Principles of Corporate Finance
Chapter 8. Introduction to risk, return, and the opportunity cost of capital
Ciclo Profissional 2o Semestre / 2009
Gradua¸c˜ao em Ciˆencias Econˆomicas
Topics covered
1 Over a century of capital market history
2 Measuring portfolio risk
3 Calculating portfolio risk
4 How individual securities affect portfolio risk
5 Diversification and value additivity
The value of an investment of $1 in 1900
How an investment of $1 at the start of 1900 would have grown in nominal terms, assuming reinvestment of all dividends and interests payments
The value of an investment of $1 in 1900
How an investment of $1 at the start of 1900 would have grown in real terms, assuming reinvestment of all dividends and interests payments
Average rates of return from 1900 to 2006 in % per year
Definition
The risk premium is the extra return versus Treasury bills returns (riskfree rate of interest)
Risk premium of portfolio p = Return ofp−Return of Treasury bills
Using historical evidence to evaluate today’s cost of capital
Suppose there is an investment project that youknow belongs to the same risk class as Standard & Poor’s Composite Index
What rate should we use to discount this project’s forecasted cash flows?
We should use the currentlyexpected rate of return on the market index, i.e., the return rm investors would forgo (if they have correct expectations) by investing in the proposed project One way to estimaterm is to assume that the future will be like the past:
I Today’s investors expect to receive the same rate of return revealed by the averages of the past realization
I The Law of Large Numbers
Following data, we should set rm = 11.7%
Using historical evidence to evaluate today’s cost of capital
Recall that
rm=rf + risk premium of the Index The riskfree interest rate rf is observable today
Therefore, a more sensible procedure to estimate the interest rate on the Index is to estimate the risk premium of the Index using average risk premium
I Mid-2006, the interest rate on Treasury bills is about 5%
I The average risk premium of the Index is 7.6%
Therefore
r = 0.05 + 0.076 = 12.6%
Using historical evidence to evaluate today’s cost of capital
The crucial assumption is that there is a stable risk premium on the Index, so that the expectedfuturerisk premium can be measured by the average past risk premium
Even with over 100 years of data, we can’t estimate the market risk premium exactly
We cannot be sure that investors today are demanding the same reward for risk that they were 50 or 100 years ago
All this leaves a plenty of room for argument about what the risk premiumreally is
Market portfolio
Definition
A market portfoliois a portfolio consisting of a weighted sum of every asset in the market, with weights in the proportions of the value (price) of each asset in the market
Denote by K the finite family of assets in the market Denote by P0,k and Pe1,k the today’s price and tomorrow’s (random) prices of each asset kand let
P0 = X
k∈K
P0,k and Pe1= X
k∈K
Pe1,k
Denote by α the proportionP /P
Market portfolio
Assume I want to invest the amount x >0
I will invest the proportion αkx in assetk by buyingy ≡x/P0 units of asset k
The rate of returnrm of this portfolio is P
k∈KyPe1,k−P
k∈KyP0,k
x =
P
k∈KyPe1,k−P
k∈KyP0,k P
k∈KyP0,k
= X
k∈K
αkPe1,k−P0,k P0,k
= X
k∈K
αkerk
Whererek is the random rate of return of asset kdefined by Pe1,k−P0,k
P
Market portfolio
In theory, the market portfolio would necessarily include every single possible available asset, including real estate, precious metals, stamp collections, jewelry, and anything with any worth, as the theoretical market being referred to what would be the world market
In practise, good proxies for the market are indexes
I S&P500 in the US
I DAX in Germany
I TSE100 in the UK
Average risk premia in the world
Nominal return of the market portfolio minus nominal return on bills, from 1900 to 2006
The equity premium puzzle
There is a high (historic) risk premium earned in the market This seems to imply that that investors are extremely risk-averse If that is true, investors should cut back their consumption when stock prices fall and wealth decreases
But evidence suggests that when stock prices fall, investors spend at nearly the same rate
This is difficult to reconcile with high risk aversion and a high market risk premium
Mehra, S. and Prescott, E.
