Equity Valuation

• Stock Valuation Models

• Diversification
• Efficient Frontiers
• Capital Asset Pricing Model (CAPM)
• State-of-the-Art Models of StockReturns

2011-10-10 Dastidar 1
 Motivating Questions

• An analyst is trying to determine the value of a company

for the purposes of an acquisition. The company is going
to experience very high growth in the first few years, but
growth will slow in later years as new competitors enter
the market. How does the analyst value the company?

=> the Dividend Discount Model

2011-10-10 Dastidar 2
 Motivating Questions

• A manager is faced with a choice of “equity”,

“bonds”, “cash” and “international equity” in his
company’s pension fund. Which combination of
assets should the manager choose in his 401K?

 Diversification and mean-variance

analysis

2011-10-10 Dastidar 3
 Motivating Questions

• A company is investigating whether to take on a new

project using NPV analysis. How does the firm
determine the discount rate it is to use?

=> CAPM

2011-10-10 Dastidar 4
 Valuing the Firm
• The value of the firm is the discounted value of the
expected cashflow of the firm (for paying all
investors – debt and equity holders)

E[C(t)]
V  Value of firm  
t 1 (1  WACC)
t

where E[C(t)] is the expected cashflow to the firm in year t,

and WACC is the constant Weighted Average Cost of Capital
of the firm.
(We leave a detailed analysis of obtaining WACC to later and
in this class we concentrate only on all equity firms)

2011-10-10 Dastidar 5
 How to Value Equity
• Equity holders have a claim on the firm. A share of
the firm pays dividends, so we can value the share by
discounting the dividends:

E[Div(t)]
P(0)  value of a share  
t 1 (1  r*)
t

Where E[Div(t)] is the expected dividend per share at

time t and r* is the constant expected return on equity.

This is the Dividend Discount Model

2011-10-10 Dastidar 6
 Dividend Discount Model: Zero Growth

• Notation:
D( t )  E[Div( t )] Expected dividends
X( t )  E[Earn ( t )] Expected earnings
E[Div(t  1)  P(t  1)] Expected
r*  E[r ( t  1)]  1 returns
P(t)
Suppose expected dividends are constant at D,
that is D(t) = D, then

D(t) D
P(0)   
t 1 (1  r*)
t
r*

2011-10-10 Dastidar 7
 Equity Valuation (I)
• An all-equity firm has 1 mil shares outstanding,
currently priced to yield 10%. The firm earns $3 mil
net profit after tax each year, and distributes the whole
amount as a dividend. What is the value of the firm
and the price per share?

2011-10-10 Dastidar 8
 Solution

3 mil
Value of Firm   30 mil
.1
30 mil
Price per share   $30
1 mil

2011-10-10 Dastidar 9
 Equity Valuation (II)
• Suppose a new project, very similar to the firm’s
existing assets becomes available at a cost of $4 mil.
The project is forecast to earn $1.2 mil each year in net
cashflows. The firm wishes to retain its all equity
status and finance the project with a new share
issuance. How many shares should be sold, and at
what price?

2011-10-10 Dastidar 10
 Solution

• First, value the project

1.2 mil
PVof Project   12 mil
.1
Find new share issuance
Value of Stock  Value of all projects
No of new shares  New share price
 Cost of new project

2011-10-10 Dastidar 11
The following relationships must hold:
(1 mil  x ) y  PV(existin g assets) PV(new project)
(1 mil  x ) y  42 mil
x  number of new shares
and y  new price per share
xy  4 mil
Two equations, two unknowns
4 mil
y  $38 x  105,264
38
2011-10-10 Dastidar 12
 Dividend Discount Model: with Growth
• If expected dividends grow at constant rate g, and
discount rate is constant at r* then
D(1) D(0)(1  g )
P ( 0)   where
r * g r * g D(1)=D(0)(1+g)
• This can be applied to a case with multiple regimes. For example,
the first 2 years expected dividends are D(1) and D(2), and then
dividends grow at rate g thereafter
P(0)  PV (CF years 1 & 2)  PV (CF thereafter)
D(1) D(2) 1 D(2)(1  g )
  
(1  r * ) (1  r*) (1  r*) 2 r * g
2

2011-10-10 Dastidar 13
 Stock Valuation with Growth: Example
• Example: a firm’s current dividend per share is $10,
and dividends are expected to grow at 10%. If the
expected return of the firm is 12% what is the price of
the stock?

10(1  .1)
P ( 0)   $550
.12  .10

2011-10-10 Dastidar 14
 Venture Capital Example
• A venture capitalist takes an equity stake of 20% in a
start-up by investing $2m. The start-up has negative
cashflow in the first year of $2m. The start-up earns a
positive cashflow of $0.5m in year 2. From years 2-5,
the firm is able to maintain a constant monopoly
growth rate g. After year 5, the firm is expected to
grow at the industry average of 10% per annum. What
is the required monopoly growth rate for years 2-5 to
give the venture capitalist a required return of 20%?

