has yet been conducted on the academia lecture finance. Only 54 lecturers currently lecture, or
calculation of the cost of equity and the specific have in the past lectured, finance, as confirmed by the relevant
divisions at the respective universities. The effective target
application of the CAPM. audience at the universities therefore only consisted of 54 lecturers.
A total of 35 lecturers of the potential 54 respondents completed the
questionnaire. All completed questionnaires were usable, consti-
RESEARCH METHODOLOGY tuting an effective response rate of 65%. Although in terms of
absolute numbers 35 responses may seem small, it should be kept
In order to achieve the objective of the research, a two-pronged in mind that similar research conducted by PwC in 2008 yielded
approach was adopted: only 25 responses (PwC, 2008). The 35 responses originated from
ten universities, which render them representative of the general
(1) Establish academic consensus regarding the methods of choice thinking in academia regarding the topic of equity valuations. The
when calculating the cost of equity and the theory of the CAPM in fact that SAICA supported the research initiative decreases the
particular; likelihood that a higher number of responses would have been
(2) Establish how the methods of choice and the theory of the achieved in any other cost-effective way. The profile of the lecturers
CAPM that were identified in (1) compare with the methods of who participated in terms of qualifications and finance-specific
choice and the CAPM that investment practitioners apply in lecturing, renders the results useful.
practice.
In order to ascertain academic consensus regarding the
calculation of the cost of equity and the use of the CAPM, a survey Profile of participants
was conducted in 2008 at 12 universities in South Africa. The PwC
survey will serve as a reflection of the preferences of investment The participants in the academic survey were suitably qualified
practitioners in the marketplace. lecturers with ample lecturing experience on the topics of equity
valuations and investment appraisals. All the participants were
members of SAICA. In addition, 80% of the participants held
Survey design and distribution masters degrees, of whom 20% also held PhDs.
These participants represented ten of the universities in South
A draft survey was prepared and a link to an electronic database Africa, indicating that the results are a fair reflection of general
was emailed to five lecturers. Their feedback and recommendations academic thinking regarding valuations. One can therefore
were incorporated in the questionnaire and the database was conclude that the participants in the survey constituted a strong
cleared of these pilot responses. academic knowledge base to respond to the questions regarding
The final survey presented 25 questions and took approximately equity valuations.
15 minutes to complete.
SAICA emailed a link to the final, electronic, web-based
questionnaire to chartered accountants who work in academia. An Survey questions
email reminder was sent out and responses were subsequently
followed up telephonically. The academic survey was divided into three sections. The first
 Nel 5341

