Corporate Finance

Lecture: Risk and Return

SHEN Tao （沈涛） Tsinghua University
Discount Rates, Risk and Required
Returns
 Finance uses the concept of risk to “calibrate” the discount
rate, required return or cost of capital used in a present value
problem.

 We will build a model which will measure how the risk of a

particular set of cash flows should be reflected in the
required return used to value those cash flows.

 Generally, the more uncertain the cash flows of a particular

project the higher the risk, and the higher the risk, the higher
the required return.

 Therefore riskier projects have lower present values than less

risky projects with the same cash flows.
 The Minimum Required Return
 The required returns for an investment are determined by four
primary factors
1 - Expected Inflation
2 - Real Returns (Compensation for Delayed Consumption)
3 - Supply and Demand of Investment Funds
4 - Risk
 The minimum return required for any investment must be at
least as great as:
Exp. Risk Free Return = Exp. Inflation + Exp. Real Return

Incorporates Investor’s Incorporates the Required Real

Expectations of Inflation Returns influenced by the
Aggregate Supply and Demand of
Investment Funds.
 Revisiting Required Returns
 Any asset (real or financial) where there is uncertainty about the future cash flows must
have a required return greater than the Expected Return on Risk Free Assets. Therefore:

Required Return = Exp. Inflation + Exp. Real Returns + Risk Premium

Or
Required Return = Exp. Risk Free Return + Risk Premium

 Therefore, for risky assets any model we develop must take the following form:
RA = RRF + RP
 Where RA is the req. return on an asset.
 RRF is the expected return on risk free cash flows.
 RP is an additional return premium for risk.
Value of $100 Invested at the End of 1925 in U.S. Large Stocks (S&P
500), Small Stocks, World Stocks, Corporate Bonds, and Treasury Bills
Table 11.1 Realized Returns, in Percent (%) for Small
Stocks, the S&P 500, Corporate Bonds, and Treasury Bills,
Year-End 1925–1935
 Historical Risks and Returns of Stocks
 Computing Historical Returns
 Realized Returns
 Individual Investment Realized Returns
 The realized return from your investment in the stock from t to t+1 is:

Divt 1  Pt 1  Pt Divt 1 Pt 1  Pt (Eq. 11.1)

Rt 1   
Pt Pt Pt
 Dividend Yield  Capital Gain Yield
Realized Return
Problem:
 Health Management Associates (HMA) paid a one-time special dividend of
$10.00 on March 2, 2007. Suppose you bought HMA stock for $20.33 on
February 15, 2007 and sold it immediately after the dividend was paid for
$10.29. What was your realized return from holding the stock?
Realized Return
Execute:
 Using Eq. 11.1, the return from February 15, 2007 until March 2, 2007 is
equal to

Div t 1  Pt 1  Pt 10.00  10.29  20.33

R t 1    0.002, or  0.2%
Pt 20.33

 This -0.2% can be broken down into the dividend yield and the capital gain
yield:

Divt 1 10.00
Dividend Yield    0.4919, or 49.19%
Pt 20.33
Pt 1  Pt 10.29  20.33
Capital GainYield    0.4939, or  49.39%
Pt 20.33
 Year-to-Year Total Returns on Large Cap
Equities 1926-2008
60%

40%

20%
Total
Returns 0%
(%)

-20%

-40%
Large Company
Common Stocks
-60%
19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 20 20
26 31 36 41 46 51 56 61 66 71 76 81 86 91 96 01 06
 Year-to-Year Total Returns on Corporate
Bonds 1928-2008
60%
U.S. Long Term Corporate Bonds
40%

20%

Total 0%
Returns
(%)
-20%

-40%

-60%
19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 20 20
26 31 36 41 46 51 56 61 66 71 76 81 86 91 96 01 06
 Year-to-Year Total Returns on TBills
1928-2008
60%
U.S. Treasury Bills
40%

20%

Total 0%
Returns
(%)
-20%

-40%

-60%
19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 20 20
26 31 36 41 46 51 56 61 66 71 76 81 86 91 96 01 06
Historical Risks and Returns of Stocks
 Computing Historical Returns
 Individual Investment Realized Returns
 For quarterly returns (or any four compounding periods that make up an
entire year) the annual realized return, Rannual, is found by compounding:

1  Rannual  (1  R1 )(1  R2 )(1  R3 )(1  R4 ) (Eq. 11.2)

 If you hold the stock beyond the date of the first dividend, then to
compute your return we must specify how you invest any dividends
you receive in the interim.
 To focus on the returns of a single security, we assume that all
dividends are immediately reinvested and used to purchase additional
shares of the same stock or security.
Compounding Realized Returns
Problem:
 Suppose you purchased Health Management Associate’s
stock (HMA) on March 16, 2006 and held it for one
year, selling on March 15, 2007. What was your realized
return?
Date Price Dividend
16-Mar-06 21.15
10-May-06 20.70 0.06
9-Aug-06 20.62 0.06
8-Nov-06 19.39 0.06
15-Feb-07 20.33
2-Mar-07 10.29 10.00
15-Mar-07 11.07
Compounding Realized Returns
Execute (cont’d):
 The table below includes the realized return at each period.