2011-10-10 Dastidar 15
 Solution

• Write down the cashflows!

time 0 1 2 3 4 5 6
2 3 3
cashflow -10 -2 0.5 .5(1+g) .5(1+g) .5(1+g) .5(1+g) (1.10)

Discount the cashflows and solve for g

10  -2(1.2) -1  0.5(1.2) 2  0.5(1  g)(1.2) -3
0.5(1  g) 2 (1.2) 4  0.5(1  g)3 (1.2) 5
0.5(1  g)3 (1.1)
 (1.2) 5 g  .6183
.2  .1

2011-10-10 Dastidar 16
 Valuing a Stock using Earnings
• Expected earnings grow at rate g (starting at X(0) in year 0),
and a fraction a of earnings is paid as dividends, and the
discount rate is constant at r*:

D(1) a  X (1) a  X (0)(1  g )

P ( 0)   
r * g r * g r * g

Note: P/E ratio is given by: P(0) a (1  g )


X(0) (r * g )

2011-10-10 Dastidar 17
 Interpreting P/Es
• Suppose the earnings growth is constant at g, and the
expected return on equity is constant at r*.
• The value of a stock can be written as the value of a
no-growth stock + the present value of growth
opportunities (PVGO)

2011-10-10 Dastidar 18
 X(1) aX(1)
P(0)   PVGO 
r* r * g
no  growth
value
PV of earnings
X(1)[g - (1 - a)r*]
 PVGO  generated by the
r * (r *  g)
assets in place
P(0) 1 g - (1 - a)r *
 
X(1) r* r * (r *  g)

Hence a high P(0)/X(1) means either:

low r* or
high g (most of the firm’s value is in the growth potential), and
this only affects the PVGO term
2011-10-10 Dastidar 19
 Interpreting P/E’s II
• Earnings growth g depends on:
- the fraction re-invested (1-a)
- the profitability of reinvestment (RIR)
Earn Inv Earn
g  
Earn Earn Inv
 (1  a )  RIR

plowback ratio

The PVGO is also known as the economic value

added (EVA) or economic profit.

2011-10-10 Dastidar 20
 PVGO: Example
• Company A has expected earnings next year of $6
which are projected to remain always constant.
Company A pays out all its earnings every period and
has a required rate of return of 6%. What is Company
A’s price?
• Company B also has expected earnings next year of $6,
constant RIR = 6% and required rate of return of 6%.
Company B pays out 30% of its earnings each year as
dividends. What is Company B’s growth rate of
earnings and Company B’s price?

2011-10-10 Dastidar 21
Company A:
a  X(1) 1 6
P ( 0)    $100
r* .06
Company B:
g  (1  a )  RIR  .7  .06  .042
a  X(1) .3  6
P ( 0)    $100
r * g .06  .042
Company B is growing, but has the same price as Company A!

=> Only when RIR>r* does growth contribute to value!

2011-10-10 Dastidar 22
 Summary: Equity Valuation

• Valuing Equity:
- When a firm undertakes a new project, the NPV of the new
project accrues only to existing shareholders.
- The Dividend Discount Model can be used to value firms which
have different regimes of growth.
- A firm’s value can be decomposed into the present value of the
earnings generated by the assets in place plus the present value
of growth opportunities.

2011-10-10 Dastidar 23
 Diversifcation: Motivation

• Recall Finance Axioms 1+2:

- Investors prefer more to less
- Investors are risk-averse

• Can a US investor holding only a domestic US

portfolio increase her return for the same amount of
risk by holding overseas assets?

=>Diversification

2011-10-10 Dastidar 24
 Mean/Variance Utility
• The utility function for a mean/variance investor can be written:
U = E(rp) – 0.5 A p2
where E(rp) is the expected return of the investor’s portfolio and
p2 is the variance of the investor’s portfolio, and A represents
the degree of the investor’s aversion to risk.
• Example: Return Deviation E(r)-0.5*A*2
10.18% 5.00% 0.1 A = 1.4
10.70% 10.00% 0.1
11.58% 15.00% 0.1
12.80% 20.00% 0.1
14.37% 25.00% 0.1

2011-10-10 Dastidar 25
 Mean/Variance Utility

• For a given level of risk Mean-Standard Deviation Indifference Curves

(standard deviation of 0.17

returns) we prefer 0.16

higher expected returns. 0.15

• For a given level of 0.14

mean
expected returns we 0.13

prefer less risk (lower 0.12

standard deviation of 0.11

returns) 0.1

0.05 0.1 0.15 0.2 0.25 0.3

standard deviation

2011-10-10 Dastidar 26
 Mean and Variance of Portfolio Returns

• Suppose we have a portfolio of N stocks, with stock

i’s return having mean E(ri) and variance var(ri)=i2.
The covariance of stock i with stock j is given by
cov(ri, rj).
• The weight held in each stock, wi, is:

