The Capital Asset Pricing

Model (CAPM)

1
 Achievable Portfolio Combinations
The Two-Asset Case

• It is possible to construct a series of portfolios with different

risk/return characteristics just by varying the weights of the two
assets in the portfolio.
• Assets A and B are assumed to have a correlation coefficient of -
0.379 and the following individual return/risk characteristics

Expected Return Standard

Deviation
Asset A 8% 8.72%
Asset B 10% 22.69%

The following table shows the portfolio characteristics for 100 different
weighting schemes for just these two securities:

9-2
 Example of Portfolio Combinations and
Correlation
You repeat this
procedure Expected Standard Correlation
down until you Asset Return Deviation Coefficient
have determine A 8.0% 8.7% -0.379
the portfolio
B 10.0% 22.7%
characteristics
The first
for all
The 100
second Portfolio Components Portfolio Characteristics
combination
portfolios.
portfolio Expected Standard
simply99%
assumes Weight of A Weight of B Return Deviation
inNext
assumes plot1%
A and the
you
in 100% 0% 8.00% 8.7%
returns
B. Notice onthe
a 99% 1% 8.02% 8.5%
invest
graph
solely
(see in
increase the 98% 2% 8.04% 8.4%
innext
Asset
return slide)
and A
the 97% 3% 8.06% 8.2%
decrease in 96% 4% 8.08% 8.1%
portfolio risk! 95% 5% 8.10% 7.9%
94% 6% 8.12% 7.8%
93% 7% 8.14% 7.7%
92% 8% 8.16% 7.5%
91% 9% 8.18% 7.4%
90% 10% 8.20% 7.3%
89% 11% 8.22% 7.2%

9-3
 Attainable Portfolio Combinations for a
Two Asset Portfolio

Example of 12.00%
Portfolio
Expected Return of the
Combinations 10.00%
and Correlation
8.00%
Portfolio

6.00%

4.00%

2.00%

0.00%
0.0% 5.0% 10.0% 15.0% 20.0% 25.0%
Standard Deviation of Returns

9-4
 Two Asset Efficient Frontier

• Figure 8 – 10 describes five different portfolios

(A,B,C,D and E in reference to the attainable
set of portfolio combinations of this two asset
portfolio.

(See Figure 8 -10 on the following slide)

9-5
 Efficient Frontier
The Two-Asset Portfolio Combinations

8 - 10 FIGURE

A is not attainable
B,E lie on the
efficient frontier and
are attainable
A B E is the minimum
Expected Return %

variance portfolio
C (lowest risk
combination)

C, D are
E attainable but are
D dominated by
superior portfolios
that line on the line
above E
Standard Deviation (%)

9-6
 Achievable Set of Portfolio
Combinations
Getting to the ‘n’ Asset Case
• In a real world investment universe with all of the investment
alternatives (stocks, bonds, money market securities, hybrid
instruments, gold real estate, etc.) it is possible to construct
many different alternative portfolios out of risky securities.
• Each portfolio will have its own unique expected return and
risk.
• Whenever you construct a portfolio, you can measure two
fundamental characteristics of the portfolio:
– Portfolio expected return (ERp)
– Portfolio risk (σp)

9-7
 The Achievable Set of Portfolio
Combinations

• You could start by randomly assembling ten

risky portfolios.
• The results (in terms of ER p and σp )might
look like the graph on the following page:

9-8
Achievable Portfolio Combinations
The First Ten Combinations Created

ERp

10 Achievable
Risky Portfolio
Combinations

Portfolio Risk (σp)

9-9
 The Achievable Set of Portfolio
Combinations

• You could continue randomly assembling

more portfolios.
• Thirty risky portfolios might look like the graph
on the following slide:

9 - 10
Achievable Portfolio Combinations
Thirty Combinations Naively Created

ERp

30 Risky Portfolio
Combinations

Portfolio Risk (σp)

9 - 11
 Achievable Set of Portfolio
Combinations
All Securities – Many Hundreds of Different Combinations

• When you construct many hundreds of

different portfolios naively varying the weight
of the individual assets and the number of
types of assets themselves, you get a set of
achievable portfolio combinations as indicated
on the following slide:

9 - 12
Achievable Portfolio Combinations
More Possible Combinations Created

The highlighted
portfolios are
ERp ‘efficient’ in that
they offer the
highest rate of
E is the return for a given
minimum level of risk.
variance Rationale investors
portfolio Achievable Set of will choose only
Risky Portfolio from this efficient
Combinations set.

Portfolio Risk (σp)

9 - 13
Achievable Portfolio Combinations
Efficient Frontier (Set)

Efficient
ERp frontier is the
set of
achievable
portfolio
combinations
Achievable Set of that offer the
Risky Portfolio
Combinations
highest rate
of return for a
given level of
E
risk.

Portfolio Risk (σp)

9 - 14
 The New Efficient Frontier
Efficient Portfolios

9 - 1 FIGURE
Figure 9 – 1
illustrates
Efficient Frontier three
ER
achievable
portfolio
combinations
B
that are
A ‘efficient’ (no
other
achievable
MVP portfolio that
offers the
same risk,
Risk offers a higher
return.)

9 - 15
 Underlying Assumption
Investors are Rational and Risk-Averse

• We assume investors are risk-averse wealth maximizers.

• This means they will not willingly undertake fair gamble.
– A risk-averse investor prefers the risk-free situation.
– The corollary of this is that the investor needs a risk premium to be
induced into a risky situation.
– Evidence of this is the willingness of investors to pay insurance
premiums to get out of risky situations.
• The implication of this, is that investors will only choose portfolios
that are members of the efficient set (frontier).

9 - 16
 Risk-free Investing

• When we introduce the presence of a risk-free

investment, a whole new set of portfolio
combinations becomes possible.
• We can estimate the return on a portfolio
made up of RF asset and a risky asset A letting
the weight w invested in the risky asset and
the weight invested in RF as (1 – w)

9 - 17
 The New Efficient Frontier
Risk-Free Investing

– Expected return on a two asset portfolio made up

of risky asset A and RF:

[9-1] ER p  RF  w (ER A - RF)

The possible combinations of A and RF are found graphed on the following slide.

9 - 18
 The New Efficient Frontier
Attainable Portfolios Using RF and A

9 - 2 FIGURE

This means
you can 9 – 2
Equation
Rearranging 9
ER achieve
illustratesany
-2 where w=σ
portfolio
what you can
p / σA and
combination
see…portfolio
substituting in
 pA ) - w
 E(R RFA along
risk the blue
increases
[9-3] ER P[9-2]
 RF    P Equation 1 we
 coloured
in line
A  A  getdirect
an
simply by to
proportion
equation for a
changing
the amount the
RF straight line
relative
invested weight
with a in the
of RFasset.
risky and A in
constant
the two asset
slope.
portfolio.
Risk

9 - 19
 The New Efficient Frontier
Attainable Portfolios using the RF and A, and RF and T

9 - 3 FIGURE

Which risky
portfolio
ER would a
rational risk-
T
averse
investor
A choose in the
presence of a
RF
RF investment?
Portfolio A?
Tangent
Risk Portfolio T?