Date Price Dividend Return

16-Mar-06 21.15
10-May-06 20.70 0.06 -1.84%
9-Aug-06 20.62 0.06 -0.10%
8-Nov-06 19.39 0.06 -5.67%
15-Feb-07 20.33 4.85%
2-Mar-07 10.29 10.00 -0.20%
15-Mar-07 11.07 7.58%
Compounding Realized Returns
 We then determine the one-year return by
compounding.
1  Rannual  (1  R1 )(1  R2 )1  R3 (1  R4 )(1  R5 )(1  R6 )
1  Rannual  (0.982)(0.999)0.943(1.048)(0.998)(1.076)
Rannual 1.0411  1  .0411or 4.11%

 It is unlikely that anyone investing in HMA expected to

receive exactly the realized return
 Reinvestment; Tax; Transaction cost
The Distribution of Annual Returns for U.S. Large Company
Stocks (S&P 500), Small Stocks, Corporate Bonds, and
Treasury Bills, 1926–2012
Historical Risks and Returns of Stocks
 Average Annual Returns
 If we assume that the distribution of possible returns is the
same over time, the average return provides an estimate of
the return we should expect in any given year
 Average Annual Return of a Security

1
R  ( R1  R2  ...  RT ) (Eq. 11.3)
T
Arithmetic and geometric average
 (1+r1)×(1+r2)×..(1+rs)=(1+g)s
 g is the geometric average
 Time-weighted average return
 Geometric average: a better description of the long-run
historical performance.
 Arithmetic average: a better estimate of an investment’s
expected return over a future horizon based on its past
performance
 View past returns as independent draws from the same
distribution
Arithmetic and geometric average
 Which is larger? Normal distribution: GM=AM - σ2 /2
 Mathmatically: Maclaurin series; Jensen Inequality
 Economically: asymmetric effect of positive and negative
rates of returns
 -20%, 20%: gain on a small base
 20%, -20%: lose on a big base

 The larger the swings in rates of return, the greater the

discrepancy
Figure 11.3 Average Annual Returns in the U.S. for Small
Stocks, Large Stocks (S&P 500), Corporate Bonds, and Treasury
Bills, 1926–2012
What is Risk?
 Given this data on historical returns, define Risk.

 What would be required in the ideal measure of the

dispersion of observations in a set of data?

 Statistics provides a framework to think about this problem.

So for now, a little review.
 Statistics Review - Measuring Risk

Variance = Average value of squared deviations from the mean

Sum of the squared deviations from the mean
Variance =
Number of observations - 1
Var  R  
1
T 1
 1
( R  R ) 2
 ( R2  R ) 2
 ...  ( RT  R ) 2

Standard Deviation: The square root of the variance.

Standard Deviation = Variance

Statistics Review - Calculating Sample
Variance and Standard Deviation
Annual Returns for Two Stocks
Steady Corp Techie Corp
Annual Return Annual Return
Observations Observations
30.0% 50.0%
21.0% Average 15.0% 62.0% Average 21.0%
-4.0% -28.0%
12.0% Median 16.5% -12.0% Median 21.5%
39.0% 58.0%
-10.0% Variance 3.6% 1.0% Variance 14.4%
42.0% -15.0%
-5.0% Std. Dev. 19% 75.0% Std. Dev. 38%
30.0% 42.0%
-5.0% -23.0%
 Statistics Review - Calculating Variance
and Standard Deviation
Chart of Annual Stock Returns
100.0%
80.0%
60.0%
Returns

40.0%
20.0%
0.0%
1 2 3 4 5 6 7 8 9 10
-20.0%
-40.0%

Steady Corp Observations

Techie Corp Observations

 Average Annual Returns: 1926-2006
Investment Ave. Return St. Dev.
Small Stocks 21.7% 41.6%
Large Common Stocks 12.2% 20.1%
Corporate Bonds 6.4% 7.1%
U.S. Treasury Bills 4.2% 3.5%
Inflation 3.4%

Source: Global Financial Data

Measuring Risk
Var  R  
1
T 1
 ( R1  R ) 2  ( R2  R ) 2  ...  ( RT  R ) 2 

 Upside is a risk?
 Asymmetric returns: extreme values
 Benchmark?
 Alternatives
 Drawdown: DDi=max[(Dj – Di)/ Dj]; Max DD=Max(DDi)
 Lower partial standard deviation: using only “bad” returns
(negative deviations from the risk-free rate), left-tail standard
deviation
…
Historical Risks and Returns of Stocks
 The Normal Distribution
 95% Prediction Interval

Average  (2 x standard deviation)

R  (2 x SD  R  )