$ value of stock i' s position

wi 
total $ value of portfolio

2011-10-10 Dastidar 27
 Mean of a portfolio’s returns

• The return on the portfolio is given by:

N
rp   w i ri
i 1
• The expected return on this portfolio is:
N
E(rp )   w i E(ri )
i 1

2011-10-10 Dastidar 28
 Variance of a portfolio’s returns
• The variance of the return on this portfolio is:
N  N 
var(rp )   w i   w jcov(ri , rj ) 
i 1  j1 
N N N
var(rp )   w i2 σ i2  2 w i w jcov(ri , rj )
i 1 i 1 ji

• Note the covariance is: cov(ri , rj )  ρ ijσ i σ j

2011-10-10 Dastidar 29
 Example: 3 assets

• Suppose we have three assets, US JP and UK, with means and

covariances given by:
mean volatility
US 13.6% 15.4%
JP 15.0% 23.0%
UK 15.9% 24.3%
with correlation matrix:
US 1 0.27 0.50
JP 0.27 1 0.36
UK 0.50 0.36 1
If an investor holds 60% in the US, 20% in JP and 20% in UK
what is the mean and volatility of the portfolio?
2011-10-10 Dastidar 30
 Example: 3 assets

• Portfolio mean:
E(rp) = 0.6*0.136 + 0.2*0.150 +0.2*0.159
= 0.143
• Portfolio variance:
var(rp) = (0.6)2*(0.154)2 + (0.2)2*(0.230)2 + (0.2)2*(0.243)2
+ 2*0.6*0.2*0.27*0.154*0.230
+ 2*0.6*0.2*0.50*0.154*0.243
+ 2*0.2*0.2*0.36*0.230*0.243
= 0.021
p = 0.146

2011-10-10 Dastidar 31
 Risk Reduction in Diversified Portfolios
• Suppose we have two assets, US and JP. Is it possible
to hold a combination of the assets so that the risk of
the portfolio (standard deviation of portfolio returns) is
less than the risk of either asset?

2011-10-10 Dastidar 32
 Risk Reduction in Diversified Portfolios
• Let w be the weight in the US, and 1-w the weight in
JP.
• The expected return of the portfolio is:
E(rp) = w*0.136 + (1-w)*0.150
• The variance of the portfolio return is:
var(rp) = w2*(0.154)2 + (1-w)2*(0.230)2
+2*w*(1-w)*0.27*0.154*0.230
• What happens when we vary w?

2011-10-10 Dastidar 33
 We can show the portfolio in risk/return space:

0.152

0.15 JP
w mean volatility
0.0 0.150 0.230
0.148 0.1 0.149 0.212
0.2 0.147 0.195
0.3 0.146 0.179
expected return

0.146
0.4 0.144 0.166
0.5 0.143 0.155
0.144 0.6 0.142 0.147
0.7 0.140 0.143
0.8 0.139 0.143
0.142 0.9 0.137 0.146
1.0 0.136 0.154
0.14

0.138

0.136 US

0.15 0.16 0.17 0.18 0.19 0.2 0.21 0.22 0.23 0.24 0.25
standard deviation

2011-10-10 Dastidar 34
 Risk Reduction in Diversified Portfolios
• Suppose the correlation of US and JP is +/-1. What is the
risk/return of the portfolio as we vary w?

0.152 Case with  = -1

Case with  = 0.27
0.15 Case with  = 1 JP

0.148
expected return

0.146

0.144

0.142

0.14

0.138

0.136 US

0 0.05 0.1 0.15 0.2 0.25

standard deviation

2011-10-10 Dastidar 35
 Risk Reduction in Diversified Portfolios
• It can be shown that for two assets with means 1 and 2, and
volatilities 1 and 2 with correlation –1<, that:
• The portfolio having minimum variance has variance:
(1  ρ 2
) σ 2 2
1σ2
min σ p  2
2

σ 1  2 ρσ 1σ 2  σ 22
• To produce this minimum variance portfolio we hold a weight
w in asset 1 of:
σ 2  ρσ1σ 2
2
w 2
σ1 - 2ρσ1σ 2  σ 22

2011-10-10 Dastidar 36
 Risk Reduction in Diversified Portfolios
• Suppose that 1=2=, that is each stock has the same
volatility. Then:

• w=½
=> Equally weighted portfolio

• min p2 =2(1- ½(1-) ) < 2

=> Risk reduction in the portfolio

2011-10-10 Dastidar 37
 Risk Reduction in Diversified Portfolios
• Suppose we start with a typical US stock.
• Now suppose we add stocks to the portfolio, all stock
positions equally weighted.
 What is the percentage reduction
in risk we should expect from
adding stocks to our portfolio?
 (In the graph 100% represents
the typical risk of a US stock)

2011-10-10 Dastidar 38
 Portfolio Risk in Diversified Portfolios
• Suppose we have an equally weighted portfolio
(holding weights 1/N) of N stocks. What is the
variance of portfolio returns?