9 - 20
 The New Efficient Frontier
Efficient Portfolios using the Tangent Portfolio T

9 - 3 FIGURE
Clearly RF with
T (the tangent
portfolio) offers
ER a series of
portfolio
combinations
T
that dominate
A those produced
by RF and A.
Further, they
RF
dominate all but
one portfolio on
the efficient
Risk frontier!

9 - 21
 The New Efficient Frontier
Lending Portfolios

9 - 3 FIGURE
Portfolios
between RF
and T are
Lending Portfolios ‘lending’
ER
portfolios,
because they
T
are achieved by
A investing in the
Tangent
Portfolio and
RF lending funds to
the government
(purchasing a
T-bill, the RF).
Risk

9 - 22
 The New Efficient Frontier
Borrowing Portfolios

9 - 3 FIGURE
The line can be
extended to risk
levels beyond
Lending Portfolios Borrowing Portfolios ‘T’ by
ER
borrowing at RF
and investing it
T
in T. This is a
A levered
investment that
increases both
RF risk and
expected return
of the portfolio.

Risk

9 - 23
 The New Efficient Frontier
The New (Super) Efficient Frontier

9 - 4 FIGURE This is now

called
Clearlythe RFnewwith
Capital Market Line (or super)
T (the market
The optimal
efficient
portfolio) frontier
offers
ER risky portfolio
B2 of risky
a series of
(the market
portfolios.
portfolio
portfolio ‘M’)
T B
combinations
Investors can
that dominate
A2 achieve any
those produced
one of these
by RF and A.
A portfolio
RF
combinations
Further, they by
borrowing
dominate all or but
σρ investing
one portfolio in RF
on
in
thecombination
efficient
with the market
frontier!
portfolio.
9 - 24
 The New Efficient Frontier
The Implications – Separation Theorem – Market Portfolio

• All investors will only hold individually-determined combinations of:

– The risk free asset (RF) and
– The model portfolio (market portfolio)
• The separation theorem
– The investment decision (how to construct the portfolio of risky
assets) is separate from the financing decision (how much should be
invested or borrowed in the risk-free asset)
– The tangent portfolio T is optimal for every investor regardless of
his/her degree of risk aversion.
• The Equilibrium Condition
– The market portfolio must be the tangent portfolio T if everyone holds
the same portfolio
– Therefore the market portfolio (M) is the tangent portfolio (T)

9 - 25
 The New Efficient Frontier
The Capital Market Line

The CML is that

CML set of superior
The optimal
portfolio
ER risky portfolio
combinations
(the market
that are ‘M’)
portfolio
M achievable in
the presence of
the equilibrium
condition.
RF

σρ

9 - 26
The Capital Asset Pricing Model
What is it?

– An hypothesis by Professor William Sharpe

• Hypothesizes that investors require higher rates of
return for greater levels of relevant risk.
• There are no prices on the model, instead it
hypothesizes the relationship between risk and return
for individual securities.
• It is often used, however, the price securities and
investments.

9 - 27
 The Capital Asset Pricing Model
How is it Used?

– Uses include:
• Determining the cost of equity capital.
• The relevant risk in the dividend discount model to estimate a stock’s intrinsic
(inherent economic worth) value. (As illustrated below)

Estimate Investment’s Determine Investment’s Estimate the Compare to the actual

Risk (Beta Coefficient) Required Return Investment’s Intrinsic stock price in the
Value market

COVi,M D1
i 
σ M2
ki  RF  ( ERM  RF )  i P0  Is the stock
kc  g fairly priced?

9 - 28
The Capital Asset Pricing Model
Assumptions

– CAPM is based on the following assumptions:

1. All investors have identical expectations about expected
returns, standard deviations, and correlation coefficients for
all securities.
2. All investors have the same one-period investment time
horizon.
3. All investors can borrow or lend money at the risk-free rate
of return (RF).
4. There are no transaction costs.
5. There are no personal income taxes so that investors are
indifferent between capital gains an dividends.
6. There are many investors, and no single investor can affect
the price of a stock through his or her buying and selling
decisions. Therefore, investors are price-takers.
7. Capital markets are in equilibrium.

9 - 29
 Market Portfolio and Capital Market
Line
• The assumptions have the following
implications:
1. The “optimal” risky portfolio is the one that is
tangent to the efficient frontier on a line that is
drawn from RF. This portfolio will be the same for all
investors.
2. This optimal risky portfolio will be the market
portfolio (M) which contains all risky securities.

(Figure 9 – 4 illustrates the Market Portfolio ‘M’)

9 - 30
 The Capital Market Line
9 - 5 FIGURE

ER
CML

The CML is that

 ERM  RF  setThe
of achievable
market
k P  RF    P
ERM M Theportfolio
CMLishas
portfolio the
 M  combinations
standard
optimal risky
thatdeviation
are possible
portfolio, of
it
portfolio
when investing
contains returns
all risky
in as
only
the
securities two
and
RF independent
assets
lies (the(T)
tangent
market
on variable.
portfolio
the efficient
and frontier.
the risk-free
σρ asset (RF).

σM

9 - 31
The Capital Asset Pricing Model
The Market Portfolio and the Capital Market Line (CML)

– The slope of the CML is the incremental expected

return divided by the incremental risk.

ER M - RF
[9-4] Slope of the CML 
M

– This is called the market price for risk. Or

– The equilibrium price of risk in the capital market.
9 - 32
The Capital Asset Pricing Model
The Market Portfolio and the Capital Market Line (CML)

– Solving for the expected return on a portfolio in the presence of a RF

asset and given the market price for risk :

 ERM - RF 
[9-5] E ( RP )  RF    P
 σM 

– Where:
• ERM = expected return on the market portfolio M
• σM = the standard deviation of returns on the market portfolio
• σP = the standard deviation of returns on the efficient portfolio being
considered

9 - 33
 The Capital Market Line
Using the CML – Expected versus Required Returns

– In an efficient capital market investors will require

a return on a portfolio that compensates them for
the risk-free return as well as the market price for
risk.
– This means that portfolios should offer returns
along the CML.