The Equity Premium: A Puzzle.
Journal of Monetary Economics 15, 1985
The equity premium puzzle
A large number of explanations for the puzzle have been proposed The puzzle is a statistical illusion
Modifications to the assumed preferences of investors Market imperfections
The equity premium puzzle
The historical average of risk premium in U.S. is 7.6%
Survey of chief financial officers commonly suggest that they expect a market risk premium that is several percentage points below the historical average
Possible explanations:
I Growth of mutual funds has made it easier for individuals to diversify away part of their risk
I Pension funds and other financial institutions have found that they also could reduce their risk by investing part of their funds overseas
I If investors can eliminate more of their risk than in the past, they may become content with a lower return
The dividend yields
If a stock is paying dividends with a constant growth rate g, then we should have
r = DIV1
P0
+g
Dividend yields in U.S. have averaged about 4.4% since 1900 The annual growth in dividends has averaged about 5.6%
If the dividend growth was representative of what investors expected, then the expected market return over this period should
be DIV1
P0
+g= 4.4 + 5.6 = 10.0%
or 6% above the risk-free interest rate
This figure is 1.6% lower than the realizedrisk premium
The dividend yields
Average dividend yields in the U.S.A. since 1900
The dividend yields
Dividend yields fluctuated quite sharply
There are only two possible reasons for the yield changes:
I some years investors were unusually optimistic or pessimistic aboutg
I the required returnrwas unusually high or low
Following several studies, Economists have concluded that very little of the variation is related to the rate of dividend growth If they are right, the level of yields ought to be telling us something about the return that investors require
Measuring portfolio risk
The stock market has been a profitable but extremely variable investment
Wide spread of returns from investment in common
stocks
Standard statistical measures
Consider a portfolio pand denote by erp the function erp : Ω−→[0,∞)
describing the possible values of the return of the portfoliop The set Ω describes possible “states of nature”
There exists a probability Perp ∈Prob(Ω), called the law oferp and defined as follows
∀A⊂[0,∞), Perp(A) =P{erp ∈A}
IfA= [0.02,0.04], thenPerp(A) is the probability that the return rp is between 2% and 4%
Standard statistical measures
Assume that Ω is finite and let Imerp be the range (or image) oferp, i.e., the set of possible values of erp:
Imerp ={r ∈[0,∞) : ∃ω∈Ω, erp(ω) =r}
Definition
The expected value of erp is the number denoted byEP[erp] or rp defined by
EP[erp] = X
ω∈Ω
P(ω)rep(ω)
= X
r∈Imrep
rPerp(r)
Standard statistical measures
Assume that Ω is infinite and that the law of Perp has a density h
erp in the sense that
∀06a < b, Perp[a, b] = Z b
a
herp(r)dr.
Definition
The expected value of erp is the number denoted byEP[erp] or rp defined by
EP[erp] = Z ∞
0
rherp(r)dr This expectation is also denoted by
∞
Variance and standard deviation
Definition
The varianceof the return erp is the expected squared deviation from the expected return, i.e.,
var(erp) =EP
(rep−rp)2
=EP h
erp−EP[rep]2i Definition
The standard deviation (orrisk) of the return erp is the square root of the variance:
Standard deviation ofrep = q
var(erp)
When there is no ambiguity concerning the portfolio, the standard deviation is denoted by σ and the variance by σ2
Example
Consider the following game You start by investing $100 Then two coins are flipped
For each head that comes up you get back your starting balance plus 20%
For each tail that comes up you get back your starting balance less 10%
There are four equally likely outcomes:
Head + Head: You gain 40%
Head + tail: You gain 10%
Tail + head: You gain 10%
Example
The expected return is
Expected return = (0.25×40%) + (0.50×10%) + (0.25× −20%)
= +10%
Measuring expected returns and variance
To compute EP[erp] and var(erp), we should
I identify the possible outcomes
I assign a probability to each outcome Where do the probabilities come from?