N N N
1 1
σ  2
2
p
N

i 1
σ  2
2
i
N
 cov(r , r )
i 1 j i
i j

1  1 N 2   1  1 N N 
σ    σ i   1   
2
p  cov(ri , rj )
N  N i 1   N   N(N - 1) i 1 ji 
1 average   1  average 
σ  
2
 1   

N  variance  N  covariance
p

2011-10-10 Dastidar 39
 Portfolio Risk in Diversified Portfolios

• What happens when N goes to infinity?

• Variance of portfolio returns

= average covariance of returns

• Risk of portfolio = Non-diversifiable risk

2011-10-10 Dastidar 40
 Diversifiable vs Non-Diversifiable Risk

• Hence, when held in a portfolio some of the risk of a

stock disappears.

• Or, the risk contribution the stock makes to the

portfolio is LESS than the risk of the stock if held in
isolation:

 total risk in   risk that cannot be   risk that can be 

       
 a stock   diversified away   diversified away 

2011-10-10 Dastidar 41
 Diversifiable vs Non-Diversifiable Risk

• The expected return on a stock r, can be

decomposed into the risk-free return rf, plus a risk
premium :
r = rf + 

• The risk premium , is a function of undiversifiable

risk (risk that cannot be diversified away).

2011-10-10 Dastidar 42
 Summary: Diversification

• Under mean-variance utility investors seek to obtain the

highest expected return for a given variance.
• In equally weighted portfolios investors can obtain lower risk
(volatility) in portfolios of stocks than if the stocks were held
individually.
• The expected return of a stock is equal to the risk-free rate plus
a risk premium. The risk premium is a function of
undiversifiable risk, that is, risk that cannot be removed when
holding the stock in a diversified portfolio. Later on, we’ll
show that this is captured by the stock’s CAPM beta!

2011-10-10 Dastidar 43
 Asset Allocation: Efficient Frontier

• I am a US investor. What are the benefits of holding

overseas securities?
• What combination of US versus foreign stocks should
we hold?
• Does the answer change if we have a risk-free asset in
the US?

2011-10-10 Dastidar 44
 International Equity Investment Example

US UK France Germany Japan

• Means: 0.1355 0.1589 0.1519 0.1435 0.1497
: 0.1535 0.2430 0.2324 0.2038 0.2298

• Correlation matrix:
US UK France Germany Japan
US 1.0000 0.5003 0.4398 0.3681 0.2663
UK 0.5003 1.0000 0.5420 0.4265 0.3581
FR 0.4398 0.5420 1.0000 0.6032 0.3923
GR 0.3681 0.4265 0.6032 1.0000 0.3663
JP 0.2663 0.3581 0.3923 0.3663 1.0000

2011-10-10 Dastidar 45
 Combining a Risk-Free and Risky Assets
Outline:
1. Now: We investigate the case of one risky asset
(US equity) and a risk-free asset first [the last
motivating question]
2. Later: We investigate the case of two risky
assets (US and Japan) and a risk-free asset

The risk-free asset:

 May be horizon dependent
 May not exist
 Proxy: T-bills

2011-10-10 Dastidar 46
 Combining a Risk-Free and One Risky Asset
• Risk-free return: rf
• Risky return: r
• Weight on risky asset: w

• Return on portfolio:
rp  w  r  (1  w)r f  rf  w  (r  rf )

excess return
• Expected portfolio return:
E[r p ]  rf  w  E[r  rf ]

2011-10-10 Dastidar 47
 Combining a Risk-Free and One Risky Asset
• Variance of portfolio return:
σ 2p  E[(r p  E[r p ]) 2 ]  w 2  σ 2
where 2 is the variance of return r. Hence:
σp  w σ

• Combining the expected return and volatility expressions:

E[r]  rf
E[r p ]  rf  σp
σ

=> CAL (Capital Allocation Line)

2011-10-10 Dastidar 48
 The Capital Allocation Line

• Example: US E[r ]  13.55% σ 2  (15.35%) 2

rf  7.00%

• Hence: E[rp ]  0.07  SR  σ p

E[r]  rf 0.1355  0.07

SR    0.4267
σ 0.1535

• SR (Sharpe Ratio) = return premium per unit of risk

2011-10-10 Dastidar 49
 The Capital Allocation Line is:

Indifference
curves
E[rp ]

0.4267
0.07
rf

σp
2011-10-10 Dastidar 50
 Optimal Portfolios
• Mathematical Problem:

max U ( r p )  max
w w
[ E[r p ] 
1
2 A σ 2p ]
• How do you solve for the optimal w?
Find the derivative and set
equal to zero!