section dealt with specific equity valuation methods, such as the As Figure 4 illustrates, the respondents had a clear
free cash flow model and multiples. Section two focused on the preference for the CAPM adjusted for additional risk
discount rate, posing questions regarding the most appropriate
factors (78%), and the CAPM (70%). The build-up
method for calculating the cost of equity and specific questions
regarding the CAPM. The third section covered the profile of the method, which garnered 58% support, was the third most
participants. This paper will focus on the second and third sections popular alternative.
of the questionnaire, while the remaining section will form part of Considering whatever investors required (26%)
further research. It is important to bear in mind that the results are garnered little support from respondents. Only 8% of the
merely a reflection of the beliefs and opinions of chartered respondents were of the opinion that the average
accountants in academia.
The emphasis of this paper falls on seven questions in the survey
historical return on equity should be used, while 7%
that were designed to establish how the cost of equity should be regarded backing out of the dividend/earnings model or
calculated, in accordance with academic thinking, with a specific listed entities (“Other”), as suitable alternatives to
emphasis on the CAPM. The questions focused on the following: calculate the required rate of return.
(1) How frequently various methods should be used to calculate the
required rate of return;
(2) How frequently various proxies for the risk-free rate should be
Proxies used for the risk-free rate when employing
used when employing the CAPM; the CAPM
(3) Reasons for the choice in (2);
(4) Whether the risk-free rate should be adjusted for tax purposes; The second question required respondents to indicate
(5) Which tax rate should be used to adjust the risk-free rate; which instruments should be applied most frequently in
(6) How beta should be determined; practice as proxies for the Rf in the CAPM. The proxies
(7) How the CAPM should be approached when an entity has a
negative beta.
for the Rf that, according to academic thinking, should be
Questions 1 and 2 required respondents to indicate how frequently applied most frequently when using the CAPM in practice
the respective methods/risk-free rates should be used on a scale of to value an interest in an entity’s equity are presented in
0 - 4, where 0 indicates that the method is never used, 1 indicates Figure 5.
that the method is almost never used, 2 indicates the method is As is evident from Figure 5, the majority of academia
sometimes used, 3 indicates that the method is almost always favors the R153 bond (now defunct), followed by the
used, and 4 indicates that the method is always used. The survey
results for questions 1 and 2 discussed in this paper reflect those
R157 and the R186 bonds. The preference for the R153
alternatives that respondents indicated should always or almost bond, vis-à-vis the R157 and the R186 bond is a strange
always be used in practice, in other words the percentage of phenomenon, since at the time of the survey, the R153
respondents who answered 3 or 4. Question 4 was a Yes/No type had a very short term to maturity (1.7 years). The close
question. Questions 3, 5, 6 and 7 were alternative/open ended type proximity between the preference for the R186 and R196
questions where respondents could choose between alternative
bonds also seems illogical. Surely, a bond’s term to
responses or specify an alternative that was not listed.
maturity should match that of the investment. The use of
the other bonds is negligible.
RESULTS The third question required respondents to provide
reasons for their selection of proxies for the risk-free rate.
Although the literature review highlighted several The main reason presented for the choice of the specific
shortcomings of the CAPM (Nawalkha, 2007; Fama and bonds was that they were the benchmark government
French, 2004; Roll, 1977), it is used extensively in bonds (41%), followed by the fact that they offered good
practice. According to the PwC survey, investment liquidity (18%) and that they gauged the domestic market
practitioners predominantly use the CAPM, with little well (16%).
regard for the APT model. In order to determine how well Alarmingly, only 6% of the respondents correctly
these preferences are aligned with those of lecturers in indicated that the appropriate bonds should be chosen
academia, it is necessary to ascertain academic con- based on the match between their term to maturity and
sensus regarding the use of the CAPM and alternative the tenure of the particular investment. Few (3%)
models. respondents regarded high trading volumes as an
appropriate reason for choosing specific bonds.
The fourth question required respondents to indicate
Methods used to determine the required rate of return whether they adjusted the Rf for tax purposes and, if so,
the fifth question required them to indicate at what tax
The first question required respondents to indicate which rate.
methods should be applied most frequently in practice to The majority (64%) of respondents were in favor of ad-
determine the required rate of return when valuing an justing the Rf for tax purposes, while 36% was opposed
interest in an entity. The methods that should be applied to it. Of the respondents in favor of a tax adjustment, 38%
most frequently to determine the required rate of return in preferred a corporate tax rate adjustment (currently 28%),
practice when valuing an interest in an entity’s equity, 32% an adjustment at the top marginal rate for individuals
according to academic thinking, are presented in Figure (40%), and 29% the corporate tax rate plus a dividends
4. tax (34.5%).
5342 Afr. J. Bus. Manage.

Adjusted CAPM for additional risk

CAPM

Build-up method*

Whatever the investors require

Back-out of the dividend/earnings model

Average historical return on equity

Other

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Figure 4. Methods that, according to academia, should be used most frequently to determine the required rate of return when valuing
an interest in an entity’s equity.

R153 (1.7*)
2.50
2.00
1.50
1.00
Other (0.3) R157 (6.7)
0.50
-

R196 (0.2) R186 (18)

Figure 5. Instruments that, according to academia, should be used most
frequently as proxies for risk-free rates when employing the CAPM in
practice.*The numbers in parenthesis indicates the bonds’ term to
maturity.