9 - 34
 The Capital Asset Pricing Model
Expected and Required Rates of Return

9 - 6 FIGURE
B is an
C
A a portfolio that
overvalued
offers andExpected
undervalued
portfolio. expected
Required equal
portfolio.
return is lesstothan
the
Expected
Return on C
ER CML required
return
the return.
required
is greater
return.
than the required
Expected
A Selling pressure
return on A return.
will cause the price
Demand
to fall andfor
the yield
C Portfolio
to rise until
A will
Required increase driving
expected equalsup
return on A
B the required
price, andreturn.
therefore the
Expected
Return on C expected return will
RF
fall until expected
equals required
(market equilibrium
condition is
achieved.)
σρ

9 - 35
The Capital Asset Pricing Model
Risk-Adjusted Performance and the Sharpe Ratios

– William Sharpe identified a ratio that can be used to assess the risk-adjusted
performance of managed funds (such as mutual funds and pension plans).
– It is called the Sharpe ratio:

ER P - RF
[9-6] Sharpe ratio 
P

– Sharpe ratio is a measure of portfolio performance that describes how well an

asset’s returns compensate investors for the risk taken.
– It’s value is the premium earned over the RF divided by portfolio risk…so it is
measuring valued added per unit of risk.
– Sharpe ratios are calculated ex post (after-the-fact) and are used to rank
portfolios or assess the effectiveness of the portfolio manager in adding value
to the portfolio over and above a benchmark.

9 - 36
The Capital Asset Pricing Model
Sharpe Ratios and Income Trusts

– Table 9 – 1 (on the following slide) illustrates

return, standard deviation, Sharpe and beta
coefficient for four very different portfolios from
2002 to 2004.
– Income Trusts did exceedingly well during this
time, however, the recent announcement of
Finance Minister Flaherty and the subsequent
drop in Income Trust values has done much to
eliminate this historical performance.
9 - 37
 Income Trust Estimated Values

Table 9-1 Income Trusts Estimated Values

Return σP Sharpe β

Median income trusts 25.83% 18.66% 1.37 0.22

Equally weighted trust portfolio 29.97% 8.02% 3.44 0.28
S&P/TSX Composite Index 8.97% 13.31% 0.49 1.00
Scotia Capital government bond index 9.55% 6.57% 1.08 20.02

Source: Adapted from L. Kryzanowski, S. Lazrak, and I. Ratika, " The True
Cost of Income Trusts," Canadian Investment Review 19, no. 5 (Spring
2006), Table 3, p. 15.

9 - 38
Diversifiable and Non-Diversifiable Risk

• CML applies to efficient portfolios

• Volatility (risk) of individual security returns are caused by two
different factors:
– Non-diversifiable risk (system wide changes in the economy and
markets that affect all securities in varying degrees)
– Diversifiable risk (company-specific factors that affect the returns of
only one security)
• Figure 9 – 7 illustrates what happens to portfolio risk as the
portfolio is first invested in only one investment, and then slowly
invested, naively, in more and more securities.

9 - 39
 The CAPM and Market Risk
Portfolio Risk and Diversification

9 - 7 FIGURE

Total Risk (σ)

Market or
systematic
Unique (Non-systematic) Risk
risk is risk
that cannot
be eliminated
from the
portfolio by
investing the
Market (Systematic) Risk
portfolio into
more and
different
securities.
Number of Securities

9 - 40
 Relevant Risk
Drawing a Conclusion from Figure 9 - 7

• Figure 9 – 7 demonstrates that an individual securities’ volatility of

return comes from two factors:
– Systematic factors
– Company-specific factors
• When combined into portfolios, company-specific risk is diversified
away.
• Since all investors are ‘diversified’ then in an efficient market, no-
one would be willing to pay a ‘premium’ for company-specific risk.
• Relevant risk to diversified investors then is systematic risk.
• Systematic risk is measured using the Beta Coefficient.

9 - 41
 The Beta Coefficient
What is the Beta Coefficient?

• A measure of systematic (non-diversifiable)

risk
• As a ‘coefficient’ the beta is a pure number
and has no units of measure.

9 - 42
 The Beta Coefficient
How Can We Estimate the Value of the Beta Coefficient?

• There are two basic approaches to estimating

the beta coefficient:

1. Using a formula (and subjective forecasts)

2. Use of regression (using past holding period
returns)

(Figure 9 – 8 on the following slide illustrates the characteristic line used to estimate the
beta coefficient)

9 - 43
 The CAPM and Market Risk
The Characteristic Line for Security A
9 - 8 FIGURE

Security A Returns (%)

4 The
Theslope
plottedof

Market Returns (%)

the
points
regression
are the
line
coincident
is beta.
2
rates of return
earned
The lineon of
the
0 investment
best fit is
-6 -4 -2 0 2 4 6 8 andknown
the market
in
-2 finance
portfolioasover
the
characteristic
past periods.
line.
-4

-6

9 - 44
The Formula for the Beta Coefficient

Beta is equal to the covariance of the returns of

the stock with the returns of the market, divided
by the variance of the returns of the market:

COVi,M  i , M  i
[9-7] i  
2
σM M

9 - 45
 The Beta Coefficient
How is the Beta Coefficient Interpreted?

• The beta of the market portfolio is ALWAYS = 1.0

• The beta of a security compares the volatility of its returns to the volatility of the market
returns:

βs = 1.0 - the security has the same volatility as the market as a

whole

βs > 1.0 - aggressive investment with volatility of returns

greater than the market

βs < 1.0 - defensive investment with volatility of returns less

than the market

βs < 0.0 - an investment with returns that are negatively

correlated with the returns of the market

Table 9 – 2 illustrates beta coefficients for a variety of Canadian Investments

9 - 46
 Canadian BETAS
Selected

Table 9-2 Canadian BETAS

Company Industry Classification Be ta

Abitibi Consolidated Inc. Materials - Paper & Forest 1.37

Algoma Steel Inc. Materials - Steel 1.92
Bank of Montreal Financials - Banks 0.50
Bank of Nova Scotia Financials - Banks 0.54
Barrick Gold Corp. Materials - Precious Metals & Minerals 0.74
BCE Inc. Communications - Telecommunications 0.39
Bema Gold Corp. Materials - Precious Metals & Minerals 0.26
CIBC Financials - Banks 0.66
Cogeco Cable Inc. Consumer Discretionary - Cable 0.67
Gammon Lake Resources Inc. Materials - Precious Metals & Minerals 2.52
Imperial Oil Ltd. Energy - Oil & Gas: Integrated Oils 0.80

Source: Research Insight, Com pustat North Am erican database, June 2006.

9 - 47
 The Beta of a Portfolio

The beta of a portfolio is simply the weighted average of the betas of

the individual asset betas that make up the portfolio.