2011-10-10 Dastidar 51
 Optimal Portfolios
• Optimal portfolio solution:
E [ r ]  rf
w*  2
Aσ
• What is the intuition for the solution in terms of
• Security properties?
• Risk aversion?

2011-10-10 Dastidar 52
 Optimal Portfolios

• Numerical Example
A w* E[rp ] (in %) σ p (in % )
- 1.0 2.78 25.21% 42.67%
- 2.0 1.39 16.10% 21.34%
- 3.0 0.93 13.09% 14.28%
- 4.0 0.70 11.58% 10.75%

E[r p ]  rf  w  E[r  rf ] Ex: 0.2521 = .07 + 2.78´(.1355-.07)

σp  w σ 0.4267 = 2.78´0.1535

2011-10-10 Dastidar 53
 Optimal Portfolios

• Leverage:

E[rp ]
L
Portfolio K: long in the risky asset
long in risk free
K
Portfolio L: borrow risk free asset to

rf invest in risky asset

Lending region Borrowing region

σ σp
2011-10-10 Dastidar 54
 Optimal Portfolios

Application

What is the typical investor’s risk aversion?

Assumptions:
- Risky portfolio = S&P500
- Risk-free asset = T-bills

2011-10-10 Dastidar 55
 Optimal Portfolios

We know for the US:

E[r]  rf  6.55%, σ  15.35%

Also, the market value of stocks is about 4

times as large as that of short-term (< 6
months) T-bills:
4
w *   .8
4 1
2011-10-10 Dastidar 56
 Optimal Portfolios

Then, what A is implied for the optimal w*?

E[ r ]  rf .0655
0 .8  w *  2
 2
Aσ A ( 0 .1535 )

 A  3 . 47

2011-10-10 Dastidar 57
 Recap/Review

• The CAL describes the menu of possible risk/return

trade-offs between a risk-free and risky asset. Its slope
equals the Sharpe ratio.
• The optimal portfolio combination with a risky and risk-
free asset for any investor is the tangency point of that
investor’s indifference curves with the CAL.
• Now, we move to asset allocation between two and more
risky securities and a risk-free asset.
• Later: look at the G5 asset allocation case.

2011-10-10 Dastidar 58
 Two Risky Assets

Concepts:

Frontier Portfolio:
Feasible Portfolio that has the lowest possible
standard deviation for a given expected return

Efficient Portfolio:
Feasible Portfolio that has the highest expected
return for a given standard deviation

2011-10-10 Dastidar 59
 Two Risky Assets

Mean-standard deviation frontier for US and Japan

0.18

0.17 Efficient Frontier

0.16 short US
0.15 JP

0.14
Er

US
0.13
Minimum
0.12 Variance short JP
Portfolio
0.11

0.1
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4


2011-10-10 Dastidar 60
 Two Risky Assets

Terminology:
• The minimum standard deviation or minimum variance
frontier is the locus of the portfolios in expected return-
standard deviation space that have the minimum variance
for each expected return. For two assets, it consists of all
possible portfolio combinations.
• The efficient frontier of risky assets is the set of efficient
portfolios. It is the upper portion of the minimum variance
frontier starting at the global minimum variance portfolio.

2011-10-10 Dastidar 61
 Two Risky Assets and One Risk-Free Asset
• With two risky assets, different CALs are available. What is the
optimal risky portfolio?

• The “tangency portfolio” provides the “steepest” CAL. In other

words, it maximizes the Sharpe ratio for all possible risky
portfolios. This portfolio is known as the minimum variance
efficient portfolio or MVE.

• Everyone will hold the same risky portfolio (MVE) and mix it
with the risk-free asset according to his/her preferences. The
efficient set is the set of portfolios on the CAL emanating from
rf going through MVE.

2011-10-10 Dastidar 62
 Two Risky Assets and One Risk Free Asset

US and Japan with Risk Free Asset

0.16
B
0.15 JP The MVE maximises the SR.
A
All other points (eg points A,
0.14 Efficient Set MVE
US B) have lower SR’s
0.13

0.12

0.11 w Erp p SR
Er

0.1
MVE 0.6822 0.1400 0.1427 0.4905
A 0.1901 0.1470 0.1959 0.3930
B -0.3732 0.1550 0.3053 0.2784
0.09

0.08
w=weight in the US
0.07 SR = Sharpe Ratio
0.06
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4


2011-10-10 Dastidar 63
 Two Risky Assets and One Risk Free Asset

Mathematical Problem:
• Find the portfolio that maximizes the Sharpe ratio (the
E [ rp ]  rf
slope of the CAL): max w p

• where E[rp ]  w E[r1 ]  1  w  E[r2 ]

