Title: Ch 11 Return and Risk: The Capital Asset Pricing Model CAPM
1Ch 11 Return and Risk The Capital Asset Pricing
Model (CAPM)
- 1 Expected Return and Variance for a single
asset - Portfolios
- Expected return and variance for a portfolio
- Efficient set with two assets
- Efficient set with N assets
- Riskless borrowing and lending
- Market equilibrium
- Diversification and Portfolio Risk
- Diversification with N assets
- Systematic Risk and Beta
- 4 The Security Market Line and CAPM
- 5 Summary
2Main Idea
- What is an appropriate measure of risk?
- If you hold single asset, standard deviation of
the asset is a good measure of the risk. - If you hold a widely diversified portfolio, the
beta of the asset is a good measure of the risk
of the asset. - Average return of the asset is a good measure of
the expected return in both cases. -
31. Expected Return and Variance
- Goal To derive a relation between return and
risk in the form of - E(Ri) Rf ?i E(Rm) Rf Â
- where
- Rf ? risk free rate (e.g., T-bill rate)
- Rm ? return on market portfolio
- (e.g., a value-weighted portfolio of all
TSE stocks) - ?i Measure of risk of asset i
- ? Cov (asseti, market portfolio) / Var
(market portfolio) - In order to measure and develop models for the
relation between risk and return, we need some
formal statistical measures.
41.1 Calculating the Expected Return
- Example 1 Average of two scores 80 and 60
(8060)/2 70. - Example 2 You feel youd get 80 and 60 in a
finance course with equal probability of 0.5 and
0.5. What do you expect to get?
- In Example 1, it is as if equal probabilities
are assigned to calculate the mean mean
(0.580) (0.560) (8060)/2 70. - We can generalize mean sum of (probpossible
outcome), i.e., - E(x) sum (Pi
xi) ? ?I (Pi xi)
5Example Calculating the Expected Return
- Example 3 You feel youd get 80 and 60 in a
finance course with probability of 0.4 and 0.6.
What do you expect to get? -
- Qu What are E(x2), E(x3 - 3x), or in general,
E anything ? - E(x2) sum (Pi xi2)
- E(x3 - 3x) sum ( Pi (xi3 - 3xi) )
- E (anything) sum (Pi possible outcome i of
anything), i.e., - E f(x) sum Pi f(xi) ? ?I Pi f(xi)
61.2 Calculating variance
- Example 1 You expect to get 80 and 60 in a
finance course with probability of 0.4 and 0.6.
What is variance of your score? - variance(x) ? sx2 ? E (x - E(x))2 . Why
squared? - Recall E anything sum (Pi possible outcome
i of anything). - sx2 ? E (x - E(x) )2 sum Pi possible
outcome of (xi - E(x))2
71.2 Calculating variance
- Why standard deviation?
- Variance isn't in the same units as the
mean--it's in (unit)2. It is often useful to
work with standard deviation which is in the
units as the mean. - 1.3 Calculating covariance and correlation
- A measure of how random variables move together
is covariance. - If we have two random variables, X and Y, their
covariance is defined as cov (x, y) ? E (x -
E(x)) (y - E(y))
81.3 Calculating covariance and correlation
- Example 1 You feel youd get 80 and 60 in a
finance and 70 and 90 in an accounting course
with probability of 0.4 and 0.6. Find out the
mean, variance and standard deviation of finance
(x) and accounting (y) scores. - E(x) 68 sx2 96 sx 9.8
- E(y) 82 sy2 96 sy 9.8
- What is the covariance of your finance and
accounting scores? - Recall E anything sum (Pi possible outcome
i of anything). - cov (x, y) ? E (x - E(x)) (y - E(y))
- sum Pi possible outcome of
(xi - E(x)) (yi - E(y)) - ?I Pi (xi - E(x)) (yi -
E(y))
91.2 Calculating covariance and correlation
- cov (x, y) ?I Pi (xi - E(x)) (yi - E(y))
corr (x, y) cov (x, y) / (sxsy) . Why using
correlation? corr (x, y) always lies between -1
and 1. corr (x, y) 1 perfectly positive
correlation corr (x, y) -1 perfectly
negative correlation corr (x, y) 0
independent
10Qu 12.7 Calculation of mean, variance covariance
Using the following returns, calculate the
average returns, the variances, the standard
deviations, covariance and correlation for stocks
X and Y.