Most analysts start by observing the past
It is reasonable to consider that portfolios with histories of high variability also have the least predictable future performance
Measuring expected returns and variance
Consider a sample of observed returns {erm(t)}16t6N whereN is the number of observations One may consider the following estimates
EP[erm]≈ 1 N
N
X
t=1
erm(t)
var(erm)≈ 1 N −1
N
X
t=1
erm(t)−EP[erm]2
Measuring risk
The annual standard deviations and variances (in nominal terms) observed over the period 1900-2006
Measuring risk
The risk (standard deviation of annual returns) of markets around the world, 1900-2006
Measuring risk
Annualized standard deviation of the Dow Jones Industrial Average, 1900-2006
Measuring risk
Standard deviations for selected U.S. common stocks, July 2001-June 2006 (figures in percent per year)
During this period, the market portfolio’s standard deviation was about 16%
Measuring risk
Standard deviation for selected foreign stocks and market indexes, July 2001–June 2006 (figures in percent per year)
Some of the stocks are much more volatile than others
But individual stocks are for the most part more variable that the
Diversification reduces risk
The risk (standard deviation) of randomly selected portfolios containing different numbers of New York Stock Exchange stocks
Diversification reduces risk
Diversification reduces risk because prices of different stocks do not move exactly together
Different kinds of risk
Definition
The risk that potentially can be eliminated by diversification is called unique risk
Unique risk may be called unsystematic risk residual risk specific risk diversifiable risk
Unique risk stems from the fact that many of the perils that surround an individual company are peculiar to that company and perhaps its immediate competitors
Different kinds of risk
Definition
The risk that cannot be avoided, regardless of how much you diversy, is called market risk
Market risk may be called systematic risk undiversifiable risk
Market risk stems from the fact that there are economywide perils that threaten all businesses
Different kinds of risk
Calculating portfolio risk
Definition
Consider two portfolios with returns re1 and er2 respectively. The covariance of the two portfolios is defined as follows
cov(er1,re2) =EP
(re1−EP[er1])(er2−EP[er2])
=EP[(re1−r1)(er2−r2)]
When there is no ambiguity concerning the two portfolios, the covariance is denoted byσ12
The covariance is symmetric, i.e.,σ12=σ21 Observe that
cov(er1,re2) =EP[er1er2]−EP[er1]EP[er2] Moreover, we haveσ =σ2, i.e.,
Calculating portfolio risk
Definition
Consider two portfolios with returns re1 and er2 respectively. The
correlation coefficientρ(er1,er2) of the two portfolios is defined as follows cov(re1,er2) =ρ(er1,re2)p
var(re1)p var(er2) When there is no ambiguity concerning the two portfolios, the correlation coefficient is denoted by ρ12, in particular we have
σ12=ρ12σ1σ2
Calculating portfolio risk
From Cauchy-Schwartz, we have ρ12∈[−1,+1]:
if the returns of the portfolios are positively and perfectly correlated then ρ12= 1
if the returns of the portfolios are wholly unrelated thenρ12= 0 if the returns of the portfolios are negatively and perfectly correlated then ρ12=−1
If two stocks are perfectly negatively correlated then there is always a portfolio strategy that will completely eliminate risk
Calculating portfolio risk
Consider a portfoliop composed of two portfolios with returnsre1
and er2
We let xi ∈[0,1] the proportion holding of portfolioi, x1+x2= 1
in the sense that the returnerp of portfoliop is defined by rep=x1er1+x2er2
The expected return rp of portfoliop is given by
rp =x1r1+x2r2 or EP[rep] =x1EP[er1] +x2EP[er2]
Calculating portfolio risk