Calculating beta when employing the CAPM judgment, presumably referring to forward looking betas
vis-à-vis trailing betas.
The sixth question required respondents to indicate how The seventh question required respondents to indicate
beta should be calculated when using the CAPM. The how the CAPM should be approached when entities are
sources of information for the calculation of beta that, confronted with negative betas, which strangely enough
according to academic thinking, should be used most are published by service providers (Bradfield, 2003;
frequently are presented in Figure 6. Arnott et al., 2009). As illustrated in Figure 7, the majority
As is evident from Figure 6, the majority of academia (57%) of the respondents indicated that a different beta
favors UCT/Cadiz financial services and McGregor BFA should be estimated, while an alarming 30% indicated
as service providers (both scored a 1.20). that the negative beta should be used.
According to academia, the second best option would A further 10% of the respondents indicated that beta
be to calculate historical betas in-house, followed by should be adjusted to zero and 3% indicated that a
either using data from Reuters or Bloomberg. Only 3% of different approach should be adopted to estimate the cost
the respondents indicated that beta should be based on of equity.
 Nel 5343

UCT/Cadiz
financial
services
2.00

1.50
Other, please McGregor
specify 1.00 BFA
0.50

Calculating
Bloomberg historical
betas

Reuters
Figure 6. Sources for the calculation of beta that, according to
academia, should be used most frequently when employing the CAPM
in practice.

A dif f erent beta should be estimated

The negative beta should be used

Beta should be adjusted to zero

Other, please specif y

0% 20% 40% 60% 80% 100%

Figure 7. Approach to the use of beta that, according to academia, should be adopted when confronted with a negative beta in
practice.

Gap analysis Similarities

The results of the academic survey, when compared to Although academia and investment practitioners place a
that of the PwC survey, revealed that academic thinking different emphasis (a 35% gap) on the use of the CAPM,
and investment practitioners’ preferences regarding the the majority of academia (74% on average) and invest-
use of the CAPM and alternative methods for the ment practitioners (100%) agree that the CAPM should
calculation of the cost of equity differ more frequently be used most frequently in practice. Academia and
than they concur. Six of the eight factors that were investment practitioners also seem to agree regarding the
compared indicated significant differences, in other words calculation of beta. The majority (97%) of respondents
gaps of more than 20%. Despite their differing opinions, from academia indicated that beta should be calculated
academia and investment practitioners agree on certain based on historical figures, a sentiment which is shared
issues. by investment practitioners (100%). Although academia
5344 Afr. J. Bus. Manage.

Table 1. Survey results.

Factor Academia** (%) Practice** (%) Gap % Similar/Different

Cost of equity
CAPM 74 100 35 Different
APT 58 3 (1 833) Different

Risk-free rate
R153 58 17 (241) Different
R157 51 67 31 Different

Reason*
Benchmark government bond 41 NI
Adjusted for taxation 64 4 (1 500) Different
Which tax rate* NI NA
Company tax rate 38 NI NA

Beta
Historical figures 97 100 3 Similar

Service provider
Bloomberg 5 43 760 Different
UCT/Cadiz financial services 32 37 16 Similar
McGregor BFA 32 33 3 Similar

Negative beta*
Estimate a different beta 57 NI NA
*These factors were not included (NI) in both surveys and therefore, a direct comparison was somewhat obscured. **The frequency
scores were converted to percentages for comparative purposes.

and investment practitioners differed regarding their on the item than investment practitioners. The calculation
preferred service provider for the calculation of beta, they of beta by using historical figures, for example, displays a
indicated similar support for UCT/Cadiz financial risk- relatively small gap (3%), which means that academia
services (32 and 37%, respectively) and McGregor BFA and investment practitioners agree that the calculation of
(32 and 33%, respectively) as alternative service beta should be based on historical figures. Table 1
providers. The results of the respective surveys are confirms that academia and investment practitioners’
compared in Table 1. preferences are not well aligned, that is they disagree
The use of historical betas as an estimate for future more than they agree.
betas is somewhat controversial. Empirical evidence
suggests that individual stocks are not stable over time,
which may cloud the validity of using historical betas as Differences
measures of risk (Blume, 1971; Baesel, 1974; Roenfeldt,
1978). Beta should rather be estimated based on the Although the respondents from academia and practice
future risk perspective of the asset being valued. seem to agree that the CAPM is the most preferred
The Gap column in Table 1 indicates the extent of the method of calculating the cost of equity, they differ rather
difference in emphasis between academia and significantly regarding other considerations. A significant
investment practitioners with regard to the calculation of portion of academia (58%) regards APT as an alternative
the cost of equity, the risk-free rate and beta. The per- to the CAPM, whereas investment practitioners (3%)
centage indicated in the Gap column reflects the extent to place very little emphasis on APT. This constitutes a very
which academia and investment practitioners disagree large (1 833%) gap between academia and practice,
regarding the item in question. A positive percentage indicating that academia places a far higher premium on
indicates that the item is used more frequently in practice APT than investment practitioners do.
than academia would advocate, while a negative per- An interesting discrepancy between academia and
centage implies that academia places a greater emphasis investment practitioners lies in the use of the R153 and
 Nel 5345