 About two-thirds of all possible outcomes fall within one

standard deviation above or below the average
Figure 11.5 Normal Distribution
Confidence Intervals
Problem:
 We found the average return for the S&P 500 from 2005-2009 to be 3.1% with
a standard deviation of 24.1%. What is a 95% confidence interval for 2010’s
return?
Confidence Intervals
Execute:
 Average ±(2  standard deviation) = 3.1% – (2  24.1%) to 3.1% + (2 
24.1% )
= –45.1% to 51.3%.
 Even though the average return from 2005 to 2009 was 3.1%, the S&P 500 was
volatile, so if we want to be 95% confident of 2010’s return, the best we can say
is that it will lie between –45.1% and +51.3%.
Normal Distribution
 Is normal distribution a realistic description of returns?
 Why normal?
 R is normally distributed
 R+R
 R*R
Summary of Tools for Working with
Historical Returns
Historical Tradeoff between Risk and
Return
 The Returns of Large Portfolios
 Investments with higher volatility, as measured by standard
deviation, tend to have higher average returns
The Historical Tradeoff Between Risk and Return
in Large Portfolios, 1926–2011
Historical Tradeoff between Risk and
Return
 The Returns of Individual Stocks
 Larger stocks have lower volatility overall
 Even the largest stocks are typically more volatile than a
portfolio of large stocks
 The standard deviation of an individual security doesn’t explain
the size of its average return
 All individual stocks have lower returns and/or higher risk than
the portfolios in Figure
Intermediate Risk & Return
 We have until now, we have used three basic understandings
of risk and return.

1 -Risk is measured by volatility, hence standard deviation.

2 -The greater the risk (or s ) the greater the expected return.
3 -The opportunity cost of capital (discount rate) is the
required return on projects that have similar risk.

 This is not good enough.

 Thinking about Risk & Return
 Risk is the Uncertainty about future outcomes.

 Looking at the past risk and return (ex-post), while helpful in

understanding possible future risk, can be misleading when
defining risk premiums for future (ex-ante) expected
returns.

 Not all risk is the same

 How do Expected or Required Returns change with as risk

increases?
 Two Types of Risk
 Required Returns do not increase linearly as overall
uncertainty increases!
 Some Risks are unique to the firm:
 Fire at a Production Facility
 Death of a CEO
 Failure of a new product
 Successful R&D project
 We refer to these Firm Specific Risks as “Independent” or
“Unsystematic” risks
 These Unsystematic risks offset across many firms because
they are unrelated.
 Common or Systematic Risk
 Some Risks effect all firms:
 Recession
 Changes in Tax Law
 Increased Interest Rates
 Some New Technologies
 We refer to these Non-Firm Specific risks as “Common” or
“Systematic” risks
 These Common risks generally impact all firms, but the level
of impact differs across many firms.
 Individual firms may be more or less exposed to Systematic
risk.
 Diversification
 Definition of a Portfolio - Any combination of assets and their
weightings.

 Diversification is the averaging out of risks in a large portfolio

of assets.

 When we combine many stocks in a large portfolio, only the

unsystematic risk will be eliminated by diversification, the
volatility of the portfolio will decline until only the systematic
risk, which affects all stocks, remain.
 Diversification
 If the variation in the returns of individual assets are highly
positively correlated, the standard deviation of a portfolio of
these assets will be not significantly less than that of the
individual assets.
 If the variation in the returns of individual assets are
unrelated, the standard deviation of a portfolio of these assets
will be significantly less than that of the individual stocks.
 If the variation in the returns of individual assets are
negatively correlated, the standard deviation of a portfolio of
these assets will be even more significantly less than that of
the individual stocks.
Figure 12.4 Volatility of an Equally Weighted
Portfolio versus the Number of Stocks
Portfolio diversification using data on
NYSE stocks
 Risk Premium over Systematic Risk
 Question: Which risk(s) will affect the expected return?
Systematic risk? Unsystematic risk? Or total risk?
 Answer: Since investors are able to diversify their portfolios, they
do not care about unsystematic risk when valuing the risk of an
individual asset. In another word, only systematic risk will
get priced.
 Suppose, on the contrary, that the market compensates investors
with an additional risk premium for diversifiable risk, you can buy
the same stocks and put them into a portfolio so that the
unsystematic risk is eliminated. As a result, you earn additional
risk premium without taking additional risk! => Arbitrage!!!
 How to Quantify the Expected Return
 Expected or required return on Risk Free Asset = RRF.
 All other assets must have a higher expected or required return
RA = RRF + RP
 Where RP represents the risk premium or additional return
required to compensate investors for the additional risk.
 Now, it is a simple matter of estimating RRF and RP.
 RRF is easy, we will use the Treasury rate.
 RP is more difficult.
 We need a model for RP that is easy to use, yet allows for the
differentiation of risk required by all the possible assets and
projects.
 Where are We Headed
 We will use what we have learned about risk and return to
develop a single model for the required return (Risk
Premium) on an asset.
 This model is based on the idea that investors will not “pay”
for risk they can diversify away.
 A more complete understanding of diversification is required
to illustrate the logic driving investor decisions on risk and
return.
 The following slides demonstrate how correlation of return
and portfolio risk are related.
Intermediate Statistics
 Correlations measure the degree to which movements in the
price of one stock is reflected in another stock.
 Using the correlation coefficient between two securities we
can calculate the covariance between two securities and
therefore the resulting variance and standard deviation of a
portfolio of the securities
 We will use covariance to develop the risk of a portfolio of
two assets.
 This logic can be easily to extended to multiple assets.
 Portfolio Returns From Combining
Techie and Steady
Chart of Stock Returns
100.0%

