Lecture 5

Capital Asset Pricing Model

Risk and return
• High return implies high risk and vice versa.
• Risk in finance is measured by volatility or
uncertainty of outcome.
• More specifically the risk of individual assets
is measured by standard deviation of return.
• Standard Deviation is of course a statistical
measure of volatility or uncertainty.
 Portfolio diversification
• It is possible to improve return relative to risk by
holding two or more assets in a portfolio.
• In fact, as a general rule, provided the assets
involved are sufficiently diverse, it is possible to
show that volatility risk, as measured by
standard deviation of portfolio return, will be
reduced as additional assets are added.
Portfolio diversification
• This phenomenon of risk reduction has a limit,
which is reached when the portfolio contains
between around 20 assets. This limit is known
as the limit of diversification
• It is normally assumed that the rational risk-
averse investor would hold such a diversified
portfolio, as it is possible to improve risk without
reduction in return
 Portfolio risk is reduced by
diversification.

• Additional assets when added to the portfolio reduce

portfolio risk, up to a limit, beyond which no further
reduction is possible.
Diversifiable vs. Undiversifiable Risk
(a.k.a. Unsystematic vs. Systematic Risk)
• The market does not reward diversifiable risk,
but it does reward undiversifiable risk.
• By diversifying the portfolio, the investor is not
exposed to the full standard deviation risk of an
individual asset.
• The residual risk that remains once a portfolio
is fully diversified is known as Undiversifiable
Risk
Beta (β) - a measure of Systematic risk

• β is the risk that arises from the fact that all

firms face uncertainty because of
macroeconomic factors beyond their control,
and in particular the business cycle.
• We typically measure this risk by comparison of
stock volatility with the volatility of the market
portfolio, a theoretical portfolio consisting of all
risky assets.
• In practical terms, this is represented by a stock
market index such as the DJI, FTSE 100 etc.
Returns volatility over the business cycle

Annual returns data for Coke, IBM and the Dow Jones Index
Learning Activity
• In layman’s terms how would you describe
volatility? Why is it important?
• How would you measure volatility in a very
simple way?
• How would you measure it in a more
sophisticated way?
Beta (β) - a measure of Systematic risk
• We need a measure of risk that looks at volatility
of individual stock return relative to the market
return. NOT Absolute volatility.
• An obvious candidate is the covariance
measure, which was introduced in portfolio
theory.
• However, it is difficult to compare one set of
covariances with another as they don’t come in
standardised units.
Beta (β) - a measure of Systematic risk

• Hence in order to compare volatility of return

over the business cycle with the market index,
we use beta.
• This is essentially still the covariance of the
individual stock with the market index, but
which is then standardised by dividing the
variance of the market return.
A Stock Index – The Dow Jones
Covariance over the Business Cycle
• The previous slide is a simplified version of how
stock returns vary over the cycle relative to the
index.
• We can use this data to calculate covariances.
• Here are the deviations from the mean.
Deviations from the Mean
Calculating Covariance

1. For each stock return take the deviation from

the mean.
2. Multiply one set of deviations by the other set.
3. Multiply each co-deviation by probability.
4. Take the sum of (3) to get the covariance.
Covariances
Learning Activity
• How do the steps involved in calculating
variance compare with the steps involved in
calculating covariance?
• Calculate the variance of the DJ index.
Variance is similar to Covariance
Beta (β) - a measure of Systematic risk
We are finally in a position to calculate Beta.
• BCoke = Cov(RCoke, RDJI)/Var(RDJI)
= 0.6%/ 0.64%
= 0.9375 is approx 0.94

• BIBM = Cov(RIBM, RDJI)/Var(RDJI)