the R157 bonds as proxies for the Rf. Academia (58%) the calculation of beta, they agree that beta should be
favors the R153, while only 17% of investment practi- calculated from historical figures. However, academia
tioners favor the R153 as a proxy for the Rf, constituting and investment practitioners disagree regarding the
a (241%) gap. Similarly, investment practitioners use the relative frequency of use of the CAPM vis-à-vis the APT
R157 (67%) fairly frequently in practice, whereas 51% of model. All the investment practitioners indicated that they
the respondents from academia favor the R157. used the CAPM frequently, compared to academic sup-
This discrepancy highlights an important principle when port of 74%, constituting a 35% gap between academia
selecting an appropriate Rf. The tenure of the bond and practice. The most significant gap between theory
should match the term of the investment, that is the and practice is with regard to the use of the APT model.
choice of the appropriate Rf should be gleaned from the Academia (58%) seems to place significant emphasis on
yield curve. Academia’s preference for the R153 bond, APT, compared to investment practitioners (3%) who
which at the time of the survey had a 1.7 year term to seem to have less regard for APT, constituting a
maturity, above the R157 bond, with a 6.7 year term to significant (1 833%) gap.
maturity, seems rather short sighted. Similarly, very few The most concerning discrepancy between academia
investment practitioners (4%) adjust the Rf for tax in and investment practitioners is perhaps the use of an
practice, whereas 64% of the respondents from appropriate Rf in the CAPM. Academia favored the R153,
academia were in favor of such an adjustment. The gap which, at the time of the survey, had a 1.7 year term to
between academia and practice is therefore, a vast (1 maturity, compared to investment practitioners who
500%), indicating that the adjustment is made far less favored the R157, which, at that stage, had a 6.7 year
frequently in practice than academia may suggest. term to maturity. Academia seems to ignore the
Academia (5%) and investment practitioners (43%) also underlying principle for selecting an appropriate Rf, which
differ in their regard for Bloomberg as a service provider, is to match the maturity of the bond with the lifespan of
with a gap of 760%, indicating that investment the asset being valued.
practitioners have a higher regard for Bloomberg as a This is confirmed by the fact that the majority of
service provider than academia. academia indicated that the reason for the choice of an
appropriate Rf was based on the fact that the R153 was
the benchmark government bond. The gaps between
SUMMARY AND CONCLUSIONS academia and practice with regard to the R153 and R157
were (241%) and 31%, respectively. A strange
The research aimed to establish whether there is a gap discrepancy surfaced between academia and practice
between the calculation of the cost of equity, as regarding the adjustment of the Rf for tax purposes. The
advocated by academia, and the calculations that leading majority of academia (64%) is of the opinion that the Rf
financial analysts and corporate financiers apply in should be adjusted for tax purposes, compared to only4%
practice in South Africa. of investment practitioners, constituting a (1 500%) gap. It
The reader should bear in mind that the research was is not clear why investment practitioners tend to apply a
based on the beliefs and opinions of chartered pre-tax Rf in practice. Post-tax cash flows should be
accountants, which constitutes a specific target audience discounted at a post-tax rate.
within the broader academic environ-ment. One could be Although academia and investment practitioners agree
inclined to argue that, since there are members of the that beta should be calculated from historical figures,
academic community who lecture valuations who are not discounting future cash flows at a rate that encapsulates
chartered accountants, the target audience was narrowly historic data seems odd.
defined, which may have ob-scured the generalisation of The more accurate approach would perhaps be to
the results. However, since valuations is a key application adopt a forward perspective on beta and use that in the
area in finance and in the SAICA syllabus in particular the CAPM. Academia and investment practitioners also
research results contribute to the continued development disagree regarding the preferred service provider for the
of the academic environment responsible for the future calculation of beta.
training of chartered accountants. Although the broader Despite investment practitioners indicating Bloomberg
academic environment may have similar concerns regar- as their preferred service provider, only 5% of academia
ding valuations, which were not included in this study, the regarded Bloomberg as a frequent service provider,
author intends to investigate the matter with further constituting a 760% gap. Although the majority of
research. academia indicated that when confronted with a negative
Although the research results revealed that there is a beta, an alternative beta should be estimated, a
significant gap between theory and practice, academia significant portion (30%) of academia indicated that the
and investment practitioners seem to agree regarding the negative beta should be used in the CAPM, while a
use of the CAPM and the calculation of beta. Academia further 10% stated that beta should be adjusted to zero.
and investment practitioners agree that the CAPM is the This is a rather strange phenomenon, since neither of
best approach to calculate the cost of equity. In terms of these alternatives would render the CAPM useful. The
5346 Afr. J. Bus. Manage.