80.0%

60.0%
Returns

40.0%

20.0%

0.0%
1 2 3 4 5 6 7 8 9 10
-20.0%

-40.0%
Steady Corp Observations
Techie Corp Observations
Portfolio Observations
 Correlation of Sample Returns
Techie vs. Steady
Scatter Diagram of Stock Returns

100.0%

Correlation
Coefficient = 0.31
50.0%

0.0%
-100.0% -50.0% 0.0% 50.0% 100.0%

-50.0%

-100.0%
 Return of a Portfolio with Two
Securities
Rp  xa Ra  xb Rb
Where
R p = Return of the Portfolio
xa = Weighting of Security a
xb = Weighting of Security b
Ra = Return of Security a
Rb = Return of Security b
In words, portfolio return is the weighted average of returns of
the components. This relation holds for more than two
securities.
R p   xi Ri where  xi  1
i i
Variance of a Portfolio of Two Securities
s p2  xa2s a2  2x a xb a ,b s a s b  xb2s b2

This is the relationship for the variance of a portfolio

of two securities a and b where:
s p2 = Variance of the Portfolio
xa = Weighting of Security a
xb = Weighting of Security b
s a2 = Variance of Security a
s b2 = Variance of Security b
 a,b  Correlation Coefficien t for a and b
Note :  a, bs a s b = Covariance of a and b = s a ,b
The Volatility of a Portfolio
 Expected return of a portfolio is equal to the weighted
average expected return of its stocks
 Risk of the portfolio is lower than the weighted average of
the individual stocks’ volatility, unless all the stocks all have
perfect positive correlation with each other
 Diversification
 Portfolio Return & Risk
Risk and Return for a Portfolio of Two stocks

Company Steady Corp.Techie Corp. Average

Exp Return 15.0% 21.0% 18.00%

Variance 3.6% 14.0% 8.81%
Standard Deviation 19.0% 38.0% 28.50%

Example One – Complete Positive Correlation

Complete Correlation
Weighting Steady 100% 75% 50% 25% 0%
Weighting Techie 0% 25% 50% 75% 100%
S.D. Steady 19% 19% 19% 19% 19%
S.D. Techie 38% 38% 38% 38% 38%
Corr A,B 1 1 1 1 1

Exp Return Portfolio 15.0% 16.5% 18.0% 19.5% 21.0%

Variance Portfolio 3.6% 5.6% 8.1% 11.1% 14.4%
S.D. Portfolio 19.0% 23.8% 28.5% 33.3% 38.0%

No Reduction in Risk From Diversification.

 Portfolio Return & Risk
Risk and Return for a Portfolio of Two stocks

Company Steady Corp.Techie Corp. Average

Exp Return 15.0% 21.0% 18.00%

Variance 3.6% 14.0% 8.81%
Standard Deviation 19.0% 38.0% 28.50%
Example Two - No Correlation
No Correlation
Weighting A 100% 75% 50% 25% 0%
Weighting B 0% 25% 50% 75% 100%
S.D. A 19% 19% 19% 19% 19%
S.D. B 38% 38% 38% 38% 38%
Corr A,B 0 0 0 0 0

Exp Return Portfolio 15.0% 16.5% 18.0% 19.5% 21.0%

Variance Portfolio 3.6% 2.9% 4.5% 8.3% 14.4%
S.D. Portfolio 19.0% 17.1% 21.2% 28.9% 38.0%

25% Reduction in Risk From Diversification

 Portfolio Return & Risk
Risk and Return for a Portfolio of Two stocks

Company Steady Corp.Techie Corp. Average

Exp Return 15.0% 21.0% 18.00%

Variance 3.6% 14.0% 8.81%
Standard Deviation 19.0% 38.0% 28.50%
Example Three - Complete Negative Correlation
Complete Negative Correlation
Weighting Steady 100% 75% 50% 25% 0%
Weighting Techie 0% 25% 50% 75% 100%
S.D. Steady 19% 19% 19% 19% 19%
S.D. Techie 38% 38% 38% 38% 38%
Corr A,B -1 -1 -1 -1 -1

Exp Return Portfolio 15.0% 16.5% 18.0% 19.5% 21.0%

Variance Portfolio 3.6% 0.2% 0.9% 5.6% 14.4%
S.D. Portfolio 19.0% 4.8% 9.5% 23.8% 38.0%

Very Significant Reduction In Risk from Diversification.