σ p  w 2 σ12  (1  w ) 2 σ 22  2 w (1  w ) ρσ1σ 2

• Solution:
E[r1  rf ]  σ 22  E[r2  rf ]  σ12
w MVE  with σ12  ρσ1σ 2
E[r1  rf ]σ 2  E[r2  rf ]σ1  E[r1  rf ]  E[r2  rf ]σ12
2 2

2011-10-10 Dastidar 64
 Two Risky Assets and One Risk Free Asset

Numerical Example (US=1 and Japan=2):

• Plug in: E [ r1 ]  13 . 55 %
E[r 2 ]  14.97%
σ12  (0.1535)2 ρ  0.2663
σ 22  (0.2298) 2 rf  0.07

w MVE  0.6822
• Results:
σMVE  0.1427
E[r MVE ]  0.1400
SR MVE  0.4905

2011-10-10 Dastidar 65
 Two Risky Assets and One Risk Free Asset

• Note the individual Sharpe ratios:

6.55% 7.97%
SR 1   0.4267 SR 2   0.3468
15.35% 22.98%
SR MVE  0.4905
• Conclusion: determines the optimal risky
portfolio w MVE

2011-10-10 Dastidar 66
 Two Risky Assets and One Risk Free Asset

• Retracing our steps:

 The CAL gives the menu of risk free and risky portfolio
combinations.
 Efficient portfolios on the Mean Variance Frontier provide
the highest expected return for each level of volatility.
 For a given risk free rate, the mean variance efficient
portfolio has the highest Sharpe ratio among all possible
risky portfolios.

2011-10-10 Dastidar 67
 Two Risky Assets and One Risk Free Asset

• Conclusions:

The “best” CAL will combine the risk free asset with
the MVE portfolio (two-fund separation)
=> this means we are back in the one-risky asset world

We can use our old utility maximization and the “best”
CAL to obtain our optimal asset allocation over two
risky portfolios and the risk free asset.

2011-10-10 Dastidar 68
 Two Risky Assets and One Risk Free Asset

• Numerical example:
A  3.47
 E[rMVE ]  rf
then w 
A  2MVE
0.1400  0.07

3.47  (0.1427) 2

 99.07%
2011-10-10 Dastidar 69
 Two Risky Assets and One Risk Free Asset
• Put 99.07% of your portfolio in MVE, put 0.93% of
your portfolio in the risk free asset.

• Portfolio allocations to original assets

• US: 0.9907 x 0.6822 = 0.6759
• Japan: 0.9907 x (1-0.6822) = 0.3148

2011-10-10 Dastidar 70
 US and Japan with Risk Free Asset
0.144

0.143

0.142

0.141

0.14

0.139
Er

0.138
Mean-Variance Frontier
Efficient Frontier
0.137 US
MVE
0.136 A=3.47
Optimal Utility = 0.1047
0.135 Utility = .1050
Utility = .1040
0.134
0.14 0.142 0.144 0.146 0.148 0.15 0.152 0.154

2011-10-10 Dastidar 71
 Two Risky Assets and One Risk Free

Application: The power of international diversification

• SR(US)= 0.4267
• SR(JP) = 0.3468 SR MVE  0.4905

• The Sharpe ratio will improve when adding a foreign asset if:
SR (new)  SR(US )  ρ

or E[rnew ]  rf  ( E[rUS ]  rf )  ρ  new
  US

Foreign risk Hurdle premium

premium

• Adding Japan will improve the 0.3468  0

.4267  0
.2663
 
Sharpe ratio as long as: 0.1136
2011-10-10 Dastidar 72
 Multiple Risky Securities

• Minimum variance frontier for N securities

 N 2 2 N N 
 i i   i j
minimum
min
{ w 1 ,..., w N }  i 1
w σ  w w cov[ ri , r ]
j   variance
i 1 j  i 
N N
such that  wi 1
i 1
feasible
portfolio  w i E[ ri ]  r  target
return
i 1

By varying r, we trace out the frontier

• Although analytical solutions are available, using Solver (Excel) is popular

2011-10-10 Dastidar 73
 Multiple Risky Securities

2011-10-10 Dastidar 74
 Recap/Review
• The CAL describes the menu of possible risk/return trade-offs between a risk-
free and risky asset. Its slope equals the Sharpe ratio.
• The optimal portfolio combination with a risky and risk-free asset for any
investor is the tangency point of that investor’s indifference curves with the
CAL.
• For a (possibly large set) of securities, one can compute a set of efficient
portfolios. These portfolios provide the highest expected return for any level of
standard deviation. They comprise the upper part of the mean standard
deviation frontier.
• If there is a risk-free asset, the mean-variance efficient portfolio (MVE) is the
risky portfolio with the highest Sharpe ratio.
• To find the optimal portfolio’s combination of the risk-free asset and the MVE,
plot the CAL that is tangent to the mean standard deviation frontier and find the
tangency point with the highest possible indifference curve.