[9-8]  P  wA  A  wB  B  ...  wn  n

Weights of individual assets are found by dividing the value of the

investment by the value of the total portfolio.

9 - 48
 The CAPM and Market Risk
The Security Market Line (SML)

– The SML is the hypothesized relationship between return (the

dependent variable) and systematic risk (the beta coefficient).
– It is a straight line relationship defined by the following formula:

[9-9] ki  RF  ( ERM  RF )  i

– Where:
ki = the required return on security ‘i’
ERM – RF = market premium for risk
Βi = the beta coefficient for security ‘i’

(See Figure 9 - 9 on the following slide for the graphical representation)

9 - 49
 The CAPM and Market Risk
The Security Market Line (SML)

9 - 9 FIGURE

ER ki  RF  ( ERM  RF )  i
M TheSML
The SMLis
ERM uses
usedtheto
beta
predict
coefficient
requiredas
the measure
returns for
of relevant
individual
RF
risk.
securities

βM = 1 β

9 - 50
 The CAPM and Market Risk
The SML and Security Valuation

9 - 10 FIGURE

Similarly,
Required
A is an returns
B is an
ER ki  RF  ( ERM  RF )  i are forecast using
undervalued
overvalued
this equation.
security
security. because
SML its expected return
You can see
Investor’s willthat
sell
is greater than the
thelock
to required
in gains,return
required return.
Expected A on any
but the security
selling is
Return A
a functionwill
Investors
pressure will
of its
Required
Return A B
systematic
‘flock’
cause to
theA market
and
risk bid
(β)
RF andthe
up
price market
toprice
fall,
factors (RF
causing expected
the and
markettopremium
return
expected fall
return
till itto
for risk)
equals
rise untiltheit equals
βA βB β required
the required return.
return.

9 - 51
 The CAPM in Summary
The SML and CML

– The CAPM is well entrenched and widely used by

investors, managers and financial institutions.
– It is a single factor model because it based on the
hypothesis that required rate of return can be
predicted using one factor – systematic risk
– The SML is used to price individual investments and
uses the beta coefficient as the measure of risk.
– The CML is used with diversified portfolios and uses
the standard deviation as the measure of risk.

9 - 52
 Challenges to CAPM
• Empirical tests suggest:
– CAPM does not hold well in practice:
• Ex post SML is an upward sloping line
• Ex ante y (vertical) – intercept is higher that RF
• Slope is less than what is predicted by theory
– Beta possesses no explanatory power for predicting stock returns
(Fama and French, 1992)
• CAPM remains in widespread use despite the foregoing.
– Advantages include – relative simplicity and intuitive logic.
• Because of the problems with CAPM, other models have been
developed including:
– Fama-French (FF) Model
– Abitrage Pricing Theory (APT)

9 - 53
Alternative Asset Pricing Models
The Fama – French Model

– A pricing model that uses three factors to relate

expected returns to risk including:
1. A market factor related to firm size.
2. The market value of a firm’s common equity (MVE)
3. Ratio of a firm’s book equity value to its market value of
equity. (BE/MVE)
– This model has become popular, and many think
it does a better job than the CAPM in explaining
ex ante stock returns.

9 - 54
Alternative Asset Pricing Models
The Arbitrage Pricing Theory

– A pricing model that uses multiple factors to relate expected returns

to risk by assuming that asset returns are linearly related to a set of
indexes, which proxy risk factors that influence security returns.

[9-10] ERi  a0  bi1 F1  bi1 F1  ...  bin Fn

– It is based on the no-arbitrage principle which is the rule that two

otherwise identical assets cannot sell at different prices.
– Underlying factors represent broad economic forces which are
inherently unpredictable.

9 - 55
Alternative Asset Pricing Models
The Arbitrage Pricing Theory – the Model

– Underlying factors represent broad economic forces which are inherently

unpredictable.

[9-10] ERi  a0  bi1 F1  bi1 F1  ...  bin Fn

– Where:
• ERi = the expected return on security i
• a0 = the expected return on a security with zero systematic risk
• bi = the sensitivity of security i to a given risk factor
• Fi = the risk premium for a given risk factor

– The model demonstrates that a security’s risk is based on its sensitivity to

broad economic forces.

9 - 56
Alternative Asset Pricing Models
The Arbitrage Pricing Theory – Challenges

– Underlying factors represent broad economic forces

which are inherently unpredictable.
– Ross and Roll identify five systematic factors:
1. Changes in expected inflation
2. Unanticipated changes in inflation
3. Unanticipated changes in industrial production
4. Unanticipated changes in the default-risk premium
5. Unanticipated changes in the term structure of interest rates

• Clearly, something that isn’t forecast, can’t be used

to price securities today…they can only be used to
explain prices after the fact.

9 - 57
 Summary and Conclusions

In this chapter you have learned:

– How the efficient frontier can be expanded by introducing risk-free

borrowing and lending leading to a super efficient frontier called the
Capital Market Line (CML)
– The Security Market Line can be derived from the CML and provides a
way to estimate a market-based, required return for any security or
portfolio based on market risk as measured by the beta.
– That alternative asset pricing models exist including the Fama-French
Model and the Arbitrage Pricing Theory.

9 - 58
Calculating a Beta Coefficient Using Ex Ante Returns

• Ex Ante means forecast…

• You would use ex ante return data if historical rates of return are
somehow not indicative of the kinds of returns the company will
produce in the future.
• A good example of this is Air Canada or American Airlines, before
and after September 11, 2001. After the World Trade Centre
terrorist attacks, a fundamental shift in demand for air travel
occurred. The historical returns on airlines are not useful in
estimating future returns.

9 - 59
 The Beta Coefficient
• Under the theory of the Capital Asset Pricing Model total risk is
partitioned into two parts:
– Systematic risk
– Unsystematic risk – diversifiable risk

Total Risk of the Investment

Systematic Risk Unsystematic Risk

• Systematic risk is non-diversifiable risk.

• Systematic risk is the only relevant risk to the diversified investor
• The beta coefficient measures systematic risk

9 - 60
 The Beta Coefficient
The Formula

Covariance of Returns between stock ' i' returns and the market
Beta 
Variance of the Market Returns

COVi,M  i , M  i
[9-7] i  
2
σM M

9 - 61
 The Term – “Relevant Risk”
• What does the term “relevant risk” mean in the context of the CAPM?
– It is generally assumed that all investors are wealth maximizing risk averse
people
– It is also assumed that the markets where these people trade are highly
efficient
– In a highly efficient market, the prices of all the securities adjust instantly to
cause the expected return of the investment to equal the required return
– When E(r) = R(r) then the market price of the stock equals its inherent worth
(intrinsic value)
– In this perfect world, the R(r) then will justly and appropriately compensate
the investor only for the risk that they perceive as relevant…
– Hence investors are only rewarded for systematic risk.