 Risk Reduction From Positive Correlations When
Combining Techie and Steady
Chart of Stock Returns
100.0%

80.0%

60.0%
Returns

40.0%

20.0%

0.0%
1 2 3 4 5 6 7 8 9 10
-20.0%

-40.0%
Steady Corp Observations
Techie Corp Observations
Positive Correlations
 Risk Reduction From Negative Correlation When
Combining Techie and Steady
Chart of Stock Returns
100.0%

80.0%

60.0%
Returns

40.0%

20.0%

0.0%
1 2 3 4 5 6 7 8 9 10
-20.0%

-40.0%
Steady Corp Observations
Techie Corp Observations
Negative Correlation
 Portfolio Return & Risk
Risk and Return for a Portfolio of Two stocks

Company Steady Corp.Techie Corp. Average

Correlation Coefficient =.31
Exp Return 15.0% 21.0% 18.00%
Variance 3.6% 14.0% 8.81%
Standard Deviation 19.0% 38.0% 28.50%
Example Four - Realistic Example of Positive Correlation
Weighting Steady 100% 75% 50% 25% 0%
Weighting Techie 0% 25% 50% 75% 100%
S.D. Steady 19% 19% 19% 19% 19%
S.D. Techie 38% 38% 38% 38% 38%
Corr A,B 0.31 0.31 0.31 0.31 0.31

Exp Return Portfolio 15.0% 16.5% 18.0% 19.5% 21.0%

Variance Portfolio 3.6% 3.8% 5.6% 9.2% 14.4%
S.D. Portfolio 19.0% 19.4% 23.7% 30.3% 38.0%

Stocks Are Generally Positively Correlated,Yet There Is Still A

Benefit (17% Risk Reduction) From Diversification
Risk Aversion
 Speculation: take risky investment for positive risk premium
 Gamble: take risky investment for zero risk premium (fair
game). A risk-averse investor will reject it.
 constant absolute risk aversion (CARA) utility:
 u(r)= -exp(-Ar)
 “A”: coefficient of absolute risk aversion
 “r”: random return, normal distribution
 certainty equivalent rate of return rce:
 u(rce) = E[u(r)] = -exp{-A[E(r)-A σ2 /2]}
 rce = E(r)-A σ2 /2
Risk Aversion
rce = E(r)-A σ2 /2
 rce: earned with certainty
 r: earned with risk
 A σ2 /2 : premium to take risk
 A: how much you hate per unit of risk
 σ2 : how much risk

 risk-free portfolio: rce=rf

 Risk neutral investor: A=0
 Risk averse investor will not invest if rf > E(r)-A σ2 /2
Risk Aversion
rce = E(r)-A σ2 /2
 mean-variance criterion: portfolio A dominates B if
 E(rA)≥ E(rB) and 𝜎𝐴 ≤ 𝜎𝐵
Capital Allocation
 a choice among broad investment classes, rather than among
the specific securities within each asset class. Stock, bonds,
cash (or money market fund)
 A portfolio C with one risky asset and a risk-free asset. A
proportion, y, in the risky portfolio with return rP, and 1- y in
the risk-free asset with return rf : rC= yrp+(1-y)rf
 E(rC)=rf +y[E(rp)-rf]
 σC = y σp
[E(rp)−rf]
 E(rC)=rf + σC
σp
Capital Allocation
[E(rp)−rf]
 E(rC)=rf + σC
σp
[E(rp)−rf]
 : Sharpe ratio
σp
 trade-off between reward (the risk premium) and risk
 slope
Leverage
Leverage
 Borrow and invest, y>1
 Borrow at rf
 Borrow at rB>rf
Asset Allocation
 CAL, the graph of all feasible risk–return combinations
available from different asset allocation choices
 Is there an optimal choice?
 Max: E(r)-A σ2 /2= rf +y[E(rp)-rf] –y2Aσp2/2
[E(rp)−rf]
 y*=
Aσp2
Indifference curves
Finding the optimal complete portfolio
by using indifference curves
Portfolios of Two Risky Assets
 E(rp)=wDE(rD)+wEE(rE)
 σp2 =wD2 σD2 + wE2 σE2 + 2wDwECov(rD ,rE)
 Cov(rD ,rE)=ρDE σDσE
 Hold others constant, smaller the ρDE , the better portfolio
 expected return: weighted average
 SD: less than weighted average of the component SD
 If ρDE =-1, perfect hedge. σp2 can be zero.
Quick summary
 no benefit from diversification when ρ =1
 As ρ is smaller,
 the same SD, the higher E(r);
 the same E(r) (so the weight is fixed), the lower the SD.
 Covariance is more important than variance of each risky
asset
 It is NOT a CAL.
Two Risky Assets and a Risk-Free Asset
Two Risky Assets and a Risk-Free Asset
 Point A: minimum variance portfolio of two risky assets
 CAL: risk-return for a risky portfolio and a risk free asset
(the same as before, one risky asset and a risk-free asset)
 Without risk free asset, point B may be not better than point
A. High return, high risk
 With risk free asset, there is a point on CAL(B) that
dominates point A.
 CAL(B) dominate CAL(A), also a higher Sharpe ratio
 With rf, the optimal risky portfolio is the one with highest
Sharp ratio
Two Risky Assets and a Risk-Free Asset
 Find optimal risky portfolio with risk free asset
E(rp)−rf
Max , with
σp
 E(rp)=wDE(rD)+wEE(rE)
 σp =[wD2 σD2 + wE2 σE2 + 2wDwECov(rD ,rE)]1/2
 wD+wE=1