= 1.2%/ 0.64%
= 1.875 is approx. 1.88
 Beta (β) - a measure of Systematic risk
• Beta of Coke is around the same as market beta.
• By definition βmarket = βDJI = 1
• Beta of IBM is much higher
• This means that IBM has much more market risk
• Consistent with inherent volatility of IT industry
Beta (β) - a measure of Systematic risk
• Once we know Beta, we are able to calculate
the return we should expect for a particular
securities using the Capital Asset Pricing
Model (CAPM).
• The CAPM relates individual share return to the
following three variables:
– Risk, as measured by beta (βi)
– The risk-free rate (rf)
– The rate of return on the market (r m)
Beta (β) - a measure of Systematic risk
• The equation that links all these together is
known as the security market line (SML). The
SML is as follows:
E(ri) = rf + βi(E(rm) – rf)
 Security Market Line

The security market line of the CAPM – beta versus return

Beta (β) - a measure of Systematic risk
• The SML is just a straight line relating return on
an individual security to the beta of that
security.
• The intercept of the line will be at r f, the risk-
free rate, as by definition this is where risk and
beta are equal to zero.
• The other point that allows us to find the line on
a graph is where β=1.
Beta (β) - a measure of Systematic risk

• This is the value of beta for the market portfolio,

βM
• In our case βM = βDJI as this is our chosen index
• The slope of the line is equal to the term in
parentheses, (E(rm) – rf), which is known as the
market risk premium.
Learning activity
1. Sketch the SML diagram.
2. If Rf is 2% and E(Rm) is 12% what is
E(RCoke) and E(RIBM), given their betas?
3. Add Coke and IBM to your sketch.
Equilibrium and arbitrage
• Your answers to part 3 of the previous
questions should have been 11.4% and 20.8%.
• But what if Coke over-achieved at 15% while
IBM ‘only’ yielded a similar 15%?
• In order to imagine the market’s response, we
have to entertain a fiction.
Equilibrium and Arbitrage
• Supposing the future value of a share is
known with certainty.
• Ignoring dividends, the overall % return is
• (Pt+1 – Pt)/Pt
• If Pt+1is fixed, the only thing that can vary
is the current price.
What happens to share price of Coke if
it’s undervalued?
• For Coke,
– suppose Pt+1 = 115, Pt = 100
• Learning Activity
– What happens next in an efficient
market?
If Coke is undervalued
• The return on coke is 15%. This is higher
than justified by the Beta risk (11.4%).
• Coke shares are bought, driving up the price.
• Eventually the process comes to an end
when Pt= 115/1.114 = 103.2316
• Check that this price would indeed yield
11.4%.
• So price rises from 100 to 103 (approx.)
What happens to share price of IBM if
it’s overvalued?
• For IBM,
– suppose Pt+1 = 115, Pt = 100
• Learning Activity
– What happens next in an efficient
market?
If IBM is overvalued
• The return on IBM is 15%. This is lower than
justified by the Beta risk (20.8%).
• IBM shares are sold, driving down the price.
• Eventually the process comes to an end
when Pt = 115/1.208 = 95.2
• Check that this price would indeed yield
20.8%
• So the price falls from 100 to 95.2 (approx.)
Key Points
• Risk in finance is measured by volatility or
uncertainty of outcome.
• High return implies high risk and vice versa
• The risk of individual assets is measured by
standard deviation of return.
• Provided the assets involved are sufficiently
diverse, volatility risk, as measured by standard
deviation of portfolio return, will be reduced as
additional assets are added.
Key Points
• Normally assumed that the rational risk-averse
investor would hold such a diversified portfolio.
• The market does not reward diversifiable risk,
but it does reward undiversifiable risk.
Key Points
• In order to examine volatility of return over the
business cycle compared to the market index,
we use beta.
• This is 1) essentially the covariance of the
individual stock return with the market index
which is then
• 2) Standardised by dividing by the covariance
of the market return with itself.
Key Points
• The expected return on a share or equity can
then be calculated using the SML
• The SML is as follows:
– E(ri) = rf + βi(E(rm) – rf)
• Expected return depends on the RFA plus
Stock Beta times the Market Risk Premium
Key Points
• CAPM is based on the assumptions of perfect
competition from economics.
– Perfect Information
– Frictionless markets
– Insignificant transaction costs etc.
• CAPM and Equilibrium - If risk and return are out of
line with the equilibrium suggested by the SML,
they will be brought back in line by arbitrage
mechanism