The variance σp2 of portfoliop is given by
var(rep) =x21var(er1) +x22var(er2) + 2x1x2cov(er1,re2) or equivalently
σ2p =x21σ12+x2σ22+ 2(x1x2ρ12σ1σ2)
Calculating portfolio risk
Consider the general case of a portfolio p composed of a finite family of stocks with returns
(eri)16i6N with proportions (xi)16i6N, i.e.,
N
X
i=1
xi= 1 and xi >0, ∀i We have
rep= X
16i6N
xieri
implying that
rp= X
16i6N
xiri and
σp= X X
xixjσij
Calculating portfolio risk
Calculating portfolio risk
Assume that the portfolio p is composed with the same proportion of each stock, i.e., xi = 1/N for each i
σp2 = N 1
N 2"
PN i=1σ2i
N
#
+ (N2−N) 1
N
2 P
i6=jσij
N2−N
= 1
N ×average variance +
1− 1 N
×average covariance Passing to the limit
N→∞lim σ2p = average covariance we see that only the average covariance matters
Calculating portfolio risk
Consider the general case of a portfolio p composed of a finite family of stocks with returns
(eri)16i6N with proportions (xi)16i6N, i.e.,
rep= X
16i6N
xieri
The risk of the portfolio is measured by the variance σp = var(erp) One may wonder what is the relative contribution of a specific stockj to the risk of portfoliop
How individual securities affect portfolio risk
Recall that the risk σp2 of portfoliop is given by σp2 =
N
X
i=1 N
X
k=1
xixkσik
Therefore, the contribution of stock j is
∂ σ2p
∂xj
= 2 [x1σj1+x2σj2+. . .+xNσjN]
= 2
N
X
i=1
xicov(rej,rei)
= 2 cov erj,
N
X
i=1
xirei
!
= 2 cov(erj,erp)
How individual securities affect portfolio risk
The relative contribution of stockj to the standard deviationσp of the portfolio pis
∂σp/∂xj
σp = 1
2σp2
∂ σp2
∂xj
= cov(erj,rep) var(erp)
= σjp
σp2
How individual securities affect portfolio risk
Definition
Consider a portfolio pwith return erp and a stock iwith returneri. The beta of stock irelative to portfolio p is denotedβi(p) and defined by
βi(p) = σip
σp2 = cov(eri,rep) var(erp) Therefore, we have
cov(rei,erp) =βi(p) var(rep)
The beta of a stock
Recall that the market portfolio mis defined by the returnrem given by
erm =
N
X
i=1
αieri where αi= Vi PN
k=1Vk and Vi is the value (or price) of equity i
Definition
The beta of stock irelative the market portfoliom is denoted byβi
An intuition
Imagine that a stock iis perfectly correlated with the market portfolio m, i.e., there exists a correlation coefficient αim such that
eri−ri =αim(erm−rm) In that case, we have βi=αim
if the market rises 1% then the price of stockiwill rise by β%
if the market falls 2% then the price of stock iwill fall 2×β%
In general stocks are not perfectly correlated with market returns, each firm is subject to unique risk
Examples
Betas for selected U.S. common stocks, July 2001-June 2006
Example
Disney had a beta of 1.26
If the future resembles the past, this means thaton average when the market rises an extra 1%
Disney’s stock prices will rise by an extra 1.26%
When the market falls an extra 2%, Disney’s stock prices will fall an extra 2.52%
Foreign betas
Betas for foreign stocks, July 2001–June 2006
Beta is measured relative to the stock’s home market
Diversification and value additivity
Diversification reduces risk for investors, and therefore makes sense for them
Does it make sense for firms to diversify their activities? Is a diversified firm more attractive to investors than an undiversified one?
If the answer is yes, then it contradicts the additivity of the net present value since the value of a diversified package would be greater than the sum of the parts
If investors can diversify then firms don’t have to do diversify since investors will not pay any extrafor firms that diversify
I It is very easy for an investor to diversify: he can buy shares of mutual funds that hold diversified portfolios
I A firm cannot diversify so easily