When you use actual data (i.e., no probabilities)
and calculate moments, follow these rules 1.
E(X) sum all xi and divide by T (no of
observations) 2. var E (X - E(X))2 sum
all (xi - E(X))2 divide by (T-1) 3. cov E
(X-E(X))(Y-E(Y)) sum all
(xi-E(X))(yi-E(Y)) and divide by (T-1)
11Qu Calculation of mean, variance covariance
12Qu. Calculation of mean, variance covariance
with Probability distribution
13Evidence on Covariance and Correlation
- (1) Stock returns are serially uncorrelated.
- If stock returns are high one year then, you
can't use this information to predict whether
returns in the subsequent year will be high or
low. This evidence is important to market
efficiency. - (2) Most stocks are positively correlated to the
market portfolio important evidence for our
discussion of asset pricing model. - market portfolio a value-weighted portfolio of
all the stocks in the economy (or as a proxy, all
stocks on the NYSE or TSE).
14Statistical Review Conclusion
- The main thing is to have an intuitive
understanding of what these statistics mean. - Arithmetic mean is a measure of how much you
can expect to receive if you hold a stock for a
year. - The variance and standard deviation are
measures of how variable the returns are likely
to be. The higher the variance or standard
deviation the greater the variation. - Covariance and correlation are measures of
whether two variables move together or in
opposite directions. - Move together positive
- Move in opposite directions negative
- Independent zero.
152. Diversification
- 2.1 Mean and variance of portfolio
- Suppose we have 1,000 to invest, and there are
two risky assets  - Â
- We could invest it all in asset 1 or all in asset
2. However, we may do better off by taking a
combination of asset 1 and asset 2, i.e.,
diversification may provide better result than by
taking 1 or 2 alone. Let us see - Suppose correlation between asset 1 and 2s
return is 0.5. Let x1 be the proportion of our
wealth invested in asset 1. (What is proportion
of wealth invested asset 2?)
162.1 Mean and variance of a portfolio
- Mean return and variance of a portfolio with two
assets - When x10, we invest entire wealth in asset 2.
The portfolio has a mean return of 8 and ? of
0.05. - When x11, we invest entire wealth in asset 1.
The portfolio has a mean return of 5 and ? of
0.03. -
- Â
172.1 Mean and variance of a portfolio
- What will happen in-between?
- E(Rp) x1 E(R1) (1 - x1) E(R2) (1)
- ?p2 x12 ?12 (1 - x1)2 ?22 2x1(1 - x1) cov
(R1, R2) (2) - Note Recall ?12 ? Corr (R1, R2) cov (R1,
R2)/(?1 ?2). - Sometimes, we use ?12 ?1 ?2 instead of cov (R1,
R2). - Suppose that x10.1. We can calculate the mean
return and ? of the portfolio using equations (1)
and (2). We can do the same calculation for
x10.2, 0.3,.,x11, and fill up the following
table. - Using these results, we can draw the following
chart
182.1 Mean and variance of a portfolio
192.1 Mean and variance of a portfolio
202.2 Relation between ? and investment
opportunity set
- Figure 1 shows investment opportunity set,
i.e., mean and standard deviations for all
possible combinations of assets 1 and 2. - For Figure 1, we assume corr 0.5. Let ?
denote correlation. Correlation lies between -1
and 1. How investment does investment
opportunity set change when ? changes?
212.2 Relation between ? and investment
opportunity set
Result The lower the ?, the greater the benefit
from diversification.
222.3 Efficient frontier with N assets
C
B minimum variance portfolio
mean
ABC minimum variance frontier
B
BC efficient frontier
Shaded area investment opportunity set
A
standard deviation
Diversification with N assets covariance among
assets affects the variance of a portfolio with N
assets. We will look at this issue at section
3.1.
232.4 Riskless Borrowing and Lending
- With a risk-free asset available and the
efficient frontier identified, an investor
chooses the capital allocation line with the
steepest slope.