2011-10-10 Dastidar 75
 Recap/Review
• The CAPM arises from imposing equilibrium on a frictionless
mean-variance world with rational investors holding the same
views on the “input list.”
 Prediction 1: Everyone should hold the market portfolio and
mix it with the risk-free asset.
=> Can you beat index funds?
• Prediction 2: The expected return on a stock is a linear function
of its beta.
=> Candidate for discount rate in NPV computations
=> Are stocks over- or under-priced?

Next CAPM!!
2011-10-10 Dastidar 76
 Optimality Conditions for Frontier Portfolios
Experiment:
• Suppose we have an optimal portfolio p.
• Consider adding a little bit of a new asset, say with a
fraction w:
• New portfolio: (1-w) old portfolio
w new asset

2011-10-10 Dastidar 77
 Optimality Conditions for Frontier Portfolios

It can be shown that for small w:

SR
 0 if SR NEW  SR p  ρ
w
with  the correlation between our portfolio and the new asset,
SRNEW the Sharpe Ratio of the new asset, and SRp the SR of our
present portfolio
We increase our Sharpe ratio

Conclusion: We will keep adding the asset until:

() SR NEW  SR p  

2011-10-10 Dastidar 78
 Optimality Conditions for Frontier Portfolios
Rewrite (*):
E[rNEW ]  rf E[rp ]  rf E[rNEW ]  rf E[rp ]  rf
  ρ or 
σ NEW σp cov[rNEW , rp ] σ 2p
The relevant risk of a security is its covariance
with the portfolio

Or, rewriting: E[rNEW  rf ] E[rp ]  rf


ρ  σ NEW σp

(1-)NEW is diversified away

2011-10-10 Dastidar 79
 Optimality Conditions for Frontier Portfolios

 In particular, this is true for:

p: any minimum variance frontier portfolio.
i: any security or portfolio

COV[ri , rp ]
Implication: E[ri ]  rf 
 2

 E[rp ]  rf 
p

=> WOW!
2011-10-10 Dastidar 80
 Optimality Conditions for Frontier Portfolios

 Retracing our steps and taking a leap forward

(1) The efficient frontier is a set of “dominant” portfolios

in risk/return space. Non-efficient portfolios would not
be held by any mean-variance investor.

(2) If there exists a risk-free asset, one portfolio of risky

securities offers the best risk-return trade-off; the
minimum variance efficient (MVE) portfolio.

2011-10-10 Dastidar 81
 Optimality Conditions for Frontier Portfolios

Question: If everybody is a mean-variance investor

facing the same frontier, what portfolio must the MVE
be for there to be no excess demand or supply for any
security?

=> Market Portfolio!

cov[ri , rM ]
Implication: E[ri ]  rf  2
 (E[rM ]  rf )
σM
where M is the market portfolio
=>The CAPM Relation
2011-10-10 Dastidar 82
 CAPM – Introduction

Capital Asset Pricing Model (CAPM)

Equilibrium model that underlies all modern financial

theory

Derived using principles of diversification with simplified

assumptions

Markowitz, Sharpe, Lintner and Mossin are researchers

credited with its development

2011-10-10 Dastidar 83
 CAPM – Assumptions

• Single-period investment horizon

• Individual investors are price takers
• Investments are limited to traded financial assets
• No taxes and transaction costs
• Information is costless and available to all investors
• Investors are rational mean-variance optimizers
• Homogeneous expectations

2011-10-10 Dastidar 84
 CAPM – Equilibrium
The CAPM is an equilibrium model:

• Economic equilibrium, where demand = supply and a

situation where no investor wants to change their
portfolio
• Riskier assets must have higher expected returns. If
not, nobody would hold them in equilibrium
• The CAPM describes the kind of risk-return
relationships that we might see in equilibrium

2011-10-10 Dastidar 85
 CAPM – Equilibrium

The equilibrium argument:

• In equilibrium, a security’s price must be such that all the shares are held (supply =
demand)
• Example: suppose everyone’s optimal portfolio assigns wIBM = 0
=> This cannot be an equilibrium, someone must hold them!
• What happens? If no-one wants to hold it, IBM must be overpriced
=>IBM’s expected return is too low
• What happens? Price falls (expected returns increase), until investors want to hold
exactly the number of IBM shares outstanding.
• Find prices so that supply = demand (equilibrium)

2011-10-10 Dastidar 86
 CAPM – Equilibrium

Resulting Equilibrium Conditions

All investors will hold the same portfolio for risky assets –
the market portfolio
The market portfolio contains all securities and the
proportion of each security is its market value as a
percentage of total market value
The risk premium on the market depends on the average
risk aversion of all market participants
The risk premium on an individual security is a function
of its covariance with the market portfolio
2011-10-10 Dastidar 87
 CAPM Predictions

The CAPM: Wrong but useful!