NOTE: The amount of systematic risk varies by investment. High systematic risk
occurs when R-square is high, and the beta coefficient is greater than 1.0

9 - 62
 The Proportion of Total Risk that is Systematic

• Every investment in the financial markets vary with respect to the

percentage of total risk that is systematic.

• Some stocks have virtually no systematic risk.

– Such stocks are not influenced by the health of the economy in general…
their financial results are predominantly influenced by company-specific
factors.
– An example is cigarette companies…people consume cigarettes because
they are addicted…so it doesn’t matter whether the economy is healthy
or not…they just continue to smoke.
• Some stocks have a high proportion of their total risk that is
systematic
– Returns on these stocks are strongly influenced by the health of the
economy.
– Durable goods manufacturers tend to have a high degree of systematic
risk.

9 - 63
The Formula Approach to Measuring the Beta

Cov(k i k M )
Beta 
Var(k M )
You need to calculate the covariance of the returns between the
stock and the market…as well as the variance of the market
returns. To do this you must follow these steps:
• Calculate the expected returns for the stock and the market
• Using the expected returns for each, measure the variance
and standard deviation of both return distributions
• Now calculate the covariance
• Use the results to calculate the beta

9 - 64
 Ex ante Return Data
A Sample

A set of estimates of possible returns and their respective probabilities looks as

follows:

Possible Since the beta

Future State Possible Possible relates the stock
By observation
returns to the
of the Returns on Returns on market
you can returns,
see the
Economy Probability the Stock the Market the
rangegreater range
is much
of stock returns
Boom 25.0% 28.0% 20.0% greater for the
changing in the
stock than theas
same direction
Normal 50.0% 17.0% 11.0% market
the market and they
Recession 25.0% -14.0% -4.0% indicates
move in the the beta
will
same be direction.
greater
than 1 and will be
positive.
(Positively
correlated to the
market returns.)
9 - 65
 The Total of the Probabilities must Equal 100%

This means that we have considered all of the possible outcomes in this discrete
probability distribution

Possible
Future State Possible Possible
of the Returns on Returns on
Economy Probability the Stock the Market
Boom 25.0% 28.0% 20.0%
Normal 50.0% 17.0% 11.0%
Recession 25.0% -14.0% -4.0%
100.0%

9 - 66
 Measuring Expected Return on the Stock
From Ex Ante Return Data
The expected return is weighted average returns from the given ex
ante data

(1) (2) (3) (4)

Possible
Future State Possible
of the Returns on
Economy Probability the Stock (4) = (2)*(3)
Boom 25.0% 28.0% 0.07
Norm al 50.0% 17.0% 0.085
Recession 25.0% -14.0% -0.035
Expected return on the Stock = 12.0%

9 - 67
Measuring Expected Return on the Market From Ex
Ante Return Data
The expected return is weighted average returns from the given ex
ante data

(1) (2) (3) (4)

Possible
Future State Possible
of the Returns on
Economy Probability the Market (4) = (2)*(3)
Boom 25.0% 20.0% 0.05
Norm al 50.0% 11.0% 0.055
Recession 25.0% -4.0% -0.01
Expected return on the Market = 9.5%

9 - 68
 Measuring Variances, Standard Deviations of the
Forecast Stock Returns
Using the expected return, calculate the deviations away from the mean, square those deviations and
then weight the squared deviations by the probability of their occurrence. Add up the weighted and
squared deviations from the mean and you have found the variance!

(1) (2) (3) (4) (5) (6) (7)

Possible Weighted
Future State Possible and
of the Returns on Squared Squared
Economy Probability the Stock (4) = (2)*(3) Deviations Deviations Deviations
Boom 25.0% 0.28 0.07 0.16 0.0256 0.0064
Norm al 50.0% 0.17 0.085 0.05 0.0025 0.00125
Recession 25.0% -0.14 -0.035 -0.26 0.0676 0.0169
Expected return (stock) = 12.0% Variance (stock)= 0.02455
Standard Deviation (stock) = 15.67%

9 - 69
Measuring Variances, Standard Deviations of the
Forecast Market Returns
Now do this for the possible returns on the market

(1) (2) (3) (4) (5) (6) (7)

Possible Weighted
Future State Possible and
of the Returns on Squared Squared
Economy Probability the Market (4) = (2)*(3) Deviations Deviations Deviations
Boom 25.0% 0.2 0.05 0.105 0.011025 0.002756
Norm al 50.0% 0.11 0.055 0.015 0.000225 0.000113
Recession 25.0% -0.04 -0.01 -0.135 0.018225 0.004556
Expected return (market) = 9.5% Variance (market) = 0.007425
Standard Deviation (market)= 8.62%

9 - 70
 Covariance
From Chapter 8 you know the formula for the covariance between the
returns on the stock and the returns on the market is:

n _ _
[8-12] COV AB   Prob i (k A,i  ki )(k B ,i - k B )
i 1

Covariance is an absolute measure of the degree of ‘co-movement’ of

returns.

9 - 71
 Correlation Coefficient
Correlation is covariance normalized by the product of the standard deviations of both
securities. It is a ‘relative measure’ of co-movement of returns on a scale from -1 to +1.

The formula for the correlation coefficient between the returns on the stock and the returns
on the market is:

COVAB
[8-13]  AB 
 A B
The correlation coefficient will always have a value in the range of +1 to -1.
+1 – is perfect positive correlation (there is no diversification potential when combining these two
securities together in a two-asset portfolio.)
- 1 - is perfect negative correlation (there should be a relative weighting mix of these two securities in a
two-asset portfolio that will eliminate all portfolio risk)

9 - 72
 Measuring Covariance
from Ex Ante Return Data

Using the expected return (mean return) and given data measure the deviations for
both the market and the stock and multiply them together with the probability of
occurrence…then add the products up.