CML
return
efficient frontier
rf
?P
24- Note that the risk-return relation of a porfolio
of risk free asset and a risky asset Q is
represented by a straight line between the risk
free rate on y-axis and the risk asset Q. - (proof optional material)
252.4 Riskless Borrowing and Lending
- With the capital allocation line identified, an
investor chooses a point along the linesome
combination of the risk-free asset and a risky
portfolio M.
CML
return
efficient frontier
M
rf
?P
26The Separation Property
-
- The Separation Property states that investors can
separate their risk aversion from their choice of
the risky portfolio. - Implications portfolio choice can be separated
into two tasks (1) determine the optimal risky
portfolio, and (2) selecting a point on the CML.
CML
return
efficient frontier
M
rf
?P
27Market equilibrium
- In a world with homogeneous expectations, the
portfolio of risky asset is the same for all
investors. - In capital market equilibrium, demand equal
supply. - The portfolio of risky asset in equilibrium is
called the market portfolio. - market portfolio A portfolio of all stocks in
the market. Portfolio weight of stock i is
equal to the proportion of stock is market value
to the market value of all stocks in the market
portfolio. Â - If total value of stock 1 is 10 billion and the
total value of the market portfolio is 1,000
billion. Then x110/1,0001. We denote the
market portfolio by M.
283. Diversification and portfolio risk
- We know how investors behave in a world with risk
free asset, and with homogeneous expectations. - Investors will hold the same risky portfolio M,
and risk free asset in their portfolio. The
proportions of the risky and risk free assets are
dependent on the investors risk aversion. - The risky portfolio M is the market portfolio.
- Next Qu what is the risk-return relation among
assets in portfolio M?
293. Diversification and portfolio risk
3.1. diversification with N assets Qu how much
risk can we eliminate by diversification?
303.1 diversification
- diversification diversification eliminates
some, but not all of the risk. - Systematic risk risk that influences overall
stock market, such as GNP, or interest rate. It
can't be diversified away - Unsystematic risk risk that influences
single industries, or individual firms such RD
results or a change in CEO. - A stock thus has two components of risk
systematic and unsystematic risk. One can
eliminate most of unsystematic risk with about
15-30 stocks.
313.2 A principle
- A principle The reward for bearing risk
depends only on the systematic risk of an
investment. - The market does not reward bearing unsystematic
risk, since these risk can be diversified away in
a reasonably large portfolio. - Hence it is systematic risk which is
important. Suppose stock A and B have the same
expected return. A has a higher variance, but
lower systematic risk than stock B. The stock A
may be much more desirable than stock B with a
lower variance.
323.3 Systematic risk and beta (ß)
- So, it is systematic risk, not the total risk
of a stock, which is important. We can say that
a stock has a high risk if it has large
systematic risk. But how do we know that a stock
has large systematic risk? I.E., how do we
measure systematic risk of a stock? - Answer Beta (ß),
- where ?i ? Cov (Ri, Rm) / Var (Rm), and Rm is
return on the market portfolio.
333.3 Systematic risk and beta (ß)
- Basic intuition about ß
- ß measures how much systematic risk a stock has
relative to the market portfolio (or an average
asset). By definition, ß of the market portfolio
is 1. - Recall that systematic risk influences
overall stock market. If a stocks return
co-moves a lot with overall stock market, this
stock has high systematic risk. That is why you
see Cov (Ri, Rm) in numerator of definition of ß.
34- Interpretation of ß
- (1) It is reasonable to say that the market
portfolio has (almost) systematic risk only,
since diversification eliminates (nearly) all of
unsystematic risk. - Consider the extent to which the variance of the
market portfolio change if we change the amount
of the stock in the portfolio. - That is ß, i.e., the contribution of the stock to
the variance of the market portfolio (or in
mathematical term (? ?m2 / ?xi)).
35- ß measures sensitivity of a stocks return to
movements in overall market. By definition, ßm
Cov (Rm, Rm) / Var (Rm) 1. That is, ß of
market portfolio is 1. - Thus stocks with a ß gt 1 tend to be sensitive to
movements in the market--they magnify these
movements. Stocks with a ß lt 1 are relatively
insensitive to movements in the market.