first will result in a rate of return which is lower than the Damodaran A (2007). Valuation Approaches and Metrics: A Survey of
nd
the Theory and Evidence. 2 edition. Hanover Now Publishers Inc.
Rf, which no rational investor would be interested in; and
Fama EF (1970). Efficient Capital Markets: A Review of Theory and
the second will result in a rate of return which is equal to Empirical Work. J. Financ., 25(2): 383-417.
the Rf, which is also not plausible. This paper has Fama EF, French KR (1996). The CAPM is Wanted, Dead or Alive. J.
highlighted the need for academia and investment Financ., 51(5): 1947-1958.
practitioners in South Africa to converge on mainstream Fama EF, French KR (2004). The Capital Asset Pricing Model: Theory
and evidence. J. Econ. Perspect., 18(3): 26-46.
valuation practices, a phenomenon that is common in Fernández P (2002). Valuation methods and shareholders’ value
developed markets. To this end, academia would do well creation. San Diego Academic press: 587-596.
to inform their students that the underlying concept when Goedhart M, Koller T, Wessels D (2005). The right role for multiples in
valuation. The McKinsey Quarterly. March. New York McKinsey and
choosing an appropriate Rf, for example, is to match the
Company, Inc.: 7-11.
term to maturity of the bond with the lifespan of the asset Harrington DR, Korajczyk RA (1993). The CAPM controversy: Policy
being valued, and that the benchmark government bond and strategy implications for investment management, an overview.
is not necessarily appropriate. As is evident from the Working paper. Charlottesville, 1-10.
Jegadeesh N (1992). Does Market Risk Really Explain the Size Effect?
survey results, this is not advocated in lecture halls.
J. Finan. Quant. Anal., 27(3): 337-351.
Similarly, academia should perhaps present beta with Lintner J (1965). The Valuation of Risk Assets and the Selection of
greater clarity, that is explain why historical figures are Risky Investments in Stock Portfolios and Capital Budgets. Rev.
used, but emphasize that this is, essentially, a second Econ. Stat., 47(1): 13 – 37.
best theory and that a negative beta cannot be applied in Magni CA (2005). Firm value and the mis-use of CAPM for valuation
and decision making. Appl. Econ. Res. Bull. [Peer-reviewed working
the CAPM, despite the fact that service providers publish paper series]. Modena University of Modena: 1-18.
negative betas. Magni CA (2007). Project valuation and investment decisions: CAPM
Similarly, investment practitioners should perhaps take versus arbitrage. Appl. Finan. Econ. Lett. 3(2): 1-9.
Mirza N (2005). The death of CAPM: A critical review. Lahore J. Econ.,
note of the APT as an alternative to the CAPM.
10(2): 35-54.
Investment practitioners should also consider adjusting Mossin J (1966). Equilibrium in a Capital Asset Market. Econometrica.
the Rf for tax purposes, as suggested by academia. 34(4): 768-783.
The research results provide an insight and guideline to Nawalkha SK (2007). Is the arbitrage pricing theory dead? Working
finance lecturers and investment practitioners, in terms of paper. Amherst University of Massachusetts: 1-40.
Nel WS (2009). Methods of choice in the valuation of ordinary
the perception in academia regarding the calculation of shareholders’ equity. Meditari Account. Res., 17(2):117-135.
the cost of equity and the application of the CAPM. The Nel WS (2010). A South African perspective on the multiples of choice
results indicate that, although academia and investment in the valuation of ordinary shareholders’ equity: From theory to