(1) (2) (3) (4) (5) (6) (7) (8) "(9)

Possible Possible Deviations Deviations

Future Returns Possible from the from the
State of the on the (4) = Returns on mean for mean for
Economy Prob. Stock (2)*(3) the Market (6)=(2)*(5) the stock the market (8)=(2)(6)(7)
Boom 25.0% 28.0% 0.07 20.0% 0.05 16.0% 10.5% 0.0042
Normal 50.0% 17.0% 0.085 11.0% 0.055 5.0% 1.5% 0.000375
Recession 25.0% -14.0% -0.035 -4.0% -0.01 -26.0% -13.5% 0.008775
E(k stock) = 12.0% E(k market) = 9.5% Covariance = 0.01335

9 - 73
 The Beta Measured
Using Ex Ante Covariance (stock, market) and Market Variance

Now you can substitute the values for covariance and the variance of the
returns on the market to find the beta of the stock:

Cov S, M .01335
Beta    1 .8
Var M .007425

• A beta that is greater than 1 means that the investment is aggressive…its

returns are more volatile than the market as a whole.
• If the market returns were expected to go up by 10%, then the stock
returns are expected to rise by 18%. If the market returns are expected
to fall by 10%, then the stock returns are expected to fall by 18%.
9 - 74
 Lets Prove the Beta of the Market is 1.0

Let us assume we are comparing the possible market returns against

itself…what will the beta be?
(1) (2) (3) (4) (5) (6) (6) (7) (8)

Possible Possible Possible Deviations Deviations

Future Returns Cov Returns .007425 from the from the
State of the Beta (4) =
on the
`M, M
on the   1.0 mean for mean for (8)=(2)(6)(7
Economy Prob. Market Var
(2)*(3) Market
M .007425
(6)=(2)*(5) the stock the market )
Boom 25.0% 20.0% 0.05 20.0% 0.05 10.5% 10.5% 0.002756
Normal 50.0% 11.0% 0.055 11.0% 0.055 1.5% 1.5% 0.000113
Recession 25.0% -4.0% -0.01 -4.0% -0.01 -13.5% -13.5% 0.004556
E(k M) = 9.5% E(kM) = 9.5% Covariance = 0.007425

Since
Sincethe
thevariance
varianceof
ofthe
thereturns
returnsononthe
themarket
marketisis==.007425
.007425…the
…thebeta
betafor
for
the market is indeed equal to 1.0 !!!
the market is indeed equal to 1.0 !!!

9 - 75
 Proving the Beta of Market = 1
If you now place the covariance of the market with itself value in the
beta formula you get:

Cov MM .007425
Beta    1.0
Var(R M ) .007425

The beta coefficient of the market will always be

1.0 because you are measuring the market returns
against market returns.

9 - 76
How Do We use Expected and Required Rates of
Return?
Once you have estimated the expected and required rates of return, you can plot them on
the SML and see if the stock is under or overpriced.

% Return
E(Rs) = 5.0%

R(ks) = 4.76%
SML
E(kM)= 4.2%

Risk-free Rate = 3%

BM= 1.0 Bs = 1.464

Since E(r)>R(r) the stock is underpriced.

9 - 77
 How Do We use Expected and Required Rates of
Return?
• The stock is fairly priced if the expected return = the required return.
• This is what we would expect to see ‘normally’ or most of the time in an efficient market where securities
are properly priced.

% Return

E(Rs) = R(Rs) 4.76%

SML
E(RM)= 4.2%

Risk-free Rate = 3%

BM= 1.0 BS = 1.464

9 - 78
 Use of the Forecast Beta
• We can use the forecast beta, together with an estimate of the risk-free rate and
the market premium for risk to calculate the investor’s required return on the
stock using the CAPM:

Required Return  RF  βi [E (k M )  RF]

• This is a ‘market-determined’ return based on the current risk-free rate (RF) as

measured by the 91-day, government of Canada T-bill yield, and a current estimate
of the market premium for risk (kM – RF)

9 - 79
 Conclusions

• Analysts can make estimates or forecasts for the returns on stock

and returns on the market portfolio.
• Those forecasts can be analyzed to estimate the beta coefficient for
the stock.
• The required return on a stock can then be calculated using the
CAPM – but you will need the stock’s beta coefficient, the expected
return on the market portfolio and the risk-free rate.
• The required return is then using in Dividend Discount Models to
estimate the ‘intrinsic value’ (inherent worth) of the stock.

9 - 80
 Calculating the Beta using
Trailing Holding Period Returns
APPENDIX 2

81
The Regression Approach to Measuring the Beta
• You need to gather historical data about the stock and the market
• You can use annual data, monthly data, weekly data or daily data.
However, monthly holding period returns are most commonly used.
• Daily data is too ‘noisy’ (short-term random volatility)
• Annual data will extend too far back in to time
• You need at least thirty (30) observations of historical data.
• Hopefully, the period over which you study the historical returns of the
stock is representative of the normal condition of the firm and its
relationship to the market.
• If the firm has changed fundamentally since these data were produced
(for example, the firm may have merged with another firm or have
divested itself of a major subsidiary) there is good reason to believe
that future returns will not reflect the past…and this approach to beta
estimation SHOULD NOT be used….rather, use the ex ante approach.
9 - 82
 Historical Beta Estimation
The Approach Used to Create the Characteristic Line

In this example, we have regressed the quarterly returns on the stock against the
quarterly returns of a surrogate for the market (TSE 300 total return composite
index) and then using Excel…used the charting feature to plot the historical
points and add a regression trend line.
The ‘cloud’ of plotted points
Period HPR(Stock)
represents HPR(TSE
‘diversifiable 300)
or company C h a r a c t e r is t ic Lin e ( R e g r e s s io n )
2006.4
specific’ risk-4.0% 1.2%
in the securities returns 30.0%
2006.3 -16.0%
that can be eliminated from-7.0%
a portfolio 25.0%
2006.2 32.0%
through diversification. 12.0%
Since 20.0%
2006.1company-specific
16.0% risk can
8.0%be

Returns on Stock
15.0%
eliminated,-22.0%
2005.4 investors don’t require
-11.0%
compensation
2005.3 15.0%for it according
16.0% to 10.0%
2005.2Markowitz Portfolio Theory.
28.0% 13.0% 5.0%
2005.1 19.0% 7.0% 0.0%
2004.4 -16.0% -4.0% -40.0% -20.0% -5.0%0.0% 20.0% 40.0%
The regression line is a line of ‘best
2004.3 8.0% 16.0%
fit’ that describes the inherent -10.0%
2004.2 -3.0% -11.0%
relationship between the returns on -15.0%
2004.1 34.0% 25.0%
the stock and the returns on the
Returns on TSE 300
market. The slope is the beta
coefficient.

9 - 83
 Characteristic Line
• The characteristic line is a regression line that represents the relationship
between the returns on the stock and the returns on the market over a past
period of time. (It will be used to forecast the future, assuming the future will be
similar to the past.)

• The slope of the Characteristic Line is the Beta Coefficient.

• The degree to which the characteristic line explains the variability in the
dependent variable (returns on the stock) is measured by the coefficient of
determination. (also known as the R2 (r-squared or coefficient of determination)).

• If the coefficient of determination equals 1.00, this would mean that all of the
points of observation would lie on the line. This would mean that the
characteristic line would explain 100% of the variability of the dependent variable.