 Upper case R is the excess return

Two Risky Assets and a Risk-Free Asset
 optimal complete portfolio with an optimal risky portfolio (P)
and risk free asset
[E(rp)−rf]
 The same as before: y*=
Aσp2
 What if the two risky assets have correlation -1?
Two risky assets have correlation -1?
 You can construct a risk free portfolio with these
two risky assets
 If the portfolio return is different from rf, then
arbitrage opportunity. Drives the return equals
to rf.
 If the portfolio return is equal to rf, then the
risk free asset is redundant and the portfolio
choice is the same as the case without risk free
assets.
Markowitz Portfolio Selection Model
 generalize the portfolio construction problem to the case
of many risky securities and a risk-free asset.
1. For a targeted return level, find minimum variance
portfolio.
 For two risky assets, portfolio return and weight is a
one-to-one mapping
2. Find the risky portfolio has the highest Sharpe ratio
3. Mix the optimal risky portfolio with risk free asset
 global minimum variance portfolio
 efficient frontier of risky assets: the part above global minimum
point
 Inside the frontier: individual assets or portfolio
 Horizontal distance between individual assets and frontier:
idiosyncratic risk
More details
 Example: 50 risky securities
 Input list: 50 E(r), 50 SD, 50*49/2 or 1225 Cov
 Calculate portfolio return and variance using the weight wi for
security i
 For each targeted return, find the optimal weight that
minimizes the portfolio variance. Can be solved analytically (e.g,
Fischer Black 1972 Journal of Business)
 Repeat and find frontier
 Introduce risk free assets and find the risky portfolio with
highest Sharpe ratio
Assets, Portfolios and Risk
 We have established a number of ideas let’s summarize:
 Diversification can reduce portfolio risk
 Risk averse investors will construct diversified portfolios
represented by the set of efficient portfolios.
 There is a line which represents the most efficient portfolios (a
combination of borrowing or lending at the risk free rate and the
most efficient portfolio.)
 Since unsystematic risk is the risk that is being diversified away, the
only risk that remains is risk of the general market.
 This optimum risk return combination of risky assets (the market
portfolio) can be described as the intersection of this line with the
efficient portfolios.
Measuring Systematic Risk
 Market Risk and Beta
 We compare a stock’s historical returns to the
market’s historical returns to determine a stock’s beta
(β)
 The sensitivity of an investment to fluctuations in
the market portfolio
 Use excess returns – security return less the risk-
free rate
 The percentage change in the stock’s return that we
expect for each 1% change in the market’s return
Measuring Systematic Risk
 The beta of the overall market portfolio is 1
 Many industries and companies have betas
higher/lower than 1
 Differences in betas by industry are related to
the sensitivity of each industry’s profits to the
general health of the economy
Total Risk Versus Systematic Risk
Problem:
 Suppose that in the coming year, you expect
SysCo’s stock to have a standard deviation of 30%
and a beta of 1.2, and UniCo’s stock to have a
standard deviation of 41% and a beta of 0.6.
 Which stock carries more total risk?
 Which has more systematic risk?
Total Risk Versus Systematic Risk
Execute:
Standard Beta (β)
Deviation (Total (Systematic
Risk) Risk)
SysCo 30% 1.2
UniCo 41% 0.6