practitioners agree on the use of the CAPM and certain practice. Afr. J. Bus. Manage., 4(6):930-941.
Perold AF (2009). The Capital Asset Pricing Model. J. Econ. Perspect.,
service providers, they disagree significantly on the use 18(3): 3-24.
of APT, the choice of an appropriate Rf and adjusting the PricewaterhouseCoopers (PwC) (2008). Valuation Methodology Survey.
Rf for tax purposes. [Online]. http: //www.pwc.com. Accessed: 9 April 2010.
Ralston N (2003). Overview: Equity valuation in a global context. AIMR
conference proceedings. April 2003. CFA Institute: 1-4.
REFERENCES Roenfeldt R (1978). Further Evidence on the Stationarity of Beta
Coefficients. J. Finan. Quant. Anal., 13(1): 11 – 21.
Ammann M, Verhofen M (2007). Testing conditional asset pricing Roll R (1977). A critique of the asset pricing theory’s tests: Part I: On
models using a Markov chain Monte Carlo approach. Eur. Finan. the past and potential testability of the theory. J. Finan. Econ., 4(2):
Manage., 14(3): 391-418. 129-176.
Arnott R, Kalesnik V, Moghtader P, Scholl C (2009). Beyond cap- Ross SA (1976). The arbitrage theory of capital asset pricing. J. Econ.
weight: The search for efficient beta. J. Indexes, 13(1):1-25. Theory, 13(3): 341-360.
Baesel JB (1974). On the Assessment of Risk: Some Further Sharpe W (1964). Capital Asset Prices: A Theory of Market Equilibrium
Consideration. J. Financ., 29(5): 1491-1494. under Conditions of Risk. J. Financ., 19(3): 425-442.
Bello ZY (2008). A statistical comparison of the CAPM to the Fama- Smith VK, Goudzwaard MB (1970). Survey of investment management:
French three factor model and the Cahart’s model. Global J. Financ. Teaching and practice. J. Financ., 25(2): 329-347.
Banking, 2(2): 14-24. South African Institute of Chartered Accountants (SAICA) (2008). SA
Berkman H, Bradhury ME, Ferguson J (2000). The accuracy of price Competency Framework. Johannesburg SAICA.
earnings and discounted cash flow methods of IPO equity valuations. Triantis A (2005). Realizing the potential of real options: Does theory
J. Int. Financ. Manage. Account., 11(2): 71-83. meet practice? J. Appl. Corp. Finan., 17(2): 8-16.
Bernstein WJ (2008). Fifty years from now. [Online]. Upton RM (1949). Conference on the teaching of business finance. J.
www.efficientfrontier.com. Accessed: 9 July 2010. Financ., 4(3): Summary statement.
Blume EM (1971). On the Assessment of Risk. J. Financ., 6(1): 1-10. Wendt PF (1966). What should we teach in an investment course? J.
Bradfield D (2003). Investment basics XLVI. On estimating the beta Financ., 21(2): 416-422.
coefficient. Invest. Anal. J., 57(1): 47-53.
Bruner F, Conroy RM, Estrada J, Kritzman M, Li W (2002). Introduction
to ‘Valuation in emerging markets’. Emerg. Markets Rev., 3(4): 300-
324.
Burton J (1998). Revisiting the capital asset pricing model. Dow Jones
Asset Manage., 20–28.
Courteau L, Kao JL, O’Keefe T, Richardson GD (2003). Gains to
valuation accuracy of direct valuation over sector multiplier
approaches. Working paper. Turkey Eastern Mediterranean
University: 1-46.
 Nel 5347

APPENDIX

List of respondents to the PWC survey

ABN AMRO Bank of South Africa

Bridge Capital
Deutsche Bank
Ernst and Young
Ethos
HSBC Bank
iCapital
Investec
Java Capital
KPMG
Kumba Resources
McGregor BFA
Morgan Stanley South Africa
Nedbank Capital
NM Rothschild and Sons (South Africa)
Old Mutual Asset Managers
PricewaterhouseCoopers
PSG Capital
Rand Merchant Bank
SABMiller
Sanlam
Sasol
Standard Bank

View publication stats