• The alpha is the vertical intercept of the regression (characteristic line). Many
stock analysts search out stocks with high alphas.

9 - 84
 Low R 2

• An R2 that approaches 0.00 (or 0%) indicates that the characteristic

(regression) line explains virtually none of the variability in the dependent
variable.
• This means that virtually of the risk of the security is ‘company-specific’.
• This also means that the regression model has virtually no predictive
ability.
• In this case, you should use other approaches to value the stock…do not
use the estimated beta coefficient.

(See the following slide for an illustration of a low r-square)

9 - 85
Characteristic Line for Imperial
Tobacco
An Example of Volatility that is Primarily Company-Specific
Characteristic
Returns on
Line for Imperial
Imperial
Tobacco
Tobacco %
• High alpha
• High alpha
• R-square is very
• R-square is very
low ≈ 0.02
low ≈ 0.02
• Beta is largely
• Beta is largely
irrelevant
irrelevant

Returns on
the Market %
(S&P TSX)

9 - 86
 High R 2

• An R2 that approaches 1.00 (or 100%) indicates that the characteristic

(regression) line explains virtually all of the variability in the dependent
variable.
• This means that virtually of the risk of the security is ‘systematic’.
• This also means that the regression model has a strong predictive ability.
… if you can predict what the market will do…then you can predict the
returns on the stock itself with a great deal of accuracy.

9 - 87
Characteristic Line General Motors
A Positive Beta with Predictive Power

Characteristic
Returns on
Line for GM
General
Motors % (high R2)
• • Positive
Positivealpha
alpha
• • R-square
R-squareisis
very
veryhigh
high≈≈0.9
0.9
• • Beta
Betaisispositive
positive
and
and closeto
close to1.0
1.0

Returns on
the Market %
(S&P TSX)

9 - 88
An Unusual Characteristic Line
A Negative Beta with Predictive Power

Returns on a Characteristic Line for a stock

Stock % that will provide excellent
portfolio diversification
• • Positive
Positivealpha
alpha
(high R2) • • R-square
R-squareisis
very
veryhigh
high
• • Beta
Betaisisnegative
negative
<0.0
<0.0 and>>-1.0
and -1.0

Returns on
the Market %
(S&P TSX)

9 - 89
 Diversifiable Risk
(Non-systematic Risk)

• Volatility in a security’s returns caused by company-specific factors

(both positive and negative) such as:
– a single company strike
– a spectacular innovation discovered through the company’s R&D program
– equipment failure for that one company
– management competence or management incompetence for that particular
firm
– a jet carrying the senior management team of the firm crashes (this could be
either a positive or negative event, depending on the competence of the
management team)
– the patented formula for a new drug discovered by the firm.
• Obviously, diversifiable risk is that unique factor that influences only the
one firm.

9 - 90
 OK – lets go back and look at raw data gathering
and data normalization

• A common source for stock of information is Yahoo.com

• You will also need to go to the library a use the TSX Review (a monthly
periodical) – to obtain:
– Number of shares outstanding for the firm each month
– Ending values for the total return composite index (surrogate for the market)
• You want data for at least 30 months.
• For each month you will need:
– Ending stock price
– Number of shares outstanding for the stock
– Dividend per share paid during the month for the stock
– Ending value of the market indicator series you plan to use (ie. TSE 300 total
return composite index)

9 - 91
Demonstration Through Example

The following slides will be based on

Alcan Aluminum (AL.TO)

92
Five Year Stock Price Chart for AL.TO

9 - 93
 Spreadsheet Data From Yahoo
Process:

– Go to http://ca.finance.yahoo.com
– Use the symbol lookup function to search for the
company you are interested in studying.
– Use the historical quotes button…and get 30
months of historical data.
– Use the download in spreadsheet format feature
to save the data to your hard drive.
9 - 94
 Spreadsheet Data From Yahoo
Alcan Example

The raw downloaded data should look like this:

Date Open High Low Close Volume

01-May-02 57.46 62.39 56.61 59.22 753874
01-Apr-02 62.9 63.61 56.25 57.9 879210
01-Mar-02 64.9 66.81 61.68 63.03 974368
01-Feb-02 61.65 65.67 58.75 64.86 836373
02-Jan-02 57.15 62.37 54.93 61.85 989030
03-Dec-01 56.6 60.49 55.2 57.15 833280
01-Nov-01 49 58.02 47.08 56.69 779509

9 - 95
 Spreadsheet Data From Yahoo
Alcan Example

The raw downloaded data should look like this:

Date Open High Low Close Volume

01-May-02 57.46 62.39 56.61 59.22 753874
01-Apr-02 62.9 63.61 56.25 57.9 879210

Volume
Volumeofof
Opening
Openingprice
priceper
pershare,
share,thethe trading
tradingdone
done
The
Theday, ininthe
day, highest
highestprice
priceper
pershare
shareduring
duringthethe thestock
stockon on
month
monthand the
and month,
month,the
thelowest
lowestprice
priceper
pershare
share theTSETSEininthe the
year month
year achieved
achievedduring
duringthe
themonth
monthand andthe
the monthinin
closing
closingprice
priceper
pershare
shareatatthe
theend
end numbers
numbersofof
ofofthe
themonth
month board
boardlots
9lots
- 96
 Spreadsheet Data From Yahoo
Alcan Example

From Yahoo, the only information you can use is the closing price per
share and the date. Just delete the other columns.

Date Close
01-May-02 59.22
01-Apr-02 57.9
01-Mar-02 63.03
01-Feb-02 64.86
02-Jan-02 61.85

9 - 97
 Acquiring the Additional Information You Need Alcan
Example

In addition to the closing price of the stock on a per share basis, you will need
to find out how many shares were outstanding at the end of the month
and whether any dividends were paid during the month.

You will also want to find the end-of-the-month value of the S&P/TSX Total
Return Composite Index (look in the green pages of the TSX Review)

You can find all of this in The TSX Review periodical.

9 - 98
 Raw Company Data
Alcan Example

Closing Price Cash

Issued for Alcan Dividends
Date Capital AL.TO per Share
01-May-02 321,400,589 $59.22 $0.00
01-Apr-02 321,400,589 $57.90 $0.15
01-Mar-02 321,400,589 $63.03 $0.00
01-Feb-02 321,400,589 $64.86 $0.00
02-Jan-02 160,700,295 $123.70 $0.30
01-Dec-01 160,700,295 $119.30 $0.00

Number of shares doubled and share price fell by half between

January and February 2002 – this is indicative of a 2 for 1 stock split.
9 - 99
 Normalizing the Raw Company Data
Alcan Example

Closing
Price for Cash
Issued Alcan Dividends Adjustment Normalized Normalized
Date Capital AL.TO per Share Factor Stock Price Dividend
01-May-02 321,400,589 $59.22 $0.00 1.00 $59.22 $0.00
01-Apr-02 321,400,589 $57.90 $0.15 1.00 $57.90 $0.15
01-Mar-02 321,400,589 $63.03 $0.00 1.00 $63.03 $0.00
01-Feb-02 321,400,589 $64.86 $0.00 1.00 $64.86 $0.00
02-Jan-02 160,700,295 $123.70 $0.30 0.50 $61.85 $0.15
01-Dec-01 145,000,500 $111.40 $0.00 0.45 $50.26 $0.00

The adjustment factor is just the value in the issued

capital cell divided by 321,400,589.