363.3 Systematic risk and beta (ß)..
- High ß stock Low ß stock
- ß regression coefficient
373.3 Systematic risk and beta (ß)..
Beta Coefficients for Selected Companies
Canadian Co. Beta Bank of Nova
Scotia 0.65 Bombardier 0.71 Canadian
Utilities 0.50 C-MAC Industries 1.85 Investors
Group 1.22 Maple Leaf Foods 0.83 Nortel
Networks 1.61 Rogers Communication 1.26
- U.S. Co Beta American
Electric Power .65 - Exxon .80
- IBM .95
- Wal-Mart 1.15
- General Motors 1.05
- Harley-Davidson 1.20
- Papa Johns 1.45
- America Online 1.65
Source (Canadian) Scotia Capital markets and
(US) Value Line Investment Survey, May 8, 1998.
383.3 Systematic risk and beta (ß)..
- Portfolio beta is equal to the weighted average
of individual stocks ß. - Example
- Amount PortfolioStock Invested
Weights Beta - (1) (2) (3) (4) (3) ? (4)
- Haskell Mfg. 6,000 50 0.90 0.450
- Cleaver, Inc. 4,000 33 1.10 0.367
- Rutherford Co 2,000 17 1.30 0.217
- Portfolio 12,000 100 1.034
394. The security market line
- 4.1 The security market line
- In equilibrium, the reward-to-risk ratio is
constant for all assets and equal to E(RA) - Rf
/ ßA.  - To see this, consider two stocks O and U.
- Stock U gives higher return relative to its
level of risk, making it a more attractive asset.
- People will buy stock U and sell stock O. This
(adjustment) process will continue until both
stocks have the same reward/risk ratio.
404.1 The security market line..
- This is the fundamental relation between risk
and return. -
- This relation describes a straight line with
vertical intercept equal to Rf and the slope of
the line equal to the risk/return ratio. This
line is called the Security Market Line (SML). - Since this relation also applies to market
portfolio, and by definition ßm 1, we have. - slope (E(Rm)-Rf )/ ßm E(Rm)-Rf
414.1 The Security Market Line (SML)..
Asset expectedreturn
E (Rm) Rf
E (Rm)
Rf
Assetbeta
bm1.0
Graphically, this relation says that if we plot
expected return against beta, all stocks will
fall on the Security Market Line.
424.2 Capital Asset Pricing Model (CAPM)
- We now know about risk/return ratio, E(Ri) -
Rf /?i, -
- (1) it is same for all assets,
- (2) it is given as slope of the security market
line, and - (3) slope of the security market line is equal to
E(Rm) - Rf. - ? for any asset i, the risk/return ratio is equal
to E(Rm) - Rf - Â
- E(Ri) - Rf / ?i E(Rm) - Rf
- Rearranging this relation gives,
- E(Ri) Rf ?i E(Rm) - Rf
434.2 Capital Asset Pricing Model (CAPM)
- The Capital Asset Pricing Model (CAPM) - an
equilibrium model of the relation between risk
and return. - E(Ri ) Rf ?i ? E(Rm ) - Rf
- An assets expected return has three components.
- The risk-free rate - the pure time value of
money - The market risk premium - the reward for bearing
systematic risk - The beta coefficient - a measure of the amount
of systematic risk of asset i relative to the
market portfolio.
44The Security Market Line (SML)
Asset Expectedreturn (E(Ri)
C
E (RC)
E (Ri) - Rf Bi
?
E (RD)
D
E (RB)
?
E (RA)
?
Rf
Asset Beta
455. Summary
- I. Total risk the variance (or the standard
deviation) of an assets return. - II. The benefit from diversification
diversification eliminates some but not all of
risk via the combination of assets into a
portfolio. The lower the correlation among
assets, the greater the benefit from
diversification. - III. Systematic and unsystematic risks
Systematic risks are unanticipated events that
have economy-wide effects. - Unsystematic risks are unanticipated events that
affect single assets or small groups of assets. - IV. Diversification eliminates (most)
unsystematic risk, but the systematic risk
remains. - This observation leads to a principle the reward
for bearing risk depends only on its level of
systematic risk. Beta measures a stocks
systematic risk. - V. In equilibrium, the reward-to-risk ratio is
same for all assets, and equal to the slope of
SML (security market line). - VI. The capital asset pricing model E(Ri) Rf
E(Rm) - Rf ????i.