 Total risk is measured by standard deviation; therefore, UniCo’s stock has more
total risk.
 Systematic risk is measured by beta. SysCo has a higher beta, and so has more
systematic risk.
Monthly Excess Returns for Apple Stock and for the S&P
500, January 2008-December 2012
Measuring Systematic Risk
 Estimating Beta from Historical Returns
 Apple’s stock for example (Figure):
 The overall tendency is for Apple to have a high
return when the market is up and a low return when
the market is down
 Apple tends to move in the same direction as the
market, but its movements are larger
 The pattern suggests that Apple’s beta is greater than
one
Measuring Systematic Risk
 In practice, we use linear regression to estimate the relation
 The output is the best-fitting line that represents the historical
relation between the stock and the market
 The slope of this line is our estimate of beta
 Tells us how much the stock’s excess return changed for a 1%
change in the market’s excess return
Scatterplot of Monthly Returns for Apple versus the S&P
500, January 2008 through December 2012
Putting It All Together: The Capital
Asset Pricing Model
 One of our goals in this chapter is to compute the cost of
equity capital
 The best available expected return offered in the market on a
similar investment
 To compute the cost of equity capital, we need to know the
relation between the stock’s risk and its expected return
Capital Asset Pricing Model
 CAPM, centerpiece of modern financial economics
 prediction of the relationship between the risk of an asset and
its expected return
 simplifying assumptions that lead to a set of predictions
concerning equilibrium expected returns on risky assets
 Later can add complexity to the hypothesized environment
Assumptions
 There are many investors, each with wealth that is small compared
to the total. Investors are price-takers.
 All investors plan for one identical holding period.
 Investments are limited to a universe of publicly traded financial
assets.
 Investors pay no taxes on returns and no transaction costs.
 All investors are rational mean-variance optimizers, meaning that
they all use the Markowitz portfolio selection model.
 Homogeneous expectations: given a set of security prices and the
risk-free interest rate, all investors use the same expected returns
and covariance matrix of security returns to generate the efficient
frontier and the unique optimal risky portfolio.
Results
 All investors will choose to hold a risky market portfolio (M) of
risky assets.
 The market portfolio will be on the efficient frontier, and will be
the tangency portfolio to the optimal capital allocation line. All
investors hold M as their optimal risky portfolio, differing only in
the amount invested in it versus in the risk-free asset. The capital
market line (CML).
 The risk premium on the market portfolio will be proportional to
its risk and the degree of risk aversion of the representative
investor
 the risk premium on individual securities is
E(ri) – rf =βi[E(rM) - rf];
CAPM
 Why Hold the Market Portfolio?
 The same Markowitz analysis; the same universe of securities; the
same time horizon; the same input list (or, homogeneous expectations)
 The same optimal risky portfolio. Equilibrium: price such that all
securities must be held by someone
 If all investors hold an identical risky portfolio, this portfolio has to
be the market portfolio.
 Risk Premium of the Market Portfolio
 each individual investor chooses a proportion y, allocated to the
E(𝑟𝑀 )−𝑟𝑓
optimal portfolio M, such that y=
A𝜎 2 𝑀
 Take average
 GE’s contribution to the variance of the market portfolio

 GE’s contribution to risk premium

Expected Returns on Individual
Securities
 A basic principle of equilibrium is that all investments should offer
the same reward-to-risk ratio. Pressure on security prices until the
ratios were equalized.

 Cov(rGE ,rM): the contribution of GE stock to the variance of the

market portfolio. Scale it by total market variance and get the beta.
 expected return–beta relationship holds for any individual asset, it
must hold for any combination of assets
Security Market Line
 The expected return–beta relationship can be portrayed
graphically as the security market line (SML)
 CML: graphs the risk premiums as a function of portfolio
standard deviation
Security Market Line
 “fairly priced” assets plot exactly on the SML
 Underpriced: in excess of the fair return stipulated by the
SML, plot above the SML
 Alpha: difference between fair and actually expected rates of
return
 Start with a passive market-index portfolio, increase the
weights of securities with positive alphas and decrease the
weights of securities with negative alphas
 Capital Asset Pricing Model (CAPM)
 This model states that the expected return for all assets is
determined by the asset’s correlation to the “market
portfolio” and the risk free return.
 Using our earlier relationship RA = RRF + RP we now have a
model to describe the risk premium.
RP    (M arket Risk Premium)
Therefore : RP    ( RM  RRF )
Where :
  The Beta coefficien t of a linear regression between
the Asset Returns and the M arket Portfolio Returns
RM  The Exp Return on the M arket Portfolio
RRF  The Risk Free Return
CAPM : RA  RRF    ( RM  RRF )
 What is the Market Portfolio
 In theory, the Market Portfolio is all risky assets that are
available to investors.
 This would include every stock, bond, project and property
available in the world.
 In practice, we use a proxy for this portfolio.
 The most commonly used proxy is the S&P 500 since this
portfolio represents approximately 70% of the market value
of all publicly traded stocks in the U.S.
 It is important to recognize that some assets are being
ignored here, but if we are trying to evaluate a U.S. based
project for a U.S. Corporation the S&P 500 works fairly
well.
 CAPM and Beta
 The Capital Asset Pricing Model (CAPM) states that the he
expected return on any asset can be described by a single
factor (Beta).
 This factor represents the correlation between the asset and
the market portfolio.
 This correlation is measured as the Beta of the asset.
 There is a linear relationship between Beta and the expected
return of an asset (Security Market Line).
 The next three slides describe this relationship graphically.
 An example enhances the understanding of this logic.
 We will use approximate historical returns to illustrate RRF =
4.0% RM = 12.0%
 The Security Market Line
Expected
return