9 - 100
 Calculating the HPR on the stock from the
Normalized Data

Normalized Normalized ( P1  P0 )  D1
HPR 
Date Stock Price Dividend HPR P0
$59.22 - $57.90  $0.00
01-May-02 $59.22 $0.00 2.28% 
$57.90
01-Apr-02 $57.90 $0.15 -7.90%
 2.28%
01-Mar-02 $63.03 $0.00 -2.82%
01-Feb-02 $64.86 $0.00 4.87%
02-Jan-02 $61.85 $0.15 23.36%
01-Dec-01 $50.26 $0.00

Use $59.22 as the ending price, $57.90 as the

beginning price and during the month of May, no
dividend was declared.

9 - 101
 Now Put the data from the S&P/TSX Total Return
Composite Index in

Ending
Normalized Normalized TSX
Date Stock Price Dividend HPR Value
01-May-02 $59.22 $0.00 2.28% 16911.33
01-Apr-02 $57.90 $0.15 -7.90% 16903.36
01-Mar-02 $63.03 $0.00 -2.82% 17308.41
01-Feb-02 $64.86 $0.00 4.87% 16801.82
02-Jan-02 $61.85 $0.15 23.36% 16908.11
01-Dec-01 $50.26 $0.00 16881.75

You
Youwill
willfind
findthe
theTotal
TotalReturn
ReturnS&P/TSX
S&P/TSXComposite
Composite
Index
Indexvalues
valuesininTSX
TSXReview
Reviewfound
foundininthe
thelibrary.
library.
9 - 102
 Now Calculate the HPR on the Market Index

( P1  P0 )
HPR 
P0
16,911.33 - 16,903.36
 Ending
16,903.36
Normalized Normalized
 0.05%
TSX HPR on
Date Stock Price Dividend HPR Value the TSX
01-May-02 $59.22 $0.00 2.28% 16911.33 0.05%
01-Apr-02 $57.90 $0.15 -7.90% 16903.36 -2.34%
01-Mar-02 $63.03 $0.00 -2.82% 17308.41 3.02%
01-Feb-02 $64.86 $0.00 4.87% 16801.82 -0.63%
02-Jan-02 $61.85 $0.15 23.36% 16908.11 0.16%
01-Dec-01 $50.26 $0.00 16881.75

Again,
Again,you
yousimply
simplyuse
usethe
theHPR
HPRformula
formulausing
usingthe
the
ending
endingvalues
valuesfor
forthe
thetotal
totalreturn
returncomposite
compositeindex.
index.
9 - 103
 Regression In Excel

• If you haven’t already…go to the tools menu…

down to add-ins and check off the VBA Analysis
Pac
• When you go back to the tools menu, you
should now find the Data Analysis bar, under
that find regression, define your dependent
and independent variable ranges, your output
range and run the regression.
9 - 104
 Regression
Defining the Data Ranges

Ending
Normalized Normalized TSX HPR on
Date Stock Price Dividend HPR Value the TSX
01-May-02 $59.22 $0.00 2.28% 16911.33 0.05%
01-Apr-02 $57.90 $0.15 -7.90% 16903.36 -2.34%
01-Mar-02 $63.03 $0.00 -2.82% 17308.41 3.02%
01-Feb-02 $64.86 $0.00 4.87% 16801.82 -0.63%
02-Jan-02 $61.85 $0.15 23.36% 16908.11 0.16%
01-Dec-01 $50.26 $0.00 16881.75

The
Thedependent
independent
variable
independent
dependent variable
variableisisisthe
variable the
isthe returns
the returns
returnson
returnson
on the
on the
the Stock.
the Market.
Stock.
Market.

9 - 105
Now Use the Regression Function in Excel to regress
the returns of the stock against the returns of the
market
SUMMARY OUTPUT

Regression Statistics R-square is the

Multiple R 0.05300947
R Square 0.00281 coefficient of
Adjusted R Square
Standard Error
-0.2464875
5.79609628
determination =
Observations 6 0.0028=.3%
ANOVA
df SS MS F Significance F
Regression 1 0.3786694 0.37866937 0.011271689 0.920560274
Residual 4 134.37893 33.5947321
Total 5 134.7576

CoefficientsStandard Error t Stat P-value Lower 95% Upper 95% Lower 95.0%Upper 95.0%
Intercept 59.3420816 2.8980481 20.4765686 3.3593E-05 51.29579335 67.38836984 51.2957934 67.38837
X Variable 1 3.55278937 33.463777 0.10616821 0.920560274 -89.35774428 96.46332302 -89.3577443 96.46332

Beta The alpha is the

Coefficient vertical intercept.
is the X-
Variable 1

9 - 106
 Finalize Your Chart
Alcan Example

• You can use the charting feature in Excel to create a scatter plot of
the points and to put a line of best fit (the characteristic line)
through the points.
• In Excel, you can edit the chart after it is created by placing the
cursor over the chart and ‘right-clicking’ your mouse.
• In this edit mode, you can ask it to add a trendline (regression line)
• Finally, you will want to interpret the Beta (X-coefficient) the alpha
(vertical intercept) and the coefficient of determination.

9 - 107
 The Beta
Alcan Example

• Obviously the beta (X-coefficient) can simply

be read from the regression output.
– In this case it was 3.56 making Alcan’s returns
more than 3 times as volatile as the market as a
whole.
– Of course, in this simple example with only 5
observations, you wouldn’t want to draw any
serious conclusions from this estimate.
9 - 108
 Copyright
Copyright © 2007 John Wiley & Sons Canada, Ltd. All rights reserved.
Reproduction or translation of this work beyond that permitted by Access
Copyright (the Canadian copyright licensing agency) is unlawful. Requests for
further information should be addressed to the Permissions Department, John
Wiley & Sons Canada, Ltd. The purchaser may make back-up copies for his or
her own use only and not for distribution or resale. The author and the publisher
assume no responsibility for errors, omissions, or damages caused by the use of
these files or programs or from the use of the information contained herein.

19 - 109