464.1 The security market line..
- Example
- Asset A has an expected return of 12 and a
beta of 1.40. Asset B has an expected return of
8 and a beta of 0.80. Are these assets valued
correctly relative to each other if the risk-free
rate is 5? - For A (.12 - .05)/1.40 ________
- For B (.08 - .05)/0.80 ________
- What would the risk-free rate have to be for
these assets to be correctly valued? - (.12 - Rf)/1.40 (.08 - Rf)/0.80
- Rf ________
474.1 The security market line..
- Example
- Asset A has an expected return of 12 and a
beta of 1.40. Asset B has an expected return of
8 and a beta of 0.80. Are these assets valued
correctly relative to each other if the risk-free
rate is 5? - For A (.12 - .05)/1.40 .05
- For B (.08 - .05)/0.80 .0375
- What would the risk-free rate have to be for
these assets to be correctly valued? - (.12 - Rf)/1.40 (.08 - Rf)/0.80
- Rf .02666
487. Questions
- 1. Assume the historic market risk premium has
been about 8.5. The risk-free rate is currently
5. GTX Corp. has a beta of .85. What return
should you expect from an investment in GTX? - E(RGTX) 5 _______ ? .85 12.225
- 2. What is the effect of diversification?
- 3. What does SML say?
- The slope of the SML ______ .
497. Questions..
- Assume the historic market risk premium has been
about 8.5. The risk-free rate is currently 5.
GTX Corp. has a beta of .85. What return should
you expect from an investment in GTX? - E(RGTX) 5 8.5 ? .85 12.225
- What is the effect of diversification?
Diversification reduces unsystematic risk. - 3. Return-to-risk ratio is same for all assets.
The slope of the SML E(Rm ) - Rf .
507. Qu.
- Consider the following information
- State of Prob. of State Stock A Stock B Stock
CEconomy of Economy Return Return Return - Boom 0.35 0.14 0.15 0.33
- Bust 0.65 0.12 0.03 -0.06
- What is the expected return on an equally
weighted portfolio of these three stocks? - What is the variance of a portfolio invested 15
percent each in A and B, and 70 percent in C?
517. Qu.
- Expected returns on an equal-weighted portfolio
- a. Boom ERp (.14 .15 .33)/3 .2067
- Bust ERp (.12 .03 - .06)/3 .0300
- so the overall portfolio expected return must
be - ERp .35(.2067) .65(.0300) .0918
527. Qu.
- b. Boom ERp __ (.14) .15(.15) .70(.33)
____ - Bust ERp .15(.12) .15(.03)
.70(-.06) ____ - ERp .35(____) .65(____) ____
- so
- 2p .35(____ - ____)2 .65(____ - ____)2
- _____
-
537. Qu.
- b. Boom ERp .15(.14) .15(.15) .70(.33)
.2745 - Bust ERp .15(.12) .15(.03)
.70(-.06) -.0195 - ERp .35(.2745) .65(-.0195) .0834
- so
- 2p .35(.2745 - .0834)2 .65(-.0195 -
.0834)2 - .01278 .00688 .01966
54Qu
- Using information from the previous chapter on
capital market history, determine the return on a
portfolio that is equally invested in Canadian
stocks and long-term bonds. - What is the return on a portfolio that is equally
invested in small-company stocks and Treasury
bills?
55Qu
- Solution
- The average annual return on common stocks over
the period 1948-1999 was 13.2 percent, and the
average annual return on long-term bonds was 7.6
percent. So, the return on a portfolio with half
invested in common stocks and half in long-term
bonds would have been - ERp1 .50(13.2) .50(7.6) 10.4
If on the other hand, one would have invested in
the common stocks of small firms and in Treasury
bills in equal amounts over the same period,
ones portfolio return would have been
ERp2 .50(14.8) .50(3.8) 9.3.