 Benchmark for evaluating portfolio managers

 Impetus for Index Funds
 Starting point for estimates of expected
returns for asset classes, even for active
portfolio managers
 Basis for cost of capital computations

2011-10-10 Dastidar 88
 Portfolio Choice: The Capital Market Line

• The MVE on the frontier of all risky assets is the market

portfolio! The capital allocation line going through the
market portfolio is called the Capital Market Line (CML)

• Everybody holds the same risky portfolio, the market

portfolio, and mixes it with the risk-free asset.

 Index funds are great.

2011-10-10 Dastidar 89
 Portfolio Choice: The Capital Market Line

E[rM ]

rf

σM
2011-10-10 Dastidar 90
 Portfolio Choice: The Capital Market Line

• Each investor’s optimal portfolio weights wi* depend

on the investor’s risk aversion Ai:
1 E[rM ]rf
w  
*
i
Ai σ2M
• The CAPM assumes there is no net borrowing or
lending; all the assets in “positive supply” are in M.
Let the invested wealth of an individual be denoted by
Wi and let there be K individuals.

2011-10-10 Dastidar 91
 Portfolio Choice: The Capital Market Line

Then:
K
 K
 E[rM ]  rf
 A i  w i  Wi   Wi 
 
*
   σ 2
i 1
dollars in risky asset 
i 1
 M
for individual i total
wealth

Let A= “wealth-weighted” average risk aversion

Then: E[rM ]  rf  A  
2
M

The average investor invests 100% of his

portfolio in the market!
2011-10-10 Dastidar 92
 Cost of Capital: The Security Market Line
• For each individual asset, we have:

E[ri ]  rf  β i E[rM ]  rf 
where
cov(ri , rM )
βi 
σ 2M
• The relation between E[ri] and i is called the Security Market
Line (SML)

2011-10-10 Dastidar 93
 The Capital Market Line Versus the Security Market Line

E[rM ] E[rM ]
E[rQ ]  E[rK ]
E[rQ ]  E[rK ]
rf

σM 

E[rM ]  rf
σM E[rM ]  rf

2011-10-10 Dastidar 94
 Cost of Capital: The Security Market Line

Numerical Example:
[Link] reports the following beta estimates (May
2006):
IBM: 1.57
GD: 0.46
Assume the risk-free rate is 5%, the standard deviation
of the market portfolio is 15.5% and the average risk
aversion is 3.47.

You hold a portfolio consisting of 50% IBM and 50% GD.

What is your estimate for the expected return on this
portfolio?
2011-10-10 Dastidar 95
 Cost of Capital: The Security Market Line

Computations:
1) Market risk premium:
E[rM ]  rf  A  σ 2M  3.47  (.155) 2  8.34%
2) Expected individual stock returns:
E[rIBM ]=0.05  1.57  8.34%  0.1810
E[rGD ]  0.05  0.46  8.34%  0.0884
3) Hence portfolio return:
E[rp ]=0.5  0.1810  0.5  0.0884  13.47%

2011-10-10 Dastidar 96
 Cost of Capital: The Security Market Line

Mis-pricing:
The SML when some E[rm ]
securities are
mispriced (non-zero )
according to the CAPM rf

=1

2011-10-10 Dastidar 97
 Cost of Capital: The Security Market Line

E[rM ]

rf

=1

2011-10-10 Dastidar 98
 Cost of Capital: The Security Market Line

Numerical Example:
[Link] reports the following beta estimates (May
2006):
IBM: 1.57
GD: 0.46
Assume the risk-free rate is 5%, the standard deviation
of the market portfolio is 15.5% and the average risk
aversion is 3.47.

You hold a portfolio consisting of 50% IBM and 50% GD.

What is your estimate for the expected return on this
portfolio?
2011-10-10 Dastidar 99
 Cost of Capital: The Security Market Line

Computations:
1) Market risk premium:
E[rM ]  rf  A  σ 2M  3.47  (.155) 2  8.34%
2) Expected individual stock returns:
E[rIBM ]=0.05  1.57  8.34%  0.1810
E[rGD ]  0.05  0.46  8.34%  0.0884
3) Hence portfolio return:
E[rp ]=0.5  0.1810  0.5  0.0884  13.47%

2011-10-10 Dastidar 100

 Recap/Review
• The CAPM arises from imposing equilibrium on a frictionless
mean-variance world with rational investors holding the same
views on the “input list.”
 Prediction 1: Everyone should hold the market portfolio and
mix it with the risk-free asset.
=> Can you beat index funds?
• Prediction 2: The expected return on a stock is a linear function
of its beta.
=> Candidate for discount rate in NPV computations
=> Are stocks over- or under-priced?

2011-10-10 Dastidar 101