Expected
market
return
12%
1 Exp Return = 12% .
Risk
free
rate
4%

0 1.0  Beta
CAPM Exp R A  RRF   (RM  RRF )
Exp R A  .04  1(.12  .04)  12%
 The Security Market Line
Expected
return

Expected
market
return
12%
1.2 Exp Return = 13.6% .
Risk
free
rate
4%

0 1.0  Beta
CAPM Exp R A  RRF   (RM  RRF )
Exp R A  .04  1.2(.12  .04)  13.6%
Average Betas for Stocks by Industry and the
Betas of a Selected Company in Each Industry
A Negative Beta Stock
Problem:
 Suppose the stock of Bankruptcy Auction Services, Inc. (BAS) has a negative
beta of -0.30. How does its expected return compare to the risk-free rate,
according to the CAPM? Does your result make sense?
A Negative Beta Stock
Solution:
Plan:
 We can use the CAPM equation to compute the expected return of this
negative beta stock just like we would a positive beta stock. We don’t have the
risk-free rate or the market risk premium, but the problem doesn’t ask us for
the exact expected return, just whether or not it will be more or less than the
risk-free rate.
A Negative Beta Stock
Execute:
 Because the expected return of the market is higher than the risk-free rate, the
expected return of Bankruptcy Auction Services (BAS) will be below the risk-
free rate. As long as the market risk premium is positive (as long as people
demand a higher return for investing in the market than for a risk-free
investment), then the second term in CAPM will have to be negative if the beta
is negative.
A Negative Beta Stock
Execute (cont’d):
 For example, if the risk-free rate is 4% and the market risk premium is 6%,
 E[RBAS] = 4% - 0.30(6%) = 2.2%.
 This result seems odd—why would investors be willing to accept a 2.2%
expected return on this stock when they can invest in a safe investment and
earn 4%?
 The answer is that a savvy investor will not hold BAS alone; instead, the investor
will hold it in combination with other securities as part of a well-diversified
portfolio.
 These other securities will tend to rise and fall with the market.
A Negative Beta Stock
Evaluate (cont’d):
 But because BAS has a negative beta, its correlation with the market is negative,
which means that BAS tends to perform well when the rest of the market is
doing poorly.
 Therefore, by holding BAS, an investor can reduce the overall market risk of the
portfolio. In a sense, BAS is “recession insurance” for a portfolio, and investors
will pay for this insurance by accepting a lower return.
 CAPM and Opportunity Cost of Capital
 Since all assets or projects are ultimately valued by the
secondary market, we can use CAPM to represent the
Opportunity Cost of Capital of a company or a project.
 In this context the Opportunity Cost of Capital refers to the
required return for a project that has risk similar to the risk
represented by the Beta in the calculation.
 This model meets our earlier criteria of being simple and
robust.
 It is simple and usually gives sensible answers.
 It distinguishes between diversifiable and non-diversifiable
risk.
 For a risk free project we would use the risk free rate.
 For a risky project we use the risk free rate + the required risk
premium from CAPM.
 Estimating Cost Of Capital With CAPM
Risk Free Rate = 4.0% Expected Market Risk Premium 8.0%
Medtronic
Estimated equity beta = .73
Cost of (equity) capital= risk free rate + Beta x exp. market risk premium

9.8% = 4.0% + .73 x 8.0%

Target
Estimated equity beta = 1.19
Cost of (equity) capital= risk free rate + Beta x exp. market risk premium

13.5% = 4.0% + 1.19 x 8.0%

Best Buy
Estimated equity beta = 1.54
Cost of (equity) capital= risk free rate + Beta x exp. market risk premium

16.3% = 4.0% + 1.54 x 8.0%

Beta of a Portfolio with Two Securities
 p  xa  a  xb b
Where
 p = Beta of the Portfolio
xa = Weighting of Security a
xb = Weighting of Security b
 a = Beta of Security a
 b = Beta of Security b
In words, portfolio beta is the weighted average of betas of the
components. This relation holds for more than two securities.

 p   xi i where  xi  1
i i
Some Cautions about Using CAPM
 Use current Risk Free Rates
 The model assumes there is some long run level of market
premium, therefore always use some “long-run” measure of
the market premium.
 Betas can be looked up on the internet or purchased from
data providers.
 Use the Beta that is most closely associated with the
cyclicality and risk components of the particular project
being considered, This may differ from the company’s Beta.
 How Valid is CAPM
 Evidence shows that CAPM is probably too simple, yet it
does a pretty good job empirically.
 It does capture the two basic ideas of the rent for money, RRF ,
and the investors concern with market instead of diversifiable
risk.
 Long-run average returns are significantly related to beta.
 However beta is not a complete explanation.
 Low beta stocks have earned higher rates of return than
predicted by the model.
 So have small company stocks.
 Overall a good rule of thumb